TensorFlow Datasets

TFDS provides a collection of ready-to-use datasets for use with TensorFlow, Jax, and other Machine Learning frameworks.

It handles downloading and preparing the data deterministically and constructing a tf.data.Dataset (or np.array).

View on TensorFlow.org

Run in Google Colab

View source on GitHub

Download notebook

Installation

TFDS exists in two packages:

• pip install tensorflow-datasets: The stable version, released every few months.
• pip install tfds-nightly: Released every day, contains the last versions of the datasets.

This colab uses tfds-nightly:

pip install -q tfds-nightly tensorflow matplotlib

import matplotlib.pyplot as plt
import numpy as np
import tensorflow as tf

import tensorflow_datasets as tfds

Find available datasets

All dataset builders are subclass of tfds.core.DatasetBuilder. To get the list of available builders, use tfds.list_builders() or look at our catalog.

tfds.list_builders()

['abstract_reasoning',
 'accentdb',
 'aeslc',
 'aflw2k3d',
 'ag_news_subset',
 'ai2_arc',
 'ai2_arc_with_ir',
 'amazon_us_reviews',
 'anli',
 'arc',
 'bair_robot_pushing_small',
 'bccd',
 'beans',
 'big_patent',
 'bigearthnet',
 'billsum',
 'binarized_mnist',
 'binary_alpha_digits',
 'blimp',
 'bool_q',
 'c4',
 'caltech101',
 'caltech_birds2010',
 'caltech_birds2011',
 'cars196',
 'cassava',
 'cats_vs_dogs',
 'celeb_a',
 'celeb_a_hq',
 'cfq',
 'cherry_blossoms',
 'chexpert',
 'cifar10',
 'cifar100',
 'cifar10_1',
 'cifar10_corrupted',
 'citrus_leaves',
 'cityscapes',
 'civil_comments',
 'clevr',
 'clic',
 'clinc_oos',
 'cmaterdb',
 'cnn_dailymail',
 'coco',
 'coco_captions',
 'coil100',
 'colorectal_histology',
 'colorectal_histology_large',
 'common_voice',
 'coqa',
 'cos_e',
 'cosmos_qa',
 'covid19sum',
 'crema_d',
 'curated_breast_imaging_ddsm',
 'cycle_gan',
 'dart',
 'davis',
 'deep_weeds',
 'definite_pronoun_resolution',
 'dementiabank',
 'diabetic_retinopathy_detection',
 'div2k',
 'dmlab',
 'dolphin_number_word',
 'downsampled_imagenet',
 'drop',
 'dsprites',
 'dtd',
 'duke_ultrasound',
 'e2e_cleaned',
 'efron_morris75',
 'emnist',
 'eraser_multi_rc',
 'esnli',
 'eurosat',
 'fashion_mnist',
 'flic',
 'flores',
 'food101',
 'forest_fires',
 'fuss',
 'gap',
 'geirhos_conflict_stimuli',
 'gem',
 'genomics_ood',
 'german_credit_numeric',
 'gigaword',
 'glue',
 'goemotions',
 'gpt3',
 'gref',
 'groove',
 'gtzan',
 'gtzan_music_speech',
 'hellaswag',
 'higgs',
 'horses_or_humans',
 'howell',
 'i_naturalist2017',
 'imagenet2012',
 'imagenet2012_corrupted',
 'imagenet2012_real',
 'imagenet2012_subset',
 'imagenet_a',
 'imagenet_r',
 'imagenet_resized',
 'imagenet_v2',
 'imagenette',
 'imagewang',
 'imdb_reviews',
 'irc_disentanglement',
 'iris',
 'kitti',
 'kmnist',
 'lambada',
 'lfw',
 'librispeech',
 'librispeech_lm',
 'libritts',
 'ljspeech',
 'lm1b',
 'lost_and_found',
 'lsun',
 'lvis',
 'malaria',
 'math_dataset',
 'mctaco',
 'mlqa',
 'mnist',
 'mnist_corrupted',
 'movie_lens',
 'movie_rationales',
 'movielens',
 'moving_mnist',
 'multi_news',
 'multi_nli',
 'multi_nli_mismatch',
 'natural_questions',
 'natural_questions_open',
 'newsroom',
 'nsynth',
 'nyu_depth_v2',
 'ogbg_molpcba',
 'omniglot',
 'open_images_challenge2019_detection',
 'open_images_v4',
 'openbookqa',
 'opinion_abstracts',
 'opinosis',
 'opus',
 'oxford_flowers102',
 'oxford_iiit_pet',
 'para_crawl',
 'patch_camelyon',
 'paws_wiki',
 'paws_x_wiki',
 'pet_finder',
 'pg19',
 'piqa',
 'places365_small',
 'plant_leaves',
 'plant_village',
 'plantae_k',
 'qa4mre',
 'qasc',
 'quac',
 'quickdraw_bitmap',
 'race',
 'radon',
 'reddit',
 'reddit_disentanglement',
 'reddit_tifu',
 'resisc45',
 'robonet',
