html_url
stringlengths
48
51
title
stringlengths
5
268
comments
stringlengths
70
51.8k
body
stringlengths
0
29.8k
comment_length
int64
16
1.52k
text
stringlengths
164
54.1k
embeddings
sequence
https://github.com/huggingface/datasets/issues/228
Not able to access the XNLI dataset
Thanks for reporting this bug ! As it seems to be just a cache problem, I closed your PR. I think we might just need to clear and reload the `xnli` cache @srush ?
When I try to access the XNLI dataset, I get the following error. The option of plain_text get selected automatically and then I get the following error. ``` FileNotFoundError: [Errno 2] No such file or directory: '/home/sasha/.cache/huggingface/datasets/xnli/plain_text/1.0.0/dataset_info.json' Traceback: File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/ScriptRunner.py", line 322, in _run_script exec(code, module.__dict__) File "/home/sasha/nlp_viewer/run.py", line 86, in <module> dts, fail = get(str(option.id), str(conf_option.name) if conf_option else None) File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 591, in wrapped_func return get_or_create_cached_value() File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 575, in get_or_create_cached_value return_value = func(*args, **kwargs) File "/home/sasha/nlp_viewer/run.py", line 72, in get builder_instance = builder_cls(name=conf) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 610, in __init__ super(GeneratorBasedBuilder, self).__init__(*args, **kwargs) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 152, in __init__ self.info = DatasetInfo.from_directory(self._cache_dir) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/info.py", line 157, in from_directory with open(os.path.join(dataset_info_dir, DATASET_INFO_FILENAME), "r") as f: ``` Is it possible to see if the dataset_info.json is correctly placed?
34
Not able to access the XNLI dataset When I try to access the XNLI dataset, I get the following error. The option of plain_text get selected automatically and then I get the following error. ``` FileNotFoundError: [Errno 2] No such file or directory: '/home/sasha/.cache/huggingface/datasets/xnli/plain_text/1.0.0/dataset_info.json' Traceback: File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/ScriptRunner.py", line 322, in _run_script exec(code, module.__dict__) File "/home/sasha/nlp_viewer/run.py", line 86, in <module> dts, fail = get(str(option.id), str(conf_option.name) if conf_option else None) File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 591, in wrapped_func return get_or_create_cached_value() File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 575, in get_or_create_cached_value return_value = func(*args, **kwargs) File "/home/sasha/nlp_viewer/run.py", line 72, in get builder_instance = builder_cls(name=conf) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 610, in __init__ super(GeneratorBasedBuilder, self).__init__(*args, **kwargs) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 152, in __init__ self.info = DatasetInfo.from_directory(self._cache_dir) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/info.py", line 157, in from_directory with open(os.path.join(dataset_info_dir, DATASET_INFO_FILENAME), "r") as f: ``` Is it possible to see if the dataset_info.json is correctly placed? Thanks for reporting this bug ! As it seems to be just a cache problem, I closed your PR. I think we might just need to clear and reload the `xnli` cache @srush ?
[ -0.14086730778217316, -0.1892593950033188, -0.03353660926222801, 0.4693676829338074, 0.40544605255126953, 0.11200098693370819, 0.05603647604584694, 0.35449883341789246, -0.10948893427848816, 0.29277610778808594, -0.14041024446487427, 0.16212044656276703, 0.050227634608745575, -0.028046416118741035, 0.050541575998067856, -0.03532493859529495, -0.16912731528282166, 0.15102030336856842, 0.13207347691059113, 0.005323123186826706, -0.17077098786830902, 0.04558214917778969, -0.269063800573349, 0.258741557598114, -0.28670641779899597, -0.08528313040733337, 0.09144605696201324, 0.38597825169563293, -0.30816975235939026, -0.5349295139312744, 0.04508195072412491, -0.14433525502681732, 0.24077387154102325, 0.19576595723628998, -0.00011105803423561156, 0.049332939088344574, 0.26291322708129883, -0.05734888091683388, -0.5286890268325806, -0.14541779458522797, -0.35688552260398865, -0.1231442391872406, -0.014699206687510014, -0.3132307231426239, -0.12878146767616272, -0.2209601253271103, 0.4022443890571594, -0.3689225912094116, 0.27180513739585876, 0.31527572870254517, 0.2039080262184143, 0.20004284381866455, 0.1218271255493164, -0.0547143891453743, 0.049337755888700485, 0.13519403338432312, -0.13144788146018982, -0.01397060789167881, 0.2392851859331131, 0.1932622790336609, 0.1636674851179123, 0.45788300037384033, -0.1711951047182083, -0.11164625734090805, 0.2552073299884796, 0.24384237825870514, -0.03029867634177208, -0.16049787402153015, 0.3209938704967499, 0.23313458263874054, 0.44375672936439514, -0.43328917026519775, -0.433499276638031, -0.11148592084646225, -0.0038185573648661375, -0.2649720311164856, 0.24886402487754822, 0.1001942902803421, -0.31770753860473633, 0.1554601490497589, -0.11993837356567383, -0.3665497303009033, -0.36813774704933167, 0.14540188014507294, 0.26279720664024353, 0.3802296817302704, -0.05968890339136124, 0.0670861005783081, -0.061199408024549484, -0.1958378702402115, 0.05281243845820427, -0.08949365466833115, -0.16413246095180511, 0.3026692867279053, -0.22698448598384857, 0.12764638662338257, 0.13528870046138763, 0.057152315974235535, 0.11984661221504211, 0.36280354857444763, 0.2032436579465866, -0.18401408195495605, -0.37554052472114563, 0.06257743388414383, -0.2645843029022217, 0.42281022667884827, 0.29570093750953674, 0.14088575541973114, 0.38271573185920715, -0.008944886736571789, 0.22189567983150482, -0.36490076780319214, -0.2945748567581177, -0.16017690300941467, -0.11438383162021637, 0.2509688138961792, 0.3615806996822357, -0.1595381647348404, -0.24764789640903473, -0.0595756396651268, -0.12639573216438293, 0.01934901811182499, 0.20292766392230988, 0.5645591020584106, 0.05364596098661423, 0.18948371708393097, 0.03513291850686073, 0.32177770137786865, 0.053703561425209045, -0.5332473516464233, -0.17430844902992249, 0.216541588306427, -0.30879613757133484, 0.14569483697414398, 0.310526579618454, -0.3313242197036743, 0.30016303062438965, -0.11575473099946976, -0.30328500270843506, 0.0032360709737986326, 0.12947247922420502, -0.1749178171157837, 0.10203266888856888, 0.07796114683151245, 0.09318613260984421, 0.10305090993642807, 0.19042626023292542, -0.0799684152007103, 0.06550320982933044, -0.0332774892449379, -0.047192495316267014, -0.48522502183914185, 0.025507068261504173, 0.13068552315235138, 0.07165248692035675, -0.009505627676844597, -0.08786679059267044, -0.05430307611823082, 0.19623017311096191, -0.22172188758850098, -0.11173594743013382, 0.11289877444505692, -0.02556431293487549, -0.27134573459625244, 0.21264219284057617, 0.5540777444839478, -0.313611775636673, 0.2092668116092682, -0.19852620363235474, -0.1348922848701477, 0.16316285729408264, 0.36284351348876953, -0.11672932654619217, 0.07758820801973343, -0.3848819434642792, 0.32663220167160034, 0.6447758674621582, -0.5202512145042419, -0.5076890587806702, 0.281650573015213, -0.22235873341560364, -0.07156804203987122, 0.5067176222801208, 0.24071164429187775, -0.06453099101781845, -0.11318366974592209, 0.04700065031647682, 0.11700887978076935, 0.04672114923596382, -0.08850125223398209, -0.11148851364850998, -0.24508868157863617, 0.3402372896671295, 0.11302019655704498, -0.030680567026138306, 0.132090762257576, -0.029109716415405273, 0.16131943464279175, 0.3716827630996704, -0.2910832464694977, 0.151186004281044, 0.2840609550476074, 0.30056697130203247, 0.03720917925238609, -0.1425778716802597, -0.11960358172655106, -0.2946610152721405, 0.23998598754405975, -0.09203005582094193, -0.020116332918405533, -0.30888083577156067, 0.00025334686506539583, -0.11308930814266205, -0.0409194678068161, -0.38762593269348145, -0.2783125638961792, 0.19049637019634247, 0.45270541310310364, 0.22734399139881134, -0.01965280994772911, -0.28994232416152954, 0.31739184260368347, -0.4826754033565521, 0.002716717543080449, -0.6354674696922302, 0.12806348502635956, -0.09318788349628448, -0.14903484284877777, -0.16396395862102509, 0.1639254093170166, 0.023549996316432953, -0.0028481101617217064, -0.09274742752313614, 0.2124699354171753, -0.2380760759115219, 0.08131196349859238, 0.2638709843158722, 0.10236212611198425, 0.03820393979549408, -0.5709121823310852, 0.337582528591156, 0.471760094165802, 0.22842155396938324, 0.08941147476434708, -0.2693725526332855, 0.11830697953701019, -0.0825481116771698, 0.0903005376458168, 0.043728042393922806, 0.01720329001545906, 0.46662530303001404, 0.014810673892498016, 0.06998161971569061, -0.21279384195804596, 0.3003251254558563, 0.21648988127708435, 0.46217772364616394, -0.08612532913684845, -0.5364862680435181, 0.08613129705190659, 0.279600590467453, 0.06051954627037048, -0.09864404052495956, -0.01093630027025938, -0.3510042428970337, -0.04163467511534691, 0.055063143372535706, 0.22897891700267792, 0.28139010071754456, 0.22780916094779968, -0.06535445898771286, 0.08311142027378082, 0.09837327152490616, -0.2621801793575287, 0.1807290017604828, -0.25792714953422546, 0.10123013705015182, 0.26987096667289734, 0.19980137050151825, 0.03205357864499092, -0.03564421087503433, -0.5033636093139648, 0.07216988503932953, 0.44549039006233215, -0.2703445553779602, -0.17282365262508392, -0.1369037926197052, -0.6854809522628784, 0.060650646686553955, 0.07638580352067947, -0.11054092645645142, -0.34322455525398254, -0.02222832664847374, 0.18775984644889832, 0.21456144750118256, -0.07199577987194061, -0.09505292773246765, 0.07554131001234055, 0.06864757835865021, -0.10940901190042496, -0.41933175921440125, -0.3149597942829132, -0.20777536928653717, 0.04685645550489426, 0.27079495787620544, -0.02280697040259838, 0.24694304168224335, -0.2802794277667999, 0.06392866373062134, -0.4144374430179596, -0.025489769876003265, 0.023476963862776756, 0.029296601191163063, -0.0004743135941680521, -0.03406819328665733, 0.48188820481300354, 0.47751712799072266, -0.062323737889528275, 0.34758952260017395, 0.04154730215668678, -0.17423060536384583, 0.2578032612800598, -0.21523937582969666, -0.23448677361011505, -0.04180895909667015, -0.5249137282371521, -0.35636940598487854, -0.3282740116119385, 0.08275364339351654, -0.13988102972507477, 0.0803988054394722, 0.31809043884277344, -0.1464771330356598, 0.2062617689371109, -0.03888341411948204, 0.19586744904518127, -0.07545904070138931, -0.2256142646074295, 0.3382493555545807, -0.16139675676822662, -0.2159951776266098, 0.23609325289726257, 0.06490588188171387, 0.21008798480033875, -0.04124198481440544, -0.23060105741024017, 0.17212596535682678, -0.22557012736797333, 0.28446343541145325, -0.1305036097764969, -0.15192051231861115, 0.34998252987861633, -0.010504299774765968, -0.0014597176341339946, 0.004380505066365004, 0.01188093051314354, 0.11780867725610733, 0.010027376934885979, 0.571620762348175, 0.20394249260425568, 0.41344502568244934, 0.14451955258846283, 0.3723565638065338, 0.44866421818733215, 0.026759330183267593, 0.32839006185531616, -0.26213935017585754, 0.0669078454375267, -0.06588185578584671, -0.3156568706035614, 0.11945436894893646, 0.17475633323192596, -0.17332448065280914, -0.10797219723463058, 0.138683021068573, 0.38894420862197876, -0.33151981234550476, -0.13302171230316162, -0.2874072194099426, -0.08548351377248764, -0.14637787640094757, 0.02025029808282852, 0.011118978261947632, 0.12911097705364227, 0.2093753218650818, -0.34255450963974, -0.3424661159515381, -0.11213251948356628, 0.4633776545524597, 0.0023362934589385986, 0.15231388807296753, -0.2744362950325012, -0.28418034315109253, -0.5618224143981934, 0.48062583804130554, 0.11800606548786163, 0.21206669509410858, -0.2758864164352417, 0.12500028312206268, 0.2663460671901703, 0.022575542330741882, 0.3283258378505707, -0.3004832863807678, 0.25636452436447144, 0.08078701049089432, -0.09893045574426651, -0.23588041961193085, 0.1696392297744751, -0.15016189217567444, 0.1985001564025879, 0.33587154746055603, 0.19395440816879272, -0.160841166973114, -0.11107291281223297, -0.016944129019975662, 0.10838181525468826, -0.1749693900346756, -0.21759386360645294, -0.017783721908926964, -0.0037044950295239687, -0.16341423988342285, -0.328482985496521, -0.1715681403875351, 0.24727433919906616, 0.13296154141426086, 0.33589187264442444, 0.04687206819653511, -0.07427357882261276, 0.17847338318824768, 0.22041575610637665, 0.33928728103637695, -0.09613542258739471, -0.15330880880355835, 0.1652642786502838, 0.2498362809419632, 0.07175866514444351, 0.49867498874664307, 0.28984472155570984, -0.3622654974460602, -0.014100667089223862, -0.22840864956378937, 0.12723597884178162, 0.2903362810611725, -0.25874918699264526, 0.10405367612838745, 0.034366361796855927, -0.00782689731568098, 0.0616951584815979, -0.09590546786785126, 0.25131773948669434, 0.03154374659061432, -0.1259423941373825, -0.3986915946006775, 0.32729730010032654, 0.10116904973983765, 0.003972935024648905, 0.4540020525455475, -0.14631907641887665, -0.22761699557304382, 0.023734675720334053, 0.05723154917359352, 0.9369785785675049, 0.12992772459983826, 0.22923718392848969, 0.3930838406085968, 0.16831186413764954, 0.5260001420974731, -0.30351364612579346, 0.36880162358283997, -0.38715463876724243, -0.08039484173059464, -0.0029480515513569117, -0.11370399594306946, -0.011664546094834805, -0.1774902492761612, -0.08031443506479263, 0.23407699167728424, -0.15633440017700195, 0.00611140625551343, -0.09859159588813782, 0.2628976106643677, -0.3349466025829315, -0.09046969562768936, -0.46061474084854126, 0.0847524031996727, -0.20800389349460602, 0.3349795937538147, -0.10503844916820526, -0.1498464047908783, -0.38530439138412476, 0.1126159206032753, -0.42020905017852783, 0.37589675188064575, -0.0859742984175682, 0.12252394109964371, -0.07361681759357452, -0.3276995122432709, -0.05533306673169136, 0.6137094497680664, 0.5807364583015442, 0.0501268096268177, -0.33917272090911865, 0.007756409700959921, -0.21130119264125824, -0.023090843111276627, -0.09389214962720871, 0.15532875061035156, 0.33820468187332153, -0.04085510969161987, -0.30917733907699585, 0.5142865777015686, -0.11258584260940552, 0.016242269426584244, -0.2366236448287964, -0.25668174028396606, -0.24029716849327087, -0.043224040418863297, -0.1230141818523407, 0.01992630772292614, -0.12974593043327332, 0.025179490447044373, 0.09974995255470276, 0.012665953487157822, -0.23290403187274933, -0.2464853972196579, -0.1339302361011505, -0.1444869339466095, 0.18300442397594452, 0.325381875038147, 0.047445446252822876, 0.1761949509382248, 0.6140017509460449, 0.22167590260505676, -0.0501692108809948, -0.24814759194850922, 0.06540445238351822, 0.15567706525325775, -0.2556737959384918, -0.21238085627555847, 0.10414279252290726, 0.17469249665737152, 0.096544049680233, 0.2774945795536041, 0.18646995723247528, -0.006257296539843082, -0.14283394813537598, -0.448702335357666, -0.5422465205192566, 0.1517493724822998, -0.19254674017429352, 0.13901548087596893, -0.1516946256160736, 0.2596282958984375, 0.11874712258577347, -0.2652503252029419, -0.3370278477668762, -0.09570170193910599, -0.32341060042381287, -0.005745049100369215, 0.3934481143951416, 0.05905287712812424, 0.19687820971012115, 0.04367268458008766, 0.13710691034793854, -0.036700136959552765, -0.2849011719226837, -0.15982723236083984, 0.048877786844968796, 0.10002417862415314, 0.1371709108352661, -0.4027869701385498, -0.11658159643411636, -0.28826212882995605, -0.08812721073627472, -0.1604430079460144, 0.05028767138719559, 0.29853424429893494, 0.03162437304854393, 0.2532196044921875, -0.31334546208381653, 0.058008331805467606, -0.2136867344379425, 0.1573789119720459, -0.15283744037151337, 0.18410615622997284, 0.05986139923334122, 0.12613047659397125, 0.01767575554549694, 0.009883997030556202, -0.550724446773529, -0.06257440149784088, -0.0907987728714943, 0.1829756200313568, 0.11525699496269226, -0.4182232618331909, 0.22805631160736084, -0.14199501276016235, 0.1714201122522354, 0.19626367092132568, -0.03331765905022621, 0.2832377552986145, 0.18114230036735535, 0.18628519773483276, -0.23866654932498932, 0.20860911905765533, 0.215780109167099, 0.09463916718959808, -0.09388977289199829, 0.04307954013347626, -0.049454621970653534, -0.02566336840391159, 0.2692066729068756, -0.1024499237537384, 0.3867937922477722, -0.4072207510471344, -0.052489642053842545, 0.11946655809879303, 0.07139203697443008, -0.05252387002110481, 0.11530407518148422, 0.3662637770175934, 0.16459278762340546, 0.7827573418617249, -0.24703241884708405, 0.14657171070575714, -0.18498532474040985, 0.3281104862689972, -0.24878236651420593, -0.3345986008644104, -0.33597898483276367, 0.11348120868206024, -0.294269859790802, -0.12617403268814087, -0.18528421223163605, 0.2995415925979614, -0.10181374847888947, 0.14273475110530853, -0.23577985167503357, 0.39215415716171265, -0.14883604645729065, -0.04539475217461586, -0.5625836849212646, -0.22663353383541107, 0.10585590451955795, -0.1797053962945938, 0.025090746581554413, 0.01599329523742199, -0.009149348363280296, 0.1309439241886139, -0.2652382552623749, -0.40185388922691345, 0.3093724250793457, 0.40477555990219116, -0.016356229782104492, -0.2793296277523041, 0.41067326068878174, 0.29058921337127686, 0.06595078855752945, 0.20533108711242676, 0.3192925453186035, 0.42760297656059265, 0.4022080600261688, 0.06650746613740921, 0.20000669360160828, -0.10694017261266708, -0.25286200642585754, -0.013881637714803219, 0.05252344533801079, 0.30146437883377075, 0.05845767632126808, 0.31253117322921753, 0.14970730245113373, -0.0062344688922166824, 0.14302103221416473, 0.04569145664572716, 0.3462026119232178, 0.09222348034381866, -0.015020581893622875, -0.14654797315597534, -0.24552859365940094, -0.1894637942314148, -0.1285674124956131, -0.5581473708152771, 0.1378030627965927, 0.21281974017620087, 0.2150430977344513, 0.25495705008506775, -0.004208188503980637, 0.04900941625237465, -0.06657014787197113, 0.13284258544445038, 0.5250786542892456, -0.03401148319244385, -0.27033546566963196, -0.32096952199935913, -0.5026273727416992, 0.16537779569625854, -0.02528504841029644, -0.4208659529685974, -0.24853651225566864, 0.12115411460399628, -0.17552419006824493, -0.2355567365884781, 0.24101929366588593, 0.1014041155576706, 0.06959380954504013, 0.33923885226249695, -0.22689945995807648, -0.33274635672569275, 0.2696407437324524, 0.0531267449259758, -0.17090114951133728, -0.17286787927150726, -0.0946459174156189, -0.14867638051509857, 0.049000054597854614, -0.04603859409689903, 0.3904182016849518, 0.007113414816558361, 0.20573078095912933, 0.19779559969902039, 0.19981218874454498, 0.48119643330574036, 0.06806674599647522, -0.0404394194483757, -0.3731052577495575, -0.11173686385154724, -0.20125903189182281, -0.041570909321308136, 0.3111204504966736, 0.5081033706665039, -0.358381986618042, -0.00834614597260952, -0.38156816363334656, 0.1278298944234848, -0.16659346222877502, -0.01748354732990265, -0.2955944538116455, 0.3703540861606598, -0.2835775315761566, 0.23727823793888092, 0.16417279839515686, -0.04555123671889305, -0.029033057391643524, 0.10304306447505951, -0.33316659927368164, 0.0036254317965358496, 0.5232228636741638, -0.2101983278989792, -0.23654088377952576, 0.09007932245731354, 0.19331854581832886, 0.11350388824939728, 0.01034716609865427, -0.33786872029304504, 0.17566758394241333, 0.18242225050926208, -0.0053519015200436115, -0.24062062799930573, -0.08169142156839371, 0.0652153491973877, 0.11355213075876236, -0.06532713025808334, -0.1427069753408432, 0.15939858555793762, -0.07285217195749283, -0.013610105030238628, -0.08653882890939713 ]
https://github.com/huggingface/datasets/issues/228
Not able to access the XNLI dataset
Update: The dataset_info.json error is gone, but we have a new one instead: ``` ConnectionError: Couldn't reach https://www.nyu.edu/projects/bowman/xnli/XNLI-1.0.zip ``` I am not able to reproduce on my side unfortunately. Any idea @srush ?
When I try to access the XNLI dataset, I get the following error. The option of plain_text get selected automatically and then I get the following error. ``` FileNotFoundError: [Errno 2] No such file or directory: '/home/sasha/.cache/huggingface/datasets/xnli/plain_text/1.0.0/dataset_info.json' Traceback: File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/ScriptRunner.py", line 322, in _run_script exec(code, module.__dict__) File "/home/sasha/nlp_viewer/run.py", line 86, in <module> dts, fail = get(str(option.id), str(conf_option.name) if conf_option else None) File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 591, in wrapped_func return get_or_create_cached_value() File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 575, in get_or_create_cached_value return_value = func(*args, **kwargs) File "/home/sasha/nlp_viewer/run.py", line 72, in get builder_instance = builder_cls(name=conf) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 610, in __init__ super(GeneratorBasedBuilder, self).__init__(*args, **kwargs) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 152, in __init__ self.info = DatasetInfo.from_directory(self._cache_dir) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/info.py", line 157, in from_directory with open(os.path.join(dataset_info_dir, DATASET_INFO_FILENAME), "r") as f: ``` Is it possible to see if the dataset_info.json is correctly placed?
33
Not able to access the XNLI dataset When I try to access the XNLI dataset, I get the following error. The option of plain_text get selected automatically and then I get the following error. ``` FileNotFoundError: [Errno 2] No such file or directory: '/home/sasha/.cache/huggingface/datasets/xnli/plain_text/1.0.0/dataset_info.json' Traceback: File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/ScriptRunner.py", line 322, in _run_script exec(code, module.__dict__) File "/home/sasha/nlp_viewer/run.py", line 86, in <module> dts, fail = get(str(option.id), str(conf_option.name) if conf_option else None) File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 591, in wrapped_func return get_or_create_cached_value() File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 575, in get_or_create_cached_value return_value = func(*args, **kwargs) File "/home/sasha/nlp_viewer/run.py", line 72, in get builder_instance = builder_cls(name=conf) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 610, in __init__ super(GeneratorBasedBuilder, self).__init__(*args, **kwargs) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 152, in __init__ self.info = DatasetInfo.from_directory(self._cache_dir) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/info.py", line 157, in from_directory with open(os.path.join(dataset_info_dir, DATASET_INFO_FILENAME), "r") as f: ``` Is it possible to see if the dataset_info.json is correctly placed? Update: The dataset_info.json error is gone, but we have a new one instead: ``` ConnectionError: Couldn't reach https://www.nyu.edu/projects/bowman/xnli/XNLI-1.0.zip ``` I am not able to reproduce on my side unfortunately. Any idea @srush ?
[ -0.14086730778217316, -0.1892593950033188, -0.03353660926222801, 0.4693676829338074, 0.40544605255126953, 0.11200098693370819, 0.05603647604584694, 0.35449883341789246, -0.10948893427848816, 0.29277610778808594, -0.14041024446487427, 0.16212044656276703, 0.050227634608745575, -0.028046416118741035, 0.050541575998067856, -0.03532493859529495, -0.16912731528282166, 0.15102030336856842, 0.13207347691059113, 0.005323123186826706, -0.17077098786830902, 0.04558214917778969, -0.269063800573349, 0.258741557598114, -0.28670641779899597, -0.08528313040733337, 0.09144605696201324, 0.38597825169563293, -0.30816975235939026, -0.5349295139312744, 0.04508195072412491, -0.14433525502681732, 0.24077387154102325, 0.19576595723628998, -0.00011105803423561156, 0.049332939088344574, 0.26291322708129883, -0.05734888091683388, -0.5286890268325806, -0.14541779458522797, -0.35688552260398865, -0.1231442391872406, -0.014699206687510014, -0.3132307231426239, -0.12878146767616272, -0.2209601253271103, 0.4022443890571594, -0.3689225912094116, 0.27180513739585876, 0.31527572870254517, 0.2039080262184143, 0.20004284381866455, 0.1218271255493164, -0.0547143891453743, 0.049337755888700485, 0.13519403338432312, -0.13144788146018982, -0.01397060789167881, 0.2392851859331131, 0.1932622790336609, 0.1636674851179123, 0.45788300037384033, -0.1711951047182083, -0.11164625734090805, 0.2552073299884796, 0.24384237825870514, -0.03029867634177208, -0.16049787402153015, 0.3209938704967499, 0.23313458263874054, 0.44375672936439514, -0.43328917026519775, -0.433499276638031, -0.11148592084646225, -0.0038185573648661375, -0.2649720311164856, 0.24886402487754822, 0.1001942902803421, -0.31770753860473633, 0.1554601490497589, -0.11993837356567383, -0.3665497303009033, -0.36813774704933167, 0.14540188014507294, 0.26279720664024353, 0.3802296817302704, -0.05968890339136124, 0.0670861005783081, -0.061199408024549484, -0.1958378702402115, 0.05281243845820427, -0.08949365466833115, -0.16413246095180511, 0.3026692867279053, -0.22698448598384857, 0.12764638662338257, 0.13528870046138763, 0.057152315974235535, 0.11984661221504211, 0.36280354857444763, 0.2032436579465866, -0.18401408195495605, -0.37554052472114563, 0.06257743388414383, -0.2645843029022217, 0.42281022667884827, 0.29570093750953674, 0.14088575541973114, 0.38271573185920715, -0.008944886736571789, 0.22189567983150482, -0.36490076780319214, -0.2945748567581177, -0.16017690300941467, -0.11438383162021637, 0.2509688138961792, 0.3615806996822357, -0.1595381647348404, -0.24764789640903473, -0.0595756396651268, -0.12639573216438293, 0.01934901811182499, 0.20292766392230988, 0.5645591020584106, 0.05364596098661423, 0.18948371708393097, 0.03513291850686073, 0.32177770137786865, 0.053703561425209045, -0.5332473516464233, -0.17430844902992249, 0.216541588306427, -0.30879613757133484, 0.14569483697414398, 0.310526579618454, -0.3313242197036743, 0.30016303062438965, -0.11575473099946976, -0.30328500270843506, 0.0032360709737986326, 0.12947247922420502, -0.1749178171157837, 0.10203266888856888, 0.07796114683151245, 0.09318613260984421, 0.10305090993642807, 0.19042626023292542, -0.0799684152007103, 0.06550320982933044, -0.0332774892449379, -0.047192495316267014, -0.48522502183914185, 0.025507068261504173, 0.13068552315235138, 0.07165248692035675, -0.009505627676844597, -0.08786679059267044, -0.05430307611823082, 0.19623017311096191, -0.22172188758850098, -0.11173594743013382, 0.11289877444505692, -0.02556431293487549, -0.27134573459625244, 0.21264219284057617, 0.5540777444839478, -0.313611775636673, 0.2092668116092682, -0.19852620363235474, -0.1348922848701477, 0.16316285729408264, 0.36284351348876953, -0.11672932654619217, 0.07758820801973343, -0.3848819434642792, 0.32663220167160034, 0.6447758674621582, -0.5202512145042419, -0.5076890587806702, 0.281650573015213, -0.22235873341560364, -0.07156804203987122, 0.5067176222801208, 0.24071164429187775, -0.06453099101781845, -0.11318366974592209, 0.04700065031647682, 0.11700887978076935, 0.04672114923596382, -0.08850125223398209, -0.11148851364850998, -0.24508868157863617, 0.3402372896671295, 0.11302019655704498, -0.030680567026138306, 0.132090762257576, -0.029109716415405273, 0.16131943464279175, 0.3716827630996704, -0.2910832464694977, 0.151186004281044, 0.2840609550476074, 0.30056697130203247, 0.03720917925238609, -0.1425778716802597, -0.11960358172655106, -0.2946610152721405, 0.23998598754405975, -0.09203005582094193, -0.020116332918405533, -0.30888083577156067, 0.00025334686506539583, -0.11308930814266205, -0.0409194678068161, -0.38762593269348145, -0.2783125638961792, 0.19049637019634247, 0.45270541310310364, 0.22734399139881134, -0.01965280994772911, -0.28994232416152954, 0.31739184260368347, -0.4826754033565521, 0.002716717543080449, -0.6354674696922302, 0.12806348502635956, -0.09318788349628448, -0.14903484284877777, -0.16396395862102509, 0.1639254093170166, 0.023549996316432953, -0.0028481101617217064, -0.09274742752313614, 0.2124699354171753, -0.2380760759115219, 0.08131196349859238, 0.2638709843158722, 0.10236212611198425, 0.03820393979549408, -0.5709121823310852, 0.337582528591156, 0.471760094165802, 0.22842155396938324, 0.08941147476434708, -0.2693725526332855, 0.11830697953701019, -0.0825481116771698, 0.0903005376458168, 0.043728042393922806, 0.01720329001545906, 0.46662530303001404, 0.014810673892498016, 0.06998161971569061, -0.21279384195804596, 0.3003251254558563, 0.21648988127708435, 0.46217772364616394, -0.08612532913684845, -0.5364862680435181, 0.08613129705190659, 0.279600590467453, 0.06051954627037048, -0.09864404052495956, -0.01093630027025938, -0.3510042428970337, -0.04163467511534691, 0.055063143372535706, 0.22897891700267792, 0.28139010071754456, 0.22780916094779968, -0.06535445898771286, 0.08311142027378082, 0.09837327152490616, -0.2621801793575287, 0.1807290017604828, -0.25792714953422546, 0.10123013705015182, 0.26987096667289734, 0.19980137050151825, 0.03205357864499092, -0.03564421087503433, -0.5033636093139648, 0.07216988503932953, 0.44549039006233215, -0.2703445553779602, -0.17282365262508392, -0.1369037926197052, -0.6854809522628784, 0.060650646686553955, 0.07638580352067947, -0.11054092645645142, -0.34322455525398254, -0.02222832664847374, 0.18775984644889832, 0.21456144750118256, -0.07199577987194061, -0.09505292773246765, 0.07554131001234055, 0.06864757835865021, -0.10940901190042496, -0.41933175921440125, -0.3149597942829132, -0.20777536928653717, 0.04685645550489426, 0.27079495787620544, -0.02280697040259838, 0.24694304168224335, -0.2802794277667999, 0.06392866373062134, -0.4144374430179596, -0.025489769876003265, 0.023476963862776756, 0.029296601191163063, -0.0004743135941680521, -0.03406819328665733, 0.48188820481300354, 0.47751712799072266, -0.062323737889528275, 0.34758952260017395, 0.04154730215668678, -0.17423060536384583, 0.2578032612800598, -0.21523937582969666, -0.23448677361011505, -0.04180895909667015, -0.5249137282371521, -0.35636940598487854, -0.3282740116119385, 0.08275364339351654, -0.13988102972507477, 0.0803988054394722, 0.31809043884277344, -0.1464771330356598, 0.2062617689371109, -0.03888341411948204, 0.19586744904518127, -0.07545904070138931, -0.2256142646074295, 0.3382493555545807, -0.16139675676822662, -0.2159951776266098, 0.23609325289726257, 0.06490588188171387, 0.21008798480033875, -0.04124198481440544, -0.23060105741024017, 0.17212596535682678, -0.22557012736797333, 0.28446343541145325, -0.1305036097764969, -0.15192051231861115, 0.34998252987861633, -0.010504299774765968, -0.0014597176341339946, 0.004380505066365004, 0.01188093051314354, 0.11780867725610733, 0.010027376934885979, 0.571620762348175, 0.20394249260425568, 0.41344502568244934, 0.14451955258846283, 0.3723565638065338, 0.44866421818733215, 0.026759330183267593, 0.32839006185531616, -0.26213935017585754, 0.0669078454375267, -0.06588185578584671, -0.3156568706035614, 0.11945436894893646, 0.17475633323192596, -0.17332448065280914, -0.10797219723463058, 0.138683021068573, 0.38894420862197876, -0.33151981234550476, -0.13302171230316162, -0.2874072194099426, -0.08548351377248764, -0.14637787640094757, 0.02025029808282852, 0.011118978261947632, 0.12911097705364227, 0.2093753218650818, -0.34255450963974, -0.3424661159515381, -0.11213251948356628, 0.4633776545524597, 0.0023362934589385986, 0.15231388807296753, -0.2744362950325012, -0.28418034315109253, -0.5618224143981934, 0.48062583804130554, 0.11800606548786163, 0.21206669509410858, -0.2758864164352417, 0.12500028312206268, 0.2663460671901703, 0.022575542330741882, 0.3283258378505707, -0.3004832863807678, 0.25636452436447144, 0.08078701049089432, -0.09893045574426651, -0.23588041961193085, 0.1696392297744751, -0.15016189217567444, 0.1985001564025879, 0.33587154746055603, 0.19395440816879272, -0.160841166973114, -0.11107291281223297, -0.016944129019975662, 0.10838181525468826, -0.1749693900346756, -0.21759386360645294, -0.017783721908926964, -0.0037044950295239687, -0.16341423988342285, -0.328482985496521, -0.1715681403875351, 0.24727433919906616, 0.13296154141426086, 0.33589187264442444, 0.04687206819653511, -0.07427357882261276, 0.17847338318824768, 0.22041575610637665, 0.33928728103637695, -0.09613542258739471, -0.15330880880355835, 0.1652642786502838, 0.2498362809419632, 0.07175866514444351, 0.49867498874664307, 0.28984472155570984, -0.3622654974460602, -0.014100667089223862, -0.22840864956378937, 0.12723597884178162, 0.2903362810611725, -0.25874918699264526, 0.10405367612838745, 0.034366361796855927, -0.00782689731568098, 0.0616951584815979, -0.09590546786785126, 0.25131773948669434, 0.03154374659061432, -0.1259423941373825, -0.3986915946006775, 0.32729730010032654, 0.10116904973983765, 0.003972935024648905, 0.4540020525455475, -0.14631907641887665, -0.22761699557304382, 0.023734675720334053, 0.05723154917359352, 0.9369785785675049, 0.12992772459983826, 0.22923718392848969, 0.3930838406085968, 0.16831186413764954, 0.5260001420974731, -0.30351364612579346, 0.36880162358283997, -0.38715463876724243, -0.08039484173059464, -0.0029480515513569117, -0.11370399594306946, -0.011664546094834805, -0.1774902492761612, -0.08031443506479263, 0.23407699167728424, -0.15633440017700195, 0.00611140625551343, -0.09859159588813782, 0.2628976106643677, -0.3349466025829315, -0.09046969562768936, -0.46061474084854126, 0.0847524031996727, -0.20800389349460602, 0.3349795937538147, -0.10503844916820526, -0.1498464047908783, -0.38530439138412476, 0.1126159206032753, -0.42020905017852783, 0.37589675188064575, -0.0859742984175682, 0.12252394109964371, -0.07361681759357452, -0.3276995122432709, -0.05533306673169136, 0.6137094497680664, 0.5807364583015442, 0.0501268096268177, -0.33917272090911865, 0.007756409700959921, -0.21130119264125824, -0.023090843111276627, -0.09389214962720871, 0.15532875061035156, 0.33820468187332153, -0.04085510969161987, -0.30917733907699585, 0.5142865777015686, -0.11258584260940552, 0.016242269426584244, -0.2366236448287964, -0.25668174028396606, -0.24029716849327087, -0.043224040418863297, -0.1230141818523407, 0.01992630772292614, -0.12974593043327332, 0.025179490447044373, 0.09974995255470276, 0.012665953487157822, -0.23290403187274933, -0.2464853972196579, -0.1339302361011505, -0.1444869339466095, 0.18300442397594452, 0.325381875038147, 0.047445446252822876, 0.1761949509382248, 0.6140017509460449, 0.22167590260505676, -0.0501692108809948, -0.24814759194850922, 0.06540445238351822, 0.15567706525325775, -0.2556737959384918, -0.21238085627555847, 0.10414279252290726, 0.17469249665737152, 0.096544049680233, 0.2774945795536041, 0.18646995723247528, -0.006257296539843082, -0.14283394813537598, -0.448702335357666, -0.5422465205192566, 0.1517493724822998, -0.19254674017429352, 0.13901548087596893, -0.1516946256160736, 0.2596282958984375, 0.11874712258577347, -0.2652503252029419, -0.3370278477668762, -0.09570170193910599, -0.32341060042381287, -0.005745049100369215, 0.3934481143951416, 0.05905287712812424, 0.19687820971012115, 0.04367268458008766, 0.13710691034793854, -0.036700136959552765, -0.2849011719226837, -0.15982723236083984, 0.048877786844968796, 0.10002417862415314, 0.1371709108352661, -0.4027869701385498, -0.11658159643411636, -0.28826212882995605, -0.08812721073627472, -0.1604430079460144, 0.05028767138719559, 0.29853424429893494, 0.03162437304854393, 0.2532196044921875, -0.31334546208381653, 0.058008331805467606, -0.2136867344379425, 0.1573789119720459, -0.15283744037151337, 0.18410615622997284, 0.05986139923334122, 0.12613047659397125, 0.01767575554549694, 0.009883997030556202, -0.550724446773529, -0.06257440149784088, -0.0907987728714943, 0.1829756200313568, 0.11525699496269226, -0.4182232618331909, 0.22805631160736084, -0.14199501276016235, 0.1714201122522354, 0.19626367092132568, -0.03331765905022621, 0.2832377552986145, 0.18114230036735535, 0.18628519773483276, -0.23866654932498932, 0.20860911905765533, 0.215780109167099, 0.09463916718959808, -0.09388977289199829, 0.04307954013347626, -0.049454621970653534, -0.02566336840391159, 0.2692066729068756, -0.1024499237537384, 0.3867937922477722, -0.4072207510471344, -0.052489642053842545, 0.11946655809879303, 0.07139203697443008, -0.05252387002110481, 0.11530407518148422, 0.3662637770175934, 0.16459278762340546, 0.7827573418617249, -0.24703241884708405, 0.14657171070575714, -0.18498532474040985, 0.3281104862689972, -0.24878236651420593, -0.3345986008644104, -0.33597898483276367, 0.11348120868206024, -0.294269859790802, -0.12617403268814087, -0.18528421223163605, 0.2995415925979614, -0.10181374847888947, 0.14273475110530853, -0.23577985167503357, 0.39215415716171265, -0.14883604645729065, -0.04539475217461586, -0.5625836849212646, -0.22663353383541107, 0.10585590451955795, -0.1797053962945938, 0.025090746581554413, 0.01599329523742199, -0.009149348363280296, 0.1309439241886139, -0.2652382552623749, -0.40185388922691345, 0.3093724250793457, 0.40477555990219116, -0.016356229782104492, -0.2793296277523041, 0.41067326068878174, 0.29058921337127686, 0.06595078855752945, 0.20533108711242676, 0.3192925453186035, 0.42760297656059265, 0.4022080600261688, 0.06650746613740921, 0.20000669360160828, -0.10694017261266708, -0.25286200642585754, -0.013881637714803219, 0.05252344533801079, 0.30146437883377075, 0.05845767632126808, 0.31253117322921753, 0.14970730245113373, -0.0062344688922166824, 0.14302103221416473, 0.04569145664572716, 0.3462026119232178, 0.09222348034381866, -0.015020581893622875, -0.14654797315597534, -0.24552859365940094, -0.1894637942314148, -0.1285674124956131, -0.5581473708152771, 0.1378030627965927, 0.21281974017620087, 0.2150430977344513, 0.25495705008506775, -0.004208188503980637, 0.04900941625237465, -0.06657014787197113, 0.13284258544445038, 0.5250786542892456, -0.03401148319244385, -0.27033546566963196, -0.32096952199935913, -0.5026273727416992, 0.16537779569625854, -0.02528504841029644, -0.4208659529685974, -0.24853651225566864, 0.12115411460399628, -0.17552419006824493, -0.2355567365884781, 0.24101929366588593, 0.1014041155576706, 0.06959380954504013, 0.33923885226249695, -0.22689945995807648, -0.33274635672569275, 0.2696407437324524, 0.0531267449259758, -0.17090114951133728, -0.17286787927150726, -0.0946459174156189, -0.14867638051509857, 0.049000054597854614, -0.04603859409689903, 0.3904182016849518, 0.007113414816558361, 0.20573078095912933, 0.19779559969902039, 0.19981218874454498, 0.48119643330574036, 0.06806674599647522, -0.0404394194483757, -0.3731052577495575, -0.11173686385154724, -0.20125903189182281, -0.041570909321308136, 0.3111204504966736, 0.5081033706665039, -0.358381986618042, -0.00834614597260952, -0.38156816363334656, 0.1278298944234848, -0.16659346222877502, -0.01748354732990265, -0.2955944538116455, 0.3703540861606598, -0.2835775315761566, 0.23727823793888092, 0.16417279839515686, -0.04555123671889305, -0.029033057391643524, 0.10304306447505951, -0.33316659927368164, 0.0036254317965358496, 0.5232228636741638, -0.2101983278989792, -0.23654088377952576, 0.09007932245731354, 0.19331854581832886, 0.11350388824939728, 0.01034716609865427, -0.33786872029304504, 0.17566758394241333, 0.18242225050926208, -0.0053519015200436115, -0.24062062799930573, -0.08169142156839371, 0.0652153491973877, 0.11355213075876236, -0.06532713025808334, -0.1427069753408432, 0.15939858555793762, -0.07285217195749283, -0.013610105030238628, -0.08653882890939713 ]
https://github.com/huggingface/datasets/issues/225
[ROUGE] Different scores with `files2rouge`
@Colanim unfortunately there are different implementations of the ROUGE metric floating around online which yield different results, and we had to chose one for the package :) We ended up including the one from the google-research repository, which does minimal post-processing before computing the P/R/F scores. If I recall correctly, files2rouge relies on the Perl, script, which among other things normalizes all numbers to a special token: in the case you presented, this should account for a good chunk of the difference. We may end up adding in more versions of the metric, but probably not for a while (@lhoestq correct me if I'm wrong). However, feel free to take a stab at adding it in yourself and submitting a PR if you're interested!
It seems that the ROUGE score of `nlp` is lower than the one of `files2rouge`. Here is a self-contained notebook to reproduce both scores : https://colab.research.google.com/drive/14EyAXValB6UzKY9x4rs_T3pyL7alpw_F?usp=sharing --- `nlp` : (Only mid F-scores) >rouge1 0.33508031962733364 rouge2 0.14574333776191592 rougeL 0.2321187823256159 `files2rouge` : >Running ROUGE... =========================== 1 ROUGE-1 Average_R: 0.48873 (95%-conf.int. 0.41192 - 0.56339) 1 ROUGE-1 Average_P: 0.29010 (95%-conf.int. 0.23605 - 0.34445) 1 ROUGE-1 Average_F: 0.34761 (95%-conf.int. 0.29479 - 0.39871) =========================== 1 ROUGE-2 Average_R: 0.20280 (95%-conf.int. 0.14969 - 0.26244) 1 ROUGE-2 Average_P: 0.12772 (95%-conf.int. 0.08603 - 0.17752) 1 ROUGE-2 Average_F: 0.14798 (95%-conf.int. 0.10517 - 0.19240) =========================== 1 ROUGE-L Average_R: 0.32960 (95%-conf.int. 0.26501 - 0.39676) 1 ROUGE-L Average_P: 0.19880 (95%-conf.int. 0.15257 - 0.25136) 1 ROUGE-L Average_F: 0.23619 (95%-conf.int. 0.19073 - 0.28663) --- When using longer predictions/gold, the difference is bigger. **How can I reproduce same score as `files2rouge` ?** @lhoestq
124
[ROUGE] Different scores with `files2rouge` It seems that the ROUGE score of `nlp` is lower than the one of `files2rouge`. Here is a self-contained notebook to reproduce both scores : https://colab.research.google.com/drive/14EyAXValB6UzKY9x4rs_T3pyL7alpw_F?usp=sharing --- `nlp` : (Only mid F-scores) >rouge1 0.33508031962733364 rouge2 0.14574333776191592 rougeL 0.2321187823256159 `files2rouge` : >Running ROUGE... =========================== 1 ROUGE-1 Average_R: 0.48873 (95%-conf.int. 0.41192 - 0.56339) 1 ROUGE-1 Average_P: 0.29010 (95%-conf.int. 0.23605 - 0.34445) 1 ROUGE-1 Average_F: 0.34761 (95%-conf.int. 0.29479 - 0.39871) =========================== 1 ROUGE-2 Average_R: 0.20280 (95%-conf.int. 0.14969 - 0.26244) 1 ROUGE-2 Average_P: 0.12772 (95%-conf.int. 0.08603 - 0.17752) 1 ROUGE-2 Average_F: 0.14798 (95%-conf.int. 0.10517 - 0.19240) =========================== 1 ROUGE-L Average_R: 0.32960 (95%-conf.int. 0.26501 - 0.39676) 1 ROUGE-L Average_P: 0.19880 (95%-conf.int. 0.15257 - 0.25136) 1 ROUGE-L Average_F: 0.23619 (95%-conf.int. 0.19073 - 0.28663) --- When using longer predictions/gold, the difference is bigger. **How can I reproduce same score as `files2rouge` ?** @lhoestq @Colanim unfortunately there are different implementations of the ROUGE metric floating around online which yield different results, and we had to chose one for the package :) We ended up including the one from the google-research repository, which does minimal post-processing before computing the P/R/F scores. If I recall correctly, files2rouge relies on the Perl, script, which among other things normalizes all numbers to a special token: in the case you presented, this should account for a good chunk of the difference. We may end up adding in more versions of the metric, but probably not for a while (@lhoestq correct me if I'm wrong). However, feel free to take a stab at adding it in yourself and submitting a PR if you're interested!
[ -0.14572615921497345, -0.16974303126335144, -0.1972297579050064, 0.3397459387779236, -0.11949753761291504, -0.4242601990699768, -0.4203995168209076, 0.1476738452911377, -0.2700440585613251, 0.17250607907772064, -0.08286702632904053, 0.1803876757621765, 0.12033016234636307, 0.008093580603599548, -0.1003899946808815, -0.09797486662864685, 0.165655255317688, 0.03882870450615883, 0.09248090535402298, -0.3720676898956299, 0.06655961275100708, 0.5696011781692505, -0.03133227676153183, 0.2104341685771942, 0.05402207747101784, 0.2702381908893585, 0.28511351346969604, 0.3035299479961395, -0.14858005940914154, 0.028436783701181412, 0.12800203263759613, 0.03064163215458393, -0.02369801700115204, 0.22214896976947784, -0.00010886792006203905, -0.2347191721200943, 0.10260970890522003, -0.0031664110720157623, -0.27572351694107056, -0.30260902643203735, -0.034557461738586426, 0.06436412036418915, -0.1229282021522522, -0.07039352506399155, 0.08584654331207275, -0.04489225521683693, 0.04877280816435814, 0.20954246819019318, 0.4293985068798065, 0.1475730836391449, 0.22625432908535004, -0.30243781208992004, -0.18534952402114868, 0.024081915616989136, 0.5763341188430786, 0.16267599165439606, -0.11448317766189575, 0.44726425409317017, 0.005699495319277048, -0.3526380658149719, -0.24019040167331696, -0.008563567884266376, 0.34304094314575195, -0.17255988717079163, 0.13255919516086578, 0.19315466284751892, 0.14858555793762207, -0.04594054073095322, -0.07953529059886932, 0.24472841620445251, -0.2702822983264923, 0.03221165016293526, -0.22717157006263733, -0.2778386175632477, -0.07631559669971466, -0.1997607946395874, 0.11689180135726929, -0.08144606649875641, 0.1642860472202301, 0.012247875332832336, 0.04413227364420891, 0.29542678594589233, 0.04963891953229904, 0.007164150010794401, -0.05405605211853981, -0.10784615576267242, 0.05084570124745369, 0.09846562147140503, 0.29527974128723145, -0.050543639808893204, -0.5163149237632751, 0.01077284850180149, -0.2593677043914795, 0.0749606341123581, -0.30228114128112793, -0.03785572201013565, 0.3144964575767517, 0.4659734070301056, -0.20087116956710815, 0.4996274709701538, 0.26919713616371155, -0.0934174507856369, -0.13220743834972382, -0.03944722190499306, -0.13782373070716858, 0.10813210904598236, -0.028945928439497948, 0.020693473517894745, 0.08811821043491364, 0.2246810793876648, -0.13358263671398163, -0.012008018791675568, 0.39822256565093994, -0.5637689232826233, -0.4777999222278595, 0.10259132832288742, -0.26103466749191284, -0.4323975145816803, -0.6891934871673584, -0.0673605352640152, -0.255431205034256, -0.11924232542514801, 0.09890846163034439, 0.22720065712928772, -0.17488402128219604, 0.29229849576950073, 0.046788230538368225, 0.18339158594608307, -0.28950363397598267, 0.01966216042637825, -0.18399901688098907, 0.06357946246862411, -0.31439271569252014, 0.13452430069446564, 0.17056682705879211, 0.11274684220552444, 0.31805747747421265, 0.25605344772338867, 0.27351316809654236, 0.03864440321922302, 0.3298662304878235, -0.2839505672454834, -0.022634783759713173, -0.21111683547496796, -0.2113915979862213, 0.1974538117647171, 0.027129076421260834, 0.09184830635786057, -0.00870169885456562, -0.07489427924156189, -0.18084076046943665, 0.1442372053861618, 0.46661806106567383, 0.20657984912395477, 0.14965085685253143, -0.11055408418178558, 0.04069235175848007, 0.11693812906742096, -0.24016346037387848, -0.5936912298202515, 0.20406076312065125, -0.1365523636341095, -0.1270110011100769, -0.05817752704024315, 0.10981972515583038, -0.03458227589726448, 0.09312087297439575, -0.11743129789829254, 0.3731990456581116, 0.06358262896537781, 0.26031258702278137, 0.45348334312438965, -0.0408313013613224, 0.011863946914672852, -0.1628563404083252, -0.07088572531938553, 0.4665316641330719, -0.43853142857551575, -0.22229626774787903, 0.3355002701282501, -0.16507026553153992, -0.06518586724996567, 0.23904691636562347, 0.03927561268210411, -0.3886531591415405, -0.16600807011127472, 0.17745089530944824, 0.19752459228038788, 0.07779872417449951, -0.048396892845630646, -0.2970068156719208, 0.13495878875255585, 0.316625714302063, 0.06547749787569046, -0.23182860016822815, -0.29935044050216675, -0.11140508949756622, -0.06439077854156494, 0.5141965746879578, 0.13280688226222992, -0.27513760328292847, 0.10499175637960434, -0.2721599340438843, 0.29369354248046875, -0.05169543996453285, 0.11857372522354126, 0.4230367839336395, -0.023375488817691803, -0.7951249480247498, -0.16027650237083435, 0.4370328485965729, -0.05248735845088959, -0.21229121088981628, -0.4714091122150421, 0.24053311347961426, -0.23946474492549896, 0.22328263521194458, 0.05504881590604782, 0.1873215287923813, 0.030354246497154236, 0.02407083846628666, 0.09660714119672775, -0.23642665147781372, -0.15861894190311432, -0.17153561115264893, 0.20367735624313354, -0.22109828889369965, -0.007432829588651657, -0.01054032240062952, 0.35960862040519714, 0.31164121627807617, 0.0875801146030426, 0.03730439767241478, 0.2634182274341583, 0.05015571042895317, 0.09804999828338623, 0.3672510087490082, 0.14480578899383545, -0.0007639330578967929, 0.18896162509918213, -0.1226600706577301, 0.6879815459251404, 0.01646246388554573, 0.03225119411945343, -0.3400397300720215, 0.42591556906700134, -0.21703946590423584, 0.20069162547588348, 0.4093106985092163, -0.07173748314380646, -0.05009761080145836, 0.057780198752880096, -0.006800386123359203, -0.01715300790965557, 0.7467252016067505, 0.1952691376209259, 0.2502553164958954, 0.0029566255398094654, -0.32202571630477905, 0.25961771607398987, 0.39914047718048096, -0.015342896804213524, -0.02085353247821331, -0.11995424330234528, 0.19371458888053894, -0.09338841587305069, -0.036720160394907, 0.2507389783859253, 0.3810695707798004, 0.19513395428657532, 0.1557074785232544, -0.13119934499263763, -0.3382363021373749, -0.34380224347114563, 0.03513277322053909, -0.18210364878177643, 0.08744674175977707, 0.49638640880584717, 0.10160709917545319, 0.21628661453723907, -0.3020774722099304, -0.5422855019569397, -0.25364869832992554, -0.22231446206569672, 0.06132601946592331, 0.16164496541023254, -0.14391355216503143, -0.3678353428840637, -0.15366077423095703, -0.091903455555439, -0.2918891906738281, -0.27327603101730347, 0.23009741306304932, 0.020875325426459312, -0.07249978184700012, -0.21893787384033203, 0.023232346400618553, 0.14874784648418427, -0.3703943192958832, 0.1783805787563324, -0.017374185845255852, -0.0280428696423769, -0.27953794598579407, 0.2228175401687622, -0.05679844319820404, -0.019304806366562843, -0.03147659823298454, -0.27918699383735657, -0.450835645198822, 0.16968795657157898, -0.37626758217811584, -0.010960331186652184, 0.06464394927024841, -0.0032369038090109825, 0.05967245250940323, -0.25061726570129395, -0.07057500630617142, 0.14640170335769653, 0.09220732748508453, -0.17413024604320526, -0.3507431149482727, -0.08965101093053818, 0.0367850661277771, 0.014213897287845612, -0.19140632450580597, -0.2775213420391083, 0.07937490195035934, -0.1433197259902954, 0.6197946667671204, 0.3270145356655121, 0.005700475536286831, 0.38587889075279236, -0.2369561791419983, 0.00001117925057769753, 0.1717422902584076, 0.5837682485580444, -0.260583758354187, -0.04442865401506424, 0.1986089050769806, -0.29971224069595337, -0.4421885013580322, -0.27306097745895386, 0.18830086290836334, 0.03232232853770256, -0.008429203182458878, -0.1786167025566101, 0.031004473567008972, -0.13094371557235718, -0.04899758845567703, -0.21304526925086975, -0.05581812188029289, 0.3378271460533142, -0.18365980684757233, -0.2421921342611313, -0.0644744485616684, 0.21673358976840973, 0.02994077280163765, 0.286546915769577, 0.4599354863166809, -0.4840276837348938, 0.16461147367954254, -0.14009366929531097, 0.12755979597568512, 0.0440831296145916, 0.5117963552474976, 0.15338170528411865, 0.13323144614696503, 0.0008751187124289572, -0.14067673683166504, 0.09818065166473389, 0.4896935820579529, 0.05618814751505852, -0.14490316808223724, 0.3308490514755249, 0.20873454213142395, 0.24980413913726807, -0.11164992302656174, -0.04191485047340393, -0.07632182538509369, -0.007912087254226208, 0.10350774228572845, 0.07728636264801025, -0.0560007318854332, 0.017136143520474434, 0.07414616644382477, -0.0036134684924036264, -0.15379202365875244, 0.21246269345283508, 0.07286833226680756, 0.3913024961948395, -0.11124511808156967, -0.2594476044178009, -0.005773492157459259, -0.4689764380455017, -0.005583834368735552, 0.03185344487428665, -0.021543025970458984, -0.15841111540794373, -0.17287009954452515, 0.05907157063484192, -0.3485897481441498, 0.26234281063079834, 0.16069422662258148, 0.23712962865829468, 0.16907602548599243, -0.13415367901325226, -0.5272517800331116, -0.2941480278968811, -0.3322207033634186, 0.18113477528095245, 0.5960766077041626, 0.1151440367102623, -0.13822154700756073, -0.1728234589099884, 0.10999204963445663, -0.05185002088546753, -0.148247629404068, 0.10066864639520645, -0.33127129077911377, 0.20241938531398773, 0.15132899582386017, -0.2806151509284973, 0.04583752155303955, 0.3753645420074463, -0.2438097596168518, 0.12480254471302032, 0.06615769863128662, -0.14999350905418396, 0.21233011782169342, 0.13495221734046936, 0.33930888772010803, -0.10402209311723709, -0.2439710795879364, -0.29319724440574646, 0.30135512351989746, -0.351443886756897, 0.3701418936252594, 0.12124495208263397, 0.00033597677247598767, 0.30029138922691345, -0.09358716756105423, 0.10684158653020859, 0.31012898683547974, -0.1816413700580597, 0.1664065271615982, -0.6221524477005005, 0.11270283907651901, -0.4461793601512909, -0.21020959317684174, 0.3684735894203186, -0.07535848766565323, 0.28507164120674133, -0.1410282403230667, 0.20845304429531097, 0.23064105212688446, -0.4770975112915039, -0.07949839532375336, -0.16209371387958527, -0.19412653148174286, 0.2774370312690735, -0.020417680963873863, 0.7596281170845032, 0.43547335267066956, 0.5633998513221741, 0.2302163988351822, -0.14735105633735657, 0.38939759135246277, -0.5955860018730164, 0.19127987325191498, -0.3415153920650482, -0.37023451924324036, 0.09574269503355026, 0.10443361103534698, -0.06511300802230835, 0.2145492285490036, 0.2076512575149536, 0.29537633061408997, 0.3931225538253784, -0.0991114005446434, -0.10828062146902084, 0.17655280232429504, 0.12343021482229233, 0.18754109740257263, -0.2870902717113495, 0.19913658499717712, 0.34644386172294617, 0.004576204810291529, -0.05969381704926491, 0.09596779942512512, -0.7420679330825806, 0.0225160401314497, 0.04249162599444389, 0.10044686496257782, -0.09559981524944305, 0.35474157333374023, 0.11958033591508865, -0.22432821989059448, -0.048718634992837906, 0.3672381341457367, 0.15987472236156464, -0.01591307297348976, 0.007195963989943266, 0.21002458035945892, -0.3194086253643036, 0.3263314664363861, 0.24231000244617462, -0.038803331553936005, 0.031407807022333145, -0.22021394968032837, -0.07871294766664505, -0.07076095044612885, 0.15572413802146912, -0.16848191618919373, -0.548475444316864, 0.18215550482273102, -0.33470356464385986, 0.0072069619782269, -0.02991681545972824, 0.10068981349468231, -0.09960504621267319, -0.05343042314052582, 0.16725927591323853, -0.05686292424798012, -0.06265318393707275, -0.03488534316420555, 0.09307224303483963, -0.3328576385974884, -0.05046382173895836, 0.4744253158569336, -0.1367390751838684, -0.046009812504053116, 0.06487401574850082, -0.330565869808197, 0.03526663780212402, -0.2431761920452118, 0.03892473876476288, -0.08075796067714691, -0.19715382158756256, -0.10430055856704712, -0.007580338045954704, -0.2291494905948639, 0.25272709131240845, 0.25843632221221924, 0.2495664805173874, 0.21303711831569672, 0.032852903008461, -0.3630848824977875, -0.09832476824522018, 0.08107944577932358, 0.1679994910955429, -0.034733790904283524, -0.01961880549788475, -0.15958917140960693, 0.25884079933166504, 0.2306310385465622, -0.35463520884513855, -0.15017235279083252, -0.08997538685798645, 0.12681719660758972, -0.2210235744714737, -0.31873929500579834, 0.02248654142022133, 0.1468537151813507, 0.1430858075618744, -0.008150210604071617, -0.05782909318804741, -0.2477169781923294, -0.20516584813594818, 0.08322594314813614, -0.04834677651524544, -0.09744031727313995, -0.13200196623802185, 0.21839942038059235, 0.10436245799064636, -0.14063531160354614, -0.2313222587108612, -0.13392621278762817, 0.35739850997924805, 0.12174567580223083, -0.08553913235664368, 0.1825161725282669, -0.14262963831424713, 0.10866586863994598, 0.018823904916644096, -0.025384819135069847, 0.07357653230428696, -0.026512816548347473, 0.109457828104496, -0.2018422782421112, -0.25641098618507385, -0.07424065470695496, -0.2803899943828583, 0.14717309176921844, 0.10414627939462662, -0.0529727004468441, 0.008782899007201195, -0.18328943848609924, 0.2844220995903015, 0.36497530341148376, 0.11780496686697006, -0.09963652491569519, -0.013917454518377781, 0.22528259456157684, -0.22220176458358765, 0.09425390511751175, 0.44961801171302795, 0.043986886739730835, 0.06563520431518555, -0.02828390896320343, 0.15857845544815063, -0.20769856870174408, -0.018284499645233154, -0.05924854055047035, -0.27442365884780884, -0.018169712275266647, 0.17280763387680054, -0.19119587540626526, 0.25152382254600525, 0.0736309215426445, 0.1553812474012375, -0.37914833426475525, 0.18718206882476807, 0.3507705330848694, -0.0025792052038013935, 0.19415226578712463, -0.2856432795524597, 0.14951284229755402, -0.02367577515542507, 0.09032062441110611, -0.27035999298095703, 0.3451123535633087, 0.0853993147611618, -0.02180611714720726, 0.08363576978445053, 0.16912171244621277, -0.10833396762609482, -0.33761361241340637, -0.0048829554580152035, -0.0880928635597229, -0.07047468423843384, -0.5443993806838989, -0.46642449498176575, -0.22646990418434143, -0.3945833742618561, 0.07105030119419098, -0.16874152421951294, 0.38178619742393494, 0.02951897494494915, 0.2477216273546219, -0.09672477096319199, -0.1826779842376709, -0.46176472306251526, 0.010245916433632374, 0.42088058590888977, -0.046756893396377563, 0.17743198573589325, 0.21671564877033234, 0.41431307792663574, 0.3571798503398895, 0.16161072254180908, 0.23874562978744507, 0.18882715702056885, -0.2336151897907257, -0.2279312014579773, 0.03246549516916275, 0.08932047337293625, -0.19615767896175385, 0.17421679198741913, 0.2599768042564392, 0.04339063540101051, 0.3002179265022278, 0.132206529378891, -0.2410813421010971, 0.07742050290107727, 0.2649875283241272, -0.0035750248935073614, -0.21412977576255798, 0.29473695158958435, -0.03916145861148834, 0.018682796508073807, -0.33147430419921875, 0.3797481954097748, -0.5077443718910217, -0.08658923208713531, -0.5439392328262329, 0.15140385925769806, 0.17864570021629333, -0.12435408681631088, 0.10393213480710983, 0.26889467239379883, 0.41731399297714233, -0.127236008644104, 0.3387746214866638, -0.28413739800453186, -0.46132057905197144, -0.26536378264427185, -0.04585053399205208, 0.039508700370788574, 0.27052944898605347, -0.43093833327293396, 0.28873541951179504, -0.29741501808166504, 0.051450327038764954, 0.2745601534843445, 0.42833009362220764, 0.24691952764987946, 0.06649734079837799, -0.33822908997535706, 0.1559215635061264, 0.19840335845947266, -0.08824482560157776, -0.11956939101219177, 0.13063283264636993, -0.17390663921833038, 0.019816448912024498, 0.1097150519490242, 0.08555615693330765, -0.3465365171432495, -0.2886541783809662, 0.4207555949687958, 0.33852237462997437, 0.27506640553474426, 0.26808562874794006, -0.37258148193359375, -0.38539496064186096, 0.03100375458598137, -0.12375269085168839, -0.0627094954252243, 0.062282588332891464, -0.06584226340055466, 0.3876439929008484, 0.025063853710889816, 0.07477404177188873, 0.07386086881160736, 0.5041282176971436, 0.05595526471734047, 0.0017199635040014982, -0.22703063488006592, -0.026206299662590027, 0.09364594519138336, 0.06342308223247528, -0.6129956841468811, 0.40744340419769287, 0.0005166334449313581, 0.18714416027069092, -0.2337079793214798, -0.2538995146751404, 0.37504610419273376, -0.3584226965904236, 0.27204254269599915, -0.09970097243785858, 0.34275397658348083, 0.2217620611190796, 0.29680484533309937, 0.04679165408015251, -0.10589522123336792, 0.0717540830373764, 0.29574862122535706, 0.162244513630867, 0.23034587502479553, -0.2334207147359848, -0.08380375802516937, -0.0016766850603744388, 0.29279467463493347, 0.0649266391992569, -0.29197120666503906, 0.4134342670440674, -0.184107705950737 ]
https://github.com/huggingface/datasets/issues/225
[ROUGE] Different scores with `files2rouge`
Thank you for your kind answer. As a side question : Isn't it better to have a package that normalize more ? I understand to idea of having a package that does minimal post-processing for transparency. But it means that people using different architecture (with different tokenizers for example) will have difference in ROUGE scores even if their predictions are actually similar. The goal of `nlp` is to have _one package to rule them all_, right ? I will look into it but I'm not sure I have the required skill for this ^^
It seems that the ROUGE score of `nlp` is lower than the one of `files2rouge`. Here is a self-contained notebook to reproduce both scores : https://colab.research.google.com/drive/14EyAXValB6UzKY9x4rs_T3pyL7alpw_F?usp=sharing --- `nlp` : (Only mid F-scores) >rouge1 0.33508031962733364 rouge2 0.14574333776191592 rougeL 0.2321187823256159 `files2rouge` : >Running ROUGE... =========================== 1 ROUGE-1 Average_R: 0.48873 (95%-conf.int. 0.41192 - 0.56339) 1 ROUGE-1 Average_P: 0.29010 (95%-conf.int. 0.23605 - 0.34445) 1 ROUGE-1 Average_F: 0.34761 (95%-conf.int. 0.29479 - 0.39871) =========================== 1 ROUGE-2 Average_R: 0.20280 (95%-conf.int. 0.14969 - 0.26244) 1 ROUGE-2 Average_P: 0.12772 (95%-conf.int. 0.08603 - 0.17752) 1 ROUGE-2 Average_F: 0.14798 (95%-conf.int. 0.10517 - 0.19240) =========================== 1 ROUGE-L Average_R: 0.32960 (95%-conf.int. 0.26501 - 0.39676) 1 ROUGE-L Average_P: 0.19880 (95%-conf.int. 0.15257 - 0.25136) 1 ROUGE-L Average_F: 0.23619 (95%-conf.int. 0.19073 - 0.28663) --- When using longer predictions/gold, the difference is bigger. **How can I reproduce same score as `files2rouge` ?** @lhoestq
94
[ROUGE] Different scores with `files2rouge` It seems that the ROUGE score of `nlp` is lower than the one of `files2rouge`. Here is a self-contained notebook to reproduce both scores : https://colab.research.google.com/drive/14EyAXValB6UzKY9x4rs_T3pyL7alpw_F?usp=sharing --- `nlp` : (Only mid F-scores) >rouge1 0.33508031962733364 rouge2 0.14574333776191592 rougeL 0.2321187823256159 `files2rouge` : >Running ROUGE... =========================== 1 ROUGE-1 Average_R: 0.48873 (95%-conf.int. 0.41192 - 0.56339) 1 ROUGE-1 Average_P: 0.29010 (95%-conf.int. 0.23605 - 0.34445) 1 ROUGE-1 Average_F: 0.34761 (95%-conf.int. 0.29479 - 0.39871) =========================== 1 ROUGE-2 Average_R: 0.20280 (95%-conf.int. 0.14969 - 0.26244) 1 ROUGE-2 Average_P: 0.12772 (95%-conf.int. 0.08603 - 0.17752) 1 ROUGE-2 Average_F: 0.14798 (95%-conf.int. 0.10517 - 0.19240) =========================== 1 ROUGE-L Average_R: 0.32960 (95%-conf.int. 0.26501 - 0.39676) 1 ROUGE-L Average_P: 0.19880 (95%-conf.int. 0.15257 - 0.25136) 1 ROUGE-L Average_F: 0.23619 (95%-conf.int. 0.19073 - 0.28663) --- When using longer predictions/gold, the difference is bigger. **How can I reproduce same score as `files2rouge` ?** @lhoestq Thank you for your kind answer. As a side question : Isn't it better to have a package that normalize more ? I understand to idea of having a package that does minimal post-processing for transparency. But it means that people using different architecture (with different tokenizers for example) will have difference in ROUGE scores even if their predictions are actually similar. The goal of `nlp` is to have _one package to rule them all_, right ? I will look into it but I'm not sure I have the required skill for this ^^
[ -0.14572615921497345, -0.16974303126335144, -0.1972297579050064, 0.3397459387779236, -0.11949753761291504, -0.4242601990699768, -0.4203995168209076, 0.1476738452911377, -0.2700440585613251, 0.17250607907772064, -0.08286702632904053, 0.1803876757621765, 0.12033016234636307, 0.008093580603599548, -0.1003899946808815, -0.09797486662864685, 0.165655255317688, 0.03882870450615883, 0.09248090535402298, -0.3720676898956299, 0.06655961275100708, 0.5696011781692505, -0.03133227676153183, 0.2104341685771942, 0.05402207747101784, 0.2702381908893585, 0.28511351346969604, 0.3035299479961395, -0.14858005940914154, 0.028436783701181412, 0.12800203263759613, 0.03064163215458393, -0.02369801700115204, 0.22214896976947784, -0.00010886792006203905, -0.2347191721200943, 0.10260970890522003, -0.0031664110720157623, -0.27572351694107056, -0.30260902643203735, -0.034557461738586426, 0.06436412036418915, -0.1229282021522522, -0.07039352506399155, 0.08584654331207275, -0.04489225521683693, 0.04877280816435814, 0.20954246819019318, 0.4293985068798065, 0.1475730836391449, 0.22625432908535004, -0.30243781208992004, -0.18534952402114868, 0.024081915616989136, 0.5763341188430786, 0.16267599165439606, -0.11448317766189575, 0.44726425409317017, 0.005699495319277048, -0.3526380658149719, -0.24019040167331696, -0.008563567884266376, 0.34304094314575195, -0.17255988717079163, 0.13255919516086578, 0.19315466284751892, 0.14858555793762207, -0.04594054073095322, -0.07953529059886932, 0.24472841620445251, -0.2702822983264923, 0.03221165016293526, -0.22717157006263733, -0.2778386175632477, -0.07631559669971466, -0.1997607946395874, 0.11689180135726929, -0.08144606649875641, 0.1642860472202301, 0.012247875332832336, 0.04413227364420891, 0.29542678594589233, 0.04963891953229904, 0.007164150010794401, -0.05405605211853981, -0.10784615576267242, 0.05084570124745369, 0.09846562147140503, 0.29527974128723145, -0.050543639808893204, -0.5163149237632751, 0.01077284850180149, -0.2593677043914795, 0.0749606341123581, -0.30228114128112793, -0.03785572201013565, 0.3144964575767517, 0.4659734070301056, -0.20087116956710815, 0.4996274709701538, 0.26919713616371155, -0.0934174507856369, -0.13220743834972382, -0.03944722190499306, -0.13782373070716858, 0.10813210904598236, -0.028945928439497948, 0.020693473517894745, 0.08811821043491364, 0.2246810793876648, -0.13358263671398163, -0.012008018791675568, 0.39822256565093994, -0.5637689232826233, -0.4777999222278595, 0.10259132832288742, -0.26103466749191284, -0.4323975145816803, -0.6891934871673584, -0.0673605352640152, -0.255431205034256, -0.11924232542514801, 0.09890846163034439, 0.22720065712928772, -0.17488402128219604, 0.29229849576950073, 0.046788230538368225, 0.18339158594608307, -0.28950363397598267, 0.01966216042637825, -0.18399901688098907, 0.06357946246862411, -0.31439271569252014, 0.13452430069446564, 0.17056682705879211, 0.11274684220552444, 0.31805747747421265, 0.25605344772338867, 0.27351316809654236, 0.03864440321922302, 0.3298662304878235, -0.2839505672454834, -0.022634783759713173, -0.21111683547496796, -0.2113915979862213, 0.1974538117647171, 0.027129076421260834, 0.09184830635786057, -0.00870169885456562, -0.07489427924156189, -0.18084076046943665, 0.1442372053861618, 0.46661806106567383, 0.20657984912395477, 0.14965085685253143, -0.11055408418178558, 0.04069235175848007, 0.11693812906742096, -0.24016346037387848, -0.5936912298202515, 0.20406076312065125, -0.1365523636341095, -0.1270110011100769, -0.05817752704024315, 0.10981972515583038, -0.03458227589726448, 0.09312087297439575, -0.11743129789829254, 0.3731990456581116, 0.06358262896537781, 0.26031258702278137, 0.45348334312438965, -0.0408313013613224, 0.011863946914672852, -0.1628563404083252, -0.07088572531938553, 0.4665316641330719, -0.43853142857551575, -0.22229626774787903, 0.3355002701282501, -0.16507026553153992, -0.06518586724996567, 0.23904691636562347, 0.03927561268210411, -0.3886531591415405, -0.16600807011127472, 0.17745089530944824, 0.19752459228038788, 0.07779872417449951, -0.048396892845630646, -0.2970068156719208, 0.13495878875255585, 0.316625714302063, 0.06547749787569046, -0.23182860016822815, -0.29935044050216675, -0.11140508949756622, -0.06439077854156494, 0.5141965746879578, 0.13280688226222992, -0.27513760328292847, 0.10499175637960434, -0.2721599340438843, 0.29369354248046875, -0.05169543996453285, 0.11857372522354126, 0.4230367839336395, -0.023375488817691803, -0.7951249480247498, -0.16027650237083435, 0.4370328485965729, -0.05248735845088959, -0.21229121088981628, -0.4714091122150421, 0.24053311347961426, -0.23946474492549896, 0.22328263521194458, 0.05504881590604782, 0.1873215287923813, 0.030354246497154236, 0.02407083846628666, 0.09660714119672775, -0.23642665147781372, -0.15861894190311432, -0.17153561115264893, 0.20367735624313354, -0.22109828889369965, -0.007432829588651657, -0.01054032240062952, 0.35960862040519714, 0.31164121627807617, 0.0875801146030426, 0.03730439767241478, 0.2634182274341583, 0.05015571042895317, 0.09804999828338623, 0.3672510087490082, 0.14480578899383545, -0.0007639330578967929, 0.18896162509918213, -0.1226600706577301, 0.6879815459251404, 0.01646246388554573, 0.03225119411945343, -0.3400397300720215, 0.42591556906700134, -0.21703946590423584, 0.20069162547588348, 0.4093106985092163, -0.07173748314380646, -0.05009761080145836, 0.057780198752880096, -0.006800386123359203, -0.01715300790965557, 0.7467252016067505, 0.1952691376209259, 0.2502553164958954, 0.0029566255398094654, -0.32202571630477905, 0.25961771607398987, 0.39914047718048096, -0.015342896804213524, -0.02085353247821331, -0.11995424330234528, 0.19371458888053894, -0.09338841587305069, -0.036720160394907, 0.2507389783859253, 0.3810695707798004, 0.19513395428657532, 0.1557074785232544, -0.13119934499263763, -0.3382363021373749, -0.34380224347114563, 0.03513277322053909, -0.18210364878177643, 0.08744674175977707, 0.49638640880584717, 0.10160709917545319, 0.21628661453723907, -0.3020774722099304, -0.5422855019569397, -0.25364869832992554, -0.22231446206569672, 0.06132601946592331, 0.16164496541023254, -0.14391355216503143, -0.3678353428840637, -0.15366077423095703, -0.091903455555439, -0.2918891906738281, -0.27327603101730347, 0.23009741306304932, 0.020875325426459312, -0.07249978184700012, -0.21893787384033203, 0.023232346400618553, 0.14874784648418427, -0.3703943192958832, 0.1783805787563324, -0.017374185845255852, -0.0280428696423769, -0.27953794598579407, 0.2228175401687622, -0.05679844319820404, -0.019304806366562843, -0.03147659823298454, -0.27918699383735657, -0.450835645198822, 0.16968795657157898, -0.37626758217811584, -0.010960331186652184, 0.06464394927024841, -0.0032369038090109825, 0.05967245250940323, -0.25061726570129395, -0.07057500630617142, 0.14640170335769653, 0.09220732748508453, -0.17413024604320526, -0.3507431149482727, -0.08965101093053818, 0.0367850661277771, 0.014213897287845612, -0.19140632450580597, -0.2775213420391083, 0.07937490195035934, -0.1433197259902954, 0.6197946667671204, 0.3270145356655121, 0.005700475536286831, 0.38587889075279236, -0.2369561791419983, 0.00001117925057769753, 0.1717422902584076, 0.5837682485580444, -0.260583758354187, -0.04442865401506424, 0.1986089050769806, -0.29971224069595337, -0.4421885013580322, -0.27306097745895386, 0.18830086290836334, 0.03232232853770256, -0.008429203182458878, -0.1786167025566101, 0.031004473567008972, -0.13094371557235718, -0.04899758845567703, -0.21304526925086975, -0.05581812188029289, 0.3378271460533142, -0.18365980684757233, -0.2421921342611313, -0.0644744485616684, 0.21673358976840973, 0.02994077280163765, 0.286546915769577, 0.4599354863166809, -0.4840276837348938, 0.16461147367954254, -0.14009366929531097, 0.12755979597568512, 0.0440831296145916, 0.5117963552474976, 0.15338170528411865, 0.13323144614696503, 0.0008751187124289572, -0.14067673683166504, 0.09818065166473389, 0.4896935820579529, 0.05618814751505852, -0.14490316808223724, 0.3308490514755249, 0.20873454213142395, 0.24980413913726807, -0.11164992302656174, -0.04191485047340393, -0.07632182538509369, -0.007912087254226208, 0.10350774228572845, 0.07728636264801025, -0.0560007318854332, 0.017136143520474434, 0.07414616644382477, -0.0036134684924036264, -0.15379202365875244, 0.21246269345283508, 0.07286833226680756, 0.3913024961948395, -0.11124511808156967, -0.2594476044178009, -0.005773492157459259, -0.4689764380455017, -0.005583834368735552, 0.03185344487428665, -0.021543025970458984, -0.15841111540794373, -0.17287009954452515, 0.05907157063484192, -0.3485897481441498, 0.26234281063079834, 0.16069422662258148, 0.23712962865829468, 0.16907602548599243, -0.13415367901325226, -0.5272517800331116, -0.2941480278968811, -0.3322207033634186, 0.18113477528095245, 0.5960766077041626, 0.1151440367102623, -0.13822154700756073, -0.1728234589099884, 0.10999204963445663, -0.05185002088546753, -0.148247629404068, 0.10066864639520645, -0.33127129077911377, 0.20241938531398773, 0.15132899582386017, -0.2806151509284973, 0.04583752155303955, 0.3753645420074463, -0.2438097596168518, 0.12480254471302032, 0.06615769863128662, -0.14999350905418396, 0.21233011782169342, 0.13495221734046936, 0.33930888772010803, -0.10402209311723709, -0.2439710795879364, -0.29319724440574646, 0.30135512351989746, -0.351443886756897, 0.3701418936252594, 0.12124495208263397, 0.00033597677247598767, 0.30029138922691345, -0.09358716756105423, 0.10684158653020859, 0.31012898683547974, -0.1816413700580597, 0.1664065271615982, -0.6221524477005005, 0.11270283907651901, -0.4461793601512909, -0.21020959317684174, 0.3684735894203186, -0.07535848766565323, 0.28507164120674133, -0.1410282403230667, 0.20845304429531097, 0.23064105212688446, -0.4770975112915039, -0.07949839532375336, -0.16209371387958527, -0.19412653148174286, 0.2774370312690735, -0.020417680963873863, 0.7596281170845032, 0.43547335267066956, 0.5633998513221741, 0.2302163988351822, -0.14735105633735657, 0.38939759135246277, -0.5955860018730164, 0.19127987325191498, -0.3415153920650482, -0.37023451924324036, 0.09574269503355026, 0.10443361103534698, -0.06511300802230835, 0.2145492285490036, 0.2076512575149536, 0.29537633061408997, 0.3931225538253784, -0.0991114005446434, -0.10828062146902084, 0.17655280232429504, 0.12343021482229233, 0.18754109740257263, -0.2870902717113495, 0.19913658499717712, 0.34644386172294617, 0.004576204810291529, -0.05969381704926491, 0.09596779942512512, -0.7420679330825806, 0.0225160401314497, 0.04249162599444389, 0.10044686496257782, -0.09559981524944305, 0.35474157333374023, 0.11958033591508865, -0.22432821989059448, -0.048718634992837906, 0.3672381341457367, 0.15987472236156464, -0.01591307297348976, 0.007195963989943266, 0.21002458035945892, -0.3194086253643036, 0.3263314664363861, 0.24231000244617462, -0.038803331553936005, 0.031407807022333145, -0.22021394968032837, -0.07871294766664505, -0.07076095044612885, 0.15572413802146912, -0.16848191618919373, -0.548475444316864, 0.18215550482273102, -0.33470356464385986, 0.0072069619782269, -0.02991681545972824, 0.10068981349468231, -0.09960504621267319, -0.05343042314052582, 0.16725927591323853, -0.05686292424798012, -0.06265318393707275, -0.03488534316420555, 0.09307224303483963, -0.3328576385974884, -0.05046382173895836, 0.4744253158569336, -0.1367390751838684, -0.046009812504053116, 0.06487401574850082, -0.330565869808197, 0.03526663780212402, -0.2431761920452118, 0.03892473876476288, -0.08075796067714691, -0.19715382158756256, -0.10430055856704712, -0.007580338045954704, -0.2291494905948639, 0.25272709131240845, 0.25843632221221924, 0.2495664805173874, 0.21303711831569672, 0.032852903008461, -0.3630848824977875, -0.09832476824522018, 0.08107944577932358, 0.1679994910955429, -0.034733790904283524, -0.01961880549788475, -0.15958917140960693, 0.25884079933166504, 0.2306310385465622, -0.35463520884513855, -0.15017235279083252, -0.08997538685798645, 0.12681719660758972, -0.2210235744714737, -0.31873929500579834, 0.02248654142022133, 0.1468537151813507, 0.1430858075618744, -0.008150210604071617, -0.05782909318804741, -0.2477169781923294, -0.20516584813594818, 0.08322594314813614, -0.04834677651524544, -0.09744031727313995, -0.13200196623802185, 0.21839942038059235, 0.10436245799064636, -0.14063531160354614, -0.2313222587108612, -0.13392621278762817, 0.35739850997924805, 0.12174567580223083, -0.08553913235664368, 0.1825161725282669, -0.14262963831424713, 0.10866586863994598, 0.018823904916644096, -0.025384819135069847, 0.07357653230428696, -0.026512816548347473, 0.109457828104496, -0.2018422782421112, -0.25641098618507385, -0.07424065470695496, -0.2803899943828583, 0.14717309176921844, 0.10414627939462662, -0.0529727004468441, 0.008782899007201195, -0.18328943848609924, 0.2844220995903015, 0.36497530341148376, 0.11780496686697006, -0.09963652491569519, -0.013917454518377781, 0.22528259456157684, -0.22220176458358765, 0.09425390511751175, 0.44961801171302795, 0.043986886739730835, 0.06563520431518555, -0.02828390896320343, 0.15857845544815063, -0.20769856870174408, -0.018284499645233154, -0.05924854055047035, -0.27442365884780884, -0.018169712275266647, 0.17280763387680054, -0.19119587540626526, 0.25152382254600525, 0.0736309215426445, 0.1553812474012375, -0.37914833426475525, 0.18718206882476807, 0.3507705330848694, -0.0025792052038013935, 0.19415226578712463, -0.2856432795524597, 0.14951284229755402, -0.02367577515542507, 0.09032062441110611, -0.27035999298095703, 0.3451123535633087, 0.0853993147611618, -0.02180611714720726, 0.08363576978445053, 0.16912171244621277, -0.10833396762609482, -0.33761361241340637, -0.0048829554580152035, -0.0880928635597229, -0.07047468423843384, -0.5443993806838989, -0.46642449498176575, -0.22646990418434143, -0.3945833742618561, 0.07105030119419098, -0.16874152421951294, 0.38178619742393494, 0.02951897494494915, 0.2477216273546219, -0.09672477096319199, -0.1826779842376709, -0.46176472306251526, 0.010245916433632374, 0.42088058590888977, -0.046756893396377563, 0.17743198573589325, 0.21671564877033234, 0.41431307792663574, 0.3571798503398895, 0.16161072254180908, 0.23874562978744507, 0.18882715702056885, -0.2336151897907257, -0.2279312014579773, 0.03246549516916275, 0.08932047337293625, -0.19615767896175385, 0.17421679198741913, 0.2599768042564392, 0.04339063540101051, 0.3002179265022278, 0.132206529378891, -0.2410813421010971, 0.07742050290107727, 0.2649875283241272, -0.0035750248935073614, -0.21412977576255798, 0.29473695158958435, -0.03916145861148834, 0.018682796508073807, -0.33147430419921875, 0.3797481954097748, -0.5077443718910217, -0.08658923208713531, -0.5439392328262329, 0.15140385925769806, 0.17864570021629333, -0.12435408681631088, 0.10393213480710983, 0.26889467239379883, 0.41731399297714233, -0.127236008644104, 0.3387746214866638, -0.28413739800453186, -0.46132057905197144, -0.26536378264427185, -0.04585053399205208, 0.039508700370788574, 0.27052944898605347, -0.43093833327293396, 0.28873541951179504, -0.29741501808166504, 0.051450327038764954, 0.2745601534843445, 0.42833009362220764, 0.24691952764987946, 0.06649734079837799, -0.33822908997535706, 0.1559215635061264, 0.19840335845947266, -0.08824482560157776, -0.11956939101219177, 0.13063283264636993, -0.17390663921833038, 0.019816448912024498, 0.1097150519490242, 0.08555615693330765, -0.3465365171432495, -0.2886541783809662, 0.4207555949687958, 0.33852237462997437, 0.27506640553474426, 0.26808562874794006, -0.37258148193359375, -0.38539496064186096, 0.03100375458598137, -0.12375269085168839, -0.0627094954252243, 0.062282588332891464, -0.06584226340055466, 0.3876439929008484, 0.025063853710889816, 0.07477404177188873, 0.07386086881160736, 0.5041282176971436, 0.05595526471734047, 0.0017199635040014982, -0.22703063488006592, -0.026206299662590027, 0.09364594519138336, 0.06342308223247528, -0.6129956841468811, 0.40744340419769287, 0.0005166334449313581, 0.18714416027069092, -0.2337079793214798, -0.2538995146751404, 0.37504610419273376, -0.3584226965904236, 0.27204254269599915, -0.09970097243785858, 0.34275397658348083, 0.2217620611190796, 0.29680484533309937, 0.04679165408015251, -0.10589522123336792, 0.0717540830373764, 0.29574862122535706, 0.162244513630867, 0.23034587502479553, -0.2334207147359848, -0.08380375802516937, -0.0016766850603744388, 0.29279467463493347, 0.0649266391992569, -0.29197120666503906, 0.4134342670440674, -0.184107705950737 ]
https://github.com/huggingface/datasets/issues/225
[ROUGE] Different scores with `files2rouge`
You're right, there's a pretty interesting trade-off here between robustness and sensitivity :) The flip side of your argument is that we also still want the metric to be sensitive to model mistakes. How we think about number normalization for example has evolved a fair bit since the Perl script was written: at the time, ROUGE was used mostly to evaluate short-medium text summarization systems, where there were only a few numbers in the input and it was assumed that the most popular methods in use at the time would get those right. However, as your example showcases, that assumption does not hold any more, and we do want to be able to penalize a model that generates a wrong numerical value. Also, we think that abstracting away tokenization differences is the role of the model/tokenizer: if you use the 🤗Tokenizers library for example, it will handle that for you ;) Finally, there is a lot of active research on developing model-powered metrics that are both more sensitive and more robust than ROUGE. Check out for example BERTscore, which is implemented in this library!
