Datasets

9,722 machine learning datasets
🔔   Share your dataset with the ML community!
Filter by Modality (clear)
Texts Images 787 Videos 250 Medical 143 Time series 129 Audio 126 Graphs 117 Tabular 110 3D 100 Speech 76 Biomedical 62 RGB-D 55 Point cloud 44 Environment 36 Tables 33 Biology 32 RGB Video 30 Tracking 27 LiDAR 25 Actions 23 3d meshes 21 Dialog 18 Stereo 15 Hyperspectral images 14 MRI 13 Physics 13 EEG 11 Interactive 10 Music 8 6D 7 Financial 6 Parallel 6 Replay data 5 Ranking 4 fMRI 3 Cad 2 Midi 2 PSG 1 Lyrics 0
Filter by Task
Question Answering 145 Text Classification 70 Text Generation 66 Language Modelling 54 Visual Question Answering (VQA) 51 Machine Translation 43 Named Entity Recognition (NER) 41 Text Summarization 40 Reading Comprehension 39 Information Retrieval 35 Natural Language Inference 34 Sentiment Analysis 34 Natural Language Understanding 30 Relation Extraction 29 Common Sense Reasoning 27 Abstractive Text Summarization 22 Misinformation 22 Code Generation 20 Entity Linking 20 Hate Speech Detection 20 Classification 19 Coreference Resolution 19 Image Captioning 18 Stance Detection 17 Machine Reading Comprehension 16 Open-Domain Question Answering 16 Document Summarization 15 Word Embeddings 15 Image Retrieval 14 Video Question Answering 14 Data-to-Text Generation 13 Fake News Detection 13 Relation Classification 13 Data Augmentation 12 Domain Adaptation 12 Paraphrase Identification 12 Retrieval 12 Visual Reasoning 12 Dialogue Generation 11 Link Prediction 11 Part-Of-Speech Tagging 11 Semantic Parsing 11 Speech Recognition 11 Token Classification 11 Emotion Recognition in Conversation 10 Joint Entity and Relation Extraction 10 Math Word Problem Solving 10 Multi-Task Learning 10 Paraphrase Generation 10 Sequence-to-sequence Language Modeling 10 Video Captioning 10 Automatic Post-Editing 9 Few-Shot Learning 9 Knowledge Graphs 9 Mathematical Reasoning 9 Recommendation Systems 9 Sarcasm Detection 9 Visual Question Answering 9 Zero-shot Text Search 9 Aspect-Based Sentiment Analysis (ABSA) 8 Cross-Lingual Transfer 8 Dialogue State Tracking 8 Document Classification 8 Emotion Classification 8 Emotion Recognition 8 NER 8 Question Generation 8 Semantic Textual Similarity 8 Task-Oriented Dialogue Systems 8 Translation 8 Video Retrieval 8 Entity Typing 7 Explanation Generation 7 Fact Verification 7 Image Classification 7 Multi-Document Summarization 7 Multiple-choice 7 Passage Retrieval 7 Sentence Embeddings 7 Text Simplification 7 Text-to-Image Generation 7 Topic Models 7 Ad-hoc video search 6 Dialogue Understanding 6 Event Extraction 6 Intent Detection 6 Medical Visual Question Answering 6 Optical Character Recognition (OCR) 6 Referring Expression Segmentation 6 Sentence Classification 6 Sentiment Classification 6 Slot Filling 6 Stance Classification 6 Visual Dialog 6 Word Sense Disambiguation 6 Zero-Shot Learning 6 AMR Parsing 5 AMR-to-Text Generation 5 Abusive Language 5 Chinese Reading Comprehension 5 Code Completion 5 Conversational Question Answering 5 Conversational Response Selection 5 Cross-Modal Retrieval 5 Dependency Parsing 5 Extractive Text Summarization 5 Goal-Oriented Dialog 5 Image-to-Text Retrieval 5 Knowledge Base Question Answering 5 Mathematical Question Answering 5 Meeting Summarization 5 Multimodal Deep Learning 5 Nested Named Entity Recognition 5 Response Generation 5 Scene Text Recognition 5 Semantic Role Labeling 5 Spoken Language Understanding 5 Table-to-Text Generation 5 Toxic Comment Classification 5 Twitter Sentiment Analysis 5 Video Grounding 5 Vision and Language Navigation 5 Vision-Language Navigation 5 Abstractive Dialogue Summarization 4 Answer Selection 4 Automated Theorem Proving 4 Chart Question Answering 4 Citation Recommendation 4 Code Search 4 Composed Image Retrieval (CoIR) 4 Cross-Lingual NER 4 Dialogue Act Classification 4 Dialogue Evaluation 4 Discourse Parsing 4 Discourse Segmentation 4 Extreme Summarization 4 Fact Checking 4 Few-Shot Relation Classification 4 Few-shot NER 4 Gesture Generation 4 Image Generation 4 KG-to-Text Generation 4 Language Identification 4 Logical Reasoning 4 Low Resource Named Entity Recognition 4 Moment Retrieval 4 Multi-Label Classification 4 Multi-Label Text Classification 4 Multi-task Language Understanding 4 Multimodal Reasoning 4 Multimodal Sentiment Analysis 4 Natural Questions 4 Nested Mention Recognition 4 News Classification 4 Node Classification 4 Object Detection 4 Open Information Extraction 4 Open Intent Discovery 4 Open-Domain Dialog 4 Paper