 'rock_paper_scissors',
 'rock_you',
 's3o4d',
 'salient_span_wikipedia',
 'samsum',
 'savee',
 'scan',
 'scene_parse150',
 'scicite',
 'scientific_papers',
 'sentiment140',
 'shapes3d',
 'siscore',
 'smallnorb',
 'snli',
 'so2sat',
 'speech_commands',
 'spoken_digit',
 'squad',
 'stanford_dogs',
 'stanford_online_products',
 'star_cfq',
 'starcraft_video',
 'stl10',
 'story_cloze',
 'sun397',
 'super_glue',
 'svhn_cropped',
 'tao',
 'ted_hrlr_translate',
 'ted_multi_translate',
 'tedlium',
 'tf_flowers',
 'the300w_lp',
 'tiny_shakespeare',
 'titanic',
 'trec',
 'trivia_qa',
 'tydi_qa',
 'uc_merced',
 'ucf101',
 'vctk',
 'vgg_face2',
 'visual_domain_decathlon',
 'voc',
 'voxceleb',
 'voxforge',
 'waymo_open_dataset',
 'web_nlg',
 'web_questions',
 'wider_face',
 'wiki40b',
 'wiki_bio',
 'wiki_table_questions',
 'wiki_table_text',
 'wikiann',
 'wikihow',
 'wikipedia',
 'wikipedia_toxicity_subtypes',
 'wine_quality',
 'winogrande',
 'wmt13_translate',
 'wmt14_translate',
 'wmt15_translate',
 'wmt16_translate',
 'wmt17_translate',
 'wmt18_translate',
 'wmt19_translate',
 'wmt_t2t_translate',
 'wmt_translate',
 'wordnet',
 'wsc273',
 'xnli',
 'xquad',
 'xsum',
 'xtreme_pawsx',
 'xtreme_xnli',
 'yelp_polarity_reviews',
 'yes_no',
 'youtube_vis',
 'huggingface:acronym_identification',
 'huggingface:ade_corpus_v2',
 'huggingface:adversarial_qa',
 'huggingface:aeslc',
 'huggingface:afrikaans_ner_corpus',
 'huggingface:ag_news',
 'huggingface:ai2_arc',
 'huggingface:air_dialogue',
 'huggingface:ajgt_twitter_ar',
 'huggingface:allegro_reviews',
 'huggingface:allocine',
 'huggingface:alt',
 'huggingface:amazon_polarity',
 'huggingface:amazon_reviews_multi',
 'huggingface:amazon_us_reviews',
 'huggingface:ambig_qa',
 'huggingface:amttl',
 'huggingface:anli',
 'huggingface:app_reviews',
 'huggingface:aqua_rat',
 'huggingface:aquamuse',
 'huggingface:ar_cov19',
 'huggingface:ar_res_reviews',
 'huggingface:ar_sarcasm',
 'huggingface:arabic_billion_words',
 'huggingface:arabic_pos_dialect',
 'huggingface:arabic_speech_corpus',
 'huggingface:arcd',
 'huggingface:arsentd_lev',
 'huggingface:art',
 'huggingface:arxiv_dataset',
 'huggingface:aslg_pc12',
 'huggingface:asnq',
 'huggingface:asset',
 'huggingface:assin',
 'huggingface:assin2',
 'huggingface:atomic',
 'huggingface:autshumato',
 'huggingface:bbc_hindi_nli',
 'huggingface:bc2gm_corpus',
 'huggingface:best2009',
 'huggingface:bianet',
 'huggingface:bible_para',
 'huggingface:big_patent',
 'huggingface:billsum',
 'huggingface:bing_coronavirus_query_set',
 'huggingface:biomrc',
 'huggingface:blended_skill_talk',
 'huggingface:blimp',
 'huggingface:blog_authorship_corpus',
 'huggingface:bn_hate_speech',
 'huggingface:bookcorpus',
 'huggingface:bookcorpusopen',
 'huggingface:boolq',
 'huggingface:bprec',
 'huggingface:break_data',
 'huggingface:brwac',
 'huggingface:bsd_ja_en',
 'huggingface:bswac',
 'huggingface:c3',
 'huggingface:c4',
 'huggingface:cail2018',
 'huggingface:caner',
 'huggingface:capes',
 'huggingface:catalonia_independence',
 'huggingface:cawac',
 'huggingface:cc100',
 'huggingface:cc_news',
 'huggingface:cdsc',
 'huggingface:cdt',
 'huggingface:cfq',
 'huggingface:chr_en',
 'huggingface:cifar10',
 'huggingface:cifar100',
 'huggingface:circa',
 'huggingface:civil_comments',
 'huggingface:clickbait_news_bg',
 'huggingface:climate_fever',
 'huggingface:clinc_oos',
 'huggingface:clue',
 'huggingface:cmrc2018',
 'huggingface:cnn_dailymail',
 'huggingface:coached_conv_pref',
 'huggingface:coarse_discourse',
 'huggingface:codah',
 'huggingface:code_search_net',
 'huggingface:com_qa',
 'huggingface:common_gen',
 'huggingface:commonsense_qa',
 'huggingface:compguesswhat',