It seems that the ROUGE score of `nlp` is lower than the one of `files2rouge`. Here is a self-contained notebook to reproduce both scores : https://colab.research.google.com/drive/14EyAXValB6UzKY9x4rs_T3pyL7alpw_F?usp=sharing --- `nlp` : (Only mid F-scores) >rouge1 0.33508031962733364 rouge2 0.14574333776191592 rougeL 0.2321187823256159 `files2rouge` : >Running ROUGE... =========================== 1 ROUGE-1 Average_R: 0.48873 (95%-conf.int. 0.41192 - 0.56339) 1 ROUGE-1 Average_P: 0.29010 (95%-conf.int. 0.23605 - 0.34445) 1 ROUGE-1 Average_F: 0.34761 (95%-conf.int. 0.29479 - 0.39871) =========================== 1 ROUGE-2 Average_R: 0.20280 (95%-conf.int. 0.14969 - 0.26244) 1 ROUGE-2 Average_P: 0.12772 (95%-conf.int. 0.08603 - 0.17752) 1 ROUGE-2 Average_F: 0.14798 (95%-conf.int. 0.10517 - 0.19240) =========================== 1 ROUGE-L Average_R: 0.32960 (95%-conf.int. 0.26501 - 0.39676) 1 ROUGE-L Average_P: 0.19880 (95%-conf.int. 0.15257 - 0.25136) 1 ROUGE-L Average_F: 0.23619 (95%-conf.int. 0.19073 - 0.28663) --- When using longer predictions/gold, the difference is bigger. **How can I reproduce same score as `files2rouge` ?** @lhoestq
184
[ROUGE] Different scores with `files2rouge` It seems that the ROUGE score of `nlp` is lower than the one of `files2rouge`. Here is a self-contained notebook to reproduce both scores : https://colab.research.google.com/drive/14EyAXValB6UzKY9x4rs_T3pyL7alpw_F?usp=sharing --- `nlp` : (Only mid F-scores) >rouge1 0.33508031962733364 rouge2 0.14574333776191592 rougeL 0.2321187823256159 `files2rouge` : >Running ROUGE... =========================== 1 ROUGE-1 Average_R: 0.48873 (95%-conf.int. 0.41192 - 0.56339) 1 ROUGE-1 Average_P: 0.29010 (95%-conf.int. 0.23605 - 0.34445) 1 ROUGE-1 Average_F: 0.34761 (95%-conf.int. 0.29479 - 0.39871) =========================== 1 ROUGE-2 Average_R: 0.20280 (95%-conf.int. 0.14969 - 0.26244) 1 ROUGE-2 Average_P: 0.12772 (95%-conf.int. 0.08603 - 0.17752) 1 ROUGE-2 Average_F: 0.14798 (95%-conf.int. 0.10517 - 0.19240) =========================== 1 ROUGE-L Average_R: 0.32960 (95%-conf.int. 0.26501 - 0.39676) 1 ROUGE-L Average_P: 0.19880 (95%-conf.int. 0.15257 - 0.25136) 1 ROUGE-L Average_F: 0.23619 (95%-conf.int. 0.19073 - 0.28663) --- When using longer predictions/gold, the difference is bigger. **How can I reproduce same score as `files2rouge` ?** @lhoestq You're right, there's a pretty interesting trade-off here between robustness and sensitivity :) The flip side of your argument is that we also still want the metric to be sensitive to model mistakes. How we think about number normalization for example has evolved a fair bit since the Perl script was written: at the time, ROUGE was used mostly to evaluate short-medium text summarization systems, where there were only a few numbers in the input and it was assumed that the most popular methods in use at the time would get those right. However, as your example showcases, that assumption does not hold any more, and we do want to be able to penalize a model that generates a wrong numerical value. Also, we think that abstracting away tokenization differences is the role of the model/tokenizer: if you use the 🤗Tokenizers library for example, it will handle that for you ;) Finally, there is a lot of active research on developing model-powered metrics that are both more sensitive and more robust than ROUGE. Check out for example BERTscore, which is implemented in this library!
[ -0.14572615921497345, -0.16974303126335144, -0.1972297579050064, 0.3397459387779236, -0.11949753761291504, -0.4242601990699768, -0.4203995168209076, 0.1476738452911377, -0.2700440585613251, 0.17250607907772064, -0.08286702632904053, 0.1803876757621765, 0.12033016234636307, 0.008093580603599548, -0.1003899946808815, -0.09797486662864685, 0.165655255317688, 0.03882870450615883, 0.09248090535402298, -0.3720676898956299, 0.06655961275100708, 0.5696011781692505, -0.03133227676153183, 0.2104341685771942, 0.05402207747101784, 0.2702381908893585, 0.28511351346969604, 0.3035299479961395, -0.14858005940914154, 0.028436783701181412, 0.12800203263759613, 0.03064163215458393, -0.02369801700115204, 0.22214896976947784, -0.00010886792006203905, -0.2347191721200943, 0.10260970890522003, -0.0031664110720157623, -0.27572351694107056, -0.30260902643203735, -0.034557461738586426, 0.06436412036418915, -0.1229282021522522, -0.07039352506399155, 0.08584654331207275, -0.04489225521683693, 0.04877280816435814, 0.20954246819019318, 0.4293985068798065, 0.1475730836391449, 0.22625432908535004, -0.30243781208992004, -0.18534952402114868, 0.024081915616989136, 0.5763341188430786, 0.16267599165439606, -0.11448317766189575, 0.44726425409317017, 0.005699495319277048, -0.3526380658149719, -0.24019040167331696, -0.008563567884266376, 0.34304094314575195, -0.17255988717079163, 0.13255919516086578, 0.19315466284751892, 0.14858555793762207, -0.04594054073095322, -0.07953529059886932, 0.24472841620445251, -0.2702822983264923, 0.03221165016293526, -0.22717157006263733, -0.2778386175632477, -0.07631559669971466, -0.1997607946395874, 0.11689180135726929, -0.08144606649875641, 0.1642860472202301, 0.012247875332832336, 0.04413227364420891, 0.29542678594589233, 0.04963891953229904, 0.007164150010794401, -0.05405605211853981, -0.10784615576267242, 0.05084570124745369, 0.09846562147140503, 0.29527974128723145, -0.050543639808893204, -0.5163149237632751, 0.01077284850180149, -0.2593677043914795, 0.0749606341123581, -0.30228114128112793, -0.03785572201013565, 0.3144964575767517, 0.4659734070301056, -0.20087116956710815, 0.4996274709701538, 0.26919713616371155, -0.0934174507856369, -0.13220743834972382, -0.03944722190499306, -0.13782373070716858, 0.10813210904598236, -0.028945928439497948, 0.020693473517894745, 0.08811821043491364, 0.2246810793876648, -0.13358263671398163, -0.012008018791675568, 0.39822256565093994, -0.5637689232826233, -0.4777999222278595, 0.10259132832288742, -0.26103466749191284, -0.4323975145816803, -0.6891934871673584, -0.0673605352640152, -0.255431205034256, -0.11924232542514801, 0.09890846163034439, 0.22720065712928772, -0.17488402128219604, 0.29229849576950073, 0.046788230538368225, 0.18339158594608307, -0.28950363397598267, 0.01966216042637825, -0.18399901688098907, 0.06357946246862411, -0.31439271569252014, 0.13452430069446564, 0.17056682705879211, 0.11274684220552444, 0.31805747747421265, 0.25605344772338867, 0.27351316809654236, 0.03864440321922302, 0.3298662304878235, -0.2839505672454834, -0.022634783759713173, -0.21111683547496796, -0.2113915979862213, 0.1974538117647171, 0.027129076421260834, 0.09184830635786057, -0.00870169885456562, -0.07489427924156189, -0.18084076046943665, 0.1442372053861618, 0.46661806106567383, 0.20657984912395477, 0.14965085685253143, -0.11055408418178558, 0.04069235175848007, 0.11693812906742096, -0.24016346037387848, -0.5936912298202515, 0.20406076312065125, -0.1365523636341095, -0.1270110011100769, -0.05817752704024315, 0.10981972515583038, -0.03458227589726448, 0.09312087297439575, -0.11743129789829254, 0.3731990456581116, 0.06358262896537781, 0.26031258702278137, 0.45348334312438965, -0.0408313013613224, 0.011863946914672852, -0.1628563404083252, -0.07088572531938553, 0.4665316641330719, -0.43853142857551575, -0.22229626774787903, 0.3355002701282501, -0.16507026553153992, -0.06518586724996567, 0.23904691636562347, 0.03927561268210411, -0.3886531591415405, -0.16600807011127472, 0.17745089530944824, 0.19752459228038788, 0.07779872417449951, -0.048396892845630646, -0.2970068156719208, 0.13495878875255585, 0.316625714302063, 0.06547749787569046, -0.23182860016822815, -0.29935044050216675, -0.11140508949756622, -0.06439077854156494, 0.5141965746879578, 0.13280688226222992, -0.27513760328292847, 0.10499175637960434, -0.2721599340438843, 0.29369354248046875, -0.05169543996453285, 0.11857372522354126, 0.4230367839336395, -0.023375488817691803, -0.7951249480247498, -0.16027650237083435, 0.4370328485965729, -0.05248735845088959, -0.21229121088981628, -0.4714091122150421, 0.24053311347961426, -0.23946474492549896, 0.22328263521194458, 0.05504881590604782, 0.1873215287923813, 0.030354246497154236, 0.02407083846628666, 0.09660714119672775, -0.23642665147781372, -0.15861894190311432, -0.17153561115264893, 0.20367735624313354, -0.22109828889369965, -0.007432829588651657, -0.01054032240062952, 0.35960862040519714, 0.31164121627807617, 0.0875801146030426, 0.03730439767241478, 0.2634182274341583, 0.05015571042895317, 0.09804999828338623, 0.3672510087490082, 0.14480578899383545, -0.0007639330578967929, 0.18896162509918213, -0.1226600706577301, 0.6879815459251404, 0.01646246388554573, 0.03225119411945343, -0.3400397300720215, 0.42591556906700134, -0.21703946590423584, 0.20069162547588348, 0.4093106985092163, -0.07173748314380646, -0.05009761080145836, 0.057780198752880096, -0.006800386123359203, -0.01715300790965557, 0.7467252016067505, 0.1952691376209259, 0.2502553164958954, 0.0029566255398094654, -0.32202571630477905, 0.25961771607398987, 0.39914047718048096, -0.015342896804213524, -0.02085353247821331, -0.11995424330234528, 0.19371458888053894, -0.09338841587305069, -0.036720160394907, 0.2507389783859253, 0.3810695707798004, 0.19513395428657532, 0.1557074785232544, -0.13119934499263763, -0.3382363021373749, -0.34380224347114563, 0.03513277322053909, -0.18210364878177643, 0.08744674175977707, 0.49638640880584717, 0.10160709917545319, 0.21628661453723907, -0.3020774722099304, -0.5422855019569397, -0.25364869832992554, -0.22231446206569672, 0.06132601946592331, 0.16164496541023254, -0.14391355216503143, -0.3678353428840637, -0.15366077423095703, -0.091903455555439, -0.2918891906738281, -0.27327603101730347, 0.23009741306304932, 0.020875325426459312, -0.07249978184700012, -0.21893787384033203, 0.023232346400618553, 0.14874784648418427, -0.3703943192958832, 0.1783805787563324, -0.017374185845255852, -0.0280428696423769, -0.27953794598579407, 0.2228175401687622, -0.05679844319820404, -0.019304806366562843, -0.03147659823298454, -0.27918699383735657, -0.450835645198822, 0.16968795657157898, -0.37626758217811584, -0.010960331186652184, 0.06464394927024841, -0.0032369038090109825, 0.05967245250940323, -0.25061726570129395, -0.07057500630617142, 0.14640170335769653, 0.09220732748508453, -0.17413024604320526, -0.3507431149482727, -0.08965101093053818, 0.0367850661277771, 0.014213897287845612, -0.19140632450580597, -0.2775213420391083, 0.07937490195035934, -0.1433197259902954, 0.6197946667671204, 0.3270145356655121, 0.005700475536286831, 0.38587889075279236, -0.2369561791419983, 0.00001117925057769753, 0.1717422902584076, 0.5837682485580444, -0.260583758354187, -0.04442865401506424, 0.1986089050769806, -0.29971224069595337, -0.4421885013580322, -0.27306097745895386, 0.18830086290836334, 0.03232232853770256, -0.008429203182458878, -0.1786167025566101, 0.031004473567008972, -0.13094371557235718, -0.04899758845567703, -0.21304526925086975, -0.05581812188029289, 0.3378271460533142, -0.18365980684757233, -0.2421921342611313, -0.0644744485616684, 0.21673358976840973, 0.02994077280163765, 0.286546915769577, 0.4599354863166809, -0.4840276837348938, 0.16461147367954254, -0.14009366929531097, 0.12755979597568512, 0.0440831296145916, 0.5117963552474976, 0.15338170528411865, 0.13323144614696503, 0.0008751187124289572, -0.14067673683166504, 0.09818065166473389, 0.4896935820579529, 0.05618814751505852, -0.14490316808223724, 0.3308490514755249, 0.20873454213142395, 0.24980413913726807, -0.11164992302656174, -0.04191485047340393, -0.07632182538509369, -0.007912087254226208, 0.10350774228572845, 0.07728636264801025, -0.0560007318854332, 0.017136143520474434, 0.07414616644382477, -0.0036134684924036264, -0.15379202365875244, 0.21246269345283508, 0.07286833226680756, 0.3913024961948395, -0.11124511808156967, -0.2594476044178009, -0.005773492157459259, -0.4689764380455017, -0.005583834368735552, 0.03185344487428665, -0.021543025970458984, -0.15841111540794373, -0.17287009954452515, 0.05907157063484192, -0.3485897481441498, 0.26234281063079834, 0.16069422662258148, 0.23712962865829468, 0.16907602548599243, -0.13415367901325226, -0.5272517800331116, -0.2941480278968811, -0.3322207033634186, 0.18113477528095245, 0.5960766077041626, 0.1151440367102623, -0.13822154700756073, -0.1728234589099884, 0.10999204963445663, -0.05185002088546753, -0.148247629404068, 0.10066864639520645, -0.33127129077911377, 0.20241938531398773, 0.15132899582386017, -0.2806151509284973, 0.04583752155303955, 0.3753645420074463, -0.2438097596168518, 0.12480254471302032, 0.06615769863128662, -0.14999350905418396, 0.21233011782169342, 0.13495221734046936, 0.33930888772010803, -0.10402209311723709, -0.2439710795879364, -0.29319724440574646, 0.30135512351989746, -0.351443886756897, 0.3701418936252594, 0.12124495208263397, 0.00033597677247598767, 0.30029138922691345, -0.09358716756105423, 0.10684158653020859, 0.31012898683547974, -0.1816413700580597, 0.1664065271615982, -0.6221524477005005, 0.11270283907651901, -0.4461793601512909, -0.21020959317684174, 0.3684735894203186, -0.07535848766565323, 0.28507164120674133, -0.1410282403230667, 0.20845304429531097, 0.23064105212688446, -0.4770975112915039, -0.07949839532375336, -0.16209371387958527, -0.19412653148174286, 0.2774370312690735, -0.020417680963873863, 0.7596281170845032, 0.43547335267066956, 0.5633998513221741, 0.2302163988351822, -0.14735105633735657, 0.38939759135246277, -0.5955860018730164, 0.19127987325191498, -0.3415153920650482, -0.37023451924324036, 0.09574269503355026, 0.10443361103534698, -0.06511300802230835, 0.2145492285490036, 0.2076512575149536, 0.29537633061408997, 0.3931225538253784, -0.0991114005446434, -0.10828062146902084, 0.17655280232429504, 0.12343021482229233, 0.18754109740257263, -0.2870902717113495, 0.19913658499717712, 0.34644386172294617, 0.004576204810291529, -0.05969381704926491, 0.09596779942512512, -0.7420679330825806, 0.0225160401314497, 0.04249162599444389, 0.10044686496257782, -0.09559981524944305, 0.35474157333374023, 0.11958033591508865, -0.22432821989059448, -0.048718634992837906, 0.3672381341457367, 0.15987472236156464, -0.01591307297348976, 0.007195963989943266, 0.21002458035945892, -0.3194086253643036, 0.3263314664363861, 0.24231000244617462, -0.038803331553936005, 0.031407807022333145, -0.22021394968032837, -0.07871294766664505, -0.07076095044612885, 0.15572413802146912, -0.16848191618919373, -0.548475444316864, 0.18215550482273102, -0.33470356464385986, 0.0072069619782269, -0.02991681545972824, 0.10068981349468231, -0.09960504621267319, -0.05343042314052582, 0.16725927591323853, -0.05686292424798012, -0.06265318393707275, -0.03488534316420555, 0.09307224303483963, -0.3328576385974884, -0.05046382173895836, 0.4744253158569336, -0.1367390751838684, -0.046009812504053116, 0.06487401574850082, -0.330565869808197, 0.03526663780212402, -0.2431761920452118, 0.03892473876476288, -0.08075796067714691, -0.19715382158756256, -0.10430055856704712, -0.007580338045954704, -0.2291494905948639, 0.25272709131240845, 0.25843632221221924, 0.2495664805173874, 0.21303711831569672, 0.032852903008461, -0.3630848824977875, -0.09832476824522018, 0.08107944577932358, 0.1679994910955429, -0.034733790904283524, -0.01961880549788475, -0.15958917140960693, 0.25884079933166504, 0.2306310385465622, -0.35463520884513855, -0.15017235279083252, -0.08997538685798645, 0.12681719660758972, -0.2210235744714737, -0.31873929500579834, 0.02248654142022133, 0.1468537151813507, 0.1430858075618744, -0.008150210604071617, -0.05782909318804741, -0.2477169781923294, -0.20516584813594818, 0.08322594314813614, -0.04834677651524544, -0.09744031727313995, -0.13200196623802185, 0.21839942038059235, 0.10436245799064636, -0.14063531160354614, -0.2313222587108612, -0.13392621278762817, 0.35739850997924805, 0.12174567580223083, -0.08553913235664368, 0.1825161725282669, -0.14262963831424713, 0.10866586863994598, 0.018823904916644096, -0.025384819135069847, 0.07357653230428696, -0.026512816548347473, 0.109457828104496, -0.2018422782421112, -0.25641098618507385, -0.07424065470695496, -0.2803899943828583, 0.14717309176921844, 0.10414627939462662, -0.0529727004468441, 0.008782899007201195, -0.18328943848609924, 0.2844220995903015, 0.36497530341148376, 0.11780496686697006, -0.09963652491569519, -0.013917454518377781, 0.22528259456157684, -0.22220176458358765, 0.09425390511751175, 0.44961801171302795, 0.043986886739730835, 0.06563520431518555, -0.02828390896320343, 0.15857845544815063, -0.20769856870174408, -0.018284499645233154, -0.05924854055047035, -0.27442365884780884, -0.018169712275266647, 0.17280763387680054, -0.19119587540626526, 0.25152382254600525, 0.0736309215426445, 0.1553812474012375, -0.37914833426475525, 0.18718206882476807, 0.3507705330848694, -0.0025792052038013935, 0.19415226578712463, -0.2856432795524597, 0.14951284229755402, -0.02367577515542507, 0.09032062441110611, -0.27035999298095703, 0.3451123535633087, 0.0853993147611618, -0.02180611714720726, 0.08363576978445053, 0.16912171244621277, -0.10833396762609482, -0.33761361241340637, -0.0048829554580152035, -0.0880928635597229, -0.07047468423843384, -0.5443993806838989, -0.46642449498176575, -0.22646990418434143, -0.3945833742618561, 0.07105030119419098, -0.16874152421951294, 0.38178619742393494, 0.02951897494494915, 0.2477216273546219, -0.09672477096319199, -0.1826779842376709, -0.46176472306251526, 0.010245916433632374, 0.42088058590888977, -0.046756893396377563, 0.17743198573589325, 0.21671564877033234, 0.41431307792663574, 0.3571798503398895, 0.16161072254180908, 0.23874562978744507, 0.18882715702056885, -0.2336151897907257, -0.2279312014579773, 0.03246549516916275, 0.08932047337293625, -0.19615767896175385, 0.17421679198741913, 0.2599768042564392, 0.04339063540101051, 0.3002179265022278, 0.132206529378891, -0.2410813421010971, 0.07742050290107727, 0.2649875283241272, -0.0035750248935073614, -0.21412977576255798, 0.29473695158958435, -0.03916145861148834, 0.018682796508073807, -0.33147430419921875, 0.3797481954097748, -0.5077443718910217, -0.08658923208713531, -0.5439392328262329, 0.15140385925769806, 0.17864570021629333, -0.12435408681631088, 0.10393213480710983, 0.26889467239379883, 0.41731399297714233, -0.127236008644104, 0.3387746214866638, -0.28413739800453186, -0.46132057905197144, -0.26536378264427185, -0.04585053399205208, 0.039508700370788574, 0.27052944898605347, -0.43093833327293396, 0.28873541951179504, -0.29741501808166504, 0.051450327038764954, 0.2745601534843445, 0.42833009362220764, 0.24691952764987946, 0.06649734079837799, -0.33822908997535706, 0.1559215635061264, 0.19840335845947266, -0.08824482560157776, -0.11956939101219177, 0.13063283264636993, -0.17390663921833038, 0.019816448912024498, 0.1097150519490242, 0.08555615693330765, -0.3465365171432495, -0.2886541783809662, 0.4207555949687958, 0.33852237462997437, 0.27506640553474426, 0.26808562874794006, -0.37258148193359375, -0.38539496064186096, 0.03100375458598137, -0.12375269085168839, -0.0627094954252243, 0.062282588332891464, -0.06584226340055466, 0.3876439929008484, 0.025063853710889816, 0.07477404177188873, 0.07386086881160736, 0.5041282176971436, 0.05595526471734047, 0.0017199635040014982, -0.22703063488006592, -0.026206299662590027, 0.09364594519138336, 0.06342308223247528, -0.6129956841468811, 0.40744340419769287, 0.0005166334449313581, 0.18714416027069092, -0.2337079793214798, -0.2538995146751404, 0.37504610419273376, -0.3584226965904236, 0.27204254269599915, -0.09970097243785858, 0.34275397658348083, 0.2217620611190796, 0.29680484533309937, 0.04679165408015251, -0.10589522123336792, 0.0717540830373764, 0.29574862122535706, 0.162244513630867, 0.23034587502479553, -0.2334207147359848, -0.08380375802516937, -0.0016766850603744388, 0.29279467463493347, 0.0649266391992569, -0.29197120666503906, 0.4134342670440674, -0.184107705950737 ]
https://github.com/huggingface/datasets/issues/224
[Feature Request/Help] BLEURT model -> PyTorch
Hitting this error when using bleurt with PyTorch ... ``` UnrecognizedFlagError: Unknown command line flag 'f' ``` ... and I'm assuming because it was built for TF specifically. Is there a way to use this metric in PyTorch?
Hi, I am interested in porting google research's new BLEURT learned metric to PyTorch (because I wish to do something experimental with language generation and backpropping through BLEURT). I noticed that you guys don't have it yet so I am partly just asking if you plan to add it (@thomwolf said you want to do so on Twitter). I had a go of just like manually using the checkpoint that they publish which includes the weights. It seems like the architecture is exactly aligned with the out-of-the-box BertModel in transformers just with a single linear layer on top of the CLS embedding. I loaded all the weights to the PyTorch model but I am not able to get the same numbers as the BLEURT package's python api. Here is my colab notebook where I tried https://colab.research.google.com/drive/1Bfced531EvQP_CpFvxwxNl25Pj6ptylY?usp=sharing . If you have any pointers on what might be going wrong that would be much appreciated! Thank you muchly!
38
[Feature Request/Help] BLEURT model -> PyTorch Hi, I am interested in porting google research's new BLEURT learned metric to PyTorch (because I wish to do something experimental with language generation and backpropping through BLEURT). I noticed that you guys don't have it yet so I am partly just asking if you plan to add it (@thomwolf said you want to do so on Twitter). I had a go of just like manually using the checkpoint that they publish which includes the weights. It seems like the architecture is exactly aligned with the out-of-the-box BertModel in transformers just with a single linear layer on top of the CLS embedding. I loaded all the weights to the PyTorch model but I am not able to get the same numbers as the BLEURT package's python api. Here is my colab notebook where I tried https://colab.research.google.com/drive/1Bfced531EvQP_CpFvxwxNl25Pj6ptylY?usp=sharing . If you have any pointers on what might be going wrong that would be much appreciated! Thank you muchly! Hitting this error when using bleurt with PyTorch ... ``` UnrecognizedFlagError: Unknown command line flag 'f' ``` ... and I'm assuming because it was built for TF specifically. Is there a way to use this metric in PyTorch?
[ 0.0004881111963186413, -0.19440069794654846, 0.09777987748384476, 0.1410355120897293, 0.2860071063041687, -0.2194754034280777, 0.29816076159477234, 0.17576171457767487, 0.1684710532426834, 0.15842966735363007, -0.22391483187675476, 0.20470838248729706, -0.47158339619636536, 0.24365317821502686, -0.039452627301216125, -0.09374258667230606, -0.18991369009017944, -0.0336320698261261, 0.10520436614751816, -0.015585321001708508, -0.09069841355085373, 0.2365889698266983, 0.04565054550766945, -0.028711872175335884, -0.3174610435962677, 0.35047978162765503, 0.05522030219435692, 0.05592724308371544, -0.14533686637878418, -0.032764680683612823, 0.34637650847435, -0.1924695372581482, 0.2557666301727295, 0.7926076054573059, -0.00012355620856396854, -0.21305102109909058, 0.0671902596950531, -0.011289668269455433, -0.2755616307258606, -0.30580487847328186, 0.4654068648815155, -0.2523367404937744, 0.0014709261013194919, -0.20474672317504883, -0.02709219604730606, -0.06222663074731827, -0.20242886245250702, -0.08055948466062546, 0.00865213107317686, 0.3162076473236084, 0.04068648815155029, 0.20291867852210999, 0.0005700011388398707, 0.07062750309705734, 0.09595873951911926, -0.10698021948337555, -0.11937190592288971, 0.58992600440979, 0.1905047595500946, -0.039019327610731125, -0.1993449628353119, 0.16511228680610657, 0.15574391186237335, 0.22187571227550507, 0.7811315655708313, 0.039632346481084824, -0.10514849424362183, 0.00786141399294138, 0.05174289271235466, 0.35052216053009033, -0.05701661854982376, -0.23309923708438873, -0.24215492606163025, 0.1831979602575302, 0.0682874470949173, -0.2815706729888916, -0.2522395849227905, 0.052199121564626694, 0.04698903486132622, -0.277474045753479, -0.38262972235679626, -0.354801207780838, -0.25620731711387634, 0.1344299465417862, -0.2618776559829712, 0.39998820424079895, -0.06487245857715607, -0.05895066261291504, 0.1828380972146988, 0.05480595678091049, 0.2097087800502777, -0.012297527864575386, 0.35755327343940735, 0.3185586929321289, 0.04500043764710426, -0.2273324579000473, 0.15664002299308777, -0.14627201855182648, 0.06553870439529419, -0.24303418397903442, -0.23485715687274933, 0.12916594743728638, -0.31121826171875, 0.009082162752747536, -0.005181606393307447, 0.4701381027698517, 0.038915470242500305, 0.374993234872818, 0.35039106011390686, 0.2842979431152344, -0.029754703864455223, -0.07869115471839905, -0.06778515875339508, -0.25943997502326965, 0.2883804142475128, 0.1613946408033371, -0.21710920333862305, -0.31106966733932495, -0.14794117212295532, -0.058148421347141266, -0.1836254745721817, -0.2136000096797943, 0.3854411840438843, 0.24474015831947327, -0.3576725125312805, -0.11658446490764618, 0.5851030349731445, 0.05812997743487358, -0.21005220711231232, 0.0859040915966034, -0.09296050667762756, -0.09709098935127258, -0.29787346720695496, 0.08617319166660309, 0.20618456602096558, -0.1500411033630371, 0.20931997895240784, -0.10786949098110199, 0.6836867332458496, 0.1358180046081543, 0.18622805178165436, 0.3334335684776306, 0.0832948237657547, -0.1664726585149765, -0.24997089803218842, -0.4027537703514099, 0.25848880410194397, 0.09364023804664612, -0.17199403047561646, 0.06458297371864319, -0.1143132746219635, -0.3340193033218384, -0.2537497878074646, 0.0032773467246443033, -0.2937202751636505, -0.13937215507030487, 0.04673707112669945, 0.3141675591468811, 0.0852745845913887, -0.29998788237571716, 0.021875780075788498, -0.2990495562553406, -0.2519974410533905, 0.09707092493772507, 0.2996882200241089, -0.34366199374198914, -0.37325894832611084, -0.4258343577384949, 0.11343003809452057, 0.38207101821899414, 0.2269522100687027, 0.1458989530801773, -0.150264173746109, -0.11249770224094391, 0.18048417568206787, -0.32851162552833557, 0.3918120563030243, 0.03860608488321304, -0.1812959611415863, -0.23008054494857788, 0.10167451202869415, -0.2862389087677002, -0.21849651634693146, 0.06395995616912842, -0.18907776474952698, -0.12197019904851913, 0.3215979337692261, 0.2692633867263794, -0.15223905444145203, -0.11927420645952225, -0.23923201858997345, -0.2757667303085327, 0.2382388859987259, 0.10660869628190994, 0.38597196340560913, -0.01850191131234169, 0.15808351337909698, 0.5193024277687073, 0.01872292347252369, -0.4093894362449646, -0.045770637691020966, -0.13258512318134308, 0.3818318545818329, -0.2926696240901947, 0.39089134335517883, -0.3206767737865448, -0.12131356447935104, 0.23700399696826935, -0.22047209739685059, 0.3617452085018158, 0.19992277026176453, 0.1151999905705452, -0.1020440086722374, -0.04495696723461151, 0.34095510840415955, 0.039177969098091125, -0.11112784594297409, -0.14549992978572845, -0.23571942746639252, -0.19972261786460876, -0.20018041133880615, -0.33275464177131653, -0.2919321358203888, -0.05433240905404091, 0.42354923486709595, 0.20417174696922302, -0.03849977254867554, -0.08839336782693863, -0.08542531728744507, 0.4208585023880005, 0.13108496367931366, 0.21697218716144562, -0.12121205031871796, -0.2075587511062622, 0.13657163083553314, -0.2814391255378723, 0.24335339665412903, 0.2878023087978363, 0.3453332781791687, -0.08362112194299698, 0.17048504948616028, 0.2630353271961212, -0.12292072921991348, -0.05530737340450287, -0.059550024569034576, 0.23119066655635834, 0.2989153265953064, 0.26676926016807556, 0.2532309293746948, 0.07388028502464294, -0.28169959783554077, 0.05441161245107651, 0.01285078376531601, 0.01236666738986969, 0.41856783628463745, -0.2691681981086731, -0.14407919347286224, -0.1935265213251114, 0.20795486867427826, 0.06539348512887955, 0.37455567717552185, -0.026033461093902588, 0.042965568602085114, 0.3506181538105011, -0.2904781103134155, -0.2164357751607895, -0.12244254350662231, -0.02304842695593834, 0.17220070958137512, 0.10130861401557922, 0.4216384291648865, 0.12463139742612839, -0.10538551211357117, 0.07629330456256866, -0.05406509339809418, 0.11428827792406082, -0.19989176094532013, -0.10426616668701172, 0.04114838317036629, 0.08948389440774918, 0.16858097910881042, -0.019061096012592316, 0.06301760673522949, 0.12986312806606293, -0.2475425899028778, 0.36891698837280273, 0.03347448259592056, 0.3673946261405945, -0.1088700219988823, -0.30087098479270935, -0.25574791431427, -0.18844951689243317, 0.008498410694301128, 0.25391143560409546, 0.09166216850280762, -0.14201869070529938, 0.05402098968625069, 0.46499115228652954, -0.0766248106956482, -0.14190183579921722, -0.029270092025399208, -0.3652303218841553, -0.34538912773132324, 0.00930763315409422, 0.041948579251766205, 0.042096201330423355, 0.09645827859640121, 0.08965020626783371, -0.03133614733815193, -0.23312753438949585, -0.7229449152946472, 0.0043877772986888885, -0.006776303518563509, 0.28053438663482666, 0.21111781895160675, -0.08313439786434174, -0.3391117453575134, 0.004089529160410166, 0.05235622450709343, -0.5004478693008423, -0.005801171064376831, 0.3323717713356018, -0.3564039468765259, 0.08894725888967514, -0.07864081859588623, -0.10566962510347366, -0.09344351291656494, -0.07754146307706833, -0.07092245668172836, 0.1984407901763916, 0.05370669811964035, -0.05838702246546745, -0.020465658977627754, 0.4024656116962433, 0.5086859464645386, -0.04803271219134331, -0.16854068636894226, 0.24041227996349335, 0.34235885739326477, -0.3283780515193939, -0.30375558137893677, -0.028277447447180748, 0.06587962806224823, 0.029017644003033638, 0.005569435190409422, -0.30144375562667847, -0.4165251851081848, -0.1035078838467598, -0.0752798318862915, 0.05142509564757347, 0.20570331811904907, 0.024882502853870392, 0.02475367859005928, 0.07888978719711304, -0.2982487380504608, -0.20026558637619019, -0.019337110221385956, 0.14835581183433533, 0.709057629108429, -0.1582220494747162, -0.17756019532680511, 0.148504376411438, 0.8652082681655884, 0.01293997373431921, -0.45980173349380493, 0.10806480795145035, 0.17440655827522278, 0.09269385784864426, 0.01570582203567028, -0.21939250826835632, 0.2692297101020813, -0.30339282751083374, -0.1727745234966278, 0.3044235408306122, 0.10612781345844269, -0.12558585405349731, -0.27442967891693115, -0.1240757554769516, 0.5246572494506836, -0.1307867467403412, 0.043277062475681305, 0.08086630702018738, 0.027793755754828453, -0.05913775786757469, -0.13328935205936432, -0.33599796891212463, -0.11665191501379013, 0.28862085938453674, 0.0849209576845169, -0.01689912937581539, 0.0468011312186718, 0.4622920751571655, -0.11969182640314102, -0.5215772986412048, 0.4779229760169983, 0.14861808717250824, -0.6269400119781494, 0.030342500656843185, -0.09077303111553192, -0.01497798040509224, -0.004339572042226791, -0.30573028326034546, 0.08299003541469574, 0.022056449204683304, -0.02300361730158329, 0.04754474014043808, -0.26395514607429504, 0.1110740676522255, -0.13539715111255646, -0.3203109800815582, 0.47493600845336914, 0.2835911214351654, -0.35798653960227966, 0.06025190278887749, 0.26615551114082336, 0.3918269872665405, -0.16820651292800903, -0.04425893351435661, -0.3871445953845978, -0.24752135574817657, 0.0511750727891922, -0.16052784025669098, 0.05983075127005577, 0.2689394950866699, 0.06398972123861313, 0.15757103264331818, 0.1485833078622818, -0.052385203540325165, 0.15332184731960297, 0.06338650733232498, 0.0676540806889534, -0.03543031960725784, 0.03659791126847267, 0.1509351283311844, -0.05382094159722328, 0.16688130795955658, 0.31179079413414, -0.41491469740867615, -0.49683740735054016, 0.1397722214460373, 0.26618656516075134, 0.21740740537643433, 0.03491002693772316, 0.07774903625249863, -0.10712707787752151, 0.32527950406074524, 0.31367388367652893, 0.17274154722690582, 0.32259780168533325, 0.15361587703227997, 0.31555402278900146, 0.11419893801212311, -0.18386094272136688, 0.3391725420951843, -0.11920905858278275, 0.1223263293504715, 0.34898853302001953, 0.029947733506560326, -0.027535313740372658, 0.06459622830152512, 0.5371320843696594, 0.9507373571395874, 0.07198932021856308, 0.10562974214553833, 0.3453923463821411, -0.215912863612175, 0.3318781554698944, -0.13465960323810577, -0.0547708123922348, -0.20902861654758453, -0.27686455845832825, -0.02873704954981804, -0.009400848299264908, -0.41684210300445557, -0.3024942874908447, -0.24513137340545654, 0.17041294276714325, -0.3722742795944214, -0.2893781065940857, 0.2739180028438568, 0.28952452540397644, -0.012750701978802681, 0.026801927015185356, 0.28437095880508423, 0.072120800614357, -0.24628637731075287, 0.3137527406215668, -0.06892000138759613, -0.13711807131767273, -0.12174974381923676, 0.11128230392932892, -0.34428122639656067, 0.1823102980852127, -0.14361557364463806, -0.04318654537200928, -0.13794079422950745, -0.2727847993373871, 0.04846378415822983, 0.04304049164056778, 0.760071873664856, 0.019664937630295753, -0.3796810209751129, 0.22250451147556305, -0.0693352073431015, 0.18749578297138214, 0.12328057736158371, -0.11628574132919312, 0.36192774772644043, -0.21870289742946625, -0.08358759433031082, -0.1212911605834961, 0.23607264459133148, -0.44642966985702515, 0.16874219477176666, -0.15053196251392365, -0.16223572194576263, 0.011513803154230118, 0.24560847878456116, 0.23653115332126617, 0.11821123957633972, -0.10112264752388, 0.05867791548371315, 0.1474025547504425, -0.23920838534832, 0.03270941600203514, 0.04361796751618385, -0.31197041273117065, 0.029099145904183388, -0.041438158601522446, -0.07832906395196915, -0.009219806641340256, 0.22514589130878448, 0.18119817972183228, -0.08502651005983353, 0.23313820362091064, -0.5456887483596802, 0.4800319969654083, -0.397786021232605, 0.16316542029380798, 0.08292680978775024, -0.03360690176486969, 0.08675359189510345, 0.7250290513038635, 0.4419586956501007, 0.13836151361465454, -0.45515573024749756, 0.10337221622467041, -0.3853459060192108, 0.05548091605305672, -0.18755128979682922, 0.3016716539859772, -0.21308331191539764, 0.16312801837921143, -0.051664385944604874, 0.19557583332061768, -0.19422313570976257, -0.019208339974284172, 0.12870511412620544, -0.13616466522216797, 0.05789308622479439, -0.36982840299606323, 0.18015773594379425, 0.07099325209856033, -0.2820509076118469, -0.12980453670024872, 0.07917574793100357, -0.0011406919220462441, -0.49501484632492065, 0.16216841340065002, 0.23006001114845276, -0.015350005589425564, -0.3231397867202759, -0.445538192987442, -0.05745205655694008, -0.10607441514730453, 0.021053338423371315, 0.14171026647090912, 0.044790253043174744, 0.3292350172996521, -0.059124648571014404, -0.04719117656350136, -0.10355976969003677, -0.05455062538385391, 0.4670523703098297, 0.22028197348117828, 0.11532914638519287, 0.012021305970847607, 0.343756765127182, 0.001460233354009688, 0.13209153711795807, 0.029761342331767082, 0.006226769182831049, 0.06046155467629433, 0.297881543636322, 0.14499522745609283, 0.19285275042057037, -0.36993929743766785, 0.2290143072605133, 0.14820770919322968, 0.12221922725439072, -0.16198107600212097, 0.17190201580524445, 0.03464852273464203, -0.23638753592967987, -0.2088804394006729, -0.3177430033683777, 0.08207542449235916, -0.17358551919460297, 0.13484296202659607, -0.2804165184497833, 0.4191928207874298, 0.31389856338500977, -0.09825564920902252, -0.24227790534496307, -0.22451092302799225, 0.068136066198349, 0.16164055466651917, -0.18700948357582092, 0.3133464753627777, 0.40764516592025757, 0.11467631161212921, 0.2044929563999176, -0.4011959135532379, -0.14249007403850555, 0.06567215174436569, -0.42258694767951965, -0.08957450091838837, 0.322716623544693, 0.11632955074310303, -0.05320845544338226, -0.13574358820915222, 0.17436078190803528, -0.07144564390182495, -0.1919138878583908, 0.39544880390167236, -0.3155261278152466, -0.21161696314811707, -0.08746592700481415, 0.3562181890010834, 0.1253105103969574, 0.09618951380252838, 0.10229621827602386, -0.08662384748458862, 0.20117226243019104, -0.05531727522611618, 0.033217016607522964, 0.15262672305107117, -0.1583862155675888, 0.12446334213018417, -0.041024547070264816, 0.2651278078556061, -0.7464026808738708, 0.29261302947998047, 0.28785309195518494, 0.14978215098381042, -0.21726630628108978, 0.11766485124826431, -0.013434577733278275, -0.05370185524225235, 0.31687065958976746, 0.03299375995993614, 0.19358164072036743, -0.37752780318260193, -0.036858413368463516, 0.21869459748268127, 0.001962229609489441, 0.1600332111120224, 0.1604834347963333, 0.49909543991088867, -0.18967555463314056, 0.057537224143743515, -0.10264860838651657, -0.14954738318920135, 0.41688254475593567, -0.3948409855365753, 0.1670200079679489, -0.061144664883613586, 0.8726226091384888, 0.2835284173488617, -0.15029354393482208, -0.2694928050041199, -0.08044777065515518, -0.21827715635299683, 0.08427560329437256, 0.08718162029981613, -0.12014015018939972, 0.07406158000230789, -0.05157200247049332, 0.0068646748550236225, 0.1870853453874588, 0.18445274233818054, 0.24675065279006958, 0.3651895225048065, -0.27606531977653503, 0.31659817695617676, -0.3954794108867645, 0.15575827658176422, 0.17428401112556458, 0.06074550375342369, -0.048489246517419815, -0.1410810500383377, 0.02348649501800537, -0.11504437774419785, 0.2646307647228241, -0.22524157166481018, 0.0511217936873436, -0.18277782201766968, -0.23931053280830383, -0.3419874906539917, -0.050036877393722534, 0.03890341520309448, -0.031736768782138824, -0.12461547553539276, -0.06375423818826675, -0.09631937742233276, -0.12806999683380127, -0.050185468047857285, -0.29143261909484863, 0.39763715863227844, 0.37085506319999695, 0.15045668184757233, -0.07252829521894455, 0.7684538960456848, 0.20320641994476318, 0.0067163254134356976, 0.3011976480484009, 0.20140594244003296, -0.11478094011545181, -0.20222105085849762, -0.03750785440206528, 0.19382244348526, -0.2518008053302765, -0.26196151971817017, -0.42612990736961365, 0.6186969876289368, 0.10939059406518936, -0.47760042548179626, -0.08888913691043854, -0.2265508770942688, -0.22605809569358826, -0.13082464039325714, 0.26198363304138184, 0.4220677614212036, 0.07908167690038681, 0.032228779047727585, -0.3935739994049072, 0.07538677752017975, 0.5604329109191895, -0.42825907468795776, -0.1913510411977768, -0.012695955112576485, 0.18095515668392181, 0.2391919046640396, -0.3084866404533386, -0.7673497200012207, -0.1244048923254013, 0.11889972537755966, 0.16897816956043243, -0.1666523665189743, 0.23522721230983734, 0.13245004415512085, 0.2679491937160492, -0.2229093462228775, 0.5535248517990112, -0.07835781574249268, -0.13849279284477234, -0.41262975335121155, -0.09996604919433594 ]
https://github.com/huggingface/datasets/issues/224
[Feature Request/Help] BLEURT model -> PyTorch
We currently provide a wrapper on the TensorFlow implementation: https://huggingface.co/metrics/bleurt We have long term plans to better handle model-based metrics, but they probably won't be implemented right away @adamwlev it would still be cool to add the BLEURT checkpoints to the transformers repo if you're interested, but that would best be discussed there :) closing for now
Hi, I am interested in porting google research's new BLEURT learned metric to PyTorch (because I wish to do something experimental with language generation and backpropping through BLEURT). I noticed that you guys don't have it yet so I am partly just asking if you plan to add it (@thomwolf said you want to do so on Twitter). I had a go of just like manually using the checkpoint that they publish which includes the weights. It seems like the architecture is exactly aligned with the out-of-the-box BertModel in transformers just with a single linear layer on top of the CLS embedding. I loaded all the weights to the PyTorch model but I am not able to get the same numbers as the BLEURT package's python api. Here is my colab notebook where I tried https://colab.research.google.com/drive/1Bfced531EvQP_CpFvxwxNl25Pj6ptylY?usp=sharing . If you have any pointers on what might be going wrong that would be much appreciated! Thank you muchly!
57
[Feature Request/Help] BLEURT model -> PyTorch Hi, I am interested in porting google research's new BLEURT learned metric to PyTorch (because I wish to do something experimental with language generation and backpropping through BLEURT). I noticed that you guys don't have it yet so I am partly just asking if you plan to add it (@thomwolf said you want to do so on Twitter). I had a go of just like manually using the checkpoint that they publish which includes the weights. It seems like the architecture is exactly aligned with the out-of-the-box BertModel in transformers just with a single linear layer on top of the CLS embedding. I loaded all the weights to the PyTorch model but I am not able to get the same numbers as the BLEURT package's python api. Here is my colab notebook where I tried https://colab.research.google.com/drive/1Bfced531EvQP_CpFvxwxNl25Pj6ptylY?usp=sharing . If you have any pointers on what might be going wrong that would be much appreciated! Thank you muchly! We currently provide a wrapper on the TensorFlow implementation: https://huggingface.co/metrics/bleurt We have long term plans to better handle model-based metrics, but they probably won't be implemented right away @adamwlev it would still be cool to add the BLEURT checkpoints to the transformers repo if you're interested, but that would best be discussed there :) closing for now
[ 0.039603687822818756, -0.2945491373538971, 0.06959482282400131, 0.1643860787153244, 0.27279895544052124, -0.281115859746933, 0.3806675672531128, 0.1402260959148407, 0.14548733830451965, 0.16574127972126007, -0.21055951714515686, 0.29347607493400574, -0.45525041222572327, 0.22415998578071594, 0.06959410011768341, -0.10988398641347885, -0.050977613776922226, 0.03715744987130165, 0.10635921359062195, -0.13577429950237274, -0.10876180976629257, 0.16499051451683044, 0.05934987589716911, -0.15254680812358856, -0.29161548614501953, 0.3176138997077942, 0.01924951560795307, 0.025706706568598747, -0.20944389700889587, -0.057742226868867874, 0.2948087155818939, -0.06041427329182625, 0.2058350145816803, 0.7945739030838013, -0.0001224943553097546, -0.2677384912967682, 0.13256369531154633, -0.025316229090094566, -0.34068188071250916, -0.22112901508808136, 0.38395175337791443, -0.3060535788536072, 0.08232928812503815, -0.17544619739055634, 0.04278844594955444, -0.034065131098032, -0.24324031174182892, -0.04265596717596054, 0.18749302625656128, 0.21665802597999573, 0.04224839061498642, 0.2304355949163437, 0.001225774409249425, 0.11258160322904587, 0.13914106786251068, -0.1055869311094284, -0.10437052696943283, 0.5303353667259216, 0.27584055066108704, -0.062410712242126465, -0.19226355850696564, 0.23816871643066406, 0.07783757895231247, 0.1014229878783226, 0.67167067527771, 0.03906068950891495, -0.055595315992832184, -0.06814554333686829, 0.005765378475189209, 0.325133740901947, -0.11480867117643356, -0.2151494324207306, -0.3067013621330261, 0.08728159219026566, 0.07104223221540451, -0.2673066258430481, -0.217970609664917, 0.1069880798459053, 0.1001448780298233, -0.26901230216026306, -0.41902434825897217, -0.3648310899734497, -0.19585224986076355, 0.07973828911781311, -0.27265140414237976, 0.3813110291957855, 0.009816587902605534, -0.15502581000328064, 0.1310456246137619, 0.14506642520427704, 0.388072669506073, -0.07437730580568314, 0.34713467955589294, 0.27463966608047485, -0.016506530344486237, -0.3021688461303711, 0.26859283447265625, -0.13609075546264648, -0.0050388239324092865, -0.2247709184885025, -0.16654133796691895, 0.20244823396205902, -0.2630820572376251, 0.07759540528059006, -0.042992301285266876, 0.4014991819858551, -0.00024039715935941786, 0.3728131353855133, 0.3213631212711334, 0.40310537815093994, 0.10458318144083023, -0.0196261964738369, -0.07396192848682404, -0.0863809883594513, 0.2585074305534363, 0.19309858977794647, -0.22396430373191833, -0.2521967589855194, -0.06776212900876999, -0.020372889935970306, -0.2209862470626831, -0.2402275651693344, 0.3333805501461029, 0.19345851242542267, -0.4989130198955536, -0.02318115346133709, 0.500805675983429, 0.006021072622388601, -0.1655692756175995, 0.10502753406763077, -0.1418713480234146, -0.12604549527168274, -0.30947232246398926, 0.07590857893228531, 0.25282445549964905, -0.16018180549144745, 0.2812947630882263, -0.07687186449766159, 0.4981268048286438, 0.04750412330031395, 0.303552508354187, 0.37072402238845825, 0.09624874591827393, -0.14412473142147064, -0.20200657844543457, -0.3485706150531769, 0.18558332324028015, 0.05072656646370888, -0.13151124119758606, -0.030972514301538467, -0.11490382254123688, -0.2794417142868042, -0.07502064108848572, 0.05097339302301407, -0.2576737105846405, -0.05764283612370491, 0.02539902739226818, 0.3534557819366455, 0.09787617623806, -0.24719707667827606, 0.0006365206791087985, -0.24852794408798218, -0.16766782104969025, 0.06729394197463989, 0.39564624428749084, -0.18060347437858582, -0.3170183300971985, -0.5069208145141602, 0.1681722104549408, 0.43953683972358704, 0.18799060583114624, 0.16617587208747864, -0.2475460171699524, -0.03294246271252632, 0.255615234375, -0.3426561951637268, 0.273783802986145, 0.11720772832632065, -0.3056721091270447, -0.23019106686115265, 0.08945814520120621, -0.30444779992103577, -0.2189071625471115, 0.08569785207509995, -0.14614678919315338, -0.06445030868053436, 0.275614470243454, 0.26867637038230896, -0.176509827375412, -0.012937691062688828, -0.2883857786655426, -0.2562456429004669, 0.24622762203216553, 0.1788952648639679, 0.376988023519516, -0.037900928407907486, 0.15235064923763275, 0.5387821197509766, 0.06697946041822433, -0.38386011123657227, -0.030025457963347435, -0.15700002014636993, 0.5144119262695312, -0.4600304961204529, 0.38164013624191284, -0.21484945714473724, -0.11701182276010513, 0.174919992685318, -0.3829159140586853, 0.346668004989624, 0.15802419185638428, 0.05973011255264282, -0.08232391625642776, 0.004818162880837917, 0.32041260600090027, -0.04485839605331421, -0.0798354372382164, -0.19225312769412994, -0.2519007623195648, -0.15173889696598053, -0.2314920872449875, -0.2959872782230377, -0.18502207100391388, -0.040659524500370026, 0.22060562670230865, 0.3695570230484009, -0.021372150629758835, -0.14678028225898743, -0.0849674642086029, 0.3680128753185272, 0.1280672252178192, 0.16879628598690033, -0.03653458133339882, -0.20773832499980927, 0.16967558860778809, -0.18458376824855804, 0.25811630487442017, 0.27523818612098694, 0.4071276783943176, 0.026199225336313248, 0.28724709153175354, 0.2546691298484802, -0.10023169964551926, -0.10496433079242706, -0.20701256394386292, 0.29886436462402344, 0.30754226446151733, 0.23850306868553162, 0.2911818027496338, 0.09369325637817383, -0.2713991105556488, 0.09760583192110062, -0.06041184812784195, 0.17113199830055237, 0.39449194073677063, -0.13852529227733612, -0.08198616653680801, -0.16408947110176086, 0.14725767076015472, 0.14154811203479767, 0.36072519421577454, -0.05262818560004234, 0.01779124140739441, 0.23726047575473785, -0.4145682752132416, -0.28494346141815186, -0.08580286800861359, 0.05542213097214699, 0.1426684409379959, 0.06415930390357971, 0.41493237018585205, 0.05140382796525955, -0.018801694735884666, 0.04474780336022377, -0.12116041779518127, 0.15478195250034332, -0.2346343845129013, -0.04626809060573578, 0.08961518853902817, 0.0987381786108017, 0.09015540778636932, 0.02513960562646389, 0.07456024736166, 0.11431242525577545, -0.14357633888721466, 0.3518407940864563, 0.13265931606292725, 0.4257064461708069, -0.23266634345054626, -0.3643445074558258, -0.25728604197502136, -0.2122785449028015, 0.03864797204732895, 0.30082911252975464, 0.07326618582010269, -0.08552351593971252, 0.2842048704624176, 0.5591581463813782, -0.049217186868190765, -0.10149654746055603, -0.06357868760824203, -0.2979995608329773, -0.2619437575340271, 0.013049578294157982, -0.047072287648916245, 0.021043088287115097, -0.012132041156291962, 0.1273242086172104, -0.07290884852409363, -0.16482117772102356, -0.7918716669082642, -0.007139210589230061, 0.06195380538702011, 0.29495614767074585, 0.12773050367832184, -0.13784657418727875, -0.32705438137054443, -0.0012700125807896256, 0.017259567975997925, -0.408931165933609, -0.02308211289346218, 0.33340123295783997, -0.24032610654830933, 0.020099246874451637, -0.033198025077581406, -0.19501246511936188, -0.1217612773180008, -0.03714706376194954, 0.035058025270700455, 0.1774720847606659, 0.03728392347693443, -0.127597376704216, 0.11328990012407303, 0.3214528560638428, 0.42720866203308105, -0.15338580310344696, -0.03843601047992706, 0.17423078417778015, 0.4179566204547882, -0.28755009174346924, -0.3364487290382385, -0.11314144730567932, -0.042941126972436905, 0.09721902012825012, 0.020036805421113968, -0.33385977149009705, -0.3707350492477417, -0.13501016795635223, -0.15471291542053223, 0.16762395203113556, 0.21132682263851166, 0.08064763247966766, 0.13571633398532867, 0.08123902231454849, -0.1643408089876175, -0.11949481815099716, -0.05120996758341789, 0.1652880311012268, 0.6772114038467407, -0.2268800139427185, -0.13615913689136505, 0.09984993934631348, 0.8651852011680603, 0.14755184948444366, -0.4560469388961792, 0.03625066578388214, 0.317102313041687, 0.025822997093200684, -0.030955852940678596, -0.2910692095756531, 0.18813341856002808, -0.2465851604938507, -0.28432103991508484, 0.28277817368507385, 0.10243114829063416, -0.2974281907081604, -0.30319005250930786, -0.20264524221420288, 0.5068922638893127, -0.12234535813331604, 0.07363094389438629, 0.11307591199874878, 0.069334477186203, -0.004519649315625429, -0.06540888547897339, -0.2938373386859894, -0.12536659836769104, 0.2407931536436081, -0.0076101720333099365, 0.09480417519807816, -0.03745785728096962, 0.44763731956481934, 0.0032587493769824505, -0.5484263300895691, 0.3487663269042969, 0.052934758365154266, -0.5662445425987244, 0.1476328819990158, -0.07294195890426636, -0.039310675114393234, -0.04653794690966606, -0.32263532280921936, 0.07470045238733292, 0.03616858646273613, -0.14837047457695007, -0.06120492145419121, -0.25290003418922424, 0.09232498705387115, -0.2258288711309433, -0.473783403635025, 0.5314671993255615, 0.3373509347438812, -0.38272055983543396, 0.0932154580950737, 0.22271127998828888, 0.31544363498687744, -0.20287898182868958, -0.014692763797938824, -0.2976086437702179, -0.2209090143442154, 0.0764065682888031, -0.07465646415948868, -0.026433484628796577, 0.22570636868476868, -0.05120127275586128, -0.027302389964461327, 0.12710921466350555, -0.10163131356239319, 0.1812262237071991, 0.07489588856697083, 0.008387451060116291, -0.026714080944657326, 0.15502206981182098, 0.11128564924001694, -0.07816680520772934, 0.19248151779174805, 0.33512261509895325, -0.41935354471206665, -0.5140190124511719, 0.11239608377218246, 0.15022096037864685, 0.173361137509346, 0.09620607644319534, 0.09976094961166382, -0.010332492180168629, 0.40235885977745056, 0.2830176055431366, 0.1623782366514206, 0.2686340808868408, 0.13129739463329315, 0.4590051472187042, 0.06577198207378387, -0.14670079946517944, 0.3369625210762024, -0.18430663645267487, 0.02922503463923931, 0.3064761459827423, -0.09339799731969833, -0.026042673736810684, 0.1751045137643814, 0.4582523703575134, 0.9547620415687561, 0.11388445645570755, 0.07662276178598404, 0.29241102933883667, -0.12001796811819077, 0.3372708559036255, -0.24702514708042145, 0.05289674922823906, -0.3119492828845978, -0.3774302899837494, -0.03756656497716904, -0.0892936959862709, -0.40337929129600525, -0.2799660563468933, -0.20358042418956757, 0.11634510010480881, -0.34473127126693726, -0.2265588641166687, 0.30436331033706665, 0.32432425022125244, 0.07537943869829178, 0.02156975120306015, 0.28700342774391174, 0.06190463900566101, -0.20935139060020447, 0.3146809935569763, -0.07800006121397018, -0.23030881583690643, -0.23614363372325897, 0.10131241381168365, -0.18740463256835938, 0.26554375886917114, -0.05719010904431343, -0.006722102407366037, 0.010055076330900192, -0.31991976499557495, 0.07066038250923157, 0.06933687627315521, 0.7521307468414307, 0.11961162090301514, -0.3720661401748657, 0.14824442565441132, -0.1703070104122162, 0.13719111680984497, 0.13435465097427368, -0.16840821504592896, 0.36602407693862915, -0.14322522282600403, -0.16689880192279816, -0.08256862312555313, 0.2166527360677719, -0.284328430891037, 0.16792264580726624, -0.17392045259475708, -0.11232814937829971, -0.07144258171319962, 0.4080495834350586, 0.32134348154067993, 0.0798187181353569, -0.20158758759498596, 0.09625999629497528, 0.22282849252223969, -0.27824896574020386, 0.08793029934167862, 0.13651283085346222, -0.23687335848808289, -0.0030594279523938894, 0.02106083557009697, -0.045443978160619736, -0.07609265297651291, 0.2099473774433136, 0.18963417410850525, -0.12245622277259827, 0.2027595192193985, -0.5506261587142944, 0.5572553277015686, -0.4292545020580292, 0.11070135980844498, -0.13716751337051392, -0.022318292409181595, 0.03415057808160782, 0.6142624616622925, 0.46294936537742615, 0.17757490277290344, -0.4233265221118927, 0.12258624285459518, -0.4270438253879547, 0.10487641394138336, -0.18003785610198975, 0.3510751724243164, -0.2739531695842743, -0.006761754862964153, 0.006709955167025328, 0.3898065984249115, -0.22701643407344818, -0.02636091411113739, 0.12607716023921967, -0.0717165470123291, -0.021651985123753548, -0.4830065965652466, 0.11179355531930923, 0.04559294134378433, -0.23183292150497437, -0.10764314234256744, 0.08287975192070007, -0.04295343905687332, -0.4671594202518463, 0.15198083221912384, 0.18790121376514435, 0.05123540386557579, -0.2913530170917511, -0.3445456326007843, -0.09238725155591965, -0.18161232769489288, -0.0012822888093069196, -0.035988420248031616, 0.0712982565164566, 0.3768668472766876, 0.16384656727313995, 0.032255060970783234, -0.2014027237892151, -0.058261677622795105, 0.38416898250579834, 0.11524153500795364, 0.062210772186517715, -0.08224421739578247, 0.391283243894577, -0.023472551256418228, 0.026852861046791077, 0.05031924694776535, 0.027711622416973114, 0.08725422620773315, 0.4256473481655121, 0.0977623388171196, 0.1573861986398697, -0.41123902797698975, 0.2807864546775818, 0.17748767137527466, 0.11828775703907013, -0.1377708464860916, 0.18009938299655914, 0.10431331396102905, -0.22499226033687592, -0.23325417935848236, -0.19721581041812897, 0.12463871389627457, -0.1126861721277237, 0.16435369849205017, -0.1148824617266655, 0.35515153408050537, 0.3632740080356598, -0.009765957482159138, -0.24801579117774963, -0.30713972449302673, 0.04222560673952103, 0.054887983947992325, -0.18486233055591583, 0.27694791555404663, 0.4547139108181, 0.06889302283525467, 0.10064522922039032, -0.4465751051902771, -0.09266699850559235, 0.08696885406970978, -0.38432368636131287, -0.0882023498415947, 0.34193098545074463, 0.19567836821079254, -0.15060929954051971, -0.20592087507247925, 0.1579788625240326, -0.11409661918878555, -0.13242574036121368, 0.3602489233016968, -0.2412531077861786, -0.14182522892951965, -0.05407173931598663, 0.32524821162223816, 0.02627568691968918, 0.07756449282169342, 0.05787629634141922, -0.10660640895366669, 0.06521860510110855, -0.006445768754929304, 0.10747730731964111, 0.06735896319150925, -0.01296466775238514, 0.1285707950592041, -0.11017922312021255, 0.20703832805156708, -0.6235055923461914, 0.3515489995479584, 0.3406660854816437, 0.04048692807555199, -0.23865248262882233, 0.13110820949077606, -0.06537609547376633, -0.26793140172958374, 0.3332059681415558, -0.07992029935121536, 0.20084725320339203, -0.3434510827064514, 0.0821337103843689, 0.28266283869743347, -0.054147426038980484, 0.06991565972566605, 0.2067144364118576, 0.40170302987098694, -0.08867084234952927, -0.05176202580332756, -0.059803254902362823, -0.17698349058628082, 0.386995792388916, -0.44275879859924316, 0.1662399023771286, -0.09579470008611679, 0.957119345664978, 0.34746190905570984, -0.24923880398273468, -0.2494019716978073, -0.1470511257648468, -0.34628573060035706, 0.10768326371908188, 0.17763465642929077, -0.0383404865860939, -0.003949153237044811, -0.06391032040119171, 0.002883369568735361, 0.05237790569663048, 0.15503762662410736, 0.2236464023590088, 0.5317481756210327, -0.35456860065460205, 0.29938849806785583, -0.22832456231117249, 0.10329020023345947, 0.2751064598560333, 0.13227225840091705, -0.12725012004375458, -0.017683451995253563, -0.009351537562906742, -0.04927100986242294, 0.026652870699763298, -0.32910633087158203, 0.0641792044043541, -0.14669622480869293, -0.2652188837528229, -0.21753916144371033, -0.04903123155236244, 0.03659207746386528, 0.04084506258368492, -0.06513180583715439, 0.032620519399642944, -0.12107454985380173, -0.10738180577754974, -0.05225258693099022, -0.31600624322891235, 0.3485419452190399, 0.3667267858982086, 0.08986850082874298, -0.02784149907529354, 0.6427874565124512, 0.15500769019126892, -0.1114492267370224, 0.20141595602035522, 0.1498376429080963, -0.11773566156625748, -0.23597478866577148, 0.022950435057282448, 0.30525508522987366, -0.11438208818435669, -0.34389615058898926, -0.4947918951511383, 0.6355995535850525, 0.1798558235168457, -0.5693950057029724, -0.17015694081783295, -0.2464664876461029, -0.1722552329301834, -0.11563298851251602, 0.3047386407852173, 0.3367748260498047, 0.1070006936788559, -0.017187241464853287, -0.367851197719574, 0.06726916134357452, 0.4848317503929138, -0.45635780692100525, -0.07332684844732285, -0.11960632354021072, 0.16673564910888672, 0.2647438049316406, -0.19481337070465088, -0.9240177273750305, -0.1374066323041916, 0.14410839974880219, -0.014072578400373459, -0.17872865498065948, 0.29899731278419495, 0.2460012286901474, 0.29371097683906555, -0.2554593086242676, 0.5315845012664795, -0.04953821375966072, -0.09967701882123947, -0.4590684473514557, -0.21204893290996552 ]
https://github.com/huggingface/datasets/issues/224
[Feature Request/Help] BLEURT model -> PyTorch
Hi there. We ran into the same problem this year (converting BLEURT to PyTorch) and thanks to @adamwlev found his colab notebook which didn't work but served as a good starting point. Finally, we **made it work** by doing just two simple conceptual fixes: 1. Transposing 'kernel' layers instead of 'dense' ones when copying params from the original model; 2. Taking pooler_output as a cls_state in forward function of the BleurtModel class. Plus few minor syntactical fixes for the outdated parts. The result is still not exactly the same, but is very close to the expected one (1.0483 vs 1.0474). Find the fixed version here (fixes are commented): https://colab.research.google.com/drive/1KsCUkFW45d5_ROSv2aHtXgeBa2Z98r03?usp=sharing
Hi, I am interested in porting google research's new BLEURT learned metric to PyTorch (because I wish to do something experimental with language generation and backpropping through BLEURT). I noticed that you guys don't have it yet so I am partly just asking if you plan to add it (@thomwolf said you want to do so on Twitter). I had a go of just like manually using the checkpoint that they publish which includes the weights. It seems like the architecture is exactly aligned with the out-of-the-box BertModel in transformers just with a single linear layer on top of the CLS embedding. I loaded all the weights to the PyTorch model but I am not able to get the same numbers as the BLEURT package's python api. Here is my colab notebook where I tried https://colab.research.google.com/drive/1Bfced531EvQP_CpFvxwxNl25Pj6ptylY?usp=sharing . If you have any pointers on what might be going wrong that would be much appreciated! Thank you muchly!
109
[Feature Request/Help] BLEURT model -> PyTorch Hi, I am interested in porting google research's new BLEURT learned metric to PyTorch (because I wish to do something experimental with language generation and backpropping through BLEURT). I noticed that you guys don't have it yet so I am partly just asking if you plan to add it (@thomwolf said you want to do so on Twitter). I had a go of just like manually using the checkpoint that they publish which includes the weights. It seems like the architecture is exactly aligned with the out-of-the-box BertModel in transformers just with a single linear layer on top of the CLS embedding. I loaded all the weights to the PyTorch model but I am not able to get the same numbers as the BLEURT package's python api. Here is my colab notebook where I tried https://colab.research.google.com/drive/1Bfced531EvQP_CpFvxwxNl25Pj6ptylY?usp=sharing . If you have any pointers on what might be going wrong that would be much appreciated! Thank you muchly! Hi there. We ran into the same problem this year (converting BLEURT to PyTorch) and thanks to @adamwlev found his colab notebook which didn't work but served as a good starting point. Finally, we **made it work** by doing just two simple conceptual fixes: 1. Transposing 'kernel' layers instead of 'dense' ones when copying params from the original model; 2. Taking pooler_output as a cls_state in forward function of the BleurtModel class. Plus few minor syntactical fixes for the outdated parts. The result is still not exactly the same, but is very close to the expected one (1.0483 vs 1.0474). Find the fixed version here (fixes are commented): https://colab.research.google.com/drive/1KsCUkFW45d5_ROSv2aHtXgeBa2Z98r03?usp=sharing
[ 0.031052321195602417, -0.210395947098732, 0.11031059920787811, 0.2494821548461914, 0.2850295305252075, -0.33688655495643616, 0.2980252504348755, 0.14864562451839447, 0.09189027547836304, 0.16699321568012238, -0.30923911929130554, 0.3939807415008545, -0.4846566319465637, 0.2546461224555969, 0.0468798466026783, -0.08954044431447983, -0.06799984723329544, 0.05894608795642853, 0.10282053053379059, -0.1792570799589157, -0.11254629492759705, 0.15591686964035034, 0.11106353998184204, -0.10564228892326355, -0.1948649138212204, 0.3722066283226013, -0.06696424633264542, 0.11051246523857117, -0.23483715951442719, -0.05581812188029289, 0.23994246125221252, -0.09313458204269409, 0.19716285169124603, 0.7407878637313843, -0.00012301238893996924, -0.2967880070209503, 0.10562025010585785, -0.013817878440022469, -0.3956652581691742, -0.22906675934791565, 0.39341676235198975, -0.3064045011997223, 0.1475396454334259, -0.1882179230451584, -0.023236092180013657, -0.046829745173454285, -0.2829227149486542, -0.11125563085079193, 0.13091984391212463, 0.24891793727874756, 0.03568284213542938, 0.12069782614707947, -0.040036510676145554, 0.1071796640753746, 0.17364279925823212, -0.19217023253440857, -0.025240423157811165, 0.5683518052101135, 0.3351757526397705, -0.019422370940446854, -0.2249428927898407, 0.17774081230163574, 0.011493615806102753, 0.10073024034500122, 0.6617962718009949, 0.040606625378131866, -0.050766535103321075, -0.04632411152124405, 0.0426529236137867, 0.3301610052585602, -0.1137453019618988, -0.19089460372924805, -0.25507697463035583, 0.12239739298820496, 0.07248994708061218, -0.24157017469406128, -0.23092706501483917, 0.1607206016778946, 0.10856480151414871, -0.24158084392547607, -0.3910423219203949, -0.3491125702857971, -0.14091525971889496, 0.13794539868831635, -0.29824990034103394, 0.30928897857666016, -0.06806906312704086, -0.12711910903453827, 0.1534782499074936, 0.0853557288646698, 0.34188586473464966, -0.09555454552173615, 0.35434094071388245, 0.197038471698761, 0.0863436758518219, -0.18238526582717896, 0.2225731611251831, 0.04210389405488968, -0.06244329363107681, -0.30931419134140015, -0.12535369396209717, 0.20691783726215363, -0.28140294551849365, 0.011796685867011547, -0.027306264266371727, 0.32033535838127136, 0.04041891172528267, 0.3597738742828369, 0.28627604246139526, 0.37072038650512695, 0.07752560079097748, -0.04536338523030281, -0.08907069265842438, -0.1548776626586914, 0.24097353219985962, 0.19222790002822876, -0.2800590693950653, -0.3287023603916168, -0.03376986086368561, -0.05419690161943436, -0.21769686043262482, -0.10590724647045135, 0.3486950099468231, 0.13636167347431183, -0.49230241775512695, -0.06439258903265, 0.6397172808647156, 0.04687034338712692, -0.18193763494491577, 0.07984860241413116, -0.14149200916290283, -0.09508480876684189, -0.3073863685131073, 0.06758333742618561, 0.21250446140766144, -0.19385316967964172, 0.2788226306438446, 0.03200077638030052, 0.5187747478485107, 0.12528865039348602, 0.16171297430992126, 0.433599591255188, 0.1615477353334427, -0.19241005182266235, -0.2416660189628601, -0.2961932420730591, 0.22628918290138245, 0.026455754414200783, -0.11240141093730927, 0.08938439935445786, -0.12872271239757538, -0.3263464868068695, -0.15209659934043884, 0.023960167542099953, -0.28712764382362366, -0.01919068768620491, -0.010953431949019432, 0.27803850173950195, 0.10987827181816101, -0.24877294898033142, -0.032197196036577225, -0.3951019048690796, -0.12442508339881897, 0.09461572021245956, 0.40959542989730835, -0.20389001071453094, -0.4032553732395172, -0.427376925945282, 0.1044914498925209, 0.4528108537197113, 0.3346205949783325, 0.09447552263736725, -0.2226802259683609, -0.20602363348007202, 0.1822289079427719, -0.3569924831390381, 0.24966193735599518, 0.19206252694129944, -0.27465394139289856, -0.2299157977104187, 0.13340447843074799, -0.2530921995639801, -0.21538260579109192, 0.025945203378796577, -0.15083996951580048, -0.06364273279905319, 0.3645898997783661, 0.11044348031282425, -0.18229316174983978, -0.023145990446209908, -0.37692394852638245, -0.1707116812467575, 0.24087874591350555, 0.14524509012699127, 0.311216801404953, 0.04421854764223099, 0.13978594541549683, 0.5452297925949097, 0.1094927117228508, -0.37678155303001404, 0.03864974156022072, -0.20870788395404816, 0.4394530653953552, -0.38880041241645813, 0.36291447281837463, -0.1322948932647705, -0.043620940297842026, 0.22862307727336884, -0.3252674341201782, 0.44982925057411194, 0.1517198532819748, 0.13739261031150818, -0.11641106754541397, -0.11465409398078918, 0.285736620426178, -0.041558247059583664, -0.09858439117670059, -0.17538607120513916, -0.2840970754623413, -0.22045330703258514, -0.14794200658798218, -0.25469720363616943, -0.23914504051208496, -0.0517391674220562, 0.19210873544216156, 0.326083242893219, -0.06349676847457886, -0.16392351686954498, -0.038680560886859894, 0.3862401843070984, 0.18508075177669525, 0.21768563985824585, -0.07578461617231369, -0.22854949533939362, 0.11393800377845764, -0.2082347273826599, 0.1406581550836563, 0.19605635106563568, 0.38518384099006653, 0.03162413835525513, 0.40312567353248596, 0.2144811451435089, -0.11146383732557297, -0.0912846252322197, -0.1220182403922081, 0.281228244304657, 0.3246676027774811, 0.26691174507141113, 0.27549681067466736, 0.12106359004974365, -0.24934859573841095, 0.10694184899330139, -0.04537427797913551, 0.10155370831489563, 0.35155993700027466, -0.16831788420677185, 0.06462013721466064, -0.16824601590633392, 0.16156993806362152, 0.29566434025764465, 0.25689876079559326, -0.05976742506027222, -0.04502660036087036, 0.2192205786705017, -0.4693129062652588, -0.2722737193107605, 0.011139390990138054, 0.07040068507194519, 0.08245235681533813, 0.07317251712083817, 0.4600593149662018, 0.1447228491306305, -0.016832523047924042, 0.027731165289878845, -0.08601713180541992, 0.14787453413009644, -0.11266202479600906, -0.0631483644247055, 0.12503015995025635, 0.06865395605564117, 0.07814257591962814, 0.10393846035003662, 0.0552339144051075, 0.14249446988105774, -0.18085716664791107, 0.3385641872882843, -0.0062665375880897045, 0.4345986247062683, -0.24935321509838104, -0.3888389468193054, -0.22619493305683136, -0.24832548201084137, -0.0348907895386219, 0.32378777861595154, -0.013398610055446625, -0.13628646731376648, 0.19924983382225037, 0.40451472997665405, -0.08564754575490952, -0.05584089457988739, -0.11347301304340363, -0.26379653811454773, -0.36608439683914185, -0.0012672677403315902, -0.06414478272199631, 0.06512652337551117, -0.08705322444438934, 0.0358789823949337, -0.1546071171760559, -0.19300711154937744, -0.7795087695121765, 0.06021299958229065, 0.03927171975374222, 0.3000189960002899, 0.20597082376480103, -0.13490086793899536, -0.3109205365180969, 0.04444722831249237, 0.10181453078985214, -0.37420234084129333, -0.026990210637450218, 0.4176819622516632, -0.23497271537780762, 0.07670922577381134, -0.03322916105389595, -0.2210606187582016, -0.12886711955070496, -0.022567683830857277, -0.055178869515657425, 0.12109819054603577, 0.06527827680110931, -0.20432546734809875, 0.1393243372440338, 0.3219202160835266, 0.41521912813186646, -0.09877268224954605, -0.06713975220918655, 0.22492286562919617, 0.42780089378356934, -0.2569456994533539, -0.30135196447372437, -0.1849202960729599, -0.0005778985796496272, 0.01725437119603157, 0.00037719690590165555, -0.3237830102443695, -0.3601541519165039, -0.11460626125335693, -0.1512562334537506, 0.2021995335817337, 0.16915777325630188, 0.06157870590686798, 0.22429554164409637, 0.06467705219984055, -0.21677853167057037, -0.16279307007789612, -0.07023095339536667, 0.2430409938097, 0.6987826824188232, -0.23153941333293915, -0.09995323419570923, 0.08970408141613007, 0.7524693012237549, 0.13057473301887512, -0.42318663001060486, 0.06254071742296219, 0.26073357462882996, 0.06086154282093048, -0.1139405220746994, -0.2671453654766083, 0.15565015375614166, -0.31454017758369446, -0.27640223503112793, 0.3156626522541046, 0.11460095643997192, -0.2804214358329773, -0.26112988591194153, -0.20049650967121124, 0.4979616701602936, -0.17563240230083466, 0.06963695585727692, 0.03468484431505203, -0.049977947026491165, 0.02282228320837021, -0.11708009988069534, -0.27406203746795654, -0.1145453080534935, 0.20600132644176483, 0.022086406126618385, 0.09041740745306015, 0.03668021038174629, 0.4527599513530731, 0.015574460849165916, -0.7020943760871887, 0.4291401505470276, 0.09122661501169205, -0.6415274143218994, 0.20079024136066437, -0.15091952681541443, -0.0830678790807724, -0.05323386937379837, -0.3633192777633667, 0.1148812398314476, 0.03658551722764969, -0.06340611726045609, -0.024902384728193283, -0.28089115023612976, 0.019936570897698402, -0.26069679856300354, -0.31099650263786316, 0.6108303070068359, 0.3093995153903961, -0.3783533275127411, 0.051323823630809784, 0.23827768862247467, 0.44776102900505066, -0.16089589893817902, 0.035822778940200806, -0.35688716173171997, -0.21575260162353516, 0.09272798895835876, -0.11704522371292114, -0.03972245380282402, 0.2884732186794281, -0.12854865193367004, -0.07872021198272705, 0.24400511384010315, -0.08774770051240921, 0.1997196227312088, 0.04970068112015724, 0.10960323363542557, -0.09482958167791367, 0.17061850428581238, 0.05049712210893631, 0.06471368670463562, 0.2465960830450058, 0.41444987058639526, -0.5077797770500183, -0.47135013341903687, 0.21681654453277588, 0.16121618449687958, 0.17137634754180908, 0.10935606062412262, 0.10491982847452164, -0.13601084053516388, 0.4354987144470215, 0.27228695154190063, 0.18568502366542816, 0.23380282521247864, 0.03917381912469864, 0.36824846267700195, 0.11878345906734467, -0.19488820433616638, 0.392410010099411, -0.1706746369600296, 0.014421051368117332, 0.22706560790538788, 0.0664922297000885, -0.012271588668227196, 0.12193950265645981, 0.433799684047699, 0.9371414184570312, 0.14222003519535065, 0.09444385766983032, 0.29919978976249695, -0.13560324907302856, 0.3273807764053345, -0.04595360532402992, 0.028853828087449074, -0.3419894278049469, -0.26656433939933777, 0.04363759979605675, -0.04958946630358696, -0.4452959895133972, -0.16226650774478912, -0.19859610497951508, 0.14732925593852997, -0.3593909740447998, -0.2568361461162567, 0.2827571630477905, 0.28083696961402893, 0.10884791612625122, 0.03093196451663971, 0.27702635526657104, 0.08055379241704941, -0.1751106232404709, 0.26924410462379456, -0.08487474918365479, -0.1843997985124588, -0.2621048390865326, 0.13207845389842987, -0.18024882674217224, 0.3719131648540497, -0.1005956381559372, -0.044855352491140366, -0.0609016977250576, -0.34924888610839844, 0.07375746220350266, 0.032762810587882996, 0.7054664492607117, 0.09231635928153992, -0.30709201097488403, 0.08027929812669754, -0.062138140201568604, 0.17744220793247223, 0.19616711139678955, -0.14955559372901917, 0.35956010222435, -0.13380089402198792, -0.16223973035812378, -0.1679966151714325, 0.14697028696537018, -0.374942421913147, 0.2212458848953247, -0.16459938883781433, -0.045548468828201294, -0.11824917793273926, 0.403385728597641, 0.3345358073711395, 0.15933948755264282, -0.2553207576274872, 0.07714977860450745, 0.20328013598918915, -0.2602018713951111, 0.04581964761018753, 0.11606188863515854, -0.2364220768213272, 0.0402359813451767, -0.009904541075229645, 0.0215242151170969, -0.0013386906357482076, 0.23494325578212738, 0.17420607805252075, -0.11456732451915741, 0.2359016239643097, -0.6396331787109375, 0.5488190650939941, -0.3684135973453522, 0.17571543157100677, -0.13607390224933624, -0.055178482085466385, 0.12224484235048294, 0.5733034610748291, 0.4357832372188568, 0.21459347009658813, -0.3986988365650177, 0.18546845018863678, -0.38894757628440857, 0.08773009479045868, -0.1864100843667984, 0.36033186316490173, -0.17637784779071808, 0.002971276640892029, -0.1169334277510643, 0.23120097815990448, -0.20673426985740662, 0.03993750363588333, 0.19829219579696655, -0.030683159828186035, 0.009558257646858692, -0.4418540298938751, 0.027486858889460564, 0.03583341836929321, -0.23903535306453705, -0.06395769864320755, 0.11508987843990326, -0.022896211594343185, -0.4367917478084564, 0.1588522344827652, 0.23203124105930328, 0.009141032584011555, -0.3205920159816742, -0.3893786370754242, -0.13044008612632751, -0.1896279901266098, 0.06611356139183044, 0.027370635420084, 0.06494433432817459, 0.2162601351737976, 0.2079443484544754, 0.02030543051660061, -0.1601991057395935, -0.0488763302564621, 0.4102197289466858, 0.1664804369211197, 0.009113945998251438, -0.03434094041585922, 0.4261152148246765, -0.028062203899025917, 0.19161005318164825, 0.024117888882756233, -0.025750068947672844, 0.16407063603401184, 0.4158232510089874, 0.1315034031867981, 0.15301229059696198, -0.39151957631111145, 0.31367576122283936, 0.25202053785324097, 0.20041392743587494, -0.20123513042926788, 0.2113131284713745, 0.08532010018825531, -0.2471274733543396, -0.1780572086572647, -0.20906932651996613, 0.05017305538058281, -0.13712023198604584, 0.17602647840976715, -0.11124162375926971, 0.36048200726509094, 0.3827468752861023, -0.061240170150995255, -0.2361864596605301, -0.2675482928752899, 0.05084855481982231, 0.044831693172454834, -0.21709851920604706, 0.3127981722354889, 0.5844197273254395, 0.04400474950671196, 0.0522158220410347, -0.38256508111953735, -0.11984249949455261, 0.008464219979941845, -0.2900800406932831, -0.108846016228199, 0.3588646650314331, 0.20460079610347748, -0.10920122265815735, -0.09902559965848923, 0.15128213167190552, -0.07170280814170837, -0.12134362757205963, 0.3947110176086426, -0.4014695882797241, -0.16226020455360413, -0.03691441938281059, 0.35680654644966125, -0.026629585772752762, 0.06826424598693848, 0.10880744457244873, -0.21613892912864685, 0.07733331620693207, -0.015025531873106956, -0.018275009468197823, 0.09222159534692764, 0.07474453747272491, 0.10134053230285645, -0.09447559714317322, 0.15066193044185638, -0.7600210309028625, 0.350068062543869, 0.2666800618171692, 0.07755817472934723, -0.19197683036327362, 0.15109218657016754, -0.11532808840274811, -0.2361515462398529, 0.3830999732017517, -0.016696663573384285, 0.25774872303009033, -0.32524359226226807, -0.04142824187874794, 0.40660130977630615, -0.02356482483446598, 0.043433256447315216, 0.11910467594861984, 0.45571115612983704, -0.13713513314723969, -0.08016843348741531, -0.08494994044303894, -0.171061173081398, 0.33477291464805603, -0.39211305975914, 0.22192439436912537, 0.0573049820959568, 0.8864086270332336, 0.33054134249687195, -0.20438863337039948, -0.2726922631263733, -0.09083352237939835, -0.29800716042518616, 0.04675344377756119, 0.10341760516166687, -0.08217316120862961, 0.027585677802562714, -0.027783647179603577, 0.006058313883841038, 0.12412019073963165, 0.14217644929885864, 0.26756593585014343, 0.2982499301433563, -0.3599912226200104, 0.2530737519264221, -0.2842271327972412, 0.15552231669425964, 0.18207035958766937, 0.06622276455163956, -0.23234108090400696, -0.007762405555695295, 0.07117771357297897, -0.10512084513902664, 0.05717313662171364, -0.23002105951309204, 0.07925614714622498, -0.21019116044044495, -0.2341051697731018, -0.2609878182411194, 0.06782612204551697, -0.0144274290651083, 0.004517834167927504, -0.08634735643863678, 0.023151902481913567, -0.11450551450252533, -0.09457802027463913, 0.021241702139377594, -0.28783363103866577, 0.37281298637390137, 0.3003579080104828, 0.12165672332048416, 0.057156745344400406, 0.5850401520729065, 0.12948991358280182, -0.17101722955703735, 0.2642612159252167, 0.1361713409423828, -0.10343613475561142, -0.1814618855714798, -0.04337970167398453, 0.26375091075897217, -0.050897132605314255, -0.26359444856643677, -0.5351043939590454, 0.5506895184516907, 0.11400321125984192, -0.5117970108985901, -0.22895289957523346, -0.16178414225578308, -0.18558932840824127, -0.10991483926773071, 0.24824410676956177, 0.38117265701293945, 0.1388428509235382, -0.028627056628465652, -0.4515787959098816, 0.004064623266458511, 0.46507027745246887, -0.4754208028316498, -0.12336935102939606, -0.12123037129640579, 0.1840689480304718, 0.29907095432281494, -0.12919843196868896, -0.829527735710144, -0.1543300449848175, 0.09583055227994919, -0.0380876250565052, -0.13884186744689941, 0.23325426876544952, 0.11724211275577545, 0.2540245056152344, -0.29439646005630493, 0.5520857572555542, 0.04639892652630806, -0.1617400199174881, -0.47914403676986694, -0.19192816317081451 ]
https://github.com/huggingface/datasets/issues/223
[Feature request] Add FLUE dataset
@mariamabarham I put all the datasets on this drive: https://1drv.ms/u/s!Ao2Rcpiny7RFinDypq7w-LbXcsx9?e=iVsEDh Some information : • For FLUE, the quote used is > @misc{le2019flaubert, > title={FlauBERT: Unsupervised Language Model Pre-training for French}, > author={Hang Le and Loïc Vial and Jibril Frej and Vincent Segonne and Maximin Coavoux and Benjamin Lecouteux and Alexandre Allauzen and Benoît Crabbé and Laurent Besacier and Didier Schwab}, > year={2019}, > eprint={1912.05372}, > archivePrefix={arXiv}, > primaryClass={cs.CL} > } • The Github repo of FLUE is avaible here : https://github.com/getalp/Flaubert/tree/master/flue Information related to the different tasks of FLUE : **1. Classification** Three dataframes are available: - Book - DVD - Music For each of these dataframes is available a set of training and test data, and a third one containing unlabelled data. Citation : >@dataset{prettenhofer_peter_2010_3251672, author = {Prettenhofer, Peter and Stein, Benno}, title = {{Webis Cross-Lingual Sentiment Dataset 2010 (Webis- CLS-10)}}, month = jul, year = 2010, publisher = {Zenodo}, doi = {10.5281/zenodo.3251672}, url = {https://doi.org/10.5281/zenodo.3251672} } **2. Paraphrasing** French part of the PAWS-X dataset (https://github.com/google-research-datasets/paws). Three dataframes are available: - train - dev - test Citation : > @InProceedings{pawsx2019emnlp, > title = {{PAWS-X: A Cross-lingual Adversarial Dataset for Paraphrase Identification}}, > author = {Yang, Yinfei and Zhang, Yuan and Tar, Chris and Baldridge, Jason}, > booktitle = {Proc. of EMNLP}, > year = {2019} > } **3. Natural Language Inference** French part of the XNLI dataset (https://github.com/facebookresearch/XNLI). Three dataframes are available: - train - dev - test For the dev and test datasets, extra columns compared to the train dataset were available so I left them in the dataframe (I didn't know if these columns could be useful for other tasks or not). In the context of the FLUE benchmark, only the columns gold_label, sentence1 and sentence2 are useful. Citation : > @InProceedings{conneau2018xnli, > author = "Conneau, Alexis > and Rinott, Ruty > and Lample, Guillaume > and Williams, Adina > and Bowman, Samuel R. > and Schwenk, Holger > and Stoyanov, Veselin", > title = "XNLI: Evaluating Cross-lingual Sentence Representations", > booktitle = "Proceedings of the 2018 Conference on Empirical Methods > in Natural Language Processing", > year = "2018", > publisher = "Association for Computational Linguistics", > location = "Brussels, Belgium", **4. Parsing** The dataset used by the FLUE authors for this task is not freely available. Users of your library will therefore not be able to access it. Nevertheless, I think maybe it is useful to add a link to the site where to request this dataframe: http://ftb.linguist.univ-paris-diderot.fr/telecharger.php?langue=en (personally it was sent to me less than 48 hours after I requested it). **5. Word Sense Disambiguation Tasks** 5.1 Verb Sense Disambiguation Two dataframes are available: train and test For both dataframes, 4 columns are available: document, sentence, lemma and word. I created the document column thinking that there were several documents in the dataset but afterwards it turns out that there were not: several sentences but only one document. It's up to you to keep it or not when importing these two dataframes. The sentence column is used to determine to which sentence the word in the word column belongs. It is in the form of a dictionary {'id': 'd000.s001', 'idx': '1'}. I thought for a while to keep only the idx because the id doesn't matter any more information. Nevertheless for the test dataset, the dictionary has an extra value indicating the source of the sentence. I don't know if it's useful or not, that's why I left the dictionary just in case. The user is free to do what he wants with it. Citation : > Segonne, V., Candito, M., and Crabb ́e, B. (2019). Usingwiktionary as a resource for wsd: the case of frenchverbs. InProceedings of the 13th International Confer-ence on Computational Semantics-Long Papers, pages259–270 5.2 Noun Sense Disambiguation Two dataframes are available: 2 train and 1 test I confess I didn't fully understand the procedure for this task. Citation : > @dataset{loic_vial_2019_3549806, > author = {Loïc Vial}, > title = {{French Word Sense Disambiguation with Princeton > WordNet Identifiers}}, > month = nov, > year = 2019, > publisher = {Zenodo}, > version = {1.0}, > doi = {10.5281/zenodo.3549806}, > url = {https://doi.org/10.5281/zenodo.3549806} > } Finally, additional information about FLUE is available in the FlauBERT publication : https://arxiv.org/abs/1912.05372 (p. 4). Hoping to have provided you with everything you need to add this benchmark :)
Hi, I think it would be interesting to add the FLUE dataset for francophones or anyone wishing to work on French. In other requests, I read that you are already working on some datasets, and I was wondering if FLUE was planned. If it is not the case, I can provide each of the cleaned FLUE datasets (in the form of a directly exploitable dataset rather than in the original xml formats which require additional processing, with the French part for cases where the dataset is based on a multilingual dataframe, etc.).