generation 4 Reading Comprehension (Few-Shot) 4 Reading Comprehension (One-Shot) 4 Reading Comprehension (Zero-Shot) 4 Referring Expression Comprehension 4 Scene Text Detection 4 Speech Synthesis 4 Story Generation 4 Text-To-SQL 4 Text-to-Code Generation 4 Text-to-Video Generation 4 Video Understanding 4 Visual Commonsense Reasoning 4 Weakly-Supervised Named Entity Recognition 4 Zero-Shot Video Question Answer 4 Zero-Shot Video Retrieval 4 2D Object Detection 3 Anomaly Detection 3 Argument Mining 3 Aspect Category Detection 3 Aspect Category Polarity 3 Aspect Term Extraction and Sentiment Classification 3 Aspect-Category-Opinion-Sentiment Quadruple Extraction 3 Audio to Text Retrieval 3 Automatic Speech Recognition (ASR) 3 Bias Detection 3 Binary Classification 3 Biomedical Information Retrieval 3 Chinese Named Entity Recognition 3 Citation Intent Classification 3 Click-Through Rate Prediction 3 Code Classification 3 Code Repair 3 Cross Document Coreference Resolution 3 Cross-Lingual Question Answering 3 Decision Making 3 Definition Extraction 3 Dense Video Captioning 3 Entity Disambiguation 3 Entity Resolution 3 Entity Retrieval 3 Event Coreference Resolution 3 Explainable artificial intelligence 3 Fairness 3 Few-Shot Image Classification 3 Formal Logic 3 Gender Bias Detection 3 Generative Question Answering 3 Graph Classification 3 Intent Classification 3 Language Acquisition 3 Long-range modeling 3 Medical Concept Normalization 3 Motion Synthesis 3 Multi-Domain Recommender Systems 3 Multi-class Classification 3 Multilingual NLP 3 Multimodal Intent Recognition 3 Multiple Choice Question Answering (MCQA) 3 Natural Language Visual Grounding 3 Negation Detection 3 Object Counting 3 Open Intent Detection 3 Opinion Mining 3 Phrase Grounding 3 Recognizing Emotion Cause in Conversations 3 Referring Expression 3 Review Generation 3 Scientific Document Summarization 3 Semantic Segmentation 3 Semantic Similarity 3 Stochastic Optimization 3 Term Extraction 3 Text Retrieval 3 Text to Audio Retrieval 3 Text-To-Speech Synthesis 3 Time Series Forecasting 3 Topic Classification 3 Translation deu-eng 3 Translation eng-deu 3 Unsupervised Extractive Summarization 3 Unsupervised Text Classification 3 Video-Text Retrieval 3 Word Alignment 3 Zero-Shot Composed Image Retrieval (ZS-CIR) 3 Zero-Shot Text Classification 3 Zero-shot Named Entity Recognition (NER) 3 2 2D Semantic Segmentation 2 3D Anomaly Detection 2 3D Face Animation 2 AbbreviationDetection 2 Action Recognition 2 Active Learning 2 Adversarial Robustness 2 Aggression Identification 2 Answer Generation 2 Argument Retrieval 2 Arithmetic Reasoning 2 Aspect Extraction 2 Astronomy 2 Bayesian Inference 2 Binary Relation Extraction 2 Binary text classification 2 Causal Inference 2 Causal Language Modeling 2 Chatbot 2 Chunking 2 Citation Prediction 2 Claim Verification 2 Code Comment Generation 2 Code Documentation Generation 2 Code Summarization 2 Code Translation 2 Commonsense Knowledge Base Construction 2 Community Question Answering 2 Conditional Text Generation 2 ContextNER 2 Continual Pretraining 2 Conversational Search 2 Cross-Lingual Abstractive Summarization 2 Cross-Lingual POS Tagging 2 Cross-Lingual Paraphrase Identification 2 Dark Humor Detection 2 Dialog Relation Extraction 2 Document Ranking 2 Dynamic Link Prediction 2 Embeddings Evaluation 2 End-To-End Dialogue Modelling 2 Entity Alignment 2 Event Argument Extraction 2 Extractive Document Summarization 2 FG-1-PG-1 2 Fact-based Text Editing 2 Factual Visual Question Answering 2 Few-Shot NLI 2 Few-Shot Text Classification 2 General Knowledge 2 Goal-Oriented Dialogue Systems 2 Grammatical Error Correction 2 Graph Embedding 2 Graph Generation 2 Handwriting Recognition 2 Headline Generation 2 Humor Detection 2 Implicit Discourse Relation Classification 2 Instance Segmentation 2 Instruction Following 2 Intent Discovery 2 Interpretable Machine Learning 2 Key Information Extraction 2 Knowledge Probing 2 Learning-To-Rank 2 Lemmatization 2 Lexical Entailment 2 Low-Resource Neural Machine Translation 2 MNLI-m 2 MNLI-mm 2 MRPC 2 Masked Language Modeling 2 Medical Code Prediction 2 Medical Diagnosis 2 Medical Report Generation 2 Memorization 2 Model Compression 2 Mortality Prediction 2 Motion Captioning 2 Multi-Label Learning 2 Multi-domain Dialogue State Tracking 2 Multi-hop Question Answering 2 Multi-modal Dialogue Generation 2 Multilingual Named Entity Recognition 2 Multilingual text classification 2 Multimodal Abstractive Text Summarization 2 Multimodal Machine