 'huggingface:conceptnet5',
 'huggingface:conll2000',
 'huggingface:conll2002',
 'huggingface:conll2003',
 'huggingface:conv_ai',
 'huggingface:conv_ai_2',
 'huggingface:conv_ai_3',
 'huggingface:coqa',
 'huggingface:cord19',
 'huggingface:cornell_movie_dialog',
 'huggingface:cos_e',
 'huggingface:cosmos_qa',
 'huggingface:counter',
 'huggingface:covid_qa_castorini',
 'huggingface:covid_qa_deepset',
 'huggingface:covid_qa_ucsd',
 'huggingface:covid_tweets_japanese',
 'huggingface:craigslist_bargains',
 'huggingface:crawl_domain',
 'huggingface:crd3',
 'huggingface:crime_and_punish',
 'huggingface:crows_pairs',
 'huggingface:cs_restaurants',
 'huggingface:curiosity_dialogs',
 'huggingface:daily_dialog',
 'huggingface:dane',
 'huggingface:danish_political_comments',
 'huggingface:dart',
 'huggingface:datacommons_factcheck',
 'huggingface:dbpedia_14',
 'huggingface:dbrd',
 'huggingface:deal_or_no_dialog',
 'huggingface:definite_pronoun_resolution',
 'huggingface:dengue_filipino',
 'huggingface:dialog_re',
 'huggingface:diplomacy_detection',
 'huggingface:disaster_response_messages',
 'huggingface:discofuse',
 'huggingface:discovery',
 'huggingface:doc2dial',
 'huggingface:docred',
 'huggingface:doqa',
 'huggingface:dream',
 'huggingface:drop',
 'huggingface:duorc',
 'huggingface:dutch_social',
 'huggingface:dyk',
 'huggingface:e2e_nlg',
 'huggingface:e2e_nlg_cleaned',
 'huggingface:ecb',
 'huggingface:ehealth_kd',
 'huggingface:eitb_parcc',
 'huggingface:eli5',
 'huggingface:emea',
 'huggingface:emo',
 'huggingface:emotion',
 'huggingface:emotone_ar',
 'huggingface:empathetic_dialogues',
 'huggingface:enriched_web_nlg',
 'huggingface:eraser_multi_rc',
 'huggingface:esnli',
 'huggingface:eth_py150_open',
 'huggingface:ethos',
 'huggingface:euronews',
 'huggingface:europa_eac_tm',
 'huggingface:europa_ecdc_tm',
 'huggingface:event2Mind',
 'huggingface:evidence_infer_treatment',
 'huggingface:exams',
 'huggingface:factckbr',
 'huggingface:fake_news_english',
 'huggingface:fake_news_filipino',
 'huggingface:farsi_news',
 'huggingface:fever',
 'huggingface:finer',
 'huggingface:flores',
 'huggingface:flue',
 'huggingface:fquad',
 'huggingface:freebase_qa',
 'huggingface:gap',
 'huggingface:gem',
 'huggingface:generated_reviews_enth',
 'huggingface:generics_kb',
 'huggingface:german_legal_entity_recognition',
 'huggingface:germaner',
 'huggingface:germeval_14',
 'huggingface:giga_fren',
 'huggingface:gigaword',
 'huggingface:glucose',
 'huggingface:glue',
 'huggingface:gnad10',
 'huggingface:go_emotions',
 'huggingface:google_wellformed_query',
 'huggingface:grail_qa',
 'huggingface:great_code',
 'huggingface:guardian_authorship',
 'huggingface:gutenberg_time',
 'huggingface:hans',
 'huggingface:hansards',
 'huggingface:hard',
 'huggingface:harem',
 'huggingface:has_part',
 'huggingface:hate_offensive',
 'huggingface:hate_speech18',
 'huggingface:hate_speech_filipino',
 'huggingface:hate_speech_offensive',
 'huggingface:hate_speech_pl',
 'huggingface:hate_speech_portuguese',
 'huggingface:hatexplain',
 'huggingface:hausa_voa_ner',
 'huggingface:hausa_voa_topics',
 'huggingface:hda_nli_hindi',
 'huggingface:head_qa',
 'huggingface:health_fact',
 'huggingface:hebrew_projectbenyehuda',
 'huggingface:hebrew_sentiment',
 'huggingface:hebrew_this_world',
 'huggingface:hellaswag',
 'huggingface:hind_encorp',
 'huggingface:hindi_discourse',
 'huggingface:hippocorpus',
 'huggingface:hkcancor',
 'huggingface:hope_edi',
 'huggingface:hotpot_qa',
 'huggingface:hover',
 'huggingface:hrenwac_para',
 'huggingface:hrwac',
 'huggingface:humicroedit',
 'huggingface:hybrid_qa',
 'huggingface:hyperpartisan_news_detection',