723
[Feature request] Add FLUE dataset Hi, I think it would be interesting to add the FLUE dataset for francophones or anyone wishing to work on French. In other requests, I read that you are already working on some datasets, and I was wondering if FLUE was planned. If it is not the case, I can provide each of the cleaned FLUE datasets (in the form of a directly exploitable dataset rather than in the original xml formats which require additional processing, with the French part for cases where the dataset is based on a multilingual dataframe, etc.). @mariamabarham I put all the datasets on this drive: https://1drv.ms/u/s!Ao2Rcpiny7RFinDypq7w-LbXcsx9?e=iVsEDh Some information : • For FLUE, the quote used is > @misc{le2019flaubert, > title={FlauBERT: Unsupervised Language Model Pre-training for French}, > author={Hang Le and Loïc Vial and Jibril Frej and Vincent Segonne and Maximin Coavoux and Benjamin Lecouteux and Alexandre Allauzen and Benoît Crabbé and Laurent Besacier and Didier Schwab}, > year={2019}, > eprint={1912.05372}, > archivePrefix={arXiv}, > primaryClass={cs.CL} > } • The Github repo of FLUE is avaible here : https://github.com/getalp/Flaubert/tree/master/flue Information related to the different tasks of FLUE : **1. Classification** Three dataframes are available: - Book - DVD - Music For each of these dataframes is available a set of training and test data, and a third one containing unlabelled data. Citation : >@dataset{prettenhofer_peter_2010_3251672, author = {Prettenhofer, Peter and Stein, Benno}, title = {{Webis Cross-Lingual Sentiment Dataset 2010 (Webis- CLS-10)}}, month = jul, year = 2010, publisher = {Zenodo}, doi = {10.5281/zenodo.3251672}, url = {https://doi.org/10.5281/zenodo.3251672} } **2. Paraphrasing** French part of the PAWS-X dataset (https://github.com/google-research-datasets/paws). Three dataframes are available: - train - dev - test Citation : > @InProceedings{pawsx2019emnlp, > title = {{PAWS-X: A Cross-lingual Adversarial Dataset for Paraphrase Identification}}, > author = {Yang, Yinfei and Zhang, Yuan and Tar, Chris and Baldridge, Jason}, > booktitle = {Proc. of EMNLP}, > year = {2019} > } **3. Natural Language Inference** French part of the XNLI dataset (https://github.com/facebookresearch/XNLI). Three dataframes are available: - train - dev - test For the dev and test datasets, extra columns compared to the train dataset were available so I left them in the dataframe (I didn't know if these columns could be useful for other tasks or not). In the context of the FLUE benchmark, only the columns gold_label, sentence1 and sentence2 are useful. Citation : > @InProceedings{conneau2018xnli, > author = "Conneau, Alexis > and Rinott, Ruty > and Lample, Guillaume > and Williams, Adina > and Bowman, Samuel R. > and Schwenk, Holger > and Stoyanov, Veselin", > title = "XNLI: Evaluating Cross-lingual Sentence Representations", > booktitle = "Proceedings of the 2018 Conference on Empirical Methods > in Natural Language Processing", > year = "2018", > publisher = "Association for Computational Linguistics", > location = "Brussels, Belgium", **4. Parsing** The dataset used by the FLUE authors for this task is not freely available. Users of your library will therefore not be able to access it. Nevertheless, I think maybe it is useful to add a link to the site where to request this dataframe: http://ftb.linguist.univ-paris-diderot.fr/telecharger.php?langue=en (personally it was sent to me less than 48 hours after I requested it). **5. Word Sense Disambiguation Tasks** 5.1 Verb Sense Disambiguation Two dataframes are available: train and test For both dataframes, 4 columns are available: document, sentence, lemma and word. I created the document column thinking that there were several documents in the dataset but afterwards it turns out that there were not: several sentences but only one document. It's up to you to keep it or not when importing these two dataframes. The sentence column is used to determine to which sentence the word in the word column belongs. It is in the form of a dictionary {'id': 'd000.s001', 'idx': '1'}. I thought for a while to keep only the idx because the id doesn't matter any more information. Nevertheless for the test dataset, the dictionary has an extra value indicating the source of the sentence. I don't know if it's useful or not, that's why I left the dictionary just in case. The user is free to do what he wants with it. Citation : > Segonne, V., Candito, M., and Crabb ́e, B. (2019). Usingwiktionary as a resource for wsd: the case of frenchverbs. InProceedings of the 13th International Confer-ence on Computational Semantics-Long Papers, pages259–270 5.2 Noun Sense Disambiguation Two dataframes are available: 2 train and 1 test I confess I didn't fully understand the procedure for this task. Citation : > @dataset{loic_vial_2019_3549806, > author = {Loïc Vial}, > title = {{French Word Sense Disambiguation with Princeton > WordNet Identifiers}}, > month = nov, > year = 2019, > publisher = {Zenodo}, > version = {1.0}, > doi = {10.5281/zenodo.3549806}, > url = {https://doi.org/10.5281/zenodo.3549806} > } Finally, additional information about FLUE is available in the FlauBERT publication : https://arxiv.org/abs/1912.05372 (p. 4). Hoping to have provided you with everything you need to add this benchmark :)
[ -0.09875038266181946, 0.2994914650917053, -0.19066588580608368, 0.08413012325763702, -0.09124937653541565, 0.10564146190881729, 0.4507474899291992, -0.1468384861946106, -0.2177068144083023, 0.004500423092395067, -0.20793592929840088, -0.17496609687805176, -0.15123456716537476, 0.2744063436985016, 0.07923620194196701, -0.3693346083164215, -0.001966479467228055, 0.03353632986545563, 0.08688843250274658, -0.34479328989982605, -0.10768900066614151, 0.17830157279968262, -0.049756765365600586, 0.19358907639980316, -0.00017912268231157213, 0.0912407636642456, -0.22928310930728912, -0.15842866897583008, -0.15099956095218658, -0.022321274504065514, 0.13173715770244598, 0.2434057593345642, 0.07313663512468338, 0.38489067554473877, -0.0001029056147672236, -0.21261043846607208, 0.1268319934606552, -0.12616340816020966, 0.02900724485516548, 0.03222448006272316, 0.022003227844834328, -0.22676856815814972, -0.339968204498291, 0.03778326138854027, -0.1280571073293686, -0.25360530614852905, -0.18752926588058472, -0.5613649487495422, -0.13439111411571503, 0.5067586302757263, 0.24277840554714203, -0.17855919897556305, 0.134841650724411, 0.014460093341767788, 0.1603563278913498, 0.27052637934684753, -0.32320067286491394, -0.1657283902168274, 0.6291846632957458, 0.1499028354883194, 0.03446395695209503, 0.2998955547809601, 0.2246345728635788, -0.3710678517818451, -0.20697060227394104, -0.131522998213768, -0.052246399223804474, -0.3681866526603699, 0.02792334370315075, 0.35658735036849976, 0.6327366232872009, -0.08152041584253311, -0.5868775248527527, -0.09462109953165054, 0.24371430277824402, -0.07554955035448074, -0.10097488760948181, -0.13449408113956451, 0.22803421318531036, 0.19478920102119446, 0.3044317662715912, -0.40359675884246826, -0.06974329799413681, 0.09901520609855652, 0.08143537491559982, 0.009417501278221607, -0.05463426932692528, -0.017630966380238533, 0.11922582983970642, -0.16042017936706543, 0.010838354006409645, 0.16477154195308685, -0.06016891077160835, 0.36071035265922546, -0.2149062603712082, -0.32067176699638367, 0.06903496384620667, -0.05932775139808655, 0.3171031177043915, 0.07181364297866821, -0.6085295677185059, 0.033939849585294724, -0.4754466116428375, -0.13653351366519928, 0.23075523972511292, 0.14107748866081238, 0.1131884902715683, -0.2693191170692444, 0.16951462626457214, -0.34914833307266235, 0.1283833235502243, -0.008936644531786442, -0.006787311285734177, -0.025768866762518883, -0.2819310128688812, 0.06205982342362404, -0.09378901869058609, -0.345057874917984, 0.00953434593975544, 0.028470728546380997, 0.19382193684577942, -0.2102639228105545, -0.10387522727251053, 0.19578363001346588, 0.0845956951379776, 0.31471458077430725, -0.18962477147579193, 0.1951976716518402, -0.18190455436706543, -0.6139771342277527, -0.15016688406467438, 0.04668831825256348, -0.25893235206604004, 0.033654000610113144, 0.11798012256622314, -0.30589014291763306, 0.045036595314741135, 0.10600323230028152, -0.035521797835826874, 0.2437460571527481, 0.1896088570356369, -0.33063414692878723, -0.14126448333263397, -0.23433828353881836, -0.23844799399375916, 0.08291588723659515, 0.006255154497921467, 0.29035255312919617, -0.27429598569869995, -0.13271577656269073, 0.0041048466227948666, -0.07771909981966019, -0.5493535995483398, 0.2553773522377014, 0.12873241305351257, -0.2849130630493164, -0.05030606687068939, 0.7471105456352234, -0.18387609720230103, -0.2807393968105316, 0.06724273413419724, -0.05907456949353218, -0.09857053309679031, -0.064385786652565, 0.03518470376729965, 0.3334689438343048, -0.5780584216117859, -0.14078933000564575, -0.21537387371063232, -0.304147332906723, 0.2633821666240692, 0.24018250405788422, -0.013686264865100384, 0.1278267800807953, -0.27479520440101624, -0.02918989583849907, 0.054755859076976776, -0.05104229599237442, -0.007747017312794924, 0.1915837526321411, -0.17146135866641998, 0.2518332004547119, 0.005482812412083149, 0.07924211770296097, -0.015348897315561771, -0.07730896025896072, 0.2999677360057831, 0.354370653629303, -0.26174861192703247, -0.07039618492126465, -0.03619963675737381, -0.4247271418571472, 0.16641639173030853, 0.18734177947044373, 0.07433381676673889, -0.0027770772576332092, 0.14446164667606354, -0.23539018630981445, 0.1300501674413681, -0.06374722719192505, 0.22386756539344788, 0.2670317590236664, 0.13562442362308502, 0.5249225497245789, 0.06076227128505707, 0.04892594739794731, -0.1597839891910553, -0.015664072707295418, 0.10479103028774261, 0.40204331278800964, 0.31528958678245544, -0.1573580801486969, -0.025230038911104202, -0.1231057196855545, -0.24866054952144623, -0.3672913908958435, 0.2465508133172989, 0.2638358771800995, 0.019912902265787125, 0.33392006158828735, -0.3577461242675781, 0.286694198846817, -0.22801648080348969, -0.2566807270050049, -0.18812040984630585, 0.2682737112045288, -0.0802951380610466, 0.02375096268951893, -0.1958562135696411, 0.3590033948421478, -0.10002176463603973, -0.06508874893188477, -0.10964313894510269, -0.16664429008960724, 0.21401207149028778, 0.144977867603302, 0.516533374786377, 0.2998848855495453, 0.2574201822280884, -0.2457236349582672, -0.1387692093849182, 0.18688954412937164, 0.06296176463365555, -0.20906668901443481, 0.14302556216716766, 0.23790739476680756, 0.1825200319290161, 0.05111707001924515, 0.06885788589715958, -0.024305209517478943, 0.396051824092865, 0.2139659970998764, -0.0781375914812088, -0.1446685791015625, 0.3003721833229065, -0.06700244545936584, 0.2817316949367523, -0.07894700020551682, -0.39791548252105713, -0.02149793691933155, 0.36849814653396606, 0.022609781473875046, 0.11486007273197174, 0.1395164579153061, -0.15276995301246643, -0.10070925205945969, 0.21678510308265686, -0.48139727115631104, 0.1276506930589676, 0.26268768310546875, 0.22171537578105927, -0.16540615260601044, 0.2598189413547516, 0.03460780531167984, 0.04588068276643753, -0.25480157136917114, -0.10957546532154083, -0.039889175444841385, 0.1586690992116928, 0.07674125581979752, -0.34561771154403687, -0.4028256833553314, -0.002589689800515771, 0.08198960870504379, 0.0495547316968441, -0.1989218145608902, 0.07890422642230988, -0.05405409634113312, 0.23634356260299683, -0.3461315631866455, -0.26083245873451233, -0.09982673823833466, 0.24856841564178467, -0.3377685248851776, -0.006428166758269072, 0.04294425994157791, -0.1900705248117447, 0.323914498090744, -0.03564959019422531, -0.003153191413730383, -0.24567262828350067, 0.005888942629098892, -0.4746786653995514, 0.29337143898010254, -0.0538148432970047, -0.31592947244644165, 0.38174834847450256, -0.07545962184667587, 0.14767509698867798, -0.330128937959671, -0.510715126991272, 0.1697303205728531, -0.12999436259269714, 0.5753856897354126, 0.12542341649532318, 0.02576473169028759, 0.013468258082866669, -0.12582671642303467, 0.02396153286099434, -0.054652757942676544, -0.16415326297283173, -0.1144847422838211, -0.08873166143894196, 0.4223630428314209, 0.03869732469320297, -0.4531417191028595, -0.4203663468360901, -0.17231138050556183, 0.25293299555778503, -0.16396576166152954, -0.0030103481840342283, 0.22998593747615814, -0.13780352473258972, 0.13826419413089752, 0.00990669522434473, 0.13187921047210693, -0.26472780108451843, 0.027846159413456917, 0.2004290670156479, -0.31068673729896545, -0.2774637043476105, 0.061187196522951126, -0.09069826453924179, 0.0852126032114029, -0.24416622519493103, -0.36258020997047424, 0.2580146789550781, -0.095871701836586, 0.27385035157203674, -0.19152098894119263, 0.07490350306034088, 0.19146353006362915, 0.16966593265533447, -0.06745770573616028, 0.11042970418930054, 0.007872041314840317, -0.13218477368354797, 0.010557126253843307, 0.08699101954698563, -0.10921625047922134, -0.08538511395454407, -0.011631248518824577, 0.294559508562088, -0.011590355075895786, 0.06750230491161346, 0.28061443567276, 0.026751255616545677, 0.23155401647090912, -0.3706284165382385, -0.4429991841316223, 0.35534632205963135, 0.040035977959632874, 0.04863426089286804, 0.4195977449417114, 0.21355532109737396, -0.21425575017929077, -0.20989418029785156, 0.052577827125787735, 0.18030858039855957, -0.07134514302015305, 0.04076283425092697, -0.10820154845714569, 0.1665843427181244, -0.24493445456027985, -0.21670763194561005, -0.17305929958820343, -0.10115182399749756, 0.25474417209625244, 0.32768532633781433, 0.09362087398767471, -0.07407199591398239, -0.22788797318935394, 0.23817911744117737, -0.3003154993057251, -0.04435830935835838, 0.08408188074827194, 0.2546118497848511, -0.2716084420681, 0.33530527353286743, 0.03002925030887127, 0.1542760580778122, 0.21860188245773315, -0.6904952526092529, -0.49393826723098755, 0.13585588335990906, 0.30606648325920105, 0.15567129850387573, 0.17673593759536743, -0.18371880054473877, -0.22654828429222107, 0.3084178864955902, 0.24551022052764893, -0.16246266663074493, 0.029782148078083992, 0.19417041540145874, 0.4270161986351013, -0.01719934307038784, -0.04692133143544197, 0.0955205038189888, -0.16586390137672424, -0.0973813533782959, 0.3946688771247864, -0.029312407597899437, 0.0171725545078516, 0.08575302362442017, 0.13707469403743744, -0.033564046025276184, 0.20169410109519958, 0.2797219455242157, -0.059395693242549896, -0.16069503128528595, 0.1521374136209488, 0.08683421462774277, -0.04038185626268387, -0.4850473403930664, 0.0786108747124672, 0.19921673834323883, -0.02088051661849022, -0.5251500010490417, 0.08761663734912872, -0.2386714518070221, 0.4695502817630768, 0.16024017333984375, 0.2618008852005005, 0.3291909396648407, -0.27755218744277954, -0.21814991533756256, 0.03038795292377472, 0.28314700722694397, 0.4071323871612549, -0.03894438594579697, -0.43501943349838257, -0.7765524983406067, -0.0014824357349425554, -0.09129580855369568, 0.016643473878502846, 0.412848562002182, 0.4393831193447113, -0.13270750641822815, 0.24752601981163025, 0.07675985991954803, 0.9190419912338257, 0.23129484057426453, 0.041550472378730774, 0.07719279080629349, -0.08021669834852219, 0.19294792413711548, -0.3251575529575348, -0.1349240243434906, 0.049671582877635956, -0.2305435836315155, 0.005984198302030563, 0.2619851529598236, 0.34716567397117615, 0.24996443092823029, -0.23360663652420044, 0.1004996970295906, 0.0962321013212204, -0.19594235718250275, 0.29610997438430786, 0.30820369720458984, 0.10677450150251389, -0.43125560879707336, 0.0338464230298996, 0.2304360717535019, 0.04379310458898544, 0.203846275806427, -0.07254540920257568, -0.05668436363339424, 0.10696342587471008, 0.1314011514186859, -0.4548586905002594, -0.07609447836875916, -0.5070955157279968, 0.1264173835515976, -0.4729512631893158, 0.08941642940044403, 0.20291359722614288, -0.08584196865558624, 0.4214228689670563, 0.471717894077301, 0.0409158319234848, 0.022924842312932014, 0.4762382209300995, 0.06284089386463165, -0.08770652860403061, -0.23555314540863037, 0.2755264937877655, -0.05053960159420967, -0.2636352777481079, 0.3230833113193512, -0.2016753852367401, -0.1362142264842987, -0.38254687190055847, -0.11632005125284195, -0.10659167915582657, -0.045728541910648346, -0.17525918781757355, 0.41036558151245117, 0.17232730984687805, -0.19534604251384735, 0.1480228304862976, 0.3497506380081177, -0.23559977114200592, -0.19416159391403198, 0.10732927918434143, -0.09747235476970673, -0.3136829435825348, -0.22665981948375702, 0.2490626871585846, -0.6465169191360474, 0.2945064306259155, 0.4389958381652832, -0.15036335587501526, -0.235343798995018, 0.08640551567077637, 0.4400276243686676, -0.47015345096588135, -0.3849065899848938, 0.005976988468319178, -0.14805977046489716, -0.04269803687930107, 0.2869281768798828, 0.06661021709442139, -0.20611144602298737, 0.1507674604654312, -0.09458687901496887, -0.36929088830947876, -0.028940768912434578, 0.14872445166110992, 0.1253795176744461, 0.2711096704006195, -0.22384385764598846, 0.035785287618637085, 0.07909552752971649, -0.3972320258617401, -0.10924999415874481, -0.10658673197031021, 0.003047720994800329, 0.24276535212993622, 0.05568705126643181, 0.014138440601527691, 0.034010618925094604, 0.19847683608531952, -0.34704363346099854, -0.07309707254171371, -0.1833736151456833, -0.11463214457035065, 0.07452717423439026, 0.1349310278892517, 0.12282261997461319, -0.0779469907283783, -0.03505604341626167, -0.10303755104541779, -0.14566507935523987, 0.11736589670181274, 0.07653243839740753, -0.07971994578838348, 0.2564331293106079, -0.09329713881015778, 0.011541368439793587, -0.1612447053194046, 0.09220381081104279, 0.2833353877067566, 0.04205577075481415, 0.28873154520988464, 0.16807034611701965, 0.185028076171875, -0.13068309426307678, 0.08269313722848892, 0.2942381203174591, -0.004716777242720127, 0.3498413860797882, -0.13883361220359802, 0.037948984652757645, 0.28711938858032227, 0.3044240474700928, -0.07962300628423691, 0.02395438589155674, -0.10387226939201355, -0.038740165531635284, 0.34125956892967224, 0.28051453828811646, -0.32523113489151, 0.1633782982826233, 0.10603304952383041, -0.14815393090248108, 0.1985509693622589, 0.21697619557380676, 0.3297952115535736, 0.29217633605003357, 0.4217361509799957, -0.15993660688400269, -0.135613352060318, 0.010873841121792793, 0.4031977951526642, 0.32408249378204346, -0.037428487092256546, -0.1339367926120758, 0.09704098850488663, 0.26424261927604675, 0.010727941989898682, 0.27406206727027893, -0.13993868231773376, 0.3285216987133026, 0.19472725689411163, 0.2168826162815094, -0.012068397365510464, -0.23323258757591248, 0.44506603479385376, 0.30710768699645996, 0.2817721366882324, 0.22357037663459778, 0.1095094084739685, -0.1401677280664444, -0.1730019450187683, -0.22349786758422852, -0.04635627567768097, 0.08016705513000488, -0.20201176404953003, -0.0010114895412698388, 0.4020216166973114, -0.34331071376800537, -0.1807057410478592, -0.22531409561634064, -0.0533670075237751, -0.0010704357409849763, -0.2768154442310333, 0.11281409114599228, 0.13130497932434082, -0.3100358247756958, 0.06996656209230423, 0.029710793867707253, 0.026648210361599922, -0.061105478554964066, 0.27800217270851135, 0.010399550199508667, 0.17701537907123566, 0.04772055149078369, 0.4197734296321869, -0.13973736763000488, 0.0833040326833725, 0.016874702647328377, 0.31802260875701904, -0.20493727922439575, -0.01801319420337677, 0.0879959985613823, 0.32789111137390137, 0.06277621537446976, 0.24596007168293, 0.3205837607383728, 0.12034466117620468, -0.19261616468429565, -0.015043454244732857, 0.1718025505542755, -0.1916522979736328, -0.4136282205581665, 0.012699572369456291, -0.03840823099017143, 0.08607997000217438, -0.37970486283302307, -0.03105136565864086, -0.12854604423046112, -0.1348855346441269, 0.3256213665008545, -0.25858423113822937, -0.10838176310062408, -0.07899197190999985, 0.13846126198768616, 0.1661757528781891, -0.20290355384349823, 0.48555782437324524, 0.23405683040618896, -0.16686509549617767, -0.2174423485994339, -0.4381386935710907, 0.051264770328998566, 0.08014970272779465, 0.17446118593215942, -0.12456273287534714, 0.06258347630500793, 0.3831974267959595, 0.11679419130086899, -0.04277163743972778, -0.06421230733394623, 0.16319985687732697, 0.323292076587677, -0.3479464054107666, 0.06293880939483643, 0.2859337627887726, 0.4084407091140747, -0.030207807198166847, -0.018503136932849884, -0.07329116761684418, 0.4125908315181732, 0.15054409205913544, -0.13348494470119476, -0.24621181190013885, 0.13039027154445648, 0.1272660344839096, 0.35571905970573425, 0.226509228348732, 0.5275251865386963, -0.22152596712112427, -0.037972234189510345, -0.2685338258743286, -0.22323250770568848, -0.06925426423549652, 0.1755630373954773, 0.0285445898771286, 0.06362161040306091, -0.11477115005254745, -0.31324565410614014, -0.24616052210330963, 0.20301197469234467, 0.17591361701488495, 0.2730841636657715, -0.5236436128616333, 0.21069760620594025, 0.0956701785326004, -0.3745821714401245, -0.18295758962631226, 0.1459924578666687, 0.041923653334379196, -0.036862343549728394, -0.0974668338894844, -0.05345079302787781, 0.043697141110897064, -0.17672143876552582, -0.23924490809440613, -0.11105199158191681, 0.13255852460861206, 0.3138338327407837, 0.11591622233390808, -0.36679214239120483, 0.17310629785060883, 0.28740885853767395, 0.22819067537784576, 0.2834445834159851, -0.026607995852828026, -0.020219719037413597, 0.03794557601213455, -0.2808721661567688, 0.3004859387874603, 0.11556500941514969, 0.09651967138051987, 0.1939515322446823, -0.22160030901432037 ]
https://github.com/huggingface/datasets/issues/222
Colab Notebook breaks when downloading the squad dataset
The notebook forces version 0.1.0. If I use the latest, things work, I'll run the whole notebook and create a PR. But in the meantime, this issue gets fixed by changing: `!pip install nlp==0.1.0` to `!pip install nlp`
When I run the notebook in Colab https://colab.research.google.com/github/huggingface/nlp/blob/master/notebooks/Overview.ipynb breaks when running this cell: ![image](https://user-images.githubusercontent.com/338917/83311709-ffd1b800-a1dd-11ea-8394-3a87df0d7f8b.png)
38
Colab Notebook breaks when downloading the squad dataset When I run the notebook in Colab https://colab.research.google.com/github/huggingface/nlp/blob/master/notebooks/Overview.ipynb breaks when running this cell: ![image](https://user-images.githubusercontent.com/338917/83311709-ffd1b800-a1dd-11ea-8394-3a87df0d7f8b.png) The notebook forces version 0.1.0. If I use the latest, things work, I'll run the whole notebook and create a PR. But in the meantime, this issue gets fixed by changing: `!pip install nlp==0.1.0` to `!pip install nlp`
[ -0.2037728726863861, 0.1056739091873169, -0.03363512456417084, 0.06406715512275696, -0.1007537916302681, -0.16913969814777374, 0.1365310400724411, 0.10918202251195908, -0.34509164094924927, 0.12806232273578644, -0.1643519550561905, 0.5187447667121887, 0.18093840777873993, 0.08213856816291809, 0.16695700585842133, 0.07651351392269135, 0.06449409574270248, 0.4605701267719269, 0.04686928912997246, 0.15668264031410217, -0.21310822665691376, 0.46249136328697205, -0.4079592227935791, 0.15370342135429382, -0.2386358678340912, -0.16774101555347443, -0.20612645149230957, 0.2790915071964264, -0.5138835310935974, -0.2758502960205078, 0.2979658246040344, 0.1417236477136612, -0.008439108729362488, 0.181135892868042, -0.00011696519504766911, -0.23673918843269348, 0.12079619616270065, -0.05185142159461975, -0.3485310673713684, -0.3542286157608032, -0.008578303270041943, -0.4436560571193695, 0.1595696359872818, -0.3419606387615204, 0.14092938601970673, 0.39214780926704407, 0.3511866331100464, 0.3706689774990082, 0.3492598831653595, 0.3874809443950653, 0.15486425161361694, 0.3137628734111786, 0.2158096879720688, -0.05447574704885483, 0.1636972576379776, -0.28066226840019226, -0.29767879843711853, 0.3243989050388336, 0.5822193622589111, -0.1685716062784195, 0.0035511699970811605, 0.10590330511331558, -0.09550755470991135, 0.05590488761663437, 0.12201980501413345, 0.16388826072216034, -0.2535199224948883, -0.42911791801452637, 0.2994450628757477, 0.12129367142915726, 0.06077487766742706, -0.20016886293888092, -0.018151946365833282, 0.0014587988844141364, 0.2585044205188751, -0.34449490904808044, 0.21898534893989563, 0.3701309561729431, -0.3430841863155365, 0.07960135489702225, 0.0347072035074234, -0.0327628068625927, -0.21706049144268036, 0.2308540791273117, 0.05665071681141853, 0.3647041618824005, -0.06477519124746323, 0.0102596590295434, 0.25631603598594666, 0.0780712366104126, 0.22043900191783905, 0.2324579954147339, -0.1993965208530426, -0.012577174231410027, -0.020614471286535263, -0.19665202498435974, -0.15251649916172028, 0.3759245276451111, 0.2353895902633667, -0.35357996821403503, 0.19688540697097778, -0.06267093867063522, 0.1999603509902954, 0.19693231582641602, 0.023211361840367317, -0.012032229453325272, 0.35832932591438293, -0.041698556393384933, 0.39903318881988525, 0.40184152126312256, -0.02484694868326187, 0.0016649233875796199, 0.04790088161826134, -0.1774967461824417, -0.2790907323360443, -0.12938424944877625, 0.06468912959098816, -0.18971899151802063, -0.21751675009727478, -0.016830136999487877, -0.1500985473394394, 0.2782168388366699, 0.033172883093357086, 0.1946285516023636, -0.06817444413900375, -0.23874127864837646, 0.10871995240449905, -0.022551873698830605, -0.36746883392333984, -0.17430199682712555, 0.0099496366456151, -0.0019470612751320004, -0.18764030933380127, 0.10968411713838577, 0.2619291841983795, 0.18368901312351227, 0.47783488035202026, -0.15345916152000427, 0.09960615634918213, -0.11781244724988937, 0.052396032959222794, -0.265323281288147, -0.10134149342775345, 0.3144093453884125, 0.31551674008369446, -0.22006887197494507, -0.2717234194278717, -0.5757358074188232, 0.04511936753988266, 0.19760005176067352, -0.032033734023571014, -0.012889334931969643, -0.1250273734331131, 0.10841792821884155, -0.28446757793426514, -0.07026506960391998, -0.32433924078941345, 0.054985299706459045, -0.2155044972896576, -0.25893187522888184, 0.04896066337823868, -0.15393143892288208, -0.2456502914428711, -0.0019288924522697926, -0.05847962945699692, 0.18696649372577667, -0.03239218518137932, -0.0988168716430664, 0.16054698824882507, 0.008413736708462238, -0.0025034071877598763, 0.2557321786880493, -0.14600422978401184, -0.08713645488023758, 0.06068825721740723, 0.19247420132160187, 0.38259610533714294, -0.31424084305763245, -0.7157808542251587, 0.12331607937812805, -0.07620780169963837, -0.28646746277809143, -0.09852341562509537, 0.1813635528087616, 0.4203037619590759, -0.18550942838191986, 0.07004693150520325, 0.4979853630065918, -0.17399728298187256, 0.10833067446947098, -0.2697705328464508, -0.17382092773914337, -0.12889860570430756, 0.050152070820331573, 0.11421242356300354, -0.017191285267472267, -0.03910699114203453, 0.7044438123703003, 0.16681362688541412, -0.09890922158956528, -0.20148754119873047, 0.059921689331531525, 0.3261888325214386, -0.3010641634464264, -0.08822111785411835, -0.14634902775287628, -0.5062059164047241, 0.010047032497823238, -0.3460903465747833, 0.21325331926345825, 0.09683506935834885, -0.21112985908985138, -0.05918259173631668, -0.06821393221616745, -0.012183494865894318, -0.17810164391994476, 0.10006795078516006, 0.020486848428845406, 0.04886997863650322, -0.04860610514879227, 0.09019891917705536, 0.39791053533554077, -0.1955254077911377, 0.4248863458633423, -0.2051776647567749, 0.18991997838020325, -0.3483101427555084, -0.3624226748943329, 0.06926047801971436, 0.3132472336292267, 0.0034315469674766064, -0.17059659957885742, -0.10006596893072128, 0.2572607398033142, 0.08105327188968658, -0.07469108700752258, 0.10122845321893692, -0.3666817843914032, 0.3960622251033783, 0.060668669641017914, -0.040356315672397614, -0.06082799285650253, 0.012146322056651115, -0.035369791090488434, -0.02880089171230793, 0.22457733750343323, -0.18602238595485687, -0.04142305254936218, 0.1730562448501587, 0.20941908657550812, 0.03926527872681618, -0.11988618224859238, -0.20445379614830017, 0.13204503059387207, 0.32316672801971436, 0.23681366443634033, 0.04118005558848381, 0.01091484259814024, -0.3629995286464691, 0.21057966351509094, 0.49385952949523926, 0.02746618166565895, 0.2480994313955307, 0.01987282745540142, 0.0485268160700798, 0.002645396161824465, 0.3425407409667969, 0.08806036412715912, 0.3751700222492218, 0.1141345351934433, -0.11013163626194, -0.0038268642965704203, -0.1853577345609665, -0.3078567385673523, 0.13142085075378418, 0.08304600417613983, 0.12990693747997284, -0.16973982751369476, 0.30302736163139343, 0.014846326783299446, -0.3646789491176605, -0.11107052117586136, -0.07071533054113388, 0.23503388464450836, 0.10903188586235046, 0.30596378445625305, -0.14340345561504364, -0.610624372959137, -0.20321539044380188, -0.11518248170614243, -0.1861470639705658, -0.18630284070968628, -0.05477512627840042, 0.39906999468803406, 0.39377883076667786, 0.1321960985660553, 0.359119713306427, 0.20930492877960205, -0.11759864538908005, 0.18299828469753265, 0.17196907103061676, -0.04389533773064613, -0.42944610118865967, 0.03136978670954704, 0.38624444603919983, 0.29799824953079224, 0.1085411012172699, -0.2365487515926361, -0.249202162027359, 0.015472673811018467, -0.42980483174324036, 0.2603362798690796, -0.33625665307044983, 0.2794792056083679, 0.09976360201835632, 0.24782203137874603, -0.5951433777809143, -0.10310769081115723, 0.16498945653438568, -0.11449159681797028, -0.12290865927934647, -0.09292823076248169, -0.036882806569337845, -0.4055459499359131, -0.2671477794647217, -0.20519211888313293, -0.4359723627567291, -0.2693944275379181, 0.33582839369773865, 0.01575196348130703, 0.21157871186733246, 0.37397632002830505, 0.09294790029525757, 0.3100525140762329, 0.04953617602586746, -0.2053631842136383, 0.1208367794752121, 0.03579412400722504, 0.08852697908878326, -0.25054535269737244, -0.49854177236557007, 0.09977488964796066, 0.05173466354608536, 0.037645015865564346, -0.2449919432401657, -0.4161094129085541, -0.5225147604942322, 0.12612779438495636, 0.18824434280395508, -0.0012073491234332323, -0.07886319607496262, 0.20694324374198914, -0.1973298341035843, 0.10955783724784851, -0.01307738572359085, -0.24611149728298187, -0.12782561779022217, 0.08309205621480942, 0.3247514069080353, -0.15073645114898682, 0.10587984323501587, 0.14839668571949005, 0.19341632723808289, 0.3145638704299927, -0.07250470668077469, 0.11436870694160461, 0.12013480067253113, 0.2739352881908417, -0.2105095386505127, -0.3405756950378418, 0.3424797058105469, 0.010134987533092499, -0.009619178250432014, 0.34545671939849854, -0.059728506952524185, -0.5129151940345764, -0.1813618242740631, 0.10224588960409164, -0.25864049792289734, -0.356606662273407, 0.10800483822822571, 0.4365466833114624, -0.0429641529917717, 0.21928995847702026, 0.17693999409675598, -0.006432090885937214, -0.2671232521533966, -0.3938911557197571, 0.2829701602458954, -0.23305200040340424, 0.10045378655195236, 0.18663017451763153, -0.45402929186820984, -0.6995073556900024, 0.12911401689052582, 0.020111963152885437, 0.297396719455719, -0.008754974231123924, 0.23194825649261475, 0.13225001096725464, -0.0280307587236166, 0.2421714961528778, 0.08653072267770767, -0.12931282818317413, 0.055536653846502304, 0.02404213882982731, -0.2520790696144104, 0.14389976859092712, -0.3590138256549835, 0.5358407497406006, 0.8802660703659058, 0.3770967423915863, -0.16434834897518158, 0.05778162181377411, 0.5292020440101624, -0.032791152596473694, 0.18958882987499237, 0.10804035514593124, -0.30443358421325684, -0.32131001353263855, -0.14403389394283295, 0.0569625198841095, -0.04829384759068489, -0.0072516039945185184, 0.0971345454454422, -0.10260337591171265, 0.2945190668106079, -0.13049434125423431, -0.18037919700145721, 0.49008452892303467, -0.04314805567264557, 0.0689837783575058, 0.034631360322237015, -0.20050600171089172, -0.148512065410614, 0.23588094115257263, 0.27820107340812683, -0.21470865607261658, -0.25108450651168823, 0.1910683661699295, -0.29935526847839355, 0.12639929354190826, 0.5574209690093994, -0.01270495168864727, -0.08767660707235336, -0.14137129485607147, -0.010357515886425972, 0.37668436765670776, 0.00820449460297823, 0.3475843667984009, 0.13797618448734283, -0.08299735188484192, 0.15213744342327118, 0.27438995242118835, -0.010234685614705086, 0.060648851096630096, -0.1266959309577942, -0.13302616775035858, -0.26653531193733215, 0.43179652094841003, 0.03132355958223343, 0.8313065767288208, -0.19805532693862915, 0.18259085714817047, 0.216595858335495, 0.03458208218216896, 0.7005776762962341, 0.09818413853645325, 0.29737457633018494, -0.16050338745117188, -0.2612914741039276, -0.02691715583205223, 0.03417295590043068, 0.2771373987197876, 0.0035689116921275854, -0.6976837515830994, 0.3543381690979004, 0.5069895386695862, -0.06665916740894318, 0.14088737964630127, 0.20467053353786469, -0.20412826538085938, -0.3309400975704193, -0.6929689049720764, 0.14991536736488342, -0.24720880389213562, 0.2120819091796875, -0.2318272739648819, -0.016606898978352547, -0.22074270248413086, -0.055851973593235016, -0.39808860421180725, 0.09784320741891861, -0.49289411306381226, -0.149103581905365, 0.3097960650920868, -0.2716412842273712, -0.015284676104784012, 0.31808653473854065, 0.2980518043041229, 0.21503795683383942, -0.04117601364850998, 0.055800847709178925, -0.009488862007856369, -0.4288063049316406, 0.06961584091186523, 0.12072650343179703, 0.0707867369055748, -0.23157912492752075, -0.011699111200869083, -0.15616832673549652, 0.19199395179748535, -0.14730100333690643, 0.37787121534347534, 0.2653465270996094, 0.11379894614219666, 0.15317033231258392, -0.11881103366613388, -0.042882852256298065, -0.1843665987253189, -0.26054760813713074, 0.10958398133516312, -0.10283757001161575, -0.10809288918972015, 0.1768660545349121, 0.1704011708498001, -0.16747742891311646, -0.018127845600247383, 0.2444475144147873, -0.10914622247219086, 0.05315729230642319, 0.21790307760238647, 0.20292866230010986, -0.04807581752538681, -0.2186717391014099, 0.16416016221046448, -0.47109168767929077, -0.20756664872169495, 0.019297951832413673, 0.0043054916895926, -0.17338953912258148, -0.05642080679535866, 0.6419909596443176, 0.2973116338253021, 0.03289369121193886, -0.12059967964887619, -0.3671324849128723, -0.22014698386192322, 0.204116091132164, 0.22043579816818237, -0.12179027497768402, 0.33465540409088135, 0.33405187726020813, 0.2858063876628876, 0.0965169370174408, -0.22786809504032135, 0.28883522748947144, -0.3742843568325043, 0.17683275043964386, 0.03506293520331383, -0.2467697411775589, 0.16113634407520294, -0.13581202924251556, -0.05651438608765602, 0.5478607416152954, 0.06339484453201294, -0.21986687183380127, -0.1331758052110672, 0.06871269643306732, 0.11896977573633194, -0.10746382176876068, -0.3171968460083008, -0.06813284754753113, -0.0985211580991745, -0.04035235941410065, 0.07765252888202667, -0.06821112334728241, 0.10933578759431839, -0.06905275583267212, 0.2667797803878784, 0.2946946918964386, 0.24825841188430786, 0.07181059569120407, -0.05436825752258301, 0.21100760996341705, 0.12927617132663727, 0.25017687678337097, 0.21401181817054749, 0.03418063744902611, 0.14100219309329987, -0.00031006732024252415, -0.07852956652641296, 0.15759477019309998, 0.2667442858219147, -0.40233513712882996, 0.10853870213031769, -0.06369985640048981, -0.04671769216656685, 0.3891511559486389, 0.05957886576652527, -0.16276578605175018, 0.2553037405014038, 0.07318639010190964, -0.14528235793113708, 0.06295536458492279, 0.5381837487220764, 0.1647595912218094, 0.07420119643211365, 0.4867258071899414, 0.11547704041004181, -0.0481448620557785, 0.14421556890010834, 0.10595760494470596, 0.23524418473243713, 0.25333428382873535, 0.011607774533331394, -0.0835575982928276, 0.013890804722905159, 0.11324302107095718, 0.0884743183851242, 0.05775366351008415, 0.1947413682937622, 0.008706332184374332, -0.002557041822001338, 0.1844155490398407, -0.13208450376987457, 0.09944796562194824, 0.29769665002822876, -0.036573950201272964, -0.21635518968105316, 0.044106319546699524, 0.27920857071876526, 0.06218254193663597, 0.13555841147899628, 0.8394278883934021, -0.535256028175354, -0.25631362199783325, -0.11296948045492172, 0.1396200954914093, -0.36160218715667725, -0.006911978591233492, -0.3097970485687256, 0.15970878303050995, -0.13192488253116608, -0.018767697736620903, -0.154642254114151, -0.04683639481663704, 0.30263879895210266, 0.29672548174858093, -0.1764538735151291, -0.11941305547952652, -0.21040664613246918, 0.0963275209069252, -0.15803036093711853, -0.04321933165192604, 0.5650871992111206, -0.461553692817688, 0.15587083995342255, -0.11572029441595078, 0.5128489136695862, 0.19595342874526978, 0.23445644974708557, -0.17214412987232208, -0.1699351668357849, -0.040043801069259644, 0.021636266261339188, -0.09688611328601837, -0.019116709008812904, 0.4978405833244324, 0.2487422525882721, 0.1886567920446396, 0.030080614611506462, -0.11569540947675705, 0.19563622772693634, 0.15852294862270355, -0.26282206177711487, 0.05865886062383652, -0.10240868479013443, -0.107417032122612, -0.14664077758789062, -0.2744995951652527, -0.04305624961853027, -0.3473242223262787, 0.5429143905639648, -0.29940956830978394, -0.10698141902685165, 0.04497699439525604, -0.09464602917432785, 0.04844537749886513, -0.004604424815624952, 0.5350642204284668, 0.40321555733680725, 0.025475937873125076, -0.21520118415355682, -0.42776626348495483, -0.3889569044113159, 0.21612274646759033, -0.2563641667366028, 0.011096286587417126, 0.05963536724448204, 0.18237698078155518, -0.10123986750841141, 0.413977712392807, -0.0789305716753006, 0.041846632957458496, -0.12649931013584137, -0.0653214305639267, -0.47243624925613403, -0.1160053014755249, 0.28529104590415955, -0.2760682702064514, -0.0696646124124527, -0.22366367280483246, 0.15214872360229492, -0.20585936307907104, -0.07175639271736145, -0.0071328203193843365, -0.3104133903980255, -0.29001426696777344, -0.1044817566871643, 0.07206235080957413, 0.17017768323421478, 0.10107682645320892, 0.019608771428465843, -0.3210943639278412, -0.04657581076025963, -0.22404059767723083, -0.2719593346118927, 0.07557530701160431, 0.22568345069885254, 0.29754921793937683, -0.05231276899576187, 0.3451109528541565, -0.16596171259880066, 0.5020354390144348, 0.10082227736711502, 0.0367015078663826, -0.012219380587339401, 0.001554313930682838, -0.12299061566591263, 0.08524483442306519, -0.14766062796115875, 0.45889899134635925, -0.27856770157814026, 0.17531083524227142, -0.4899500608444214, -0.31665006279945374, 0.38322725892066956, -0.4803469479084015, -0.09981098771095276, 0.001730389310978353, 0.20958471298217773, 0.1394142061471939, -0.30026519298553467, -0.3976467251777649, -0.02579435706138611, -0.1576421856880188, -0.1750725954771042, -0.4597724676132202, 0.13977788388729095, -0.20215091109275818, 0.051265932619571686, -0.08261678367853165, 0.4182424247264862, 0.07982735335826874, -0.2558419108390808, 0.015247183851897717, -0.15567591786384583 ]
https://github.com/huggingface/datasets/issues/222
Colab Notebook breaks when downloading the squad dataset
When you install `nlp` for the first time on a Colab runtime, it updates the `pyarrow` library that was already on colab. This update shows this message on colab: ``` WARNING: The following packages were previously imported in this runtime: [pyarrow] You must restart the runtime in order to use newly installed versions. ``` You just have to restart the runtime and it should be fine. If you don't restart, then it breaks like in your first message
When I run the notebook in Colab https://colab.research.google.com/github/huggingface/nlp/blob/master/notebooks/Overview.ipynb breaks when running this cell: ![image](https://user-images.githubusercontent.com/338917/83311709-ffd1b800-a1dd-11ea-8394-3a87df0d7f8b.png)
78
Colab Notebook breaks when downloading the squad dataset When I run the notebook in Colab https://colab.research.google.com/github/huggingface/nlp/blob/master/notebooks/Overview.ipynb breaks when running this cell: ![image](https://user-images.githubusercontent.com/338917/83311709-ffd1b800-a1dd-11ea-8394-3a87df0d7f8b.png) When you install `nlp` for the first time on a Colab runtime, it updates the `pyarrow` library that was already on colab. This update shows this message on colab: ``` WARNING: The following packages were previously imported in this runtime: [pyarrow] You must restart the runtime in order to use newly installed versions. ``` You just have to restart the runtime and it should be fine. If you don't restart, then it breaks like in your first message
[ -0.16229388117790222, 0.288650244474411, -0.016385361552238464, 0.0890752375125885, -0.09513921290636063, -0.13527429103851318, 0.24155989289283752, 0.04239322990179062, -0.44447314739227295, 0.05577054247260094, -0.34253883361816406, 0.5006802678108215, 0.19029919803142548, 0.051721375435590744, 0.2317516803741455, 0.12949404120445251, 0.05272120609879494, 0.45624205470085144, -0.005169594660401344, 0.1515726000070572, -0.13718967139720917, 0.4322420358657837, -0.3422400653362274, 0.19298385083675385, -0.28372883796691895, -0.20665064454078674, -0.338755339384079, 0.3366294801235199, -0.38382747769355774, -0.36793914437294006, 0.3007683753967285, 0.11652277410030365, -0.010650957003235817, 0.242111474275589, -0.00011587276821956038, -0.2633945643901825, 0.07170131057500839, 0.050973426550626755, -0.3235515058040619, -0.1854395866394043, 0.1579292118549347, -0.276764452457428, 0.3587700128555298, -0.36968234181404114, 0.23197370767593384, 0.2630355656147003, 0.19682104885578156, 0.4131794273853302, 0.3784221112728119, 0.25509005784988403, 0.15432634949684143, 0.3623243570327759, 0.15149927139282227, -0.049902793020009995, 0.3280055522918701, -0.24015401303768158, -0.24223317205905914, 0.5541875958442688, 0.5191044807434082, -0.17601603269577026, -0.06431899964809418, -0.02365473285317421, -0.15008224546909332, 0.10818630456924438, 0.09065601974725723, 0.2510839104652405, -0.30428025126457214, -0.3595670759677887, 0.14779271185398102, 0.07521253079175949, 0.010056298226118088, -0.2550983428955078, 0.08777561783790588, -0.08877317607402802, 0.2548277974128723, -0.46778053045272827, 0.22016583383083344, 0.2952525019645691, -0.3333261013031006, 0.04739907756447792, 0.06930729746818542, -0.09359437972307205, -0.1557636708021164, 0.18317195773124695, 0.11445876210927963, 0.3580796420574188, -0.056636981666088104, 0.011706068180501461, 0.257982462644577, 0.11046643555164337, 0.2554868161678314, 0.26135551929473877, -0.1273597925901413, -0.12093061208724976, -0.04816731810569763, -0.10802503675222397, -0.11227375268936157, 0.35357746481895447, 0.1700029969215393, -0.2848016023635864, 0.18919986486434937, -0.08284584432840347, 0.1804109811782837, 0.18949389457702637, -0.02586640976369381, 0.011831995099782944, 0.17808897793293, -0.22734493017196655, 0.4025895893573761, 0.26510098576545715, -0.1325010061264038, -0.01960589736700058, 0.06968578696250916, -0.0828315019607544, -0.1812393218278885, -0.20027397572994232, 0.11146128177642822, -0.1733042597770691, -0.21198949217796326, -0.060525644570589066, -0.21012011170387268, 0.10313393175601959, 0.030671831220388412, 0.22637084126472473, -0.04862583428621292, -0.3248872756958008, -0.008808780461549759, 0.00972329918295145, -0.3067673146724701, -0.1658766269683838, -0.020702529698610306, -0.01581350527703762, -0.3158728778362274, 0.29279258847236633, 0.25611117482185364, 0.014170718379318714, 0.4369165897369385, -0.06674621254205704, 0.11256535351276398, -0.06156915798783302, -0.03654184192419052, -0.2505078613758087, -0.13767024874687195, 0.19454501569271088, 0.28504177927970886, -0.13704141974449158, -0.2746259272098541, -0.5182225108146667, -0.029475443065166473, 0.1341782659292221, -0.09137090295553207, -0.10038149356842041, -0.14188021421432495, 0.11342848837375641, -0.2824956476688385, -0.0845007672905922, -0.24448548257350922, -0.011238967068493366, -0.2775983214378357, -0.32704630494117737, 0.06488179415464401, -0.11334086954593658, -0.11958078294992447, -0.17336565256118774, -0.051390986889600754, 0.2277742475271225, -0.15503674745559692, -0.06079227849841118, 0.1266467124223709, 0.08688955754041672, 0.05646294355392456, 0.36917585134506226, -0.22387829422950745, 0.058817826211452484, -0.030296456068754196, 0.27381184697151184, 0.47548240423202515, -0.13133639097213745, -0.5511142611503601, 0.022248180583119392, -0.09273975342512131, -0.22001181542873383, -0.17768460512161255, 0.18534061312675476, 0.435928076505661, -0.17135454714298248, 0.12053326517343521, 0.4776642620563507, -0.1279914379119873, 0.08927565068006516, -0.25258946418762207, -0.20198699831962585, 0.02106177993118763, 0.0683601051568985, 0.1030823290348053, -0.04792466014623642, 0.01084702368825674, 0.6572165489196777, 0.1523803323507309, -0.12794378399848938, -0.29248911142349243, 0.012023300863802433, 0.26373133063316345, -0.15930862724781036, -0.08615025877952576, -0.10207698494195938, -0.4976184070110321, -0.018914980813860893, -0.38266417384147644, 0.22172406315803528, 0.11234783381223679, -0.12611830234527588, -0.019689487293362617, -0.029196416959166527, 0.013085149228572845, -0.10262636095285416, 0.11551812291145325, 0.09121518582105637, 0.03127596154808998, 0.035521961748600006, 0.01103814784437418, 0.3079177439212799, -0.21081526577472687, 0.40704047679901123, -0.1642879843711853, 0.339485228061676, -0.3726192116737366, -0.45434340834617615, 0.038831572979688644, 0.3175658583641052, -0.005407249089330435, -0.1858677864074707, -0.08476964384317398, 0.3038392663002014, -0.03310577571392059, -0.0015203795628622174, -0.04197312891483307, -0.21679817140102386, 0.41755712032318115, -0.03388013690710068, 0.0402371846139431, -0.15333673357963562, 0.011028697714209557, -0.11775404959917068, 0.12128852307796478, 0.15278571844100952, -0.2726379632949829, -0.04674198850989342, 0.119327113032341, 0.08081535995006561, 0.0008183441823348403, -0.12340391427278519, -0.12160493433475494, 0.20139923691749573, 0.2504895031452179, 0.23211480677127838, 0.05561455711722374, 0.003987410105764866, -0.34483665227890015, 0.16886170208454132, 0.6388939023017883, 0.009611033834517002, 0.19339968264102936, 0.0026634298264980316, -0.05535617470741272, -0.09187658131122589, 0.3654562830924988, -0.1382661908864975, 0.44734662771224976, 0.08536034822463989, -0.11062626540660858, 0.04611613228917122, -0.20016132295131683, -0.3595986068248749, 0.08434499055147171, 0.15042081475257874, 0.33935287594795227, -0.16112491488456726, 0.3281324803829193, 0.18927942216396332, -0.4165046513080597, -0.19667786359786987, -0.11324004828929901, 0.16031929850578308, 0.0463760606944561, 0.3105352818965912, -0.2258267104625702, -0.6363139748573303, -0.21384714543819427, -0.22468766570091248, -0.17365486919879913, -0.2121385931968689, -0.06543543934822083, 0.41710540652275085, 0.3795734941959381, 0.10482976585626602, 0.39083635807037354, 0.29194238781929016, -0.1407936066389084, 0.15146809816360474, 0.07684767991304398, -0.11653324216604233, -0.4451967775821686, 0.035596415400505066, 0.4403403103351593, 0.3027222454547882, 0.18668515980243683, -0.18262286484241486, -0.14730432629585266, 0.04165108501911163, -0.47399404644966125, 0.2552632987499237, -0.28294262290000916, 0.25056612491607666, 0.005636756774038076, 0.2146395444869995, -0.5915745496749878, -0.2966303825378418, 0.17837487161159515, -0.19174470007419586, -0.06382713466882706, -0.07141555100679398, -0.09075440466403961, -0.36869946122169495, -0.2321414053440094, -0.19074231386184692, -0.3892238736152649, -0.2519965171813965, 0.30632463097572327, 0.04709045588970184, 0.1246846467256546, 0.34518328309059143, -0.057847023010253906, 0.4203668534755707, -0.006661214865744114, -0.04496514052152634, 0.06837761402130127, 0.03224172443151474, 0.06294324994087219, -0.2506495714187622, -0.39018315076828003, 0.08632972836494446, 0.0814005583524704, 0.07147722691297531, -0.32992270588874817, -0.44783109426498413, -0.4516487717628479, 0.10525130480527878, 0.2513573467731476, 0.009098992682993412, -0.09579433500766754, 0.2081669270992279, -0.180908203125, 0.05214687064290047, 0.12245649844408035, -0.35239335894584656, -0.20721876621246338, 0.15541191399097443, 0.30154314637184143, -0.15515124797821045, 0.14002054929733276, -0.03658068925142288, 0.24720655381679535, 0.12030598521232605, -0.16269680857658386, 0.24349345266819, 0.047038301825523376, 0.2754901349544525, -0.26920491456985474, -0.3264657258987427, 0.3509715497493744, -0.13879068195819855, -0.0719209536910057, 0.3527228832244873, -0.1739964485168457, -0.5435376167297363, -0.17621362209320068, 0.08835059404373169, -0.35202929377555847, -0.35120049118995667, 0.1408773958683014, 0.4890921115875244, -0.1283901184797287, 0.16154631972312927, 0.10040314495563507, 0.1366296112537384, -0.31285786628723145, -0.2801976501941681, 0.146880105137825, -0.3128332197666168, 0.015015988610684872, 0.08403727412223816, -0.5422755479812622, -0.6152955889701843, 0.09041346609592438, 0.02460479363799095, 0.16473153233528137, 0.010636893101036549, 0.16953375935554504, 0.16979213058948517, -0.07759524881839752, 0.18881633877754211, 0.2393738180398941, -0.21169398725032806, 0.10137899219989777, 0.10286115854978561, -0.2493128627538681, 0.004477797541767359, -0.4466743469238281, 0.4395610988140106, 0.8527015447616577, 0.2162889689207077, -0.16088302433490753, 0.14705929160118103, 0.49671176075935364, 0.08200792223215103, 0.18961025774478912, 0.13085366785526276, -0.2746916711330414, -0.3481544256210327, -0.20279675722122192, 0.02792227827012539, -0.20913065969944, 0.03567702695727348, -0.04792383685708046, -0.05737583711743355, 0.33631280064582825, -0.05814103037118912, -0.11765913665294647, 0.5659457445144653, -0.05071761831641197, 0.13834697008132935, -0.03656565770506859, -0.16751043498516083, -0.0553867369890213, 0.20562461018562317, 0.31446409225463867, -0.1491711288690567, -0.11714614182710648, 0.2581256330013275, -0.17916445434093475, 0.1546858847141266, 0.5725124478340149, 0.03370540216565132, -0.015727074816823006, -0.25497835874557495, -0.027279751375317574, 0.3534434139728546, 0.038461118936538696, 0.39194387197494507, 0.024101948365569115, -0.09691433608531952, 0.22327563166618347, 0.36032676696777344, 0.006347659509629011, 0.005739286541938782, -0.05591950565576553, -0.1315789520740509, -0.16340439021587372, 0.5269415974617004, 0.09646090120077133, 0.9199802875518799, -0.1351892203092575, 0.23280107975006104, 0.39558592438697815, 0.005457806866616011, 0.8213491439819336, 0.036442551761865616, 0.29021504521369934, -0.22001081705093384, -0.10888057202100754, 0.03861979767680168, -0.027458952739834785, 0.14550721645355225, -0.15644876658916473, -0.6769128441810608, 0.24205642938613892, 0.45106393098831177, 0.08140242844820023, 0.05928398296236992, 0.22673897445201874, -0.04052628576755524, -0.279349148273468, -0.6700126528739929, 0.16516225039958954, -0.23910804092884064, 0.25718119740486145, -0.31535476446151733, 0.006178461946547031, -0.12149997055530548, -0.019443398341536522, -0.41948917508125305, 0.1596047580242157, -0.3940165340900421, -0.10973135381937027, 0.25963592529296875, -0.24909469485282898, -0.003666806733235717, 0.19052797555923462, 0.3236684501171112, 0.18673281371593475, -0.035921741276979446, -0.027197666466236115, -0.07232914865016937, -0.34440332651138306, 0.13889990746974945, 0.07803338021039963, 0.13850174844264984, -0.23670165240764618, -0.08687042444944382, 0.028654184192419052, 0.240557461977005, -0.1788833737373352, 0.2510046362876892, 0.2675101161003113, 0.09652291983366013, 0.038775768131017685, 0.02177240699529648, 0.04931182786822319, -0.12440291792154312, -0.25639116764068604, 0.08975756168365479, -0.10483017563819885, -0.046367671340703964, 0.2597413957118988, 0.1653759777545929, -0.15297961235046387, -0.09329468756914139, 0.16011503338813782, -0.12814654409885406, 0.09165698289871216, 0.18983635306358337, 0.261301726102829, -0.07720876485109329, -0.17520982027053833, 0.04547451436519623, -0.5221540927886963, -0.3000229299068451, -0.07522933185100555, 0.026011241599917412, -0.05449257045984268, 0.07419297844171524, 0.6213568449020386, 0.3452061116695404, 0.13891078531742096, -0.09827134758234024, -0.3008250296115875, -0.13825076818466187, 0.22200581431388855, 0.3104008734226227, -0.10538150370121002, 0.43308305740356445, 0.42102181911468506, 0.23336516320705414, 0.17018412053585052, -0.23408561944961548, 0.3082623779773712, -0.34559938311576843, 0.1343727558851242, -0.04723036661744118, -0.27821439504623413, 0.18948988616466522, -0.06866831332445145, -0.050933126360177994, 0.5490370988845825, 0.1302432417869568, -0.1948457807302475, -0.05549868941307068, 0.06485958397388458, 0.08955344557762146, -0.10635147988796234, -0.3555722236633301, -0.08047781139612198, 0.0717417299747467, 0.007463438436388969, 0.11996138095855713, -0.1320907026529312, 0.1513807475566864, -0.0682617723941803, 0.3728257417678833, 0.24534714221954346, 0.13897696137428284, 0.11990634351968765, 0.018168259412050247, 0.23148885369300842, 0.164936825633049, 0.2641076147556305, 0.1112995371222496, -0.00837633851915598, 0.10891076177358627, 0.01216722559183836, -0.044086556881666183, 0.06559478491544724, 0.19642257690429688, -0.3878045380115509, 0.07796381413936615, -0.2351732850074768, -0.03739141300320625, 0.48941361904144287, 0.13260631263256073, -0.18248459696769714, 0.24811656773090363, 0.06972337514162064, -0.11593460291624069, 0.022276416420936584, 0.448589026927948, 0.1829523742198944, -0.019730819389224052, 0.4950132668018341, 0.1573646366596222, -0.003335085231810808, 0.126298725605011, 0.04812293499708176, 0.2687559127807617, 0.36889493465423584, 0.00048571894876658916, -0.06775577366352081, -0.06217401474714279, 0.03523796796798706, 0.17615684866905212, 0.06021834909915924, 0.1521511822938919, -0.06740184128284454, -0.12788811326026917, 0.15209031105041504, -0.28901150822639465, 0.06382237374782562, 0.30662307143211365, -0.05420028790831566, -0.25578412413597107, 0.1104951724410057, 0.37385624647140503, -0.02909010462462902, 0.08440849184989929, 0.8333593010902405, -0.3544100224971771, -0.320932537317276, -0.10203983634710312, 0.2567324638366699, -0.34500885009765625, 0.09901536256074905, -0.2724238336086273, 0.103946752846241, -0.083628349006176, 0.08783449977636337, -0.17857900261878967, -0.13201075792312622, 0.3258434534072876, 0.2735811471939087, -0.2538753151893616, -0.1599980890750885, -0.16739647090435028, 0.05449562519788742, -0.1816900223493576, 0.07494228333234787, 0.5572718381881714, -0.33135437965393066, 0.09935491532087326, -0.2330409735441208, 0.5280357599258423, 0.289602667093277, 0.20487409830093384, -0.1909986138343811, -0.1523386687040329, -0.040742795914411545, 0.045050814747810364, -0.014735683798789978, -0.06881915777921677, 0.49519580602645874, 0.10088562965393066, 0.1896582394838333, 0.04977743327617645, -0.13975879549980164, 0.28492435812950134, 0.13173885643482208, -0.21514587104320526, 0.037305768579244614, -0.10045892745256424, 0.00375389470718801, -0.17565642297267914, -0.30538445711135864, -0.0506778284907341, -0.3146499693393707, 0.5148186087608337, -0.14559194445610046, -0.11733570694923401, -0.04240399971604347, -0.1393558830022812, 0.05286143347620964, 0.0952523797750473, 0.6466189622879028, 0.5057281255722046, 0.15907450020313263, -0.1942807137966156, -0.45061254501342773, -0.4104742407798767, 0.26898157596588135, -0.30991464853286743, 0.11366542428731918, 0.05045677721500397, 0.192466601729393, -0.16570279002189636, 0.47900041937828064, -0.026662949472665787, 0.1169876754283905, -0.24810971319675446, -0.1550626903772354, -0.5336962938308716, -0.015051491558551788, 0.11213742196559906, -0.3295290172100067, -0.16424046456813812, -0.04330587759613991, 0.21823424100875854, -0.342174232006073, -0.07783357799053192, -0.06188115477561951, -0.3067428767681122, -0.2668483555316925, -0.0754532441496849, 0.25476428866386414, 0.10352242738008499, 0.13317273557186127, 0.07666145265102386, -0.31117329001426697, 0.02930278889834881, -0.38363730907440186, -0.14511217176914215, 0.04734329879283905, 0.1732354611158371, 0.19779276847839355, 0.0003893404791597277, 0.36838871240615845, -0.19359782338142395, 0.5045951008796692, 0.09660834074020386, 0.06460680812597275, -0.0029810687992721796, -0.09048990160226822, -0.07207760214805603, 0.08612105995416641, -0.08781018853187561, 0.42042768001556396, -0.239137202501297, 0.25775447487831116, -0.5476999878883362, -0.2630959749221802, 0.4129124581813812, -0.45274871587753296, -0.04186322167515755, -0.004355814307928085, 0.1946285367012024, 0.11911741644144058, -0.1933596134185791, -0.4065580666065216, -0.09077273309230804, -0.10226752609014511, -0.1282307356595993, -0.4465615451335907, 0.18074031174182892, -0.12650586664676666, -0.009618858806788921, -0.11334388703107834, 0.2060467004776001, 0.0443296954035759, -0.2754810154438019, -0.07351337373256683, -0.18137361109256744 ]
https://github.com/huggingface/datasets/issues/222
Colab Notebook breaks when downloading the squad dataset
Thanks for reporting the second one ! We'll update the notebook to fix this one :)
When I run the notebook in Colab https://colab.research.google.com/github/huggingface/nlp/blob/master/notebooks/Overview.ipynb breaks when running this cell: ![image](https://user-images.githubusercontent.com/338917/83311709-ffd1b800-a1dd-11ea-8394-3a87df0d7f8b.png)
16
Colab Notebook breaks when downloading the squad dataset When I run the notebook in Colab https://colab.research.google.com/github/huggingface/nlp/blob/master/notebooks/Overview.ipynb breaks when running this cell: ![image](https://user-images.githubusercontent.com/338917/83311709-ffd1b800-a1dd-11ea-8394-3a87df0d7f8b.png) Thanks for reporting the second one ! We'll update the notebook to fix this one :)
[ -0.20812645554542542, 0.08529706299304962, -0.0881832093000412, 0.16447612643241882, -0.02311563305556774, -0.10333478450775146, 0.3287060260772705, 0.08260199427604675, -0.3126959502696991, 0.21142807602882385, -0.39940711855888367, 0.3362332880496979, 0.2900560796260834, 0.16409459710121155, 0.1548372507095337, 0.16451476514339447, 0.09830828756093979, 0.37275925278663635, -0.06697428226470947, 0.14713570475578308, -0.2285151332616806, 0.45175933837890625, -0.36850640177726746, 0.06003290042281151, -0.2734021842479706, -0.20931944251060486, -0.2376537173986435, 0.2732951045036316, -0.42369070649147034, -0.21809439361095428, 0.28793662786483765, 0.18390311300754547, -0.11174608021974564, 0.25990378856658936, -0.00011433532927185297, -0.24122992157936096, -0.02032240480184555, -0.06533642113208771, -0.3073406517505646, -0.16067147254943848, -0.0806206613779068, -0.23325729370117188, 0.22001750767230988, -0.4098556935787201, 0.13190816342830658, 0.41221585869789124, 0.22275245189666748, 0.2862115800380707, 0.29398587346076965, 0.2952936589717865, 0.1804482638835907, 0.40403494238853455, 0.1271446943283081, -0.1776515394449234, 0.3254966139793396, -0.2722633183002472, -0.24315351247787476, 0.4100196361541748, 0.62740558385849, -0.08831368386745453, -0.06562738120555878, 0.13427621126174927, -0.12808893620967865, 0.12494787573814392, 0.0890844389796257, 0.247712180018425, -0.28943362832069397, -0.4838736951351166, 0.3436744809150696, 0.1162116751074791, 0.1510484665632248, -0.12315115332603455, 0.05892585963010788, 0.05899565666913986, 0.17138387262821198, -0.2932756245136261, 0.26158586144447327, 0.3350566327571869, -0.2227061539888382, 0.13434632122516632, 0.0062896632589399815, 0.08987358212471008, -0.1818213015794754, 0.13155514001846313, -0.027880113571882248, 0.16018423438072205, -0.14802993834018707, -0.0957624614238739, 0.22563086450099945, 0.011602974496781826, 0.24657776951789856, 0.09906629472970963, -0.16215696930885315, -0.09483583271503448, -0.10899645835161209, -0.24236777424812317, -0.2443491518497467, 0.24999375641345978, 0.2372865527868271, -0.2415652871131897, 0.3187464475631714, -0.029161114245653152, 0.24501091241836548, 0.1847769170999527, -0.031999148428440094, -0.14745502173900604, 0.25958141684532166, -0.10221584886312485, 0.424021452665329, 0.4162791669368744, -0.188274547457695, -0.09184856712818146, -0.024418745189905167, -0.17938989400863647, -0.2249310165643692, -0.09151216596364975, 0.04073415696620941, -0.240535169839859, -0.2889058589935303, 0.07782135903835297, -0.13684917986392975, 0.2789805233478546, 0.06914211809635162, 0.24263975024223328, -0.0014099458931013942, -0.4101408123970032, 0.03285670280456543, -0.010167072527110577, -0.269529789686203, -0.22026406228542328, -0.030173251405358315, -0.09110955893993378, -0.16445393860340118, 0.2610430419445038, 0.24966926872730255, 0.1362549066543579, 0.5002054572105408, -0.0945059061050415, 0.2376812994480133, -0.20367708802223206, 0.03617650642991066, -0.42523878812789917, -0.09555714577436447, 0.35759806632995605, 0.3498214781284332, -0.20395493507385254, -0.22149436175823212, -0.5597004294395447, 0.10842265188694, 0.07068196684122086, -0.03556067496538162, -0.022770734503865242, -0.037244897335767746, 0.15114104747772217, -0.20522868633270264, 0.02430853061378002, -0.31279516220092773, 0.04790053516626358, -0.28003236651420593, -0.14217375218868256, 0.013601941987872124, -0.03400660678744316, -0.21538719534873962, -0.031970128417015076, -0.016684800386428833, 0.2605360746383667, -0.10019432753324509, 0.00802699662744999, 0.20574188232421875, -0.04609893262386322, -0.04037706181406975, 0.3128235936164856, -0.15650904178619385, -0.009524085558950901, -0.010018914006650448, 0.2053341418504715, 0.3080557882785797, -0.25814899802207947, -0.7119438052177429, 0.05824512243270874, 0.005247133318334818, -0.26256227493286133, -0.13132742047309875, 0.23769648373126984, 0.5601803064346313, -0.19879060983657837, 0.11458522826433182, 0.463533878326416, -0.18903332948684692, 0.0986376628279686, -0.1500820815563202, -0.15123161673545837, -0.12750139832496643, 0.12364066392183304, 0.2079048752784729, -0.006193924695253372, 0.07666529715061188, 0.5572873950004578, 0.16624747216701508, -0.12651027739048004, -0.3530397415161133, 0.07210400700569153, 0.2784351110458374, -0.3214610517024994, -0.09803232550621033, -0.10858321934938431, -0.564651608467102, 0.0625721663236618, -0.2864466905593872, 0.08351272344589233, 0.13649076223373413, -0.23094579577445984, -0.05809482932090759, -0.06265661865472794, -0.07724513113498688, -0.03654155880212784, 0.13375908136367798, -0.09502655267715454, -0.03345930948853493, 0.02000843733549118, 0.11398904025554657, 0.34533587098121643, -0.2902797758579254, 0.5099155902862549, 0.008631368167698383, 0.32485419511795044, -0.2755908966064453, -0.3000109791755676, 0.13371913135051727, 0.09635943174362183, 0.08106745034456253, -0.1677868515253067, -0.14541833102703094, 0.29110413789749146, 0.19663825631141663, 0.03105844184756279, 0.14562585949897766, -0.2731461226940155, 0.4598880112171173, 0.0589737594127655, -0.035778433084487915, -0.13375787436962128, 0.08443685621023178, -0.19134971499443054, -0.0462985523045063, 0.2320578247308731, -0.2428794652223587, -0.04818783327937126, 0.11958283931016922, 0.18583989143371582, 0.04129072278738022, -0.13352665305137634, -0.15656760334968567, 0.09857837855815887, 0.23880967497825623, 0.20704007148742676, -0.0013626334257423878, -0.043376024812459946, -0.34334415197372437, 0.19563604891300201, 0.46204379200935364, -0.005481822416186333, 0.20487989485263824, -0.06295570731163025, -0.06925434619188309, 0.0003484062326606363, 0.4217231273651123, 0.04121619090437889, 0.4216999113559723, 0.08651922643184662, -0.09068899601697922, 0.0007989125442691147, -0.034902870655059814, -0.3570866882801056, 0.16904474794864655, 0.14286892116069794, 0.21061287820339203, -0.1635545939207077, 0.2298244833946228, -0.036555781960487366, -0.3995288610458374, -0.10759498924016953, -0.08823610842227936, 0.13155031204223633, 0.134910449385643, 0.32000789046287537, -0.11699051409959793, -0.5118405222892761, -0.07171627879142761, -0.18816889822483063, -0.09109179675579071, -0.2565028965473175, -0.04709171503782272, 0.5061331987380981, 0.3253307342529297, 0.013788546435534954, 0.34854307770729065, 0.34163376688957214, -0.0021225805394351482, 0.3129972815513611, 0.11849908530712128, -0.04346577450633049, -0.28942444920539856, 0.07920368015766144, 0.4227302074432373, 0.24148355424404144, 0.21288242936134338, -0.3029240071773529, -0.15140984952449799, -0.14659254252910614, -0.2868509888648987, 0.29411792755126953, -0.19982601702213287, 0.2345810979604721, -0.023697806522250175, 0.19926971197128296, -0.5368195176124573, -0.12628576159477234, 0.1590670794248581, -0.010400688275694847, -0.08413945138454437, -0.052356019616127014, -0.016661860048770905, -0.44588083028793335, -0.20343199372291565, -0.17404311895370483, -0.39175131916999817, -0.24149514734745026, 0.1878700703382492, -0.11721895635128021, 0.1827535480260849, 0.23959913849830627, 0.06552764773368835, 0.40171757340431213, -0.03590831533074379, -0.24859145283699036, 0.01844647154211998, -0.12903641164302826, 0.20763593912124634, -0.32397714257240295, -0.5115101337432861, 0.10641656070947647, 0.1879417598247528, 0.003200552426278591, -0.269090861082077, -0.43559908866882324, -0.37786343693733215, 0.04990605264902115, 0.11913418024778366, 0.023840611800551414, -0.2214810848236084, 0.005934615153819323, -0.262151300907135, 0.04465436190366745, -0.010199661366641521, -0.27398860454559326, -0.11105107516050339, 0.07669281959533691, 0.3078038692474365, -0.19712676107883453, 0.11028679460287094, 0.031528323888778687, 0.25294429063796997, 0.3045099973678589, -0.11898044496774673, 0.18053847551345825, 0.1001780554652214, 0.21791480481624603, -0.2815535366535187, -0.4495451748371124, 0.33795133233070374, -0.16195085644721985, -0.06860939413309097, 0.3591623902320862, -0.09100767970085144, -0.5044241547584534, -0.17165987193584442, 0.07266765832901001, -0.26782119274139404, -0.42276066541671753, 0.0708649754524231, 0.44430264830589294, -0.14769385755062103, 0.11258655786514282, 0.21931618452072144, 0.019851617515087128, -0.22919775545597076, -0.3125333786010742, 0.3360981345176697, -0.2867804765701294, 0.0014976866077631712, 0.15573765337467194, -0.4171317219734192, -0.6907586455345154, 0.16188770532608032, -0.05036431550979614, 0.11516404896974564, -0.11067402362823486, 0.2906992435455322, 0.10505291819572449, 0.03925468772649765, 0.3141641616821289, 0.1167532429099083, -0.15859462320804596, 0.043577902019023895, -0.0749518871307373, -0.19444702565670013, 0.03445063903927803, -0.4641059339046478, 0.5842936038970947, 0.7916568517684937, 0.39100465178489685, -0.04272050783038139, 0.14425745606422424, 0.5085098743438721, -0.03295798599720001, 0.15193398296833038, 0.08288562297821045, -0.1406870186328888, -0.3030998110771179, -0.12402430921792984, 0.05947885662317276, -0.027399616315960884, -0.06065034493803978, 0.03376628831028938, -0.20056161284446716, 0.30888572335243225, -0.07161439210176468, -0.02678716741502285, 0.5213574171066284, -0.019524168223142624, 0.08624453842639923, 0.10341482609510422, -0.10577677190303802, -0.11154305189847946, 0.4208716154098511, 0.3046403229236603, -0.20926716923713684, -0.26748088002204895, 0.2845481038093567, -0.49008163809776306, 0.09635606408119202, 0.5848206877708435, -0.010691383853554726, -0.031013529747724533, -0.173979252576828, 0.005066107027232647, 0.3778679668903351, 0.05310966074466705, 0.40148675441741943, 0.1770641803741455, -0.02953946217894554, 0.1776052564382553, 0.257417768239975, -0.09426901489496231, 0.018869956955313683, 0.014036518521606922, -0.14121687412261963, -0.23207958042621613, 0.4488869309425354, 0.07712177187204361, 0.8232890367507935, -0.15118750929832458, 0.06643517315387726, 0.17773118615150452, 0.005394114181399345, 0.5558367371559143, -0.015880292281508446, 0.22724485397338867, -0.13967911899089813, -0.4450143277645111, -0.022555632516741753, -0.10461302846670151, 0.28905540704727173, -0.02663286216557026, -0.7058285474777222, 0.21495872735977173, 0.26579418778419495, 0.152628093957901, 0.020415710285305977, 0.05576921999454498, -0.14507761597633362, -0.2428600937128067, -0.5719939470291138, 0.2210027426481247, -0.1751994639635086, 0.23505178093910217, -0.28548693656921387, 0.030615437775850296, -0.2551967203617096, 0.04343552514910698, -0.4520660638809204, 0.0669051930308342, -0.48229414224624634, 0.01302406843751669, 0.1449773907661438, -0.2509017288684845, 0.0438167080283165, 0.3447951376438141, 0.387530118227005, 0.17426331341266632, -0.04343319684267044, -0.01864875666797161, -0.03557238727807999, -0.30340930819511414, -0.06309951096773148, -0.016941864043474197, 0.005116539541631937, -0.2645793557167053, -0.06806758791208267, -0.07059979438781738, 0.30341148376464844, -0.149382084608078, 0.35796964168548584, 0.17621520161628723, 0.12822259962558746, -0.0667857676744461, -0.05379918962717056, -0.11600887775421143, -0.17481020092964172, -0.2913317084312439, 0.13214072585105896, -0.12050139904022217, -0.005069730803370476, 0.29780760407447815, 0.18357300758361816, -0.12276756018400192, -0.03555604815483093, 0.29132315516471863, -0.2046392858028412, 0.10926849395036697, 0.2187437266111374, 0.257244348526001, -0.12845048308372498, -0.28012746572494507, 0.3003308176994324, -0.49323758482933044, -0.17429210245609283, 0.11607957631349564, -0.17402034997940063, -0.12756571173667908, -0.1307649314403534, 0.640363335609436, 0.4149587154388428, -0.010212292894721031, -0.06361141055822372, -0.36319300532341003, -0.20702525973320007, 0.20271000266075134, 0.2500632107257843, 0.025946613401174545, 0.22087998688220978, 0.19487746059894562, 0.21906648576259613, 0.06693252176046371, -0.2703333795070648, 0.3732132017612457, -0.4460934102535248, 0.05523935705423355, -0.03002094104886055, -0.2469731718301773, 0.23904557526111603, -0.1518963873386383, -0.033737555146217346, 0.5021684765815735, 0.06566688418388367, -0.2611507773399353, -0.06452959030866623, 0.06806589663028717, 0.008530630730092525, -0.12818129360675812, -0.26094186305999756, 0.004728295374661684, -0.030353548005223274, -0.06656641513109207, 0.1719762682914734, -0.05886862426996231, 0.27002182602882385, -0.12606291472911835, 0.33104458451271057, 0.20877602696418762, 0.15549905598163605, 0.08469528704881668, 0.013366960920393467, 0.20232300460338593, 0.12882868945598602, 0.1213655173778534, 0.15800145268440247, -0.07078959047794342, 0.2570587694644928, 0.024878190830349922, -0.0761776939034462, -0.00784143153578043, 0.3335093557834625, -0.33863362669944763, 0.15942133963108063, 0.0026587408501654863, -0.1664644181728363, 0.35182270407676697, -0.012282634153962135, -0.12231257557868958, 0.1025400385260582, 0.13919833302497864, -0.10381652414798737, -0.05152597650885582, 0.5680354833602905, 0.20062652230262756, 0.030121691524982452, 0.5001781582832336, 0.0856856182217598, -0.040030963718891144, 0.21090205013751984, 0.04401477053761482, 0.31992045044898987, 0.4771459102630615, 0.07485100626945496, -0.11321007460355759, 0.02573900856077671, 0.23483602702617645, 0.119077168405056, 0.05644070729613304, 0.16241006553173065, -0.045043546706438065, -0.05026019364595413, 0.1110038161277771, -0.17792028188705444, 0.08935945481061935, 0.4992899000644684, -0.10340747237205505, -0.2284979522228241, 0.042943719774484634, 0.10159790515899658, 0.027173789218068123, 0.12443218380212784, 0.7389979362487793, -0.4879820644855499, -0.2675541937351227, -0.19579735398292542, 0.14382757246494293, -0.42155125737190247, 0.08674067258834839, -0.31011196970939636, 0.06441112607717514, -0.09468219429254532, -0.001960538327693939, -0.1672723889350891, -0.057342562824487686, 0.39590680599212646, 0.2621210515499115, -0.19068565964698792, 0.03612079471349716, -0.1225869357585907, 0.10783549398183823, -0.1399857997894287, -0.012021173723042011, 0.4424280524253845, -0.30636051297187805, 0.26280564069747925, 0.01434061024338007, 0.607215940952301, 0.18633602559566498, 0.20774558186531067, -0.10181638598442078, -0.2245774120092392, 0.06042488291859627, -0.08757278323173523, 0.012050090357661247, -0.055068276822566986, 0.44311821460723877, 0.08779612183570862, 0.26833054423332214, 0.08101984113454819, -0.1402062326669693, 0.2214479297399521, 0.12705087661743164, -0.18602342903614044, 0.03443298488855362, -0.29817211627960205, -0.09205471724271774, -0.14600889384746552, -0.2923966646194458, -0.17628195881843567, -0.3040124773979187, 0.46308040618896484, -0.23748058080673218, -0.15882515907287598, 0.09287690371274948, 0.026172177866101265, 0.07179015129804611, 0.00036960927536711097, 0.5908599495887756, 0.32058534026145935, -0.0013036095770075917, -0.2988019287586212, -0.505959153175354, -0.3325136601924896, 0.19119028747081757, -0.16133596003055573, 0.13938631117343903, 0.0850100889801979, 0.18678312003612518, -0.023113863542675972, 0.45860812067985535, 0.16402588784694672, -0.08083473891019821, -0.21005991101264954, -0.06280995160341263, -0.5517159104347229, -0.13538356125354767, 0.2651447355747223, -0.2606634199619293, -0.16066469252109528, -0.16178712248802185, 0.2897636890411377, -0.15786424279212952, 0.008336832746863365, -0.04479416087269783, -0.2731548547744751, -0.3319765031337738, -0.24862343072891235, 0.15242038667201996, 0.12959279119968414, 0.1229705810546875, 0.007585069164633751, -0.333141952753067, 0.12094756960868835, -0.28892117738723755, -0.1068279892206192, 0.011426792480051517, 0.21625438332557678, 0.25030094385147095, 0.028534671291708946, 0.26356276869773865, -0.09871052950620651, 0.49516594409942627, 0.15883544087409973, 0.02855929546058178, -0.16628432273864746, -0.08736907690763474, -0.07266167551279068, 0.059975843876600266, -0.04567224904894829, 0.40819424390792847, -0.3233315944671631, 0.2676670253276825, -0.40840721130371094, -0.27963095903396606, 0.3924647569656372, -0.3721633851528168, -0.07345762848854065, -0.017433730885386467, 0.21171969175338745, 0.1328282356262207, -0.13608190417289734, -0.3291107714176178, -0.07494794577360153, -0.1162363588809967, -0.290026992559433, -0.45486053824424744, 0.26036718487739563, -0.07690273970365524, 0.06409204751253128, -0.10680948942899704, 0.3408884108066559, 0.12940163910388947, -0.2829599678516388, -0.16729772090911865, -0.21795837581157684 ]
https://github.com/huggingface/datasets/issues/222
Colab Notebook breaks when downloading the squad dataset
This trick from @thomwolf seems to be the most reliable solution to fix this colab notebook issue: ```python # install nlp !pip install -qq nlp==0.2.0 # Make sure that we have a recent version of pyarrow in the session before we continue - otherwise reboot Colab to activate it import pyarrow if int(pyarrow.__version__.split('.')[1]) < 16: import os os.kill(os.getpid(), 9) ```
When I run the notebook in Colab https://colab.research.google.com/github/huggingface/nlp/blob/master/notebooks/Overview.ipynb breaks when running this cell: ![image](https://user-images.githubusercontent.com/338917/83311709-ffd1b800-a1dd-11ea-8394-3a87df0d7f8b.png)
60
Colab Notebook breaks when downloading the squad dataset When I run the notebook in Colab https://colab.research.google.com/github/huggingface/nlp/blob/master/notebooks/Overview.ipynb breaks when running this cell: ![image](https://user-images.githubusercontent.com/338917/83311709-ffd1b800-a1dd-11ea-8394-3a87df0d7f8b.png) This trick from @thomwolf seems to be the most reliable solution to fix this colab notebook issue: ```python # install nlp !pip install -qq nlp==0.2.0 # Make sure that we have a recent version of pyarrow in the session before we continue - otherwise reboot Colab to activate it import pyarrow if int(pyarrow.__version__.split('.')[1]) < 16: import os os.kill(os.getpid(), 9) ```
[ -0.21085861325263977, 0.31306004524230957, -0.049380555748939514, 0.03972147777676582, -0.08726848661899567, -0.18556547164916992, 0.2631244361400604, 0.15955470502376556, -0.4475891888141632, 0.10873615741729736, -0.36074137687683105, 0.5732998847961426, 0.19588497281074524, 0.08197644352912903, 0.17324349284172058, 0.19502787292003632, -0.008885027840733528, 0.39630353450775146, -0.0574667751789093, 0.1644325852394104, -0.2231677919626236, 0.34456774592399597, -0.3933258354663849, 0.10321398824453354, -0.22195515036582947, -0.19702434539794922, -0.15914255380630493, 0.29760849475860596, -0.3919450044631958, -0.21676313877105713, 0.27606290578842163, 0.171592578291893, -0.05999332666397095, 0.17929427325725555, -0.00011473387712612748, -0.20161810517311096, -0.030263392254710197, -0.07803525030612946, -0.37296247482299805, -0.23220141232013702, 0.0352734811604023, -0.33603212237358093, 0.3040676414966583, -0.4480278491973877, 0.13219597935676575, 0.457571804523468, 0.28862401843070984, 0.35761648416519165, 0.2969406843185425, 0.27599769830703735, 0.1642182320356369, 0.3264269530773163, 0.13450834155082703, 0.0051154508255422115, 0.23249927163124084, -0.253113329410553, -0.2693358063697815, 0.42635276913642883, 0.5949574112892151, -0.2634858787059784, -0.13266980648040771, 0.10558602958917618, -0.16910985112190247, 0.15007245540618896, 0.11090785264968872, 0.22836945950984955, -0.3441576063632965, -0.4259018301963806, 0.22296011447906494, 0.060684092342853546, 0.02652840130031109, -0.2624766528606415, 0.05482767894864082, 0.00587764848023653, 0.2763140797615051, -0.30670446157455444, 0.1578662097454071, 0.35337406396865845, -0.37504372000694275, 0.11998128890991211, 0.024275969713926315, 0.03381447121500969, -0.19935056567192078, 0.18583522737026215, 0.05883850157260895, 0.34322941303253174, -0.024360867217183113, -0.013874880969524384, 0.21716885268688202, 0.08212471753358841, 0.34461045265197754, 0.21561779081821442, -0.03924629092216492, -0.00300064985640347, -0.1363028585910797, -0.2019573450088501, -0.1412385106086731, 0.23442406952381134, 0.23353025317192078, -0.32346588373184204, 0.28888341784477234, 0.006386871449649334, 0.291126012802124, 0.19484837353229523, -0.058985672891139984, -0.11289190500974655, 0.21247808635234833, -0.1368216872215271, 0.3969639539718628, 0.3778131902217865, -0.10544412583112717, 0.0028237695805728436, -0.044340550899505615, -0.1690577268600464, -0.1508910208940506, -0.10000840574502945, 0.12246556580066681, -0.17493809759616852, -0.28056758642196655, 0.023782242089509964, -0.2251797765493393, 0.2988462746143341, 0.04263237118721008, 0.24172692000865936, -0.06156959384679794, -0.37630075216293335, 0.06695275008678436, -0.025022301822900772, -0.342669278383255, -0.20411615073680878, -0.002302370499819517, -0.019037814810872078, -0.16797620058059692, 0.09720977395772934, 0.3000754117965698, 0.16170348227024078, 0.40994635224342346, -0.11885719001293182, 0.17195625603199005, -0.10903291404247284, 0.03236806392669678, -0.29709795117378235, -0.09300390630960464, 0.2763929069042206, 0.38804471492767334, -0.24145789444446564, -0.22902123630046844, -0.5571026802062988, 0.04567011818289757, 0.16808415949344635, -0.01252466905862093, -0.00839775800704956, -0.1567080169916153, 0.14246954023838043, -0.27026307582855225, -0.048106614500284195, -0.3073243498802185, 0.010950199328362942, -0.25604739785194397, -0.18128392100334167, 0.03165725991129875, -0.07667271792888641, -0.1764993965625763, -0.07915328443050385, 0.0010234855581074953, 0.19804790616035461, -0.1631060093641281, -0.03464553877711296, 0.12622420489788055, 0.03138430416584015, 0.047351155430078506, 0.23515069484710693, -0.18265122175216675, 0.054192036390304565, 0.02884005196392536, 0.2262764871120453, 0.41118142008781433, -0.20428447425365448, -0.6995335817337036, -0.007170574273914099, -0.08847211301326752, -0.24175718426704407, -0.13286618888378143, 0.24760857224464417, 0.4465605914592743, -0.14351379871368408, 0.21046629548072815, 0.5739885568618774, -0.16754527390003204, 0.14042899012565613, -0.22272223234176636, -0.17239806056022644, -0.06703212857246399, 0.18465404212474823, 0.14324404299259186, -0.10967379063367844, 0.011931713670492172, 0.6371195912361145, 0.23053820431232452, -0.11815875768661499, -0.23211504518985748, -0.004396210424602032, 0.266366571187973, -0.30782315135002136, -0.10001255571842194, -0.05420028045773506, -0.5955294370651245, 0.030195750296115875, -0.2725133001804352, 0.19959977269172668, -0.012548255734145641, -0.1999819576740265, -0.032262708991765976, 0.0027677142061293125, -0.02147478237748146, -0.041503798216581345, 0.14903366565704346, -0.10393455624580383, 0.03776148334145546, 0.016410430893301964, 0.13712584972381592, 0.33913031220436096, -0.23257866501808167, 0.4468238651752472, -0.003963978961110115, 0.2574145197868347, -0.28839758038520813, -0.3253236711025238, 0.1558016985654831, 0.24388819932937622, 0.030585385859012604, -0.12488038092851639, -0.06957951188087463, 0.21655449271202087, 0.1029067263007164, -0.08505760878324509, 0.06705182045698166, -0.33819106221199036, 0.37754982709884644, 0.0862097442150116, -0.05293237045407295, -0.07910645753145218, 0.07668917626142502, -0.07673238962888718, 0.018533140420913696, 0.22335097193717957, -0.3123072385787964, -0.07031028717756271, 0.1040065661072731, 0.2196219265460968, 0.06820753216743469, -0.13731420040130615, -0.1277415156364441, 0.21088215708732605, 0.22927701473236084, 0.24448943138122559, 0.015721233561635017, 0.038441430777311325, -0.365306556224823, 0.27149054408073425, 0.4710390865802765, 0.009044397622346878, 0.28053152561187744, -0.0276138037443161, -0.06783606112003326, -0.01764831319451332, 0.37193480134010315, 0.023189768195152283, 0.3294188678264618, 0.11665694415569305, -0.11657097190618515, 0.011138915084302425, -0.10210765898227692, -0.29719653725624084, 0.1116592213511467, 0.12786372005939484, 0.31787392497062683, -0.10101091861724854, 0.3305140733718872, -0.03437226638197899, -0.34309184551239014, -0.20008893311023712, -0.09329705685377121, 0.24215565621852875, 0.1284601092338562, 0.2591359615325928, -0.1354101151227951, -0.6098309755325317, -0.05507965758442879, -0.16726858913898468, -0.09036113321781158, -0.22800911962985992, -0.023899033665657043, 0.4561629593372345, 0.27615320682525635, 0.0735805481672287, 0.29388198256492615, 0.2140803337097168, 0.03674903139472008, 0.08619970083236694, 0.15598119795322418, -0.11963631957769394, -0.3185093402862549, 0.07188037782907486, 0.39522644877433777, 0.2435242235660553, 0.08686131983995438, -0.24151480197906494, -0.18731831014156342, -0.09849010407924652, -0.3265630602836609, 0.20327259600162506, -0.23567287623882294, 0.305990606546402, -0.017747370526194572, 0.23887331783771515, -0.6264249086380005, -0.21012341976165771, 0.102808378636837, -0.12800568342208862, -0.11295291036367416, -0.038872890174388885, -0.005858159624040127, -0.36101648211479187, -0.24185672402381897, -0.14751099050045013, -0.5018797516822815, -0.24258741736412048, 0.2574925124645233, -0.0037880917079746723, 0.19984523952007294, 0.2799191474914551, 0.07124092429876328, 0.31753048300743103, 0.0995088666677475, -0.23881179094314575, 0.06216589733958244, 0.022067038342356682, 0.23596280813217163, -0.24667267501354218, -0.5061872601509094, 0.06595058739185333, 0.05682883411645889, 0.03246128559112549, -0.2706150412559509, -0.40027639269828796, -0.46518129110336304, 0.1409190148115158, 0.23564451932907104, -0.0490809828042984, -0.13650156557559967, 0.13784299790859222, -0.18993276357650757, 0.0716857761144638, 0.015007883310317993, -0.23039905726909637, -0.17598633468151093, 0.13017944991588593, 0.283284068107605, -0.2543376088142395, 0.04595978558063507, 0.15705278515815735, 0.22293885052204132, 0.11176846921443939, -0.08313345909118652, 0.15639039874076843, 0.09553366899490356, 0.1974157989025116, -0.1802193522453308, -0.42029011249542236, 0.33444926142692566, -0.11916878819465637, -0.09323793649673462, 0.34363627433776855, -0.15984097123146057, -0.6042137145996094, -0.12987609207630157, 0.0671723410487175, -0.3866874873638153, -0.46388736367225647, 0.16150765120983124, 0.5007522106170654, -0.13982784748077393, 0.25034603476524353, 0.18707406520843506, 0.008182377554476261, -0.34842967987060547, -0.37131014466285706, 0.22920924425125122, -0.26665517687797546, 0.04256046190857887, 0.2036813497543335, -0.4245845675468445, -0.671054482460022, 0.1856251060962677, 0.07730073481798172, 0.1294669657945633, 0.03373365476727486, 0.16176460683345795, 0.14359386265277863, 0.010684072971343994, 0.20230939984321594, 0.23272551596164703, -0.0438249297440052, 0.08600233495235443, 0.05781114473938942, -0.2841724753379822, 0.02707941271364689, -0.39344522356987, 0.5534505248069763, 0.8214972019195557, 0.21728621423244476, -0.12233015149831772, 0.1616934835910797, 0.5125229954719543, -0.02566499076783657, 0.16802288591861725, 0.13292120397090912, -0.22827917337417603, -0.3713432550430298, -0.14193807542324066, 0.10894763469696045, -0.05841979384422302, 0.07994838804006577, 0.09687230736017227, -0.069400854408741, 0.24518509209156036, -0.07987092435359955, -0.05815791338682175, 0.5733702778816223, -0.04309709742665291, 0.17280927300453186, 0.07421620935201645, -0.1856633722782135, -0.13516667485237122, 0.3057752847671509, 0.2549995481967926, -0.04839896783232689, -0.23486243188381195, 0.22553203999996185, -0.31759563088417053, 0.16910557448863983, 0.5100574493408203, -0.028745876625180244, -0.008707603439688683, -0.26932424306869507, 0.009692074730992317, 0.37018486857414246, 0.08201959729194641, 0.2566707730293274, 0.07617219537496567, -0.08871478587388992, 0.20067395269870758, 0.23740258812904358, -0.12386104464530945, 0.024681931361556053, -0.15152467787265778, -0.19545307755470276, -0.20539431273937225, 0.43531280755996704, 0.09283686429262161, 0.7878345251083374, -0.23235654830932617, 0.161381334066391, 0.2782195508480072, 0.02919670380651951, 0.6408743262290955, -0.03151613846421242, 0.33673030138015747, -0.18474169075489044, -0.2963068187236786, 0.05368519201874733, -0.021155234426259995, 0.1532122790813446, 0.014196829870343208, -0.6053224802017212, 0.23526068031787872, 0.3834214508533478, 0.02982729859650135, 0.08889079093933105, 0.13620369136333466, -0.11280205845832825, -0.32899412512779236, -0.6385911107063293, 0.1790301501750946, -0.3138791620731354, 0.19821497797966003, -0.30542823672294617, 0.0033893792424350977, -0.25723743438720703, 0.05614417791366577, -0.3699455261230469, 0.15100449323654175, -0.5142829418182373, -0.04811447486281395, 0.2063174545764923, -0.30066680908203125, -0.10240282863378525, 0.3006395101547241, 0.17832356691360474, 0.20467104017734528, 0.06039503589272499, 0.06155625358223915, -0.058590568602085114, -0.30445775389671326, 0.0837138444185257, 0.03717808425426483, 0.07622430473566055, -0.20884190499782562, -0.08287814259529114, -0.07315327972173691, 0.27861639857292175, -0.23776978254318237, 0.3863205313682556, 0.2796514928340912, 0.18206222355365753, 0.020237624645233154, -0.0653483048081398, -0.03768164664506912, -0.17243234813213348, -0.28082212805747986, 0.10838403552770615, -0.1088259220123291, -0.06522026658058167, 0.2896266281604767, 0.14065705239772797, -0.1736132651567459, -0.003592601977288723, 0.27035456895828247, -0.17901048064231873, 0.1063859760761261, 0.20714977383613586, 0.2732163369655609, -0.11991070210933685, -0.21892830729484558, 0.1924758404493332, -0.5636963248252869, -0.1794547438621521, 0.02946115843951702, -0.07871992886066437, -0.02636171504855156, -0.08183631300926208, 0.6315823793411255, 0.30250149965286255, 0.025656579062342644, -0.07841416448354721, -0.40588271617889404, -0.1844428926706314, 0.20544080436229706, 0.17861147224903107, -0.1519642174243927, 0.3767312169075012, 0.33319732546806335, 0.19334232807159424, 0.048748236149549484, -0.23890827596187592, 0.29427918791770935, -0.4090551733970642, 0.16265296936035156, -0.05745452642440796, -0.31670522689819336, 0.1807575672864914, -0.16345800459384918, -0.0068308706395328045, 0.5193843841552734, 0.1276141107082367, -0.2376260906457901, -0.08323689550161362, 0.06300925463438034, 0.020876627415418625, -0.10651818662881851, -0.3156403601169586, -0.07924584299325943, -0.03797832876443863, -0.03086882457137108, 0.09227948635816574, -0.11091037839651108, 0.21215200424194336, -0.04508569836616516, 0.25605422258377075, 0.22376392781734467, 0.1853877753019333, 0.0336076095700264, -0.026624931022524834, 0.2572001814842224, 0.10353605449199677, 0.20876544713974, 0.17264555394649506, 0.038589559495449066, 0.22618462145328522, -0.04137681424617767, -0.16636726260185242, 0.07988878339529037, 0.31233179569244385, -0.3684277832508087, 0.06631996482610703, -0.12071537971496582, -0.08334951847791672, 0.39146047830581665, 0.09454567730426788, -0.22146672010421753, 0.24221819639205933, 0.12057948112487793, -0.20341120660305023, 0.059440746903419495, 0.5620881915092468, 0.19853399693965912, 0.11759109050035477, 0.5118358135223389, 0.0033112105447798967, -0.00012639109627343714, 0.1494714915752411, 0.11183463782072067, 0.2903251349925995, 0.4005499482154846, -0.059446487575769424, -0.10892564803361893, -0.04077715054154396, 0.10463763773441315, 0.12247217446565628, 0.10486497730016708, 0.18972009420394897, -0.0031352308578789234, -0.08938498795032501, 0.09117848426103592, -0.21281373500823975, 0.07399344444274902, 0.3456501066684723, -0.07070387154817581, -0.2760522663593292, 0.061198823153972626, 0.22912675142288208, 0.062462594360113144, 0.0791783481836319, 0.7251800298690796, -0.3464801013469696, -0.22814327478408813, -0.02409789338707924, 0.16770732402801514, -0.3900534212589264, 0.11122927069664001, -0.25787633657455444, 0.103654645383358, -0.07326433807611465, 0.066691093146801, -0.15407682955265045, -0.13330821692943573, 0.35017332434654236, 0.29399898648262024, -0.17318843305110931, -0.024733027443289757, -0.10646089166402817, 0.16021236777305603, -0.10827048122882843, -0.023033371195197105, 0.38537707924842834, -0.34098494052886963, 0.1951875239610672, -0.1848583221435547, 0.5654686689376831, 0.1751050055027008, 0.18691018223762512, -0.25897058844566345, -0.1922297328710556, -0.04654531180858612, -0.009189892560243607, -0.1291232705116272, -0.10915734618902206, 0.4811042845249176, 0.17956197261810303, 0.25972673296928406, 0.07102730870246887, -0.139082133769989, 0.2577846944332123, 0.19828630983829498, -0.28091341257095337, 0.14965374767780304, -0.2773092985153198, 0.04808846116065979, -0.1703658252954483, -0.2510802149772644, -0.10992174595594406, -0.26709237694740295, 0.4196454882621765, -0.26302987337112427, -0.08880841732025146, 0.08305705338716507, -0.06837189197540283, 0.06529135257005692, 0.03694263845682144, 0.5547953844070435, 0.3388822674751282, 0.017960943281650543, -0.2560867965221405, -0.4043077528476715, -0.396300345659256, 0.22965623438358307, -0.22898265719413757, -0.07774502038955688, 0.08669266104698181, 0.2964426875114441, -0.05007990077137947, 0.49343326687812805, 0.014593523927032948, 0.044202107936143875, -0.1833363026380539, -0.0094634173437953, -0.5380405187606812, -0.06271463632583618, 0.2135578691959381, -0.33248093724250793, -0.07470705360174179, -0.25559201836586, 0.2263031154870987, -0.34457752108573914, 0.011229896917939186, -0.02247164398431778, -0.37018170952796936, -0.24126404523849487, -0.08207903057336807, 0.17296883463859558, 0.07968755066394806, 0.06808504462242126, 0.012300046160817146, -0.2333475947380066, 0.03424548730254173, -0.2586528956890106, -0.20849592983722687, 0.04084810987114906, 0.20121580362319946, 0.2290244847536087, 0.08602162450551987, 0.3633444905281067, -0.09533372521400452, 0.5152751207351685, 0.09685762971639633, 0.04516284167766571, -0.04169957712292671, -0.02424672245979309, -0.12852980196475983, 0.13830940425395966, -0.22996185719966888, 0.4407966136932373, -0.3113744556903839, 0.2767091393470764, -0.4855019748210907, -0.34006908535957336, 0.3633520305156708, -0.5322362780570984, -0.18764877319335938, 0.04002312198281288, 0.22818858921527863, 0.08981290459632874, -0.17658764123916626, -0.2977270185947418, -0.024630555883049965, -0.11291196942329407, -0.23598967492580414, -0.4163486957550049, 0.23685169219970703, -0.18495258688926697, 0.04134831577539444, -0.09677967429161072, 0.2848268151283264, 0.03882342576980591, -0.24422185122966766, -0.21315838396549225, -0.2190767526626587 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
I like this feature! I think the first question we should decide on is how to convert all datasets into the same format. In T5, the authors decided to format every dataset into a text-to-text format. If the dataset had "multiple" inputs like MNLI, the inputs were concatenated. So in MNLI the input: > - **Hypothesis**: The St. Louis Cardinals have always won. > > - **Premise**: yeah well losing is i mean i’m i’m originally from Saint Louis and Saint Louis Cardinals when they were there were uh a mostly a losing team but was flattened to a single input: > mnli hypothesis: The St. Louis Cardinals have always won. premise: > yeah well losing is i mean i’m i’m originally from Saint Louis and Saint Louis Cardinals > when they were there were uh a mostly a losing team but. This flattening is actually a very simple operation in `nlp` already. You would just need to do the following: ```python def flatten_inputs(example): return {"input": "mnli hypothesis: " + example['hypothesis'] + " premise: " + example['premise']} t5_ready_mnli_ds = mnli_ds.map(flatten_inputs, remove_columns=[<all columns except output>]) ``` So I guess converting the datasets into the same format can be left to the user for now. Then the question is how we can merge the datasets. I would probably be in favor of a simple ```python dataset.add() ``` function that checks if the dataset is of the same format and if yes merges the two datasets. Finally, how should the sampling be implemented? **Examples-proportional mixing** corresponds to just merging the datasets and shuffling. For the other two sampling approaches we would need some higher-level features, maybe even a `dataset.sample()` function for merged datasets. What are your thoughts on this @thomwolf @lhoestq @ghomasHudson @enzoampil ?