Translation 2 Multimodal Text Prediction 2 Multiview Contextual Commonsense Inference 2 Named Entity Recognition 2 Network Embedding 2 New Product Sales Forecasting 2 News Annotation 2 News Recommendation 2 Node Clustering 2 Object Localization 2 Object Recognition 2 Out of Distribution (OOD) Detection 2 Out-of-Distribution Detection 2 POS 2 Paper generation (Conclusion-to-title) 2 Paper generation (Title-to-abstract) 2 Paper generation (abstract-to-conclusion) 2 Person Re-Identification 2 Person Retrieval 2 Person Search 2 Probing Language Models 2 Product Recommendation 2 Program Synthesis 2 Prosody Prediction 2 QNLI 2 QQP 2 Quantization 2 Query-Based Extractive Summarization 2 Reinforcement Learning (RL) 2 Relational Reasoning 2 Rumour Detection 2 SQL Parsing 2 SST-2 2 Science Question Answering 2 Scientific Concept Extraction 2 Scientific Results Extraction 2 Semantic Textual Similarity within Bi-Encoder 2 Semi Supervised Learning for Image Captioning 2 Semi-Supervised Text Classification 2 Sentence Embedding 2 Sentence-Embedding 2 Sequential Recommendation 2 Sign Language Translation 2 Speaker Diarization 2 Speaker Identification 2 Speech Emotion Recognition 2 Speech-to-Speech Translation 2 Spoken Dialogue Systems 2 Story Completion 2 Table Detection 2 Temporal Action Localization 2 Temporal Information Extraction 2 Temporal Relation Classification 2 Temporal/Casual QA 2 Text Categorization 2 Text Style Transfer 2 Text based Person Retrieval 2 Text to Video Retrieval 2 Text-to-Music Generation 2 Timex normalization 2 Translation eng-hrv 2 Translation eng-srp_Cyrl 2 Transliteration 2 Unsupervised Domain Adaptation 2 Unsupervised Machine Translation 2 Unsupervised Opinion Summarization 2 ValNov 2 Visual Entailment 2 Visual Navigation 2 Visual Relationship Detection 2 Visual Storytelling 2 Weakly Supervised Classification 2 Word Translation 2 Zero-Shot Cross-Lingual Transfer 2 Zero-Shot Image Classification 2 Zero-shot Relation Classification 2 Zero-shot Relation Triplet Extraction 2 dialog state tracking 2 knowledge editing 2 multimodal generation 2 text annotation 2 text similarity 2 text-based games 2 text2text-generation 2 2D Human Pose Estimation 1 3D Object Classification 1 3D Object Detection 1 3D Object Recognition 1 3D Shape Reconstruction 1 3D dense captioning 1 Abstract Algebra 1 Abuse Detection 1 Accented Speech Recognition 1 Action Anticipation 1 Action Classification 1 Action Detection 1 Action Segmentation 1 Actionable Phrase Detection 1 Ad-Hoc Information Retrieval 1 Adversarial Attack 1 Adversarial Text 1 Aesthetic Image Captioning 1 Aesthetics Quality Assessment 1 Age And Gender Classification 1 Analogical Similarity 1 Analytic Entailment 1 Anatomy 1 Annotated Code Search 1 Argument Pair Extraction (APE) 1 Aspect Category Sentiment Analysis 1 Aspect Sentiment Triplet Extraction 1 Aspect-Based Sentiment Analysis 1 Aspect-oriented Opinion Extraction 1 Attribute Value Extraction 1 Audio Classification 1 Audio Super-Resolution 1 Audio captioning 1 Audio-Visual Speech Recognition 1 Authorship Verification 1 Auto Debugging 1 AutoML 1 Automated Essay Scoring 1 Automatic Lyrics Transcription 1 Automatic Speech Recognition 1 Autonomous Driving 1 Bandwidth Extension 1 Blackout Poetry Generation 1 Board Games 1 Bridging Anaphora Resolution 1 Business Ethics 1 Caption Generation 1 Card Games 1 Causal Discovery 1 Causal Emotion Entailment 1 Causal Identification 1 Cell Segmentation 1 Cell Tracking 1 Chat-based Image Retrieval 1 Chemical Entity Recognition 1 Chemical Indexing 1 Claim Extraction with Stance Classification (CESC) 1 Claim-Evidence Pair Extraction (CEPE) 1 Clinical Knowledge 1 Clinical Note Phenotyping 1 Cloze (multi-choices) (Few-Shot) 1 Cloze (multi-choices) (One-Shot) 1 Cloze (multi-choices) (Zero-Shot) 1 Clustering 1 Clustering Algorithms Evaluation 1 CoLA 1 CodeSearchNet - Java 1 College Biology 1 College Chemistry 1 College Computer Science 1 College Mathematics 1 College Medicine 1 College Physics 1 Common Sense Reasoning (Zero-Shot) 1 Commonsense Causal Reasoning 1 Complex Query Answering 1 Component Classification 1 Composed Video Retrieval (CoVR) 1 Compositional Zero-Shot Learning 1 Computational Phenotyping 1 Computed Tomography (CT) 1 Source Code Summarization 1
Filter by Language (clear)