 'huggingface:id_clickbait',
 'huggingface:id_liputan6',
 'huggingface:id_nergrit_corpus',
 'huggingface:id_newspapers_2018',
 'huggingface:id_panl_bppt',
 'huggingface:id_puisi',
 'huggingface:igbo_english_machine_translation',
 'huggingface:igbo_monolingual',
 'huggingface:igbo_ner',
 'huggingface:ilist',
 'huggingface:imdb',
 'huggingface:imdb_urdu_reviews',
 'huggingface:imppres',
 'huggingface:indic_glue',
 'huggingface:indonlu',
 'huggingface:inquisitive_qg',
 'huggingface:interpress_news_category_tr',
 'huggingface:irc_disentangle',
 'huggingface:isixhosa_ner_corpus',
 'huggingface:isizulu_ner_corpus',
 'huggingface:iwslt2017',
 'huggingface:jeopardy',
 'huggingface:jfleg',
 'huggingface:jigsaw_toxicity_pred',
 'huggingface:jnlpba',
 'huggingface:journalists_questions',
 'huggingface:kannada_news',
 'huggingface:kd_conv',
 'huggingface:kde4',
 'huggingface:kelm',
 'huggingface:kilt_tasks',
 'huggingface:kilt_wikipedia',
 'huggingface:kinnews_kirnews',
 'huggingface:kor_3i4k',
 'huggingface:kor_hate',
 'huggingface:kor_ner',
 'huggingface:kor_nli',
 'huggingface:kor_nlu',
 'huggingface:kor_qpair',
 'huggingface:kor_sae',
 'huggingface:kor_sarcasm',
 'huggingface:labr',
 'huggingface:lama',
 'huggingface:lambada',
 'huggingface:large_spanish_corpus',
 'huggingface:lc_quad',
 'huggingface:lener_br',
 'huggingface:liar',
 'huggingface:librispeech_asr',
 'huggingface:librispeech_lm',
 'huggingface:limit',
 'huggingface:lince',
 'huggingface:linnaeus',
 'huggingface:liveqa',
 'huggingface:lj_speech',
 'huggingface:lm1b',
 'huggingface:lst20',
 'huggingface:mac_morpho',
 'huggingface:makhzan',
 'huggingface:math_dataset',
 'huggingface:math_qa',
 'huggingface:matinf',
 'huggingface:mc_taco',
 'huggingface:md_gender_bias',
 'huggingface:med_hop',
 'huggingface:medal',
 'huggingface:medical_dialog',
 'huggingface:medical_questions_pairs',
 'huggingface:menyo20k_mt',
 'huggingface:meta_woz',
 'huggingface:metooma',
 'huggingface:metrec',
 'huggingface:mkb',
 'huggingface:mkqa',
 'huggingface:mlqa',
 'huggingface:mlsum',
 'huggingface:mnist',
 'huggingface:mocha',
 'huggingface:movie_rationales',
 'huggingface:mrqa',
 'huggingface:ms_marco',
 'huggingface:ms_terms',
 'huggingface:msr_genomics_kbcomp',
 'huggingface:msr_sqa',
 'huggingface:msr_text_compression',
 'huggingface:msr_zhen_translation_parity',
 'huggingface:msra_ner',
 'huggingface:mt_eng_vietnamese',
 'huggingface:muchocine',
 'huggingface:multi_booked',
 'huggingface:multi_news',
 'huggingface:multi_nli',
 'huggingface:multi_nli_mismatch',
 'huggingface:multi_para_crawl',
 'huggingface:multi_re_qa',
 'huggingface:multi_woz_v22',
 'huggingface:multi_x_science_sum',
 'huggingface:mutual_friends',
 'huggingface:mwsc',
 'huggingface:myanmar_news',
 'huggingface:narrativeqa',
 'huggingface:narrativeqa_manual',
 'huggingface:natural_questions',
 'huggingface:ncbi_disease',
 'huggingface:nchlt',
 'huggingface:ncslgr',
 'huggingface:nell',
 'huggingface:neural_code_search',
 'huggingface:news_commentary',
 'huggingface:newsgroup',
 'huggingface:newsph',
 'huggingface:newsph_nli',
 'huggingface:newsqa',
 'huggingface:newsroom',
 'huggingface:nkjp-ner',
 'huggingface:nli_tr',
 'huggingface:norwegian_ner',
 'huggingface:nq_open',
 'huggingface:nsmc',
 'huggingface:numer_sense',
 'huggingface:numeric_fused_head',
 'huggingface:oclar',
 'huggingface:offcombr',
 'huggingface:offenseval2020_tr',
 'huggingface:offenseval_dravidian',
 'huggingface:ofis_publik',
 'huggingface:ohsumed',
 'huggingface:ollie',
 'huggingface:omp',
 'huggingface:onestop_english',
 'huggingface:open_subtitles',
 'huggingface:openbookqa',
 'huggingface:openwebtext',
 'huggingface:opinosis',