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
291
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. I like this feature! I think the first question we should decide on is how to convert all datasets into the same format. In T5, the authors decided to format every dataset into a text-to-text format. If the dataset had "multiple" inputs like MNLI, the inputs were concatenated. So in MNLI the input: > - **Hypothesis**: The St. Louis Cardinals have always won. > > - **Premise**: yeah well losing is i mean i’m i’m originally from Saint Louis and Saint Louis Cardinals when they were there were uh a mostly a losing team but was flattened to a single input: > mnli hypothesis: The St. Louis Cardinals have always won. premise: > yeah well losing is i mean i’m i’m originally from Saint Louis and Saint Louis Cardinals > when they were there were uh a mostly a losing team but. This flattening is actually a very simple operation in `nlp` already. You would just need to do the following: ```python def flatten_inputs(example): return {"input": "mnli hypothesis: " + example['hypothesis'] + " premise: " + example['premise']} t5_ready_mnli_ds = mnli_ds.map(flatten_inputs, remove_columns=[<all columns except output>]) ``` So I guess converting the datasets into the same format can be left to the user for now. Then the question is how we can merge the datasets. I would probably be in favor of a simple ```python dataset.add() ``` function that checks if the dataset is of the same format and if yes merges the two datasets. Finally, how should the sampling be implemented? **Examples-proportional mixing** corresponds to just merging the datasets and shuffling. For the other two sampling approaches we would need some higher-level features, maybe even a `dataset.sample()` function for merged datasets. What are your thoughts on this @thomwolf @lhoestq @ghomasHudson @enzoampil ?
[ -0.06062903627753258, -0.4335702657699585, -0.05662935599684715, -0.015486029908061028, -0.16559092700481415, 0.04524800553917885, 0.18558642268180847, 0.18404991924762726, 0.20416028797626495, -0.14907842874526978, -0.2486170083284378, 0.2844129800796509, -0.18000783026218414, 0.47155192494392395, 0.38028332591056824, -0.4318713843822479, 0.07262514531612396, -0.14835745096206665, -0.3279532492160797, 0.34652191400527954, 0.04343181475996971, 0.12348002195358276, -0.18072380125522614, 0.0299021378159523, -0.35130956768989563, -0.1616334766149521, -0.38507139682769775, 0.07486508786678314, 0.014117106795310974, 0.03146897628903389, 0.06988030672073364, 0.39470648765563965, -0.04164378345012665, 0.18248321115970612, -0.00011498481762828305, -0.1062702164053917, 0.11025461554527283, -0.10874617844820023, 0.09373293071985245, -0.03407808765769005, 0.13381743431091309, -0.3790651261806488, 0.0580892339348793, -0.16541606187820435, 0.20267705619335175, 0.1457466334104538, 0.19343526661396027, 0.051303885877132416, 0.367055743932724, -0.21853314340114594, 0.08402930945158005, 0.264430969953537, -0.13485004007816315, 0.3218308687210083, -0.06517403572797775, 0.02047020196914673, 0.0448855496942997, 0.2399408519268036, 0.6195868849754333, -0.4785766303539276, -0.027776921167969704, 0.22522149980068207, -0.011841867119073868, -0.01231606025248766, 0.12501268088817596, -0.06387169659137726, -0.2911248803138733, -0.23895953595638275, -0.3664337396621704, 0.24455289542675018, -0.04346790909767151, -0.21901525557041168, -0.22686216235160828, -0.5487819910049438, -0.02522245980799198, -0.016629504039883614, -0.3268791437149048, 0.06366660445928574, -0.16956321895122528, -0.0531393326818943, -0.16527917981147766, -0.07770996540784836, -0.0644056648015976, 0.13073258101940155, 0.29616186022758484, 0.54425448179245, 0.03871045261621475, 0.05356590077280998, 0.29034173488616943, 0.03676622360944748, -0.11274208873510361, -0.27517837285995483, 0.3076985478401184, 0.07872676849365234, -0.3209530711174011, -0.3482196033000946, 0.0695451870560646, -0.3125922679901123, 0.3218032717704773, 0.0821099802851677, 0.13770300149917603, 0.06355174630880356, -0.13061770796775818, 0.21757495403289795, 0.3832533359527588, 0.008566953241825104, -0.07508304715156555, -0.1686057150363922, 0.007640494499355555, -0.14267051219940186, 0.16614370048046112, 0.1513284593820572, 0.11848543584346771, -0.003999081440269947, -0.4497523605823517, -0.062116317451000214, -0.0710136741399765, 0.05218537896871567, -0.125784233212471, -0.5080181360244751, -0.11419256031513214, -0.09674448519945145, -0.11398378014564514, -0.17604196071624756, -0.26717662811279297, 0.17692509293556213, 0.0036277135368436575, 0.271809458732605, -0.16389134526252747, 0.039084356278181076, -0.03798045590519905, 0.01682962477207184, -0.4351600408554077, -0.08235824108123779, 0.2814766466617584, 0.08760984987020493, 0.05325400456786156, 0.1616659313440323, -0.041585758328437805, 0.06550058722496033, 0.33225369453430176, 0.02868466079235077, 0.13818179070949554, -0.04580089449882507, 0.17860114574432373, -0.26701077818870544, -0.010550275444984436, 0.15709304809570312, -0.3184621036052704, -0.12241021543741226, -0.15247073769569397, -0.20893213152885437, 0.05303450673818588, 0.010931397788226604, -0.17490306496620178, -0.2891993820667267, -0.45561420917510986, 0.724881649017334, 0.015413427725434303, -0.08319629728794098, -0.19418594241142273, 0.13860651850700378, -0.15149907767772675, -0.0932871624827385, 0.22360262274742126, 0.11550799757242203, -0.07817365229129791, -0.2775519788265228, 0.019100885838270187, -0.08028440177440643, 0.012737736105918884, 0.3045516610145569, -0.06466145068407059, 0.25489547848701477, 0.06150002405047417, 0.20962871611118317, 0.30092233419418335, -0.39963504672050476, -0.00243769446387887, 0.34459447860717773, -0.055489517748355865, 0.38467511534690857, 0.10268256068229675, 0.1290329247713089, 0.002214512089267373, -0.1920386701822281, 0.2654210329055786, 0.9147469401359558, -0.3811213970184326, 0.06711094081401825, -0.1982174664735794, -0.3537554442882538, 0.5127950310707092, 0.2841843068599701, 0.011131275445222855, -0.34456244111061096, -0.13308767974376678, -0.004994895774871111, 0.07843354344367981, 0.0030070876237004995, 0.037370651960372925, -0.02651955932378769, -0.3109759986400604, 0.07546776533126831, -0.19495461881160736, -0.11524031311273575, -0.4247467815876007, -0.10035908222198486, 0.04577982425689697, 0.3210239112377167, 0.2605731785297394, -0.15339244902133942, 0.22681379318237305, -0.308633416891098, -0.15567155182361603, -0.2565403878688812, 0.133099764585495, -0.02675182744860649, -0.014473753049969673, -0.20608538389205933, -0.0964525043964386, 0.32367777824401855, 0.057791974395513535, -0.0758986547589302, -0.49590304493904114, 0.24824398756027222, -0.07776182144880295, 0.020057570189237595, 0.07954919338226318, 0.6802397966384888, -0.1879967302083969, -0.07588937133550644, 0.18547311425209045, 0.26534074544906616, -0.316218763589859, 0.25385767221450806, 0.22005999088287354, 0.33087342977523804, 0.2140360027551651, -0.06979589909315109, 0.10537411272525787, -0.06378648430109024, -0.04165637493133545, -0.2234189808368683, -0.19158977270126343, 0.43757492303848267, -0.1759514957666397, 0.46365246176719666, -0.03217552229762077, 0.029152704402804375, -0.22143839299678802, -0.016903018578886986, -0.23426444828510284, 0.38474106788635254, 0.31314292550086975, -0.20771268010139465, 0.33080998063087463, -0.046972617506980896, -0.44120222330093384, 0.2577674984931946, 0.2447437047958374, 0.06818360090255737, 0.09573749452829361, -0.0095518808811903, -0.045534197241067886, -0.3349422514438629, -0.06029944494366646, 0.34204792976379395, 0.5666441321372986, 0.2699577510356903, 0.19837455451488495, 0.0683957040309906, -0.26644691824913025, -0.11241111904382706, 0.06929817795753479, 0.16281509399414062, 0.012806706130504608, 0.2764880955219269, 0.03590720519423485, 0.07019875943660736, -0.01635843515396118, -0.09048422425985336, 0.13983875513076782, -0.18486683070659637, -0.15955105423927307, 0.013510562479496002, -0.23286226391792297, -0.6612979769706726, -0.2655377984046936, 0.059566009789705276, -0.306205689907074, -0.3195211887359619, 0.1293628066778183, 0.017231810837984085, -0.14742325246334076, 0.1712544560432434, 0.23219631612300873, 0.6074469685554504, -0.3907056152820587, -0.15059229731559753, -0.06880822032690048, -0.3318817913532257, -0.14265254139900208, 0.07787410914897919, 0.5498902201652527, 0.3654058277606964, 0.4354557693004608, 0.17507511377334595, 0.0016456637531518936, 0.040315527468919754, -0.4197554886341095, 0.0945982038974762, -0.33445635437965393, 0.16005390882492065, 0.06284310668706894, -0.16257432103157043, 0.1456708014011383, -0.42311903834342957, -0.08997166901826859, 0.02438977360725403, -0.09786473959684372, 0.017521336674690247, -0.03967570140957832, 0.02787160500884056, -0.22343942523002625, -0.30245503783226013, -0.5716316103935242, -0.4520237445831299, 0.3431037962436676, -0.27985358238220215, 0.14475825428962708, 0.06941834837198257, -0.2256755381822586, 0.14136268198490143, -0.17480264604091644, 0.07033822685480118, -0.11356339603662491, -0.10608216375112534, 0.04219083860516548, -0.34368380904197693, -0.24148795008659363, -0.38526856899261475, -0.026993276551365852, 0.17016905546188354, -0.00152994180098176, -0.060109417885541916, -0.23663626611232758, 0.07347795367240906, -0.21928182244300842, 0.11132589727640152, 0.2957318425178528, 0.13768994808197021, -0.16666559875011444, 0.13377735018730164, -0.04997418820858002, -0.1868762969970703, 0.20780999958515167, 0.3281921148300171, 0.38636523485183716, 0.15270382165908813, 0.0017185022588819265, 0.27756547927856445, 0.6705085635185242, 0.34821486473083496, 0.04389125853776932, 0.09031037241220474, 0.25889790058135986, 0.060461245477199554, 0.1286640167236328, -0.18067032098770142, 0.26686421036720276, -0.20002704858779907, 0.5133733749389648, 0.1512204259634018, 0.18066474795341492, -0.10541228204965591, -0.21315033733844757, 0.1212698370218277, -0.35108914971351624, -0.19965611398220062, 0.3020293116569519, -0.3192823529243469, 0.02158193290233612, -0.05448956787586212, -0.3992002308368683, -0.2533683478832245, -0.04955455660820007, 0.017426956444978714, 0.35206419229507446, -0.05508057028055191, 0.19885952770709991, -0.4190734624862671, -0.09170380979776382, -0.4600537121295929, 0.19697119295597076, 0.08974127471446991, 0.33363261818885803, -0.0685618594288826, -0.1117725521326065, -0.08517778664827347, 0.13111233711242676, 0.5710859894752502, -0.5576238632202148, -0.2217063456773758, 0.17702382802963257, 0.1506173461675644, -0.08356287330389023, -0.12241483479738235, -0.38056671619415283, -0.02671450562775135, 0.46636226773262024, 0.42542871832847595, -0.40828508138656616, -0.3596426248550415, 0.2669551372528076, -0.16865219175815582, -0.020688604563474655, -0.1301172524690628, -0.20354849100112915, -0.21533064544200897, -0.1732441633939743, 0.026581766083836555, 0.224521204829216, 0.16896389424800873, -0.15442778170108795, 0.15072369575500488, -0.04602191969752312, 0.2372140884399414, 0.21118490397930145, -0.012833614833652973, -0.05038972198963165, 0.26388493180274963, 0.06503301858901978, 0.07246547937393188, -0.030610794201493263, -0.2702561616897583, 0.10697295516729355, -0.13292747735977173, 0.06653204560279846, 0.27353882789611816, -0.07887053489685059, 0.7340443134307861, 0.15070149302482605, 0.3863653242588043, 0.17446643114089966, -0.10423766821622849, 0.02026783674955368, -0.24745485186576843, 0.21436667442321777, 0.20877225697040558, 0.24122455716133118, -0.18776948750019073, -0.19011402130126953, 0.5577901601791382, 0.33030593395233154, -0.01569283939898014, 0.4144621193408966, -0.47154343128204346, -0.2955446243286133, 0.31731584668159485, 0.037460945546627045, 0.7747870087623596, -0.055943287909030914, 0.3117327392101288, 0.08896934241056442, -0.4208640456199646, 0.5706256031990051, -0.4346333146095276, 0.15796615183353424, -0.09969193488359451, -0.1840035617351532, 0.030260540544986725, -0.10041110217571259, -0.026499580591917038, 0.2533988058567047, -0.3821479082107544, 0.3142699897289276, 0.4203949272632599, -0.2569931149482727, -0.031891271471977234, 0.5367088913917542, -0.05316861718893051, -0.392829567193985, -0.05631422623991966, 0.041811753064394, -0.32738640904426575, 0.08863253891468048, -0.1278614103794098, -0.3133986294269562, 0.23929034173488617, 0.15863969922065735, -0.3130236566066742, 0.10464929044246674, -0.06307310611009598, 0.14268863201141357, 0.2242191582918167, -0.035015594214200974, 0.2476469874382019, -0.13931423425674438, 0.04299277812242508, -0.13076256215572357, -0.06671774387359619, 0.03300900757312775, -0.10206156224012375, 0.07684865593910217, 0.0985979214310646, -0.16361387073993683, 0.5141544342041016, -0.003341516712680459, 0.008826995268464088, -0.2976301908493042, 0.2612648606300354, -0.38956359028816223, -0.27695053815841675, -0.024676501750946045, 0.18370667099952698, 0.09997130930423737, -0.29114818572998047, 0.614385187625885, 0.08731109648942947, 0.035768404603004456, 0.040252137929201126, 0.13858215510845184, -0.23400503396987915, 0.5710294246673584, -0.24325881898403168, -0.014289749786257744, 0.08371656388044357, -0.01370172668248415, -0.11303979903459549, -0.4529513418674469, -0.020737536251544952, 0.3857307434082031, 0.07245323061943054, -0.02220546081662178, 0.06766720116138458, 0.13055844604969025, -0.060833822935819626, -0.017283082008361816, -0.2517116963863373, -0.2992796301841736, 0.10248854756355286, 0.24880781769752502, 0.24665313959121704, 0.1551617830991745, -0.04194330424070358, -0.24314841628074646, -0.13975226879119873, 0.546168327331543, -0.0309030469506979, 0.26203468441963196, -0.05144910886883736, 0.38082554936408997, 0.3815823197364807, 0.5050745606422424, -0.2507598102092743, 0.07350707799196243, 0.0664883404970169, 0.19243568181991577, 0.4419764280319214, -0.26780688762664795, -0.03233399987220764, 0.2462565153837204, -0.1350870430469513, 0.1086004227399826, -0.3244999647140503, -0.15105067193508148, -0.20546050369739532, 0.1055411621928215, 0.19726678729057312, 0.0291514340788126, 0.047509465366601944, -0.033540014177560806, -0.13116863369941711, -0.10678375512361526, 0.021673746407032013, -0.37187132239341736, -0.12796542048454285, 0.23094786703586578, -0.11810000240802765, 0.036770887672901154, 0.0795673057436943, 0.135007843375206, 0.004463276360183954, -0.09252642840147018, 0.3175368309020996, 0.0924285426735878, 0.06634542346000671, -0.1891515552997589, -0.18550240993499756, 0.04238743335008621, 0.23217354714870453, 0.18918286263942719, 0.0023197997361421585, 0.08651178330183029, -0.17059411108493805, -0.16673603653907776, 0.21267834305763245, 0.23364511132240295, 0.01032536756247282, -0.09136737138032913, 0.1436547338962555, 0.09549912810325623, -0.054431382566690445, -0.005216205958276987, 0.22590750455856323, 0.059953827410936356, -0.007213828153908253, 0.0664544403553009, 0.19826142489910126, 0.3934859037399292, -0.18172912299633026, 0.04547881707549095, 0.03525976091623306, -0.3973194360733032, 0.1519799828529358, -0.04009997844696045, 0.0002020208485191688, 0.0052862572483718395, 0.06006694957613945, 0.03778809681534767, 0.1249130517244339, -0.3215548098087311, 0.328936904668808, 0.26297348737716675, 0.025910967960953712, 0.03557557612657547, 0.17994649708271027, -0.028497587889432907, -0.15546953678131104, 0.27964526414871216, 0.0755624920129776, 0.43249422311782837, 0.05860938876867294, 0.8388553261756897, -0.292815238237381, -0.5137995481491089, 0.11349032074213028, 0.4244176149368286, -0.0020870622247457504, -0.19419610500335693, 0.05413691699504852, 0.11889059096574783, -0.3490162491798401, 0.0717269703745842, -0.06716357171535492, -0.12387624382972717, 0.5954530835151672, -0.003620558651164174, 0.16622339189052582, -0.22186878323554993, -0.17409615218639374, 0.28185051679611206, 0.23936374485492706, -0.11452365666627884, 0.11490674316883087, 0.0892239511013031, -0.1297861486673355, -0.15155139565467834, 0.556912362575531, 0.2587621510028839, 0.052571024745702744, -0.42290517687797546, 0.014232473447918892, 0.09556431323289871, 0.10881666094064713, 0.07581375539302826, 0.08942315727472305, 0.4531348645687103, -0.13934700191020966, 0.2778196334838867, -0.011475368402898312, -0.01169667486101389, -0.15413127839565277, 0.033524803817272186, -0.48072144389152527, -0.34382179379463196, 0.3919150233268738, 0.14652658998966217, -0.1690302938222885, 0.10381017625331879, 0.26207050681114197, -0.7270407676696777, -0.2905248701572418, 0.41981545090675354, -0.028902510181069374, -0.03651430830359459, -0.20693455636501312, 0.06619687378406525, 0.18844294548034668, 0.5728138089179993, 0.2763862907886505, 0.32753780484199524, -0.3885541558265686, -0.11769860982894897, -0.5355226993560791, -0.17271281778812408, 0.19988888502120972, -0.04911639541387558, -0.18681901693344116, 0.14797575771808624, -0.02143285796046257, 0.3246808648109436, -0.2616150379180908, 0.10892476886510849, 0.16829875111579895, 0.07349096983671188, -0.5057322382926941, 0.11960281431674957, 0.20687872171401978, 0.04922573268413544, 0.19890868663787842, -0.05705815553665161, -0.08499018847942352, 0.1317051500082016, -0.07632080465555191, -0.08576175570487976, 0.005866107530891895, 0.24516701698303223, 0.16404074430465698, -0.026463065296411514, 0.08232589066028595, 0.46307995915412903, -0.013872740790247917, -0.19936496019363403, -0.01980419084429741, 0.044574081897735596, -0.40713468194007874, -0.07028967887163162, -0.0931830182671547, 0.19522559642791748, -0.39534664154052734, 0.2937242090702057, 0.008450246416032314, -0.11665989458560944, 0.07646239548921585, 0.028759559616446495, -0.15167859196662903, -0.0498308390378952, 0.08785472065210342, 0.19569411873817444, 0.2983796000480652, 0.7401276230812073, -0.005243984051048756, 0.12361162155866623, -0.1726134717464447, -0.06902076303958893, 0.02740965597331524, -0.05322415009140968, -0.2343159168958664, -0.10208047926425934, 0.06130256503820419, 0.16821062564849854, 0.2172335386276245, -0.6970354914665222, -0.37758535146713257, 0.0372028648853302, -0.05564756318926811, 0.013424401171505451, 0.7778041958808899, 0.13142213225364685, -0.16813673079013824, -0.16086938977241516, -0.3078276216983795, 0.011136811226606369, -0.2971329689025879, 0.048108696937561035, -0.4513055384159088 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
I agree that we should leave the flattening of the dataset to the user for now. Especially because although the T5 framing seems obvious, there are slight variations on how the T5 authors do it in comparison to other approaches such as gpt-3 and decaNLP. In terms of sampling, Examples-proportional mixing does seem the simplest to implement so would probably be a good starting point. Temperature-scaled mixing would probably most useful, offering flexibility as it can simulate the other 2 methods by setting the temperature parameter. There is a [relevant part of the T5 repo](https://github.com/google-research/text-to-text-transfer-transformer/blob/03c94165a7d52e4f7230e5944a0541d8c5710788/t5/data/utils.py#L889-L1118) which should help with implementation. According to the T5 authors, equal-mixing performs worst. Among the other two methods, tuning the K value (the artificial dataset size limit) has a large impact.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
126
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. I agree that we should leave the flattening of the dataset to the user for now. Especially because although the T5 framing seems obvious, there are slight variations on how the T5 authors do it in comparison to other approaches such as gpt-3 and decaNLP. In terms of sampling, Examples-proportional mixing does seem the simplest to implement so would probably be a good starting point. Temperature-scaled mixing would probably most useful, offering flexibility as it can simulate the other 2 methods by setting the temperature parameter. There is a [relevant part of the T5 repo](https://github.com/google-research/text-to-text-transfer-transformer/blob/03c94165a7d52e4f7230e5944a0541d8c5710788/t5/data/utils.py#L889-L1118) which should help with implementation. According to the T5 authors, equal-mixing performs worst. Among the other two methods, tuning the K value (the artificial dataset size limit) has a large impact.
[ -0.07244475185871124, -0.4672178328037262, -0.04016014188528061, -0.05503220856189728, -0.30456024408340454, 0.07474454492330551, 0.1744341254234314, 0.06729564815759659, 0.26532408595085144, -0.11934687197208405, -0.2597549855709076, 0.3054884970188141, -0.16949862241744995, 0.3780531883239746, 0.35094764828681946, -0.44998228549957275, -0.02154984138906002, -0.14634419977664948, -0.2921181321144104, 0.3242069482803345, 0.05431612208485603, 0.11644241213798523, -0.15435409545898438, 0.06619985401630402, -0.2752164602279663, -0.1766953021287918, -0.34399890899658203, 0.12340977042913437, 0.08483152836561203, 0.07976388186216354, 0.135843887925148, 0.39544492959976196, -0.018593138083815575, 0.12798544764518738, -0.00011141275172121823, -0.09476138651371002, 0.09925314784049988, -0.09712187945842743, 0.15263663232326508, 0.03999250382184982, 0.17799226939678192, -0.3429226577281952, 0.06340292096138, -0.14008669555187225, 0.13159745931625366, 0.24719113111495972, 0.1972431093454361, 0.0887654721736908, 0.4500884711742401, -0.35335245728492737, 0.10489175468683243, 0.3131382465362549, -0.24382935464382172, 0.30103328824043274, -0.08568546175956726, -0.15280304849147797, 0.011506573297083378, 0.28885066509246826, 0.6218153238296509, -0.5280055403709412, -0.015143528580665588, 0.1048830896615982, -0.046287767589092255, 0.04999246075749397, 0.13104334473609924, -0.06663867831230164, -0.2823977470397949, -0.2535237967967987, -0.362751841545105, 0.27751269936561584, -0.12239691615104675, -0.06288987398147583, -0.19258670508861542, -0.5656949281692505, -0.0037438415456563234, 0.05627412721514702, -0.32901331782341003, 0.08781738579273224, -0.2160569280385971, 0.015035547316074371, -0.1249791756272316, -0.12686848640441895, -0.00769107835367322, 0.11148174852132797, 0.3081294000148773, 0.5256875157356262, 0.07436881959438324, 0.13465355336666107, 0.35248732566833496, 0.10151586681604385, -0.14089243113994598, -0.26514482498168945, 0.2907789349555969, 0.13547062873840332, -0.3288772702217102, -0.37709712982177734, 0.15168611705303192, -0.2695583701133728, 0.28027069568634033, 0.09671389311552048, 0.14681462943553925, 0.05229099094867706, -0.08265113830566406, 0.2643831968307495, 0.3891301155090332, 0.012632107362151146, -0.09676504880189896, -0.11974373459815979, 0.023958921432495117, -0.10591050237417221, 0.19155815243721008, 0.13948890566825867, 0.1446608155965805, 0.009005075320601463, -0.43491920828819275, -0.11233676970005035, -0.19096234440803528, 0.09654742479324341, -0.18051883578300476, -0.5517175793647766, -0.0857008546590805, -0.10183689743280411, -0.02041568234562874, -0.12436402589082718, -0.2886069715023041, 0.1530837118625641, -0.07474686950445175, 0.1504199355840683, -0.18043993413448334, 0.020788371562957764, -0.07014717906713486, 0.06636720895767212, -0.4629811644554138, -0.0349682979285717, 0.2592132091522217, 0.10559343546628952, 0.0714486837387085, 0.11740325391292572, -0.020503787323832512, 0.16794022917747498, 0.4261094629764557, 0.03080563247203827, 0.028727732598781586, -0.031046995893120766, 0.14761915802955627, -0.30910658836364746, -0.031393490731716156, 0.23604942858219147, -0.35183122754096985, -0.0972793698310852, -0.14454928040504456, -0.1775629073381424, 0.1374158263206482, 0.04549885913729668, -0.21582357585430145, -0.26927077770233154, -0.4415574073791504, 0.7260340452194214, -0.013284183107316494, -0.023931831121444702, -0.17176946997642517, 0.25947973132133484, -0.2092253714799881, -0.11802950501441956, 0.18940463662147522, 0.09559904038906097, -0.021370014175772667, -0.3190780580043793, -0.10603832453489304, -0.0609193779528141, -0.01561582088470459, 0.32130640745162964, -0.15178422629833221, 0.19926604628562927, 0.10687829554080963, 0.17462316155433655, 0.2618967592716217, -0.429456889629364, -0.05689621344208717, 0.35273483395576477, -0.06147163733839989, 0.3841184377670288, 0.050875429064035416, 0.1725042760372162, -0.0096141267567873, -0.21683627367019653, 0.24459543824195862, 0.9837128520011902, -0.40412476658821106, 0.002475675893947482, -0.1859396994113922, -0.42441004514694214, 0.5637452602386475, 0.34826216101646423, 0.06482468545436859, -0.2935229539871216, -0.1886729747056961, 0.034367501735687256, 0.04671807214617729, 0.09230990707874298, 0.07490944862365723, -0.08023253828287125, -0.264489084482193, 0.010359781794250011, -0.2241344302892685, -0.16983987390995026, -0.42165300250053406, -0.008736447431147099, -0.036962490528821945, 0.3787113130092621, 0.37277838587760925, -0.14035247266292572, 0.23759692907333374, -0.30643683671951294, -0.11149374395608902, -0.3117159307003021, 0.1445588767528534, -0.06806720793247223, 0.012566040270030499, -0.23941943049430847, -0.06527691334486008, 0.2053757905960083, 0.09740204364061356, -0.1050228551030159, -0.42300838232040405, 0.2914789319038391, -0.04073332995176315, 0.02150796912610531, -0.013075762428343296, 0.6146174669265747, -0.21612344682216644, -0.11658541858196259, 0.20798267424106598, 0.29159414768218994, -0.31383341550827026, 0.21099944412708282, 0.30932140350341797, 0.35455596446990967, 0.19584274291992188, -0.06140482798218727, 0.0992705374956131, -0.01221090741455555, -0.08991345763206482, -0.1710817515850067, -0.0809541642665863, 0.4797084927558899, -0.20128418505191803, 0.357545405626297, -0.021909892559051514, 0.03489631786942482, -0.2077239751815796, -0.05692208185791969, -0.19069355726242065, 0.3881004750728607, 0.31067466735839844, -0.22799581289291382, 0.33478012681007385, -0.07678527384996414, -0.4003094732761383, 0.24344253540039062, 0.21397216618061066, 0.013479609042406082, 0.09026569128036499, -0.023068252950906754, 0.03124084696173668, -0.31019872426986694, -0.06050260365009308, 0.33754435181617737, 0.5514166355133057, 0.28949636220932007, 0.17737045884132385, 0.016729174181818962, -0.21758238971233368, -0.07269266247749329, 0.04261374846100807, 0.16537733376026154, -0.06318686902523041, 0.14198394119739532, -0.041099656373262405, 0.09890086948871613, 0.026658065617084503, -0.10134867578744888, 0.12281966209411621, -0.1445389986038208, -0.0426812469959259, -0.03750555217266083, -0.15475648641586304, -0.5833346843719482, -0.1911756545305252, 0.052576445043087006, -0.35153117775917053, -0.29605937004089355, 0.20875594019889832, 0.05685420706868172, -0.1775723397731781, 0.1679806262254715, 0.19975072145462036, 0.6398574113845825, -0.3923532962799072, -0.17517153918743134, -0.052481550723314285, -0.2999000549316406, -0.11523331701755524, 0.13736408948898315, 0.43582600355148315, 0.3023831248283386, 0.4983600080013275, 0.23070314526557922, -0.03287999704480171, 0.12016382813453674, -0.38628503680229187, 0.11135093122720718, -0.35678210854530334, 0.29216957092285156, 0.09976132214069366, -0.10114610940217972, 0.09706760197877884, -0.3661125600337982, -0.07395517826080322, -0.09430358558893204, -0.061669375747442245, -0.040394268929958344, 0.008099166676402092, 0.026680780574679375, -0.22710460424423218, -0.22857140004634857, -0.5969745516777039, -0.4918062090873718, 0.2876461148262024, -0.25232386589050293, 0.09606632590293884, 0.04323139786720276, -0.28764525055885315, 0.20857641100883484, -0.14133451879024506, 0.038155581802129745, -0.18621976673603058, -0.07839281857013702, -0.020862577483057976, -0.3604215681552887, -0.17928342521190643, -0.42282944917678833, -0.01258440688252449, 0.19307354092597961, -0.07771410048007965, -0.06733416765928268, -0.2891920208930969, 0.032976455986499786, -0.10656126588582993, 0.08936779201030731, 0.2331579476594925, 0.14413036406040192, -0.20325018465518951, 0.07359974086284637, -0.027181902900338173, -0.1418783813714981, 0.27341705560684204, 0.39811569452285767, 0.3655804395675659, 0.14558780193328857, -0.025387275964021683, 0.21974174678325653, 0.749504566192627, 0.2815074324607849, 0.08701778203248978, 0.10910969972610474, 0.29363659024238586, -0.0391642227768898, 0.09498507529497147, -0.1791396588087082, 0.2571568787097931, -0.11236868798732758, 0.49735623598098755, 0.17449788749217987, 0.15857039391994476, -0.20294590294361115, -0.27151674032211304, 0.019823536276817322, -0.2972192168235779, -0.21401064097881317, 0.26668107509613037, -0.27441754937171936, -0.0028799285646528006, -0.10535620898008347, -0.41696470975875854, -0.23437300324440002, -0.08540177345275879, -0.020185409113764763, 0.28638899326324463, 0.04233977198600769, 0.1747763305902481, -0.3722774088382721, -0.024910731241106987, -0.45011916756629944, 0.13502992689609528, 0.05351858586072922, 0.3062896430492401, -0.12003669887781143, -0.10522803664207458, -0.09981377422809601, 0.15174618363380432, 0.5571807622909546, -0.5467596650123596, -0.15157882869243622, 0.08499211072921753, 0.08082804828882217, -0.13787327706813812, -0.07683352380990982, -0.3483755886554718, -0.013060607016086578, 0.4500731825828552, 0.3170127868652344, -0.4330320656299591, -0.3620240390300751, 0.22607167065143585, -0.1992088109254837, -0.07860645651817322, -0.1529192179441452, -0.20009315013885498, -0.18559397757053375, -0.12708818912506104, 0.014104447327554226, 0.2638648748397827, 0.20318469405174255, -0.09490784257650375, 0.18410947918891907, -0.02212691493332386, 0.26804253458976746, 0.2252078354358673, 0.033784057945013046, -0.08329121023416519, 0.23447169363498688, 0.14794892072677612, 0.07315298914909363, -0.027882205322384834, -0.38055896759033203, 0.07879547774791718, -0.08251626044511795, 0.11951764672994614, 0.30150675773620605, 0.0060751475393772125, 0.6325337886810303, 0.08441141247749329, 0.3359811305999756, 0.28056100010871887, -0.10613907873630524, 0.08464562892913818, -0.21934041380882263, 0.1391848623752594, 0.15871590375900269, 0.251853883266449, -0.09183443337678909, -0.23388001322746277, 0.5279419422149658, 0.2838381826877594, -0.050370555371046066, 0.427849143743515, -0.4882172644138336, -0.2839677333831787, 0.2870749235153198, 0.0938035249710083, 0.7551211714744568, -0.021681303158402443, 0.30531156063079834, 0.11571823060512543, -0.3540651798248291, 0.6180203557014465, -0.3570944368839264, 0.10530250519514084, -0.07974356412887573, -0.19833412766456604, 0.03314225748181343, -0.05398932471871376, -0.12255187332630157, 0.26841989159584045, -0.3492085039615631, 0.3354414403438568, 0.4309316873550415, -0.26445460319519043, -0.045097801834344864, 0.5335502624511719, -0.02422507107257843, -0.32665371894836426, -0.020114202052354813, 0.07147373259067535, -0.3004840910434723, 0.06917193531990051, -0.14425411820411682, -0.2498892992734909, 0.21163950860500336, 0.17763052880764008, -0.2500084936618805, 0.059739045798778534, -0.035752054303884506, 0.15424463152885437, 0.21908216178417206, 0.020608417689800262, 0.2647167444229126, -0.2386106550693512, -0.07471293956041336, -0.1682818979024887, -0.13552431762218475, 0.07205867767333984, -0.19097544252872467, 0.04476720839738846, 0.1156843975186348, -0.09888534992933273, 0.5525388717651367, -0.07500877231359482, -0.10550376772880554, -0.2602421045303345, 0.2500440776348114, -0.4098784923553467, -0.21422797441482544, 0.020000578835606575, 0.17416426539421082, 0.151570662856102, -0.35633978247642517, 0.5705782175064087, 0.06015285849571228, 0.05220349133014679, 0.0529620386660099, 0.16527265310287476, -0.255143404006958, 0.472382128238678, -0.2441769391298294, 0.006948657799512148, 0.05782732367515564, 0.009747098200023174, -0.10201357305049896, -0.3523837625980377, -0.06964684277772903, 0.31335705518722534, 0.0930527001619339, -0.042641378939151764, 0.03194182366132736, 0.1327233910560608, -0.11251021921634674, 0.019939400255680084, -0.17062978446483612, -0.36914920806884766, 0.14089350402355194, 0.24945800006389618, 0.27114856243133545, 0.18250638246536255, -0.11808470636606216, -0.20866401493549347, -0.10900367796421051, 0.510397732257843, 0.0016729355556890368, 0.2612942159175873, -0.1301575005054474, 0.3655259311199188, 0.42578962445259094, 0.49408066272735596, -0.29024457931518555, 0.03717358037829399, 0.03744088113307953, 0.09641584753990173, 0.4827969968318939, -0.26365846395492554, 0.02177978679537773, 0.2817062437534332, -0.13145826756954193, 0.05857205390930176, -0.34479478001594543, -0.16829554736614227, -0.2200629711151123, 0.11335418373346329, 0.10334287583827972, -0.020654331892728806, 0.07225687056779861, -0.04305393248796463, -0.14204154908657074, -0.0688854455947876, 0.045749809592962265, -0.49925723671913147, -0.08053167909383774, 0.22027722001075745, -0.13718007504940033, -0.011609738692641258, 0.020350176841020584, 0.23941479623317719, -0.008112763054668903, -0.1258602738380432, 0.2913135290145874, 0.13309305906295776, 0.12393437325954437, -0.19249922037124634, -0.1676720231771469, 0.019878797233104706, 0.21255220472812653, 0.17324033379554749, -0.05693621560931206, 0.19535771012306213, -0.1702476590871811, -0.14223447442054749, 0.24023385345935822, 0.12741270661354065, 0.04383215308189392, -0.17915229499340057, 0.17719930410385132, 0.14164860546588898, -0.09756139665842056, 0.06519678235054016, 0.2262757122516632, 0.08455821126699448, 0.12077242881059647, 0.09019804745912552, 0.12194262444972992, 0.4357938766479492, -0.25440526008605957, -0.01790304109454155, 0.14192256331443787, -0.35142165422439575, 0.134095698595047, -0.04542116820812225, -0.07775188982486725, 0.0030958708375692368, 0.11714547127485275, -0.07418793439865112, 0.12427593767642975, -0.35630735754966736, 0.35232338309288025, 0.27313128113746643, -0.01798049360513687, 0.06044257432222366, 0.17766675353050232, 0.005221328232437372, -0.1391511708498001, 0.35451436042785645, 0.1029016301035881, 0.40711700916290283, 0.038383081555366516, 0.9214756488800049, -0.28261396288871765, -0.4910770058631897, 0.07427056133747101, 0.4407656788825989, 0.013904890045523643, -0.20885884761810303, 0.06338087469339371, 0.07670687884092331, -0.34909191727638245, 0.09109161794185638, 0.028142251074314117, -0.16176636517047882, 0.6003194451332092, -0.09266402572393417, 0.19700372219085693, -0.12051048874855042, -0.18638728559017181, 0.304165244102478, 0.2193450629711151, -0.08482000231742859, 0.0386621430516243, 0.08103229850530624, -0.09659130126237869, -0.20828381180763245, 0.5394885540008545, 0.13788539171218872, -0.030670013278722763, -0.4813675880432129, 0.026541778817772865, 0.13031253218650818, 0.16194455325603485, -0.04226190224289894, 0.16966284811496735, 0.4215094745159149, -0.1417669951915741, 0.29074084758758545, 0.0038775827270001173, -0.09474679827690125, -0.1848486214876175, 0.07130422443151474, -0.4407307803630829, -0.35878992080688477, 0.38067784905433655, 0.14892977476119995, -0.181696817278862, 0.12254220247268677, 0.27388209104537964, -0.6083036661148071, -0.22651855647563934, 0.288765549659729, -0.1661260724067688, -0.0840122401714325, -0.2262103259563446, 0.08732547610998154, 0.05002930760383606, 0.6215307116508484, 0.23861190676689148, 0.3424454927444458, -0.3124419152736664, -0.11013630777597427, -0.5452612638473511, -0.12234305590391159, 0.1547335684299469, -0.12682484090328217, -0.16537843644618988, 0.17525863647460938, -0.0254004318267107, 0.3293077051639557, -0.26487964391708374, 0.18954212963581085, 0.13836850225925446, 0.08010149002075195, -0.45095735788345337, 0.14279389381408691, 0.08015494048595428, 0.06859370321035385, 0.2718912363052368, -0.10980717837810516, -0.1711379438638687, 0.08312534540891647, -0.06507854163646698, -0.0660625696182251, -0.06892748922109604, 0.2686029374599457, 0.161920428276062, 0.03138852119445801, 0.0771801620721817, 0.43555647134780884, 0.004979622550308704, -0.20629273355007172, 0.0015175103908404708, 0.001496192067861557, -0.35154417157173157, -0.19136925041675568, -0.07274625450372696, 0.16833744943141937, -0.3356747627258301, 0.2971017360687256, -0.029855260625481606, -0.11843444406986237, 0.12275110185146332, 0.061915524303913116, -0.19488686323165894, -0.0924559235572815, 0.10430461913347244, 0.15145082771778107, 0.21517972648143768, 0.7912331819534302, -0.03391629084944725, 0.0512431338429451, -0.10069747269153595, -0.010347816161811352, 0.04032700136303902, -0.09624476730823517, -0.22648485004901886, -0.07545948773622513, 0.03443226218223572, 0.08696621656417847, 0.10204672813415527, -0.6406654715538025, -0.37819764018058777, 0.037047937512397766, -0.040623005479574203, -0.0003257919743191451, 0.7422398924827576, 0.09145575016736984, -0.19132667779922485, -0.07554019242525101, -0.32336464524269104, 0.008202070370316505, -0.2920856773853302, 0.16921234130859375, -0.34073689579963684 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
I agree with going with temperature-scaled mixing for its flexibility! For the function that combines the datasets, I also find `dataset.add()` okay while also considering that users may want it to be easy to combine a list of say 10 data sources in one go. `dataset.sample()` should also be good. By the looks of it, we're planning to have as main parameters: `temperature`, and `K`. On converting the datasets to the same format, I agree that we can leave these to the users for now. But, I do imagine it'd be an awesome feature for the future to have this automatically handled, based on a chosen *approach* to formatting :smile: E.g. T5, GPT-3, decaNLP, original raw formatting, or a contributed way of formatting in text-to-text.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
125
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. I agree with going with temperature-scaled mixing for its flexibility! For the function that combines the datasets, I also find `dataset.add()` okay while also considering that users may want it to be easy to combine a list of say 10 data sources in one go. `dataset.sample()` should also be good. By the looks of it, we're planning to have as main parameters: `temperature`, and `K`. On converting the datasets to the same format, I agree that we can leave these to the users for now. But, I do imagine it'd be an awesome feature for the future to have this automatically handled, based on a chosen *approach* to formatting :smile: E.g. T5, GPT-3, decaNLP, original raw formatting, or a contributed way of formatting in text-to-text.
[ -0.091631680727005, -0.42115867137908936, -0.04513263702392578, -0.08009923994541168, -0.2404671013355255, 0.059044934809207916, 0.1587897539138794, 0.05346130207180977, 0.3097868263721466, -0.15817700326442719, -0.2333139330148697, 0.31752073764801025, -0.19275881350040436, 0.35245901346206665, 0.420116126537323, -0.47216013073921204, 0.027789130806922913, -0.2041647732257843, -0.3079892098903656, 0.3535069525241852, 0.007799830753356218, 0.11438106000423431, -0.17745055258274078, 0.08323544263839722, -0.29802966117858887, -0.1659754514694214, -0.3557685315608978, 0.13368478417396545, 0.09495007991790771, 0.019252635538578033, 0.17773206532001495, 0.3762139678001404, -0.035789526998996735, 0.06485811620950699, -0.00011105542216682807, -0.11337100714445114, 0.21172407269477844, -0.08244482427835464, 0.09594827890396118, -0.04149753972887993, 0.14164188504219055, -0.3843332529067993, 0.00042363835382275283, -0.0986679345369339, 0.18147023022174835, 0.16644401848316193, 0.19007395207881927, 0.049618761986494064, 0.33850231766700745, -0.29403969645500183, 0.12200936675071716, 0.3003103733062744, -0.16042497754096985, 0.25943705439567566, -0.07904578000307083, -0.07974854856729507, 0.0491224080324173, 0.223652645945549, 0.6150226593017578, -0.549126386642456, -0.0011971908388659358, 0.12823158502578735, -0.0015522981993854046, -0.02403174340724945, 0.1823764592409134, -0.04576147720217705, -0.3587007522583008, -0.2435752898454666, -0.3750150501728058, 0.23614202439785004, -0.11858659237623215, -0.15378664433956146, -0.21855023503303528, -0.5235368609428406, 0.02771906368434429, 0.01679692603647709, -0.2812062203884125, 0.029973171651363373, -0.1675623655319214, -0.0484490692615509, -0.16560260951519012, -0.12260448187589645, 0.013107526116073132, 0.15340298414230347, 0.34610816836357117, 0.5681384801864624, 0.06152743473649025, 0.16792145371437073, 0.40035420656204224, 0.001665790448896587, -0.1527254581451416, -0.31980255246162415, 0.29201093316078186, 0.0985112413764, -0.38790443539619446, -0.33713603019714355, 0.14897991716861725, -0.18864686787128448, 0.2458423674106598, 0.1551024168729782, 0.10689643025398254, 0.08710207790136337, -0.11532768607139587, 0.27482134103775024, 0.3207070231437683, 0.06586887687444687, -0.0684720128774643, -0.11937577277421951, 0.02774689719080925, -0.08491197973489761, 0.21637968719005585, 0.12413576245307922, 0.21670301258563995, -0.02785881608724594, -0.4467780292034149, -0.1466723531484604, -0.17592625319957733, 0.1282651722431183, -0.17640095949172974, -0.5241801142692566, -0.05227171257138252, -0.10353740304708481, 0.0894397646188736, -0.04895605519413948, -0.2832755446434021, 0.1021508276462555, -0.07685330510139465, 0.20829147100448608, -0.0627482682466507, 0.074934221804142, -0.07943777740001678, 0.12596295773983002, -0.4375811517238617, -0.014975243248045444, 0.3017207384109497, 0.11207002401351929, 0.117110975086689, 0.15991495549678802, -0.028330618515610695, 0.10989777743816376, 0.3795182704925537, -0.016750650480389595, 0.05097763240337372, -0.007164126727730036, 0.13988077640533447, -0.32601600885391235, -0.0488303005695343, 0.17612838745117188, -0.34295201301574707, -0.14655040204524994, -0.13508203625679016, -0.2072175294160843, 0.14267127215862274, 0.052464962005615234, -0.19705772399902344, -0.2726626694202423, -0.46978330612182617, 0.727302610874176, -0.06460906565189362, -0.04315299168229103, -0.13948529958724976, 0.27761343121528625, -0.24226990342140198, -0.09580105543136597, 0.19651871919631958, 0.09004755318164825, -0.008327420800924301, -0.3287341892719269, -0.09379526972770691, -0.11021561175584793, -0.032542262226343155, 0.26917794346809387, -0.173434779047966, 0.20143714547157288, 0.04872151464223862, 0.20125973224639893, 0.272591769695282, -0.4674473702907562, -0.06602437794208527, 0.31485679745674133, -0.02159835770726204, 0.37245628237724304, 0.047157011926174164, 0.22150468826293945, -0.009119083173573017, -0.19109752774238586, 0.27087295055389404, 0.9994990825653076, -0.39096641540527344, -0.005298844538629055, -0.1448642611503601, -0.36753779649734497, 0.5181847810745239, 0.3499695062637329, 0.06550005078315735, -0.3326672613620758, -0.14244520664215088, -0.023778248578310013, 0.08096607029438019, 0.0770779550075531, 0.08081123232841492, -0.023396294564008713, -0.3496009409427643, 0.025720782577991486, -0.21601709723472595, -0.1317535638809204, -0.39827924966812134, -0.013192736543715, -0.011376888491213322, 0.3362923562526703, 0.2629062235355377, -0.1139773577451706, 0.23644974827766418, -0.2684759795665741, -0.13135114312171936, -0.3461124300956726, 0.1502704620361328, -0.061266470700502396, 0.05107888951897621, -0.26668673753738403, -0.03675738722085953, 0.3218361437320709, 0.11384926736354828, -0.11466022580862045, -0.5048952698707581, 0.27984175086021423, -0.08512625843286514, 0.03201049566268921, -0.015921080484986305, 0.6123459339141846, -0.13566870987415314, -0.09030520170927048, 0.18309910595417023, 0.3067479729652405, -0.3286167085170746, 0.2714979946613312, 0.3248352110385895, 0.2580251395702362, 0.18337446451187134, -0.023683473467826843, 0.11496029794216156, -0.03480027616024017, -0.03136909753084183, -0.2011304348707199, -0.10553065687417984, 0.4779740571975708, -0.1793975532054901, 0.403445303440094, -0.027947627007961273, 0.027526505291461945, -0.19954346120357513, -0.03600264713168144, -0.22701162099838257, 0.3675665557384491, 0.30155739188194275, -0.19315452873706818, 0.38591375946998596, -0.09095422923564911, -0.34883978962898254, 0.17814479768276215, 0.20084072649478912, 0.013798702508211136, 0.09303024411201477, -0.016065409407019615, -0.060315217822790146, -0.3105754256248474, -0.06443580985069275, 0.4159683883190155, 0.6467362642288208, 0.2788434624671936, 0.14651672542095184, 0.05391921103000641, -0.23704113066196442, -0.07217692583799362, 0.027097187936306, 0.179716095328331, -0.02205779403448105, 0.13218168914318085, 0.028972819447517395, 0.05366856977343559, -0.019798092544078827, -0.14596939086914062, 0.13913565874099731, -0.18290100991725922, -0.02678830549120903, 0.03504466637969017, -0.19006137549877167, -0.5519418120384216, -0.2311933934688568, 0.04605177044868469, -0.3759758174419403, -0.30405372381210327, 0.12783901393413544, 0.025094620883464813, -0.19186332821846008, 0.18358418345451355, 0.2177858203649521, 0.5667129755020142, -0.4283859133720398, -0.25661712884902954, -0.1000010073184967, -0.25852158665657043, -0.10058066248893738, 0.124039426445961, 0.41809260845184326, 0.30088621377944946, 0.4967922568321228, 0.2893137037754059, -0.01446940191090107, 0.020839491859078407, -0.38075923919677734, 0.047637008130550385, -0.2881385087966919, 0.2660781741142273, 0.10908100008964539, -0.23446282744407654, 0.1544492095708847, -0.41126111149787903, -0.027231905609369278, -0.06777665764093399, -0.05139271914958954, -0.04911769554018974, -0.01519888173788786, 0.020708847790956497, -0.17686617374420166, -0.22578772902488708, -0.6239806413650513, -0.4909670650959015, 0.2893373370170593, -0.2813090682029724, 0.06138443946838379, 0.016535194590687752, -0.26014024019241333, 0.19389960169792175, -0.15522293746471405, 0.037598881870508194, -0.1745579093694687, -0.10408509522676468, 0.046573419123888016, -0.3616587817668915, -0.16067248582839966, -0.4154684245586395, -0.11438293009996414, 0.23246842622756958, -0.08507445454597473, -0.042574670165777206, -0.23878824710845947, 0.048095088452100754, -0.1275748461484909, 0.09415790438652039, 0.22289180755615234, 0.17515960335731506, -0.22287115454673767, 0.09907539188861847, -0.03677685931324959, -0.26462939381599426, 0.1948920339345932, 0.33807289600372314, 0.4272564649581909, 0.1778102070093155, 0.016580432653427124, 0.17279298603534698, 0.7245492339134216, 0.2591904103755951, 0.05576436594128609, 0.150517076253891, 0.30706313252449036, 0.03356916829943657, 0.14654755592346191, -0.23375490307807922, 0.24164269864559174, -0.16472817957401276, 0.5056423544883728, 0.09843334555625916, 0.17190153896808624, -0.2043496072292328, -0.28388074040412903, -0.012782560661435127, -0.2579224109649658, -0.1879766881465912, 0.26838430762290955, -0.22223544120788574, 0.020278485491871834, -0.05537892505526543, -0.3724340796470642, -0.2528824508190155, -0.042795561254024506, 0.017948057502508163, 0.3158904016017914, -0.030992304906249046, 0.15221843123435974, -0.3451536297798157, -0.0411999449133873, -0.4830806255340576, 0.1510532945394516, 0.08815208077430725, 0.2757003605365753, -0.08164861053228378, -0.07314757257699966, -0.1406983882188797, 0.10453727096319199, 0.5943599343299866, -0.5221494436264038, -0.16565921902656555, 0.09090524911880493, 0.0804269015789032, -0.09018903225660324, -0.08775381743907928, -0.2720242440700531, -0.04363035038113594, 0.44831594824790955, 0.36958083510398865, -0.36920517683029175, -0.31362152099609375, 0.25652027130126953, -0.18059851229190826, -0.031666871160268784, -0.199013814330101, -0.18238654732704163, -0.1892317533493042, -0.1420619934797287, 0.032154180109500885, 0.18435074388980865, 0.21136410534381866, -0.10405104607343674, 0.1773526966571808, -0.0519658625125885, 0.25296083092689514, 0.20239883661270142, 0.013077356852591038, -0.04030042141675949, 0.22695480287075043, 0.1543976366519928, 0.03185997158288956, -0.12105312943458557, -0.2936055660247803, 0.09633102267980576, -0.06996113061904907, 0.11004971712827682, 0.2714962363243103, -0.0030551734380424023, 0.6472854614257812, 0.05717718228697777, 0.31057435274124146, 0.2889772355556488, -0.12562067806720734, 0.059007491916418076, -0.2301301211118698, 0.14177140593528748, 0.2295093685388565, 0.2583223283290863, -0.1393100768327713, -0.19662682712078094, 0.5935571193695068, 0.31629860401153564, -0.07326284050941467, 0.38970208168029785, -0.4620021879673004, -0.2712932527065277, 0.25234127044677734, 0.027918420732021332, 0.6917434930801392, -0.02594769559800625, 0.26716285943984985, 0.06503325700759888, -0.3845381736755371, 0.6287998557090759, -0.39091193675994873, 0.1575743556022644, -0.07772570103406906, -0.2354031801223755, 0.032601434737443924, -0.09457404166460037, -0.0697193443775177, 0.26830849051475525, -0.3717157244682312, 0.34969374537467957, 0.47733697295188904, -0.20441009104251862, -0.05117984488606453, 0.5518012642860413, -0.029192624613642693, -0.40027934312820435, -0.10723701119422913, 0.08906105160713196, -0.2812136709690094, 0.10603366792201996, -0.163366436958313, -0.2763771712779999, 0.2486419677734375, 0.11350016295909882, -0.27409279346466064, 0.11336208134889603, -0.03516485169529915, 0.1855694204568863, 0.26483961939811707, 0.016313914209604263, 0.2512062191963196, -0.17889244854450226, -0.06662282347679138, -0.15494702756404877, -0.08863630890846252, 0.1312912255525589, -0.14661245048046112, 0.05025870352983475, 0.03055552951991558, -0.20452803373336792, 0.5105487108230591, -0.07615143805742264, -0.10789072513580322, -0.24377934634685516, 0.244764506816864, -0.43671685457229614, -0.1542506217956543, 0.11010678857564926, 0.168225958943367, 0.21321596205234528, -0.3162398636341095, 0.6001379489898682, 0.08641368895769119, -0.009584017097949982, 0.055802371352910995, 0.17730115354061127, -0.23682624101638794, 0.4851782023906708, -0.2578953504562378, -0.004609816242009401, 0.052440181374549866, 0.010878310538828373, -0.16225440800189972, -0.37275442481040955, -0.03353330120444298, 0.2981375455856323, 0.0863661840558052, -0.016165120527148247, -0.002981772879138589, 0.10210438817739487, -0.07975661009550095, -0.0030079155694693327, -0.15544813871383667, -0.3322069048881531, 0.11653779447078705, 0.14724449813365936, 0.1947060227394104, 0.16701354086399078, -0.05395311862230301, -0.24001483619213104, -0.1859925240278244, 0.46923404932022095, 0.04061912000179291, 0.2234194427728653, -0.1361563801765442, 0.3405205011367798, 0.42906028032302856, 0.5344378352165222, -0.28160959482192993, 0.07217354327440262, 0.032790567725896835, 0.08643452078104019, 0.48312050104141235, -0.23492644727230072, 0.07466652989387512, 0.27867862582206726, -0.11924906075000763, 0.04919916018843651, -0.34023603796958923, -0.17272847890853882, -0.25473615527153015, 0.09835324436426163, 0.11223389953374863, -0.0018388299504294991, 0.00541333481669426, -0.061273135244846344, -0.07910183817148209, -0.08158956468105316, 0.016571475192904472, -0.3894312381744385, -0.06932209432125092, 0.15564671158790588, -0.11541114002466202, 0.015997860580682755, 0.05923876166343689, 0.20302237570285797, 0.009793905541300774, -0.05242832377552986, 0.2911643087863922, 0.1270984262228012, 0.1111840307712555, -0.19097968935966492, -0.17631743848323822, 0.004692028276622295, 0.21382662653923035, 0.14815224707126617, -0.02907835692167282, 0.16041800379753113, -0.2343686819076538, -0.14009541273117065, 0.2342924177646637, 0.14868582785129547, 0.020620213821530342, -0.12043069303035736, 0.10565533488988876, 0.14062292873859406, -0.12489914149045944, -0.028531499207019806, 0.1781449317932129, 0.11084775626659393, 0.136631578207016, 0.11059911549091339, 0.1587066799402237, 0.4499141573905945, -0.21001610159873962, 0.025973504409193993, 0.1152660921216011, -0.36859607696533203, 0.12917551398277283, -0.1154741570353508, -0.04045641049742699, 0.0053083086386322975, 0.14748279750347137, -0.011129156686365604, 0.13959269225597382, -0.3216474950313568, 0.30607345700263977, 0.30969545245170593, -0.02371235564351082, 0.058645982295274734, 0.23911476135253906, 0.014807509258389473, -0.13310997188091278, 0.3046792149543762, 0.01857638545334339, 0.31923991441726685, 0.042773354798555374, 0.8754273653030396, -0.27222734689712524, -0.45505979657173157, 0.10648257285356522, 0.4365844428539276, 0.032111313194036484, -0.21552824974060059, 0.15521696209907532, 0.09118610620498657, -0.4059763550758362, 0.12097818404436111, 0.05810832232236862, -0.16374598443508148, 0.6574726700782776, -0.11649464815855026, 0.16466756165027618, -0.15811505913734436, -0.14994023740291595, 0.35239651799201965, 0.19673563539981842, -0.09543825685977936, 0.010715498588979244, 0.11992550641298294, -0.11069333553314209, -0.16985248029232025, 0.5139837861061096, 0.2285720407962799, 0.03479337692260742, -0.4124753475189209, 0.037352025508880615, 0.1071360632777214, 0.14290305972099304, -0.07321161776781082, 0.13803668320178986, 0.4449816346168518, -0.18126708269119263, 0.2542765736579895, 0.03678722679615021, -0.07882057130336761, -0.1354016214609146, 0.043384332209825516, -0.38847383856773376, -0.39733758568763733, 0.45656660199165344, 0.10967732965946198, -0.11010126024484634, 0.04526491090655327, 0.3207000195980072, -0.6264183521270752, -0.2374822050333023, 0.3614813983440399, -0.13628998398780823, -0.050093259662389755, -0.24483540654182434, 0.08715888112783432, 0.08407583087682724, 0.6122644543647766, 0.2146088182926178, 0.353493869304657, -0.40171435475349426, -0.14038510620594025, -0.5742546319961548, -0.21603325009346008, 0.17194122076034546, -0.07661692798137665, -0.25840142369270325, 0.14676734805107117, -0.06859195977449417, 0.2895055115222931, -0.2140117883682251, 0.11786480993032455, 0.17773236334323883, 0.027057355269789696, -0.43053895235061646, 0.11578340828418732, 0.05879133567214012, 0.004421604797244072, 0.31976118683815, -0.11052566021680832, -0.17421768605709076, 0.08242619782686234, -0.056964173913002014, -0.09088997542858124, -0.06792833656072617, 0.2378150075674057, 0.16779834032058716, 0.09648943692445755, 0.08136232197284698, 0.46605226397514343, -0.04932507500052452, -0.20165443420410156, -0.008785558864474297, 0.012513049878180027, -0.3770976960659027, -0.16616779565811157, -0.07718373835086823, 0.1801462024450302, -0.334835022687912, 0.254971444606781, -0.016634074971079826, -0.08482349663972855, 0.11921888589859009, 0.018418781459331512, -0.18934886157512665, -0.049457862973213196, 0.0742049515247345, 0.19506916403770447, 0.22280582785606384, 0.7319663763046265, -0.010562333278357983, 0.08575459569692612, -0.0929856076836586, -0.03421347588300705, 0.0708661824464798, -0.052112411707639694, -0.20773270726203918, -0.13613715767860413, 0.024237556383013725, 0.09727275371551514, 0.13568124175071716, -0.6967564225196838, -0.33582937717437744, 0.014331682585179806, -0.022863570600748062, 0.03177705034613609, 0.7183232307434082, 0.06665355712175369, -0.14486569166183472, -0.1175316646695137, -0.29869967699050903, 0.05342429131269455, -0.2820819318294525, 0.1892062872648239, -0.34770864248275757 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
This is an interesting discussion indeed and it would be nice to make multi-task easier. Probably the best would be to have a new type of dataset especially designed for that in order to easily combine and sample from the multiple datasets. This way we could probably handle the combination of datasets with differing schemas as well (unlike T5).
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
59
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. This is an interesting discussion indeed and it would be nice to make multi-task easier. Probably the best would be to have a new type of dataset especially designed for that in order to easily combine and sample from the multiple datasets. This way we could probably handle the combination of datasets with differing schemas as well (unlike T5).