English Chinese 205 German 127 French 112 Spanish 94 Russian 86 Portuguese 61 Italian 59 Japanese 56 Arabic 54 Hindi 54 Korean 40 Turkish 39 Vietnamese 39 Dutch 32 Persian 31 Bengali 30 Czech 30 Tamil 30 Danish 29 Polish 28 Indonesian 27 Finnish 24 Romanian 24 Marathi 22 Multilingual 22 Telugu 21 Hungarian 19 Urdu 19 Greek 18 Swedish 18 Thai 18 Estonian 17 Bulgarian 15 Gujarati 15 Hebrew 15 Malayalam 15 Slovak 14 Swahili 14 Basque 13 Croatian 13 Punjabi 13 Ukrainian 13 Latvian 11 Mandarin Chinese 11 Norwegian 11 Slovenian 11 Amharic 10 Catalan 10 Kazakh 10 Lithuanian 10 Serbian 10 Kannada 9 Albanian 8 Armenian 8 Assamese 8 Irish 7 Oriya (macrolanguage) 7 Sanskrit 7 Sinhala 7 Tagalog 7 Welsh 7 Yoruba 7 Burmese 6 Georgian 6 Hausa 6 Icelandic 6 Igbo 6 Iranian Persian 6 Kurdish 6 Macedonian 6 Maltese 6 Mongolian 6 Somali 6 Afrikaans 5 Azerbaijani 5 Galician 5 Guarani 5 Haitian 5 Malay (individual language) 5 Norwegian Bokmål 5 Oromo 5 Sindhi 5 Uzbek 5 American Sign Language 4 Bambara 4 Belarusian 4 Breton 4 Egyptian Arabic 4 Filipino 4 Latin 4 Malagasy 4 Nigerian Pidgin 4 Norwegian Nynorsk 4 Odia 4 Scottish Gaelic 4 Serbo-Croatian 4 Tigrinya 4 Wolof 4 Bangala 3 Cebuano 3 Central Khmer 3 Central Kurdish 3 Chechen 3 Esperanto 3 Fulah 3 Ganda 3 Iloko 3 Javanese 3 Kirghiz 3 Lao 3 Lingala 3 Nepali (macrolanguage) 3 Quechua 3 South Azerbaijani 3 Standard Arabic 3 Sundanese 3 Upper Sorbian 3 Western Panjabi 3 Aragonese 2 Bashkir 2 Bavarian 2 Bhojpuri 2 Bishnupriya 2 Bosnian 2 Dhivehi 2 Erzya 2 Faroese 2 Goan Konkani 2 Jejueo 2 Kabyle 2 Kinyarwanda 2 Luo (Kenya and Tanzania) 2 Maithili 2 Malay (macrolanguage) 2 Modern Greek 2 Moroccan Arabic 2 Nepali (individual language) 2 Nyanja 2 Romansh 2 Russia Buriat 2 Swati 2 Tajik 2 Tatar 2 Tibetan 2 Tsonga 2 Tswana 2 Uighur 2 Waray (Philippines) 2 Xhosa 2 Yiddish 2 Yue Chinese 2 Akkadian 1 Akuntsu 1 Ancient Greek 1 Ancient Hebrew 1 Apurinã 1 Argentine Sign Language 1 Assyrian Neo-Aramaic 1 Asturian 1 Avaric 1 Aymara 1 Bemba (Zambia) 1 Central Bikol 1 Central Pashto 1 Chavacano 1 Chukot 1 Church Slavic 1 Chuvash 1 Congo Swahili 1 Coptic 1 Cornish 1 Dimli (individual language) 1 Dogri (macrolanguage) 1 Eastern Mari 1 Ewe 1 Fon 1 Geez 1 German Sign Language 1 Gothic 1 Gulf Arabic 1 Halh Mongolian 1 Ido 1 Interlingue 1 Inuktitut 1 Kabuverdianu 1 Kachin 1 Kalaallisut 1 Kalmyk 1 Karachay-Balkar 1 Karelian 1 Khunsari 1 Komi 1 Komi-Permyak 1 Komi-Zyrian 1 Krio 1 Lezghian 1 Limburgan 1 Literary Chinese 1 Livvi 1 Lojban 1 Lombard 1 Low German 1 Lower Sorbian 1 Lozi 1 Lunda 1 Luo (Cameroon) 1 Lushai 1 Luxembourgish 1 Manipuri 1 Manx 1 Maori 1 Mazanderani 1 Mbyá Guaraní 1 Mesopotamian Arabic 1 Minangkabau 1 Mingrelian 1 Mirandese 1 Moksha 1 Mundurukú 1 Najdi Arabic 1 Nayini 1 Neapolitan 1 Newari 1 Nigerian Fulfulde 1 North Azerbaijani 1 North Levantine Arabic 1 Northern Frisian 1 Northern Kurdish 1 Northern Luri 1 Northern Sami 1 Northern Uzbek 1 Occitan (post 1500) 1 Old French 1 Old Russian 1 Old Turkish 1 Ossetian 1 Pampanga 1 Pedi 1 Piemontese 1 Plateau Malagasy 1 Pushto 1 Rundi 1 Sardinian 1 Shan 1 Shona 1 Sicilian 1 Skolt Sami 1 Soi 1 South Levantine Arabic 1 Southern Pashto 1 Southern Sotho 1 Standard Latvian 1 Swedish Sign Language 1 Swiss German 1 Swiss-German Sign Language 1 Tonga (Zambia) 1 Tosk Albanian 1 Tupinambá 1 Turkmen 1 Tuvinian 1 Twi 1 Venetian 1 Volapük 1 Walloon 1 Warlpiri 1 West Central Oromo 1 Western Frisian 1 Western Mari 1 Wu Chinese 1 Yakut 1 Zulu 1 Abkhazian 0 Achinese 0 Adyghe 0 Afar 0 Akan 0 Arpitan 0 Bangladeshi Sign Language 0 Banjar 0 Bislama 0 Bodo (India) 0 Buginese 0 Chamorro 0 Cherokee 0 Cheyenne 0 Choctaw 0 Corsican 0 Cree 0 Creek 0 Crimean Tatar 0 Dogri (individual language) 0 Dzongkha 0 Extremaduran 0 Fiji Hindi 0 Fijian 0 Friulian 0 Gagauz 0 Gan Chinese 0 Gilaki 0 Greek Sign Language 0 Hakha Chin 0 Hakka Chinese 0 Hawaiian 0 Herero 0 Hiri Motu 0 Interlingua (International Auxiliary Language Association) 0 Inupiaq 0 Jamaican Creole English 0 Kabardian 0 Kanuri 0 Kara-Kalpak 0 Kashmiri 0 Kashubian 0 Kikuyu 0 Kongo 0 Kuanyama 0 Kölsch 0 Ladino 0 Lak 0 Latgalian 0 Ligurian 0 Marshallese 0 Min Dong Chinese 0 Modern Greek (1453-) 0 Narom 0 Nauru 0 Navajo 0 Naxi 0 Ndonga 0 Northern Huishui Hmong 0 Novial 0 Official Aramaic (700-300 BCE) 0 Old English (ca. 450-1100) 0 Pali 0 Pangasinan 0 Papiamento 0 Pennsylvania German 0 Pfaelzisch 0 Picard 0 Pitcairn-Norfolk 0 Pontic 0 Portuguse 0 Rajasthani 0 Rusyn 0 Saidi Arabic 0 Samoan 0 Sango 0 Santali 0 Saterfriesisch 0 Scots 0 Sichuan Yi 0 Silesian 0 Sranan Tongo 0 Swahili (macrolanguage) 0 Tahitian 0 Tai 0 Tetum 0 Tok Pisin 0 Tonga (Tonga Islands) 0 Tulu 0 Tumbuka 0 Tunisian Arabic 0 Turkish Sign Language 0 Udmurt 0 Venda 0 Veps 0 Vlaams 0 Vlax Romani 0 Votic 0 Zaza 0 Zeeuws 0 Zhuang 0