 'huggingface:opus100',
 'huggingface:opus_books',
 'huggingface:opus_dgt',
 'huggingface:opus_dogc',
 'huggingface:opus_elhuyar',
 'huggingface:opus_euconst',
 'huggingface:opus_finlex',
 'huggingface:opus_fiskmo',
 'huggingface:opus_gnome',
 'huggingface:opus_infopankki',
 'huggingface:opus_memat',
 'huggingface:opus_montenegrinsubs',
 'huggingface:opus_openoffice',
 'huggingface:opus_paracrawl',
 'huggingface:opus_rf',
 'huggingface:opus_tedtalks',
 'huggingface:opus_ubuntu',
 'huggingface:opus_wikipedia',
 'huggingface:opus_xhosanavy',
 'huggingface:orange_sum',
 'huggingface:oscar',
 'huggingface:para_crawl',
 'huggingface:para_pat',
 'huggingface:paws',
 'huggingface:paws-x',
 'huggingface:pec',
 'huggingface:peer_read',
 'huggingface:peoples_daily_ner',
 'huggingface:per_sent',
 'huggingface:persian_ner',
 'huggingface:pg19',
 'huggingface:php',
 'huggingface:piaf',
 'huggingface:pib',
 'huggingface:piqa',
 'huggingface:pn_summary',
 'huggingface:poem_sentiment',
 'huggingface:polemo2',
 'huggingface:poleval2019_cyberbullying',
 'huggingface:poleval2019_mt',
 'huggingface:polsum',
 'huggingface:polyglot_ner',
 'huggingface:prachathai67k',
 'huggingface:pragmeval',
 'huggingface:proto_qa',
 'huggingface:psc',
 'huggingface:ptb_text_only',
 'huggingface:pubmed',
 'huggingface:pubmed_qa',
 'huggingface:py_ast',
 'huggingface:qa4mre',
 'huggingface:qa_srl',
 'huggingface:qa_zre',
 'huggingface:qangaroo',
 'huggingface:qanta',
 'huggingface:qasc',
 'huggingface:qed',
 'huggingface:qed_amara',
 'huggingface:quac',
 'huggingface:quail',
 'huggingface:quarel',
 'huggingface:quartz',
 'huggingface:quora',
 'huggingface:quoref',
 'huggingface:race',
 'huggingface:re_dial',
 'huggingface:reasoning_bg',
 'huggingface:recipe_nlg',
 'huggingface:reclor',
 'huggingface:reddit',
 'huggingface:reddit_tifu',
 'huggingface:refresd',
 'huggingface:reuters21578',
 'huggingface:roman_urdu',
 'huggingface:ronec',
 'huggingface:ropes',
 'huggingface:rotten_tomatoes',
 'huggingface:s2orc',
 'huggingface:samsum',
 'huggingface:sanskrit_classic',
 'huggingface:saudinewsnet',
 'huggingface:scan',
 'huggingface:scb_mt_enth_2020',
 'huggingface:schema_guided_dstc8',
 'huggingface:scicite',
 'huggingface:scielo',
 'huggingface:scientific_papers',
 'huggingface:scifact',
 'huggingface:sciq',
 'huggingface:scitail',
 'huggingface:scitldr',
 'huggingface:search_qa',
 'huggingface:selqa',
 'huggingface:sem_eval_2010_task_8',
 'huggingface:sem_eval_2014_task_1',
 'huggingface:sem_eval_2020_task_11',
 'huggingface:sent_comp',
 'huggingface:senti_lex',
 'huggingface:senti_ws',
 'huggingface:sentiment140',
 'huggingface:sepedi_ner',
 'huggingface:sesotho_ner_corpus',
 'huggingface:setimes',
 'huggingface:setswana_ner_corpus',
 'huggingface:sharc',
 'huggingface:sharc_modified',
 'huggingface:sick',
 'huggingface:silicone',
 'huggingface:simple_questions_v2',
 'huggingface:siswati_ner_corpus',
 'huggingface:smartdata',
 'huggingface:sms_spam',
 'huggingface:snips_built_in_intents',
 'huggingface:snli',
 'huggingface:snow_simplified_japanese_corpus',
 'huggingface:so_stacksample',
 'huggingface:social_bias_frames',
 'huggingface:social_i_qa',
 'huggingface:sofc_materials_articles',
 'huggingface:sogou_news',
 'huggingface:spanish_billion_words',
 'huggingface:spc',
 'huggingface:species_800',
 'huggingface:spider',
 'huggingface:squad',
 'huggingface:squad_adversarial',
 'huggingface:squad_es',
 'huggingface:squad_it',
 'huggingface:squad_kor_v1',
 'huggingface:squad_kor_v2',
 'huggingface:squad_v1_pt',
 'huggingface:squad_v2',
 'huggingface:squadshifts',
 'huggingface:srwac',
 'huggingface:stereoset',
 'huggingface:stsb_mt_sv',