[ -0.07760483026504517, -0.43013614416122437, -0.051281098276376724, -0.007648616097867489, -0.24152909219264984, 0.11484125256538391, 0.1743364930152893, 0.09231007844209671, 0.285217821598053, -0.11796212196350098, -0.24373626708984375, 0.27350908517837524, -0.19464397430419922, 0.4169623553752899, 0.39820030331611633, -0.456034392118454, 0.007996998727321625, -0.21966670453548431, -0.2606221139431, 0.3799695670604706, 0.041827648878097534, 0.0955592468380928, -0.1473454236984253, 0.06278283149003983, -0.29021939635276794, -0.1371731162071228, -0.34358662366867065, 0.08050740510225296, 0.10446877777576447, 0.03401508182287216, 0.19900530576705933, 0.37377819418907166, 0.008996821939945221, 0.10891860723495483, -0.00011252854892518371, -0.09094025939702988, 0.16050881147384644, -0.07122159749269485, 0.13028468191623688, -0.02305193617939949, 0.1236008033156395, -0.3596465587615967, 0.06394360214471817, -0.10716301202774048, 0.18151821196079254, 0.1867564618587494, 0.2017911970615387, 0.0839318186044693, 0.3181018829345703, -0.315500408411026, 0.09998155385255814, 0.29038387537002563, -0.1883256584405899, 0.28077974915504456, -0.05406222492456436, -0.06420397013425827, 0.0306809414178133, 0.22856402397155762, 0.6673135757446289, -0.5304595828056335, 0.0446113683283329, 0.06839294731616974, -0.027133867144584656, 0.04156392440199852, 0.11788813769817352, -0.08518829196691513, -0.3070951998233795, -0.23946762084960938, -0.3735044598579407, 0.28026261925697327, -0.07192567735910416, -0.10643959790468216, -0.23320792615413666, -0.533369243144989, 0.015733862295746803, 0.07696186006069183, -0.295244961977005, 0.059608977288007736, -0.16279707849025726, -0.01917090453207493, -0.12912997603416443, -0.12187929451465607, -0.001455985358916223, 0.1736188530921936, 0.308641642332077, 0.5555923581123352, 0.06791847199201584, 0.16606776416301727, 0.3354537785053253, 0.06970095634460449, -0.0984867662191391, -0.28932738304138184, 0.28367510437965393, 0.0658087208867073, -0.4028248190879822, -0.3567551076412201, 0.1301075965166092, -0.2654320299625397, 0.2804775834083557, 0.1396418660879135, 0.16255934536457062, 0.09564366191625595, -0.16685377061367035, 0.2518713176250458, 0.33976128697395325, -0.022479169070720673, -0.12287195026874542, -0.13603028655052185, 0.017504345625638962, -0.10887712240219116, 0.16654396057128906, 0.13610005378723145, 0.17564235627651215, -0.018662475049495697, -0.42686548829078674, -0.11984287947416306, -0.14647647738456726, 0.09820450097322464, -0.1879584938287735, -0.6031385660171509, -0.09075374156236649, -0.10779064893722534, 0.031006425619125366, -0.15766018629074097, -0.2877175807952881, 0.09265389293432236, -0.10346339643001556, 0.24397864937782288, -0.1317925602197647, 0.008540723472833633, -0.048576995730400085, 0.06551449745893478, -0.423929363489151, -0.048491448163986206, 0.2941402792930603, 0.09935479611158371, 0.048311349004507065, 0.1649857759475708, -0.0342780239880085, 0.1286846101284027, 0.4205945134162903, 0.029048264026641846, 0.014122365973889828, -0.03673812001943588, 0.1730395406484604, -0.3192977011203766, -0.011867192573845387, 0.22267630696296692, -0.3717392385005951, -0.11837562173604965, -0.16984780132770538, -0.21736453473567963, 0.09760810434818268, 0.03343379497528076, -0.21831905841827393, -0.3184991478919983, -0.4913881719112396, 0.712941586971283, -0.04474186152219772, -0.046462416648864746, -0.16467730700969696, 0.2666390538215637, -0.23226146399974823, -0.0979924350976944, 0.16326270997524261, 0.08497123420238495, 0.013149724341928959, -0.29495877027511597, -0.093130923807621, -0.1038350835442543, -0.06481102108955383, 0.3017260730266571, -0.1689356118440628, 0.21227331459522247, 0.06585868448019028, 0.13697485625743866, 0.24945685267448425, -0.45365026593208313, -0.04889877513051033, 0.33244872093200684, -0.024968815967440605, 0.35705482959747314, 0.07683292776346207, 0.21231018006801605, -0.0030687025282531977, -0.2040671557188034, 0.1949269026517868, 1.0177953243255615, -0.3952714800834656, -0.021733995527029037, -0.11022838950157166, -0.3906150162220001, 0.5425782203674316, 0.3610698878765106, 0.06735029071569443, -0.33101290464401245, -0.13180582225322723, -0.008706217631697655, 0.09087147563695908, 0.07975579053163528, 0.04832073301076889, -0.06289375573396683, -0.3434605300426483, 0.04504111781716347, -0.2458025962114334, -0.19731979072093964, -0.41920289397239685, -0.02027900703251362, 0.01695721410214901, 0.2923034131526947, 0.3340207040309906, -0.10002617537975311, 0.2713525891304016, -0.30399465560913086, -0.15376164019107819, -0.3133300840854645, 0.1197587177157402, -0.00561920739710331, 0.03799621760845184, -0.24834011495113373, -0.030675068497657776, 0.3191802203655243, 0.09926758706569672, -0.13511922955513, -0.4337625801563263, 0.3398182690143585, -0.047642629593610764, 0.04676089435815811, -0.046398308128118515, 0.6206659078598022, -0.1988632082939148, -0.091585673391819, 0.17018118500709534, 0.2487032115459442, -0.3234957754611969, 0.2560332715511322, 0.32414066791534424, 0.35003411769866943, 0.1919659674167633, -0.08836990594863892, 0.08618849515914917, -0.012060612440109253, -0.05773824825882912, -0.1945427507162094, -0.08406363427639008, 0.4658958613872528, -0.18547096848487854, 0.41023874282836914, -0.053562261164188385, 0.01475467812269926, -0.19410163164138794, -0.05506950989365578, -0.18381237983703613, 0.410701185464859, 0.3079739212989807, -0.19726142287254333, 0.33659157156944275, -0.11086128652095795, -0.3630489110946655, 0.19674399495124817, 0.1851436048746109, 0.05677737668156624, 0.07615210860967636, -0.032488953322172165, -0.01916598528623581, -0.29046374559402466, -0.11108561605215073, 0.4045727550983429, 0.6323988437652588, 0.2805095911026001, 0.11706043779850006, -0.021244529634714127, -0.250074565410614, -0.10282382369041443, 0.011356702074408531, 0.2060648500919342, -0.01684323325753212, 0.1502837836742401, 0.04689272120594978, 0.0736096128821373, 0.04397176206111908, -0.14593523740768433, 0.19349056482315063, -0.17940062284469604, -0.049568526446819305, 0.0246107317507267, -0.19285333156585693, -0.565240740776062, -0.21911394596099854, 0.043787963688373566, -0.3279033303260803, -0.30137449502944946, 0.15034201741218567, 0.053340278565883636, -0.176954984664917, 0.15528932213783264, 0.17466263473033905, 0.5224672555923462, -0.44439440965652466, -0.23384326696395874, -0.06413089483976364, -0.2828475832939148, -0.08073630928993225, 0.11607662588357925, 0.49301403760910034, 0.28585025668144226, 0.45890289545059204, 0.2643531262874603, -0.04553234577178955, 0.003795946016907692, -0.3939478099346161, 0.029010996222496033, -0.3069179952144623, 0.27917319536209106, 0.0830267146229744, -0.2171400934457779, 0.1505551040172577, -0.39499610662460327, -0.06443852931261063, -0.04882242903113365, -0.07792926579713821, -0.028693009167909622, -0.02504764124751091, 0.08972270041704178, -0.17804285883903503, -0.2269403636455536, -0.6084064245223999, -0.4707540273666382, 0.2826780378818512, -0.2821083962917328, 0.07274046540260315, 0.03876468166708946, -0.2779347598552704, 0.15081925690174103, -0.15593421459197998, 0.028028614819049835, -0.15452177822589874, -0.1129903793334961, -0.018640007823705673, -0.36685502529144287, -0.159623920917511, -0.41231489181518555, -0.07378389686346054, 0.22635845839977264, -0.0729171633720398, -0.04983562231063843, -0.22019991278648376, 0.06068329140543938, -0.10324574261903763, 0.05648218095302582, 0.17424607276916504, 0.171852707862854, -0.22455748915672302, 0.11065208166837692, -0.041232429444789886, -0.20303432643413544, 0.20850946009159088, 0.3692528307437897, 0.39855581521987915, 0.17247514426708221, -0.04534241184592247, 0.1586776077747345, 0.6988725662231445, 0.22570137679576874, 0.07469514012336731, 0.19577351212501526, 0.300357848405838, -0.008464815095067024, 0.14122061431407928, -0.23328590393066406, 0.21620100736618042, -0.1369471400976181, 0.49532949924468994, 0.11535341292619705, 0.15646055340766907, -0.1784890741109848, -0.2667299509048462, 0.05356646701693535, -0.2868155241012573, -0.18034467101097107, 0.2325088530778885, -0.2959993779659271, 0.0018163897329941392, -0.06489153206348419, -0.4627395570278168, -0.2831376791000366, -0.049076251685619354, -0.028592674061655998, 0.30236130952835083, -0.018328126519918442, 0.162306547164917, -0.3587016463279724, -0.023968331515789032, -0.35949310660362244, 0.16688688099384308, 0.09561176598072052, 0.25671523809432983, -0.08165322989225388, -0.045285165309906006, -0.07541736215353012, 0.1324964016675949, 0.5941112637519836, -0.5726916193962097, -0.11468014121055603, 0.112262062728405, 0.06943973153829575, -0.10702923685312271, -0.1124657392501831, -0.2531118094921112, -0.02277839556336403, 0.4108986258506775, 0.3724684715270996, -0.35121721029281616, -0.335947722196579, 0.2770611643791199, -0.17802587151527405, -0.06660338491201401, -0.1773483157157898, -0.1805446296930313, -0.1992291957139969, -0.11174308508634567, -0.015012826770544052, 0.22503121197223663, 0.23601850867271423, -0.09397926181554794, 0.1691286414861679, -0.11902878433465958, 0.2343083918094635, 0.1777767837047577, -0.020016608759760857, -0.08604403585195541, 0.21967290341854095, 0.19012922048568726, 0.04374001920223236, -0.12046536803245544, -0.28548550605773926, 0.09027766436338425, -0.08531799167394638, 0.0410679392516613, 0.32314565777778625, -0.01695922203361988, 0.6543559432029724, 0.0035479131620377302, 0.35406097769737244, 0.29701948165893555, -0.15183062851428986, 0.035119399428367615, -0.22525554895401, 0.15921857953071594, 0.19328385591506958, 0.26458102464675903, -0.16709142923355103, -0.22239522635936737, 0.5615090727806091, 0.323903888463974, -0.041719160974025726, 0.39511552453041077, -0.5345010757446289, -0.2746390104293823, 0.2725147008895874, 0.029301423579454422, 0.6709384322166443, -0.08567716926336288, 0.2816002070903778, 0.012769940309226513, -0.4185323417186737, 0.6093616485595703, -0.35631999373435974, 0.14552365243434906, -0.07994025200605392, -0.2315615564584732, 0.021089931949973106, -0.07193577289581299, -0.09989561140537262, 0.2743530571460724, -0.34961915016174316, 0.3014124631881714, 0.4439023435115814, -0.18051087856292725, -0.03987286239862442, 0.5531432032585144, -0.015312721021473408, -0.3570755124092102, -0.03860980644822121, 0.06670382618904114, -0.2900905907154083, 0.07109297811985016, -0.15476760268211365, -0.25153324007987976, 0.24413533508777618, 0.19273681938648224, -0.3443203568458557, 0.0698326826095581, 0.01897631399333477, 0.18553534150123596, 0.22446347773075104, -0.0021209921687841415, 0.29660794138908386, -0.16361679136753082, -0.031092317774891853, -0.15915141999721527, -0.11198915541172028, 0.12094797939062119, -0.18845416605472565, 0.004920710809528828, 0.04263301193714142, -0.19619153439998627, 0.5082732439041138, -0.05026707425713539, -0.17390115559101105, -0.2432447224855423, 0.277166485786438, -0.46236884593963623, -0.23436476290225983, 0.12246417999267578, 0.19409802556037903, 0.2181558907032013, -0.31469687819480896, 0.5614801049232483, 0.12799659371376038, -0.01107463613152504, 0.04455440863966942, 0.13490228354930878, -0.2083812952041626, 0.4753072261810303, -0.28534382581710815, -0.025862539187073708, 0.06688446551561356, -0.004448126070201397, -0.09673965722322464, -0.37176454067230225, -0.04464844986796379, 0.31770825386047363, 0.16569313406944275, -0.028410430997610092, 0.07400819659233093, 0.1382138431072235, -0.0919288694858551, -0.007477260194718838, -0.16431015729904175, -0.2981978952884674, 0.17177744209766388, 0.18262004852294922, 0.2225334793329239, 0.18226587772369385, -0.07159313559532166, -0.2359989434480667, -0.12219997495412827, 0.4769056439399719, -0.008986275643110275, 0.2460334300994873, -0.13987435400485992, 0.38459840416908264, 0.43049702048301697, 0.5406709909439087, -0.25706741213798523, 0.12530726194381714, 0.043342381715774536, 0.08136989921331406, 0.5019499659538269, -0.22554437816143036, 0.038004714995622635, 0.27418315410614014, -0.14212439954280853, 0.04171774163842201, -0.2888586223125458, -0.16853700578212738, -0.2163936346769333, 0.11692991852760315, 0.12696069478988647, 0.024143647402524948, 0.03174271434545517, -0.05489133670926094, -0.12920565903186798, -0.0581381693482399, -0.010788447223603725, -0.3781214952468872, -0.04661114141345024, 0.1347699910402298, -0.06214950606226921, 0.04813915863633156, 0.04555444419384003, 0.24881473183631897, -0.0054840268567204475, -0.0489119328558445, 0.3026912808418274, 0.11371728032827377, 0.13134901225566864, -0.19243119657039642, -0.2258504331111908, -0.02276800200343132, 0.27811509370803833, 0.16005359590053558, -0.08049104362726212, 0.15094336867332458, -0.2103305459022522, -0.14571769535541534, 0.19737198948860168, 0.10546549409627914, 0.06881015002727509, -0.0883578211069107, 0.14663808047771454, 0.10758338123559952, -0.07554765045642853, 0.013402373529970646, 0.19421444833278656, 0.0746556743979454, 0.07250716537237167, 0.10837621241807938, 0.1365598887205124, 0.47358953952789307, -0.22609271109104156, 0.0031988718546926975, 0.12324982136487961, -0.35941773653030396, 0.15418677031993866, -0.08749885112047195, -0.06692657619714737, -0.031022604554891586, 0.1314770132303238, -0.0590047650039196, 0.13142883777618408, -0.3177984356880188, 0.31435203552246094, 0.2902786135673523, -0.016615120694041252, -0.023868974298238754, 0.2444927990436554, 0.03816697373986244, -0.1479455828666687, 0.2975655496120453, 0.0424143448472023, 0.3625641465187073, 0.03182356432080269, 0.898205041885376, -0.2870635688304901, -0.4490215480327606, 0.08506857603788376, 0.44423046708106995, 0.09313908219337463, -0.21840853989124298, 0.12272954732179642, 0.10584560036659241, -0.33367180824279785, 0.12666074931621552, 0.027429604902863503, -0.16793420910835266, 0.6969411373138428, -0.08871020376682281, 0.13946978747844696, -0.1556522250175476, -0.16006018221378326, 0.3791298270225525, 0.15306228399276733, -0.08720052987337112, 0.005124494433403015, 0.11803163588047028, -0.11495508253574371, -0.1713058203458786, 0.5400868654251099, 0.19663292169570923, -0.016987159848213196, -0.441108763217926, -0.00291931233368814, 0.0928228572010994, 0.18584506213665009, -0.00310856057330966, 0.11664769053459167, 0.4275524914264679, -0.17765876650810242, 0.2757039964199066, 0.016501182690262794, -0.042451582849025726, -0.12721534073352814, 0.04194430634379387, -0.39963585138320923, -0.44053196907043457, 0.4334969222545624, 0.1266341358423233, -0.10133050382137299, 0.12239079922437668, 0.3422192633152008, -0.6254709362983704, -0.21663473546504974, 0.3283933103084564, -0.1351538449525833, -0.09978533536195755, -0.21278643608093262, 0.08299475908279419, 0.052062828093767166, 0.626800000667572, 0.2669466435909271, 0.3559400737285614, -0.36786213517189026, -0.10077381134033203, -0.5913134813308716, -0.1538391411304474, 0.15314769744873047, -0.07130128890275955, -0.22856473922729492, 0.1566452980041504, -0.04573491960763931, 0.2795036733150482, -0.1662362664937973, 0.1557861864566803, 0.22006599605083466, 0.06328561902046204, -0.4629291892051697, 0.058169737458229065, 0.15672561526298523, 0.0008710448746569455, 0.2697933614253998, -0.09562113136053085, -0.18361033499240875, 0.10721495002508163, -0.08659322559833527, -0.1330372840166092, -0.08308341354131699, 0.27850767970085144, 0.14434286952018738, 0.12579546868801117, 0.060522209852933884, 0.48311787843704224, -0.013349862769246101, -0.1953572779893875, 0.00032312431721948087, -0.017219215631484985, -0.36595985293388367, -0.13347648084163666, -0.1121906116604805, 0.16162455081939697, -0.40695810317993164, 0.31124192476272583, 0.020268002524971962, -0.11749874800443649, 0.07889232784509659, 0.08206035941839218, -0.19863153994083405, -0.10771983861923218, 0.09251314401626587, 0.1740236133337021, 0.22926786541938782, 0.7064042687416077, -0.009655475616455078, 0.08157208561897278, -0.08162645250558853, -0.02464248239994049, 0.013805198483169079, 0.0019771098159253597, -0.23659086227416992, -0.09721981734037399, 0.035987768322229385, 0.04707091301679611, 0.14540311694145203, -0.6879175305366516, -0.3390016555786133, 0.07124500721693039, 0.008100426755845547, 0.060268107801675797, 0.7683094143867493, 0.13407014310359955, -0.143321231007576, -0.10919411480426788, -0.3065090775489807, 0.028837883844971657, -0.2758442163467407, 0.1856124997138977, -0.33611053228378296 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
@thomwolf Are you suggesting making a wrapper class which can take existing datasets as arguments and do all the required sampling/combining, to present the same interface as a normal dataset? That doesn't seem too complicated to implement.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
37
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. @thomwolf Are you suggesting making a wrapper class which can take existing datasets as arguments and do all the required sampling/combining, to present the same interface as a normal dataset? That doesn't seem too complicated to implement.
[ -0.08985123783349991, -0.3969317376613617, -0.03166533261537552, -0.03735226392745972, -0.22644852101802826, 0.08738210052251816, 0.1969548612833023, 0.06602370738983154, 0.3327305018901825, -0.13962101936340332, -0.22614829242229462, 0.3318242132663727, -0.20250579714775085, 0.3645176589488983, 0.40155625343322754, -0.5019195079803467, -0.00903159286826849, -0.17320957779884338, -0.3404359221458435, 0.34348350763320923, 0.015904292464256287, 0.08975981175899506, -0.12996898591518402, 0.05399256572127342, -0.265904039144516, -0.1480942666530609, -0.3480377197265625, 0.10398421436548233, 0.09025131165981293, 0.00020966396550647914, 0.2654457688331604, 0.3834894001483917, -0.004184514284133911, 0.11468610912561417, -0.0001109990116674453, -0.10409476608037949, 0.17799261212348938, -0.061887506395578384, 0.13401608169078827, -0.022733453661203384, 0.2059173732995987, -0.3741094470024109, 0.041295375674963, -0.14299559593200684, 0.09510379284620285, 0.18651209771633148, 0.1486475020647049, 0.061315685510635376, 0.4389658272266388, -0.3271811604499817, 0.1137162521481514, 0.27153128385543823, -0.2127838283777237, 0.2160601168870926, -0.10353539884090424, -0.0817168578505516, 0.03618598356842995, 0.294839471578598, 0.6407869458198547, -0.5255510807037354, -0.026187967509031296, 0.09548573195934296, -0.02736244723200798, -0.02691737376153469, 0.16530920565128326, -0.05579410865902901, -0.35311025381088257, -0.23468150198459625, -0.39337629079818726, 0.2618134021759033, -0.19633138179779053, -0.1110711395740509, -0.22825059294700623, -0.5546971559524536, 0.043270938098430634, 0.05650657042860985, -0.3279976546764374, 0.04984421283006668, -0.16306190192699432, 0.015065819956362247, -0.16031573712825775, -0.1137707307934761, -0.035134896636009216, 0.15094515681266785, 0.3342219293117523, 0.5391830801963806, 0.0717264860868454, 0.18103110790252686, 0.3762325942516327, 0.06643190234899521, -0.1559692770242691, -0.25584742426872253, 0.26040151715278625, 0.1465931534767151, -0.300144225358963, -0.38151079416275024, 0.1029033213853836, -0.20207493007183075, 0.2648245692253113, 0.16467426717281342, 0.09348071366548538, 0.12698085606098175, -0.11747182905673981, 0.28201189637184143, 0.3259015381336212, -0.0005840566009283066, -0.05843743309378624, -0.08932558447122574, 0.0330900102853775, -0.0813913643360138, 0.16326574981212616, 0.12141142785549164, 0.17396438121795654, 0.061765316873788834, -0.4309350252151489, -0.12424056231975555, -0.20832139253616333, 0.1426924467086792, -0.126838818192482, -0.5166338086128235, -0.06742701679468155, -0.08390867710113525, 0.06903038918972015, -0.0892474427819252, -0.30100715160369873, 0.11446855962276459, -0.022935623303055763, 0.21466897428035736, -0.06969887763261795, 0.014182898215949535, -0.05495680496096611, 0.11815588921308517, -0.4216562807559967, -0.016975609585642815, 0.3122235834598541, 0.09955655783414841, 0.09712197631597519, 0.17393361032009125, -0.07775605469942093, 0.12298405170440674, 0.42270317673683167, 0.029997248202562332, 0.028817351907491684, -0.0377977229654789, 0.14484618604183197, -0.32560497522354126, -0.0064032720401883125, 0.18319672346115112, -0.3624859154224396, -0.15414012968540192, -0.09662557393312454, -0.16510240733623505, 0.12379063665866852, 0.059437792748212814, -0.22519667446613312, -0.2789829671382904, -0.3781821131706238, 0.729292631149292, -0.06950287520885468, -0.061432428658008575, -0.172153040766716, 0.29769429564476013, -0.25143373012542725, -0.07137696444988251, 0.18124252557754517, 0.07104931026697159, -0.03282120078802109, -0.38990500569343567, -0.10039648413658142, -0.10813358426094055, -0.07206184417009354, 0.26875007152557373, -0.20015212893486023, 0.2165667712688446, 0.040044065564870834, 0.16750991344451904, 0.2588682472705841, -0.4050794839859009, -0.1015973761677742, 0.2870194911956787, -0.037553898990154266, 0.385880708694458, 0.07683313637971878, 0.24547535181045532, 0.08919375389814377, -0.17639228701591492, 0.2526645064353943, 1.0071953535079956, -0.39769595861434937, -0.0693170502781868, -0.10972343385219574, -0.3863369822502136, 0.5040088295936584, 0.36682891845703125, 0.10829998552799225, -0.23684120178222656, -0.149130716919899, -0.05921320989727974, 0.06534463167190552, 0.04500467702746391, 0.10629868507385254, -0.081968292593956, -0.3427741527557373, 0.030671551823616028, -0.21699653565883636, -0.20223873853683472, -0.36724212765693665, -0.002258129185065627, -0.0650896281003952, 0.3331250250339508, 0.37277641892433167, -0.19358041882514954, 0.18135933578014374, -0.2651146352291107, -0.0785318911075592, -0.2893357574939728, 0.13996616005897522, -0.07206754386425018, 0.057969458401203156, -0.2934384346008301, -0.0477028451859951, 0.32401835918426514, 0.0663003996014595, -0.0903622955083847, -0.4725525379180908, 0.3164876103401184, -0.0660732313990593, 0.041039735078811646, -0.0641026571393013, 0.6213023066520691, -0.16827817261219025, -0.13204868137836456, 0.21507923305034637, 0.22514618933200836, -0.3285652995109558, 0.21186058223247528, 0.33674418926239014, 0.24053128063678741, 0.17107154428958893, -0.05437089875340462, 0.10397067666053772, 0.019159851595759392, -0.012393715791404247, -0.22969649732112885, -0.1085934117436409, 0.5694276094436646, -0.06802866607904434, 0.4326317608356476, -0.02600567601621151, 0.042701829224824905, -0.183145210146904, -0.11107372492551804, -0.24281013011932373, 0.38110750913619995, 0.29904991388320923, -0.21203316748142242, 0.3523603677749634, -0.05134269967675209, -0.3553398847579956, 0.1363971084356308, 0.13285301625728607, 0.0007288824999704957, 0.08132301270961761, 0.018113886937499046, -0.0639314204454422, -0.24718452990055084, -0.09759605675935745, 0.4116675555706024, 0.5995680093765259, 0.2978721857070923, 0.14136429131031036, 0.07985905557870865, -0.25959134101867676, -0.07501431554555893, 0.03891141712665558, 0.19106259942054749, -0.07945368438959122, 0.15421098470687866, 0.010218946263194084, 0.05711266025900841, -0.015599183738231659, -0.09665288031101227, 0.12611635029315948, -0.2244979590177536, -0.025746598839759827, 0.040470022708177567, -0.15040068328380585, -0.5861167907714844, -0.23188799619674683, 0.00516132777556777, -0.3123072683811188, -0.26452499628067017, 0.14035704731941223, 0.0393286794424057, -0.17840610444545746, 0.1565386801958084, 0.18521888554096222, 0.6711629033088684, -0.45885545015335083, -0.22821243107318878, -0.057080842554569244, -0.2689543068408966, -0.11032195389270782, 0.1257171481847763, 0.4261064827442169, 0.2776329219341278, 0.5188310146331787, 0.3089578151702881, -0.0348958820104599, 0.06895487755537033, -0.39623183012008667, 0.057826314121484756, -0.26608580350875854, 0.31058469414711, 0.11345035582780838, -0.22008152306079865, 0.17031830549240112, -0.3742896318435669, 0.0022690151818096638, -0.10265439748764038, -0.03408007323741913, -0.07405845075845718, -0.04390879347920418, 0.082270048558712, -0.2027611881494522, -0.1946837157011032, -0.6382985711097717, -0.4739052951335907, 0.30521366000175476, -0.21334022283554077, 0.0697750672698021, 0.0194583460688591, -0.2589375376701355, 0.17852003872394562, -0.13915301859378815, -0.013688018545508385, -0.194584921002388, -0.1289622187614441, 0.01676088571548462, -0.33759644627571106, -0.16410097479820251, -0.4051862955093384, -0.20086124539375305, 0.22619332373142242, -0.09001310169696808, -0.06370017677545547, -0.25243067741394043, 0.07718200981616974, -0.12796762585639954, 0.11129260808229446, 0.21998296678066254, 0.15117621421813965, -0.26820701360702515, 0.09364607185125351, -0.02915448322892189, -0.20805451273918152, 0.25715240836143494, 0.29497891664505005, 0.3784078359603882, 0.15610337257385254, 0.038204535841941833, 0.13332608342170715, 0.7063826322555542, 0.2446940690279007, 0.03195121884346008, 0.14577054977416992, 0.3078969717025757, 0.017254112288355827, 0.1577184498310089, -0.22609028220176697, 0.1919374316930771, -0.17036141455173492, 0.4489407241344452, 0.10616748780012131, 0.1941898614168167, -0.20405130088329315, -0.2649405896663666, 0.05006057769060135, -0.2274237722158432, -0.23034359514713287, 0.2668198049068451, -0.1887030452489853, 0.011721063405275345, -0.0904015377163887, -0.3691737651824951, -0.24698826670646667, -0.022600578144192696, 0.016348861157894135, 0.2658742070198059, -0.023195896297693253, 0.12559135258197784, -0.4113301932811737, -0.03634222224354744, -0.4307921528816223, 0.18226023018360138, 0.07883033156394958, 0.26000627875328064, -0.06049313396215439, -0.07115352153778076, -0.08951002359390259, 0.10068045556545258, 0.5856118202209473, -0.5367434620857239, -0.17423437535762787, 0.11988352239131927, 0.07640207558870316, -0.14059586822986603, -0.06939791887998581, -0.2734319567680359, -0.0443900004029274, 0.45053836703300476, 0.3297438621520996, -0.35634878277778625, -0.2867952287197113, 0.2995762228965759, -0.172275573015213, -0.09658363461494446, -0.20668724179267883, -0.1536155790090561, -0.19949345290660858, -0.13871638476848602, 0.054717954248189926, 0.18321509659290314, 0.1596936136484146, -0.1042323186993599, 0.12230967730283737, -0.11102616041898727, 0.21563111245632172, 0.19780443608760834, 0.009141841903328896, -0.04508168622851372, 0.21428607404232025, 0.1789846420288086, 0.0323692262172699, -0.14142338931560516, -0.3413662016391754, 0.12961816787719727, -0.06286953389644623, 0.12924735248088837, 0.23772482573986053, -0.015323036350309849, 0.6900312900543213, 0.04273298382759094, 0.3528764843940735, 0.31852850317955017, -0.10639546811580658, 0.009463205933570862, -0.30537208914756775, 0.13097569346427917, 0.1652955710887909, 0.2444354146718979, -0.11793605983257294, -0.22836337983608246, 0.5872110724449158, 0.3297063708305359, -0.08609943836927414, 0.3757956624031067, -0.4737700819969177, -0.31049031019210815, 0.32061317563056946, 0.048291679471731186, 0.735260009765625, -0.020149990916252136, 0.29032987356185913, 0.060059141367673874, -0.3805786073207855, 0.6396000981330872, -0.34959062933921814, 0.1266002506017685, -0.13993461430072784, -0.2060180902481079, 0.00896663311868906, -0.08151312172412872, -0.061738207936286926, 0.2745799124240875, -0.36089715361595154, 0.3065803647041321, 0.375642865896225, -0.22479835152626038, -0.026937691494822502, 0.5579147338867188, -0.028486115857958794, -0.40201860666275024, -0.0726892277598381, 0.10170169919729233, -0.2750028371810913, 0.06954915076494217, -0.16664576530456543, -0.2324201613664627, 0.2085060328245163, 0.1353929489850998, -0.2649182081222534, 0.10901321470737457, -0.021996289491653442, 0.15653330087661743, 0.2781084477901459, 0.021803172305226326, 0.27406957745552063, -0.24619391560554504, -0.08067683130502701, -0.10604254901409149, -0.08823495358228683, 0.13408689200878143, -0.22274042665958405, 0.059132907539606094, 0.07656463235616684, -0.16619180142879486, 0.5294390320777893, -0.02933022379875183, -0.10595626384019852, -0.2020445466041565, 0.23767568171024323, -0.4289614260196686, -0.1612631231546402, 0.10507849603891373, 0.1949494481086731, 0.12262489646673203, -0.3510899841785431, 0.5494361519813538, 0.0929960310459137, -0.03074311837553978, 0.05949217081069946, 0.16084764897823334, -0.2850426733493805, 0.46693816781044006, -0.20441244542598724, 0.014505140483379364, 0.06675807386636734, -0.034968819469213486, -0.09817085415124893, -0.3632011413574219, -0.03582318127155304, 0.28807687759399414, 0.14274710416793823, -0.045033156871795654, 0.008256521075963974, 0.17375124990940094, -0.06492209434509277, -0.0009053550311364233, -0.1085994690656662, -0.3678087592124939, 0.13817305862903595, 0.15505141019821167, 0.281816303730011, 0.20770123600959778, -0.10849186778068542, -0.248535618185997, -0.22846585512161255, 0.4540197253227234, 0.08564242720603943, 0.24932026863098145, -0.13096769154071808, 0.365329384803772, 0.40376004576683044, 0.5555804967880249, -0.28828105330467224, 0.047474224120378494, 0.030074216425418854, 0.09261980652809143, 0.48831906914711, -0.24070249497890472, 0.0004641004779841751, 0.2718995213508606, -0.12661142647266388, 0.06990157067775726, -0.3354584872722626, -0.1801498383283615, -0.27896958589553833, 0.10328452289104462, 0.1215909868478775, 0.004378876648843288, 0.019434597343206406, -0.025179466232657433, -0.09327104687690735, -0.07621178776025772, -0.032402656972408295, -0.42004597187042236, -0.059619929641485214, 0.1347336769104004, -0.07899252325296402, 0.04574660211801529, -0.004299607127904892, 0.21682484447956085, -0.005198767874389887, -0.04506275802850723, 0.2772456109523773, 0.08748000115156174, 0.14421898126602173, -0.1765514612197876, -0.14229975640773773, 0.006133316084742546, 0.30909624695777893, 0.10410713404417038, -0.028175780549645424, 0.17523452639579773, -0.20415693521499634, -0.07347244024276733, 0.26419714093208313, 0.19137366116046906, 0.02827771008014679, -0.06573760509490967, 0.12197926640510559, 0.15415143966674805, -0.13383087515830994, 0.037421371787786484, 0.24652636051177979, 0.05280547961592674, 0.12569332122802734, 0.13203732669353485, 0.15263573825359344, 0.46163201332092285, -0.22779308259487152, -0.009582077153027058, 0.13401435315608978, -0.3001977801322937, 0.16639317572116852, -0.1424223631620407, -0.003670151112601161, -0.01884313113987446, 0.14483997225761414, -0.030460041016340256, 0.17718985676765442, -0.3630976676940918, 0.29315492510795593, 0.31418606638908386, -0.012024300172924995, 0.04972655326128006, 0.21571582555770874, 0.01138595212250948, -0.13010439276695251, 0.30930855870246887, 0.09537271410226822, 0.3369998633861542, 0.003113327082246542, 0.8959376811981201, -0.3085145354270935, -0.49877551198005676, 0.08208099752664566, 0.4227997958660126, 0.03482503071427345, -0.19934344291687012, 0.08230806142091751, 0.12727220356464386, -0.36981821060180664, 0.07597491890192032, 0.07843468338251114, -0.18659283220767975, 0.6474289894104004, -0.11679505556821823, 0.1635960340499878, -0.170231431722641, -0.20496180653572083, 0.3429819643497467, 0.15966182947158813, -0.13731494545936584, 0.027702728286385536, 0.11107443273067474, -0.06848183274269104, -0.16262051463127136, 0.5162653923034668, 0.17436529695987701, 0.016103243455290794, -0.44874638319015503, 0.04122581332921982, 0.05003609135746956, 0.15217266976833344, -0.08710533380508423, 0.1192082092165947, 0.4416903555393219, -0.16612043976783752, 0.24096085131168365, 0.038405004888772964, -0.10613936185836792, -0.15496858954429626, 0.014972071163356304, -0.3957523703575134, -0.4279158413410187, 0.44072356820106506, 0.12134957313537598, -0.15100476145744324, 0.01822040230035782, 0.2517423629760742, -0.6123272776603699, -0.2271745800971985, 0.3545636236667633, -0.17337313294410706, -0.07675163447856903, -0.22562633454799652, 0.08382952213287354, 0.03794892504811287, 0.6115514636039734, 0.21737945079803467, 0.3595838248729706, -0.3842936158180237, -0.17158165574073792, -0.5617926716804504, -0.18688637018203735, 0.1403198093175888, -0.10462966561317444, -0.2472451776266098, 0.16251933574676514, -0.07911849766969681, 0.2967410981655121, -0.22634322941303253, 0.12328333407640457, 0.22285376489162445, -0.024598531424999237, -0.41087043285369873, 0.12129580974578857, 0.12449554353952408, 0.021147744730114937, 0.31346964836120605, -0.08089793473482132, -0.2103290855884552, 0.07324425876140594, -0.07079591602087021, -0.03564102575182915, -0.08734412491321564, 0.2631871700286865, 0.14747180044651031, 0.11267469823360443, 0.06292130053043365, 0.5086535215377808, -0.026272673159837723, -0.16295601427555084, 0.020117122679948807, -0.04937873035669327, -0.3885861337184906, -0.12324826419353485, -0.08058890700340271, 0.21543601155281067, -0.3562215566635132, 0.2585201561450958, 0.004247428849339485, -0.15498517453670502, 0.1442205160856247, 0.03260669857263565, -0.25623148679733276, -0.05825859680771828, -0.012390286661684513, 0.21576166152954102, 0.21925528347492218, 0.7426391243934631, 0.02795422449707985, 0.0870673879981041, -0.039152178913354874, -0.009825315326452255, 0.08792926371097565, -0.04098978638648987, -0.22142843902111053, -0.16324034333229065, 0.010918846353888512, 0.015441708266735077, 0.0981706976890564, -0.664665162563324, -0.36807212233543396, 0.04237062856554985, -0.007606104481965303, 0.09818395972251892, 0.7144814133644104, 0.08214665949344635, -0.14683640003204346, -0.1454656571149826, -0.27579164505004883, 0.034837283194065094, -0.24518172442913055, 0.19327987730503082, -0.3910036087036133 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
I guess we're looking at the end user writing something like: ``` python ds = nlp.load_dataset('multitask-t5',datasets=["squad","cnn_dm",...], k=1000, t=2.0) ``` Using the t5 method of combining here (or this could be a function passed in as an arg) Passing kwargs to each 'sub-dataset' might become tricky.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
45
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. I guess we're looking at the end user writing something like: ``` python ds = nlp.load_dataset('multitask-t5',datasets=["squad","cnn_dm",...], k=1000, t=2.0) ``` Using the t5 method of combining here (or this could be a function passed in as an arg) Passing kwargs to each 'sub-dataset' might become tricky.
[ -0.029287584125995636, -0.39404410123825073, -0.061029355973005295, -0.025662757456302643, -0.2608685791492462, 0.08258077502250671, 0.18105261027812958, 0.046955641359090805, 0.28684332966804504, -0.15945376455783844, -0.23304402828216553, 0.30905014276504517, -0.20747140049934387, 0.38534143567085266, 0.3700453042984009, -0.44683191180229187, -0.007145015522837639, -0.2071918249130249, -0.26221829652786255, 0.3925703465938568, 0.03699687123298645, 0.11199021339416504, -0.14147160947322845, 0.10980456322431564, -0.30104613304138184, -0.13675586879253387, -0.33219122886657715, 0.10305850207805634, 0.11648905277252197, 0.018128585070371628, 0.2122899740934372, 0.36170345544815063, -0.04755067825317383, 0.14389477670192719, -0.000114972070150543, -0.04456222802400589, 0.14054280519485474, -0.07454642653465271, 0.12798701226711273, -0.03853226453065872, 0.1727726012468338, -0.38791537284851074, 0.06748553365468979, -0.1346672773361206, 0.21632020175457, 0.15920692682266235, 0.1809331178665161, 0.06926102191209793, 0.4056929349899292, -0.32851284742355347, 0.08759284019470215, 0.2849256992340088, -0.19006940722465515, 0.34371352195739746, -0.10634847730398178, -0.10406813770532608, 0.03765282779932022, 0.25891387462615967, 0.5864786505699158, -0.5919809937477112, 0.05316491797566414, 0.08587028086185455, 0.012978089973330498, -0.03542965278029442, 0.1583544760942459, -0.028579682111740112, -0.3110811412334442, -0.22444434463977814, -0.41868332028388977, 0.18872550129890442, -0.10443755239248276, -0.15518999099731445, -0.22152848541736603, -0.5562816858291626, -0.017073102295398712, -0.0037982682697474957, -0.3029250204563141, -0.002131849527359009, -0.1731027513742447, -0.03949795290827751, -0.14767633378505707, -0.10621437430381775, -0.051074832677841187, 0.15497130155563354, 0.3348461389541626, 0.570990264415741, 0.1295994520187378, 0.17109720408916473, 0.3198663592338562, 0.05387605354189873, -0.13317269086837769, -0.2922164499759674, 0.30810171365737915, 0.16428826749324799, -0.36470818519592285, -0.3597906231880188, 0.19177627563476562, -0.25898098945617676, 0.23242586851119995, 0.09343254566192627, 0.12703311443328857, 0.08136384934186935, -0.1447870284318924, 0.26923468708992004, 0.3316325545310974, 0.03686713054776192, -0.10527118295431137, -0.12733295559883118, 0.02876242995262146, -0.1599452644586563, 0.23521646857261658, 0.09670281410217285, 0.1859353482723236, -0.022178875282406807, -0.39544323086738586, -0.16337113082408905, -0.13463498651981354, 0.14888383448123932, -0.19198086857795715, -0.5355587601661682, -0.1626536250114441, -0.13995623588562012, 0.04749845713376999, -0.13046790659427643, -0.3142296373844147, 0.07659965753555298, -0.030082346871495247, 0.18955890834331512, -0.14268624782562256, 0.12083211541175842, -0.05264069885015488, 0.0712476372718811, -0.448092520236969, -0.046076834201812744, 0.2776447534561157, 0.1095258966088295, 0.08483029901981354, 0.13384349644184113, 0.01484078448265791, 0.12548094987869263, 0.40193817019462585, 0.03929232433438301, -0.019936716184020042, -0.03314149007201195, 0.1681915819644928, -0.33171558380126953, -0.0062630330212414265, 0.1757419854402542, -0.35305988788604736, -0.14903035759925842, -0.1596110612154007, -0.18683311343193054, 0.12111642956733704, 0.024497872218489647, -0.22084037959575653, -0.3201468884944916, -0.4782353937625885, 0.676811933517456, -0.027634670957922935, -0.06473133713006973, -0.16472378373146057, 0.2757854461669922, -0.26916947960853577, -0.12027699500322342, 0.1744776964187622, 0.04582813382148743, 0.0011192123638466, -0.35605573654174805, -0.06673283874988556, -0.07920416444540024, -0.07760521024465561, 0.25847098231315613, -0.17129391431808472, 0.2623879611492157, 0.09088386595249176, 0.18576402962207794, 0.340880423784256, -0.47314998507499695, -0.022683197632431984, 0.29702964425086975, -0.05563291907310486, 0.41089123487472534, 0.07046778500080109, 0.22796069085597992, -0.08120576292276382, -0.20891065895557404, 0.27190226316452026, 1.0307589769363403, -0.3764420449733734, -0.022222435101866722, -0.09570217877626419, -0.39040130376815796, 0.5535189509391785, 0.349433034658432, 0.0771399661898613, -0.30797073245048523, -0.10218024998903275, -0.009633762761950493, 0.06673607230186462, 0.08410991728305817, 0.07473334670066833, -0.08781152963638306, -0.3393520712852478, 0.11391231417655945, -0.2102670520544052, -0.15189877152442932, -0.35301411151885986, 0.03392346575856209, 0.024168020114302635, 0.3140192925930023, 0.2573537230491638, -0.13017840683460236, 0.20869912207126617, -0.28740137815475464, -0.14865320920944214, -0.2948302924633026, 0.1073915958404541, -0.07125681638717651, 0.00028750274213962257, -0.23797935247421265, -0.03177603334188461, 0.30777207016944885, 0.13992391526699066, -0.13622407615184784, -0.4772074222564697, 0.2933182716369629, -0.06322204321622849, 0.0578543022274971, -0.05632811412215233, 0.6482886075973511, -0.16215428709983826, -0.06727863103151321, 0.1922089010477066, 0.32548752427101135, -0.3306887447834015, 0.2363162338733673, 0.3137442469596863, 0.3130963444709778, 0.16561391949653625, 0.006151152774691582, 0.11799057573080063, 0.011639749631285667, -0.001240299316123128, -0.21448661386966705, -0.09475599974393845, 0.5063753128051758, -0.1947728991508484, 0.4248306453227997, -0.023951204493641853, -0.0024468558840453625, -0.20531010627746582, -0.01521735917776823, -0.17521750926971436, 0.41539108753204346, 0.2906001806259155, -0.23537106812000275, 0.29709139466285706, -0.1605680286884308, -0.3739417791366577, 0.17212192714214325, 0.17921577394008636, 0.018867095932364464, 0.03440403193235397, -0.044534120708703995, 0.007732789497822523, -0.27543920278549194, -0.10600459575653076, 0.31868308782577515, 0.657733142375946, 0.2544582188129425, 0.13760951161384583, 0.06570440530776978, -0.20216353237628937, -0.05955953150987625, 0.0485227033495903, 0.21359162032604218, -0.04889363422989845, 0.1590908020734787, 0.03624693676829338, 0.08083868771791458, 0.06211617588996887, -0.1677285134792328, 0.13920889794826508, -0.20449765026569366, -0.11621721088886261, 0.054169945418834686, -0.18401724100112915, -0.6338436603546143, -0.18930263817310333, -0.009176885709166527, -0.33165132999420166, -0.2957566976547241, 0.13720200955867767, 0.055824633687734604, -0.11464399844408035, 0.13877330720424652, 0.19704563915729523, 0.5815650224685669, -0.4308137893676758, -0.27685585618019104, -0.036638468503952026, -0.31371790170669556, -0.10173463076353073, 0.10424836724996567, 0.43557342886924744, 0.29597708582878113, 0.5036945343017578, 0.2990524470806122, -0.06889305263757706, 0.0986972376704216, -0.4021720886230469, 0.06541711837053299, -0.35339897871017456, 0.25852999091148376, 0.13582926988601685, -0.22173327207565308, 0.14583013951778412, -0.3897809386253357, -0.05303405970335007, -0.025670049712061882, -0.07492654025554657, -0.051337700337171555, 0.05406862497329712, 0.06931398808956146, -0.19855529069900513, -0.19036796689033508, -0.6149933338165283, -0.4419180452823639, 0.30627530813217163, -0.2279975414276123, 0.08272510766983032, 0.06216733902692795, -0.2821544110774994, 0.23458468914031982, -0.08945883065462112, 0.025363445281982422, -0.17027738690376282, -0.05233953893184662, 0.010178900323808193, -0.38480693101882935, -0.14478403329849243, -0.33935168385505676, -0.0398954339325428, 0.22139869630336761, -0.06605762243270874, -0.029407810419797897, -0.27509257197380066, 0.08280285447835922, -0.08778967708349228, 0.07219158113002777, 0.26224297285079956, 0.16565261781215668, -0.20603424310684204, 0.13177087903022766, -0.08111463487148285, -0.2719358503818512, 0.21034707129001617, 0.3302367031574249, 0.3737383782863617, 0.1844935566186905, 0.03631795942783356, 0.18707454204559326, 0.7639580965042114, 0.24814258515834808, 0.06307937949895859, 0.14840303361415863, 0.26118797063827515, -0.012645027600228786, 0.16147364675998688, -0.18826884031295776, 0.18401652574539185, -0.17331543564796448, 0.52779221534729, 0.08122685551643372, 0.1695803999900818, -0.14921729266643524, -0.3321087956428528, 0.01802757754921913, -0.31423673033714294, -0.1845276653766632, 0.1901613026857376, -0.27717265486717224, 0.017156021669507027, -0.09102077782154083, -0.4362218976020813, -0.28573912382125854, -0.03737962618470192, -0.01700068823993206, 0.33567774295806885, -0.006728252395987511, 0.17077027261257172, -0.33540967106819153, -0.043419066816568375, -0.42174920439720154, 0.1442064642906189, 0.05830434709787369, 0.23728303611278534, -0.08073758333921432, -0.0721394494175911, -0.10103435069322586, 0.11182735115289688, 0.5621082782745361, -0.547377347946167, -0.1090129017829895, 0.0950562059879303, 0.07976479083299637, -0.12696290016174316, -0.11654829233884811, -0.29187890887260437, 0.020434679463505745, 0.42762985825538635, 0.37688013911247253, -0.3781234323978424, -0.2943044602870941, 0.2408280223608017, -0.25395745038986206, -0.07215525209903717, -0.19330352544784546, -0.24059031903743744, -0.21128347516059875, -0.12437959760427475, -0.005925618577748537, 0.25360995531082153, 0.23441174626350403, -0.12050193548202515, 0.18363426625728607, -0.08123110234737396, 0.26042401790618896, 0.24369831383228302, 0.01038103736937046, -0.11040768027305603, 0.24419310688972473, 0.17174606025218964, 0.03899581730365753, -0.12102355062961578, -0.35914623737335205, 0.09985493123531342, -0.02167419157922268, 0.0665019229054451, 0.25530385971069336, -0.0018160351319238544, 0.6958092451095581, 0.04607251659035683, 0.3278180658817291, 0.27770745754241943, -0.14805732667446136, -0.00294584222137928, -0.22390872240066528, 0.1858537346124649, 0.23404690623283386, 0.2490612417459488, -0.14151279628276825, -0.2325431853532791, 0.5836408734321594, 0.307031512260437, -0.0868648812174797, 0.438527375459671, -0.4941883087158203, -0.33374688029289246, 0.217090904712677, 0.02576558105647564, 0.7134877443313599, -0.022532299160957336, 0.3297561705112457, 0.11808187514543533, -0.41327327489852905, 0.5890749096870422, -0.3362227976322174, 0.12408443540334702, -0.05483739823102951, -0.17275090515613556, -0.011037340387701988, -0.07591137290000916, -0.1029583066701889, 0.20575636625289917, -0.3488745391368866, 0.3369854688644409, 0.40350931882858276, -0.2302367240190506, -0.04739286005496979, 0.5434772372245789, -0.014471161179244518, -0.34057775139808655, -0.0709589496254921, 0.0513051338493824, -0.2821183502674103, 0.06947915256023407, -0.21265488862991333, -0.25520190596580505, 0.2105037420988083, 0.15928424894809723, -0.2744204103946686, 0.06679617613554001, 0.00819696020334959, 0.2008344978094101, 0.22477585077285767, 0.035592272877693176, 0.2640017569065094, -0.1730501651763916, -0.10373202711343765, -0.17674149572849274, -0.06488734483718872, 0.10578814148902893, -0.23131312429904938, 0.03633525222539902, 0.09533362835645676, -0.15070948004722595, 0.49385008215904236, -0.07441766560077667, -0.10552598536014557, -0.22589963674545288, 0.26180681586265564, -0.4456820487976074, -0.2548675537109375, 0.13174158334732056, 0.16105304658412933, 0.18171532452106476, -0.32188642024993896, 0.5945079922676086, 0.08666719496250153, 0.05346466228365898, 0.018005885183811188, 0.1704840511083603, -0.23972651362419128, 0.4645395278930664, -0.32235202193260193, 0.00029913062462583184, 0.09918437898159027, 0.0001489784917794168, -0.15183033049106598, -0.41894495487213135, -0.053586266934871674, 0.27482616901397705, 0.12449419498443604, 0.055340979248285294, -0.016994597390294075, 0.11429919302463531, -0.03799274191260338, -0.05551130324602127, -0.11189194023609161, -0.2893208861351013, 0.1342903971672058, 0.24401602149009705, 0.17651277780532837, 0.19784779846668243, -0.06045433133840561, -0.2944519817829132, -0.13006658852100372, 0.49924972653388977, -0.016899898648262024, 0.23208317160606384, -0.15137682855129242, 0.370862752199173, 0.41062846779823303, 0.4903720021247864, -0.24310016632080078, 0.05704503133893013, 0.0985335037112236, 0.10004336386919022, 0.4676123559474945, -0.22680513560771942, 0.04738747701048851, 0.2821197211742401, -0.1461154967546463, 0.031245607882738113, -0.35195863246917725, -0.13588076829910278, -0.249692901968956, 0.12456842511892319, 0.12881578505039215, -0.028475750237703323, 0.028417155146598816, -0.1112867072224617, -0.10461067408323288, -0.05775509774684906, 0.04444015026092529, -0.36330121755599976, -0.09757297486066818, 0.16149677336215973, -0.09412873536348343, 0.08671180158853531, 0.05705199018120766, 0.20702902972698212, 0.06947289407253265, -0.07310795783996582, 0.338720440864563, 0.12007114291191101, 0.08686436712741852, -0.168158620595932, -0.14083094894886017, 0.016439862549304962, 0.29555824398994446, 0.18031619489192963, -0.06893280893564224, 0.14224347472190857, -0.2093847244977951, -0.19972921907901764, 0.2077009677886963, 0.20243225991725922, 0.03528764843940735, -0.13023793697357178, 0.10691750049591064, 0.09648512303829193, -0.06001567840576172, -0.012907186523079872, 0.23015165328979492, 0.08779217302799225, 0.07244772464036942, 0.11680996417999268, 0.11368749290704727, 0.4800713062286377, -0.2822706699371338, 0.03274897485971451, 0.09447178989648819, -0.35945484042167664, 0.1163831278681755, -0.08243781328201294, -0.08555765450000763, -0.01449133176356554, 0.12477091699838638, 0.00008960163540905342, 0.11933571100234985, -0.3342016339302063, 0.32404932379722595, 0.2672711908817291, 0.015399129129946232, 0.015001093037426472, 0.1387520134449005, 0.04229051247239113, -0.17318065464496613, 0.26796072721481323, 0.035019952803850174, 0.3751722574234009, 0.06041363254189491, 0.8366439342498779, -0.21346993744373322, -0.3915116488933563, 0.15799547731876373, 0.4222695231437683, 0.08762921392917633, -0.21886064112186432, 0.13186529278755188, 0.09811283648014069, -0.3754846453666687, 0.06979594379663467, 0.05699168145656586, -0.1820816695690155, 0.6210576891899109, -0.0859886035323143, 0.16826103627681732, -0.11792000383138657, -0.1851585954427719, 0.3580673933029175, 0.21875599026679993, -0.07344881445169449, -0.03960040211677551, 0.12822666764259338, -0.0619417242705822, -0.19880583882331848, 0.4743301272392273, 0.21330253779888153, -0.0024092611856758595, -0.48168686032295227, 0.033378731459379196, 0.1014532521367073, 0.17900317907333374, 0.0025962193030864, 0.13696503639221191, 0.4460620880126953, -0.15025974810123444, 0.20349392294883728, -0.007044513709843159, -0.03401820361614227, -0.13445653021335602, 0.06418414413928986, -0.4029557406902313, -0.4237152934074402, 0.4620606005191803, 0.11025909334421158, -0.1495259553194046, 0.11755118519067764, 0.37620994448661804, -0.629585862159729, -0.26949140429496765, 0.3343757390975952, -0.058231648057699203, -0.08502030372619629, -0.19609685242176056, 0.06549724191427231, 0.07512463629245758, 0.6126101613044739, 0.20807692408561707, 0.40820133686065674, -0.3459770977497101, -0.09671233594417572, -0.5682125091552734, -0.2036946564912796, 0.17174023389816284, -0.15298110246658325, -0.22418349981307983, 0.11838775873184204, -0.06933377683162689, 0.27628403902053833, -0.1890556365251541, 0.10309835523366928, 0.24528910219669342, 0.04982002079486847, -0.4366210103034973, 0.08494287729263306, 0.19068244099617004, 0.04040239751338959, 0.2738914489746094, -0.06394296884536743, -0.22662203013896942, 0.0473252534866333, -0.06836315244436264, -0.0969073474407196, -0.11419855803251266, 0.2681550681591034, 0.1839284598827362, 0.09595821797847748, 0.05707338824868202, 0.5114495158195496, 0.010969553142786026, -0.1574939340353012, 0.0035757713485509157, 0.04371923580765724, -0.3633822202682495, -0.13569596409797668, -0.09348233789205551, 0.1329883188009262, -0.3916027545928955, 0.3226500153541565, 0.005707671400159597, -0.07805079966783524, 0.09187620133161545, 0.0003347972815390676, -0.12278418242931366, -0.11219821870326996, 0.0437779575586319, 0.22140516340732574, 0.1861126720905304, 0.7520395517349243, -0.0126851387321949, 0.13772766292095184, -0.09494876861572266, -0.00878309179097414, 0.027268320322036743, -0.06434780359268188, -0.2542925477027893, -0.09406072646379471, 0.06744660437107086, 0.06809114664793015, 0.14320744574069977, -0.6984320878982544, -0.35412970185279846, 0.05046379566192627, 0.01615244336426258, 0.015999862924218178, 0.7800260186195374, 0.09473685920238495, -0.13276302814483643, -0.08945035934448242, -0.36224648356437683, 0.0005289010587148368, -0.22626951336860657, 0.16598664224147797, -0.3741372227668762 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
From thinking upon @thomwolf 's suggestion, I've started experimenting: ```python class MultitaskDataset(DatasetBuilder): def __init__(self, *args, **kwargs): super(MultitaskDataset, self).__init__(*args, **kwargs) self._datasets = kwargs.get("datasets") def _info(self): return nlp.DatasetInfo( description=_DESCRIPTION, features=nlp.Features({ "source": nlp.Value("string"), "target": nlp.Sequence(nlp.Value("string")) }) ) def _get_common_splits(self): '''Finds the common splits present in all self._datasets''' min_set = None for dataset in self._datasets: if min_set != None: min_set.intersection(set(dataset.keys())) else: min_set = set(dataset.keys()) return min_set .... # Maybe this?: squad = nlp.load_dataset("squad") cnn_dm = nlp.load_dataset("cnn_dailymail","3.0.0") multitask_dataset = nlp.load_dataset( 'multitask_dataset', datasets=[squad,cnn_dailymail], k=1000, t=2.0 ) ``` Does anyone know what methods of `MultitaskDataset` I would need to implement? Maybe `as_dataset` and `download_and_prepare`? Most of these should be just calling the methods of the sub-datasets. I'm assuming DatasetBuilder is better than the more specific `GeneratorBasedBuilder`, `BeamBasedBuilder`, etc.... One of the other problems is that the dataset size is unknown till you construct it (as you can pick the sub-datasets). Am hoping not to need to make changes to `nlp.load_dataset` just for this class. I'd appreciate it if anyone more familiar with nlp's internal workings could tell me if I'm on the right track!
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
177
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. From thinking upon @thomwolf 's suggestion, I've started experimenting: ```python class MultitaskDataset(DatasetBuilder): def __init__(self, *args, **kwargs): super(MultitaskDataset, self).__init__(*args, **kwargs) self._datasets = kwargs.get("datasets") def _info(self): return nlp.DatasetInfo( description=_DESCRIPTION, features=nlp.Features({ "source": nlp.Value("string"), "target": nlp.Sequence(nlp.Value("string")) }) ) def _get_common_splits(self): '''Finds the common splits present in all self._datasets''' min_set = None for dataset in self._datasets: if min_set != None: min_set.intersection(set(dataset.keys())) else: min_set = set(dataset.keys()) return min_set .... # Maybe this?: squad = nlp.load_dataset("squad") cnn_dm = nlp.load_dataset("cnn_dailymail","3.0.0") multitask_dataset = nlp.load_dataset( 'multitask_dataset', datasets=[squad,cnn_dailymail], k=1000, t=2.0 ) ``` Does anyone know what methods of `MultitaskDataset` I would need to implement? Maybe `as_dataset` and `download_and_prepare`? Most of these should be just calling the methods of the sub-datasets. I'm assuming DatasetBuilder is better than the more specific `GeneratorBasedBuilder`, `BeamBasedBuilder`, etc.... One of the other problems is that the dataset size is unknown till you construct it (as you can pick the sub-datasets). Am hoping not to need to make changes to `nlp.load_dataset` just for this class. I'd appreciate it if anyone more familiar with nlp's internal workings could tell me if I'm on the right track!
[ -0.04102901741862297, -0.3800129294395447, -0.05613596737384796, -0.05533390864729881, -0.22696149349212646, 0.03814494237303734, 0.19116327166557312, 0.028541920706629753, 0.3140309751033783, -0.16932398080825806, -0.21132950484752655, 0.3226816654205322, -0.21177850663661957, 0.3609279692173004, 0.41081151366233826, -0.40815508365631104, 0.049611303955316544, -0.23854896426200867, -0.30185458064079285, 0.3816815912723541, 0.0294326338917017, 0.11054202169179916, -0.18156084418296814, 0.11165957152843475, -0.27534806728363037, -0.14701879024505615, -0.3376018702983856, 0.1057681143283844, 0.12588585913181305, 0.00012284105469007045, 0.25411471724510193, 0.40501534938812256, -0.06453417986631393, 0.16401495039463043, -0.00011387033009668812, -0.028973601758480072, 0.23698237538337708, -0.0645473301410675, 0.1149069294333458, -0.07505529373884201, 0.20829924941062927, -0.4271640479564667, 0.05915255844593048, -0.16756828129291534, 0.17296180129051208, 0.15671925246715546, 0.1583685427904129, 0.040026649832725525, 0.39363378286361694, -0.2916634678840637, 0.10325055569410324, 0.28750234842300415, -0.15446880459785461, 0.276138573884964, -0.1251123547554016, -0.047688078135252, 0.03673836588859558, 0.2582828998565674, 0.5602094531059265, -0.6138654947280884, 0.003775055520236492, 0.11586428433656693, 0.001109798438847065, 0.004889032803475857, 0.15319642424583435, -0.020795315504074097, -0.3521723747253418, -0.21768826246261597, -0.4039084017276764, 0.18628573417663574, -0.1451549530029297, -0.20172235369682312, -0.23753727972507477, -0.5448269248008728, 0.004671125672757626, 0.006173090543597937, -0.26987800002098083, -0.01835896633565426, -0.20179829001426697, -0.06324586272239685, -0.12326794117689133, -0.08969506621360779, -0.03366624936461449, 0.15525388717651367, 0.3182492256164551, 0.5842536687850952, 0.09058979898691177, 0.16732574999332428, 0.36797139048576355, 0.04139329865574837, -0.15085774660110474, -0.28049373626708984, 0.32753702998161316, 0.1458224356174469, -0.3698492646217346, -0.30306050181388855, 0.20416416227817535, -0.23676463961601257, 0.2298242300748825, 0.10457225888967514, 0.09047602117061615, 0.11722009629011154, -0.14798396825790405, 0.2619042694568634, 0.3045753240585327, 0.04612181335687637, -0.0961487740278244, -0.09544787555932999, 0.034676067531108856, -0.08594369888305664, 0.19273564219474792, 0.11217852681875229, 0.18549828231334686, -0.012384217232465744, -0.3950420618057251, -0.1339057981967926, -0.12934337556362152, 0.16474437713623047, -0.2152850478887558, -0.4849032163619995, -0.10857424885034561, -0.11952312290668488, 0.11771396547555923, -0.10458685457706451, -0.30913496017456055, 0.09875573962926865, -0.038953203707933426, 0.26127830147743225, -0.13749782741069794, 0.1306162178516388, -0.07104407250881195, 0.08409839123487473, -0.4412373900413513, -0.07710437476634979, 0.30592894554138184, 0.15183499455451965, 0.08070240914821625, 0.16790153086185455, -0.001432355958968401, 0.12360367178916931, 0.38938167691230774, 0.015578439459204674, 0.0013976499903947115, -0.000942085636779666, 0.13964281976222992, -0.3161430060863495, 0.020422881469130516, 0.15890757739543915, -0.3474515974521637, -0.19030031561851501, -0.11934749037027359, -0.1685405820608139, 0.14057426154613495, 0.03215498477220535, -0.20059575140476227, -0.2385004311800003, -0.4981873333454132, 0.7015710473060608, -0.05721518397331238, -0.0934595912694931, -0.17591387033462524, 0.3037513494491577, -0.24646325409412384, -0.08007053285837173, 0.2119656205177307, 0.05763937905430794, 0.011350232176482677, -0.3442132771015167, -0.10541829466819763, -0.09868159890174866, -0.021547788754105568, 0.2500793933868408, -0.1620028018951416, 0.2792709171772003, 0.05273359641432762, 0.20901469886302948, 0.2798384428024292, -0.4821226894855499, -0.06264195591211319, 0.25964367389678955, -0.044114213436841965, 0.4229988157749176, 0.05807715654373169, 0.19613544642925262, -0.026013467460870743, -0.17014871537685394, 0.2901472747325897, 0.9980711340904236, -0.34383928775787354, -0.019432082772254944, -0.10914522409439087, -0.35987621545791626, 0.5201608538627625, 0.3251277208328247, 0.10994330793619156, -0.2907426655292511, -0.13588403165340424, -0.015212993137538433, 0.12174703925848007, 0.07488232105970383, 0.0830933153629303, -0.07177481055259705, -0.34809327125549316, 0.06251148879528046, -0.2146705836057663, -0.1692681908607483, -0.34028953313827515, 0.023376796394586563, 0.053741421550512314, 0.3371376097202301, 0.21158994734287262, -0.14518533647060394, 0.1958480179309845, -0.2553425133228302, -0.14328379929065704, -0.30226650834083557, 0.1462630331516266, -0.06994912773370743, 0.013663031160831451, -0.24087974429130554, -0.0318981371819973, 0.3690547049045563, 0.11622243374586105, -0.12677814066410065, -0.49740028381347656, 0.3005133271217346, -0.07422429323196411, 0.06661371141672134, -0.03259013965725899, 0.6585516929626465, -0.15063238143920898, -0.08514700084924698, 0.18119415640830994, 0.2812552750110626, -0.29423239827156067, 0.20129653811454773, 0.3036303222179413, 0.19760724902153015, 0.15173520147800446, 0.004687821492552757, 0.07329757511615753, 0.05224281921982765, 0.024796754121780396, -0.23360484838485718, -0.0985802635550499, 0.51006680727005, -0.16134679317474365, 0.4522989094257355, -0.028001848608255386, 0.02515815570950508, -0.1460135579109192, -0.02968272753059864, -0.2038654386997223, 0.3801974952220917, 0.2894110083580017, -0.18660062551498413, 0.32788321375846863, -0.16236719489097595, -0.3643944561481476, 0.1550499051809311, 0.13120156526565552, 0.012626302428543568, 0.05463068187236786, -0.04349257051944733, -0.013247989118099213, -0.27911481261253357, -0.10775308310985565, 0.3511255085468292, 0.6806014776229858, 0.23463626205921173, 0.19186392426490784, 0.06999512761831284, -0.2855550944805145, -0.07115644961595535, 0.0013606133870780468, 0.22168433666229248, -0.040608156472444534, 0.16736474633216858, 0.06225522235035896, 0.04244364798069, 0.02196422964334488, -0.15282703936100006, 0.15568017959594727, -0.18520879745483398, -0.0936625599861145, 0.10740393400192261, -0.22168287634849548, -0.5938891172409058, -0.17749539017677307, 0.0022366594057530165, -0.2878173887729645, -0.30850136280059814, 0.10392793267965317, 0.0533197857439518, -0.14587979018688202, 0.16206030547618866, 0.2049495428800583, 0.5456452369689941, -0.42458295822143555, -0.27564841508865356, -0.041723042726516724, -0.31425365805625916, -0.11091036349534988, 0.10897129774093628, 0.45676758885383606, 0.3351975381374359, 0.4890904128551483, 0.3100687563419342, -0.0816783607006073, 0.04969417303800583, -0.4016467332839966, 0.04757334664463997, -0.3254901170730591, 0.26705101132392883, 0.11735404282808304, -0.2622731924057007, 0.16214051842689514, -0.372204452753067, -0.011827545240521431, -0.08734909445047379, -0.08597221970558167, -0.021902495995163918, -0.019531656056642532, 0.059149689972400665, -0.21000787615776062, -0.15615111589431763, -0.6618933081626892, -0.4537320137023926, 0.31900379061698914, -0.2383415251970291, 0.08222025632858276, 0.059315700083971024, -0.295493483543396, 0.2223706990480423, -0.06498759984970093, 0.016360916197299957, -0.17736470699310303, -0.048844918608665466, 0.029317015781998634, -0.3534870743751526, -0.1780763864517212, -0.3723650276660919, -0.13820268213748932, 0.28022709488868713, -0.06061975285410881, -0.007123859599232674, -0.2503639757633209, 0.09986142814159393, -0.07900948077440262, 0.10570447146892548, 0.24656283855438232, 0.17851509153842926, -0.23915883898735046, 0.1344873458147049, -0.060159265995025635, -0.264729380607605, 0.24683882296085358, 0.2817917466163635, 0.3843083679676056, 0.18339203298091888, 0.05993526801466942, 0.17080555856227875, 0.6879072785377502, 0.22980043292045593, 0.06567388772964478, 0.14727602899074554, 0.254945307970047, 0.019311072304844856, 0.1698751151561737, -0.2620750665664673, 0.20821620523929596, -0.21381095051765442, 0.4608762860298157, 0.045081764459609985, 0.16560126841068268, -0.2366371750831604, -0.29566049575805664, 0.02418248914182186, -0.31385400891304016, -0.21501585841178894, 0.1974213868379593, -0.2264556586742401, 0.002459387294948101, -0.07362595200538635, -0.37921613454818726, -0.28471294045448303, -0.0331578329205513, 0.00451201805844903, 0.3330291211605072, -0.0960414856672287, 0.11804516613483429, -0.3570634722709656, -0.031005756929516792, -0.46974050998687744, 0.14145852625370026, 0.008327112533152103, 0.2584218382835388, -0.050975501537323, -0.07874026894569397, -0.08971879631280899, 0.07503262907266617, 0.6008833050727844, -0.4771219789981842, -0.12468094378709793, 0.15265151858329773, 0.1111767590045929, -0.1372944414615631, -0.12630045413970947, -0.2912557125091553, 0.046896953135728836, 0.46849584579467773, 0.37232547998428345, -0.387547105550766, -0.2634353041648865, 0.2291095107793808, -0.2160625159740448, -0.044988807290792465, -0.19089001417160034, -0.23585587739944458, -0.2560034692287445, -0.09917119145393372, 0.04424171522259712, 0.21777823567390442, 0.23202070593833923, -0.12347611039876938, 0.2110004425048828, -0.10897155106067657, 0.2612602114677429, 0.22022061049938202, 0.004205596167594194, -0.027149712666869164, 0.20108677446842194, 0.1709839105606079, 0.038425084203481674, -0.13986222445964813, -0.37141937017440796, 0.1328182816505432, 0.00464691873639822, 0.078825443983078, 0.24406211078166962, -0.007657051086425781, 0.6334800720214844, 0.07391361147165298, 0.2964062988758087, 0.2896260619163513, -0.14868782460689545, 0.017744984477758408, -0.2544415593147278, 0.19163605570793152, 0.24283245205879211, 0.2626420557498932, -0.1403989940881729, -0.2564392685890198, 0.590032696723938, 0.3069307208061218, -0.09948931634426117, 0.4368518888950348, -0.49968454241752625, -0.3505486845970154, 0.2613227069377899, 0.02210071310400963, 0.6753666996955872, 0.005988419055938721, 0.31732314825057983, 0.16142134368419647, -0.37519145011901855, 0.5796289443969727, -0.3777526915073395, 0.1610085666179657, -0.07037798315286636, -0.14920993149280548, 0.004808472935110331, -0.08049342781305313, -0.0371193066239357, 0.22904394567012787, -0.34711557626724243, 0.3315231502056122, 0.4464799761772156, -0.20433221757411957, -0.06858977675437927, 0.5156776309013367, -0.023617764934897423, -0.37427836656570435, -0.12388738989830017, 0.07793392986059189, -0.2607242166996002, 0.09877067059278488, -0.15557418763637543, -0.2480802983045578, 0.19329650700092316, 0.14641085267066956, -0.2879005968570709, 0.10499077290296555, 0.030853280797600746, 0.23830029368400574, 0.2569221258163452, 0.009896501898765564, 0.216981902718544, -0.19437547028064728, -0.054442163556814194, -0.15986984968185425, -0.06007993221282959, 0.12273019552230835, -0.20815414190292358, 0.08337409794330597, 0.08152738213539124, -0.22250565886497498, 0.5000629425048828, -0.053083084523677826, -0.10866129398345947, -0.24004176259040833, 0.2608029246330261, -0.43278154730796814, -0.18399779498577118, 0.1675369143486023, 0.16328534483909607, 0.2320406138896942, -0.321902334690094, 0.6011399030685425, 0.1171332523226738, -0.009043414145708084, 0.035241540521383286, 0.18379519879817963, -0.27487900853157043, 0.5102382302284241, -0.3566153645515442, 0.011908870190382004, 0.08584152162075043, 0.01222094614058733, -0.14734922349452972, -0.37615835666656494, -0.032249532639980316, 0.29934772849082947, 0.0852387472987175, 0.03198632597923279, -0.01090412586927414, 0.10516118258237839, -0.03416301682591438, -0.04484115540981293, -0.09733852744102478, -0.326081246137619, 0.12282904982566833, 0.18378686904907227, 0.17996643483638763, 0.17217276990413666, -0.08546771109104156, -0.2773149907588959, -0.1805693358182907, 0.48039117455482483, 0.023019086569547653, 0.20212779939174652, -0.15253081917762756, 0.35346758365631104, 0.41968733072280884, 0.5000759959220886, -0.2531454861164093, 0.055512435734272, 0.07240857928991318, 0.07157958298921585, 0.4242963194847107, -0.2460532784461975, 0.06022517755627632, 0.24505974352359772, -0.12159831076860428, 0.04352067783474922, -0.3518274128437042, -0.14629240334033966, -0.2637699544429779, 0.10850557684898376, 0.10938475281000137, 0.00038984668208286166, 0.0513630211353302, -0.09670044481754303, -0.09724676609039307, -0.06615369766950607, -0.01298169232904911, -0.3419177830219269, -0.07169052958488464, 0.1208123117685318, -0.06952336430549622, 0.043189674615859985, 0.07352322340011597, 0.1863006204366684, 0.039544690400362015, -0.07555142045021057, 0.3119194209575653, 0.09127143025398254, 0.1235489547252655, -0.16061826050281525, -0.13676145672798157, 0.020298222079873085, 0.2227635234594345, 0.15596750378608704, -0.0428767055273056, 0.11895941942930222, -0.20203469693660736, -0.15555652976036072, 0.2734249532222748, 0.22079318761825562, 0.027456453070044518, -0.1333657205104828, 0.13120338320732117, 0.11210913211107254, -0.11988253146409988, -0.025684459134936333, 0.23964129388332367, 0.08068180084228516, 0.10649394989013672, 0.12201182544231415, 0.11496614664793015, 0.4606333076953888, -0.28958824276924133, 0.049498964101076126, 0.11247263848781586, -0.3702647387981415, 0.11843647807836533, -0.13738718628883362, -0.010500718839466572, 0.02615814469754696, 0.14512287080287933, -0.004540335386991501, 0.11539991199970245, -0.24399881064891815, 0.3238443434238434, 0.30008357763290405, -0.010188991203904152, 0.03903507813811302, 0.1394425332546234, 0.021483195945620537, -0.17022641003131866, 0.2212379425764084, -0.012151825241744518, 0.3521035611629486, 0.031932782381772995, 0.8429726958274841, -0.30105841159820557, -0.39671996235847473, 0.1265188753604889, 0.40005335211753845, 0.060758356004953384, -0.23886413872241974, 0.1321716159582138, 0.10595063865184784, -0.3579234182834625, 0.06069978326559067, 0.06416893750429153, -0.15428325533866882, 0.6550420522689819, -0.1400015652179718, 0.16470621526241302, -0.17565381526947021, -0.21197979152202606, 0.33194786310195923, 0.20278297364711761, -0.07796867936849594, -0.01160579826682806, 0.12744653224945068, -0.060982391238212585, -0.19573865830898285, 0.49206677079200745, 0.25489893555641174, 0.05156726390123367, -0.4656880497932434, 0.007283030543476343, 0.09706360101699829, 0.1373627483844757, -0.09356136620044708, 0.1530831754207611, 0.45377764105796814, -0.15144921839237213, 0.20709183812141418, 0.004273405764251947, -0.058435309678316116, -0.09915976971387863, 0.08272409439086914, -0.39912086725234985, -0.4085474908351898, 0.5045669078826904, 0.060878004878759384, -0.11606679111719131, 0.08093558996915817, 0.3370767831802368, -0.6575292348861694, -0.26047569513320923, 0.34948840737342834, -0.08489368855953217, -0.03731139004230499, -0.22380822896957397, 0.07407920807600021, 0.07273732870817184, 0.6086324453353882, 0.16365914046764374, 0.37715280055999756, -0.3970233201980591, -0.10719913244247437, -0.6464769840240479, -0.21320606768131256, 0.09976034611463547, -0.16706301271915436, -0.26028192043304443, 0.13774213194847107, -0.038987454026937485, 0.27516070008277893, -0.2075423300266266, 0.05163273215293884, 0.2532150447368622, 0.02692098543047905, -0.4563051462173462, 0.07905163615942001, 0.1370769441127777, 0.02010919898748398, 0.2837491035461426, -0.07789786905050278, -0.22131957113742828, 0.06585617363452911, -0.056703925132751465, -0.050493646413087845, -0.08613960444927216, 0.277947336435318, 0.1756100058555603, 0.035993583500385284, 0.04806533083319664, 0.5123874545097351, -0.022415723651647568, -0.17422255873680115, -0.026029085740447044, 0.028573179617524147, -0.37258657813072205, -0.1118156909942627, -0.12267346680164337, 0.2113187313079834, -0.3768765926361084, 0.27004823088645935, 0.016957731917500496, -0.10174824297428131, 0.12268909811973572, -0.009531861171126366, -0.17500975728034973, -0.021082447841763496, 0.04411695525050163, 0.22659964859485626, 0.21104493737220764, 0.7624978423118591, -0.0001504886313341558, 0.17093849182128906, -0.07620440423488617, -0.06651189923286438, 0.07553702592849731, -0.11749576032161713, -0.2645830810070038, -0.15456421673297882, 0.06485782563686371, 0.047523293644189835, 0.11066404730081558, -0.7056501507759094, -0.3468599021434784, 0.016169417649507523, -0.01917714811861515, 0.05395888909697533, 0.7361039519309998, 0.07266077399253845, -0.1439816653728485, -0.13109447062015533, -0.3231861889362335, 0.02054022066295147, -0.23995502293109894, 0.1891360878944397, -0.3676553964614868 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
I think I would probably go for a `MultiDataset` wrapper around a list of `Dataset`. I'm not sure we need to give it `k` and `t` parameters at creation, it can maybe be something along the lines of: ```python squad = nlp.load_dataset("squad") cnn_dm = nlp.load_dataset("cnn_dailymail","3.0.0") multitask_dataset = nlp.MultiDataset(squad, cnn_dm) batch = multitask_dataset.sample(10, temperature=2.0, k=1000) ``` The first proof-of-concept for multi-task datasets could definitely require that the provided datasets have the same name/type for columns (if needed you easily rename/cast a column prior to instantiating the `MultiDataset`). It's good to think about it for some time though and don't overfit too much on the T5 examples (in particular for the ways/kwargs for sampling among datasets).
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
114
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. I think I would probably go for a `MultiDataset` wrapper around a list of `Dataset`. I'm not sure we need to give it `k` and `t` parameters at creation, it can maybe be something along the lines of: ```python squad = nlp.load_dataset("squad") cnn_dm = nlp.load_dataset("cnn_dailymail","3.0.0") multitask_dataset = nlp.MultiDataset(squad, cnn_dm) batch = multitask_dataset.sample(10, temperature=2.0, k=1000) ``` The first proof-of-concept for multi-task datasets could definitely require that the provided datasets have the same name/type for columns (if needed you easily rename/cast a column prior to instantiating the `MultiDataset`). It's good to think about it for some time though and don't overfit too much on the T5 examples (in particular for the ways/kwargs for sampling among datasets).