1464 dataset results for Texts AND English

GLUE (General Language Understanding Evaluation benchmark)

General Language Understanding Evaluation (GLUE) benchmark is a collection of nine natural language understanding tasks, including single-sentence tasks CoLA and SST-2, similarity and paraphrasing tasks MRPC, STS-B and QQP, and natural language inference tasks MNLI, QNLI, RTE and WNLI.

2,693 PAPERS • 45 BENCHMARKS

SST (Stanford Sentiment Treebank)

The Stanford Sentiment Treebank is a corpus with fully labeled parse trees that allows for a complete analysis of the compositional effects of sentiment in language. The corpus is based on the dataset introduced by Pang and Lee (2005) and consists of 11,855 single sentences extracted from movie reviews. It was parsed with the Stanford parser and includes a total of 215,154 unique phrases from those parse trees, each annotated by 3 human judges.

2,010 PAPERS • 13 BENCHMARKS

MultiNLI (Multi-Genre Natural Language Inference)

The Multi-Genre Natural Language Inference (MultiNLI) dataset has 433K sentence pairs. Its size and mode of collection are modeled closely like SNLI. MultiNLI offers ten distinct genres (Face-to-face, Telephone, 9/11, Travel, Letters, Oxford University Press, Slate, Verbatim, Goverment and Fiction) of written and spoken English data. There are matched dev/test sets which are derived from the same sources as those in the training set, and mismatched sets which do not closely resemble any seen at training time.

1,662 PAPERS • 8 BENCHMARKS

IMDb Movie Reviews

The IMDb Movie Reviews dataset is a binary sentiment analysis dataset consisting of 50,000 reviews from the Internet Movie Database (IMDb) labeled as positive or negative. The dataset contains an even number of positive and negative reviews. Only highly polarizing reviews are considered. A negative review has a score ≤ 4 out of 10, and a positive review has a score ≥ 7 out of 10. No more than 30 reviews are included per movie. The dataset contains additional unlabeled data.

1,569 PAPERS • 11 BENCHMARKS

SNLI (Stanford Natural Language Inference)

The SNLI dataset (Stanford Natural Language Inference) consists of 570k sentence-pairs manually labeled as entailment, contradiction, and neutral. Premises are image captions from Flickr30k, while hypotheses were generated by crowd-sourced annotators who were shown a premise and asked to generate entailing, contradicting, and neutral sentences. Annotators were instructed to judge the relation between sentences given that they describe the same event. Each pair is labeled as “entailment”, “neutral”, “contradiction” or “-”, where “-” indicates that an agreement could not be reached.

1,217 PAPERS • 3 BENCHMARKS

QNLI (Question-answering NLI)

The QNLI (Question-answering NLI) dataset is a Natural Language Inference dataset automatically derived from the Stanford Question Answering Dataset v1.1 (SQuAD). SQuAD v1.1 consists of question-paragraph pairs, where one of the sentences in the paragraph (drawn from Wikipedia) contains the answer to the corresponding question (written by an annotator). The dataset was converted into sentence pair classification by forming a pair between each question and each sentence in the corresponding context, and filtering out pairs with low lexical overlap between the question and the context sentence. The task is to determine whether the context sentence contains the answer to the question. This modified version of the original task removes the requirement that the model select the exact answer, but also removes the simplifying assumptions that the answer is always present in the input and that lexical overlap is a reliable cue. The QNLI dataset is part of GLUE benchmark.

1,046 PAPERS • 5 BENCHMARKS

Natural Questions

The Natural Questions corpus is a question answering dataset containing 307,373 training examples, 7,830 development examples, and 7,842 test examples. Each example is comprised of a google.com query and a corresponding Wikipedia page. Each Wikipedia page has a passage (or long answer) annotated on the page that answers the question and one or more short spans from the annotated passage containing the actual answer. The long and the short answer annotations can however be empty. If they are both empty, then there is no answer on the page at all. If the long answer annotation is non-empty, but the short answer annotation is empty, then the annotated passage answers the question but no explicit short answer could be found. Finally 1% of the documents have a passage annotated with a short answer that is “yes” or “no”, instead of a list of short spans.