 'huggingface:style_change_detection',
 'huggingface:super_glue',
 'huggingface:swag',
 'huggingface:swahili',
 'huggingface:swahili_news',
 'huggingface:swda',
 'huggingface:swedish_ner_corpus',
 'huggingface:swedish_reviews',
 'huggingface:tab_fact',
 'huggingface:tamilmixsentiment',
 'huggingface:tanzil',
 'huggingface:tapaco',
 'huggingface:tashkeela',
 'huggingface:taskmaster1',
 'huggingface:taskmaster2',
 'huggingface:taskmaster3',
 'huggingface:tatoeba',
 'huggingface:ted_hrlr',
 'huggingface:ted_iwlst2013',
 'huggingface:ted_multi',
 'huggingface:ted_talks_iwslt',
 'huggingface:telugu_books',
 'huggingface:telugu_news',
 'huggingface:tep_en_fa_para',
 'huggingface:thai_toxicity_tweet',
 'huggingface:thainer',
 'huggingface:thaiqa_squad',
 'huggingface:thaisum',
 'huggingface:tilde_model',
 'huggingface:times_of_india_news_headlines',
 'huggingface:tiny_shakespeare',
 'huggingface:tlc',
 'huggingface:tmu_gfm_dataset',
 'huggingface:totto',
 'huggingface:trec',
 'huggingface:trivia_qa',
 'huggingface:tsac',
 'huggingface:ttc4900',
 'huggingface:tunizi',
 'huggingface:tuple_ie',
 'huggingface:turk',
 'huggingface:turkish_movie_sentiment',
 'huggingface:turkish_ner',
 'huggingface:turkish_product_reviews',
 'huggingface:turkish_shrinked_ner',
 'huggingface:turku_ner_corpus',
 'huggingface:tweet_eval',
 'huggingface:tweet_qa',
 'huggingface:tweets_ar_en_parallel',
 'huggingface:tweets_hate_speech_detection',
 'huggingface:twi_text_c3',
 'huggingface:twi_wordsim353',
 'huggingface:tydiqa',
 'huggingface:ubuntu_dialogs_corpus',
 'huggingface:udhr',
 'huggingface:um005',
 'huggingface:un_ga',
 'huggingface:un_multi',
 'huggingface:un_pc',
 'huggingface:universal_dependencies',
 'huggingface:universal_morphologies',
 'huggingface:urdu_fake_news',
 'huggingface:urdu_sentiment_corpus',
 'huggingface:web_nlg',
 'huggingface:web_of_science',
 'huggingface:web_questions',
 'huggingface:weibo_ner',
 'huggingface:wi_locness',
 'huggingface:wiki40b',
 'huggingface:wiki_asp',
 'huggingface:wiki_atomic_edits',
 'huggingface:wiki_auto',
 'huggingface:wiki_bio',
 'huggingface:wiki_dpr',
 'huggingface:wiki_hop',
 'huggingface:wiki_lingua',
 'huggingface:wiki_movies',
 'huggingface:wiki_qa',
 'huggingface:wiki_qa_ar',
 'huggingface:wiki_snippets',
 'huggingface:wiki_source',
 'huggingface:wiki_split',
 'huggingface:wiki_summary',
 'huggingface:wikiann',
 'huggingface:wikicorpus',
 'huggingface:wikihow',
 'huggingface:wikipedia',
 'huggingface:wikisql',
 'huggingface:wikitext',
 'huggingface:wikitext_tl39',
 'huggingface:wili_2018',
 'huggingface:wino_bias',
 'huggingface:winograd_wsc',
 'huggingface:winogrande',
 'huggingface:wiqa',
 'huggingface:wisesight1000',
 'huggingface:wisesight_sentiment',
 'huggingface:wmt14',
 'huggingface:wmt15',
 'huggingface:wmt16',
 'huggingface:wmt17',
 'huggingface:wmt18',
 'huggingface:wmt19',
 'huggingface:wmt20_mlqe_task1',
 'huggingface:wmt20_mlqe_task2',
 'huggingface:wmt20_mlqe_task3',
 'huggingface:wmt_t2t',
 'huggingface:wnut_17',
 'huggingface:wongnai_reviews',
 'huggingface:woz_dialogue',
 'huggingface:wrbsc',
 'huggingface:x_stance',
 'huggingface:xcopa',
 'huggingface:xed_en_fi',
 'huggingface:xglue',
 'huggingface:xnli',
 'huggingface:xor_tydi_qa',
 'huggingface:xquad',
 'huggingface:xquad_r',
 'huggingface:xsum',
 'huggingface:xsum_factuality',
 'huggingface:xtreme',
 'huggingface:yahoo_answers_qa',
 'huggingface:yahoo_answers_topics',
 'huggingface:yelp_polarity',
 'huggingface:yelp_review_full',
 'huggingface:yoruba_bbc_topics',
 'huggingface:yoruba_gv_ner',
 'huggingface:yoruba_text_c3',
 'huggingface:yoruba_wordsim353',
 'huggingface:youtube_caption_corrections',
 'huggingface:zest']