[ -0.03095950372517109, -0.41867339611053467, -0.039658013731241226, -0.046870894730091095, -0.22470249235630035, 0.09376367926597595, 0.1678377389907837, 0.03758658096194267, 0.2795368432998657, -0.16709890961647034, -0.22087904810905457, 0.334222674369812, -0.2148641049861908, 0.3291552662849426, 0.4196424186229706, -0.4405974745750427, 0.02405465580523014, -0.2191534787416458, -0.25798019766807556, 0.37987783551216125, 0.06556864082813263, 0.10199666768312454, -0.16362515091896057, 0.05979249253869057, -0.26799318194389343, -0.1647084653377533, -0.36144736409187317, 0.12241535633802414, 0.1369815319776535, -0.01051545049995184, 0.21925088763237, 0.3853777348995209, -0.02397017739713192, 0.11684691905975342, -0.00011339465709170327, -0.07660769671201706, 0.159489244222641, -0.12002424895763397, 0.09108947962522507, 0.006583403330296278, 0.15462002158164978, -0.42770883440971375, 0.07293473184108734, -0.15843740105628967, 0.14493362605571747, 0.156585231423378, 0.14506776630878448, -0.0069006639532744884, 0.4177410304546356, -0.2556488513946533, 0.09768408536911011, 0.3037799000740051, -0.16508737206459045, 0.27079740166664124, -0.1420087367296219, -0.06699206680059433, 0.02792990952730179, 0.22085760533809662, 0.6148955821990967, -0.5899105072021484, 0.0041679940186440945, 0.07856754213571548, -0.014349795877933502, 0.007405399810522795, 0.15548309683799744, -0.02988414652645588, -0.33571839332580566, -0.26306936144828796, -0.39329877495765686, 0.2161170393228531, -0.13005776703357697, -0.1769692301750183, -0.2505916953086853, -0.5384421348571777, 0.017067506909370422, 0.03541652485728264, -0.29035377502441406, 0.031573157757520676, -0.2046494334936142, -0.05459129065275192, -0.12335360795259476, -0.10292137414216995, -0.006522749084979296, 0.15645121037960052, 0.3066312074661255, 0.5719138979911804, 0.12777699530124664, 0.1712883859872818, 0.3525916337966919, 0.02389625459909439, -0.14905187487602234, -0.2875998318195343, 0.294041246175766, 0.16211551427841187, -0.4089527428150177, -0.35077908635139465, 0.13092657923698425, -0.24935375154018402, 0.2844473123550415, 0.08810195326805115, 0.11836913228034973, 0.05591239035129547, -0.13708645105361938, 0.28888919949531555, 0.30111274123191833, 0.03557567298412323, -0.07668834179639816, -0.12673504650592804, 0.021894125267863274, -0.13308803737163544, 0.18895968794822693, 0.12626688182353973, 0.18178312480449677, -0.00866074487566948, -0.37493881583213806, -0.13579273223876953, -0.1604587584733963, 0.1738719940185547, -0.1611003428697586, -0.5166133642196655, -0.11007855087518692, -0.09324707090854645, 0.072165846824646, -0.09246832877397537, -0.29434049129486084, 0.07610074430704117, -0.048854388296604156, 0.22475813329219818, -0.0882171243429184, 0.052158620208501816, -0.06363725662231445, 0.08743686228990555, -0.4271281957626343, -0.031100211665034294, 0.3111231327056885, 0.0953138917684555, 0.07913808524608612, 0.16832078993320465, -0.009065762162208557, 0.10786441713571548, 0.42135143280029297, 0.02073696069419384, -0.005291291046887636, -0.0257207490503788, 0.2052460014820099, -0.2978520393371582, 0.013125874102115631, 0.13944831490516663, -0.34433069825172424, -0.123369961977005, -0.14080309867858887, -0.16177302598953247, 0.1051672101020813, 0.026168093085289, -0.2270698994398117, -0.26348575949668884, -0.47871291637420654, 0.7096788287162781, -0.01631808653473854, -0.04073549062013626, -0.18003593385219574, 0.28762421011924744, -0.26460000872612, -0.11316076666116714, 0.22921176254749298, 0.10030703246593475, -0.04390301555395126, -0.2998161017894745, -0.07132155448198318, -0.08353041112422943, -0.08975902199745178, 0.27883946895599365, -0.19985716044902802, 0.22274592518806458, 0.06801257282495499, 0.1548844426870346, 0.30274003744125366, -0.4629935622215271, -0.03657873347401619, 0.2691686749458313, -0.035777006298303604, 0.38466617465019226, 0.06560098379850388, 0.22492572665214539, 0.019382698461413383, -0.2001308798789978, 0.2639881670475006, 1.0421267747879028, -0.4086366593837738, 0.000942250422667712, -0.09630386531352997, -0.356860488653183, 0.5225278735160828, 0.34482264518737793, 0.05289280042052269, -0.28935718536376953, -0.14739477634429932, -0.005419716704636812, 0.07239245623350143, 0.05815136060118675, 0.09516752511262894, -0.03698837384581566, -0.3416343331336975, 0.06954765319824219, -0.22003605961799622, -0.15002980828285217, -0.44301092624664307, 0.019452031701803207, 0.04457903280854225, 0.343418151140213, 0.23521652817726135, -0.13884542882442474, 0.2276785522699356, -0.2906137704849243, -0.13999618589878082, -0.3124925494194031, 0.13172060251235962, -0.05093533545732498, 0.03107062168419361, -0.22076000273227692, -0.035360340029001236, 0.36213135719299316, 0.1493372917175293, -0.12701545655727386, -0.48026639223098755, 0.3278312385082245, -0.07594664394855499, 0.03664102405309677, -0.027285726740956306, 0.6227855682373047, -0.15069277584552765, -0.07301894575357437, 0.19060467183589935, 0.2668914794921875, -0.28597140312194824, 0.21143180131912231, 0.36122769117355347, 0.270976722240448, 0.17123492062091827, -0.034092627465724945, 0.09529859572649002, 0.023809488862752914, -0.005174029618501663, -0.22024014592170715, -0.08570516854524612, 0.523391604423523, -0.17810837924480438, 0.41875627636909485, -0.04507823288440704, 0.036600515246391296, -0.13553018867969513, -0.018147645518183708, -0.19881540536880493, 0.4192892611026764, 0.2726355791091919, -0.17454442381858826, 0.35942038893699646, -0.139204204082489, -0.34282928705215454, 0.16008496284484863, 0.155892476439476, 0.038506265729665756, 0.090784452855587, -0.003108187345787883, -0.048500318080186844, -0.26149407029151917, -0.10891565680503845, 0.3499150276184082, 0.6467790007591248, 0.2586946189403534, 0.13757553696632385, 0.013220284134149551, -0.25058889389038086, -0.06948143243789673, 0.031848080456256866, 0.1693364381790161, -0.028486372902989388, 0.14386461675167084, 0.0615905299782753, 0.049080319702625275, 0.029332265257835388, -0.17154966294765472, 0.18374226987361908, -0.16221733391284943, -0.0755690336227417, 0.07238379865884781, -0.166484996676445, -0.5669136047363281, -0.22831463813781738, 0.0277330931276083, -0.3815074563026428, -0.288953572511673, 0.10477927327156067, 0.0514129176735878, -0.12023938447237015, 0.18662749230861664, 0.20177355408668518, 0.5722048878669739, -0.3877090513706207, -0.29302388429641724, -0.02914583310484886, -0.3389907777309418, -0.11456456035375595, 0.11243952065706253, 0.45572465658187866, 0.2603592872619629, 0.4842228293418884, 0.32592618465423584, -0.03397210314869881, 0.046375956386327744, -0.3984246850013733, 0.017643919214606285, -0.3012907803058624, 0.2676163613796234, 0.07377493381500244, -0.18586884438991547, 0.15339453518390656, -0.40689218044281006, -0.005613064859062433, -0.021066704764962196, -0.07087556272745132, -0.05289338156580925, 0.004416038282215595, 0.09542001038789749, -0.18347829580307007, -0.19025947153568268, -0.6412388682365417, -0.4301626980304718, 0.2673932909965515, -0.26016321778297424, 0.08659768104553223, 0.014041529037058353, -0.2555759847164154, 0.17967815697193146, -0.10786698013544083, -0.002608138369396329, -0.16347739100456238, -0.03895273059606552, 0.04661223292350769, -0.35634148120880127, -0.1756976991891861, -0.36977848410606384, -0.11427297443151474, 0.24605445563793182, -0.06947711110115051, -0.024848200380802155, -0.287604421377182, 0.09452791512012482, -0.11205773800611496, 0.053056538105010986, 0.20907078683376312, 0.19394835829734802, -0.22071024775505066, 0.13459722697734833, -0.07369828224182129, -0.2940366268157959, 0.23027558624744415, 0.272707462310791, 0.38758036494255066, 0.152955561876297, 0.07122144103050232, 0.17559200525283813, 0.7356219291687012, 0.2542083263397217, 0.018226446583867073, 0.15264922380447388, 0.2519253194332123, -0.008517189882695675, 0.1782766580581665, -0.24685481190681458, 0.17666277289390564, -0.15473157167434692, 0.4875797629356384, 0.09875710308551788, 0.15080231428146362, -0.24501726031303406, -0.30623963475227356, 0.058926284313201904, -0.26855406165122986, -0.2171594202518463, 0.2424306422472, -0.22875826060771942, 0.014071078039705753, -0.08965998888015747, -0.41623830795288086, -0.2906382977962494, -0.05388417840003967, -0.022310050204396248, 0.3469153046607971, -0.07768424600362778, 0.15731769800186157, -0.34134066104888916, -0.04604510962963104, -0.4354317784309387, 0.14782841503620148, 0.07338923215866089, 0.28789299726486206, -0.06971324235200882, -0.02444002591073513, -0.11077526211738586, 0.10292856395244598, 0.6152384281158447, -0.5185811519622803, -0.12251337617635727, 0.10406100004911423, 0.08391167968511581, -0.1375816911458969, -0.10161527246236801, -0.26018163561820984, 0.014868251048028469, 0.4297329783439636, 0.33571842312812805, -0.37738242745399475, -0.2934282422065735, 0.17445822060108185, -0.2098122090101242, -0.08447550982236862, -0.1624630093574524, -0.25328338146209717, -0.22844815254211426, -0.12505638599395752, 0.041775889694690704, 0.2297196090221405, 0.22009973227977753, -0.06440143287181854, 0.16535010933876038, -0.10738155245780945, 0.20170758664608002, 0.21851950883865356, 0.02303250879049301, -0.04001285508275032, 0.2556944489479065, 0.19773399829864502, 0.03954583778977394, -0.1268983632326126, -0.36823809146881104, 0.06594070047140121, -0.05863909795880318, 0.06726883351802826, 0.25432857871055603, -0.03158380836248398, 0.6494697332382202, 0.04232890531420708, 0.33929699659347534, 0.29567623138427734, -0.17104026675224304, 0.00383656844496727, -0.22791878879070282, 0.1694692075252533, 0.22283922135829926, 0.22636952996253967, -0.15907929837703705, -0.2618807256221771, 0.5629376769065857, 0.2996862232685089, -0.07139312475919724, 0.45167842507362366, -0.4898039698600769, -0.3304510712623596, 0.26246377825737, 0.0708126649260521, 0.6956493854522705, -0.06268035620450974, 0.29476606845855713, 0.07930520921945572, -0.3640843331813812, 0.6168705224990845, -0.36440685391426086, 0.15062405169010162, -0.10622136294841766, -0.1969083994626999, 0.012675273232161999, -0.09358679503202438, -0.0731007382273674, 0.2685151696205139, -0.38254594802856445, 0.307796835899353, 0.42872586846351624, -0.20226293802261353, -0.02535995841026306, 0.5634382367134094, 0.056232914328575134, -0.39850568771362305, -0.15839365124702454, 0.06879642605781555, -0.30815503001213074, 0.054364562034606934, -0.16654208302497864, -0.25808972120285034, 0.2257055789232254, 0.15583239495754242, -0.30135032534599304, 0.06679005175828934, -0.03523074835538864, 0.21778255701065063, 0.21286368370056152, 0.01526521798223257, 0.2679019570350647, -0.11605995148420334, -0.03785102814435959, -0.1535062938928604, -0.08397172391414642, 0.13552704453468323, -0.19333329796791077, 0.03781455010175705, 0.018268408253788948, -0.18674054741859436, 0.4988890290260315, -0.06247991323471069, -0.11852169036865234, -0.21425122022628784, 0.2802071273326874, -0.4461525082588196, -0.1915704607963562, 0.14757540822029114, 0.17568407952785492, 0.19827388226985931, -0.304949551820755, 0.5695930123329163, 0.10945088416337967, 0.011156787164509296, 0.03410428762435913, 0.21011927723884583, -0.27115851640701294, 0.5002739429473877, -0.2811298668384552, 0.009963661432266235, 0.10483988374471664, 0.025052066892385483, -0.11265403032302856, -0.38574904203414917, -0.030786756426095963, 0.3170592486858368, 0.13859905302524567, 0.008726694621145725, 0.02852930687367916, 0.11237512528896332, -0.06501560658216476, -0.02706971764564514, -0.1449051797389984, -0.35135388374328613, 0.13187648355960846, 0.18006309866905212, 0.1875753104686737, 0.232990100979805, -0.02897302620112896, -0.2715381681919098, -0.16603919863700867, 0.44810640811920166, 0.03526046499609947, 0.2099885642528534, -0.1294512301683426, 0.3465990424156189, 0.4091411530971527, 0.5483705997467041, -0.259594589471817, 0.07455850392580032, 0.055366747081279755, 0.11098060756921768, 0.4592457711696625, -0.2511616349220276, 0.07607949525117874, 0.27714860439300537, -0.12474231421947479, 0.03135412186384201, -0.3204188346862793, -0.15355819463729858, -0.2497408092021942, 0.11093185842037201, 0.10442163795232773, 0.004201395902782679, 0.004018181003630161, -0.10565141588449478, -0.11311271041631699, -0.08425271511077881, -0.00029923015972599387, -0.38162797689437866, -0.059590961784124374, 0.12420312315225601, -0.038066547363996506, 0.06506042927503586, 0.03687990456819534, 0.1877909004688263, 0.023435764014720917, -0.10253800451755524, 0.3493967354297638, 0.11723912507295609, 0.10813342779874802, -0.17780663073062897, -0.16322283446788788, 0.0233441274613142, 0.2428954690694809, 0.1519826352596283, -0.06412450224161148, 0.12868060171604156, -0.24157734215259552, -0.12066943198442459, 0.23631323873996735, 0.1691906899213791, 0.017448076978325844, -0.10271228849887848, 0.13327576220035553, 0.11498119682073593, -0.09520771354436874, -0.035100195556879044, 0.27450188994407654, 0.07827569544315338, 0.1340380758047104, 0.14364270865917206, 0.0987599566578865, 0.47280073165893555, -0.23880413174629211, 0.009598800912499428, 0.14305821061134338, -0.352010041475296, 0.12352635711431503, -0.1303716003894806, -0.03644505888223648, -0.011556494981050491, 0.10323693603277206, -0.024381542578339577, 0.1411908119916916, -0.2700303792953491, 0.28077471256256104, 0.3052043318748474, 0.0004444957594387233, -0.033914100378751755, 0.22373414039611816, 0.05606937035918236, -0.14373543858528137, 0.24643707275390625, 0.04788106307387352, 0.3264901638031006, 0.021484870463609695, 0.8898254632949829, -0.2593217194080353, -0.44537031650543213, 0.10914997011423111, 0.4276822507381439, 0.04372868314385414, -0.20927801728248596, 0.14286822080612183, 0.10250186920166016, -0.35283270478248596, 0.073243148624897, 0.06681445986032486, -0.2198759764432907, 0.6501642465591431, -0.10448949038982391, 0.21174609661102295, -0.1332387924194336, -0.1671053022146225, 0.39403867721557617, 0.1751192808151245, -0.06507761031389236, 0.01570737548172474, 0.1436767280101776, -0.12336242198944092, -0.17897897958755493, 0.43864351511001587, 0.21286165714263916, 0.02780913934111595, -0.46195387840270996, 0.06281668692827225, 0.09056834131479263, 0.14692993462085724, -0.03955427184700966, 0.10659749805927277, 0.4349878132343292, -0.15175749361515045, 0.2386014610528946, 0.013740160502493382, -0.05349770560860634, -0.19473090767860413, 0.07981132715940475, -0.42988643050193787, -0.3788128197193146, 0.47887900471687317, 0.06758588552474976, -0.11323978006839752, 0.05943525582551956, 0.33388495445251465, -0.6175374984741211, -0.2591400146484375, 0.32679057121276855, -0.09543346613645554, -0.06627283245325089, -0.19924728572368622, 0.07281950116157532, 0.06321022659540176, 0.6397787928581238, 0.17726370692253113, 0.3605991303920746, -0.3750835955142975, -0.12910133600234985, -0.6346367597579956, -0.15605154633522034, 0.11392433196306229, -0.17616459727287292, -0.2531372606754303, 0.1707489937543869, -0.04700442776083946, 0.32274821400642395, -0.2095470130443573, 0.09758244454860687, 0.2715466618537903, 0.02710801362991333, -0.4724883437156677, 0.0400514155626297, 0.09500545263290405, 0.04796018823981285, 0.2817946672439575, -0.0757131576538086, -0.19103245437145233, 0.06382110714912415, -0.06736105680465698, -0.08385728299617767, -0.11887863278388977, 0.2494053691625595, 0.17717045545578003, 0.06206126883625984, 0.03445291891694069, 0.5532026290893555, 0.01081832405179739, -0.16162721812725067, 0.031094711273908615, 0.028595320880413055, -0.3770720362663269, -0.1337217539548874, -0.08216767758131027, 0.20839688181877136, -0.39294567704200745, 0.252956748008728, -0.007655053399503231, -0.02523532323539257, 0.1193833202123642, 0.011497118510305882, -0.20174461603164673, -0.02404620498418808, 0.08407945185899734, 0.2401800900697708, 0.2028205692768097, 0.7438049912452698, -0.014433801174163818, 0.17315638065338135, -0.08187635242938995, -0.03745248168706894, 0.06602835655212402, -0.13764019310474396, -0.24784807860851288, -0.13568295538425446, 0.0355520136654377, 0.062083568423986435, 0.12089275568723679, -0.676857590675354, -0.3293690085411072, 0.045108720660209656, -0.0008662724867463112, 0.0040007405914366245, 0.74090576171875, 0.06504032760858536, -0.11601396650075912, -0.11704631894826889, -0.3568645715713501, 0.03341871127486229, -0.25153300166130066, 0.1817403882741928, -0.371323823928833 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
The problem with changing `k` and `t` per sampling is that you'd have to somehow remember which examples you'd already returned while re-weighting the remaining examples based on the new `k` and `t`values. It seems possible but complicated (I can't really see a reason why you'd want to change the weighting of datasets after you constructed the multidataset). Wouldn't it be convenient if it implemented the dataset interface? Then if someone has code using a single nlp dataset, they can replace it with a multitask combination of more datasets without having to change other code. We would at least need to be able to pass it into a `DataLoader`.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
109
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. The problem with changing `k` and `t` per sampling is that you'd have to somehow remember which examples you'd already returned while re-weighting the remaining examples based on the new `k` and `t`values. It seems possible but complicated (I can't really see a reason why you'd want to change the weighting of datasets after you constructed the multidataset). Wouldn't it be convenient if it implemented the dataset interface? Then if someone has code using a single nlp dataset, they can replace it with a multitask combination of more datasets without having to change other code. We would at least need to be able to pass it into a `DataLoader`.
[ -0.08343596756458282, -0.37729039788246155, -0.05401250347495079, -0.0832778736948967, -0.212246835231781, 0.016944611445069313, 0.13349972665309906, 0.06816421449184418, 0.3124590218067169, -0.02525966428220272, -0.23778492212295532, 0.3599814772605896, -0.15870462357997894, 0.3497791588306427, 0.38235917687416077, -0.4798979163169861, -0.03074454888701439, -0.13304591178894043, -0.3688003420829773, 0.38237258791923523, 0.0409013107419014, 0.09933431446552277, -0.12265484035015106, 0.007462510373443365, -0.23627148568630219, -0.15221893787384033, -0.3432963490486145, 0.09290074557065964, 0.18456177413463593, -0.055031999945640564, 0.19634565711021423, 0.44995468854904175, -0.017280584201216698, 0.05232072249054909, -0.00011072369670728222, -0.08731920272111893, 0.1317603588104248, -0.1200200617313385, 0.049834415316581726, 0.0378306619822979, 0.21656598150730133, -0.41172778606414795, -0.06971439719200134, -0.18803788721561432, 0.1087680533528328, 0.21512429416179657, 0.1830592155456543, 0.024412494152784348, 0.41552063822746277, -0.32108643651008606, 0.12936988472938538, 0.3268473148345947, -0.2219880223274231, 0.2345314472913742, -0.0946439877152443, -0.15330494940280914, 0.09412910044193268, 0.3127887547016144, 0.4916842579841614, -0.547714352607727, -0.06117946654558182, 0.14151182770729065, 0.007721433416008949, 0.05961744114756584, 0.10257232934236526, -0.05151675269007683, -0.3251843750476837, -0.2065102756023407, -0.3548212945461273, 0.29408180713653564, -0.18114663660526276, -0.1491723656654358, -0.19877657294273376, -0.5856713652610779, 0.060058850795030594, -0.01082699280232191, -0.299710214138031, 0.02046123333275318, -0.14967548847198486, -0.012793189845979214, -0.12954573333263397, -0.09831388294696808, 0.07478515058755875, 0.09907804429531097, 0.3274923861026764, 0.5603525638580322, 0.0938606783747673, 0.1836351901292801, 0.3918946087360382, 0.07738862186670303, -0.21224859356880188, -0.24602682888507843, 0.28512832522392273, 0.096305251121521, -0.282671183347702, -0.34620562195777893, 0.16300152242183685, -0.263979434967041, 0.23295825719833374, 0.0578402578830719, 0.16775856912136078, 0.11276677250862122, -0.1378822773694992, 0.24946406483650208, 0.28214263916015625, 0.05985144153237343, 0.01065152883529663, -0.1180340126156807, 0.014983871020376682, -0.05733712762594223, 0.15855072438716888, 0.09276831895112991, 0.21617837250232697, -0.04496941715478897, -0.40501487255096436, -0.08261949568986893, -0.2754172086715698, 0.19549527764320374, -0.20601999759674072, -0.5297427773475647, -0.018626810982823372, -0.11480309069156647, 0.023904908448457718, -0.10028781741857529, -0.3187434673309326, 0.1601448506116867, -0.05653901770710945, 0.1746995896100998, -0.11744719743728638, 0.07287564873695374, -0.08563556522130966, 0.08851902931928635, -0.4986368715763092, 0.025194140151143074, 0.2598647475242615, 0.08004960417747498, 0.16220267117023468, 0.1625867336988449, -0.026163287460803986, 0.1506989300251007, 0.39101871848106384, 0.05960140377283096, 0.0018364964053034782, -0.05813088268041611, 0.1590106040239334, -0.29496967792510986, -0.042152684181928635, 0.1823379397392273, -0.42464184761047363, -0.15480490028858185, -0.14691752195358276, -0.07604163140058517, 0.1608911156654358, 0.045543018728494644, -0.24469682574272156, -0.29811903834342957, -0.4465659260749817, 0.7699326276779175, -0.009563101455569267, -0.11291150748729706, -0.10955788940191269, 0.20794647932052612, -0.30656126141548157, -0.08008915930986404, 0.2194669097661972, 0.03209225460886955, -0.019059257581830025, -0.4033973813056946, -0.042270831763744354, -0.037179261445999146, -0.10395511239767075, 0.26446613669395447, -0.2160889357328415, 0.1813713163137436, 0.07405558228492737, 0.17394722998142242, 0.2999924421310425, -0.46792587637901306, -0.08489713072776794, 0.23441095650196075, -0.03608766943216324, 0.32896745204925537, 0.0849846675992012, 0.1241305023431778, 0.051966071128845215, -0.1715918481349945, 0.17609839141368866, 0.9709122776985168, -0.35481831431388855, 0.024371661245822906, -0.10637873411178589, -0.38635125756263733, 0.5253028869628906, 0.2912897765636444, 0.0707937702536583, -0.2695537805557251, -0.12689059972763062, 0.0046651228331029415, 0.09227938205003738, 0.01856304332613945, 0.08767016232013702, -0.04912399500608444, -0.3308201730251312, 0.009197410196065903, -0.21637961268424988, -0.1552148163318634, -0.37339720129966736, 0.0538264624774456, -0.061896663159132004, 0.387917697429657, 0.3986716568470001, -0.11829853802919388, 0.2000272125005722, -0.3578912913799286, -0.03352352976799011, -0.28234612941741943, 0.1489841341972351, -0.013988283462822437, 0.04843536764383316, -0.22091196477413177, -0.04951252415776253, 0.23565366864204407, 0.1350266933441162, -0.15551720559597015, -0.407478004693985, 0.3045089840888977, 0.02017449028789997, -0.001504183979704976, -0.054201237857341766, 0.5833075642585754, -0.16488279402256012, -0.10873354971408844, 0.2087198793888092, 0.30474424362182617, -0.3626171052455902, 0.2941076159477234, 0.3104536235332489, 0.19204045832157135, 0.1580664962530136, 0.019926026463508606, 0.15771156549453735, -0.014536633156239986, -0.0647680014371872, -0.16084501147270203, -0.0415622852742672, 0.6416842341423035, -0.048300474882125854, 0.348207950592041, -0.0292414091527462, -0.031100578606128693, -0.1747986078262329, -0.16735880076885223, -0.2596576511859894, 0.33093762397766113, 0.2930160462856293, -0.2784145474433899, 0.36245134472846985, -0.12749864161014557, -0.4090743660926819, 0.1585981696844101, 0.19831019639968872, 0.05338175594806671, 0.12208274751901627, 0.011676471680402756, -0.05402818322181702, -0.30141109228134155, -0.10330019891262054, 0.34477943181991577, 0.6520264148712158, 0.2875113785266876, 0.16832205653190613, 0.021604279056191444, -0.2816530764102936, -0.09253256022930145, 0.002398334676399827, 0.1340308040380478, -0.029897043481469154, 0.18200212717056274, 0.10243808478116989, 0.031474363058805466, -0.08563216775655746, -0.06431403756141663, 0.16443881392478943, -0.21713672578334808, -0.03231697529554367, 0.03986555337905884, -0.17864571511745453, -0.5128594040870667, -0.29099351167678833, 0.03559592366218567, -0.3008250296115875, -0.3023710548877716, 0.18047207593917847, 0.09110572189092636, -0.22132913768291473, 0.21164411306381226, 0.237873375415802, 0.6204207539558411, -0.39749497175216675, -0.12988007068634033, -0.10952067375183105, -0.346415251493454, -0.10274600982666016, 0.1299118846654892, 0.35552820563316345, 0.27457284927368164, 0.4790077805519104, 0.2562713623046875, -0.05841077119112015, 0.1019946038722992, -0.40407443046569824, 0.026286179199814796, -0.319781631231308, 0.2502516508102417, 0.11853636056184769, -0.18916521966457367, 0.1525098979473114, -0.3981902003288269, 0.011864658445119858, -0.14020180702209473, -0.029884548857808113, -0.059307195246219635, 0.011886421591043472, 0.14863160252571106, -0.20273514091968536, -0.20275530219078064, -0.608305811882019, -0.4239635169506073, 0.245614156126976, -0.2116800844669342, 0.04542825371026993, 0.05185750499367714, -0.21421359479427338, 0.14821328222751617, -0.1654004454612732, 0.015458811074495316, -0.18778501451015472, -0.1026327908039093, -0.006060943938791752, -0.39064279198646545, -0.10505211353302002, -0.504276692867279, -0.12487275898456573, 0.2693505585193634, -0.10490928590297699, -0.10202112048864365, -0.3205796182155609, 0.05709993466734886, -0.09570715576410294, 0.016935022547841072, 0.2777612805366516, 0.10743698477745056, -0.22773592174053192, 0.03465208783745766, -0.06954718381166458, -0.21417218446731567, 0.21847239136695862, 0.4109446704387665, 0.38754042983055115, 0.11119987815618515, 0.0037969525437802076, 0.14250265061855316, 0.7288061380386353, 0.20422515273094177, -0.006215783301740885, 0.2417243868112564, 0.22308948636054993, 0.059006430208683014, 0.116936095058918, -0.14880694448947906, 0.2164953649044037, -0.1390654742717743, 0.5181331038475037, 0.13408245146274567, 0.14490394294261932, -0.23636089265346527, -0.26169630885124207, 0.0015474569518119097, -0.20625972747802734, -0.2486131191253662, 0.2672933042049408, -0.23298242688179016, -0.038977280259132385, -0.043079692870378494, -0.3545140326023102, -0.3271300494670868, -0.04302430525422096, 0.021406864747405052, 0.16934043169021606, -0.009690525941550732, 0.13491587340831757, -0.4466549754142761, 0.00731002539396286, -0.48752468824386597, 0.17487017810344696, 0.03552931547164917, 0.20631344616413116, -0.014624569565057755, -0.09552477300167084, -0.13732247054576874, 0.10531513392925262, 0.4926849603652954, -0.5116750597953796, -0.07899940013885498, 0.08137312531471252, 0.06042462959885597, -0.13401301205158234, -0.13115063309669495, -0.32275527715682983, -0.0770329162478447, 0.44155147671699524, 0.37745505571365356, -0.42304709553718567, -0.30254584550857544, 0.22415271401405334, -0.16462074220180511, -0.09548002481460571, -0.1715964525938034, -0.21112029254436493, -0.1963450312614441, -0.15463051199913025, 0.07413206249475479, 0.14330673217773438, 0.2671215236186981, -0.12531571090221405, 0.19032765924930573, -0.10813857614994049, 0.1845884919166565, 0.1798727959394455, 0.01582728885114193, -0.044192202389240265, 0.35392338037490845, 0.11021852493286133, 0.07063734531402588, -0.131255641579628, -0.3319585621356964, 0.10548872500658035, -0.13519957661628723, 0.20508283376693726, 0.22697044909000397, 0.09067510068416595, 0.5530039668083191, 0.11436271667480469, 0.3576848804950714, 0.30341124534606934, -0.0634748637676239, 0.125437930226326, -0.3138454556465149, 0.2028869092464447, 0.17584088444709778, 0.25504419207572937, -0.10684085637331009, -0.21424846351146698, 0.5952860713005066, 0.2874777913093567, -0.08879192918539047, 0.38233262300491333, -0.5090106129646301, -0.3641583025455475, 0.2897244095802307, 0.08487631380558014, 0.6568556427955627, -0.012666381895542145, 0.3372669219970703, 0.059179287403821945, -0.30404409766197205, 0.6199080348014832, -0.2759537100791931, 0.1473556011915207, -0.13500402867794037, -0.24565228819847107, 0.03667650371789932, -0.048641499131917953, -0.08241476118564606, 0.2677821218967438, -0.3512749969959259, 0.367671936750412, 0.44386735558509827, -0.1998150497674942, -0.129360631108284, 0.5204609036445618, -0.08692406862974167, -0.32492923736572266, -0.11418427526950836, 0.09938502311706543, -0.23347404599189758, 0.11657463014125824, -0.12016189098358154, -0.250545471906662, 0.22407959401607513, 0.13571149110794067, -0.2270088940858841, 0.045545440167188644, -0.06588154286146164, 0.13693247735500336, 0.2686852812767029, 0.030228247866034508, 0.1925262212753296, -0.2200125902891159, -0.02552666887640953, -0.10796097666025162, -0.10625415295362473, 0.12955379486083984, -0.16364243626594543, 0.07055362313985825, 0.13920433819293976, -0.14759700000286102, 0.5190824866294861, 0.012029717676341534, -0.11477196961641312, -0.19632397592067719, 0.2693728804588318, -0.3825458586215973, -0.17794667184352875, 0.024098346009850502, 0.0924796462059021, 0.10347113758325577, -0.30905479192733765, 0.5627151131629944, 0.12509994208812714, -0.028349393978714943, 0.06166955083608627, 0.10210768133401871, -0.22989462316036224, 0.512381911277771, -0.18565788865089417, -0.03955089673399925, 0.05415215715765953, 0.09490787237882614, -0.1626538336277008, -0.30411332845687866, -0.04040665552020073, 0.24128514528274536, 0.13462112843990326, -0.014453310519456863, 0.014386431314051151, 0.16485275328159332, -0.17030973732471466, 0.043305546045303345, -0.0753474235534668, -0.34250250458717346, 0.15653552114963531, 0.2247282713651657, 0.21894167363643646, 0.2962498068809509, -0.11109855771064758, -0.18540410697460175, -0.12255596369504929, 0.4910309612751007, 0.09181645512580872, 0.2689322233200073, -0.11989612877368927, 0.3225714862346649, 0.35474762320518494, 0.5900267958641052, -0.3123569190502167, 0.07766642421483994, 0.06387712061405182, 0.13484321534633636, 0.509900689125061, -0.2374131977558136, 0.00896961148828268, 0.3305617868900299, -0.1103556752204895, 0.07589998841285706, -0.32956811785697937, -0.1768559217453003, -0.2944238781929016, 0.08765871822834015, 0.07303505390882492, 0.012625504285097122, 0.023895010352134705, -0.03426361083984375, -0.04506850615143776, -0.16708719730377197, 0.005586294457316399, -0.4168512523174286, -0.10029652714729309, 0.20001031458377838, -0.061024196445941925, 0.0643853023648262, 0.14227387309074402, 0.22846290469169617, 0.009681159630417824, -0.15692190825939178, 0.24389781057834625, 0.1689838469028473, 0.10425107926130295, -0.12324467301368713, -0.16884909570217133, -0.002899676561355591, 0.2162175178527832, 0.08830507844686508, -0.12145940959453583, 0.19059306383132935, -0.2052348256111145, -0.09061140567064285, 0.31784725189208984, 0.18736647069454193, 0.04893608018755913, -0.18051356077194214, 0.15276871621608734, 0.16645260155200958, -0.11254414170980453, 0.0003931565152015537, 0.15967291593551636, 0.11689870804548264, 0.15863411128520966, 0.1467079520225525, 0.14573164284229279, 0.3819445073604584, -0.19101963937282562, -0.04100773110985756, 0.08764369785785675, -0.32143500447273254, 0.11080771684646606, -0.060432739555835724, -0.020613402128219604, -0.032124072313308716, 0.1805833876132965, -0.06326822191476822, 0.2051950991153717, -0.4029548764228821, 0.2911439836025238, 0.3381801247596741, -0.008856015279889107, 0.03018956072628498, 0.12258732318878174, 0.011403042823076248, -0.1338501274585724, 0.37775200605392456, 0.05748864263296127, 0.3323085904121399, 0.05188087001442909, 0.9268811941146851, -0.262957900762558, -0.44818466901779175, 0.06358586251735687, 0.5313950181007385, 0.029138533398509026, -0.32141920924186707, 0.061389267444610596, 0.06478076428174973, -0.3745325207710266, 0.16029535233974457, 0.011611673049628735, -0.15400655567646027, 0.6641730070114136, -0.15340939164161682, 0.19332946836948395, -0.16817358136177063, -0.2260017991065979, 0.28352800011634827, 0.18811418116092682, -0.1497502624988556, 0.04486066475510597, 0.07804404199123383, -0.06418509036302567, -0.12437900900840759, 0.5293443202972412, 0.1735813021659851, -0.013323095627129078, -0.4142279028892517, -0.003277060342952609, 0.1442958265542984, 0.16323933005332947, -0.1211547926068306, 0.1270752102136612, 0.47768494486808777, -0.17132702469825745, 0.2919727563858032, 0.05136556923389435, -0.14167405664920807, -0.16505229473114014, 0.05372549220919609, -0.40651339292526245, -0.4111210107803345, 0.41772323846817017, 0.1373804658651352, -0.04147876799106598, 0.06708186864852905, 0.28556162118911743, -0.5498074889183044, -0.22149886190891266, 0.2549925148487091, -0.17315106093883514, -0.0709797590970993, -0.22309322655200958, 0.09759364277124405, 0.04475836083292961, 0.647975504398346, 0.21212534606456757, 0.25204381346702576, -0.35490477085113525, -0.0999404713511467, -0.551097571849823, -0.1398828625679016, 0.17897365987300873, -0.10730237513780594, -0.27738451957702637, 0.16404426097869873, -0.036955736577510834, 0.2786957621574402, -0.18398289382457733, 0.1812703162431717, 0.21143680810928345, 0.009155137464404106, -0.4085697531700134, 0.1516011357307434, 0.01877441443502903, 0.041605688631534576, 0.3125753700733185, -0.11525017023086548, -0.1505114883184433, 0.02137487567961216, -0.02663578651845455, -0.045244038105010986, -0.0816572904586792, 0.24790668487548828, 0.1851312518119812, 0.16567835211753845, 0.013480708003044128, 0.4840938448905945, -0.07819242775440216, -0.12652726471424103, -0.020788226276636124, -0.07461056113243103, -0.4193071722984314, -0.2067510187625885, -0.09964533895254135, 0.21812154352664948, -0.27957722544670105, 0.3174087703227997, -0.009172584861516953, -0.11152952164411545, 0.15208306908607483, 0.06763501465320587, -0.10436654835939407, -0.05300767347216606, 0.1152551919221878, 0.09159164875745773, 0.1921045482158661, 0.7487502098083496, 0.06614495068788528, 0.09840822219848633, -0.08251723647117615, -0.11334831267595291, 0.08893370628356934, -0.06935162097215652, -0.18787802755832672, -0.16191406548023224, 0.01233151275664568, 0.10883473604917526, 0.0035939563531428576, -0.6421051025390625, -0.30900174379348755, 0.028569648042321205, 0.040786031633615494, 0.03615647181868553, 0.709376335144043, 0.052084099501371384, -0.1664981096982956, -0.12391171604394913, -0.3487201929092407, 0.08365879207849503, -0.3357553482055664, 0.06608200073242188, -0.3758983612060547 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
A very janky (but working) implementation of `multitask_dataset.sample()` could be something like this: ```python import nlp import torch class MultiDataset(): def __init__(self, *args, temperature=2.0, k=1000, maximum=None, scale=1): self.datasets = args self._dataloaders = {} for split in self._get_common_splits(): split_datasets = [ds[split] for ds in self.datasets] mixing_rates = self._calc_mixing_rates(split_datasets,temperature, k, maximum, scale) weights = [] for i in range(len(self.datasets)): weights += [mixing_rates[i]]*len(self.datasets[i][split]) self._dataloaders[split] = torch.utils.data.DataLoader(torch.utils.data.ConcatDataset(split_datasets), sampler=torch.utils.data.sampler.WeightedRandomSampler( num_samples=len(weights), weights = weights, replacement=True), shuffle=False) def _get_common_splits(self): '''Finds the common splits present in all self.datasets''' min_set = None for dataset in self.datasets: if min_set != None: min_set.intersection(set(dataset.keys())) else: min_set = set(dataset.keys()) return min_set def _calc_mixing_rates(self,datasets, temperature=2.0, k=1000, maximum=None, scale=1): '''Work out the weighting of each dataset based on t and k''' mixing_rates = [] for dataset in datasets: rate = len(dataset) rate *= scale if maximum: rate = min(rate, maximum) if temperature != 1.0: rate = rate ** (1.0/temperature) mixing_rates.append(rate) return mixing_rates def sample(self,n,split): batch = [] for example in self._dataloaders[split]: batch.append(example) n -= 1 if n == 0: return batch def flatten(dataset,flatten_fn): for k in dataset.keys(): if isinstance(dataset[k],nlp.Dataset): dataset[k] = dataset[k].map(flatten_fn,remove_columns=dataset[k].column_names) # Squad def flatten_squad(example): return {"source": "squad context: " + example['context'] + " question: " + example['question'],"target":example["answers"]["text"]} squad = nlp.load_dataset("squad") flatten(squad,flatten_squad) # CNN_DM def flatten_cnn_dm(example): return {"source": "cnn_dm: " + example['article'],"target":[example["highlights"]]} cnn_dm = nlp.load_dataset("cnn_dailymail", "3.0.0") flatten(cnn_dm,flatten_cnn_dm) multitask_dataset = MultiDataset(squad, cnn_dm) batch = multitask_dataset.sample(100,"train") ``` There's definitely a more sensible way than embedding `DataLoader`s inside.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
231
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. A very janky (but working) implementation of `multitask_dataset.sample()` could be something like this: ```python import nlp import torch class MultiDataset(): def __init__(self, *args, temperature=2.0, k=1000, maximum=None, scale=1): self.datasets = args self._dataloaders = {} for split in self._get_common_splits(): split_datasets = [ds[split] for ds in self.datasets] mixing_rates = self._calc_mixing_rates(split_datasets,temperature, k, maximum, scale) weights = [] for i in range(len(self.datasets)): weights += [mixing_rates[i]]*len(self.datasets[i][split]) self._dataloaders[split] = torch.utils.data.DataLoader(torch.utils.data.ConcatDataset(split_datasets), sampler=torch.utils.data.sampler.WeightedRandomSampler( num_samples=len(weights), weights = weights, replacement=True), shuffle=False) def _get_common_splits(self): '''Finds the common splits present in all self.datasets''' min_set = None for dataset in self.datasets: if min_set != None: min_set.intersection(set(dataset.keys())) else: min_set = set(dataset.keys()) return min_set def _calc_mixing_rates(self,datasets, temperature=2.0, k=1000, maximum=None, scale=1): '''Work out the weighting of each dataset based on t and k''' mixing_rates = [] for dataset in datasets: rate = len(dataset) rate *= scale if maximum: rate = min(rate, maximum) if temperature != 1.0: rate = rate ** (1.0/temperature) mixing_rates.append(rate) return mixing_rates def sample(self,n,split): batch = [] for example in self._dataloaders[split]: batch.append(example) n -= 1 if n == 0: return batch def flatten(dataset,flatten_fn): for k in dataset.keys(): if isinstance(dataset[k],nlp.Dataset): dataset[k] = dataset[k].map(flatten_fn,remove_columns=dataset[k].column_names) # Squad def flatten_squad(example): return {"source": "squad context: " + example['context'] + " question: " + example['question'],"target":example["answers"]["text"]} squad = nlp.load_dataset("squad") flatten(squad,flatten_squad) # CNN_DM def flatten_cnn_dm(example): return {"source": "cnn_dm: " + example['article'],"target":[example["highlights"]]} cnn_dm = nlp.load_dataset("cnn_dailymail", "3.0.0") flatten(cnn_dm,flatten_cnn_dm) multitask_dataset = MultiDataset(squad, cnn_dm) batch = multitask_dataset.sample(100,"train") ``` There's definitely a more sensible way than embedding `DataLoader`s inside.
[ -0.08342023193836212, -0.38811933994293213, -0.058892376720905304, -0.0579487644135952, -0.25454190373420715, 0.06940426677465439, 0.19493088126182556, 0.05141011252999306, 0.34056010842323303, -0.14826437830924988, -0.20243516564369202, 0.3717308044433594, -0.23944765329360962, 0.34754109382629395, 0.38729801774024963, -0.4599963426589966, 0.024137498810887337, -0.2675699293613434, -0.26987817883491516, 0.3728724420070648, 0.06227037310600281, 0.08812984079122543, -0.18115825951099396, 0.08036605268716812, -0.25397735834121704, -0.1634482890367508, -0.34938836097717285, 0.10462171584367752, 0.15728843212127686, -0.013107460923492908, 0.21203705668449402, 0.38126274943351746, -0.0404963418841362, 0.13177834451198578, -0.00011317195458104834, -0.07457657903432846, 0.21052570641040802, -0.07623807340860367, 0.08656402677297592, -0.06524460017681122, 0.2296265810728073, -0.4317648708820343, 0.036713067442178726, -0.14333127439022064, 0.14059743285179138, 0.1517212986946106, 0.16424991190433502, -0.0074315145611763, 0.38448086380958557, -0.27362778782844543, 0.11178207397460938, 0.28740960359573364, -0.1595599353313446, 0.2645363211631775, -0.10869432240724564, -0.04744035005569458, 0.04901286959648132, 0.2619539499282837, 0.6061143279075623, -0.6340859532356262, -0.014125113375484943, 0.10082380473613739, -0.04980522021651268, 0.031944118440151215, 0.1234842911362648, -0.08227342367172241, -0.3827718198299408, -0.23529954254627228, -0.3922519087791443, 0.2454875260591507, -0.17746670544147491, -0.1945362538099289, -0.22744792699813843, -0.5580018758773804, 0.006252939347177744, 0.041994791477918625, -0.30483847856521606, 0.050221920013427734, -0.19981621205806732, -0.06731585413217545, -0.13756787776947021, -0.08848818391561508, 0.011830191127955914, 0.17321676015853882, 0.31564927101135254, 0.5342757701873779, 0.10947076231241226, 0.17463359236717224, 0.3422691524028778, 0.04283103346824646, -0.13867327570915222, -0.28696683049201965, 0.3420218527317047, 0.12645862996578217, -0.37124958634376526, -0.30758586525917053, 0.1824493259191513, -0.2611481547355652, 0.25415703654289246, 0.11465180665254593, 0.11636385321617126, 0.09271908551454544, -0.12476256489753723, 0.2851982116699219, 0.3020448088645935, 0.015004734508693218, -0.08135770261287689, -0.08082737028598785, 0.023554407060146332, -0.13282231986522675, 0.22979427874088287, 0.12041173130273819, 0.1645006239414215, -0.030591757968068123, -0.40878111124038696, -0.06945715099573135, -0.16123688220977783, 0.1976405531167984, -0.23831014335155487, -0.48460257053375244, -0.1420590579509735, -0.08315490931272507, 0.07923731207847595, -0.09192744642496109, -0.2979576289653778, 0.10018078237771988, -0.07304764539003372, 0.24686883389949799, -0.10002510994672775, 0.07183794677257538, -0.056345127522945404, 0.0983508750796318, -0.43944042921066284, -0.06991681456565857, 0.32330989837646484, 0.14224645495414734, 0.11645498871803284, 0.17511357367038727, -0.016918161883950233, 0.1301407665014267, 0.4420311450958252, -0.0019401790341362357, 0.020958175882697105, -0.01405948493629694, 0.1952323466539383, -0.3251473009586334, 0.004883179906755686, 0.22755217552185059, -0.37265798449516296, -0.14289061725139618, -0.128090500831604, -0.1560910940170288, 0.1126977726817131, 0.03565239906311035, -0.23599126935005188, -0.2588006556034088, -0.4986608922481537, 0.6829752326011658, -0.02900400012731552, -0.04187630116939545, -0.1579708307981491, 0.2522132396697998, -0.24844463169574738, -0.1251499056816101, 0.21692012250423431, 0.06579384952783585, -0.002443592296913266, -0.31193816661834717, -0.1297091543674469, -0.08177625387907028, -0.03710135444998741, 0.22530725598335266, -0.16007858514785767, 0.2730761766433716, 0.04683201014995575, 0.2083573341369629, 0.26415303349494934, -0.5000431537628174, -0.04386594146490097, 0.266920268535614, -0.042177170515060425, 0.4183538258075714, 0.04843828082084656, 0.2128075808286667, 0.03480517864227295, -0.16171251237392426, 0.29856300354003906, 1.0328638553619385, -0.38605260848999023, 0.0045570251531898975, -0.09491441398859024, -0.3651701807975769, 0.5233176350593567, 0.3392058312892914, 0.09073849022388458, -0.3195149898529053, -0.1159660592675209, -0.053001802414655685, 0.10058603435754776, 0.08589214831590652, 0.07345230877399445, -0.07229728996753693, -0.365039199590683, 0.059784043580293655, -0.2487357258796692, -0.19382630288600922, -0.342143714427948, 0.030358711257576942, 0.11190682649612427, 0.27976635098457336, 0.24866817891597748, -0.1320004165172577, 0.22297149896621704, -0.25494563579559326, -0.12981511652469635, -0.3390691578388214, 0.1418405920267105, -0.05979080870747566, 0.03747917711734772, -0.2510697841644287, 0.0014998660190030932, 0.3411412537097931, 0.12741728127002716, -0.15057143568992615, -0.47374778985977173, 0.28402602672576904, -0.05593821406364441, 0.03936878591775894, -0.022018037736415863, 0.6338953375816345, -0.16466931998729706, -0.08810720592737198, 0.19178640842437744, 0.28774383664131165, -0.26854732632637024, 0.21729783713817596, 0.3182617425918579, 0.16927006840705872, 0.14661987125873566, -0.022832229733467102, 0.09416741877794266, 0.062102749943733215, -0.02409151755273342, -0.2137683779001236, -0.08747932314872742, 0.5368933081626892, -0.150593563914299, 0.43937888741493225, -0.06569937616586685, 0.06121264398097992, -0.16889765858650208, -0.017659280449151993, -0.18787488341331482, 0.4131902754306793, 0.3052523732185364, -0.19494245946407318, 0.34917739033699036, -0.18313777446746826, -0.37643325328826904, 0.15710029006004333, 0.089298777282238, 0.029829898849129677, 0.09827006608247757, -0.04359466955065727, -0.051643457263708115, -0.2501397132873535, -0.0670483261346817, 0.34258130192756653, 0.6638851165771484, 0.2629100978374481, 0.1862419992685318, 0.04439222440123558, -0.24098949134349823, -0.06112935021519661, 0.008607645519077778, 0.17471592128276825, -0.08582803606987, 0.16375607252120972, 0.055455055087804794, 0.03326430916786194, 0.04951440915465355, -0.13218683004379272, 0.15407127141952515, -0.18271596729755402, -0.07683458179235458, 0.09348588436841965, -0.2060638815164566, -0.5439797639846802, -0.19668960571289062, 0.022410990670323372, -0.2867783010005951, -0.28189772367477417, 0.09111542999744415, 0.09463918954133987, -0.13861504197120667, 0.15519224107265472, 0.21347860991954803, 0.534140408039093, -0.39720308780670166, -0.2968820035457611, -0.021574856713414192, -0.34398671984672546, -0.1300346851348877, 0.12559708952903748, 0.43872183561325073, 0.33325719833374023, 0.5198891162872314, 0.3033442199230194, -0.08945254236459732, 0.04793376848101616, -0.33727186918258667, 0.010944589041173458, -0.28580906987190247, 0.2852506935596466, 0.13329263031482697, -0.242193341255188, 0.16811217367649078, -0.34855756163597107, -0.025233762338757515, -0.0736948624253273, -0.08946555852890015, -0.03885059431195259, -0.002678220858797431, 0.07000242173671722, -0.17970342934131622, -0.17550243437290192, -0.6369870901107788, -0.4389748275279999, 0.29911237955093384, -0.26405516266822815, 0.09133286029100418, -0.016800886020064354, -0.2485511153936386, 0.1854192167520523, -0.11016422510147095, 0.0057688262313604355, -0.17779766023159027, -0.09301396459341049, 0.02673337236046791, -0.3839554786682129, -0.18263882398605347, -0.38308313488960266, -0.1138102188706398, 0.21733084321022034, -0.019878994673490524, -0.023589301854372025, -0.2534295916557312, 0.10554058849811554, -0.06678219884634018, 0.06013956665992737, 0.21710561215877533, 0.1514703631401062, -0.23825204372406006, 0.13472460210323334, -0.05888685584068298, -0.23463928699493408, 0.23617631196975708, 0.2706869840621948, 0.4174253046512604, 0.16836750507354736, 0.04018945246934891, 0.20723868906497955, 0.6981121897697449, 0.24865445494651794, 0.028430690988898277, 0.1425480991601944, 0.2371983677148819, -0.00511866994202137, 0.18883751332759857, -0.2463110387325287, 0.21866919100284576, -0.21620038151741028, 0.48305749893188477, 0.028078535571694374, 0.1475924551486969, -0.21976760029792786, -0.2751483619213104, 0.005366126541048288, -0.2993879020214081, -0.1995483934879303, 0.22123681008815765, -0.2730292081832886, 0.011601455509662628, -0.11817097663879395, -0.3471129834651947, -0.2714879512786865, -0.036768194288015366, 0.013553237542510033, 0.30513352155685425, -0.08174040168523788, 0.15701237320899963, -0.3513636291027069, -0.049009084701538086, -0.44961100816726685, 0.1521972417831421, 0.04184596613049507, 0.2807391583919525, -0.06174781918525696, -0.06096750497817993, -0.10397979617118835, 0.09561669826507568, 0.584717333316803, -0.5166602730751038, -0.09553302079439163, 0.12874868512153625, 0.12548890709877014, -0.1312006562948227, -0.11914404481649399, -0.2858884632587433, 0.07362567633390427, 0.4597923159599304, 0.340938001871109, -0.344413161277771, -0.2527541518211365, 0.21955060958862305, -0.17277483642101288, -0.057592738419771194, -0.18763329088687897, -0.23618759214878082, -0.23544250428676605, -0.08206864446401596, 0.055361490696668625, 0.22892586886882782, 0.22814002633094788, -0.11155752837657928, 0.19488991796970367, -0.12732572853565216, 0.24054819345474243, 0.21754497289657593, -0.002513660816475749, -0.024933401495218277, 0.18107442557811737, 0.17871487140655518, 0.04497780650854111, -0.1487913280725479, -0.3022288382053375, 0.09064873307943344, -0.05727655813097954, 0.11080967634916306, 0.22470267117023468, -0.020743204280734062, 0.6292213797569275, 0.0477001890540123, 0.3185668885707855, 0.2778654992580414, -0.19443415105342865, -0.002981480211019516, -0.2579376697540283, 0.20183652639389038, 0.2203504592180252, 0.2621207535266876, -0.11640992015600204, -0.2639242112636566, 0.5500632524490356, 0.3211117684841156, -0.0659315288066864, 0.4420504570007324, -0.5858774185180664, -0.3123209476470947, 0.2432999312877655, 0.06730388849973679, 0.6210483312606812, -0.011136581189930439, 0.286962628364563, 0.0889556035399437, -0.3830382227897644, 0.5410962104797363, -0.3828496038913727, 0.16390222311019897, -0.08226855844259262, -0.15365350246429443, 0.015216345898807049, -0.09603583067655563, -0.04071812331676483, 0.2526591122150421, -0.32957977056503296, 0.35041865706443787, 0.4120703339576721, -0.20682227611541748, -0.0675337091088295, 0.5238788723945618, -0.013963894918560982, -0.3526764214038849, -0.042020026594400406, 0.08234096318483353, -0.2836208641529083, 0.10621663182973862, -0.1444058120250702, -0.24390126764774323, 0.217010498046875, 0.15858182311058044, -0.34577375650405884, 0.08260791003704071, -0.03591394051909447, 0.22727487981319427, 0.21372590959072113, -0.03624681383371353, 0.2051752358675003, -0.17723801732063293, -0.07493184506893158, -0.1512724757194519, -0.03667193651199341, 0.1263703554868698, -0.14632152020931244, 0.04019325599074364, 0.0808703750371933, -0.218107670545578, 0.45674705505371094, -0.04643131047487259, -0.122221939265728, -0.199751615524292, 0.286871999502182, -0.4222888648509979, -0.18035922944545746, 0.1595400869846344, 0.17278672754764557, 0.2526025176048279, -0.3115995526313782, 0.556342601776123, 0.14838624000549316, -0.010223736986517906, 0.03905479237437248, 0.18057683110237122, -0.260419100522995, 0.5079047083854675, -0.36686766147613525, 0.014650780707597733, 0.0897601768374443, 0.02091219089925289, -0.13653387129306793, -0.36404022574424744, -0.04824399948120117, 0.3149685561656952, 0.1139809861779213, 0.00437669875100255, 0.06601306796073914, 0.12075348198413849, 0.013668476603925228, -0.04037683829665184, -0.16214576363563538, -0.34312960505485535, 0.11653667688369751, 0.1604318469762802, 0.17776013910770416, 0.18723249435424805, -0.04996218904852867, -0.2939021587371826, -0.16337944567203522, 0.4305417239665985, 0.044077713042497635, 0.20243090391159058, -0.13496133685112, 0.35182085633277893, 0.38145267963409424, 0.49481523036956787, -0.25801002979278564, 0.07705677300691605, 0.09505507349967957, 0.09604062885046005, 0.4429807662963867, -0.2144620716571808, 0.0520828478038311, 0.2533465623855591, -0.11877474188804626, 0.029874807223677635, -0.32196852564811707, -0.15790054202079773, -0.2378179281949997, 0.09774894267320633, 0.0700349435210228, 0.017564890906214714, 0.0414375439286232, -0.1024017259478569, -0.08863141387701035, -0.07675302028656006, -0.011173226870596409, -0.324833482503891, -0.06518688797950745, 0.14712892472743988, -0.10908284783363342, 0.04781080782413483, 0.07037947326898575, 0.19740602374076843, -0.002259120810776949, -0.08732935786247253, 0.28135186433792114, 0.12359558790922165, 0.18367743492126465, -0.16559188067913055, -0.14358817040920258, -0.0252524521201849, 0.2127145677804947, 0.1461205631494522, -0.07139448821544647, 0.16139689087867737, -0.21960103511810303, -0.15503600239753723, 0.25354981422424316, 0.2257738709449768, 0.024593526497483253, -0.10969015955924988, 0.147074893116951, 0.11975526064634323, -0.12432178854942322, -0.002117814961820841, 0.2085062712430954, 0.09803588688373566, 0.11718344688415527, 0.1034611389040947, 0.10285257548093796, 0.4659959375858307, -0.30514270067214966, 0.031375233083963394, 0.10533059388399124, -0.3483785390853882, 0.11580241471529007, -0.08272483199834824, -0.0022729961201548576, 0.0024443003349006176, 0.10375615954399109, -0.0023493068292737007, 0.14066536724567413, -0.23288759589195251, 0.3303031027317047, 0.3737261891365051, 0.06398952007293701, 0.007039366755634546, 0.1710142344236374, 0.0021040078718215227, -0.11631634831428528, 0.23376470804214478, 0.02021610736846924, 0.35285118222236633, 0.03638806566596031, 0.8677220344543457, -0.31640496850013733, -0.44532185792922974, 0.10563376545906067, 0.3968658447265625, 0.05611949414014816, -0.23128989338874817, 0.14472177624702454, 0.10362959653139114, -0.3610670864582062, 0.08872591704130173, 0.06925946474075317, -0.17860978841781616, 0.6680487990379333, -0.12029631435871124, 0.16749359667301178, -0.18092724680900574, -0.22236540913581848, 0.35287171602249146, 0.1628759801387787, -0.10737736523151398, -0.035656917840242386, 0.10721278190612793, -0.1029248759150505, -0.1998305767774582, 0.49408864974975586, 0.24655286967754364, 0.04875296354293823, -0.5133509635925293, 0.01197219081223011, 0.06946779787540436, 0.1574360728263855, -0.079464390873909, 0.10217893868684769, 0.4472753703594208, -0.17720024287700653, 0.2422116994857788, 0.01544740330427885, -0.05776221677660942, -0.1007937416434288, 0.09602212905883789, -0.4450978636741638, -0.3861231207847595, 0.4605049788951874, 0.07121484726667404, -0.1043766513466835, 0.06537472456693649, 0.3186514377593994, -0.632595419883728, -0.2643201947212219, 0.3530021905899048, -0.08635822683572769, -0.0563836432993412, -0.20915178954601288, 0.08163759112358093, 0.04211152344942093, 0.6012598872184753, 0.183181494474411, 0.34985941648483276, -0.37034693360328674, -0.11620806902647018, -0.6050464510917664, -0.1977108269929886, 0.051924604922533035, -0.15896205604076385, -0.26225268840789795, 0.14968352019786835, -0.06800574064254761, 0.3262110948562622, -0.23237282037734985, 0.09375490993261337, 0.2632231116294861, 0.1062975749373436, -0.4595511853694916, 0.05649925395846367, 0.10483251512050629, 0.030725467950105667, 0.30466967821121216, -0.10057736188173294, -0.2276274859905243, 0.046522676944732666, -0.05575620010495186, -0.06158801540732384, -0.1410764753818512, 0.2911341190338135, 0.18577349185943604, 0.08580711483955383, 0.025451239198446274, 0.524669885635376, -0.0034387216437608004, -0.14572064578533173, -0.02578272670507431, 0.05499577894806862, -0.3764565885066986, -0.12250643968582153, -0.09452313929796219, 0.1602577418088913, -0.3817414939403534, 0.2723013758659363, 0.011710545048117638, -0.045299798250198364, 0.118520587682724, 0.002989842789247632, -0.19630953669548035, -0.030838776379823685, 0.06354264914989471, 0.23153437674045563, 0.17898160219192505, 0.7314291000366211, -0.02548316866159439, 0.15572886168956757, -0.07544959336519241, -0.03531152009963989, 0.05345000699162483, -0.13339778780937195, -0.2366468906402588, -0.2042829692363739, 0.06664790213108063, 0.0596994087100029, 0.1490628719329834, -0.6794455647468567, -0.3367588222026825, 0.033881280571222305, -0.0028483958449214697, 0.06076914444565773, 0.7086073756217957, 0.07913240045309067, -0.12908105552196503, -0.11463358998298645, -0.32713255286216736, -0.00030930357752367854, -0.2582996189594269, 0.20294028520584106, -0.3788250982761383 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
Good spot! Here are my thoughts: - Aside: Adding `MultitaskModel` to transformers might be a thing to raise - even though having task-specific heads has become unfashionable in recent times in favour of text-to-text type models. - Adding the task name as an extra field also seems useful for these kind of models which have task-specific heads - There is some validation of our approach that the user should be expected to `map` datasets into a common form. - The size-proportional sampling (also called "Examples-proportional mixing") used here doesn't perform too badly in the T5 paper (it's comparable to temperature-scaled mixing in many cases but less flexible. This is only reasonable with a `K` maximum size parameter to prevent very large datasets dominating). This might be good for a first prototype using: ```python def __iter__(self): """ For each batch, sample a task, and yield a batch from the respective task Dataloader. We use size-proportional sampling, but you could easily modify this to sample from some-other distribution. """ task_choice_list = [] for i, task_name in enumerate(self.task_name_list): task_choice_list += [i] * self.num_batches_dict[task_name] task_choice_list = np.array(task_choice_list) np.random.shuffle(task_choice_list) dataloader_iter_dict = { task_name: iter(dataloader) for task_name, dataloader in self.dataloader_dict.items() } for task_choice in task_choice_list: task_name = self.task_name_list[task_choice] yield next(dataloader_iter_dict[task_name]) ``` We'd just need to pull samples from the raw datasets and not from `DataLoader`s for each task. We can assume the user has done `dataset.shuffle()` if they want to. Other sampling methods can later be implemented by changing how the `task_choice_list` is generated. This should allow more flexibility and not tie us to specific methods for sampling among datasets.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
264
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. Good spot! Here are my thoughts: - Aside: Adding `MultitaskModel` to transformers might be a thing to raise - even though having task-specific heads has become unfashionable in recent times in favour of text-to-text type models. - Adding the task name as an extra field also seems useful for these kind of models which have task-specific heads - There is some validation of our approach that the user should be expected to `map` datasets into a common form. - The size-proportional sampling (also called "Examples-proportional mixing") used here doesn't perform too badly in the T5 paper (it's comparable to temperature-scaled mixing in many cases but less flexible. This is only reasonable with a `K` maximum size parameter to prevent very large datasets dominating). This might be good for a first prototype using: ```python def __iter__(self): """ For each batch, sample a task, and yield a batch from the respective task Dataloader. We use size-proportional sampling, but you could easily modify this to sample from some-other distribution. """ task_choice_list = [] for i, task_name in enumerate(self.task_name_list): task_choice_list += [i] * self.num_batches_dict[task_name] task_choice_list = np.array(task_choice_list) np.random.shuffle(task_choice_list) dataloader_iter_dict = { task_name: iter(dataloader) for task_name, dataloader in self.dataloader_dict.items() } for task_choice in task_choice_list: task_name = self.task_name_list[task_choice] yield next(dataloader_iter_dict[task_name]) ``` We'd just need to pull samples from the raw datasets and not from `DataLoader`s for each task. We can assume the user has done `dataset.shuffle()` if they want to. Other sampling methods can later be implemented by changing how the `task_choice_list` is generated. This should allow more flexibility and not tie us to specific methods for sampling among datasets.
[ -0.005269445478916168, -0.4564429223537445, -0.04090099409222603, -0.03315446898341179, -0.2421441227197647, 0.08998850733041763, 0.2262677401304245, 0.03379306569695473, 0.3455827236175537, -0.18324913084506989, -0.17462334036827087, 0.27823737263679504, -0.21713024377822876, 0.31477978825569153, 0.39542362093925476, -0.44745877385139465, 0.05297639220952988, -0.26878687739372253, -0.2423505336046219, 0.37680408358573914, 0.02382880076766014, 0.13764438033103943, -0.16504839062690735, 0.10277920216321945, -0.34901657700538635, -0.14932696521282196, -0.29104581475257874, 0.10549043118953705, 0.14364326000213623, 0.04943356290459633, 0.24418717622756958, 0.341758668422699, -0.031972698867321014, 0.19578063488006592, -0.00011775490565923974, -0.02630443684756756, 0.22959113121032715, -0.05339064076542854, 0.1380276083946228, -0.051430992782115936, 0.13814187049865723, -0.3776620030403137, 0.06772933900356293, -0.12424936890602112, 0.19982902705669403, 0.2635387182235718, 0.18150998651981354, 0.13065983355045319, 0.34446805715560913, -0.33991119265556335, 0.05876656249165535, 0.27942803502082825, -0.20459231734275818, 0.3041360080242157, -0.08194995671510696, 0.04751138761639595, 0.0006845130119472742, 0.24062460660934448, 0.6316851377487183, -0.6122652292251587, 0.00954695139080286, 0.1181459128856659, 0.020631887018680573, -0.07559328526258469, 0.213165283203125, -0.07582761347293854, -0.3081114590167999, -0.2063165009021759, -0.4016382396221161, 0.19827313721179962, -0.13543793559074402, -0.12603098154067993, -0.17508962750434875, -0.5726636052131653, -0.09653235971927643, 0.07819868624210358, -0.28447970747947693, 0.015646301209926605, -0.19182930886745453, -0.04815220832824707, -0.13264083862304688, -0.12773177027702332, -0.03177603706717491, 0.13176578283309937, 0.35107341408729553, 0.6222614049911499, 0.1479908674955368, 0.15523575246334076, 0.38221293687820435, 0.1009347066283226, -0.08788445591926575, -0.24048112332820892, 0.3139016032218933, 0.11480113118886948, -0.393947035074234, -0.36231350898742676, 0.20420101284980774, -0.2805655300617218, 0.229351207613945, 0.06943400949239731, 0.16647212207317352, 0.08447272330522537, -0.12309946119785309, 0.26120585203170776, 0.31291526556015015, 0.004651522729545832, -0.13495342433452606, -0.09768778085708618, 0.0522310733795166, -0.07296261191368103, 0.23544931411743164, 0.10745207220315933, 0.16393166780471802, -0.01664767973124981, -0.43312522768974304, -0.08163998275995255, -0.1350042074918747, 0.1605202853679657, -0.12132003903388977, -0.5605204105377197, -0.03214458376169205, -0.1420903354883194, 0.1316172182559967, -0.09903332591056824, -0.30580490827560425, 0.13915832340717316, -0.08490332961082458, 0.2681117057800293, -0.09449848532676697, 0.13290292024612427, -0.03960457444190979, 0.04850296303629875, -0.44950082898139954, -0.06490973383188248, 0.3281114101409912, 0.2135881930589676, 0.05287676304578781, 0.13712088763713837, -0.0014475133502855897, 0.07443277537822723, 0.3536284267902374, 0.05483192205429077, -0.01000108290463686, 0.04295583814382553, 0.07790736109018326, -0.309536874294281, -0.027146149426698685, 0.2130465805530548, -0.34669846296310425, -0.13569515943527222, -0.16489844024181366, -0.23564139008522034, 0.1343444585800171, -0.014896339736878872, -0.18980444967746735, -0.25104671716690063, -0.5263700485229492, 0.6592673659324646, -0.03056156635284424, -0.05707409605383873, -0.12782634794712067, 0.32526835799217224, -0.24029608070850372, -0.078280970454216, 0.1297525018453598, 0.04584088176488876, 0.03465897589921951, -0.3742164373397827, -0.09624514728784561, -0.0856848806142807, -0.040357571095228195, 0.30118224024772644, -0.18863286077976227, 0.2752714455127716, 0.09550350159406662, 0.13118794560432434, 0.21981237828731537, -0.4872385859489441, -0.030618591234087944, 0.2935410439968109, -0.026290977373719215, 0.3781125247478485, 0.07428216934204102, 0.24824465811252594, -0.049776095896959305, -0.18438908457756042, 0.25284427404403687, 0.9806958436965942, -0.3445674777030945, 0.007813118398189545, -0.12597215175628662, -0.3848104178905487, 0.5938200950622559, 0.35658639669418335, 0.1247871145606041, -0.28234636783599854, -0.13511605560779572, 0.08289189636707306, 0.0904683917760849, 0.11593067646026611, 0.07905580848455429, -0.09205992519855499, -0.32606279850006104, 0.023914769291877747, -0.16235782206058502, -0.11649902164936066, -0.4359433650970459, 0.010673446580767632, -0.029215633869171143, 0.3597773611545563, 0.21428373456001282, -0.10928677767515182, 0.27148211002349854, -0.2791426181793213, -0.1207624077796936, -0.3025939464569092, 0.0706900805234909, -0.12522879242897034, -0.033662788569927216, -0.28421708941459656, -0.014253886416554451, 0.287822425365448, 0.17413753271102905, -0.09134731441736221, -0.462568461894989, 0.3106149733066559, -0.11350930482149124, 0.07947522401809692, -0.07837386429309845, 0.6710867285728455, -0.13304321467876434, -0.05170876532793045, 0.18674446642398834, 0.25455960631370544, -0.30537280440330505, 0.26025357842445374, 0.31741344928741455, 0.3043530583381653, 0.1962374746799469, -0.05067864805459976, 0.08027470111846924, 0.07100693136453629, -0.006333031691610813, -0.25429004430770874, -0.16883835196495056, 0.45542821288108826, -0.1798308938741684, 0.4642225503921509, -0.07356186956167221, -0.006369295064359903, -0.20803220570087433, -0.04436594247817993, -0.2130458950996399, 0.44610440731048584, 0.2914744019508362, -0.1696995198726654, 0.21452751755714417, -0.10940814763307571, -0.35217341780662537, 0.1489396095275879, 0.1390201598405838, 0.031161176040768623, 0.03513563051819801, 0.012584545649588108, 0.006687578279525042, -0.2945311665534973, -0.11442459374666214, 0.29003700613975525, 0.5758757591247559, 0.21885429322719574, 0.13711343705654144, 0.017956335097551346, -0.2088221311569214, -0.09033544361591339, 0.013694005087018013, 0.2752796411514282, 0.01692301221191883, 0.19390881061553955, 0.015641700476408005, 0.064555324614048, 0.042457692325115204, -0.20733022689819336, 0.15761017799377441, -0.20261071622371674, -0.059820372611284256, 0.03191610053181648, -0.18620608747005463, -0.5582323670387268, -0.2554794251918793, -0.010924187488853931, -0.36712130904197693, -0.2729381322860718, 0.15407246351242065, 0.043434176594018936, -0.16207271814346313, 0.1376047432422638, 0.24740619957447052, 0.6375622153282166, -0.46454888582229614, -0.2303403615951538, -0.14063890278339386, -0.3725031018257141, -0.07144709676504135, 0.07383917272090912, 0.4597984850406647, 0.28610607981681824, 0.38312146067619324, 0.33520135283470154, -0.07496794313192368, 0.1093757152557373, -0.4088931381702423, 0.010390521958470345, -0.2810269296169281, 0.24833235144615173, 0.08054898679256439, -0.21447741985321045, 0.17429642379283905, -0.3981611430644989, -0.06309691816568375, -0.027325574308633804, -0.0667390376329422, -0.004362516570836306, -0.006471315864473581, 0.08041540533304214, -0.1972562074661255, -0.1686614751815796, -0.6268010139465332, -0.4757329523563385, 0.3247363269329071, -0.2877424359321594, 0.06258340179920197, 0.09646813571453094, -0.27044111490249634, 0.12282247096300125, -0.07755067944526672, -0.006085993722081184, -0.12584719061851501, -0.042074743658304214, 0.027647005394101143, -0.33682018518447876, -0.14505459368228912, -0.37486064434051514, -0.09668385237455368, 0.2254825383424759, -0.07548344135284424, -0.014169519767165184, -0.14891423285007477, 0.10138937830924988, -0.1627759039402008, 0.14726229012012482, 0.1359441578388214, 0.17427416145801544, -0.20071294903755188, 0.1626322716474533, -0.01870718225836754, -0.32956239581108093, 0.22921361029148102, 0.3730611205101013, 0.45728474855422974, 0.17798136174678802, 0.054322630167007446, 0.16411525011062622, 0.7202289700508118, 0.2120823860168457, 0.11296989023685455, 0.1207045465707779, 0.33972135186195374, -0.09869548678398132, 0.1656608134508133, -0.2394079566001892, 0.2550591230392456, -0.17682728171348572, 0.4767408072948456, 0.04971028119325638, 0.2018960416316986, -0.2010156363248825, -0.2819044888019562, -0.030092494562268257, -0.2980363070964813, -0.2182210236787796, 0.267738401889801, -0.20839963853359222, 0.04012451693415642, -0.07382522523403168, -0.4009004831314087, -0.28047361969947815, -0.06336969882249832, 0.05634457990527153, 0.4238349199295044, -0.057890329509973526, 0.1724991351366043, -0.33453837037086487, -0.06981971859931946, -0.4045555293560028, 0.05911208689212799, 0.02873910777270794, 0.23816999793052673, -0.09146839380264282, -0.06206634268164635, -0.06574030220508575, 0.16588762402534485, 0.649579644203186, -0.4798402786254883, -0.10412192344665527, 0.08365976065397263, 0.0829395279288292, -0.04358312487602234, -0.14425024390220642, -0.24003955721855164, -0.0459776297211647, 0.4381369948387146, 0.40390917658805847, -0.3940841853618622, -0.2811397910118103, 0.19087934494018555, -0.2568363845348358, -0.12578079104423523, -0.23165123164653778, -0.21141789853572845, -0.17475780844688416, -0.14961399137973785, 0.05115014687180519, 0.2645754814147949, 0.21153362095355988, -0.10403922200202942, 0.1365634649991989, -0.01972057670354843, 0.2384251356124878, 0.21176375448703766, -0.014664326794445515, -0.09972887486219406, 0.24854622781276703, 0.12893997132778168, 0.03632725402712822, -0.16067320108413696, -0.34276074171066284, 0.07137423008680344, 0.04164772108197212, 0.13564151525497437, 0.2838734984397888, -0.008410248905420303, 0.6613308787345886, 0.07134031504392624, 0.38200563192367554, 0.29507148265838623, -0.03617784008383751, -0.0069733248092234135, -0.2792738080024719, 0.13812726736068726, 0.18902549147605896, 0.243909552693367, -0.19070561230182648, -0.17326904833316803, 0.5538422465324402, 0.2548622190952301, -0.10853856056928635, 0.38894709944725037, -0.5794231295585632, -0.284702867269516, 0.28209570050239563, 0.0664711445569992, 0.7568147778511047, -0.022754650563001633, 0.2946831285953522, 0.051072463393211365, -0.4602009356021881, 0.5372833013534546, -0.4241020083427429, 0.15177753567695618, -0.0489807091653347, -0.1196800023317337, -0.010852711275219917, -0.10564876347780228, -0.10179876536130905, 0.30794018507003784, -0.33810093998908997, 0.30630987882614136, 0.49197518825531006, -0.23215728998184204, -0.07935850322246552, 0.5174072980880737, 0.09096954762935638, -0.38823652267456055, 0.00893829483538866, 0.008923307992517948, -0.25484153628349304, 0.16929379105567932, -0.1970807909965515, -0.29695209860801697, 0.19138267636299133, 0.12744583189487457, -0.29909488558769226, 0.08042915910482407, 0.08365561068058014, 0.2613985240459442, 0.3219776153564453, 0.0013861366314813495, 0.30562058091163635, -0.21544784307479858, -0.08355750888586044, -0.18756048381328583, -0.10171805322170258, 0.16740112006664276, -0.22281645238399506, 0.07013482600450516, 0.04552379623055458, -0.18126125633716583, 0.5122030973434448, -0.077839195728302, -0.12756991386413574, -0.27518799901008606, 0.25064849853515625, -0.45339521765708923, -0.2337384670972824, 0.14939169585704803, 0.14604708552360535, 0.23558323085308075, -0.3322305679321289, 0.5851313471794128, 0.08234444260597229, -0.005585158709436655, 0.01226531807333231, 0.18602275848388672, -0.20392675697803497, 0.5055802464485168, -0.3174494206905365, 0.04855186119675636, 0.08217503875494003, -0.043954454362392426, -0.11986556649208069, -0.40590423345565796, -0.09712442010641098, 0.35435613989830017, 0.09507549554109573, 0.025454746559262276, -0.0568874292075634, 0.13590510189533234, -0.12698441743850708, -0.05918358266353607, -0.13868758082389832, -0.3926321864128113, 0.10445529967546463, 0.21058261394500732, 0.20176346600055695, 0.16465343534946442, -0.07526330649852753, -0.2580512762069702, -0.2193695306777954, 0.515089213848114, -0.026738697662949562, 0.20104695856571198, -0.16419096291065216, 0.3756382465362549, 0.42429491877555847, 0.5418840050697327, -0.20867730677127838, 0.06487943232059479, 0.061779361218214035, 0.0810442715883255, 0.3777204751968384, -0.2797590494155884, 0.04444734379649162, 0.22707724571228027, -0.18765446543693542, 0.006583585403859615, -0.27863746881484985, -0.10919912159442902, -0.2729393243789673, 0.13855066895484924, 0.14789046347141266, 0.035220373421907425, -0.0034778721164911985, -0.0839105173945427, -0.045493774116039276, -0.01851394958794117, -0.061204295605421066, -0.3870375454425812, -0.026049574837088585, 0.18790346384048462, -0.043608658015728, 0.055761706084012985, 0.01792800985276699, 0.2448154240846634, 0.042888104915618896, -0.06855159252882004, 0.3669426739215851, 0.09157957881689072, 0.1558609902858734, -0.14030097424983978, -0.24093985557556152, 0.020428678020834923, 0.23353491723537445, 0.13942976295948029, -0.009182004258036613, 0.1907387673854828, -0.23085424304008484, -0.13711199164390564, 0.16963434219360352, 0.12699483335018158, 0.0751056894659996, -0.106068454682827, 0.0985284149646759, 0.07580416649580002, -0.11297909915447235, -0.023444797843694687, 0.21591253578662872, 0.09666827321052551, 0.04326412454247475, 0.11526913195848465, 0.1418038010597229, 0.49411433935165405, -0.26620662212371826, 0.08396892249584198, 0.1088300570845604, -0.3563462495803833, 0.13451845943927765, -0.15543581545352936, -0.02022460103034973, 0.029280539602041245, 0.15330460667610168, -0.01826273277401924, 0.11684251576662064, -0.35832133889198303, 0.38128581643104553, 0.22113105654716492, -0.003973303362727165, 0.031030472368001938, 0.2010093778371811, 0.05039365217089653, -0.23988553881645203, 0.2065342515707016, 0.027465827763080597, 0.28734952211380005, 0.06632690131664276, 0.8376122117042542, -0.21268734335899353, -0.38412338495254517, 0.1474454700946808, 0.4088921844959259, 0.0464998222887516, -0.2087956815958023, 0.16752369701862335, 0.1362491101026535, -0.37386593222618103, 0.06661008298397064, 0.04919515177607536, -0.1454668790102005, 0.6858698129653931, -0.10418234020471573, 0.11888349056243896, -0.07461710274219513, -0.2131347954273224, 0.38438981771469116, 0.16913048923015594, -0.05465083569288254, -0.06694145500659943, 0.13789913058280945, -0.11123886704444885, -0.1763106733560562, 0.49055102467536926, 0.20801778137683868, 0.009057878516614437, -0.4584711492061615, 0.020228570327162743, 0.12952661514282227, 0.13868196308612823, -0.05805066600441933, 0.17673447728157043, 0.47014498710632324, -0.07212276756763458, 0.15682053565979004, -0.033703774213790894, -0.010842431336641312, -0.12366782873868942, 0.042208701372146606, -0.41784200072288513, -0.30987945199012756, 0.5140650868415833, 0.09148689359426498, -0.17002157866954803, 0.09727980196475983, 0.34294167160987854, -0.6166447997093201, -0.25685393810272217, 0.2994372546672821, -0.11552070081233978, -0.09028048813343048, -0.23393985629081726, 0.04657304286956787, 0.10411223024129868, 0.6218815445899963, 0.14806951582431793, 0.3906711935997009, -0.3981970548629761, -0.10698532313108444, -0.5736935138702393, -0.23161567747592926, 0.1399529129266739, -0.051708731800317764, -0.1787453442811966, 0.09441910684108734, -0.06492859870195389, 0.22461514174938202, -0.2561459243297577, -0.01942722126841545, 0.213270902633667, 0.09001750499010086, -0.4094817042350769, 0.16877418756484985, 0.1896282285451889, 0.06828110665082932, 0.3064979612827301, -0.032203949987888336, -0.15833881497383118, 0.20729246735572815, -0.13092076778411865, -0.1387365758419037, -0.08773770928382874, 0.24361592531204224, 0.18980973958969116, 0.04684310406446457, 0.021143220365047455, 0.4844200015068054, -0.004027991555631161, -0.2120717614889145, 0.010847182013094425, -0.0026888109277933836, -0.325647234916687, -0.17872512340545654, -0.07713478803634644, 0.20804078876972198, -0.36006852984428406, 0.29446494579315186, -0.009957379661500454, -0.13632415235042572, 0.023876400664448738, -0.008792412467300892, -0.2192276418209076, -0.025523066520690918, 0.03367040678858757, 0.23170892894268036, 0.24931339919567108, 0.6675163507461548, 0.016650671139359474, 0.093211829662323, -0.08999025821685791, -0.050489433109760284, 0.026074953377246857, -0.015022613108158112, -0.24134965240955353, -0.14494356513023376, 0.0330267958343029, 0.014686851762235165, 0.18479889631271362, -0.7851418852806091, -0.3945769667625427, -0.021979020908474922, 0.031411588191986084, 0.08482035249471664, 0.7013658881187439, 0.1230233684182167, -0.17507082223892212, -0.09732472151517868, -0.3418373763561249, 0.020175324752926826, -0.22876331210136414, 0.17516614496707916, -0.36070144176483154 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
Another thought: Multitasking over benchmarks (represented as Meta-datasets in nlp) is probably a common use case. Would be nice to pass an entire benchmark to our `MultiDataset` wrapper rather than having to pass individual components.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
35
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. Another thought: Multitasking over benchmarks (represented as Meta-datasets in nlp) is probably a common use case. Would be nice to pass an entire benchmark to our `MultiDataset` wrapper rather than having to pass individual components.