988 PAPERS • 8 BENCHMARKS

MS MARCO (Microsoft Machine Reading Comprehension Dataset)

The MS MARCO (Microsoft MAchine Reading Comprehension) is a collection of datasets focused on deep learning in search. The first dataset was a question answering dataset featuring 100,000 real Bing questions and a human generated answer. Over time the collection was extended with a 1,000,000 question dataset, a natural language generation dataset, a passage ranking dataset, keyphrase extraction dataset, crawling dataset, and a conversational search.

816 PAPERS • 7 BENCHMARKS

WikiText-2

The WikiText language modeling dataset is a collection of over 100 million tokens extracted from the set of verified Good and Featured articles on Wikipedia. The dataset is available under the Creative Commons Attribution-ShareAlike License.

806 PAPERS • 5 BENCHMARKS

AG News (AG’s News Corpus)

AG News (AG’s News Corpus) is a subdataset of AG's corpus of news articles constructed by assembling titles and description fields of articles from the 4 largest classes (“World”, “Sports”, “Business”, “Sci/Tech”) of AG’s Corpus. The AG News contains 30,000 training and 1,900 test samples per class.

764 PAPERS • 10 BENCHMARKS

Flickr30k

The Flickr30k dataset contains 31,000 images collected from Flickr, together with 5 reference sentences provided by human annotators.

727 PAPERS • 9 BENCHMARKS

MRPC (Microsoft Research Paraphrase Corpus)

Microsoft Research Paraphrase Corpus (MRPC) is a corpus consists of 5,801 sentence pairs collected from newswire articles. Each pair is labelled if it is a paraphrase or not by human annotators. The whole set is divided into a training subset (4,076 sentence pairs of which 2,753 are paraphrases) and a test subset (1,725 pairs of which 1,147 are paraphrases).

688 PAPERS • 7 BENCHMARKS

MMLU (Massive Multitask Language Understanding)

MMLU (Massive Multitask Language Understanding) is a new benchmark designed to measure knowledge acquired during pretraining by evaluating models exclusively in zero-shot and few-shot settings. This makes the benchmark more challenging and more similar to how we evaluate humans. The benchmark covers 57 subjects across STEM, the humanities, the social sciences, and more. It ranges in difficulty from an elementary level to an advanced professional level, and it tests both world knowledge and problem solving ability. Subjects range from traditional areas, such as mathematics and history, to more specialized areas like law and ethics. The granularity and breadth of the subjects makes the benchmark ideal for identifying a model’s blind spots.

678 PAPERS • 26 BENCHMARKS

CoNLL 2003
CoNLL 2003

CoNLL-2003 is a named entity recognition dataset released as a part of CoNLL-2003 shared task: language-independent named entity recognition. The data consists of eight files covering two languages: English and German. For each of the languages there is a training file, a development file, a test file and a large file with unannotated data.

635 PAPERS • 16 BENCHMARKS

TriviaQA

TriviaQA is a realistic text-based question answering dataset which includes 950K question-answer pairs from 662K documents collected from Wikipedia and the web. This dataset is more challenging than standard QA benchmark datasets such as Stanford Question Answering Dataset (SQuAD), as the answers for a question may not be directly obtained by span prediction and the context is very long. TriviaQA dataset consists of both human-verified and machine-generated QA subsets.

624 PAPERS • 5 BENCHMARKS

HotpotQA

HotpotQA is a question answering dataset collected on the English Wikipedia, containing about 113K crowd-sourced questions that are constructed to require the introduction paragraphs of two Wikipedia articles to answer. Each question in the dataset comes with the two gold paragraphs, as well as a list of sentences in these paragraphs that crowdworkers identify as supporting facts necessary to answer the question.

555 PAPERS • 2 BENCHMARKS

DBpedia

DBpedia (from "DB" for "database") is a project aiming to extract structured content from the information created in the Wikipedia project. DBpedia allows users to semantically query relationships and properties of Wikipedia resources, including links to other related datasets.

551 PAPERS • 4 BENCHMARKS

MSR-VTT

MSR-VTT (Microsoft Research Video to Text) is a large-scale dataset for the open domain video captioning, which consists of 10,000 video clips from 20 categories, and each video clip is annotated with 20 English sentences by Amazon Mechanical Turks. There are about 29,000 unique words in all captions. The standard splits uses 6,513 clips for training, 497 clips for validation, and 2,990 clips for testing.

519 PAPERS • 7 BENCHMARKS

Universal Dependencies

The Universal Dependencies (UD) project seeks to develop cross-linguistically consistent treebank annotation of morphology and syntax for multiple languages. The first version of the dataset was released in 2015 and consisted of 10 treebanks over 10 languages. Version 2.7 released in 2020 consists of 183 treebanks over 104 languages. The annotation consists of UPOS (universal part-of-speech tags), XPOS (language-specific part-of-speech tags), Feats (universal morphological features), Lemmas, dependency heads and universal dependency labels.

505 PAPERS • 15 BENCHMARKS

WikiText-103

The WikiText language modeling dataset is a collection of over 100 million tokens extracted from the set of verified Good and Featured articles on Wikipedia. The dataset is available under the Creative Commons Attribution-ShareAlike License.

490 PAPERS • 2 BENCHMARKS

CNN/Daily Mail

CNN/Daily Mail is a dataset for text summarization. Human generated abstractive summary bullets were generated from news stories in CNN and Daily Mail websites as questions (with one of the entities hidden), and stories as the corresponding passages from which the system is expected to answer the fill-in the-blank question. The authors released the scripts that crawl, extract and generate pairs of passages and questions from these websites.

462 PAPERS • 13 BENCHMARKS

FEVER (Fact Extraction and VERification)

FEVER is a publicly available dataset for fact extraction and verification against textual sources.