Load a dataset

tfds.load

The easiest way of loading a dataset is tfds.load. It will:

1. Download the data and save it as tfrecord files.
2. Load the tfrecord and create the tf.data.Dataset.

ds = tfds.load('mnist', split='train', shuffle_files=True)
assert isinstance(ds, tf.data.Dataset)
print(ds)

<_OptionsDataset shapes: {image: (28, 28, 1), label: ()}, types: {image: tf.uint8, label: tf.int64}>

Some common arguments:

• split=: Which split to read (e.g. 'train', ['train', 'test'], 'train[80%:]',...). See our split API guide.
• shuffle_files=: Control whether to shuffle the files between each epoch (TFDS store big datasets in multiple smaller files).
• data_dir=: Location where the dataset is saved ( defaults to ~/tensorflow_datasets/)
• with_info=True: Returns the tfds.core.DatasetInfo containing dataset metadata
• download=False: Disable download

tfds.builder

tfds.load is a thin wrapper around tfds.core.DatasetBuilder. You can get the same output using the tfds.core.DatasetBuilder API:

builder = tfds.builder('mnist')
# 1. Create the tfrecord files (no-op if already exists)
builder.download_and_prepare()
# 2. Load the `tf.data.Dataset`
ds = builder.as_dataset(split='train', shuffle_files=True)
print(ds)

<_OptionsDataset shapes: {image: (28, 28, 1), label: ()}, types: {image: tf.uint8, label: tf.int64}>

tfds build CLI

If you want to generate a specific dataset, you can use the tfds command line. For example:

tfds build mnist

See the doc for available flags.

Iterate over a dataset

As dict

By default, the tf.data.Dataset object contains a dict of tf.Tensors:

ds = tfds.load('mnist', split='train')
ds = ds.take(1)  # Only take a single example

for example in ds:  # example is `{'image': tf.Tensor, 'label': tf.Tensor}`
  print(list(example.keys()))
  image = example["image"]
  label = example["label"]
  print(image.shape, label)

['image', 'label']
(28, 28, 1) tf.Tensor(4, shape=(), dtype=int64)

To find out the dict key names and structure, look at the dataset documentation in our catalog. For example: mnist documentation.

As tuple (as_supervised=True)

By using as_supervised=True, you can get a tuple (features, label) instead for supervised datasets.

ds = tfds.load('mnist', split='train', as_supervised=True)
ds = ds.take(1)

for image, label in ds:  # example is (image, label)
  print(image.shape, label)

(28, 28, 1) tf.Tensor(4, shape=(), dtype=int64)

As numpy (tfds.as_numpy)

Uses tfds.as_numpy to convert:

• tf.Tensor -> np.array
• tf.data.Dataset -> Iterator[Tree[np.array]] (Tree can be arbitrary nested Dict, Tuple)

ds = tfds.load('mnist', split='train', as_supervised=True)
ds = ds.take(1)

for image, label in tfds.as_numpy(ds):
  print(type(image), type(label), label)

<class 'numpy.ndarray'> <class 'numpy.int64'> 4

As batched tf.Tensor (batch_size=-1)

By using batch_size=-1, you can load the full dataset in a single batch.

This can be combined with as_supervised=True and tfds.as_numpy to get the the data as (np.array, np.array):

image, label = tfds.as_numpy(tfds.load(
    'mnist',
    split='test',
    batch_size=-1,
    as_supervised=True,
))

print(type(image), image.shape)

<class 'numpy.ndarray'> (10000, 28, 28, 1)

Be careful that your dataset can fit in memory, and that all examples have the same shape.

Benchmark your datasets

Benchmarking a dataset is a simple tfds.benchmark call on any iterable (e.g. tf.data.Dataset, tfds.as_numpy,...).

ds = tfds.load('mnist', split='train')
ds = ds.batch(32).prefetch(1)

tfds.benchmark(ds, batch_size=32)
tfds.benchmark(ds, batch_size=32)  # Second epoch much faster due to auto-caching

************ Summary ************

Examples/sec (First included) 37648.72 ex/sec (total: 60000 ex, 1.59 sec)
Examples/sec (First only) 53.84 ex/sec (total: 32 ex, 0.59 sec)
Examples/sec (First excluded) 60009.42 ex/sec (total: 59968 ex, 1.00 sec)

************ Summary ************

Examples/sec (First included) 300819.05 ex/sec (total: 60000 ex, 0.20 sec)
Examples/sec (First only) 2555.37 ex/sec (total: 32 ex, 0.01 sec)
Examples/sec (First excluded) 320799.77 ex/sec (total: 59968 ex, 0.19 sec)

• Do not forget to normalize the results per batch size with the batch_size= kwarg.
• In the summary, the first warmup batch is separated from the other ones to capture tf.data.Dataset extra setup time (e.g. buffers initialization,...).
• Notice how the second iteration is much faster due to TFDS auto-caching.
• tfds.benchmark returns a tfds.core.BenchmarkResult which can be inspected for further analysis.

Build end-to-end pipeline

To go further, you can look:

• Our end-to-end Keras example to see a full training pipeline (with batching, shuffling,...).
• Our performance guide to improve the speed of your pipelines (tip: use tfds.benchmark(ds) to benchmark your datasets).

Visualization

tfds.as_dataframe

tf.data.Dataset objects can be converted to pandas.DataFrame with tfds.as_dataframe to be visualized on Colab.