[ -0.06561174988746643, -0.37235522270202637, -0.06819324940443039, -0.04926852136850357, -0.28527209162712097, 0.1337665617465973, 0.21648046374320984, 0.12065233290195465, 0.3021678328514099, -0.14211761951446533, -0.1687692254781723, 0.3203652799129486, -0.16379931569099426, 0.348694384098053, 0.4185856580734253, -0.41218018531799316, 0.031688325107097626, -0.21076597273349762, -0.21456031501293182, 0.4246898889541626, 0.06257869303226471, 0.08338536322116852, -0.16562999784946442, 0.10172665864229202, -0.29676759243011475, -0.18775197863578796, -0.3303470015525818, 0.12289752066135406, 0.18027456104755402, -0.006745833437889814, 0.21674251556396484, 0.3386637270450592, -0.005248019937425852, 0.07626061886548996, -0.00011275352881057188, -0.07976435124874115, 0.21804840862751007, -0.05337457358837128, 0.10312952846288681, -0.013962449505925179, 0.1930619180202484, -0.3639012575149536, 0.04667546972632408, -0.10680344700813293, 0.1687580645084381, 0.19356893002986908, 0.17223447561264038, -0.0764036476612091, 0.34126436710357666, -0.25450170040130615, 0.10234137624502182, 0.31566810607910156, -0.23991118371486664, 0.2976696491241455, -0.11138324439525604, -0.08681972324848175, 0.011975322850048542, 0.24389155209064484, 0.6718506217002869, -0.627389669418335, -0.06007206812500954, 0.057242199778556824, -0.032386068254709244, 0.04977758973836899, 0.09267529100179672, -0.13015539944171906, -0.3021050989627838, -0.2108955681324005, -0.3377222716808319, 0.25109902024269104, -0.15973052382469177, -0.13706067204475403, -0.22889825701713562, -0.5552752614021301, -0.013058805838227272, 0.05129188671708107, -0.30808937549591064, 0.03183281794190407, -0.17259611189365387, -0.006247504614293575, -0.10114488005638123, -0.12405438721179962, 0.05755409598350525, 0.0890997052192688, 0.3357069194316864, 0.5584859848022461, 0.120794378221035, 0.17407916486263275, 0.3698946237564087, 0.07315339148044586, -0.08478503674268723, -0.25911855697631836, 0.2741997539997101, 0.1001642718911171, -0.4158080220222473, -0.3103587329387665, 0.13953252136707306, -0.23301154375076294, 0.3023063540458679, 0.09700202941894531, 0.13515329360961914, 0.11098421365022659, -0.12362248450517654, 0.273695170879364, 0.340285062789917, 0.03425595164299011, -0.07769939303398132, -0.16922681033611298, 0.010786185972392559, -0.14165501296520233, 0.18027468025684357, 0.11355353146791458, 0.2074793130159378, 0.025804204866290092, -0.4392688572406769, -0.13691210746765137, -0.22847962379455566, 0.16348838806152344, -0.2595623731613159, -0.568891167640686, -0.11211668699979782, -0.09568798542022705, 0.06208180636167526, -0.13887612521648407, -0.26140186190605164, 0.13918156921863556, -0.0921638086438179, 0.2695342004299164, -0.068855419754982, 0.006288500968366861, -0.05109172686934471, 0.0879676416516304, -0.45554277300834656, -0.010129055008292198, 0.33236202597618103, 0.10064975172281265, 0.08620436489582062, 0.1457320749759674, -0.019783884286880493, 0.062232453376054764, 0.3870140314102173, 0.011977208778262138, -0.0044589294120669365, -0.012135263532400131, 0.18673965334892273, -0.31405556201934814, 0.029218709096312523, 0.18315273523330688, -0.33685457706451416, -0.0813511461019516, -0.1455327272415161, -0.1486731618642807, 0.14257434010505676, 0.032133061438798904, -0.22536812722682953, -0.2860386073589325, -0.5482081174850464, 0.6673360466957092, -0.011244986206293106, -0.0629957839846611, -0.1631249636411667, 0.3023529648780823, -0.30237749218940735, -0.09309008717536926, 0.17095202207565308, 0.013164771720767021, 0.021747848019003868, -0.27748388051986694, -0.1617572009563446, -0.07043283432722092, -0.10190138965845108, 0.31535136699676514, -0.19219155609607697, 0.30844148993492126, 0.04201120510697365, 0.17138834297657013, 0.23466965556144714, -0.4846600890159607, -0.03291591256856918, 0.3313530683517456, -0.07133078575134277, 0.4018954336643219, 0.05024728178977966, 0.22615347802639008, 0.05564466491341591, -0.17824988067150116, 0.23286080360412598, 0.9699894189834595, -0.38806721568107605, 0.009526207111775875, -0.08592972904443741, -0.39205941557884216, 0.543384313583374, 0.3569904863834381, 0.05959482118487358, -0.24156570434570312, -0.14711597561836243, -0.012278212234377861, 0.08712279796600342, 0.1098291352391243, 0.09951669722795486, -0.07516547292470932, -0.4207242727279663, 0.023374594748020172, -0.2236107438802719, -0.1986401230096817, -0.43184441328048706, 0.01817844994366169, 0.021139616146683693, 0.22960275411605835, 0.3293685019016266, -0.13701984286308289, 0.24782653152942657, -0.27949923276901245, -0.04655004292726517, -0.37557780742645264, 0.12504081428050995, -0.051529593765735626, 0.00891288835555315, -0.20238712430000305, -0.04816122353076935, 0.35677435994148254, 0.17522454261779785, -0.12952175736427307, -0.39121297001838684, 0.3071177899837494, -0.05294416472315788, 0.03812476247549057, -0.04055449739098549, 0.6141854524612427, -0.21523283421993256, -0.057151127606630325, 0.18919044733047485, 0.24514171481132507, -0.2727408707141876, 0.2884795665740967, 0.35366785526275635, 0.2753917872905731, 0.19267569482326508, -0.035916704684495926, 0.05417722091078758, 0.06028116121888161, -0.03294999524950981, -0.2539367973804474, -0.051722876727581024, 0.49388599395751953, -0.12789475917816162, 0.46705785393714905, -0.06947580724954605, 0.05351877212524414, -0.14168433845043182, -0.03650076687335968, -0.1959547996520996, 0.4411691427230835, 0.29005083441734314, -0.21228156983852386, 0.3862041234970093, -0.10933732986450195, -0.31891047954559326, 0.17448939383029938, 0.13803315162658691, 0.05124538391828537, 0.06468228250741959, -0.00907236710190773, -0.03657002002000809, -0.2727172076702118, -0.07160243391990662, 0.37076741456985474, 0.6577776670455933, 0.2386052906513214, 0.17744724452495575, 0.01481762994080782, -0.229314386844635, -0.08080202341079712, -0.031207330524921417, 0.19398164749145508, 0.026520121842622757, 0.0988892987370491, 0.013474198058247566, 0.02231714315712452, 0.048615165054798126, -0.14007893204689026, 0.15098941326141357, -0.16983912885189056, -0.009785136207938194, 0.041374027729034424, -0.22050830721855164, -0.5566989779472351, -0.21707621216773987, 0.0291738361120224, -0.3283436894416809, -0.2532232999801636, 0.1027173399925232, 0.12225008010864258, -0.13139201700687408, 0.15123379230499268, 0.23217691481113434, 0.6160982847213745, -0.4460153877735138, -0.26886478066444397, -0.08353522419929504, -0.300821989774704, -0.1132342666387558, 0.12900172173976898, 0.45876023173332214, 0.3014472424983978, 0.38529637455940247, 0.3329300582408905, -0.009069367311894894, -0.021664466708898544, -0.3423925042152405, -0.038963720202445984, -0.25041428208351135, 0.2589895725250244, 0.03648321330547333, -0.24169927835464478, 0.1447736769914627, -0.38711991906166077, -0.07622957229614258, -0.07680220901966095, -0.07060439139604568, -0.08701073378324509, 0.03128569945693016, 0.14604739844799042, -0.185271218419075, -0.15247845649719238, -0.585279107093811, -0.47434401512145996, 0.2974986433982849, -0.2947559654712677, 0.05387405306100845, -0.10809881240129471, -0.3133614659309387, 0.1933501660823822, -0.12611332535743713, -0.01943228952586651, -0.13752536475658417, -0.09073184430599213, -0.05549491196870804, -0.3396887183189392, -0.15744979679584503, -0.39757734537124634, -0.08305011689662933, 0.2783394455909729, -0.049361784011125565, -0.04247894510626793, -0.15597358345985413, 0.13113895058631897, -0.15483514964580536, 0.060068223625421524, 0.09263822436332703, 0.17417538166046143, -0.2363424152135849, 0.1292828619480133, -0.028270643204450607, -0.2620979845523834, 0.236829936504364, 0.2830857038497925, 0.37461957335472107, 0.17052774131298065, 0.022373829036951065, 0.2380325198173523, 0.701531171798706, 0.2394607663154602, 0.054874494671821594, 0.16642321646213531, 0.2817082107067108, -0.023641221225261688, 0.11926227062940598, -0.21993127465248108, 0.15194275975227356, -0.18846170604228973, 0.4781728982925415, 0.10397280007600784, 0.15760914981365204, -0.25659728050231934, -0.2707083225250244, 0.030139975249767303, -0.2531310021877289, -0.1704380214214325, 0.2132001519203186, -0.31595462560653687, 0.033423494547605515, -0.06611885875463486, -0.39819571375846863, -0.22175024449825287, -0.018919363617897034, -0.014532617293298244, 0.288003534078598, -0.042994070798158646, 0.1606457680463791, -0.33520305156707764, -0.031863775104284286, -0.4085689187049866, 0.11785263568162918, 0.09698854386806488, 0.33181220293045044, -0.08659715205430984, -0.0450323224067688, -0.08494308590888977, 0.11186149716377258, 0.552223265171051, -0.4672248065471649, -0.09274476021528244, 0.10849882662296295, 0.06989079713821411, -0.09652066975831985, -0.08832426369190216, -0.22059988975524902, 0.045805998146533966, 0.4289891719818115, 0.36070746183395386, -0.2990396022796631, -0.28864550590515137, 0.15508979558944702, -0.21957744657993317, -0.048654887825250626, -0.1847693920135498, -0.21510864794254303, -0.1798703819513321, -0.1340259164571762, 0.06568842381238937, 0.2318376898765564, 0.21855074167251587, -0.08076146990060806, 0.15293222665786743, -0.11033478379249573, 0.15788602828979492, 0.1860542595386505, -0.04371156543493271, -0.09453758597373962, 0.15243034064769745, 0.20288020372390747, 0.013335631228983402, -0.15375544130802155, -0.30702725052833557, 0.021048402413725853, 0.013778520748019218, 0.1322283297777176, 0.24650128185749054, -0.017967589199543, 0.6386029124259949, -0.04344050958752632, 0.307864785194397, 0.3434130549430847, -0.15040166676044464, -0.0601094625890255, -0.2892884910106659, 0.16420325636863708, 0.18684616684913635, 0.25749412178993225, -0.12326885014772415, -0.18717561662197113, 0.572580099105835, 0.33587220311164856, -0.06958118081092834, 0.45397359132766724, -0.5722924470901489, -0.31260812282562256, 0.24356500804424286, 0.0592392161488533, 0.577468991279602, -0.08398009091615677, 0.26166024804115295, -0.019968749955296516, -0.4101733863353729, 0.5315446853637695, -0.3737443685531616, 0.153408020734787, -0.0518803596496582, -0.24385517835617065, 0.020600950345396996, -0.08406249433755875, -0.007088113110512495, 0.3127880394458771, -0.32291707396507263, 0.3246549665927887, 0.4892112910747528, -0.2282962203025818, -0.0012170264963060617, 0.5655282735824585, 0.025864817202091217, -0.37018048763275146, 0.016942718997597694, 0.06858373433351517, -0.29029569029808044, 0.08191970735788345, -0.15320679545402527, -0.21072736382484436, 0.2203502207994461, 0.1797533482313156, -0.3520199656486511, 0.058837831020355225, 0.02680482342839241, 0.21443995833396912, 0.20536483824253082, 0.035367924720048904, 0.28780093789100647, -0.17419666051864624, -0.06097088009119034, -0.15786224603652954, -0.08689096570014954, 0.12871581315994263, -0.19305986166000366, 0.03640541061758995, 0.09449757635593414, -0.20959539711475372, 0.4564695358276367, -0.06033085286617279, -0.1451978236436844, -0.22870008647441864, 0.27023670077323914, -0.5092930793762207, -0.14957062900066376, 0.1981714367866516, 0.1473064422607422, 0.28635239601135254, -0.29994669556617737, 0.5590229630470276, 0.10002853721380234, -0.015114538371562958, 0.041535742580890656, 0.19969813525676727, -0.23633138835430145, 0.5029979944229126, -0.3028038740158081, -0.015230927616357803, 0.06073099747300148, 0.056797947734594345, -0.14584985375404358, -0.3981849253177643, -0.07032763212919235, 0.2884267568588257, 0.14057889580726624, -0.0045674690045416355, 0.05541439354419708, 0.09924281388521194, 0.05065260827541351, 0.019722815603017807, -0.200583353638649, -0.32429245114326477, 0.1335984468460083, 0.1471765637397766, 0.203813374042511, 0.22745738923549652, -0.09523485600948334, -0.30265945196151733, -0.19763019680976868, 0.4594477713108063, -0.029068946838378906, 0.1782679408788681, -0.1285366415977478, 0.34765616059303284, 0.4074634909629822, 0.5269008874893188, -0.24591681361198425, 0.11764306575059891, 0.04560895264148712, 0.1269553154706955, 0.5051094889640808, -0.21541810035705566, -0.004284779541194439, 0.3125101625919342, -0.13643251359462738, 0.06286756694316864, -0.26634275913238525, -0.16179129481315613, -0.1922886073589325, 0.10798437893390656, 0.12295644730329514, 0.025694487616419792, 0.02970549650490284, -0.1108931377530098, -0.09539131820201874, -0.04093518853187561, -0.05580670386552811, -0.3989925980567932, -0.061246778815984726, 0.09844289720058441, -0.0674811378121376, 0.07659336924552917, 0.024724137037992477, 0.2167937010526657, -0.03503909334540367, -0.12325549125671387, 0.24483513832092285, 0.1176038384437561, 0.1735721081495285, -0.15246327221393585, -0.16086594760417938, -0.036695703864097595, 0.22497326135635376, 0.14430338144302368, -0.05042271316051483, 0.19050471484661102, -0.2337227463722229, -0.07541338354349136, 0.24634672701358795, 0.10478583723306656, 0.09123043715953827, -0.08847726881504059, 0.1727253496646881, 0.10370089113712311, -0.0844673290848732, -0.020609624683856964, 0.23100921511650085, 0.059598084539175034, 0.11645391583442688, 0.13184809684753418, 0.19638805091381073, 0.46792417764663696, -0.18571552634239197, -0.0031029274687170982, 0.11268236488103867, -0.38476085662841797, 0.12622179090976715, -0.18793395161628723, -0.07245881110429764, -0.0286669060587883, 0.13434544205665588, -0.017931552603840828, 0.13674604892730713, -0.2621411383152008, 0.33800071477890015, 0.388969749212265, 0.0716753602027893, -0.007467829622328281, 0.22169528901576996, 0.06501366943120956, -0.1531662940979004, 0.25695472955703735, 0.05426430702209473, 0.30022236704826355, 0.012246371246874332, 0.83601975440979, -0.3240251839160919, -0.4433019757270813, 0.11316070705652237, 0.3656110167503357, 0.051836058497428894, -0.21638809144496918, 0.2377728819847107, 0.13523784279823303, -0.336312860250473, 0.08761145174503326, 0.06379051506519318, -0.18096208572387695, 0.7053232192993164, -0.13340464234352112, 0.1490318477153778, -0.08333388715982437, -0.2459763139486313, 0.36664894223213196, 0.11514912545681, -0.07119155675172806, -0.0004825464857276529, 0.057085711508989334, -0.11510665714740753, -0.17619846761226654, 0.5325781106948853, 0.18094295263290405, 0.018203645944595337, -0.44259923696517944, 0.010071126744151115, 0.014930172823369503, 0.1756376475095749, -0.07289597392082214, 0.17742693424224854, 0.44115424156188965, -0.18200452625751495, 0.24563774466514587, 0.0059405057691037655, -0.046177953481674194, -0.1398252695798874, 0.04684950411319733, -0.4108142852783203, -0.4485563337802887, 0.4289400577545166, 0.07663799822330475, -0.12908443808555603, 0.08759141713380814, 0.36286309361457825, -0.5635625123977661, -0.25134116411209106, 0.3186551630496979, -0.2019166350364685, -0.13163213431835175, -0.20967769622802734, 0.08235731720924377, 0.04996441677212715, 0.6422712206840515, 0.20109859108924866, 0.3915330171585083, -0.35836318135261536, -0.1792929619550705, -0.5960267782211304, -0.1786126345396042, 0.08641727268695831, -0.08518611639738083, -0.2282562553882599, 0.15490640699863434, -0.11009465157985687, 0.33917391300201416, -0.17584434151649475, 0.07582426816225052, 0.2411528080701828, 0.00003824974555755034, -0.4148184359073639, 0.0921235978603363, 0.12537753582000732, -0.017391137778759003, 0.2854139506816864, -0.10768546164035797, -0.144165500998497, 0.12176575511693954, -0.07111327350139618, -0.07582549005746841, -0.11642738431692123, 0.2526828646659851, 0.15655504167079926, 0.10558995604515076, 0.010083803907036781, 0.45129111409187317, 0.01785186491906643, -0.1485956460237503, -0.05152207240462303, 0.03749065473675728, -0.35272181034088135, -0.18147192895412445, -0.1022651344537735, 0.14909912645816803, -0.4346381425857544, 0.29576921463012695, 0.021366244181990623, -0.061808113008737564, 0.1016726940870285, 0.005677744746208191, -0.25479835271835327, -0.05856207385659218, 0.03143671900033951, 0.19942878186702728, 0.2543525993824005, 0.7323930859565735, -0.002326537389308214, 0.1040789857506752, -0.05692066624760628, 0.030946198850870132, -0.02811802551150322, -0.08862321823835373, -0.2501007914543152, -0.13463057577610016, 0.05298857390880585, 0.07657445967197418, 0.08411630243062973, -0.6738635301589966, -0.4056534767150879, 0.0536898598074913, 0.022415820509195328, 0.0578993521630764, 0.7152572274208069, 0.027555424720048904, -0.12574444711208344, -0.1139773279428482, -0.38882318139076233, 0.008580038323998451, -0.2279847264289856, 0.21922355890274048, -0.4018602967262268 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
Here's a fully working implementation based on the `__iter__` function of @zphang. - I've generated the task choice list in the constructor as it allows us to index into the MultiDataset just like a normal dataset. I'm changing `task_choice_list` into a list of `(dataset_idx, example_idx)` so each entry references a unique dataset example. The shuffling has to be done before this as we don't want to shuffle within each task (we assume this is done by the user if this is what they intend). - I'm slightly concerned this list could become very large if many large datasets were used. Can't see a way round it at the moment though. - I've used `task.info.builder_name` as the dataset name. Not sure if this is correct. - I'd love to add some of the other `Dataset` methods (map, slicing by column, etc...). Would be great to implement the whole interface so a single dataset can be simply replaced by this. - This does everything on the individual example-level. If some application required batches all from a single task in turn we can't really do that. ```python import nlp import numpy as np class MultiDataset: def __init__(self,tasks): self.tasks = tasks # Create random order of tasks # Using size-proportional sampling task_choice_list = [] for i, task in enumerate(self.tasks): task_choice_list += [i] * len(task) task_choice_list = np.array(task_choice_list) np.random.shuffle(task_choice_list) # Add index into each dataset # - We don't want to shuffle within each task counters = {} self.task_choice_list = [] for i in range(len(task_choice_list)): idx = counters.get(task_choice_list[i],0) self.task_choice_list.append((task_choice_list[i],idx)) counters[task_choice_list[i]] = idx + 1 def __len__(self): return np.sum([len(t) for t in self.tasks]) def __repr__(self): task_str = ", ".join([str(t) for t in self.tasks]) return f"MultiDataset(tasks: {task_str})" def __getitem__(self,key): if isinstance(key, int): task_idx, example_idx = self.task_choice_list[key] task = self.tasks[task_idx] example = task[example_idx] example["task_name"] = task.info.builder_name return example elif isinstance(key, slice): raise NotImplementedError() def __iter__(self): for i in range(len(self)): yield self[i] def load_multitask(*datasets): '''Create multitask datasets per split''' def _get_common_splits(datasets): '''Finds the common splits present in all self.datasets''' min_set = None for dataset in datasets: if min_set != None: min_set.intersection(set(dataset.keys())) else: min_set = set(dataset.keys()) return min_set common_splits = _get_common_splits(datasets) out = {} for split in common_splits: out[split] = MultiDataset([d[split] for d in datasets]) return out ########################################## # Dataset Flattening def flatten(dataset,flatten_fn): for k in dataset.keys(): if isinstance(dataset[k],nlp.Dataset): dataset[k] = dataset[k].map(flatten_fn,remove_columns=dataset[k].column_names) # Squad def flatten_squad(example): return {"source": "squad context: " + example['context'] + " question: " + example['question'], "target":example["answers"]["text"]} squad = nlp.load_dataset("squad") flatten(squad,flatten_squad) # CNN_DM def flatten_cnn_dm(example): return {"source": "cnn_dm: " + example['article'],"target":[example["highlights"]]} cnn_dm = nlp.load_dataset("cnn_dailymail", "3.0.0") flatten(cnn_dm,flatten_cnn_dm) ############################################# mtds = load_multitask(squad,cnn_dm) for example in mtds["train"]: print(example["task_name"],example["target"]) ``` Let me know if you have any thoughts. I've started using this in some of my projects and it seems to work. If people are happy with the general approach for a first version, I can make a pull request.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
469
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. Here's a fully working implementation based on the `__iter__` function of @zphang. - I've generated the task choice list in the constructor as it allows us to index into the MultiDataset just like a normal dataset. I'm changing `task_choice_list` into a list of `(dataset_idx, example_idx)` so each entry references a unique dataset example. The shuffling has to be done before this as we don't want to shuffle within each task (we assume this is done by the user if this is what they intend). - I'm slightly concerned this list could become very large if many large datasets were used. Can't see a way round it at the moment though. - I've used `task.info.builder_name` as the dataset name. Not sure if this is correct. - I'd love to add some of the other `Dataset` methods (map, slicing by column, etc...). Would be great to implement the whole interface so a single dataset can be simply replaced by this. - This does everything on the individual example-level. If some application required batches all from a single task in turn we can't really do that. ```python import nlp import numpy as np class MultiDataset: def __init__(self,tasks): self.tasks = tasks # Create random order of tasks # Using size-proportional sampling task_choice_list = [] for i, task in enumerate(self.tasks): task_choice_list += [i] * len(task) task_choice_list = np.array(task_choice_list) np.random.shuffle(task_choice_list) # Add index into each dataset # - We don't want to shuffle within each task counters = {} self.task_choice_list = [] for i in range(len(task_choice_list)): idx = counters.get(task_choice_list[i],0) self.task_choice_list.append((task_choice_list[i],idx)) counters[task_choice_list[i]] = idx + 1 def __len__(self): return np.sum([len(t) for t in self.tasks]) def __repr__(self): task_str = ", ".join([str(t) for t in self.tasks]) return f"MultiDataset(tasks: {task_str})" def __getitem__(self,key): if isinstance(key, int): task_idx, example_idx = self.task_choice_list[key] task = self.tasks[task_idx] example = task[example_idx] example["task_name"] = task.info.builder_name return example elif isinstance(key, slice): raise NotImplementedError() def __iter__(self): for i in range(len(self)): yield self[i] def load_multitask(*datasets): '''Create multitask datasets per split''' def _get_common_splits(datasets): '''Finds the common splits present in all self.datasets''' min_set = None for dataset in datasets: if min_set != None: min_set.intersection(set(dataset.keys())) else: min_set = set(dataset.keys()) return min_set common_splits = _get_common_splits(datasets) out = {} for split in common_splits: out[split] = MultiDataset([d[split] for d in datasets]) return out ########################################## # Dataset Flattening def flatten(dataset,flatten_fn): for k in dataset.keys(): if isinstance(dataset[k],nlp.Dataset): dataset[k] = dataset[k].map(flatten_fn,remove_columns=dataset[k].column_names) # Squad def flatten_squad(example): return {"source": "squad context: " + example['context'] + " question: " + example['question'], "target":example["answers"]["text"]} squad = nlp.load_dataset("squad") flatten(squad,flatten_squad) # CNN_DM def flatten_cnn_dm(example): return {"source": "cnn_dm: " + example['article'],"target":[example["highlights"]]} cnn_dm = nlp.load_dataset("cnn_dailymail", "3.0.0") flatten(cnn_dm,flatten_cnn_dm) ############################################# mtds = load_multitask(squad,cnn_dm) for example in mtds["train"]: print(example["task_name"],example["target"]) ``` Let me know if you have any thoughts. I've started using this in some of my projects and it seems to work. If people are happy with the general approach for a first version, I can make a pull request.
[ -0.0021333603654056787, -0.3933575451374054, -0.04490913450717926, 0.011328889057040215, -0.22390766441822052, 0.021761232987046242, 0.1978977620601654, 0.027536379173398018, 0.4143022894859314, -0.19782723486423492, -0.23707053065299988, 0.3982107639312744, -0.19453275203704834, 0.3028287887573242, 0.35159656405448914, -0.40193241834640503, -0.022123174741864204, -0.2419825792312622, -0.2949594557285309, 0.3513539731502533, 0.0604415126144886, 0.1977684646844864, -0.21249979734420776, 0.06919216364622116, -0.21596850454807281, -0.1789112240076065, -0.35519078373908997, 0.15981405973434448, 0.08680301159620285, 0.008535332046449184, 0.2525813579559326, 0.420715868473053, -0.04791150614619255, 0.17200107872486115, -0.00011538729449966922, -0.04062880948185921, 0.22244973480701447, -0.1226240023970604, 0.05334274098277092, -0.06257646530866623, 0.17279627919197083, -0.4180442988872528, 0.0764215737581253, -0.1779252290725708, 0.16830509901046753, 0.15077154338359833, 0.12851455807685852, 0.01803109422326088, 0.39563167095184326, -0.2608160078525543, 0.10167468339204788, 0.31338343024253845, -0.15474624931812286, 0.32001397013664246, -0.13475899398326874, -0.021238407120108604, 0.025054901838302612, 0.2562727928161621, 0.6043381094932556, -0.5478275418281555, -0.03985998407006264, 0.09073184430599213, 0.02038188837468624, -0.08250822871923447, 0.12837646901607513, -0.027154872193932533, -0.3430265486240387, -0.2517932951450348, -0.3898301124572754, 0.22086505591869354, -0.058725327253341675, -0.1800946146249771, -0.22417034208774567, -0.5440161228179932, -0.05742178484797478, 0.0014681450556963682, -0.20001594722270966, -0.09796605259180069, -0.20692327618598938, -0.07884522527456284, -0.08055754005908966, -0.1003153920173645, -0.05526085942983627, 0.1227867603302002, 0.28622671961784363, 0.5722639560699463, 0.09168506413698196, 0.1695953905582428, 0.41128870844841003, -0.003769771894440055, -0.10389386862516403, -0.27631959319114685, 0.2890004515647888, 0.10878820717334747, -0.46526631712913513, -0.3517702519893646, 0.20772352814674377, -0.273562490940094, 0.24205534160137177, 0.11815501004457474, 0.12676817178726196, 0.14092056453227997, -0.1725960075855255, 0.2350257784128189, 0.3101998567581177, 0.16164416074752808, -0.08544246852397919, -0.14018550515174866, -0.04084446281194687, -0.1004975438117981, 0.26056796312332153, 0.0931038111448288, 0.13976232707500458, -0.024155663326382637, -0.4383578300476074, -0.11291621625423431, -0.19020836055278778, 0.17506897449493408, -0.23685778677463531, -0.4651108682155609, -0.11562272161245346, -0.13347367942333221, 0.16438128054141998, -0.08931661397218704, -0.3263590931892395, 0.09385742247104645, -0.06360576301813126, 0.22851437330245972, -0.16430990397930145, 0.09540671855211258, -0.05092420056462288, 0.07936499267816544, -0.42001965641975403, -0.10349620133638382, 0.30415675044059753, 0.20076106488704681, 0.08877851068973541, 0.10183301568031311, 0.04135487601161003, 0.16144442558288574, 0.3386403024196625, -0.01858961023390293, 0.07116810232400894, -0.006750461645424366, 0.17005905508995056, -0.24186667799949646, 0.023633666336536407, 0.08572143316268921, -0.32762855291366577, -0.14327862858772278, -0.11624511331319809, -0.17121246457099915, 0.09555145353078842, 0.001572210225276649, -0.24700219929218292, -0.21890287101268768, -0.5296893119812012, 0.6929865479469299, 0.00901296641677618, -0.07522574812173843, -0.17376162111759186, 0.21700386703014374, -0.224966898560524, -0.11346238851547241, 0.18577522039413452, 0.13063377141952515, -0.01946037821471691, -0.2366185039281845, -0.08683909475803375, -0.1156580001115799, 0.00779843470081687, 0.25118863582611084, -0.17226935923099518, 0.31110909581184387, 0.0009650260908529162, 0.2511013150215149, 0.26969850063323975, -0.45747435092926025, -0.08002308011054993, 0.29960501194000244, -0.07061327993869781, 0.4915623664855957, 0.06509673595428467, 0.20209074020385742, 0.019948167726397514, -0.20112159848213196, 0.2741645872592926, 0.9725066423416138, -0.3398759067058563, -0.04546623304486275, -0.13535164296627045, -0.3082440197467804, 0.5843707323074341, 0.29365116357803345, 0.046059396117925644, -0.23651723563671112, -0.16255471110343933, 0.025136223062872887, 0.13693960011005402, 0.11461346596479416, 0.10209669172763824, -0.04339649900794029, -0.31103450059890747, 0.06801392883062363, -0.16123950481414795, -0.13331149518489838, -0.43702399730682373, 0.037815142422914505, 0.046109337359666824, 0.34421613812446594, 0.1861216127872467, -0.04249940812587738, 0.20138134062290192, -0.2429167479276657, -0.20070238411426544, -0.32318738102912903, 0.1282242238521576, -0.034077487885951996, -0.006764936726540327, -0.32578399777412415, -0.04058946669101715, 0.3911014497280121, 0.11225391924381256, -0.1496472954750061, -0.5206971764564514, 0.2977626919746399, -0.08318737149238586, 0.0701504647731781, 0.00024075244436971843, 0.6380759477615356, -0.09240604937076569, -0.13536162674427032, 0.18318288028240204, 0.3113521635532379, -0.2477383315563202, 0.14652997255325317, 0.3683392107486725, 0.17878352105617523, 0.1724848747253418, -0.0537877194583416, 0.04117901250720024, 0.014778828248381615, 0.06409084796905518, -0.25408563017845154, -0.04174485802650452, 0.49793940782546997, -0.20704413950443268, 0.4949685335159302, -0.05947204306721687, 0.09615954756736755, -0.07771971076726913, 0.003605868434533477, -0.21134988963603973, 0.41704630851745605, 0.2573401927947998, -0.16382719576358795, 0.3197738826274872, -0.17490975558757782, -0.38928598165512085, 0.21817058324813843, 0.1477797031402588, 0.05071171745657921, -0.023939797654747963, -0.004027349874377251, -0.004173393361270428, -0.2835718095302582, -0.08406389504671097, 0.25929099321365356, 0.672740638256073, 0.21467293798923492, 0.17579330503940582, -0.0035534724593162537, -0.21288920938968658, -0.07563218474388123, 0.01703449711203575, 0.21315525472164154, -0.01838487572968006, 0.1383441686630249, 0.06322697550058365, 0.055012695491313934, 0.02725956216454506, -0.1688724309206009, 0.16363899409770966, -0.15025147795677185, -0.11828155815601349, 0.11910311132669449, -0.240086168050766, -0.6449177861213684, -0.1902627795934677, -0.005380993243306875, -0.33743414282798767, -0.3379833400249481, 0.061977531760931015, 0.09609219431877136, -0.09279263764619827, 0.17458797991275787, 0.3390875458717346, 0.5370786786079407, -0.422608882188797, -0.2787003219127655, -0.0702023059129715, -0.3130028247833252, -0.22870545089244843, 0.10595978051424026, 0.4761104881763458, 0.3147263526916504, 0.43627244234085083, 0.33345988392829895, -0.0783115029335022, -0.011106290854513645, -0.40816861391067505, 0.07382749766111374, -0.3157113492488861, 0.25911790132522583, 0.10779257863759995, -0.16198229789733887, 0.15699651837348938, -0.36640769243240356, -0.05446052923798561, 0.025672148913145065, -0.1082521453499794, 0.005735683720558882, -0.0213202815502882, 0.06178341805934906, -0.2249409258365631, -0.20574027299880981, -0.6049152612686157, -0.3986763060092926, 0.33175206184387207, -0.2684570550918579, 0.15955175459384918, 0.11135060340166092, -0.29019060730934143, 0.2159576267004013, -0.0665777176618576, -0.03549874946475029, -0.1976851224899292, -0.02661256492137909, 0.04863744601607323, -0.3335723578929901, -0.2293011099100113, -0.3334794044494629, -0.08374940603971481, 0.28511953353881836, -0.060613855719566345, -0.01924082823097706, -0.21448449790477753, 0.13713322579860687, -0.09409260749816895, 0.11316557973623276, 0.18199342489242554, 0.2623929977416992, -0.22190843522548676, 0.1516171246767044, -0.1329754889011383, -0.337842732667923, 0.28923583030700684, 0.30653244256973267, 0.466779500246048, 0.26281359791755676, 0.059581559151411057, 0.19852980971336365, 0.7363934516906738, 0.2467268407344818, 0.030771398916840553, 0.16282251477241516, 0.24470698833465576, -0.008579007349908352, 0.1286430060863495, -0.29556092619895935, 0.24368086457252502, -0.2057005763053894, 0.489104688167572, 0.04684245213866234, 0.12142498046159744, -0.27849870920181274, -0.3198148310184479, 0.09729424864053726, -0.29605594277381897, -0.24099378287792206, 0.22886072099208832, -0.22867809236049652, -0.02966860681772232, -0.04629133269190788, -0.3796521723270416, -0.3028077781200409, -0.03775918111205101, 0.024953672662377357, 0.36506667733192444, -0.08353592455387115, 0.1451675295829773, -0.3331812918186188, -0.05883107706904411, -0.47370147705078125, 0.11649084091186523, 0.027234310284256935, 0.3008979856967926, -0.07650738954544067, -0.10251417756080627, -0.0961405336856842, 0.09850406646728516, 0.7110227942466736, -0.44423359632492065, -0.14522895216941833, 0.14357218146324158, 0.09769883751869202, -0.09330236911773682, -0.12175717204809189, -0.2865290939807892, 0.11032100766897202, 0.5207920670509338, 0.3619169294834137, -0.45472946763038635, -0.26354825496673584, 0.1425701230764389, -0.2180933952331543, -0.04638401418924332, -0.18992455303668976, -0.25059476494789124, -0.2670263350009918, -0.09710831940174103, 0.04224763065576553, 0.22509320080280304, 0.27964064478874207, -0.11629268527030945, 0.21690602600574493, -0.07204923778772354, 0.17009669542312622, 0.21983112394809723, 0.0008594237733632326, -0.02683510258793831, 0.25667092204093933, 0.14444349706172943, 0.08391444385051727, -0.0964151993393898, -0.35581180453300476, 0.12449466437101364, -0.02867567352950573, 0.13755163550376892, 0.34512564539909363, -0.06954682618379593, 0.5981538891792297, 0.15671159327030182, 0.3002188801765442, 0.22123515605926514, -0.13670718669891357, -0.013893986120820045, -0.1888938844203949, 0.23657263815402985, 0.21810705959796906, 0.20831623673439026, -0.14313721656799316, -0.24706393480300903, 0.48490485548973083, 0.25417235493659973, -0.13027384877204895, 0.4766072630882263, -0.46295738220214844, -0.3232699930667877, 0.23429550230503082, 0.08237776160240173, 0.7584076523780823, -0.02576621063053608, 0.248221293091774, 0.2218555063009262, -0.3872928023338318, 0.5620595216751099, -0.2961163818836212, 0.1522105634212494, -0.06021064519882202, -0.15607032179832458, -0.0034728043247014284, -0.10052952170372009, -0.06659572571516037, 0.19200089573860168, -0.30052128434181213, 0.3170955181121826, 0.474131315946579, -0.25450530648231506, -0.02542509138584137, 0.5520510673522949, 0.07409597188234329, -0.4448997378349304, -0.06326024979352951, 0.05981139838695526, -0.2775408625602722, 0.12795968353748322, -0.12317372113466263, -0.32199689745903015, 0.19604803621768951, 0.14368174970149994, -0.33403679728507996, 0.0841856375336647, 0.002531936625018716, 0.21790429949760437, 0.302400678396225, -0.018724597990512848, 0.24353741109371185, -0.1553886979818344, -0.06911890208721161, -0.20241792500019073, -0.08242638409137726, 0.123386912047863, -0.15316882729530334, 0.11678946018218994, 0.09097012132406235, -0.23128864169120789, 0.5451812148094177, -0.09264515340328217, -0.07114778459072113, -0.2456400841474533, 0.22469641268253326, -0.40141791105270386, -0.20007511973381042, 0.2005418986082077, 0.1613871455192566, 0.1955033540725708, -0.29072120785713196, 0.5905135273933411, 0.12673057615756989, -0.0039932942017912865, 0.01494226511567831, 0.22409683465957642, -0.23790395259857178, 0.49399104714393616, -0.33471786975860596, 0.03277881443500519, 0.1014416292309761, 0.041583284735679626, -0.10776904225349426, -0.4338432550430298, -0.0012162006460130215, 0.3342473804950714, 0.07581999897956848, 0.05143161490559578, -0.08145029097795486, 0.07534009218215942, -0.048213861882686615, -0.09353337436914444, -0.18528534471988678, -0.34339889883995056, 0.0834069550037384, 0.2625077962875366, 0.1403319537639618, 0.22585947811603546, -0.06235148757696152, -0.30122631788253784, -0.18288184702396393, 0.4282071888446808, -0.09056097269058228, 0.17181028425693512, -0.16093997657299042, 0.3156590163707733, 0.412031888961792, 0.4628569185733795, -0.2444220930337906, 0.016907453536987305, 0.022540757432579994, 0.09322317689657211, 0.3734446167945862, -0.23166696727275848, 0.07709408551454544, 0.24294039607048035, -0.13898295164108276, 0.04263535514473915, -0.34411001205444336, -0.12457136809825897, -0.26963138580322266, 0.11243976652622223, 0.1352788507938385, -0.04004857316613197, 0.004310352727770805, -0.13928335905075073, -0.13679295778274536, -0.06827126443386078, -0.03434407711029053, -0.4005204141139984, -0.10314895212650299, 0.0878286063671112, -0.04722683131694794, 0.06675835698843002, 0.061761487275362015, 0.16811570525169373, 0.052707038819789886, -0.08132447302341461, 0.38305598497390747, 0.09156324714422226, 0.07138901948928833, -0.1855536848306656, -0.16798467934131622, 0.05471836403012276, 0.27848997712135315, 0.14243140816688538, -0.04896443337202072, 0.06975319236516953, -0.21316082775592804, -0.1310151219367981, 0.21939092874526978, 0.22825856506824493, 0.05144934728741646, -0.19707201421260834, 0.13877585530281067, 0.08696145564317703, -0.0729917660355568, -0.024210074916481972, 0.25871357321739197, 0.0611560232937336, 0.0581028051674366, 0.15578068792819977, 0.13392339646816254, 0.4340778887271881, -0.18399545550346375, 0.04718602076172829, 0.16860873997211456, -0.37249672412872314, 0.05300147831439972, -0.19252823293209076, -0.12392820417881012, 0.04346950724720955, 0.17797201871871948, -0.020713338628411293, 0.11280062794685364, -0.1853484958410263, 0.2707880139350891, 0.312272310256958, 0.009717307053506374, 0.047666389495134354, 0.13717597723007202, -0.05930177494883537, -0.18108217418193817, 0.20019438862800598, 0.05756025388836861, 0.34439516067504883, 0.01869833841919899, 0.8508671522140503, -0.2761963903903961, -0.38287946581840515, 0.17217527329921722, 0.4096406400203705, 0.03172922134399414, -0.18231843411922455, 0.08489274233579636, 0.07741580158472061, -0.34510067105293274, 0.01853763498365879, 0.03847216069698334, -0.1703849881887436, 0.6866497993469238, -0.13125547766685486, 0.11843198537826538, -0.22235527634620667, -0.2760974168777466, 0.3105558454990387, 0.1714400202035904, -0.056525588035583496, 0.06068376824259758, 0.1309126913547516, -0.09606020897626877, -0.16307930648326874, 0.4281637966632843, 0.2053738236427307, 0.0710359737277031, -0.4505878984928131, -0.055379509925842285, 0.10228003561496735, 0.11845839768648148, -0.08575453609228134, 0.15496787428855896, 0.4437693655490875, -0.12202072888612747, 0.27056705951690674, -0.023743296042084694, -0.0232196357101202, -0.13216695189476013, 0.1632668823003769, -0.33133387565612793, -0.37357965111732483, 0.45984208583831787, 0.08283783495426178, -0.11625416576862335, 0.09675093740224838, 0.29102247953414917, -0.6527740359306335, -0.24476131796836853, 0.2560926377773285, -0.06182890757918358, -0.0006648338749073446, -0.1802779585123062, 0.06906047463417053, 0.08660885691642761, 0.6433348655700684, 0.16453427076339722, 0.30600109696388245, -0.40145447850227356, -0.12584948539733887, -0.6316565275192261, -0.2174752801656723, 0.1793736070394516, -0.13840211927890778, -0.1703631579875946, 0.18058380484580994, -0.007042988669127226, 0.24776317179203033, -0.26092463731765747, -0.006970847491174936, 0.3292713463306427, 0.04296679422259331, -0.519255518913269, 0.039342790842056274, 0.1607200652360916, 0.11304409056901932, 0.2652929127216339, -0.020683500915765762, -0.20508576929569244, 0.10758076608181, -0.06017373874783516, -0.0842011347413063, -0.1182250902056694, 0.2389034926891327, 0.20819246768951416, 0.0055058798752725124, 0.06280211359262466, 0.49198204278945923, 0.006552661303430796, -0.17013630270957947, -0.004910780116915703, 0.09330537170171738, -0.33859243988990784, -0.16125202178955078, -0.10824790596961975, 0.22235308587551117, -0.34201085567474365, 0.35956889390945435, -0.017937349155545235, -0.08411450684070587, 0.13189351558685303, 0.025732532143592834, -0.263685017824173, 0.028533894568681717, -0.008091885596513748, 0.2732331454753876, 0.20005369186401367, 0.7247748970985413, 0.039764754474163055, 0.17260189354419708, -0.1615489274263382, -0.041413113474845886, 0.10445640981197357, -0.17149613797664642, -0.28233709931373596, -0.11068553477525711, 0.07003909349441528, 0.03175491839647293, 0.13340802490711212, -0.7072269320487976, -0.3839826285839081, 0.0066392431035637856, -0.03267090395092964, 0.024425743147730827, 0.6573564410209656, 0.04589991644024849, -0.10253032296895981, -0.13903065025806427, -0.3231089115142822, 0.0940261036157608, -0.23133830726146698, 0.18395103514194489, -0.35463395714759827 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
Hey! Happy to jump into the discussion here. I'm still getting familiar with bits of this code, but the reasons I sampled over data loaders rather than datasets is 1) ensuring that each sampled batch corresponds to only 1 task (in case of different inputs formats/downstream models) and 2) potentially having different batch sizes per task (e.g. some tasks have very long/short inputs). How are you currently dealing with these in your PR?
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
73
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. Hey! Happy to jump into the discussion here. I'm still getting familiar with bits of this code, but the reasons I sampled over data loaders rather than datasets is 1) ensuring that each sampled batch corresponds to only 1 task (in case of different inputs formats/downstream models) and 2) potentially having different batch sizes per task (e.g. some tasks have very long/short inputs). How are you currently dealing with these in your PR?
[ -0.08960554003715515, -0.3705180287361145, -0.04412422329187393, -0.06419779360294342, -0.22869285941123962, 0.02895098552107811, 0.2676898241043091, 0.02336002327501774, 0.3950953483581543, -0.10933452099561691, -0.1705915778875351, 0.32033583521842957, -0.20159779489040375, 0.37548577785491943, 0.39844727516174316, -0.43781787157058716, -0.002577137900516391, -0.2555519938468933, -0.3273431360721588, 0.3867936432361603, 0.0030516749247908592, 0.11846121400594711, -0.14524959027767181, 0.05655597522854805, -0.2847825884819031, -0.12906894087791443, -0.33200934529304504, 0.10755182802677155, 0.09705800563097, -0.011513284407556057, 0.23895084857940674, 0.36387988924980164, -0.0029587079770863056, 0.12613248825073242, -0.0001156874859589152, -0.041392985731363297, 0.22455823421478271, -0.0885121077299118, 0.06489138305187225, -0.04569966346025467, 0.17900067567825317, -0.3761780560016632, 0.058788664638996124, -0.14258639514446259, 0.20260384678840637, 0.17821432650089264, 0.17752526700496674, 0.04718824103474617, 0.3459136486053467, -0.3254343569278717, 0.08241233974695206, 0.32058170437812805, -0.2082773745059967, 0.2640053331851959, -0.11156558990478516, -0.050212666392326355, 0.07479429244995117, 0.31859785318374634, 0.6831013560295105, -0.5693539977073669, -0.07522579282522202, 0.10948431491851807, 0.01886441931128502, -0.0638963133096695, 0.153956800699234, -0.08444599062204361, -0.30900251865386963, -0.24644383788108826, -0.436266154050827, 0.21254250407218933, -0.11863076686859131, -0.12883256375789642, -0.2360723614692688, -0.6045428514480591, -0.02912726067006588, 0.07235995680093765, -0.27344658970832825, -0.00966254435479641, -0.21529775857925415, -0.04984115809202194, -0.1738603115081787, -0.10561870783567429, -0.03110460378229618, 0.14210423827171326, 0.3534385859966278, 0.5484309196472168, 0.07717595249414444, 0.21021868288516998, 0.41617855429649353, 0.08392105251550674, -0.11613097041845322, -0.27954214811325073, 0.31732675433158875, 0.12179842591285706, -0.36284583806991577, -0.34425225853919983, 0.17837680876255035, -0.2727919816970825, 0.26535090804100037, 0.1431955099105835, 0.13283878564834595, 0.05661166459321976, -0.12779097259044647, 0.24213290214538574, 0.3584893047809601, 0.08159506320953369, -0.09061973541975021, -0.1105470359325409, 0.011070125736296177, -0.1044858917593956, 0.22195376455783844, 0.10372376441955566, 0.14533616602420807, -0.015399143099784851, -0.4716433584690094, -0.08467740565538406, -0.2308618277311325, 0.11420781910419464, -0.18811678886413574, -0.5452080368995667, -0.07727597653865814, -0.14436045289039612, 0.11034179478883743, -0.089957095682621, -0.2681083381175995, 0.11734511703252792, -0.0560462549328804, 0.25359100103378296, -0.138411283493042, 0.04875034838914871, -0.0495743565261364, 0.061328575015068054, -0.45849037170410156, -0.09332142025232315, 0.28703662753105164, 0.17207972705364227, 0.12903070449829102, 0.12023808062076569, -0.019158367067575455, 0.1441025584936142, 0.33735761046409607, 0.028596311807632446, 0.01305076852440834, 0.010992662981152534, 0.18654268980026245, -0.30493998527526855, 0.011405106633901596, 0.23146313428878784, -0.3490491211414337, -0.10899244248867035, -0.20152604579925537, -0.19108104705810547, 0.1673603653907776, 0.013931088149547577, -0.21046417951583862, -0.26141494512557983, -0.559792160987854, 0.6944108605384827, -0.03270532563328743, -0.10934837907552719, -0.14054279029369354, 0.2590895891189575, -0.16008660197257996, -0.08462852984666824, 0.1551489531993866, 0.048238445073366165, -0.04864814132452011, -0.36557167768478394, -0.07868371158838272, -0.12028457224369049, 0.05071501433849335, 0.26793599128723145, -0.22832469642162323, 0.27010542154312134, 0.018788857385516167, 0.1974085569381714, 0.24765551090240479, -0.47684645652770996, -0.03297731280326843, 0.3645942211151123, -0.044346049427986145, 0.40321362018585205, 0.10897758603096008, 0.20473597943782806, 0.042717546224594116, -0.20092621445655823, 0.2404809594154358, 1.0242562294006348, -0.3771042823791504, 0.008822514675557613, -0.12261920422315598, -0.3946356773376465, 0.5419135093688965, 0.3336593508720398, 0.08739208430051804, -0.26467546820640564, -0.11319258064031601, 0.020931541919708252, 0.14002974331378937, 0.06807202845811844, 0.09149987995624542, -0.06338773667812347, -0.3728692829608917, 0.04508441314101219, -0.21324512362480164, -0.09416227787733078, -0.36404508352279663, 0.02004147320985794, -0.07089370489120483, 0.33006349205970764, 0.27682793140411377, -0.10196209698915482, 0.29011106491088867, -0.25997957587242126, -0.135525181889534, -0.29892584681510925, 0.08099512755870819, -0.09198074042797089, 0.019625229761004448, -0.3017866015434265, -0.04352991282939911, 0.3114900290966034, 0.09314750134944916, -0.11146474629640579, -0.5201103091239929, 0.29312941431999207, -0.035998258739709854, 0.061564382165670395, -0.0562952384352684, 0.5926681160926819, -0.16117364168167114, -0.12305664271116257, 0.15402843058109283, 0.30346056818962097, -0.26907238364219666, 0.2271573394536972, 0.2872881591320038, 0.1880013644695282, 0.1648530811071396, -0.016522115096449852, 0.13337133824825287, 0.005181524436920881, -0.01253964751958847, -0.22247450053691864, -0.07112506031990051, 0.49432915449142456, -0.17993122339248657, 0.4873959720134735, -0.044664788991212845, 0.015740156173706055, -0.1863202452659607, -0.047534745186567307, -0.23269301652908325, 0.4075045883655548, 0.31928837299346924, -0.1794973760843277, 0.3725852072238922, -0.0862383246421814, -0.35132768750190735, 0.17856565117835999, 0.09627626091241837, 0.00019789105863310397, 0.021255312487483025, -0.012914730235934258, -0.009134471416473389, -0.333839476108551, -0.1100996658205986, 0.33554261922836304, 0.6644570827484131, 0.24936838448047638, 0.17727471888065338, -0.028650732710957527, -0.19699716567993164, -0.09434293955564499, 0.03606782481074333, 0.21235482394695282, -0.02482195571064949, 0.15700162947177887, -0.010826516896486282, 0.03673882409930229, -0.0019101343350484967, -0.145143523812294, 0.1448415368795395, -0.2040402889251709, -0.08474630862474442, 0.08304176479578018, -0.20195823907852173, -0.6090885996818542, -0.2475719153881073, -0.017396826297044754, -0.3290187418460846, -0.2885233461856842, 0.10992716997861862, 0.08244853466749191, -0.15863166749477386, 0.14156195521354675, 0.22592225670814514, 0.6497551202774048, -0.4458456039428711, -0.23835527896881104, -0.09807469695806503, -0.27364420890808105, -0.1178838312625885, 0.08979293704032898, 0.4389151930809021, 0.2987910807132721, 0.4678761661052704, 0.3357656002044678, -0.07294555753469467, 0.0949573740363121, -0.3517071306705475, 0.03553038462996483, -0.2930789291858673, 0.3098820447921753, 0.13900770246982574, -0.18240390717983246, 0.19600342214107513, -0.3770199716091156, -0.0679650604724884, -0.04905390739440918, -0.09277144074440002, -0.06871548295021057, -0.026184791699051857, 0.07875335216522217, -0.19790895283222198, -0.14304053783416748, -0.6475344300270081, -0.42761489748954773, 0.26241016387939453, -0.2563125491142273, 0.05381110683083534, 0.05618848279118538, -0.27896520495414734, 0.16724130511283875, -0.11408708244562149, -0.0301990769803524, -0.1667529046535492, -0.1156589612364769, -0.022316204383969307, -0.3414052426815033, -0.16686128079891205, -0.43680301308631897, -0.0881427675485611, 0.20312537252902985, -0.019464073702692986, -0.019430462270975113, -0.18683062493801117, 0.15789037942886353, -0.1286054104566574, 0.13385477662086487, 0.22555020451545715, 0.14002415537834167, -0.2490689754486084, 0.14498241245746613, -0.06924628466367722, -0.2159973382949829, 0.22665143013000488, 0.31023114919662476, 0.46727079153060913, 0.20428210496902466, 0.021375007927417755, 0.1477929800748825, 0.7359364032745361, 0.22510188817977905, 0.06165260821580887, 0.15805049240589142, 0.29719817638397217, -0.011522085405886173, 0.10848750919103622, -0.20904165506362915, 0.2638302147388458, -0.19775930047035217, 0.48094651103019714, 0.0856480523943901, 0.14452938735485077, -0.18779617547988892, -0.2808535099029541, 0.00895918533205986, -0.2754822075366974, -0.22052402794361115, 0.2475835382938385, -0.2749198377132416, 0.008521145209670067, -0.07700461894273758, -0.36289507150650024, -0.28628379106521606, -0.03926142677664757, -0.007200351916253567, 0.3148202896118164, 0.030279072001576424, 0.16325677931308746, -0.39912745356559753, -0.06987354159355164, -0.4792778193950653, 0.12462103366851807, 0.05004160851240158, 0.272379070520401, -0.07456710189580917, -0.10680718719959259, -0.08096519112586975, 0.13911333680152893, 0.6718359589576721, -0.46807432174682617, -0.1201029121875763, 0.0467565692961216, 0.08207797259092331, -0.059771839529275894, -0.12703630328178406, -0.23704156279563904, 0.020004751160740852, 0.4522341787815094, 0.3554958701133728, -0.41076284646987915, -0.2772548496723175, 0.2263064831495285, -0.16873954236507416, -0.08126653730869293, -0.22157074511051178, -0.21372391283512115, -0.17879308760166168, -0.1205180212855339, 0.022923383861780167, 0.19870980083942413, 0.2567113935947418, -0.1758774220943451, 0.1802804172039032, -0.0964084267616272, 0.27499037981033325, 0.18857261538505554, -0.00485547911375761, -0.06122709438204765, 0.19523654878139496, 0.15556934475898743, 0.07697004079818726, -0.11399330198764801, -0.2888614237308502, 0.11771688610315323, -0.09440398216247559, 0.09064515680074692, 0.27966880798339844, -0.027801048010587692, 0.6893088221549988, 0.08675303310155869, 0.3350837230682373, 0.2698398232460022, -0.0843542218208313, -0.01583169214427471, -0.25507739186286926, 0.17849336564540863, 0.1570008099079132, 0.2542004883289337, -0.15993094444274902, -0.2616676688194275, 0.5813788175582886, 0.3021725118160248, -0.07488233596086502, 0.4123932719230652, -0.5658520460128784, -0.3181343674659729, 0.3210618197917938, 0.008234641514718533, 0.6767167448997498, -0.011694325134158134, 0.29665639996528625, 0.09642128646373749, -0.2917300760746002, 0.5918285846710205, -0.3068067729473114, 0.13567307591438293, -0.08629437536001205, -0.15701019763946533, -0.0007171488250605762, -0.10220611095428467, -0.06825969368219376, 0.28566810488700867, -0.2693983018398285, 0.3406902849674225, 0.4515395164489746, -0.2593255341053009, -0.062462568283081055, 0.497019499540329, 0.040260449051856995, -0.39022523164749146, -0.05338864028453827, 0.03806700184941292, -0.26367396116256714, 0.15802551805973053, -0.13398049771785736, -0.27603402733802795, 0.2424035519361496, 0.09275715798139572, -0.2599029242992401, 0.1079489067196846, 0.002680172910913825, 0.2147812843322754, 0.3164939880371094, 0.024473637342453003, 0.2476588785648346, -0.18959233164787292, -0.15023615956306458, -0.17084495723247528, -0.06086140498518944, 0.13157792389392853, -0.14995668828487396, 0.07499933987855911, 0.10225125402212143, -0.187038391828537, 0.533451497554779, -0.06037310138344765, -0.08038851618766785, -0.2575174570083618, 0.22078347206115723, -0.4532489478588104, -0.15675626695156097, 0.15572895109653473, 0.17293323576450348, 0.17084285616874695, -0.3983147442340851, 0.5807891488075256, 0.13632529973983765, 0.016067009419202805, -0.0007994562620297074, 0.16640523076057434, -0.22039026021957397, 0.5007690191268921, -0.31078168749809265, 0.009038742631673813, 0.07070190459489822, 0.006287605036050081, -0.1292516589164734, -0.41763001680374146, -0.048088934272527695, 0.29771339893341064, 0.09809495508670807, 0.05041976273059845, -0.009112373925745487, 0.18022215366363525, -0.11553841084241867, -0.030411649495363235, -0.11281533539295197, -0.2936977744102478, 0.13766342401504517, 0.1897747814655304, 0.20098917186260223, 0.19345170259475708, -0.08014584332704544, -0.32933899760246277, -0.1923336535692215, 0.49070796370506287, -0.035176124423742294, 0.20113801956176758, -0.08926854282617569, 0.3534668982028961, 0.39224234223365784, 0.45730653405189514, -0.2312459945678711, 0.034199442714452744, 0.07259298861026764, 0.09041871130466461, 0.42385438084602356, -0.21112456917762756, 0.0549248605966568, 0.29967057704925537, -0.15968696773052216, 0.0328933410346508, -0.29686465859413147, -0.12285884469747543, -0.24775628745555878, 0.11770711094141006, 0.12449155002832413, -0.01967359147965908, 0.0004342575848568231, -0.1036360114812851, -0.05941019207239151, -0.05669474974274635, 0.016157839447259903, -0.3881988823413849, -0.07074052840471268, 0.08002275228500366, -0.056611210107803345, 0.05917535349726677, 0.0188970435410738, 0.305361270904541, 0.03774058073759079, -0.04798999801278114, 0.31572604179382324, 0.1257237046957016, 0.08717852085828781, -0.15319710969924927, -0.1950378268957138, 0.0172455832362175, 0.3017753064632416, 0.1727597862482071, -0.03927168250083923, 0.16133391857147217, -0.16719701886177063, -0.1590324342250824, 0.22609062492847443, 0.2231975495815277, 0.04009833559393883, -0.1603202074766159, 0.11157412081956863, 0.0839044377207756, -0.0976882129907608, 0.012813540175557137, 0.15069225430488586, 0.1132158488035202, 0.061027832329273224, 0.14053089916706085, 0.17595915496349335, 0.45910993218421936, -0.2538210451602936, -0.002814853098243475, 0.14726053178310394, -0.24171803891658783, 0.0918261781334877, -0.07453323900699615, -0.042423348873853683, 0.022125162184238434, 0.1766907125711441, 0.01691071316599846, 0.12707701325416565, -0.3183572590351105, 0.2580794095993042, 0.3101721704006195, 0.005789766553789377, 0.014649984426796436, 0.09429752081632614, -0.06749323755502701, -0.18665927648544312, 0.3170959949493408, 0.009613605216145515, 0.3464256227016449, 0.04529754817485809, 0.875375509262085, -0.3337633013725281, -0.42751333117485046, 0.15399423241615295, 0.44893980026245117, 0.056862931698560715, -0.2221122533082962, 0.037358108907938004, 0.1282808780670166, -0.368176132440567, 0.11881019175052643, 0.016469178721308708, -0.14899775385856628, 0.6261694431304932, -0.11506877839565277, 0.1415538340806961, -0.19208961725234985, -0.2517244517803192, 0.30882471799850464, 0.1510937213897705, -0.10871902108192444, 0.003980865702033043, 0.08623854070901871, -0.031954143196344376, -0.0897655263543129, 0.4948278069496155, 0.24426411092281342, 0.044602423906326294, -0.42176398634910583, -0.05275627225637436, 0.06980502605438232, 0.15835484862327576, -0.09271010011434555, 0.1416839212179184, 0.45747193694114685, -0.1465972512960434, 0.25981706380844116, -0.012058008462190628, -0.028776448220014572, -0.10949544608592987, 0.10569210350513458, -0.4016034007072449, -0.3663012683391571, 0.4545689523220062, 0.09633476287126541, -0.10901253670454025, 0.050898145884275436, 0.29747840762138367, -0.6373807787895203, -0.23753005266189575, 0.30419328808784485, -0.12987738847732544, -0.07157387584447861, -0.18063631653785706, 0.06053175404667854, 0.057966407388448715, 0.6411344408988953, 0.244784414768219, 0.3587520122528076, -0.41919243335723877, -0.12837129831314087, -0.5469253063201904, -0.2503894865512848, 0.1415645331144333, -0.0787108987569809, -0.263457328081131, 0.13791395723819733, -0.04050194472074509, 0.25394538044929504, -0.2701583802700043, 0.10107421875, 0.23565591871738434, 0.09575521945953369, -0.43658778071403503, 0.0883835032582283, 0.13122303783893585, 0.10667705535888672, 0.2530834674835205, -0.095354363322258, -0.1709674447774887, 0.10993629693984985, -0.12099000066518784, -0.08666301518678665, -0.08444850146770477, 0.31150922179222107, 0.22627875208854675, 0.048147547990083694, 0.051024097949266434, 0.5658936500549316, 0.008747790940105915, -0.1783836930990219, 0.01966794766485691, 0.009437673725187778, -0.3435695767402649, -0.16456227004528046, -0.08115562051534653, 0.154269739985466, -0.31995949149131775, 0.3417586386203766, -0.019736025482416153, -0.11823615431785583, 0.10428190976381302, 0.002966197207570076, -0.2295610010623932, -0.05096915736794472, 0.005003814119845629, 0.1825326532125473, 0.2513962984085083, 0.7283110022544861, 0.028254294767975807, 0.09514983743429184, -0.07251212745904922, 0.0025239847600460052, 0.031027233228087425, -0.10433628410100937, -0.2157592922449112, -0.17576730251312256, 0.044877782464027405, 0.013304026797413826, 0.12388887256383896, -0.7102765440940857, -0.36426395177841187, 0.019617583602666855, 0.013225079514086246, 0.08402092009782791, 0.7129796743392944, 0.074734166264534, -0.14252959191799164, -0.1384015828371048, -0.3290708065032959, 0.05110197141766548, -0.23991213738918304, 0.19118696451187134, -0.33836326003074646 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
The short answer is - I'm not! Everything is currently on a per-example basis. It would be fairly simple to add a `batch_size` argument which would ensure that every `batch_size` examples come from the same task. That should suit most use-cases (unless you wanted to ensure batches all came from the same task and apply something like `SortishSampler` on each task first) Your notebook was really inspiring by the way - thanks!
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
72
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. The short answer is - I'm not! Everything is currently on a per-example basis. It would be fairly simple to add a `batch_size` argument which would ensure that every `batch_size` examples come from the same task. That should suit most use-cases (unless you wanted to ensure batches all came from the same task and apply something like `SortishSampler` on each task first) Your notebook was really inspiring by the way - thanks!
[ -0.0956544578075409, -0.43038398027420044, -0.04629383236169815, -0.06745117157697678, -0.2578943371772766, 0.009947361424565315, 0.20224931836128235, 0.07624121755361557, 0.2903968095779419, -0.11092482507228851, -0.18591274321079254, 0.2946228086948395, -0.23009297251701355, 0.3789537250995636, 0.4100240468978882, -0.45200031995773315, 0.024441173300147057, -0.177206352353096, -0.29455024003982544, 0.36274445056915283, 0.015423019416630268, 0.14288891851902008, -0.1547030806541443, 0.08106373995542526, -0.31288331747055054, -0.16926202178001404, -0.34849435091018677, 0.10086295753717422, 0.1370152235031128, -0.024620696902275085, 0.2090301811695099, 0.37785276770591736, 0.0340089425444603, 0.10139340162277222, -0.00011142913717776537, -0.09728694707155228, 0.20507904887199402, -0.09062527865171432, 0.10297595709562302, -0.05900351703166962, 0.16401681303977966, -0.36792224645614624, -0.002734801033511758, -0.12695015966892242, 0.15940764546394348, 0.22511327266693115, 0.2023814469575882, 0.08568355441093445, 0.3988804221153259, -0.3097243905067444, 0.11813075840473175, 0.328694224357605, -0.18170171976089478, 0.23797333240509033, -0.08357034623622894, -0.06251093000173569, 0.06098923832178116, 0.23452819883823395, 0.6071343421936035, -0.58875572681427, -0.06404931843280792, 0.14051827788352966, 0.03854520991444588, 0.021421179175376892, 0.14441172778606415, -0.038736216723918915, -0.29622340202331543, -0.232766255736351, -0.43623578548431396, 0.2806535065174103, -0.1612735390663147, -0.13062630593776703, -0.2634523808956146, -0.5633004307746887, -0.0008519762777723372, 0.0014623303432017565, -0.33009815216064453, 0.044788017868995667, -0.21435800194740295, -0.03346629813313484, -0.1524394154548645, -0.06026735529303551, -0.04060104489326477, 0.11201122403144836, 0.3631242513656616, 0.5616530179977417, 0.08171816915273666, 0.13045640289783478, 0.3781227767467499, 0.05841761454939842, -0.19902941584587097, -0.2562060058116913, 0.2837378978729248, 0.11374102532863617, -0.34667477011680603, -0.3419145941734314, 0.17506270110607147, -0.19308975338935852, 0.3228084444999695, 0.10108928382396698, 0.0934973806142807, 0.09113769978284836, -0.10029987245798111, 0.2920592427253723, 0.3476570248603821, 0.1223694384098053, -0.056605610996484756, -0.12424670159816742, 0.03975919634103775, -0.06336721777915955, 0.2047533541917801, 0.15042757987976074, 0.1986275166273117, -0.0070348153822124004, -0.4239506423473358, -0.113590307533741, -0.18267902731895447, 0.1420101821422577, -0.21430006623268127, -0.5234178304672241, -0.029956500977277756, -0.1304442137479782, 0.08226392418146133, -0.056739676743745804, -0.2853318154811859, 0.0973653644323349, -0.08805660158395767, 0.19672279059886932, -0.12846866250038147, 0.07483378052711487, -0.08253401517868042, 0.07176294922828674, -0.4690389037132263, -0.04859967529773712, 0.30152627825737, 0.1074480339884758, 0.13133564591407776, 0.13816595077514648, -0.0377437062561512, 0.1328812688589096, 0.33942869305610657, 0.03031286410987377, -0.0003307335136923939, -0.03616715595126152, 0.15753719210624695, -0.3345581591129303, -0.04376925528049469, 0.1560826301574707, -0.4061700105667114, -0.15703776478767395, -0.12641926109790802, -0.21524955332279205, 0.13307444751262665, 0.05488738417625427, -0.23455174267292023, -0.25436821579933167, -0.4572620689868927, 0.7054932713508606, -0.04623964801430702, -0.03340712934732437, -0.1386662870645523, 0.28011998534202576, -0.23864740133285522, -0.07164394855499268, 0.15914246439933777, 0.09545451402664185, 0.0009728616569191217, -0.37805625796318054, -0.03246217966079712, -0.10533303022384644, -0.06803721189498901, 0.31936436891555786, -0.17716726660728455, 0.1892450600862503, 0.05113629996776581, 0.20222069323062897, 0.24152277410030365, -0.49662357568740845, -0.03344209864735603, 0.30526044964790344, -0.03650669381022453, 0.35196879506111145, 0.07787486910820007, 0.1823495477437973, 0.09818778187036514, -0.1924564093351364, 0.2653273344039917, 1.0035321712493896, -0.38286951184272766, 0.022055543959140778, -0.12745937705039978, -0.40109676122665405, 0.48556333780288696, 0.3276016414165497, 0.023420549929142, -0.31363874673843384, -0.1300777792930603, -0.02560476027429104, 0.09931154549121857, 0.06988128274679184, 0.07652884721755981, -0.06977523118257523, -0.31873393058776855, 0.013977481052279472, -0.15631647408008575, -0.1352730542421341, -0.4236290454864502, 0.03415070101618767, -0.06946251541376114, 0.33294907212257385, 0.32506129145622253, -0.13765071332454681, 0.21762682497501373, -0.26340484619140625, -0.10048290342092514, -0.31292086839675903, 0.1418483406305313, -0.10447479784488678, 0.0539260171353817, -0.27995923161506653, -0.07492944598197937, 0.3310256004333496, 0.0866703987121582, -0.12241551280021667, -0.46758076548576355, 0.2426157146692276, -0.10650381445884705, 0.017546098679304123, 0.00730044674128294, 0.6171316504478455, -0.14959515631198883, -0.07000638544559479, 0.20503389835357666, 0.28201720118522644, -0.28210893273353577, 0.2547476291656494, 0.2886512875556946, 0.22653183341026306, 0.1838991641998291, 0.012912996113300323, 0.1047697439789772, -0.018239282071590424, -0.03309926018118858, -0.18438954651355743, -0.12761665880680084, 0.5337746143341064, -0.11340359598398209, 0.42374780774116516, 0.01028094906359911, -0.0016261906130239367, -0.1857481747865677, -0.08734411746263504, -0.2569580674171448, 0.35327139496803284, 0.3242209851741791, -0.21554195880889893, 0.3575747311115265, -0.0927836075425148, -0.3873851001262665, 0.1576685756444931, 0.18133841454982758, 0.02245202660560608, 0.08836869895458221, 0.022484786808490753, -0.043993622064590454, -0.2892822325229645, -0.08268630504608154, 0.3609899878501892, 0.600655198097229, 0.30525386333465576, 0.17336374521255493, 0.027060076594352722, -0.30429205298423767, -0.11205693334341049, 0.015227646566927433, 0.18314243853092194, -0.030197398737072945, 0.15285487473011017, 0.041586801409721375, 0.05599971115589142, -0.048957277089357376, -0.16611476242542267, 0.09252717345952988, -0.17580796778202057, -0.08189008384943008, 0.03957559913396835, -0.16714344918727875, -0.5780373215675354, -0.2460925430059433, 0.04412265866994858, -0.32625389099121094, -0.2550867199897766, 0.15884177386760712, 0.029433995485305786, -0.16992613673210144, 0.17299647629261017, 0.2292766273021698, 0.6286781430244446, -0.43886151909828186, -0.18500018119812012, -0.10761889815330505, -0.3222369849681854, -0.15262532234191895, 0.12096085399389267, 0.41602078080177307, 0.2905067801475525, 0.4932437837123871, 0.2830926179885864, -0.026001110672950745, 0.10474313050508499, -0.37736472487449646, 0.054451633244752884, -0.27542707324028015, 0.28514426946640015, 0.11039283871650696, -0.22095873951911926, 0.15785686671733856, -0.39220502972602844, -0.05006104335188866, -0.12310875952243805, -0.09991275519132614, -0.0815800353884697, -0.026580799371004105, 0.04209123179316521, -0.1918628215789795, -0.1675719916820526, -0.6334232687950134, -0.4681072533130646, 0.34189045429229736, -0.21221688389778137, 0.10471164435148239, 0.057700417935848236, -0.26068320870399475, 0.16616985201835632, -0.15556222200393677, 0.027958815917372704, -0.16115091741085052, -0.08060179650783539, 0.027159949764609337, -0.3650360703468323, -0.16097712516784668, -0.4260775148868561, -0.13156619668006897, 0.21172581613063812, -0.08573532104492188, -0.018161123618483543, -0.25038737058639526, 0.07165082544088364, -0.1331431120634079, 0.11693111807107925, 0.2675018608570099, 0.19170334935188293, -0.2375509887933731, 0.09199431538581848, -0.030880149453878403, -0.23764334619045258, 0.2405966818332672, 0.33344313502311707, 0.39254119992256165, 0.14052002131938934, 0.030540434643626213, 0.19585371017456055, 0.7353604435920715, 0.25642937421798706, 0.023826198652386665, 0.18662676215171814, 0.2621121108531952, 0.024636724963784218, 0.1601390391588211, -0.1698179692029953, 0.25950542092323303, -0.17571234703063965, 0.44925352931022644, 0.12856188416481018, 0.18453159928321838, -0.2333080768585205, -0.24578683078289032, -0.01603509858250618, -0.2677399814128876, -0.20620539784431458, 0.24365288019180298, -0.2004666030406952, -0.005440595094114542, -0.09856291115283966, -0.38829508423805237, -0.25989192724227905, -0.032904382795095444, 0.017790714278817177, 0.28790998458862305, 0.002849042182788253, 0.16430731117725372, -0.31946712732315063, -0.055637311190366745, -0.48451465368270874, 0.14956332743167877, 0.07221481204032898, 0.2689269483089447, -0.06599996238946915, -0.1026056706905365, -0.11336396634578705, 0.08600826561450958, 0.6084613800048828, -0.4609886109828949, -0.12899239361286163, 0.029674870893359184, 0.09475585073232651, -0.12481461465358734, -0.1021866649389267, -0.26836177706718445, -0.06343196332454681, 0.46247297525405884, 0.3793349266052246, -0.3683629035949707, -0.30867573618888855, 0.25277993083000183, -0.2167631834745407, -0.05535212159156799, -0.23541618883609772, -0.21979860961437225, -0.18149450421333313, -0.15289396047592163, 0.09308241307735443, 0.22387337684631348, 0.17905868589878082, -0.10410675406455994, 0.12050031125545502, -0.07052155584096909, 0.20699188113212585, 0.2310265600681305, 0.03241977468132973, -0.024891743436455727, 0.20212391018867493, 0.1299929916858673, 0.07291918247938156, -0.128123477101326, -0.3435958921909332, 0.05205639824271202, -0.07871635258197784, 0.16748005151748657, 0.2720288336277008, 0.03894692286849022, 0.6369621157646179, 0.09319660067558289, 0.34834811091423035, 0.24644911289215088, -0.10003046691417694, 0.028285283595323563, -0.2644094228744507, 0.1170211136341095, 0.1925375908613205, 0.24081139266490936, -0.1332414746284485, -0.18242290616035461, 0.6144337058067322, 0.3210254907608032, -0.07534204423427582, 0.39156925678253174, -0.48021143674850464, -0.2988831400871277, 0.30478909611701965, 0.0476769395172596, 0.6780330538749695, -0.008658384904265404, 0.2885782718658447, 0.10952404141426086, -0.35490506887435913, 0.5664435625076294, -0.34624338150024414, 0.15785473585128784, -0.10396938771009445, -0.19105449318885803, 0.0071837687864899635, -0.08794985711574554, -0.048731107264757156, 0.26887255907058716, -0.34670794010162354, 0.34354403614997864, 0.456682413816452, -0.2356584668159485, -0.08080495148897171, 0.5609846711158752, -0.04154358431696892, -0.40955615043640137, -0.08284275233745575, 0.0758705735206604, -0.28342726826667786, 0.1413334608078003, -0.17460066080093384, -0.2745567262172699, 0.25787752866744995, 0.10646621137857437, -0.27389323711395264, 0.0917183980345726, -0.05002882331609726, 0.1728329062461853, 0.27562350034713745, 0.03175861015915871, 0.19174398481845856, -0.2028181552886963, -0.06302611529827118, -0.19644136726856232, -0.07966659218072891, 0.15450090169906616, -0.14524787664413452, 0.10553400963544846, 0.087846539914608, -0.20984157919883728, 0.5108206272125244, -0.05306711792945862, -0.10169688612222672, -0.26747050881385803, 0.2604365944862366, -0.4042939245700836, -0.17751352488994598, 0.10693441331386566, 0.13696110248565674, 0.2079698145389557, -0.2924814224243164, 0.5627287030220032, 0.09027507156133652, -0.0058631207793951035, 0.055334482342004776, 0.15226776897907257, -0.2723241150379181, 0.5453818440437317, -0.2550187110900879, 0.011216804385185242, 0.05190224200487137, 0.011895696632564068, -0.1488829255104065, -0.36376771330833435, -0.08748186379671097, 0.27418428659439087, 0.09738295525312424, -0.00017570797353982925, -0.013708430342376232, 0.14800380170345306, -0.1133343055844307, -0.020417794585227966, -0.12666310369968414, -0.35359564423561096, 0.17925116419792175, 0.2160952091217041, 0.25573790073394775, 0.21601082384586334, -0.07465722411870956, -0.24117881059646606, -0.19481340050697327, 0.4819282293319702, 0.027668677270412445, 0.2639501988887787, -0.13278523087501526, 0.35004258155822754, 0.4172840416431427, 0.542597234249115, -0.2901419997215271, 0.0194889884442091, 0.04544743895530701, 0.11256490647792816, 0.4598180949687958, -0.24196583032608032, 0.07676412165164948, 0.28588053584098816, -0.13011077046394348, 0.0701565220952034, -0.36212125420570374, -0.17070747911930084, -0.26530247926712036, 0.10153089463710785, 0.10833431780338287, 0.04718958958983421, 0.0031762130092829466, -0.019687095656991005, -0.07955950498580933, -0.0957367867231369, 0.01848340965807438, -0.3760216236114502, -0.0982823371887207, 0.15372022986412048, -0.06953983753919601, 0.042393699288368225, 0.12043562531471252, 0.20509065687656403, 0.046698667109012604, -0.06932839751243591, 0.32367485761642456, 0.1258392035961151, 0.07482589781284332, -0.15220259130001068, -0.16026081144809723, 0.007074892520904541, 0.22868448495864868, 0.1364775002002716, -0.06157491356134415, 0.1617528200149536, -0.27166223526000977, -0.13609030842781067, 0.24081115424633026, 0.14597104489803314, 0.07060426473617554, -0.15793418884277344, 0.1163192018866539, 0.14452947676181793, -0.11241628974676132, 0.0010164926061406732, 0.2037462443113327, 0.056546203792095184, 0.1279762089252472, 0.15175670385360718, 0.16456347703933716, 0.45382311940193176, -0.21429572999477386, 0.036483678966760635, 0.14220228791236877, -0.3709934949874878, 0.11530382931232452, -0.13379819691181183, -0.025103149935603142, 0.024294903501868248, 0.1567714512348175, -0.00945306196808815, 0.21674704551696777, -0.34670162200927734, 0.28668415546417236, 0.3190627992153168, -0.03136817365884781, 0.05513693019747734, 0.19727487862110138, -0.0020196037366986275, -0.16688472032546997, 0.27943548560142517, 0.07130037993192673, 0.3188149929046631, 0.055346015840768814, 0.8951666951179504, -0.24943041801452637, -0.4379865825176239, 0.08847995847463608, 0.3954363167285919, 0.024620000272989273, -0.22906756401062012, 0.10322873294353485, 0.08457647264003754, -0.33028456568717957, 0.09990270435810089, 0.041744738817214966, -0.12678350508213043, 0.6448745727539062, -0.08233334869146347, 0.14766809344291687, -0.15032020211219788, -0.23305849730968475, 0.32927995920181274, 0.1781778186559677, -0.11302302777767181, 0.013508823700249195, 0.09902048110961914, -0.10855212062597275, -0.1989584118127823, 0.5276769995689392, 0.2439906746149063, 0.05592787265777588, -0.4059012234210968, 0.007024501450359821, 0.11586368829011917, 0.14389555156230927, -0.09145320951938629, 0.10737698525190353, 0.4503694474697113, -0.14932692050933838, 0.21223920583724976, 0.03401578962802887, -0.06872895359992981, -0.09852878004312515, 0.03311347961425781, -0.45605725049972534, -0.3647691309452057, 0.45689040422439575, 0.10242580622434616, -0.08989333361387253, 0.04291737452149391, 0.3086782991886139, -0.6284584403038025, -0.18854011595249176, 0.2987588346004486, -0.13009707629680634, -0.07465958595275879, -0.2261146456003189, 0.0887196958065033, 0.07111946493387222, 0.585948646068573, 0.21885712444782257, 0.35041648149490356, -0.40086090564727783, -0.09598887711763382, -0.6091049909591675, -0.20071937143802643, 0.11933484673500061, -0.08799099177122116, -0.21377229690551758, 0.1967516392469406, -0.10631401091814041, 0.2746656835079193, -0.1899932622909546, 0.11173830181360245, 0.18231849372386932, 0.01936580054461956, -0.4103379249572754, 0.1525803804397583, 0.09875068813562393, 0.011352013796567917, 0.3067415654659271, -0.09572187066078186, -0.17017193138599396, 0.08457709103822708, -0.07569780200719833, -0.08605175465345383, -0.051421258598566055, 0.2711253762245178, 0.17782904207706451, 0.04539410024881363, 0.055525828152894974, 0.4602868854999542, -0.04975072667002678, -0.1465178281068802, -0.020975535735487938, -0.021389184519648552, -0.3999607264995575, -0.17749764025211334, -0.08906100690364838, 0.17003083229064941, -0.33483171463012695, 0.24291029572486877, -0.04528403654694557, -0.10608486086130142, 0.120690256357193, 0.012443204410374165, -0.13162140548229218, -0.03240368515253067, 0.03830329328775406, 0.18737448751926422, 0.22846567630767822, 0.736083984375, 0.03749365732073784, 0.08689601719379425, -0.09599608927965164, -0.05491280183196068, 0.054358530789613724, -0.09335985034704208, -0.2339468002319336, -0.1753934770822525, 0.0794786661863327, 0.09167445451021194, 0.07787389308214188, -0.671375572681427, -0.3535114526748657, 0.010525512509047985, 0.010181042365729809, 0.07833046466112137, 0.7260477542877197, 0.07258034497499466, -0.15513920783996582, -0.15334580838680267, -0.3561757504940033, 0.060089774429798126, -0.2716902494430542, 0.1818743348121643, -0.4164091944694519 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
@zphang is having different batch sizes per task actually helpful? Would be interesting to know as it's not something I've come across as a technique used by any MTL papers.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
30
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. @zphang is having different batch sizes per task actually helpful? Would be interesting to know as it's not something I've come across as a technique used by any MTL papers.
[ -0.09692893177270889, -0.46223804354667664, -0.07604547590017319, 0.01724984496831894, -0.30476319789886475, 0.05286415293812752, 0.22632136940956116, 0.0648956298828125, 0.27689728140830994, -0.11790042370557785, -0.17251333594322205, 0.2042180895805359, -0.25846952199935913, 0.4444766044616699, 0.4080321490764618, -0.4105130434036255, 0.014127123169600964, -0.27331623435020447, -0.2500169575214386, 0.39677613973617554, 0.05055159702897072, 0.08481944352388382, -0.17099398374557495, 0.06342416256666183, -0.24067102372646332, -0.13419800996780396, -0.3394121527671814, 0.08957803249359131, 0.1340940296649933, 0.02739320881664753, 0.14990690350532532, 0.3057299554347992, 0.027069218456745148, 0.1451931744813919, -0.00011703866766765714, -0.10317951440811157, 0.22462137043476105, -0.05497625097632408, 0.12323480099439621, -0.03875386714935303, 0.11764135956764221, -0.4690670371055603, 0.0013242801651358604, -0.09075816720724106, 0.27294179797172546, 0.19905000925064087, 0.24085493385791779, 0.05113803222775459, 0.3071523606777191, -0.30362674593925476, 0.08462025225162506, 0.2716977000236511, -0.23924730718135834, 0.3131468892097473, -0.04093201458454132, -0.03336275741457939, 0.06543541699647903, 0.2598794102668762, 0.7037147283554077, -0.5388655066490173, 0.003553200513124466, 0.11596129834651947, 0.028103794902563095, -0.021970883011817932, 0.13279123604297638, -0.07223263382911682, -0.2183319479227066, -0.2454032301902771, -0.34902623295783997, 0.20940133929252625, -0.05419535934925079, -0.08674006909132004, -0.21434028446674347, -0.5842272043228149, -0.018567940220236778, 0.05227060988545418, -0.3170750141143799, 0.13627299666404724, -0.20855510234832764, -0.07745742797851562, -0.19316919147968292, -0.10364782065153122, -0.019546326249837875, 0.10086861997842789, 0.3700273633003235, 0.5515711307525635, 0.05214443802833557, 0.16430160403251648, 0.3478100895881653, 0.02283669263124466, -0.10691897571086884, -0.24575915932655334, 0.29284951090812683, 0.12592418491840363, -0.35657069087028503, -0.31134751439094543, 0.18628418445587158, -0.2371801882982254, 0.285808265209198, 0.06318151950836182, 0.13804517686367035, 0.039516106247901917, -0.10145347565412521, 0.20296595990657806, 0.38573935627937317, 0.09933821111917496, -0.08045084774494171, -0.14554166793823242, 0.02665860950946808, -0.16947370767593384, 0.19106633961200714, 0.15759578347206116, 0.15331681072711945, -0.08121456205844879, -0.46981364488601685, -0.13959696888923645, -0.24632896482944489, 0.12414982169866562, -0.18019212782382965, -0.5548356175422668, -0.06913377344608307, -0.20435194671154022, 0.027278970927000046, -0.1281009018421173, -0.2890937924385071, 0.13472577929496765, -0.0876297578215599, 0.2149021029472351, -0.13607260584831238, 0.0736110582947731, -0.04291783645749092, 0.06125001236796379, -0.470269113779068, -0.10851015150547028, 0.2724400460720062, 0.1882496476173401, 0.0766226127743721, 0.1188499927520752, -0.04511119797825813, 0.11725517362356186, 0.3278568685054779, 0.01572209596633911, -0.0059085325337946415, 0.01859467290341854, 0.2151927798986435, -0.3085087835788727, -0.09372963011264801, 0.2627731263637543, -0.382362425327301, -0.04949548467993736, -0.22280055284500122, -0.17501120269298553, 0.15861134231090546, 0.013053442351520061, -0.13337022066116333, -0.246748149394989, -0.5358601808547974, 0.6971766948699951, 0.029376566410064697, -0.06125430762767792, -0.1539636254310608, 0.24740588665008545, -0.21105189621448517, -0.14691412448883057, 0.08802849799394608, 0.06152286008000374, -0.07157918065786362, -0.34737658500671387, -0.03984463959932327, -0.08777172863483429, 0.016240296885371208, 0.33794912695884705, -0.19569700956344604, 0.26970529556274414, 0.07623108476400375, 0.12101157009601593, 0.2956900894641876, -0.4761197566986084, -0.042139604687690735, 0.39082038402557373, -0.029753265902400017, 0.35634148120880127, 0.12997478246688843, 0.2487468123435974, -0.01919848844408989, -0.17189978063106537, 0.27164551615715027, 0.9539662599563599, -0.4012835621833801, 0.032429929822683334, -0.1284986287355423, -0.4258193075656891, 0.5435500144958496, 0.288767546415329, 0.0950731560587883, -0.29588931798934937, -0.08115857094526291, 0.05353771150112152, 0.14724500477313995, 0.11203324049711227, 0.12126505374908447, -0.06787168234586716, -0.3955807089805603, 0.023653090000152588, -0.19152028858661652, -0.10705969482660294, -0.4190424978733063, -0.0004107767890673131, -0.09667208045721054, 0.2867058515548706, 0.345079243183136, -0.0710882917046547, 0.30363792181015015, -0.31333696842193604, -0.08020184189081192, -0.3391849398612976, 0.0749736800789833, -0.11293651163578033, 0.008268136531114578, -0.302785724401474, -0.060413725674152374, 0.3214499056339264, 0.13272054493427277, -0.1588383913040161, -0.46373865008354187, 0.2198476642370224, -0.0665503740310669, 0.07297101616859436, -0.006660561077296734, 0.6013069748878479, -0.24156352877616882, -0.0950026661157608, 0.14528316259384155, 0.2679246962070465, -0.299379825592041, 0.23697495460510254, 0.27708864212036133, 0.2627179026603699, 0.21460787951946259, 0.04104560241103172, 0.18638287484645844, -0.10739437490701675, -0.06513883918523788, -0.21716973185539246, -0.07372776418924332, 0.46423739194869995, -0.13604377210140228, 0.4431437849998474, -0.04130024462938309, -0.05664600059390068, -0.23851493000984192, -0.04269196093082428, -0.20221076905727386, 0.3690052926540375, 0.3462929129600525, -0.189581960439682, 0.297355055809021, -0.101374551653862, -0.3327431380748749, 0.15535064041614532, 0.11895082145929337, 0.09320942312479019, 0.05557192862033844, -0.012828504666686058, -0.03845524787902832, -0.29319971799850464, -0.14500245451927185, 0.3609071969985962, 0.6997565031051636, 0.2174215465784073, 0.1800074726343155, 0.027522020041942596, -0.1548192799091339, -0.13634850084781647, 0.015728579834103584, 0.17999443411827087, -0.043210454285144806, 0.17419350147247314, 0.0200825035572052, 0.04398098960518837, 0.05878245085477829, -0.13080257177352905, 0.14341099560260773, -0.20149332284927368, -0.04728226736187935, 0.007520558312535286, -0.2275296151638031, -0.6512371897697449, -0.31002506613731384, 0.024343211203813553, -0.3236200213432312, -0.24214604496955872, 0.07698982208967209, -0.004847442731261253, -0.14679484069347382, 0.09195481240749359, 0.19153852760791779, 0.6404311656951904, -0.4318293631076813, -0.1751059889793396, -0.024588700383901596, -0.25083765387535095, -0.15639041364192963, 0.09369321912527084, 0.469990611076355, 0.24401967227458954, 0.4443300664424896, 0.2816877067089081, -0.06684470921754837, 0.10976660996675491, -0.36882197856903076, 0.022475531324744225, -0.23586638271808624, 0.2514665424823761, 0.1361657828092575, -0.25158631801605225, 0.19181467592716217, -0.33062925934791565, -0.05950085073709488, -0.10711812227964401, -0.0802377238869667, -0.14460338652133942, -0.017944110557436943, 0.06091868877410889, -0.21953335404396057, -0.1628664880990982, -0.6563559174537659, -0.42742374539375305, 0.29517605900764465, -0.235421821475029, 0.09205694496631622, 0.04310010001063347, -0.2766208052635193, 0.16108925640583038, -0.11539862304925919, 0.005505555309355259, -0.1104394719004631, -0.12340112775564194, -0.06430736184120178, -0.3830779492855072, -0.148684561252594, -0.3794017434120178, -0.032763343304395676, 0.21915273368358612, -0.013218380510807037, -0.06567750871181488, -0.138601154088974, 0.09972263127565384, -0.19404838979244232, 0.0751047134399414, 0.288838267326355, 0.07945176959037781, -0.3194800913333893, 0.13742317259311676, -0.04164227098226547, -0.24689322710037231, 0.16785089671611786, 0.34957805275917053, 0.4843294620513916, 0.184091717004776, 0.022530699148774147, 0.16180577874183655, 0.7138546705245972, 0.23926393687725067, 0.07667479664087296, 0.12391604483127594, 0.27780720591545105, -0.027644306421279907, 0.15621866285800934, -0.15137889981269836, 0.34679922461509705, -0.13601264357566833, 0.45005154609680176, 0.12394843995571136, 0.21397538483142853, -0.19466973841190338, -0.2623998522758484, -0.008468929678201675, -0.26845189929008484, -0.1703481525182724, 0.2559320032596588, -0.255561888217926, 0.049365218728780746, -0.06720167398452759, -0.47130417823791504, -0.26852747797966003, -0.04071032628417015, -0.04501020535826683, 0.3134006857872009, -0.07623044401407242, 0.1900920867919922, -0.3363177180290222, -0.13906937837600708, -0.4542549252510071, 0.1405397355556488, 0.0639505535364151, 0.3106072247028351, -0.04748280718922615, -0.06475035846233368, -0.08007989823818207, 0.1626020222902298, 0.6325609683990479, -0.48288866877555847, -0.14991648495197296, 0.014306745491921902, 0.10305792838335037, -0.08693014085292816, -0.1750022917985916, -0.24498899281024933, 0.004539347719401121, 0.4797534644603729, 0.37007424235343933, -0.32539159059524536, -0.280244380235672, 0.2273396998643875, -0.20513874292373657, -0.050977673381567, -0.22064723074436188, -0.2203364372253418, -0.12599420547485352, -0.0952439159154892, 0.07316458225250244, 0.17764057219028473, 0.24161885678768158, -0.15876391530036926, 0.18788036704063416, -0.020658278837800026, 0.2550891935825348, 0.1945965588092804, -0.011927830055356026, -0.09128247946500778, 0.2523726522922516, 0.11457360535860062, -0.00043937304872088134, -0.07985694706439972, -0.24603654444217682, 0.0720708891749382, -0.10971274971961975, 0.19002987444400787, 0.3117484450340271, -0.054916370660066605, 0.6821688413619995, 0.07290902733802795, 0.3269191086292267, 0.23021718859672546, -0.09951464086771011, 0.043482255190610886, -0.19248592853546143, 0.1378147006034851, 0.176866814494133, 0.22169746458530426, -0.18922339379787445, -0.22393718361854553, 0.5282303094863892, 0.36822330951690674, -0.021432960405945778, 0.3635035753250122, -0.6384041905403137, -0.2950289249420166, 0.24399586021900177, 0.03697944059967995, 0.620188295841217, 0.0012559413444250822, 0.3311546742916107, -0.028298621997237206, -0.3404570519924164, 0.5145266056060791, -0.3430119454860687, 0.1747768521308899, -0.06714142858982086, -0.13017140328884125, 0.02278958074748516, -0.10523965954780579, -0.03440069034695625, 0.3134416341781616, -0.2745059132575989, 0.33596789836883545, 0.4565870761871338, -0.2524252235889435, -0.12036897242069244, 0.5019969940185547, -0.0001571260072523728, -0.3571082055568695, -0.026804259046912193, 0.025746650993824005, -0.24538463354110718, 0.13652046024799347, -0.22022758424282074, -0.22735601663589478, 0.22728382050991058, 0.15298515558242798, -0.3146010935306549, 0.04300118237733841, -0.022225186228752136, 0.19116652011871338, 0.2674506902694702, 0.009534700773656368, 0.2880604863166809, -0.20601926743984222, -0.11121570318937302, -0.15486107766628265, -0.06647585332393646, 0.09129443764686584, -0.1101321280002594, 0.09081549942493439, 0.08554155379533768, -0.18792083859443665, 0.46513205766677856, -0.08558332920074463, -0.04987647756934166, -0.23923133313655853, 0.24239233136177063, -0.39156830310821533, -0.22097696363925934, 0.20412589609622955, 0.1990998089313507, 0.23439107835292816, -0.3496170938014984, 0.5967291593551636, 0.11620349436998367, 0.06519223004579544, 0.0013030694099143147, 0.17796778678894043, -0.1788128912448883, 0.5910230875015259, -0.31430843472480774, -0.025046851485967636, 0.040158215910196304, -0.061759501695632935, -0.03130752220749855, -0.4336966872215271, -0.07607518136501312, 0.29465752840042114, 0.14731968939304352, 0.04011399298906326, 0.07178951054811478, 0.06344687938690186, -0.024860626086592674, 0.006633422803133726, -0.08219339698553085, -0.27645134925842285, 0.21565330028533936, 0.21482659876346588, 0.176828533411026, 0.24704387784004211, -0.06320678442716599, -0.2921839654445648, -0.11573295295238495, 0.46784716844558716, -0.030525602400302887, 0.21432989835739136, -0.14435698091983795, 0.429349809885025, 0.4802251160144806, 0.5306455492973328, -0.20600438117980957, 0.08634356409311295, 0.11954691261053085, 0.09168463945388794, 0.4120355248451233, -0.11529985815286636, 0.01623450592160225, 0.3485606014728546, -0.20053036510944366, 0.0638982504606247, -0.3479441702365875, -0.11550626158714294, -0.23809675872325897, 0.1280660182237625, 0.1398671418428421, 0.02418104000389576, 0.040377579629421234, -0.10755127668380737, -0.059195809066295624, -0.025316447019577026, 0.017238257452845573, -0.3293403685092926, -0.06795470416545868, 0.10453077405691147, -0.025514516979455948, 0.015602421015501022, 0.09028758108615875, 0.3568158745765686, 0.0729781836271286, -0.09809110313653946, 0.34946945309638977, 0.13369199633598328, 0.1295308917760849, -0.1518435925245285, -0.21585892140865326, -0.04204433411359787, 0.2591365873813629, 0.17299015820026398, -0.14142635464668274, 0.12014427036046982, -0.23301859200000763, -0.12351936846971512, 0.20526495575904846, 0.2321888953447342, 0.07684151828289032, -0.12201099842786789, 0.13692405819892883, 0.06485427916049957, -0.04432749003171921, -0.004679468926042318, 0.18947133421897888, 0.0924200788140297, 0.02123846486210823, 0.16045896708965302, 0.18984226882457733, 0.4918515086174011, -0.26601511240005493, 0.025715693831443787, 0.08390598744153976, -0.34789299964904785, 0.10506028681993484, -0.15809723734855652, -0.05164264142513275, 0.019299570471048355, 0.19063155353069305, -0.053772859275341034, 0.1364482343196869, -0.2919658422470093, 0.35888415575027466, 0.25594571232795715, 0.058684926480054855, -0.01334156934171915, 0.139642134308815, -0.039308253675699234, -0.20437964797019958, 0.33291301131248474, 0.0015009588096290827, 0.32949626445770264, 0.022727901116013527, 0.8239643573760986, -0.22451326251029968, -0.4310360848903656, 0.1678004264831543, 0.4035535752773285, 0.02263057231903076, -0.24922850728034973, 0.12385416775941849, 0.10800202190876007, -0.31046244502067566, 0.1577770709991455, 0.009082038886845112, -0.11470063030719757, 0.6107502579689026, -0.08435121923685074, 0.16867586970329285, -0.18480560183525085, -0.2317778617143631, 0.36075058579444885, 0.06150565296411514, -0.11912164837121964, 0.03143597021698952, 0.014173964969813824, -0.0898064374923706, -0.13866391777992249, 0.5110018849372864, 0.19786576926708221, 0.08152861148118973, -0.3625519573688507, -0.08625281602144241, 0.12035294622182846, 0.23069070279598236, -0.1393299102783203, 0.12428078055381775, 0.5147860050201416, -0.18943631649017334, 0.2248300313949585, -0.031415846198797226, -0.027001993730664253, -0.07785646617412567, 0.05451812222599983, -0.4602070748806, -0.4399226903915405, 0.43461495637893677, 0.041092630475759506, -0.11985684186220169, 0.05708609148859978, 0.39168450236320496, -0.6262607574462891, -0.2196517139673233, 0.336073637008667, -0.15085943043231964, -0.08783341944217682, -0.24786457419395447, 0.05479937419295311, 0.061759013682603836, 0.6547300219535828, 0.2686654031276703, 0.3616766035556793, -0.36749380826950073, -0.13172876834869385, -0.5774889588356018, -0.18853674829006195, 0.09687331318855286, -0.026181375607848167, -0.28758934140205383, 0.23825061321258545, -0.11827666312456131, 0.2193918079137802, -0.12811683118343353, 0.10893595218658447, 0.17497029900550842, 0.07805264741182327, -0.42623022198677063, 0.11324404180049896, 0.09243196994066238, 0.08165384083986282, 0.27569735050201416, -0.08105088025331497, -0.14506039023399353, 0.036099594086408615, -0.12800276279449463, -0.14890070259571075, -0.09801267087459564, 0.32596859335899353, 0.2252088487148285, 0.09227728843688965, 0.03513956442475319, 0.5028653740882874, 0.028904732316732407, -0.19265539944171906, -0.07568595558404922, 0.004324793815612793, -0.35453471541404724, -0.16355323791503906, -0.05845796316862106, 0.16321328282356262, -0.4496604800224304, 0.3605158030986786, 0.049265794456005096, -0.11139561980962753, 0.09741564095020294, 0.02194923907518387, -0.0988563820719719, -0.09097585827112198, 0.08899512141942978, 0.15876562893390656, 0.28199073672294617, 0.7396939396858215, 0.01803232915699482, 0.04638202488422394, -0.09544392675161362, 0.010542252101004124, -0.019518977031111717, -0.0713992640376091, -0.21920378506183624, -0.20084305107593536, 0.07530664652585983, -0.03850206732749939, 0.08397145569324493, -0.6680580973625183, -0.40913698077201843, 0.03187616914510727, 0.04847078025341034, 0.07243166118860245, 0.7247983813285828, 0.05399908870458603, -0.18400447070598602, -0.11854583024978638, -0.3672243654727936, 0.03123132511973381, -0.2750071883201599, 0.1617082804441452, -0.3264085054397583 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
> @zphang is having different batch sizes per task actually helpful? Would be interesting to know as it's not something I've come across as a technique used by any MTL papers. I think having different batch sizes per task is particularly helpful in some scenarios where each task has different amount of data. For example, the problem I'm currently facing is one task has tens of thousands of samples while one task has a couple hundreds. I think in this case different batch size could help. But if using the same batch size is a lot simpler to implement, I guess it makes sense to go with that.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
108
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. > @zphang is having different batch sizes per task actually helpful? Would be interesting to know as it's not something I've come across as a technique used by any MTL papers. I think having different batch sizes per task is particularly helpful in some scenarios where each task has different amount of data. For example, the problem I'm currently facing is one task has tens of thousands of samples while one task has a couple hundreds. I think in this case different batch size could help. But if using the same batch size is a lot simpler to implement, I guess it makes sense to go with that.