411 PAPERS • 2 BENCHMARKS

BoolQ (Boolean Questions)

BoolQ is a question answering dataset for yes/no questions containing 15942 examples. These questions are naturally occurring – they are generated in unprompted and unconstrained settings. Each example is a triplet of (question, passage, answer), with the title of the page as optional additional context.

386 PAPERS • 2 BENCHMARKS

MATH

MATH is a new dataset of 12,500 challenging competition mathematics problems. Each problem in MATH has a full step-by-step solution which can be used to teach models to generate answer derivations and explanations.

370 PAPERS • 2 BENCHMARKS

DailyDialog

DailyDialog is a high-quality multi-turn open-domain English dialog dataset. It contains 13,118 dialogues split into a training set with 11,118 dialogues and validation and test sets with 1000 dialogues each. On average there are around 8 speaker turns per dialogue with around 15 tokens per turn.

367 PAPERS • 2 BENCHMARKS

SuperGLUE

SuperGLUE is a benchmark dataset designed to pose a more rigorous test of language understanding than GLUE. SuperGLUE has the same high-level motivation as GLUE: to provide a simple, hard-to-game measure of progress toward general-purpose language understanding technologies for English. SuperGLUE follows the basic design of GLUE: It consists of a public leaderboard built around eight language understanding tasks, drawing on existing data, accompanied by a single-number performance metric, and an analysis toolkit. However, it improves upon GLUE in several ways:

365 PAPERS • NO BENCHMARKS YET

VCTK
VCTK (CSTR VCTK Corpus)

This CSTR VCTK Corpus includes speech data uttered by 110 English speakers with various accents. Each speaker reads out about 400 sentences, which were selected from a newspaper, the rainbow passage and an elicitation paragraph used for the speech accent archive. The newspaper texts were taken from Herald Glasgow, with permission from Herald & Times Group. Each speaker has a different set of the newspaper texts selected based a greedy algorithm that increases the contextual and phonetic coverage. The details of the text selection algorithms are described in the following paper: C. Veaux, J. Yamagishi and S. King, "The voice bank corpus: Design, collection and data analysis of a large regional accent speech database," https://doi.org/10.1109/ICSDA.2013.6709856. The rainbow passage and elicitation paragraph are the same for all speakers. The rainbow passage can be found at International Dialects of English Archive: (http://web.ku.edu/~idea/readings/rainbow.htm). The elicitation paragraph

363 PAPERS • 5 BENCHMARKS

RACE (ReAding Comprehension dataset from Examinations)

The ReAding Comprehension dataset from Examinations (RACE) dataset is a machine reading comprehension dataset consisting of 27,933 passages and 97,867 questions from English exams, targeting Chinese students aged 12-18. RACE consists of two subsets, RACE-M and RACE-H, from middle school and high school exams, respectively. RACE-M has 28,293 questions and RACE-H has 69,574. Each question is associated with 4 candidate answers, one of which is correct. The data generation process of RACE differs from most machine reading comprehension datasets - instead of generating questions and answers by heuristics or crowd-sourcing, questions in RACE are specifically designed for testing human reading skills, and are created by domain experts.

347 PAPERS • 4 BENCHMARKS

OpenBookQA (OBQA)

OpenBookQA is a new kind of question-answering dataset modeled after open book exams for assessing human understanding of a subject. It consists of 5,957 multiple-choice elementary-level science questions (4,957 train, 500 dev, 500 test), which probe the understanding of a small “book” of 1,326 core science facts and the application of these facts to novel situations. For training, the dataset includes a mapping from each question to the core science fact it was designed to probe. Answering OpenBookQA questions requires additional broad common knowledge, not contained in the book. The questions, by design, are answered incorrectly by both a retrieval-based algorithm and a word co-occurrence algorithm. Additionally, the dataset includes a collection of 5,167 crowd-sourced common knowledge facts, and an expanded version of the train/dev/test questions where each question is associated with its originating core fact, a human accuracy score, a clarity score, and an anonymized crowd-worker

336 PAPERS • 3 BENCHMARKS

CommonsenseQA (CSQA)

The CommonsenseQA is a dataset for commonsense question answering task. The dataset consists of 12,247 questions with 5 choices each. The dataset was generated by Amazon Mechanical Turk workers in the following process (an example is provided in parentheses):

330 PAPERS • 1 BENCHMARK

MultiWOZ (Multi-domain Wizard-of-Oz)

The Multi-domain Wizard-of-Oz (MultiWOZ) dataset is a large-scale human-human conversational corpus spanning over seven domains, containing 8438 multi-turn dialogues, with each dialogue averaging 14 turns. Different from existing standard datasets like WOZ and DSTC2, which contain less than 10 slots and only a few hundred values, MultiWOZ has 30 (domain, slot) pairs and over 4,500 possible values. The dialogues span seven domains: restaurant, hotel, attraction, taxi, train, hospital and police.

316 PAPERS • 15 BENCHMARKS

Conceptual Captions

Automatic image captioning is the task of producing a natural-language utterance (usually a sentence) that correctly reflects the visual content of an image. Up to this point, the resource most used for this task was the MS-COCO dataset, containing around 120,000 images and 5-way image-caption annotations (produced by paid annotators).

311 PAPERS • 2 BENCHMARKS

WSC (Winograd Schema Challenge)

The Winograd Schema Challenge was introduced both as an alternative to the Turing Test and as a test of a system’s ability to do commonsense reasoning. A Winograd schema is a pair of sentences differing in one or two words with a highly ambiguous pronoun, resolved differently in the two sentences, that appears to require commonsense knowledge to be resolved correctly. The examples were designed to be easily solvable by humans but difficult for machines, in principle requiring a deep understanding of the content of the text and the situation it describes.