• Add the tfds.core.DatasetInfo as second argument of tfds.as_dataframe to visualize images, audio, texts, videos,...
• Use ds.take(x) to only display the first x examples. pandas.DataFrame will load the full dataset in-memory, and can be very expensive to display.

ds, info = tfds.load('mnist', split='train', with_info=True)

tfds.as_dataframe(ds.take(4), info)

tfds.show_examples

tfds.show_examples returns a matplotlib.figure.Figure (only image datasets supported now):

ds, info = tfds.load('mnist', split='train', with_info=True)

fig = tfds.show_examples(ds, info)

Access the dataset metadata

All builders include a tfds.core.DatasetInfo object containing the dataset metadata.

It can be accessed through:

• The tfds.load API:

ds, info = tfds.load('mnist', with_info=True)

• The tfds.core.DatasetBuilder API:

builder = tfds.builder('mnist')
info = builder.info

The dataset info contains additional informations about the dataset (version, citation, homepage, description,...).

print(info)

tfds.core.DatasetInfo(
    name='mnist',
    full_name='mnist/3.0.1',
    description="""
    The MNIST database of handwritten digits.
    """,
    homepage='http://yann.lecun.com/exdb/mnist/',
    data_path='gs://tensorflow-datasets/datasets/mnist/3.0.1',
    download_size=11.06 MiB,
    dataset_size=21.00 MiB,
    features=FeaturesDict({
        'image': Image(shape=(28, 28, 1), dtype=tf.uint8),
        'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=10),
    }),
    supervised_keys=('image', 'label'),
    splits={
        'test': <SplitInfo num_examples=10000, num_shards=1>,
        'train': <SplitInfo num_examples=60000, num_shards=1>,
    },
    citation="""@article{lecun2010mnist,
      title={MNIST handwritten digit database},
      author={LeCun, Yann and Cortes, Corinna and Burges, CJ},
      journal={ATT Labs [Online]. Available: http://yann.lecun.com/exdb/mnist},
      volume={2},
      year={2010}
    }""",
)

Features metadata (label names, image shape,...)

Access the tfds.features.FeatureDict:

info.features

FeaturesDict({
    'image': Image(shape=(28, 28, 1), dtype=tf.uint8),
    'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=10),
})

Number of classes, label names:

print(info.features["label"].num_classes)
print(info.features["label"].names)
print(info.features["label"].int2str(7))  # Human readable version (8 -> 'cat')
print(info.features["label"].str2int('7'))

10
['0', '1', '2', '3', '4', '5', '6', '7', '8', '9']
7
7

Shapes, dtypes:

print(info.features.shape)
print(info.features.dtype)
print(info.features['image'].shape)
print(info.features['image'].dtype)

{'image': (28, 28, 1), 'label': ()}
{'image': tf.uint8, 'label': tf.int64}
(28, 28, 1)
<dtype: 'uint8'>

Split metadata (e.g. split names, number of examples,...)

Access the tfds.core.SplitDict:

print(info.splits)

{'test': <SplitInfo num_examples=10000, num_shards=1>, 'train': <SplitInfo num_examples=60000, num_shards=1>}

Available splits:

print(list(info.splits.keys()))

['test', 'train']

Get info on individual split:

print(info.splits['train'].num_examples)
print(info.splits['train'].filenames)
print(info.splits['train'].num_shards)

60000
['mnist-train.tfrecord-00000-of-00001']
1

It also works with the subsplit API:

print(info.splits['train[15%:75%]'].num_examples)
print(info.splits['train[15%:75%]'].file_instructions)

36000
[FileInstruction(filename='mnist-train.tfrecord-00000-of-00001', skip=9000, take=36000, num_examples=36000)]

Troubleshooting

Manual download (if download fails)

If download fails for some reason (e.g. offline,...). You can always manually download the data yourself and place it in the manual_dir (defaults to ~/tensorflow_datasets/download/manual/.

To find out which urls to download, look into:

• For new datasets (implemented as folder): tensorflow_datasets/<type>/<dataset_name>/checksums.tsv. For example: tensorflow_datasets/text/bool_q/checksums.tsv.

  You can find the dataset source location in our catalog.

• For old datasets: tensorflow_datasets/url_checksums/<dataset_name>.txt

Fixing NonMatchingChecksumError

TFDS ensure determinism by validating the checksums of downloaded urls. If NonMatchingChecksumError is raised, might indicate:

• The website may be down (e.g. 503 status code). Please check the url.
• For Google Drive URLs, try again later as Drive sometimes rejects downloads when too many people access the same URL. See bug
• The original datasets files may have been updated. In this case the TFDS dataset builder should be updated. Please open a new Github issue or PR:
  • Register the new checksums with tfds build --register_checksums
  • Eventually update the dataset generation code.
  • Update the dataset VERSION
  • Update the dataset RELEASE_NOTES: What caused the checksums to change ? Did some examples changed ?
  • Make sure the dataset can still be built.
  • Send us a PR

Citation

If you're using tensorflow-datasets for a paper, please include the following citation, in addition to any citation specific to the used datasets (which can be found in the dataset catalog).

@misc{TFDS,
  title = { {TensorFlow Datasets}, A collection of ready-to-use datasets},
  howpublished = {\url{https://www.tensorflow.org/datasets} },
}