309 PAPERS • 2 BENCHMARKS

MPQA Opinion Corpus (Multi-Perspective Question Answering)

The MPQA Opinion Corpus contains 535 news articles from a wide variety of news sources manually annotated for opinions and other private states (i.e., beliefs, emotions, sentiments, speculations, etc.).

301 PAPERS • 3 BENCHMARKS

DROP (Discrete Reasoning Over Paragraphs)

Discrete Reasoning Over Paragraphs DROP is a crowdsourced, adversarially-created, 96k-question benchmark, in which a system must resolve references in a question, perhaps to multiple input positions, and perform discrete operations over them (such as addition, counting, or sorting). These operations require a much more comprehensive understanding of the content of paragraphs than what was necessary for prior datasets. The questions consist of passages extracted from Wikipedia articles. The dataset is split into a training set of about 77,000 questions, a development set of around 9,500 questions and a hidden test set similar in size to the development set.

293 PAPERS • 2 BENCHMARKS

COPA (Choice of Plausible Alternatives)

The Choice Of Plausible Alternatives (COPA) evaluation provides researchers with a tool for assessing progress in open-domain commonsense causal reasoning. COPA consists of 1000 questions, split equally into development and test sets of 500 questions each. Each question is composed of a premise and two alternatives, where the task is to select the alternative that more plausibly has a causal relation with the premise. The correct alternative is randomized so that the expected performance of randomly guessing is 50%.

278 PAPERS • 2 BENCHMARKS

MSVD (Microsoft Research Video Description Corpus)

The Microsoft Research Video Description Corpus (MSVD) dataset consists of about 120K sentences collected during the summer of 2010. Workers on Mechanical Turk were paid to watch a short video snippet and then summarize the action in a single sentence. The result is a set of roughly parallel descriptions of more than 2,000 video snippets. Because the workers were urged to complete the task in the language of their choice, both paraphrase and bilingual alternations are captured in the data.

278 PAPERS • 3 BENCHMARKS

LJSpeech (The LJ Speech Dataset)

This is a public domain speech dataset consisting of 13,100 short audio clips of a single speaker reading passages from 7 non-fiction books. A transcription is provided for each clip. Clips vary in length from 1 to 10 seconds and have a total length of approximately 24 hours. The texts were published between 1884 and 1964, and are in the public domain. The audio was recorded in 2016-17 by the LibriVox project and is also in the public domain.

268 PAPERS • 3 BENCHMARKS

New York Times Annotated Corpus

The New York Times Annotated Corpus contains over 1.8 million articles written and published by the New York Times between January 1, 1987 and June 19, 2007 with article metadata provided by the New York Times Newsroom, the New York Times Indexing Service and the online production staff at nytimes.com. The corpus includes:

264 PAPERS • 7 BENCHMARKS

ATIS (Airline Travel Information Systems)

The ATIS (Airline Travel Information Systems) is a dataset consisting of audio recordings and corresponding manual transcripts about humans asking for flight information on automated airline travel inquiry systems. The data consists of 17 unique intent categories. The original split contains 4478, 500 and 893 intent-labeled reference utterances in train, development and test set respectively.

263 PAPERS • 7 BENCHMARKS

OK-VQA (Outside Knowledge Visual Question Answering)

Outside Knowledge Visual Question Answering (OK-VQA) includes more than 14,000 questions that require external knowledge to answer.

255 PAPERS • 2 BENCHMARKS

NewsQA

The NewsQA dataset is a crowd-sourced machine reading comprehension dataset of 120,000 question-answer pairs.

249 PAPERS • 1 BENCHMARK

SNIPS (SNIPS Natural Language Understanding benchmark)

The SNIPS Natural Language Understanding benchmark is a dataset of over 16,000 crowdsourced queries distributed among 7 user intents of various complexity:

244 PAPERS • 6 BENCHMARKS

OntoNotes 5.0
OntoNotes 5.0

OntoNotes 5.0 is a large corpus comprising various genres of text (news, conversational telephone speech, weblogs, usenet newsgroups, broadcast, talk shows) in three languages (English, Chinese, and Arabic) with structural information (syntax and predicate argument structure) and shallow semantics (word sense linked to an ontology and coreference).

237 PAPERS • 12 BENCHMARKS

ICDAR 2013

The ICDAR 2013 dataset consists of 229 training images and 233 testing images, with word-level annotations provided. It is the standard benchmark dataset for evaluating near-horizontal text detection.

229 PAPERS • 3 BENCHMARKS

CoQA (Conversational Question Answering Challenge)

CoQA is a large-scale dataset for building Conversational Question Answering systems. The goal of the CoQA challenge is to measure the ability of machines to understand a text passage and answer a series of interconnected questions that appear in a conversation.

228 PAPERS • 2 BENCHMARKS

BIG-bench (Beyond the Imitation Game Benchmark)

The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to probe large language models and extrapolate their future capabilities. Big-bench include more than 200 tasks.

216 PAPERS • 134 BENCHMARKS

MELD (Multimodal EmotionLines Dataset)

Multimodal EmotionLines Dataset (MELD) has been created by enhancing and extending EmotionLines dataset. MELD contains the same dialogue instances available in EmotionLines, but it also encompasses audio and visual modality along with text. MELD has more than 1400 dialogues and 13000 utterances from Friends TV series. Multiple speakers participated in the dialogues. Each utterance in a dialogue has been labeled by any of these seven emotions -- Anger, Disgust, Sadness, Joy, Neutral, Surprise and Fear. MELD also has sentiment (positive, negative and neutral) annotation for each utterance.

208 PAPERS • 2 BENCHMARKS

Contact us on: hello@paperswithcode.com . Papers With Code is a free resource with all data licensed under CC-BY-SA.
Terms Data policy Cookies policy