[ -0.15286588668823242, -0.48510846495628357, -0.08828972280025482, 0.043889012187719345, -0.30230480432510376, 0.025921138003468513, 0.26860523223876953, 0.0912453681230545, 0.27223002910614014, -0.07411877810955048, -0.16743925213813782, 0.12118729948997498, -0.25727298855781555, 0.4499777555465698, 0.36345285177230835, -0.406209260225296, 0.030634230002760887, -0.2730120122432709, -0.22878915071487427, 0.4066847264766693, 0.03692016005516052, 0.05140312388539314, -0.13749800622463226, 0.04864584654569626, -0.23849353194236755, -0.15247581899166107, -0.3196727931499481, 0.05557989329099655, 0.19754455983638763, 0.04465049132704735, 0.11467429995536804, 0.30155107378959656, 0.04731499031186104, 0.1570083647966385, -0.0001164070563390851, -0.0821220800280571, 0.22991499304771423, -0.020130082964897156, 0.07947643846273422, -0.01903468184173107, 0.13090038299560547, -0.5154121518135071, -0.03749954327940941, -0.1042584776878357, 0.2734789550304413, 0.21355992555618286, 0.265973299741745, 0.03253597766160965, 0.25440800189971924, -0.29471635818481445, 0.09862415492534637, 0.3221178352832794, -0.297036737203598, 0.2996925115585327, -0.010349895805120468, -0.014742602594196796, 0.0532819889485836, 0.2629302442073822, 0.6999949216842651, -0.4993143677711487, 0.001270392327569425, 0.1606196165084839, 0.027350544929504395, 0.02736632153391838, 0.10411415249109268, -0.05754983797669411, -0.1475425660610199, -0.22595557570457458, -0.30797481536865234, 0.24689747393131256, -0.08192521333694458, -0.06607388705015182, -0.23695650696754456, -0.6513190865516663, -0.007323681842535734, 0.04993271455168724, -0.350864052772522, 0.1869450807571411, -0.1884656399488449, -0.09649728983640671, -0.15846073627471924, -0.05361097678542137, -0.007652486674487591, 0.06648122519254684, 0.3757088780403137, 0.5052570700645447, 0.013429636135697365, 0.17369171977043152, 0.34670865535736084, 0.031212300062179565, -0.10651310533285141, -0.2447148561477661, 0.2853567600250244, 0.10458478331565857, -0.37025871872901917, -0.2722153663635254, 0.16172201931476593, -0.264308363199234, 0.27126145362854004, 0.02366454340517521, 0.17093400657176971, 0.021655602380633354, -0.06886782497167587, 0.17013129591941833, 0.4097623825073242, 0.11251693964004517, -0.09519115835428238, -0.1455383002758026, 0.041093651205301285, -0.1776208132505417, 0.18310900032520294, 0.20799778401851654, 0.14859317243099213, -0.13576622307300568, -0.465637743473053, -0.08216284960508347, -0.25629356503486633, 0.14932109415531158, -0.20054614543914795, -0.5421445369720459, -0.046839382499456406, -0.23184746503829956, -0.009074853733181953, -0.11373739689588547, -0.2909088134765625, 0.14404429495334625, -0.11999616771936417, 0.2271474003791809, -0.16093473136425018, 0.0743538960814476, -0.057899102568626404, 0.0046853795647621155, -0.44629862904548645, -0.11514412611722946, 0.269219309091568, 0.1677376627922058, 0.08348635584115982, 0.0706985667347908, -0.0673728734254837, 0.11798155307769775, 0.2836701273918152, -0.004395791329443455, -0.03297993540763855, 0.039601776748895645, 0.22117310762405396, -0.27689042687416077, -0.13525167107582092, 0.3050743639469147, -0.39978766441345215, -0.02252546325325966, -0.23434102535247803, -0.19428612291812897, 0.18709291517734528, 0.02074514701962471, -0.1286899596452713, -0.22317051887512207, -0.5973212122917175, 0.6830009818077087, 0.06751760095357895, -0.0465238094329834, -0.15667112171649933, 0.17821672558784485, -0.19402417540550232, -0.17894063889980316, 0.011591852642595768, 0.08312398195266724, -0.06165257841348648, -0.34613537788391113, 0.0005165084730833769, -0.11840606480836868, 0.015426012687385082, 0.3973839581012726, -0.1924615204334259, 0.2865569591522217, 0.09445211291313171, 0.12006434053182602, 0.26640889048576355, -0.47417494654655457, -0.04938250780105591, 0.4507997930049896, -0.019094614312052727, 0.350606232881546, 0.16728410124778748, 0.26022228598594666, 0.00202055461704731, -0.17475317418575287, 0.23725663125514984, 0.8971971273422241, -0.4060545861721039, 0.08358033746480942, -0.13938412070274353, -0.42281413078308105, 0.5240347981452942, 0.31012481451034546, 0.1059497594833374, -0.3104816973209381, -0.06986027956008911, 0.04622741788625717, 0.13682259619235992, 0.0981062576174736, 0.12557409703731537, -0.05776312202215195, -0.4325704574584961, -0.0024137741420418024, -0.19192562997341156, -0.06833650171756744, -0.4964830279350281, 0.0032334611751139164, -0.15541410446166992, 0.2653368413448334, 0.373635470867157, -0.05858943983912468, 0.3415443003177643, -0.3247836232185364, -0.06826767325401306, -0.31966665387153625, 0.07443998754024506, -0.09567189961671829, 0.03928600996732712, -0.3197909891605377, -0.06206311285495758, 0.32378122210502625, 0.12032081186771393, -0.1444503515958786, -0.4255875051021576, 0.19232165813446045, -0.03772430494427681, 0.08093028515577316, 0.01669754460453987, 0.5754877924919128, -0.2911284863948822, -0.12075066566467285, 0.10930120944976807, 0.24802954494953156, -0.28013497591018677, 0.21860824525356293, 0.21097511053085327, 0.2668554484844208, 0.25490397214889526, 0.12753546237945557, 0.20218488574028015, -0.12013927847146988, -0.09720587730407715, -0.21976572275161743, -0.032139670103788376, 0.44920292496681213, -0.09584470093250275, 0.467934250831604, -0.041421566158533096, -0.10826423019170761, -0.22867219150066376, -0.042063381522893906, -0.20951521396636963, 0.3518669307231903, 0.33880478143692017, -0.1527584046125412, 0.32122376561164856, -0.08753787726163864, -0.321072518825531, 0.1529049128293991, 0.09710444509983063, 0.07538942992687225, 0.05987492576241493, -0.015521951019763947, -0.04942828044295311, -0.2951929271221161, -0.16740119457244873, 0.34958815574645996, 0.7306098341941833, 0.22298826277256012, 0.19981880486011505, -0.001651662983931601, -0.1664976179599762, -0.16180939972400665, -0.007260547950863838, 0.19259802997112274, -0.01902139000594616, 0.20632898807525635, 0.026077723130583763, 0.010703319683670998, 0.047556839883327484, -0.14258535206317902, 0.12812817096710205, -0.19433870911598206, -0.03902191296219826, 0.001226053573191166, -0.2181590497493744, -0.6630987524986267, -0.34492480754852295, 0.039668839424848557, -0.27959340810775757, -0.2296169549226761, 0.09637480229139328, 0.006200928241014481, -0.2022167593240738, 0.09744171798229218, 0.2544863224029541, 0.655793309211731, -0.420545369386673, -0.12200190126895905, -0.011811833828687668, -0.24652491509914398, -0.18224340677261353, 0.08044201135635376, 0.47870898246765137, 0.19717244803905487, 0.4226313829421997, 0.2554946839809418, -0.07075616717338562, 0.08091502636671066, -0.3518062233924866, 0.02570156194269657, -0.24721641838550568, 0.24243222177028656, 0.1162124052643776, -0.23505346477031708, 0.19698381423950195, -0.3383501172065735, -0.03175223991274834, -0.11529609560966492, -0.07352671027183533, -0.18286468088626862, -0.022766392678022385, 0.045827172696590424, -0.21434544026851654, -0.1471017599105835, -0.6761796474456787, -0.4236726462841034, 0.3288803994655609, -0.25891757011413574, 0.08312948048114777, 0.0686505064368248, -0.24256260693073273, 0.11010275036096573, -0.12082315981388092, 0.020177340134978294, -0.10749764740467072, -0.13586480915546417, -0.10517019033432007, -0.3834724426269531, -0.16408555209636688, -0.4417365789413452, -0.051664721220731735, 0.2295839935541153, 0.027339935302734375, -0.10456004738807678, -0.13729730248451233, 0.09618272632360458, -0.2293078601360321, 0.07182291895151138, 0.30790331959724426, 0.05160672217607498, -0.35603106021881104, 0.13332130014896393, -0.012046161107718945, -0.21837808191776276, 0.12953494489192963, 0.37781327962875366, 0.49136731028556824, 0.14920923113822937, 0.02477382868528366, 0.12455224990844727, 0.7539135813713074, 0.20179376006126404, 0.0599365197122097, 0.10636753588914871, 0.24882186949253082, -0.01694445125758648, 0.1358461081981659, -0.11611074954271317, 0.3977302014827728, -0.13664832711219788, 0.4263165295124054, 0.17098748683929443, 0.19975203275680542, -0.20308181643486023, -0.19826583564281464, -0.04472154378890991, -0.2616243064403534, -0.18588194251060486, 0.28222331404685974, -0.22426706552505493, 0.052601028233766556, -0.08051076531410217, -0.4398553669452667, -0.2861534357070923, -0.030420077964663506, -0.06067369133234024, 0.27411526441574097, -0.09199509024620056, 0.18165338039398193, -0.3203600347042084, -0.15978291630744934, -0.4608956575393677, 0.14598330855369568, 0.0887591615319252, 0.3017212152481079, -0.012457234784960747, -0.08777331560850143, -0.07358939945697784, 0.20257115364074707, 0.6373810768127441, -0.4487636983394623, -0.16017547249794006, -0.005617096088826656, 0.060697879642248154, -0.13317398726940155, -0.21247699856758118, -0.25151175260543823, -0.0027013730723410845, 0.46255525946617126, 0.3910320997238159, -0.3513210415840149, -0.29208076000213623, 0.1962275505065918, -0.16269095242023468, -0.055912625044584274, -0.2322782278060913, -0.24113129079341888, -0.10254275798797607, -0.10813780128955841, 0.1177346333861351, 0.15848411619663239, 0.23454289138317108, -0.16833271086215973, 0.16883866488933563, -0.03416478633880615, 0.21754121780395508, 0.17560790479183197, 0.002339516067877412, -0.08446969836950302, 0.2803274691104889, 0.11757047474384308, -0.018795425072312355, -0.052353385835886, -0.189988374710083, 0.08903053402900696, -0.16537900269031525, 0.19147714972496033, 0.3219180107116699, -0.037185173481702805, 0.6493059396743774, 0.09719598293304443, 0.3355262279510498, 0.19747301936149597, -0.05717593431472778, 0.10656388849020004, -0.20644046366214752, 0.16744384169578552, 0.1632014513015747, 0.19573573768138885, -0.2039261907339096, -0.2500295042991638, 0.5089161396026611, 0.40791377425193787, -0.00574007211253047, 0.3332831561565399, -0.6443982720375061, -0.25987672805786133, 0.2356385737657547, 0.0071029458194971085, 0.5431051850318909, 0.005898382514715195, 0.3782312273979187, -0.0436672568321228, -0.31204620003700256, 0.4807913601398468, -0.3299480974674225, 0.16691407561302185, -0.04985669627785683, -0.1427994817495346, 0.0768360123038292, -0.09651881456375122, -0.05870410054922104, 0.3432527482509613, -0.21262706816196442, 0.34454935789108276, 0.4961306154727936, -0.2248220592737198, -0.1798940896987915, 0.47908756136894226, -0.012490312568843365, -0.3797689378261566, 0.0026408659759908915, 0.029900772497057915, -0.23055540025234222, 0.162584587931633, -0.21588599681854248, -0.20611147582530975, 0.2247895896434784, 0.1379055380821228, -0.3079603612422943, -0.02638506144285202, -0.05604485794901848, 0.19251853227615356, 0.26101288199424744, -0.017163416370749474, 0.2552068829536438, -0.24698959290981293, -0.11747213453054428, -0.17436344921588898, -0.05939332768321037, 0.0534646101295948, -0.09174803644418716, 0.11164917796850204, 0.07159272581338882, -0.19615225493907928, 0.4640113115310669, -0.07571952790021896, -0.021193843334913254, -0.2242819368839264, 0.23087891936302185, -0.36356255412101746, -0.22265036404132843, 0.20926374197006226, 0.14877180755138397, 0.23195643723011017, -0.31442955136299133, 0.5743625164031982, 0.08713387697935104, 0.07795105129480362, 0.010979154147207737, 0.1568935513496399, -0.12796734273433685, 0.6329970955848694, -0.29411858320236206, -0.07013030350208282, -0.006362774409353733, -0.0899035632610321, 0.01585705578327179, -0.447561115026474, -0.06190009042620659, 0.2750469446182251, 0.13799358904361725, 0.02883337438106537, 0.09607573598623276, 0.046322837471961975, -0.07321401685476303, 0.024421853944659233, -0.05657600983977318, -0.26283901929855347, 0.21956203877925873, 0.24102330207824707, 0.1873137205839157, 0.2717297673225403, -0.06187158450484276, -0.26336997747421265, -0.10183033347129822, 0.4638352394104004, -0.013535781763494015, 0.242749884724617, -0.09981802850961685, 0.4236369729042053, 0.47125065326690674, 0.5466593503952026, -0.20703795552253723, 0.07370612025260925, 0.13231603801250458, 0.09527600556612015, 0.39160579442977905, -0.05915427953004837, 0.041305460035800934, 0.40140241384506226, -0.1892259567975998, 0.08920932561159134, -0.3487740755081177, -0.1275123655796051, -0.22972333431243896, 0.12747325003147125, 0.15703824162483215, 0.037528280168771744, 0.05876709148287773, -0.08815556019544601, -0.013388999737799168, -0.0449567586183548, 0.06431688368320465, -0.29070714116096497, -0.0705222487449646, 0.08662527054548264, 0.019151579588651657, 0.014970763586461544, 0.12828946113586426, 0.41452303528785706, 0.07437872141599655, -0.08982500433921814, 0.3278346061706543, 0.13821078836917877, 0.15724597871303558, -0.12971721589565277, -0.21711260080337524, -0.08144577592611313, 0.27161285281181335, 0.1716538965702057, -0.16723279654979706, 0.09358716011047363, -0.21851016581058502, -0.08033716678619385, 0.1846632957458496, 0.24251535534858704, 0.07674810290336609, -0.12868663668632507, 0.1766478419303894, 0.07308002561330795, -0.02340264804661274, -0.0032806992530822754, 0.13657641410827637, 0.09948007762432098, 0.03915456309914589, 0.20603764057159424, 0.2117110639810562, 0.5047963261604309, -0.23347993195056915, 0.01652541756629944, 0.06762881577014923, -0.3135482668876648, 0.09969256073236465, -0.1387844681739807, -0.016194675117731094, 0.02821793034672737, 0.2262631058692932, -0.08678135275840759, 0.15465785562992096, -0.3059605360031128, 0.3273349106311798, 0.24166017770767212, 0.035662006586790085, -0.023880939930677414, 0.12478064000606537, -0.03723427653312683, -0.20662900805473328, 0.33759385347366333, -0.028999686241149902, 0.3151319622993469, 0.009087612852454185, 0.7983553409576416, -0.23808087408542633, -0.3960985243320465, 0.1543670892715454, 0.40715575218200684, 0.008456549607217312, -0.26501303911209106, 0.10717971622943878, 0.13453185558319092, -0.26112693548202515, 0.20869074761867523, -0.047896020114421844, -0.0980670377612114, 0.5933933854103088, -0.06555312871932983, 0.1735769808292389, -0.16121479868888855, -0.26063549518585205, 0.3475784361362457, 0.008399995043873787, -0.12483474612236023, 0.058210790157318115, -0.029603544622659683, -0.09513796865940094, -0.10379590839147568, 0.506176769733429, 0.1873367726802826, 0.09599233418703079, -0.34532901644706726, -0.12983420491218567, 0.14326077699661255, 0.2453795224428177, -0.1899362951517105, 0.11646655946969986, 0.49969834089279175, -0.19961200654506683, 0.2697574496269226, -0.025633223354816437, -0.05195740610361099, -0.04092932865023613, 0.04421653226017952, -0.4889706075191498, -0.4163160026073456, 0.41672757267951965, 0.03329286351799965, -0.09751567989587784, 0.05186840519309044, 0.38108399510383606, -0.6517304182052612, -0.19317157566547394, 0.3062872290611267, -0.18151602149009705, -0.08955499529838562, -0.24177570641040802, 0.05955835431814194, 0.03920209780335426, 0.7016067504882812, 0.29243552684783936, 0.3273228704929352, -0.38327163457870483, -0.0800289511680603, -0.5715476870536804, -0.1767011433839798, 0.018201302736997604, 0.00005979567868052982, -0.2943788468837738, 0.2641196846961975, -0.12518583238124847, 0.22162336111068726, -0.09406119585037231, 0.08814559876918793, 0.14160321652889252, 0.07060176879167557, -0.432794451713562, 0.09113334119319916, 0.050961684435606, 0.0823569968342781, 0.30425798892974854, -0.07882446050643921, -0.13369077444076538, 0.05160797759890556, -0.13522709906101227, -0.17467167973518372, -0.08309821784496307, 0.3104028105735779, 0.21442173421382904, 0.108701191842556, 0.01488965842872858, 0.46078112721443176, 0.051516007632017136, -0.1857530027627945, -0.12474841624498367, -0.02235616371035576, -0.34483325481414795, -0.13399288058280945, -0.05004841834306717, 0.1708674281835556, -0.4427907466888428, 0.3610154390335083, 0.04574137181043625, -0.12132177501916885, 0.11556030064821243, 0.03411459922790527, -0.05038754642009735, -0.06718167662620544, 0.12175939977169037, 0.12575653195381165, 0.27909281849861145, 0.699252188205719, 0.026893798261880875, 0.020200515165925026, -0.08986464887857437, -0.02438351698219776, -0.03306368738412857, -0.05258188024163246, -0.20456305146217346, -0.21896381676197052, 0.0958389863371849, -0.060808517038822174, 0.10704422742128372, -0.641090452671051, -0.3868038058280945, 0.0544835589826107, 0.0603058859705925, 0.058845020830631256, 0.6827637553215027, 0.07146178185939789, -0.2238551378250122, -0.11899235099554062, -0.37256914377212524, 0.04101933166384697, -0.29618901014328003, 0.15026095509529114, -0.35550251603126526 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
I think that instead of proportional to size sampling you should specify weights or probabilities for drawing a batch from each dataset. We should also ensure that the smaller datasets are repeated so that the encoder layer doesn't overtrain on the largest dataset.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
43
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. I think that instead of proportional to size sampling you should specify weights or probabilities for drawing a batch from each dataset. We should also ensure that the smaller datasets are repeated so that the encoder layer doesn't overtrain on the largest dataset.
[ -0.10336605459451675, -0.43319350481033325, -0.027864355593919754, -0.014036810025572777, -0.2707728147506714, 0.08221656084060669, 0.15497663617134094, 0.07477432489395142, 0.3294965922832489, -0.09033020585775375, -0.16610087454319, 0.26055192947387695, -0.19721542298793793, 0.40196025371551514, 0.3530554473400116, -0.4648924469947815, -0.05361893028020859, -0.20055343210697174, -0.36640307307243347, 0.33401989936828613, 0.01564602367579937, 0.08674684911966324, -0.17946510016918182, 0.019073208793997765, -0.2719651460647583, -0.11051692813634872, -0.39493528008461, 0.10839428752660751, 0.11043456196784973, -0.013923799619078636, 0.18760821223258972, 0.34423497319221497, 0.0654187947511673, 0.05401672422885895, -0.00011230399104533717, -0.12137190252542496, 0.2175678014755249, -0.09191775321960449, 0.13065959513187408, -0.0005823251558467746, 0.07844502478837967, -0.47050103545188904, -0.007731327787041664, -0.13589666783809662, 0.20232897996902466, 0.1704597920179367, 0.2904418706893921, 0.008269364014267921, 0.3929944336414337, -0.33089274168014526, 0.09726560115814209, 0.25307050347328186, -0.15323840081691742, 0.26218849420547485, -0.1310250461101532, -0.020231416448950768, 0.10178600996732712, 0.25316593050956726, 0.6352720260620117, -0.5347999930381775, -0.03683286905288696, 0.10696182399988174, 0.025654343888163567, 0.03074496239423752, 0.1634945124387741, -0.08338377624750137, -0.31913045048713684, -0.3349868655204773, -0.4021609127521515, 0.23699480295181274, -0.1488322913646698, -0.08393018692731857, -0.2452063411474228, -0.5854566097259521, 0.0016673466889187694, 0.01647743582725525, -0.31532585620880127, 0.11600280553102493, -0.2037290334701538, -0.034372683614492416, -0.25125637650489807, -0.14696139097213745, -0.04344289004802704, 0.14494949579238892, 0.3343850076198578, 0.5547482371330261, 0.08374882489442825, 0.16105464100837708, 0.36956092715263367, 0.057474736124277115, -0.12632258236408234, -0.31179729104042053, 0.2927980422973633, 0.1079670786857605, -0.36915984749794006, -0.3550264239311218, 0.13920535147190094, -0.20298601686954498, 0.29498884081840515, 0.1029912680387497, 0.02505466900765896, 0.08368667960166931, -0.0762173980474472, 0.2560974657535553, 0.37171295285224915, 0.09638248383998871, -0.08300157636404037, -0.06299041956663132, 0.0029549780301749706, -0.06848414242267609, 0.14342010021209717, 0.17668971419334412, 0.12384498864412308, 0.0034175007604062557, -0.4794364273548126, -0.12967395782470703, -0.2586895823478699, 0.07780996710062027, -0.15980222821235657, -0.4763878881931305, -0.06893099099397659, -0.11731696873903275, 0.029383568093180656, -0.08602488040924072, -0.29834437370300293, 0.1183156818151474, -0.09277273714542389, 0.2307395190000534, -0.08128528296947479, 0.037010785192251205, -0.04833837226033211, 0.07137173414230347, -0.4358134865760803, -0.05891101062297821, 0.2667330503463745, 0.12061896920204163, 0.1062668040394783, 0.16640633344650269, 0.013973203487694263, 0.1516638994216919, 0.3776471018791199, 0.016560707241296768, 0.005950232967734337, -0.06492960453033447, 0.21329852938652039, -0.3031076192855835, 0.0006732223555445671, 0.17076852917671204, -0.3618326783180237, -0.1457025557756424, -0.19035477936267853, -0.16295617818832397, 0.09470046311616898, 0.017278753221035004, -0.1942497342824936, -0.2632852792739868, -0.44688743352890015, 0.7332516312599182, 0.026847414672374725, -0.017140312120318413, -0.18028531968593597, 0.26064786314964294, -0.22355805337429047, -0.06396771967411041, 0.20860476791858673, 0.12475001811981201, -0.016909822821617126, -0.3514385521411896, -0.04679785296320915, -0.014809887856245041, 0.03452025726437569, 0.3615563213825226, -0.2216625064611435, 0.1277332305908203, 0.06172599270939827, 0.14611102640628815, 0.21800969541072845, -0.4322618246078491, -0.09815123677253723, 0.40603432059288025, -0.01648096926510334, 0.3736615478992462, 0.11828548461198807, 0.19480179250240326, 0.12559908628463745, -0.19569601118564606, 0.24780865013599396, 1.0400980710983276, -0.42255982756614685, 0.0221929419785738, -0.16494378447532654, -0.3881131410598755, 0.5052053332328796, 0.36283379793167114, 0.09461858868598938, -0.3206782639026642, -0.1631166636943817, -0.026616493239998817, 0.11877492070198059, 0.04143574833869934, 0.10451076179742813, -0.009235205128788948, -0.4052460193634033, -0.02248690091073513, -0.22144314646720886, -0.07816132158041, -0.33089810609817505, 0.015293283388018608, -0.08741326630115509, 0.327463835477829, 0.3405655324459076, -0.10581555217504501, 0.28967177867889404, -0.29848751425743103, -0.13058112561702728, -0.33861058950424194, 0.11447030305862427, -0.03951876610517502, 0.07346755266189575, -0.23656052350997925, -0.06718111038208008, 0.3598691523075104, 0.08904359489679337, -0.1385270208120346, -0.5096871256828308, 0.30903857946395874, -0.057400286197662354, 0.06364470720291138, -0.025106757879257202, 0.5965571999549866, -0.19375915825366974, -0.12890735268592834, 0.14488190412521362, 0.22485187649726868, -0.33462101221084595, 0.1949293166399002, 0.33768758177757263, 0.1734289973974228, 0.2078784704208374, -0.08267931640148163, 0.17195749282836914, -0.029911579564213753, -0.09459716081619263, -0.14048871397972107, -0.030395962297916412, 0.5003795623779297, -0.11950497329235077, 0.44350969791412354, -0.01883564703166485, 0.015262207016348839, -0.26150989532470703, -0.09255732595920563, -0.21653763949871063, 0.3594397306442261, 0.3366412818431854, -0.19295217096805573, 0.36073416471481323, -0.052114058285951614, -0.35946568846702576, 0.14874406158924103, 0.20372191071510315, 0.08420208841562271, 0.046031318604946136, 0.020731620490550995, -0.10165785253047943, -0.27896127104759216, -0.10182236135005951, 0.46157994866371155, 0.5798782706260681, 0.2789749205112457, 0.19102512300014496, 0.0029390749987214804, -0.2511986196041107, -0.08015933632850647, 0.035356294363737106, 0.21586604416370392, -0.05848611146211624, 0.14054538309574127, 0.02464836649596691, 0.009309896267950535, 0.010679068975150585, -0.11853808909654617, 0.146829754114151, -0.18777836859226227, -0.014902847819030285, 0.06336373835802078, -0.19838860630989075, -0.6163865923881531, -0.3052627444267273, -0.015071588568389416, -0.35158029198646545, -0.28699955344200134, 0.13720452785491943, -0.0037929171230643988, -0.13886824250221252, 0.16575385630130768, 0.12699654698371887, 0.5480086803436279, -0.42024481296539307, -0.17030300199985504, -0.023589521646499634, -0.2374991625547409, -0.1251213550567627, 0.10783185809850693, 0.4728635251522064, 0.26112523674964905, 0.4312513768672943, 0.2812909781932831, -0.03624337166547775, 0.12175858020782471, -0.34444910287857056, 0.04084806144237518, -0.2936265170574188, 0.2540328800678253, 0.011808530427515507, -0.2061467468738556, 0.1346391886472702, -0.4294753670692444, -0.06051969155669212, -0.019379181787371635, -0.053349606692790985, -0.028742991387844086, -0.028220709413290024, 0.032595571130514145, -0.17166392505168915, -0.20089273154735565, -0.6588302254676819, -0.45629554986953735, 0.2835223376750946, -0.2002888023853302, 0.09491493552923203, 0.025160131976008415, -0.23395352065563202, 0.08987832814455032, -0.16715094447135925, -0.001951201120391488, -0.12760652601718903, -0.12098327279090881, 0.08369488269090652, -0.35820749402046204, -0.1671838015317917, -0.47265300154685974, -0.12426956743001938, 0.2418021857738495, -0.07113909721374512, -0.07343507558107376, -0.1877882033586502, 0.07646551728248596, -0.1948925405740738, 0.07409601658582687, 0.26539087295532227, 0.10548770427703857, -0.2660577893257141, 0.10515142232179642, 0.0029675124678760767, -0.13249844312667847, 0.2176913321018219, 0.3457125425338745, 0.4796200096607208, 0.15022975206375122, 0.1133512482047081, 0.1909075677394867, 0.6792362928390503, 0.3296077847480774, 0.007964663207530975, 0.21351252496242523, 0.3103105425834656, 0.0240740068256855, 0.197546124458313, -0.21084992587566376, 0.3423006236553192, -0.11590617150068283, 0.45432043075561523, 0.10699531435966492, 0.15114682912826538, -0.2443637251853943, -0.24342261254787445, 0.02432881109416485, -0.2064080387353897, -0.17803598940372467, 0.25939586758613586, -0.22649116814136505, 0.020161205902695656, -0.09770207852125168, -0.36491379141807556, -0.235571950674057, -0.0543934591114521, -0.03128254786133766, 0.2283366471529007, 0.011614445596933365, 0.17223303020000458, -0.3531283140182495, -0.1016591414809227, -0.4612160325050354, 0.20466774702072144, 0.057742878794670105, 0.3017749488353729, -0.019986895844340324, -0.08469349145889282, -0.1292436122894287, 0.12709137797355652, 0.5680524706840515, -0.47282952070236206, -0.22094768285751343, -0.025382792577147484, 0.09025270491838455, -0.09211981296539307, -0.10572770982980728, -0.2197122424840927, -0.025454604998230934, 0.41992616653442383, 0.306242436170578, -0.3639724850654602, -0.27792584896087646, 0.268292635679245, -0.1646316945552826, -0.07220401614904404, -0.14969608187675476, -0.21267899870872498, -0.1383240967988968, -0.17088116705417633, 0.07758388668298721, 0.1296512633562088, 0.19336868822574615, -0.13969528675079346, 0.16703510284423828, -0.10889340937137604, 0.2603107690811157, 0.18591780960559845, 0.05234595760703087, -0.027581224218010902, 0.19997946918010712, 0.17190316319465637, 0.05351273715496063, -0.10656832903623581, -0.23571926355361938, 0.12169591337442398, -0.03250064328312874, 0.07200084626674652, 0.29760995507240295, 0.03904136270284653, 0.6938974857330322, 0.04260319098830223, 0.36705532670021057, 0.22503095865249634, -0.06475987285375595, 0.08746987581253052, -0.21818426251411438, 0.1336083561182022, 0.22997237741947174, 0.2078850418329239, -0.1668774038553238, -0.2211143672466278, 0.5431400537490845, 0.29854604601860046, -0.009611266665160656, 0.33156725764274597, -0.5390052795410156, -0.251165509223938, 0.29675689339637756, 0.04016410931944847, 0.6300413012504578, -0.007101157214492559, 0.29389330744743347, -0.03298863396048546, -0.33714473247528076, 0.5710725784301758, -0.3307279646396637, 0.191995307803154, -0.1401997208595276, -0.24169863760471344, -0.002791389124467969, -0.08570393174886703, -0.07936577498912811, 0.2763814330101013, -0.3371972143650055, 0.3691793382167816, 0.4814932942390442, -0.24805763363838196, -0.10458723455667496, 0.5050843954086304, -0.056458260864019394, -0.37266093492507935, -0.11116693913936615, 0.061154939234256744, -0.2758670449256897, 0.09509270638227463, -0.11868711560964584, -0.30080288648605347, 0.2708374857902527, 0.12162364274263382, -0.24231107532978058, 0.164385586977005, -0.05043228715658188, 0.18452277779579163, 0.25303879380226135, 0.00870249979197979, 0.2706407308578491, -0.16248050332069397, -0.08966303616762161, -0.11695297807455063, -0.07923789322376251, 0.11518190801143646, -0.1497710645198822, 0.01284435298293829, 0.11990507692098618, -0.202458918094635, 0.4892711043357849, -0.009573045186698437, -0.13599494099617004, -0.21550890803337097, 0.3113108277320862, -0.38764169812202454, -0.17851822078227997, 0.10875000059604645, 0.18113873898983002, 0.22955568134784698, -0.359727680683136, 0.5780019164085388, 0.13180913031101227, 0.027301214635372162, 0.030786989256739616, 0.14182808995246887, -0.2512458562850952, 0.5130013227462769, -0.2759883999824524, -0.03809579461812973, 0.04208805784583092, 0.007354210130870342, -0.08804525434970856, -0.3510379195213318, -0.03188444301486015, 0.28335297107696533, 0.12923090159893036, 0.021556586027145386, 0.06551896780729294, 0.17462125420570374, -0.07396002858877182, -0.06540019065141678, -0.13226741552352905, -0.31944820284843445, 0.15650328993797302, 0.07593587040901184, 0.19611194729804993, 0.1613198071718216, -0.07628854364156723, -0.21275363862514496, -0.21801719069480896, 0.5112918615341187, 0.021172093227505684, 0.2578117847442627, -0.18163937330245972, 0.3082199692726135, 0.38870692253112793, 0.5697476863861084, -0.25568681955337524, 0.11726149171590805, 0.055238187313079834, 0.06974653899669647, 0.46103325486183167, -0.2023734152317047, 0.04363463446497917, 0.28928154706954956, -0.14841964840888977, 0.06235450878739357, -0.3408696949481964, -0.1428142488002777, -0.2346571832895279, 0.098805271089077, 0.15117929875850677, -0.01935713179409504, 0.009457666426897049, -0.06798819452524185, -0.0965590700507164, -0.11638068407773972, 0.013010605238378048, -0.43350884318351746, -0.06818906217813492, 0.11059043556451797, -0.037180617451667786, -0.04546308517456055, 0.07459424436092377, 0.2376604527235031, 0.0012456406839191914, -0.020070616155862808, 0.3473863899707794, 0.1378142237663269, 0.13269732892513275, -0.1667906790971756, -0.276227742433548, -0.012614493258297443, 0.2735835909843445, 0.16343896090984344, -0.07122235745191574, 0.10352052748203278, -0.2437407225370407, -0.08736572414636612, 0.2628617286682129, 0.20636162161827087, 0.07711697369813919, -0.17260317504405975, 0.15144705772399902, 0.11271830648183823, -0.07894773036241531, -0.0010418589226901531, 0.20892126858234406, 0.07174279540777206, 0.08645100891590118, 0.17023278772830963, 0.1579517424106598, 0.42820557951927185, -0.2442435771226883, -0.00755491154268384, 0.10386905074119568, -0.34369781613349915, 0.1442802995443344, -0.07468456029891968, -0.016898751258850098, -0.05834364518523216, 0.11643948405981064, -0.0426749587059021, 0.22477802634239197, -0.28217121958732605, 0.2502267062664032, 0.30582115054130554, 0.00996896531432867, -0.04212402179837227, 0.20314408838748932, 0.001802923739887774, -0.22454103827476501, 0.33291471004486084, 0.08520065248012543, 0.38552311062812805, 0.0489545539021492, 0.8560096025466919, -0.2478874772787094, -0.4657873511314392, 0.03729671612381935, 0.48590123653411865, 0.04150944948196411, -0.2708476185798645, 0.1433795690536499, 0.0959801897406578, -0.411872535943985, 0.1614198088645935, 0.07498472183942795, -0.18442420661449432, 0.6622328758239746, -0.05848821997642517, 0.16207750141620636, -0.1906120926141739, -0.17011377215385437, 0.36753880977630615, 0.14315102994441986, -0.13632845878601074, 0.07419078052043915, 0.13879097998142242, -0.10969728231430054, -0.14287304878234863, 0.520385205745697, 0.25319501757621765, 0.03317784145474434, -0.43458354473114014, -0.02207270823419094, 0.16021227836608887, 0.16314105689525604, -0.10989008098840714, 0.12353606522083282, 0.40921059250831604, -0.17104248702526093, 0.26817142963409424, 0.01345699094235897, -0.06439194828271866, -0.17000052332878113, -0.0010192003101110458, -0.4380500018596649, -0.38811805844306946, 0.41963931918144226, 0.1148868203163147, -0.051416296511888504, 0.019020548090338707, 0.24604016542434692, -0.6877148747444153, -0.2315691113471985, 0.3309122622013092, -0.13534817099571228, -0.10280837118625641, -0.24208959937095642, 0.07230065017938614, -0.015391076914966106, 0.6274802088737488, 0.29352232813835144, 0.3365947902202606, -0.40877261757850647, -0.15738660097122192, -0.5566293597221375, -0.14260350167751312, 0.10983331501483917, -0.1562364101409912, -0.2828679084777832, 0.23817521333694458, -0.09585937112569809, 0.23567338287830353, -0.21898718178272247, 0.133131742477417, 0.21304857730865479, 0.09838509559631348, -0.44912129640579224, 0.1221681758761406, 0.10001190751791, 0.04428117722272873, 0.2792526185512543, -0.09820915758609772, -0.171553835272789, 0.05372096598148346, -0.10322351008653641, -0.008148927241563797, -0.01687893643975258, 0.24552445113658905, 0.09913671761751175, 0.12277556955814362, 0.06725380569696426, 0.5235671997070312, -0.03352413326501846, -0.21915000677108765, 0.005621878430247307, -0.04023395851254463, -0.4122084975242615, -0.13063228130340576, -0.07771628350019455, 0.24383783340454102, -0.34682705998420715, 0.2638210356235504, -0.040281280875205994, -0.12499412894248962, 0.12921594083309174, 0.03196447715163231, -0.12255149334669113, -0.058099210262298584, 0.057855620980262756, 0.21735675632953644, 0.21858011186122894, 0.7121253609657288, 0.03826141729950905, 0.10666968673467636, -0.028915775939822197, -0.024187590926885605, 0.06974607706069946, -0.081906758248806, -0.22436873614788055, -0.2049364447593689, 0.025506746023893356, 0.06479831039905548, 0.07541260868310928, -0.6441549062728882, -0.32651790976524353, 0.011607500724494457, 0.004460558760911226, 0.047522660344839096, 0.7026140093803406, 0.11499752849340439, -0.14029550552368164, -0.13398154079914093, -0.2580881714820862, -0.000519195687957108, -0.3298356533050537, 0.20005030930042267, -0.30358055233955383 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
Are there any references for people doing different batch sizes per task in the literature? I've only seen constant batch sizes with differing numbers of batches for each task which seems sufficient to prevent the impact of large datasets (Read 3.5.3 of the [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) for example).
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
47
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. Are there any references for people doing different batch sizes per task in the literature? I've only seen constant batch sizes with differing numbers of batches for each task which seems sufficient to prevent the impact of large datasets (Read 3.5.3 of the [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) for example).
[ -0.06431491672992706, -0.4659213423728943, -0.08466382324695587, -0.031063778325915337, -0.2937632203102112, 0.01426788978278637, 0.2748717665672302, -0.0015312168980017304, 0.3273965120315552, -0.08819633722305298, -0.1483938992023468, 0.10962097346782684, -0.16663536429405212, 0.4398266077041626, 0.34445920586586, -0.36647844314575195, 0.03876933082938194, -0.2969399690628052, -0.2708534598350525, 0.38539090752601624, 0.03808370605111122, 0.0419313870370388, -0.13354307413101196, 0.08822571486234665, -0.2808082401752472, -0.2636704742908478, -0.38970741629600525, 0.034435611218214035, 0.13997460901737213, 0.05134955793619156, 0.23408909142017365, 0.3075520098209381, 0.04861658811569214, 0.27919286489486694, -0.000116871771751903, -0.08377692848443985, 0.23081009089946747, -0.01913638599216938, 0.11680792272090912, -0.0018161081243306398, 0.15380090475082397, -0.4783225953578949, 0.005465093534439802, -0.11083633452653885, 0.2739710509777069, 0.1568039059638977, 0.25918152928352356, 0.1064637303352356, 0.28866496682167053, -0.3447948694229126, 0.09772246330976486, 0.2746801972389221, -0.21083910763263702, 0.22634705901145935, -0.11846113204956055, -0.014193056151270866, 0.07735450565814972, 0.2606102526187897, 0.7055597901344299, -0.5410467982292175, -0.07276517897844315, 0.07944506406784058, 0.04223455488681793, 0.011969735845923424, 0.10907944291830063, -0.17149685323238373, -0.2628781199455261, -0.2975197434425354, -0.3295663595199585, 0.22085711359977722, -0.08587867021560669, -0.09882605820894241, -0.2601613998413086, -0.6264219284057617, -0.036424487829208374, 0.031553253531455994, -0.3546115458011627, 0.16949862241744995, -0.2045319676399231, -0.12913109362125397, -0.23460255563259125, -0.10024083405733109, -0.026420725509524345, 0.07539902627468109, 0.3474642038345337, 0.5395688414573669, 0.04197026789188385, 0.12721961736679077, 0.3926808834075928, -0.03826446458697319, -0.08897168189287186, -0.273771733045578, 0.25671231746673584, 0.13433721661567688, -0.39920032024383545, -0.29389309883117676, 0.1546253114938736, -0.26450952887535095, 0.3152869939804077, 0.08219613879919052, 0.0817464217543602, 0.05161236971616745, -0.1418328881263733, 0.19767364859580994, 0.4014687240123749, 0.12532833218574524, -0.09977984428405762, -0.11220309138298035, 0.04722466692328453, -0.16153109073638916, 0.21224670112133026, 0.18241579830646515, 0.1426582783460617, -0.1386050134897232, -0.5396949052810669, -0.1671905815601349, -0.27625399827957153, 0.14143963158130646, -0.18073846399784088, -0.5744894742965698, -0.020447449758648872, -0.15083526074886322, 0.020509451627731323, -0.09436854720115662, -0.272125244140625, 0.09609336405992508, -0.11833248287439346, 0.26906853914260864, -0.16005472838878632, 0.07366324961185455, -0.0505211241543293, 0.04711344093084335, -0.47076377272605896, -0.1046871468424797, 0.2876485586166382, 0.17828789353370667, 0.09426690638065338, 0.18709887564182281, -0.03195863217115402, 0.08593231439590454, 0.30259817838668823, -0.06191779673099518, -0.05103762447834015, 0.03935444727540016, 0.24605825543403625, -0.26802653074264526, -0.08793450146913528, 0.24862846732139587, -0.37916016578674316, -0.0590340830385685, -0.23170852661132812, -0.15932072699069977, 0.15850946307182312, -0.002280393149703741, -0.14439576864242554, -0.2504311800003052, -0.5782615542411804, 0.673693060874939, 0.062366727739572525, -0.05668538436293602, -0.16031385958194733, 0.22042030096054077, -0.1436537653207779, -0.18698100745677948, 0.0599549300968647, 0.018209684640169144, -0.03283495455980301, -0.33299535512924194, 0.036641210317611694, -0.03524084389209747, 0.0035879071801900864, 0.4364960193634033, -0.18516051769256592, 0.2895740568637848, 0.08100887387990952, 0.1218290776014328, 0.2664521634578705, -0.4424833357334137, -0.01151538360863924, 0.4173333942890167, 0.050326310098171234, 0.33077841997146606, 0.12340635061264038, 0.23189163208007812, 0.04421810805797577, -0.20737804472446442, 0.23588965833187103, 0.9741359949111938, -0.37452736496925354, 0.09668035060167313, -0.05804971978068352, -0.4658018946647644, 0.4411916136741638, 0.22587604820728302, 0.08976082503795624, -0.2925754487514496, -0.0739048421382904, -0.054094098508358, 0.05737728998064995, 0.0896449089050293, 0.0959465354681015, 0.047745298594236374, -0.4209628701210022, 0.0964869037270546, -0.1834496110677719, -0.050520122051239014, -0.4460892379283905, -0.010564140975475311, -0.08277671039104462, 0.1956140249967575, 0.37465623021125793, -0.09083902835845947, 0.33623936772346497, -0.32630717754364014, -0.09256357699632645, -0.34133195877075195, 0.06611599028110504, -0.1217469722032547, 0.09767425060272217, -0.3077353239059448, 0.0007396616856567562, 0.3354279100894928, 0.09459530562162399, -0.10449402779340744, -0.4976731538772583, 0.17593234777450562, -0.036460816860198975, -0.008629961870610714, 0.05814821273088455, 0.540845513343811, -0.19120092689990997, -0.08414289355278015, 0.10430680960416794, 0.2897184193134308, -0.2685346007347107, 0.23165194690227509, 0.3110857903957367, 0.26073771715164185, 0.21135883033275604, 0.11839230358600616, 0.20557305216789246, -0.10977327078580856, -0.0994366854429245, -0.22492973506450653, -0.10347114503383636, 0.4607519805431366, -0.13459277153015137, 0.5010110139846802, -0.029697518795728683, -0.07168645411729813, -0.18867127597332, -0.01730378344655037, -0.1613263040781021, 0.45697250962257385, 0.3395255506038666, -0.17075695097446442, 0.3268982470035553, -0.09833323955535889, -0.3560431897640228, 0.05383677035570145, 0.022897696122527122, 0.05386385694146156, 0.0811348408460617, -0.008003470487892628, -0.03082766942679882, -0.29537686705589294, -0.1262146532535553, 0.36534640192985535, 0.6546177268028259, 0.22161811590194702, 0.21824181079864502, -0.061494287103414536, -0.22549648582935333, -0.15221549570560455, -0.025874748826026917, 0.20214460790157318, 0.006640059873461723, 0.16049796342849731, 0.02629782445728779, -0.0057368651032447815, 0.08471197634935379, -0.10533840954303741, 0.11840701103210449, -0.225221648812294, -0.08519236743450165, 0.005404846277087927, -0.22219181060791016, -0.6311869025230408, -0.30558815598487854, 0.01507929153740406, -0.30497074127197266, -0.2313881814479828, 0.03909904509782791, 0.07374676316976547, -0.11424049735069275, 0.08188250660896301, 0.2115861177444458, 0.6373739838600159, -0.43466493487358093, -0.10998228937387466, -0.03147687390446663, -0.24741187691688538, -0.14098462462425232, 0.06759561598300934, 0.4684337377548218, 0.20970197021961212, 0.46051180362701416, 0.3219377398490906, -0.10924175381660461, 0.11420053988695145, -0.32258516550064087, 0.02633778564631939, -0.200529545545578, 0.24238120019435883, 0.04370439052581787, -0.1586247980594635, 0.176504984498024, -0.3366839587688446, -0.0731433853507042, -0.061084896326065063, -0.05285284295678139, -0.1735837310552597, -0.0702570304274559, 0.05648845434188843, -0.1488361805677414, -0.10039743036031723, -0.6727864742279053, -0.396466463804245, 0.3388642370700836, -0.2569650411605835, 0.06550893932580948, 0.025421498343348503, -0.19426141679286957, 0.07539017498493195, -0.14416226744651794, -0.01773693598806858, -0.0749676451086998, -0.12288586050271988, -0.029145220294594765, -0.4028762876987457, -0.15951478481292725, -0.373208224773407, -0.021674521267414093, 0.19834591448307037, 0.017058100551366806, -0.06768307089805603, -0.09868616610765457, 0.029117537662386894, -0.17894546687602997, 0.1095876693725586, 0.3481382429599762, 0.08324112743139267, -0.39999639987945557, 0.1514003425836563, -0.05563649907708168, -0.24908781051635742, 0.14666183292865753, 0.2859278619289398, 0.4913569986820221, 0.2199200689792633, 0.008942006155848503, 0.15517908334732056, 0.7662393450737, 0.273021399974823, 0.023623326793313026, 0.098593570291996, 0.28743666410446167, 0.025504659861326218, 0.12757132947444916, -0.1188780665397644, 0.3374466598033905, -0.17567260563373566, 0.43661925196647644, 0.06794430315494537, 0.2040228545665741, -0.16058248281478882, -0.23380205035209656, -0.03952755406498909, -0.3257651925086975, -0.17143787443637848, 0.25072619318962097, -0.2517659664154053, 0.07811910659074783, -0.14358103275299072, -0.4316737949848175, -0.29271504282951355, -0.01696156896650791, -0.06760840117931366, 0.30172327160835266, -0.08189809322357178, 0.20029865205287933, -0.3350657820701599, -0.16427893936634064, -0.38935917615890503, 0.1740121841430664, -0.0023740497417747974, 0.29987502098083496, -0.037796683609485626, -0.0142647260800004, -0.07898116856813431, 0.1751031130552292, 0.5862995386123657, -0.4357357621192932, -0.14293192327022552, -0.014257918111979961, 0.1116928979754448, -0.14096753299236298, -0.19628851115703583, -0.2515060603618622, 0.0530729703605175, 0.5074827671051025, 0.37952783703804016, -0.36115238070487976, -0.316876620054245, 0.1973589062690735, -0.19273711740970612, -0.09322429448366165, -0.2261342853307724, -0.21406060457229614, -0.16825507581233978, -0.07910788804292679, 0.14117662608623505, 0.13951686024665833, 0.15919196605682373, -0.23411113023757935, 0.14947563409805298, -0.07838361710309982, 0.2822118401527405, 0.1764063984155655, -0.018855005502700806, -0.07879506796598434, 0.24801956117153168, 0.1028497964143753, 0.01776900142431259, -0.08682423830032349, -0.1627187281847, 0.03729351982474327, -0.13130152225494385, 0.1557685136795044, 0.2967382073402405, 0.013812391087412834, 0.6570737957954407, 0.10724955052137375, 0.4018983840942383, 0.20059800148010254, -0.11568454653024673, 0.05698458105325699, -0.19889222085475922, 0.12703095376491547, 0.23611514270305634, 0.2462623417377472, -0.2228366881608963, -0.24729792773723602, 0.5878694653511047, 0.3816869258880615, -0.05445464327931404, 0.4050363600254059, -0.6244825124740601, -0.2605518400669098, 0.23180058598518372, 0.024832887575030327, 0.5990932583808899, -0.01311671081930399, 0.29962193965911865, -0.07943973690271378, -0.35769352316856384, 0.4593539535999298, -0.31158798933029175, 0.16311481595039368, -0.07899990677833557, -0.08968014270067215, 0.0702202245593071, -0.13893495500087738, 0.018397552892565727, 0.29449304938316345, -0.2356727421283722, 0.37945204973220825, 0.412104994058609, -0.2506571114063263, -0.1553010493516922, 0.4758177697658539, -0.03475132957100868, -0.3831043243408203, 0.05561250075697899, 0.02473701722919941, -0.2695634067058563, 0.1704661250114441, -0.21481584012508392, -0.20060227811336517, 0.2319066822528839, 0.1601426899433136, -0.3523414731025696, 0.02330562099814415, -0.01612582989037037, 0.22780856490135193, 0.27945661544799805, -0.04916521906852722, 0.2548290491104126, -0.1672864705324173, -0.10953901708126068, -0.12556132674217224, -0.03871764987707138, 0.10054410248994827, -0.11218155920505524, 0.03601552173495293, 0.0384792722761631, -0.1712268441915512, 0.439947247505188, -0.054046664386987686, -0.08606473356485367, -0.2332592010498047, 0.22317679226398468, -0.2976767122745514, -0.29632464051246643, 0.13378073275089264, 0.1284630447626114, 0.1511077582836151, -0.3166044056415558, 0.5542105436325073, 0.1613525152206421, 0.09315460175275803, 0.0071568782441318035, 0.163462832570076, -0.12557047605514526, 0.6486726403236389, -0.32968729734420776, -0.0107022849842906, 0.03141966462135315, -0.0907798558473587, 0.006326408125460148, -0.4751656949520111, -0.0592220276594162, 0.29011309146881104, 0.11815181374549866, 0.045331601053476334, 0.09262403100728989, 0.09316834062337875, 0.02672489918768406, -0.01084180362522602, -0.10102397948503494, -0.25960010290145874, 0.17278294265270233, 0.16765469312667847, 0.1894170194864273, 0.1972038298845291, -0.0561286024749279, -0.27054363489151, -0.10723770409822464, 0.4969044327735901, -0.01926986128091812, 0.21876248717308044, -0.11857295036315918, 0.3374299705028534, 0.49458253383636475, 0.5996086597442627, -0.1953844279050827, 0.09189900755882263, 0.15111735463142395, 0.09730435907840729, 0.41306930780410767, -0.0851142629981041, 0.015879785642027855, 0.35309088230133057, -0.20409686863422394, 0.03782385587692261, -0.29770344495773315, -0.11826516687870026, -0.19182099401950836, 0.11325038969516754, 0.12125448137521744, 0.046959806233644485, 0.028699617832899094, -0.14802177250385284, 0.009298484772443771, -0.051961302757263184, 0.061060987412929535, -0.2568275034427643, -0.07784228771924973, 0.1117902398109436, 0.00030386584694497287, 0.08506185561418533, 0.14502829313278198, 0.3739888668060303, 0.04396049305796623, -0.06393001228570938, 0.3479340374469757, 0.10780206322669983, 0.1781654953956604, -0.15824896097183228, -0.2107112854719162, -0.05424986034631729, 0.29739826917648315, 0.1429697871208191, -0.17916187644004822, 0.08633306622505188, -0.18172810971736908, -0.12257339060306549, 0.14351697266101837, 0.2863253355026245, 0.08999735862016678, -0.11031531542539597, 0.2158234864473343, 0.06141481176018715, 0.0060815452598035336, 0.0023055581841617823, 0.17192623019218445, 0.10082137584686279, -0.0049418434500694275, 0.1974211037158966, 0.17235055565834045, 0.4390959143638611, -0.2788703441619873, 0.029636230319738388, 0.09111060202121735, -0.34259939193725586, 0.09730201959609985, -0.13354741036891937, -0.0017445755656808615, -0.02151177078485489, 0.2189326286315918, 0.0036819379311054945, 0.1451462209224701, -0.31865230202674866, 0.3492981195449829, 0.28361237049102783, 0.04186604544520378, -0.03126363083720207, 0.1697317212820053, -0.1356530487537384, -0.2388625293970108, 0.2818865180015564, -0.007455060258507729, 0.29146838188171387, 0.006257666274905205, 0.8371670842170715, -0.2760741412639618, -0.40760013461112976, 0.11187463253736496, 0.4058804512023926, -0.00841414462774992, -0.24195142090320587, 0.05891656130552292, 0.13566075265407562, -0.21899603307247162, 0.18768532574176788, 0.005945628974586725, -0.15572088956832886, 0.6309080123901367, -0.086859330534935, 0.15201330184936523, -0.1394614577293396, -0.18477801978588104, 0.32048988342285156, 0.06872197240591049, -0.04573463648557663, 0.10238125920295715, -0.0006044308538548648, -0.08211388438940048, -0.1377706676721573, 0.5545238256454468, 0.27132299542427063, 0.05069422349333763, -0.394832968711853, -0.05802132934331894, 0.12167723476886749, 0.18451663851737976, -0.0989895910024643, 0.12961940467357635, 0.4572375416755676, -0.21894396841526031, 0.2153020203113556, -0.018487654626369476, -0.033718422055244446, -0.06840454041957855, 0.032773055136203766, -0.4699130654335022, -0.35822156071662903, 0.4419863522052765, 0.03924256190657616, -0.15338733792304993, 0.09869414567947388, 0.3518233597278595, -0.6634212732315063, -0.22207628190517426, 0.3930751085281372, -0.15513229370117188, -0.11853025108575821, -0.19786478579044342, 0.05703793093562126, 0.12376904487609863, 0.6506406664848328, 0.3044011890888214, 0.31558525562286377, -0.33722448348999023, -0.15078012645244598, -0.5069012641906738, -0.2103443294763565, -0.015344860963523388, -0.048915062099695206, -0.31534773111343384, 0.21218343079090118, -0.1383204162120819, 0.1627189666032791, -0.1201685220003128, 0.09354425221681595, 0.22755862772464752, 0.18092086911201477, -0.4607583284378052, 0.12454617023468018, 0.09379228204488754, 0.10597740858793259, 0.2719219923019409, -0.08295954763889313, -0.15479229390621185, 0.08311763405799866, -0.13035956025123596, -0.06441937386989594, -0.005010298918932676, 0.370147705078125, 0.13319042325019836, 0.08900167793035507, 0.067532978951931, 0.5516714453697205, 0.08134454488754272, -0.16196219623088837, -0.0657716616988182, 0.03485586494207382, -0.35132086277008057, -0.13600271940231323, -0.08042512089014053, 0.11832867562770844, -0.4727514684200287, 0.30819985270500183, 0.0680919662117958, -0.15013915300369263, 0.05180405080318451, 0.01679895631968975, -0.035492897033691406, -0.10961998254060745, 0.10881639271974564, 0.15477830171585083, 0.342180460691452, 0.7173800468444824, 0.013488725759088993, 0.051124218851327896, -0.11207658797502518, 0.016379347071051598, -0.0414472371339798, -0.1004003956913948, -0.2099619209766388, -0.2421920746564865, 0.09355363994836807, -0.043548889458179474, 0.12603606283664703, -0.5950589776039124, -0.3449791371822357, 0.0035270643420517445, 0.056977950036525726, 0.09147119522094727, 0.6922089457511902, 0.08021263778209686, -0.1998313069343567, -0.12055385857820511, -0.3673875629901886, 0.05131477117538452, -0.2942933738231659, 0.09561274200677872, -0.31406664848327637 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
Hi, regarding building T5 dataset , I think we can use datasets https://github.com/huggingface/datasets and then need something similar to tf.data.experimental.sample_from_datasets, do you know if similar functionality exist in pytorch? Which can sample multiple datasets with the given rates. thanks.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
39
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. Hi, regarding building T5 dataset , I think we can use datasets https://github.com/huggingface/datasets and then need something similar to tf.data.experimental.sample_from_datasets, do you know if similar functionality exist in pytorch? Which can sample multiple datasets with the given rates. thanks.
[ -0.03330687806010246, -0.5136847496032715, -0.02777979150414467, 0.06664847582578659, -0.25032755732536316, 0.04262993857264519, 0.1753961443901062, 0.060455743223428726, 0.20672675967216492, -0.12869608402252197, -0.2596317529678345, 0.1723335087299347, -0.16145756840705872, 0.5526631474494934, 0.36890414357185364, -0.42828017473220825, -0.033644117414951324, -0.22553163766860962, -0.3311310112476349, 0.3586278557777405, 0.16255323588848114, 0.12674058973789215, -0.13107413053512573, 0.011625542305409908, -0.3748578727245331, -0.0523345023393631, -0.3290127217769623, -0.014158261008560658, 0.07128273695707321, 0.15654438734054565, 0.140565887093544, 0.34378740191459656, -0.04573146253824234, 0.3042806088924408, -0.00012068865908076987, -0.0008576058899052441, 0.06160637363791466, -0.12618504464626312, 0.11212540417909622, -0.03864677995443344, 0.28744903206825256, -0.3399033844470978, 0.053131673485040665, -0.1707758903503418, 0.13239850103855133, 0.1373211145401001, 0.2363462746143341, 0.07320497930049896, 0.4259682595729828, -0.22620785236358643, 0.04279346019029617, 0.4667559266090393, -0.28901147842407227, 0.29696983098983765, 0.003143023233860731, -0.0374879464507103, -0.03362719714641571, 0.2170954793691635, 0.7237201929092407, -0.40428465604782104, 0.06341978907585144, 0.040230266749858856, -0.013891276903450489, -0.01136762648820877, 0.21762850880622864, -0.007717551197856665, -0.3298623859882355, -0.29819416999816895, -0.3776233494281769, 0.24901743233203888, 0.00380880874581635, -0.18128299713134766, -0.22731955349445343, -0.5786264538764954, -0.05229785293340683, -0.008305443450808525, -0.3058771789073944, 0.028880156576633453, -0.2804424464702606, -0.07985930144786835, -0.1518014520406723, -0.136544331908226, -0.11704979836940765, 0.18049681186676025, 0.37516215443611145, 0.4616799056529999, 0.06983580440282822, 0.17935635149478912, 0.2523779571056366, 0.13546842336654663, -0.05440926551818848, -0.30060818791389465, 0.2920225262641907, 0.18487867712974548, -0.35682037472724915, -0.34360867738723755, 0.07528078556060791, -0.3887386918067932, 0.3567279875278473, 0.12050828337669373, 0.16586177051067352, 0.010490135289728642, -0.14146921038627625, 0.11952217668294907, 0.39286863803863525, -0.09637804329395294, -0.16700118780136108, -0.11347788572311401, -0.025100121274590492, -0.25130605697631836, 0.07018446922302246, 0.09315354377031326, 0.1699235439300537, -0.10998431593179703, -0.36253824830055237, -0.1245318055152893, -0.13821248710155487, 0.09857913106679916, -0.03294748440384865, -0.5224343538284302, -0.21595120429992676, -0.18340986967086792, -0.03497113287448883, -0.08714708685874939, -0.2840885519981384, 0.04984131082892418, -0.0025755579117685556, 0.21522292494773865, -0.2706935703754425, 0.04158295691013336, -0.04056704789400101, 0.027274757623672485, -0.374625027179718, -0.04783232510089874, 0.2949216365814209, 0.0531737245619297, 0.1146206483244896, 0.16285595297813416, 0.19290141761302948, 0.12154766917228699, 0.44899511337280273, -0.02812281996011734, -0.010044615715742111, -0.023936940357089043, 0.14568492770195007, -0.21127949655056, -0.01912006177008152, 0.33989474177360535, -0.3035697340965271, -0.06994905322790146, -0.1554129272699356, -0.19051949679851532, 0.09611643850803375, -0.050944313406944275, -0.20126138627529144, -0.3853227496147156, -0.4683857560157776, 0.6835909485816956, 0.06396520882844925, -0.09594032913446426, -0.20830343663692474, 0.20528550446033478, -0.210993692278862, -0.08647233992815018, 0.16142477095127106, 0.06823378056287766, -0.013202696107327938, -0.37244001030921936, 0.04198368266224861, -0.100539430975914, 0.008564785122871399, 0.3029562830924988, -0.1540871262550354, 0.22199887037277222, 0.08476738631725311, 0.1121397614479065, 0.33994558453559875, -0.47918134927749634, -0.009698819369077682, 0.35266879200935364, -0.042694538831710815, 0.2965773344039917, 0.14705269038677216, 0.27371546626091003, -0.0726066529750824, -0.15075424313545227, 0.1296711415052414, 0.9934563040733337, -0.3805130124092102, -0.01721472665667534, -0.07959863543510437, -0.3645017743110657, 0.4801669716835022, 0.3585851788520813, 0.19760873913764954, -0.3657471239566803, -0.18771888315677643, -0.005162875633686781, 0.0852506160736084, 0.006785052828490734, 0.08330775052309036, 0.024007778614759445, -0.23441942036151886, 0.10306442528963089, -0.1965171843767166, -0.2335735410451889, -0.4084457457065582, 0.04143492132425308, -0.05092319846153259, 0.37805965542793274, 0.3517454266548157, -0.05343768373131752, 0.2586524784564972, -0.2782161831855774, -0.12434320896863937, -0.3101746141910553, 0.03456566855311394, -0.023268690332770348, 0.0006009777425788343, -0.26450562477111816, -0.09321241080760956, 0.26055583357810974, 0.03617459908127785, -0.09498095512390137, -0.43100589513778687, 0.42751798033714294, 0.02262916788458824, -0.03463868051767349, -0.03160082921385765, 0.531765878200531, -0.246052086353302, -0.1196654811501503, 0.10470885783433914, 0.216263085603714, -0.20524661242961884, 0.1573849618434906, 0.22795172035694122, 0.3987586498260498, 0.23988543450832367, -0.08323944360017776, 0.05731722339987755, -0.051717862486839294, -0.06435848027467728, -0.21896986663341522, -0.004287502728402615, 0.40685948729515076, -0.2748054563999176, 0.40373408794403076, -0.00748521089553833, -0.02760281041264534, -0.22626423835754395, 0.010464388877153397, -0.12387295812368393, 0.42509186267852783, 0.34055131673812866, -0.24968090653419495, 0.31807172298431396, -0.1349046528339386, -0.34771785140037537, 0.23371106386184692, 0.1268346905708313, 0.019251983612775803, 0.00021442289289552718, -0.04031013324856758, 0.015348880551755428, -0.2949434816837311, -0.1368178129196167, 0.3106767535209656, 0.6342356204986572, 0.19913384318351746, 0.1362776756286621, -0.06178681179881096, -0.18401069939136505, -0.06175155192613602, 0.07712726294994354, 0.1738758087158203, -0.07647119462490082, 0.21138271689414978, -0.013370044529438019, 0.10678864270448685, 0.04169214516878128, -0.1683126837015152, 0.25196734070777893, -0.17216432094573975, -0.16441860795021057, 0.0485866479575634, -0.17014405131340027, -0.6914840936660767, -0.11619990319013596, 0.08361675590276718, -0.29779040813446045, -0.2556682527065277, 0.1518947035074234, 0.13057976961135864, -0.09642641991376877, 0.17519380152225494, 0.16043347120285034, 0.4907016158103943, -0.3046950101852417, -0.3377006947994232, -0.0032319684978574514, -0.26299428939819336, -0.07772134244441986, 0.04284597933292389, 0.43576303124427795, 0.41830989718437195, 0.4902554750442505, 0.19849003851413727, -0.07153186947107315, -0.04201938211917877, -0.389818400144577, 0.15624652802944183, -0.36184272170066833, 0.3624500632286072, 0.0631992518901825, -0.12011870741844177, 0.07015486806631088, -0.31865787506103516, -0.014651964418590069, -0.08830571174621582, -0.1483505219221115, -0.06262721121311188, -0.09027503430843353, 0.07648718357086182, -0.13062815368175507, -0.16166265308856964, -0.6078262329101562, -0.3920362889766693, 0.19814468920230865, -0.18852901458740234, 0.07596120238304138, 0.15446610748767853, -0.284027636051178, 0.24272431433200836, -0.12732841074466705, 0.08162108063697815, -0.1322251856327057, -0.14825496077537537, -0.016929341480135918, -0.4053654968738556, -0.22016732394695282, -0.4019717872142792, -0.019565101712942123, 0.1676710844039917, 0.05245693773031235, -0.09963679313659668, -0.3910076916217804, 0.0016662951093167067, 0.027671873569488525, 0.05836048722267151, 0.25692451000213623, 0.11975838989019394, -0.28187522292137146, 0.16816358268260956, -0.14121799170970917, -0.18737497925758362, 0.2724558115005493, 0.31200867891311646, 0.36048623919487, 0.17845961451530457, -0.0021582236513495445, 0.12795639038085938, 0.9331133961677551, 0.2947480380535126, 0.0042396667413413525, 0.18400724232196808, 0.13027088344097137, 0.019158944487571716, 0.08279627561569214, -0.19149218499660492, 0.2007608264684677, -0.11971060931682587, 0.46759676933288574, 0.10902028530836105, 0.0879262387752533, -0.016810104250907898, -0.29273855686187744, 0.09595175087451935, -0.3363732397556305, -0.25714942812919617, 0.2042868286371231, -0.3424229621887207, 0.1067073792219162, -0.12691494822502136, -0.4086243510246277, -0.37843796610832214, -0.07601001113653183, -0.09919151663780212, 0.28520894050598145, 0.10526403039693832, 0.17052388191223145, -0.38959306478500366, -0.05017722770571709, -0.4277198314666748, 0.14170709252357483, 0.12198333442211151, 0.2932141125202179, -0.08066704124212265, -0.0009063371689990163, -0.06195133924484253, 0.1394704133272171, 0.5762670040130615, -0.4894067049026489, -0.09264380484819412, 0.1321844905614853, -0.09517470002174377, -0.16487520933151245, -0.029359569773077965, -0.3096030354499817, 0.05311211571097374, 0.34030410647392273, 0.341982364654541, -0.48226892948150635, -0.4431769847869873, 0.28553125262260437, -0.07070673257112503, -0.06597413122653961, -0.1669311672449112, -0.16285565495491028, -0.23329295217990875, -0.0486721433699131, -0.08596506714820862, 0.35172224044799805, 0.24646803736686707, -0.14004100859165192, 0.20227625966072083, -0.03201494738459587, 0.2508802115917206, 0.20811648666858673, -0.08595210313796997, -0.07969840615987778, 0.1945643126964569, 0.25474750995635986, 0.034911271184682846, -0.1054154708981514, -0.3044125735759735, 0.1429988592863083, -0.144871786236763, -0.1175747886300087, 0.2817609906196594, 0.062038298696279526, 0.6452472805976868, 0.07200584560632706, 0.34857362508773804, 0.34181302785873413, -0.1439763456583023, 0.0882175862789154, -0.10354948043823242, 0.20663662254810333, 0.22724968194961548, 0.2593936622142792, -0.11130714416503906, -0.2703581154346466, 0.5457836389541626, 0.3118359446525574, 0.01879643276333809, 0.46680960059165955, -0.5664708018302917, -0.2669513523578644, 0.2087346911430359, 0.043550409376621246, 0.7242982983589172, -0.06499584019184113, 0.41728365421295166, 0.14827431738376617, -0.39617758989334106, 0.578062117099762, -0.21578100323677063, 0.041985925287008286, -0.1446857899427414, -0.3205382525920868, 0.024116091430187225, -0.13038329780101776, -0.2033686637878418, 0.22062093019485474, -0.22746001183986664, 0.2735259234905243, 0.3709966242313385, -0.21724484860897064, -0.03466913849115372, 0.4872440695762634, -0.08320419490337372, -0.3102729022502899, -0.00783240795135498, -0.005010046996176243, -0.272060751914978, 0.15342167019844055, -0.15633097290992737, -0.2725106179714203, 0.1802784502506256, 0.22555729746818542, -0.36308765411376953, -0.03686443343758583, -0.04107186198234558, 0.17539910972118378, 0.2798943519592285, -0.06461749225854874, 0.22811925411224365, -0.05899675190448761, -0.04431503638625145, -0.1844000667333603, -0.13139386475086212, 0.11050407588481903, -0.25417783856391907, 0.03881372511386871, 0.1308594048023224, -0.08861711621284485, 0.47457239031791687, -0.10375456511974335, -0.14080624282360077, -0.2136504054069519, 0.20078079402446747, -0.4532198905944824, -0.2897719442844391, -0.028219075873494148, 0.04603682458400726, 0.14461523294448853, -0.31200042366981506, 0.5833739638328552, 0.020649202167987823, 0.05548401176929474, -0.032348014414310455, 0.20651693642139435, -0.17367179691791534, 0.4253312647342682, -0.35417526960372925, -0.04309452697634697, 0.05548514798283577, 0.03814886510372162, -0.049222011119127274, -0.39530766010284424, 0.07578033208847046, 0.41568753123283386, 0.1633375585079193, 0.03143133595585823, 0.15434998273849487, 0.10619046539068222, -0.18556229770183563, 0.11073166131973267, -0.07481778413057327, -0.19344598054885864, 0.12716969847679138, 0.36334651708602905, 0.16826575994491577, 0.12084873020648956, -0.06369765847921371, -0.23129519820213318, -0.18166837096214294, 0.5214853286743164, -0.014889610931277275, 0.22884058952331543, -0.15737059712409973, 0.4722113609313965, 0.42730894684791565, 0.4741526246070862, -0.19213159382343292, 0.11956674605607986, 0.1168096512556076, 0.08993370831012726, 0.5097171068191528, -0.24610301852226257, 0.04991719126701355, 0.26596829295158386, -0.21029984951019287, -0.0020567206665873528, -0.3129920959472656, -0.08803512901067734, -0.26381582021713257, 0.15460637211799622, 0.13262170553207397, 0.017633410170674324, 0.08132212609052658, -0.10233187675476074, -0.2083645462989807, -0.10318611562252045, 0.13714393973350525, -0.2286572903394699, -0.0798501968383789, 0.21865060925483704, 0.0048442562110722065, -0.01534509751945734, -0.06830450147390366, 0.32523953914642334, 0.06706135720014572, -0.003958745859563351, 0.34234219789505005, 0.15840305387973785, 0.16730649769306183, -0.1784553825855255, -0.16622091829776764, -0.027407215908169746, 0.3266676068305969, 0.09608463943004608, -0.0647890642285347, 0.058879103511571884, -0.11979356408119202, -0.18889710307121277, 0.18732775747776031, 0.2522055208683014, 0.03003431111574173, -0.027138877660036087, 0.21464169025421143, 0.03223838284611702, -0.03687275946140289, 0.001633219188079238, 0.13352461159229279, 0.005256569944322109, 0.08753132820129395, 0.1235709860920906, 0.009910017251968384, 0.48670899868011475, -0.17142648994922638, -0.040550194680690765, 0.0002593464741948992, -0.2595714330673218, 0.1623915731906891, 0.03387802094221115, -0.05490421503782272, -0.01830553449690342, 0.07210487872362137, -0.036743953824043274, 0.12226502597332001, -0.3303038775920868, 0.3012453317642212, 0.2769792377948761, -0.0871916264295578, -0.015070867724716663, 0.23038533329963684, -0.08356577157974243, -0.19793176651000977, 0.3110864758491516, 0.07949718832969666, 0.4178656339645386, -0.009903163649141788, 0.9180452823638916, -0.30068135261535645, -0.4647102355957031, 0.011499481275677681, 0.5212082266807556, 0.14811085164546967, -0.25235921144485474, 0.06891921162605286, 0.0899171531200409, -0.2977737784385681, 0.09773928672075272, -0.009851986542344093, -0.18153241276741028, 0.39744821190834045, -0.010155903175473213, 0.19723694026470184, -0.1302175521850586, -0.1488279402256012, 0.3492196500301361, 0.17988631129264832, -0.00638532405719161, 0.08323734253644943, 0.03972681239247322, -0.10503066331148148, -0.17178568243980408, 0.5120564103126526, 0.08268097788095474, -0.0702284500002861, -0.5144467949867249, 0.022896844893693924, 0.15500874817371368, 0.1871422976255417, 0.12166785448789597, 0.20464268326759338, 0.4158383011817932, -0.19626379013061523, 0.36525967717170715, -0.08221456408500671, -0.03379237651824951, -0.18054720759391785, 0.060573823750019073, -0.428354412317276, -0.464821994304657, 0.468281090259552, 0.05242592841386795, -0.14958122372627258, 0.16356684267520905, 0.24570222198963165, -0.6194172501564026, -0.2717808783054352, 0.32530948519706726, -0.1706252098083496, -0.09467201679944992, -0.1400817185640335, 0.03522844985127449, 0.015107291750609875, 0.7085170149803162, 0.2406422197818756, 0.4471988379955292, -0.3157654404640198, -0.06770242750644684, -0.5137097239494324, -0.13571318984031677, 0.1859472393989563, -0.07559226453304291, -0.18541443347930908, 0.05579252168536186, 0.04734259471297264, 0.31720688939094543, -0.08298204094171524, 0.16424746811389923, 0.1825135350227356, 0.19075433909893036, -0.5117715001106262, -0.04429322108626366, -0.023751981556415558, 0.23374812304973602, 0.24095746874809265, -0.1344553679227829, -0.21129441261291504, 0.1133662685751915, -0.11925774067640305, -0.10776684433221817, -0.12228354811668396, 0.34944555163383484, 0.057620592415332794, 0.10408420860767365, 0.0757717564702034, 0.5753722190856934, 0.06265440583229065, -0.19474078714847565, 0.04764428362250328, -0.051162369549274445, -0.29825636744499207, -0.09186097979545593, -0.07321760058403015, 0.11370586603879929, -0.3598737418651581, 0.35187703371047974, 0.013054987415671349, -0.026847273111343384, 0.11974840611219406, -0.05373810604214668, -0.07982292771339417, -0.13715869188308716, 0.06085817888379097, 0.19985537230968475, 0.25871777534484863, 0.7949309945106506, -0.1530762016773224, 0.1390356570482254, -0.1364794820547104, 0.05447755008935928, 0.10467564314603806, -0.12959639728069305, -0.21673141419887543, -0.11664749681949615, 0.05686800181865692, 0.044898051768541336, 0.1433512270450592, -0.5357409715652466, -0.28407585620880127, 0.12216445803642273, 0.005257176700979471, 0.017482200637459755, 0.7596599459648132, 0.28241315484046936, -0.1751401722431183, -0.051766667515039444, -0.1894180178642273, 0.008701004087924957, -0.22647447884082794, 0.0978517234325409, -0.28781479597091675 ]
https://github.com/huggingface/datasets/issues/216
❓ How to get ROUGE-2 with the ROUGE metric ?
For the rouge2 metric you can do ```python rouge = nlp.load_metric('rouge') with open("pred.txt") as p, open("ref.txt") as g: for lp, lg in zip(p, g): rouge.add(lp, lg) score = rouge.compute(rouge_types=["rouge2"]) ``` Note that I just did a PR to have both `.add` and `.add_batch` for metrics, that's why now this is `rouge.add(lp, lg)` and not `rouge.add([lp], [lg])`
I'm trying to use ROUGE metric, but I don't know how to get the ROUGE-2 metric. --- I compute scores with : ```python import nlp rouge = nlp.load_metric('rouge') with open("pred.txt") as p, open("ref.txt") as g: for lp, lg in zip(p, g): rouge.add([lp], [lg]) score = rouge.compute() ``` then : _(print only the F-score for readability)_ ```python for k, s in score.items(): print(k, s.mid.fmeasure) ``` It gives : >rouge1 0.7915168355671788 rougeL 0.7915168355671788 --- **How can I get the ROUGE-2 score ?** Also, it's seems weird that ROUGE-1 and ROUGE-L scores are the same. Did I made a mistake ? @lhoestq
56
❓ How to get ROUGE-2 with the ROUGE metric ? I'm trying to use ROUGE metric, but I don't know how to get the ROUGE-2 metric. --- I compute scores with : ```python import nlp rouge = nlp.load_metric('rouge') with open("pred.txt") as p, open("ref.txt") as g: for lp, lg in zip(p, g): rouge.add([lp], [lg]) score = rouge.compute() ``` then : _(print only the F-score for readability)_ ```python for k, s in score.items(): print(k, s.mid.fmeasure) ``` It gives : >rouge1 0.7915168355671788 rougeL 0.7915168355671788 --- **How can I get the ROUGE-2 score ?** Also, it's seems weird that ROUGE-1 and ROUGE-L scores are the same. Did I made a mistake ? @lhoestq For the rouge2 metric you can do ```python rouge = nlp.load_metric('rouge') with open("pred.txt") as p, open("ref.txt") as g: for lp, lg in zip(p, g): rouge.add(lp, lg) score = rouge.compute(rouge_types=["rouge2"]) ``` Note that I just did a PR to have both `.add` and `.add_batch` for metrics, that's why now this is `rouge.add(lp, lg)` and not `rouge.add([lp], [lg])`
[ 0.08258155733346939, -0.44010505080223083, -0.09260937571525574, 0.3807034492492676, -0.05949254706501961, -0.1371210515499115, -0.3506767451763153, 0.09061744809150696, 0.02350207418203354, 0.2352948784828186, -0.25593113899230957, 0.16841112077236176, -0.021551530808210373, 0.0022069502156227827, 0.09740132838487625, -0.31786614656448364, 0.037423115223646164, -0.03475522994995117, 0.16254755854606628, -0.2055857926607132, -0.15831127762794495, 0.3564314842224121, -0.06627991795539856, 0.26760706305503845, 0.05491020530462265, 0.22830094397068024, 0.18456704914569855, 0.20209920406341553, -0.41247716546058655, -0.17357990145683289, 0.14888429641723633, 0.005113252904266119, 0.23019839823246002, 0.29974085092544556, -0.0001134504855144769, -0.22685807943344116, 0.16347800195217133, -0.09964251518249512, 0.10824290663003922, -0.5389807224273682, 0.06842223554849625, -0.2803023159503937, -0.1171010285615921, -0.050359971821308136, -0.07701842486858368, -0.2652795910835266, -0.30860164761543274, -0.1595865786075592, 0.30006301403045654, 0.15584254264831543, 0.13660471141338348, -0.20554190874099731, -0.062215253710746765, -0.10612660646438599, 0.26183760166168213, -0.10174686461687088, 0.1145925521850586, 0.5414429306983948, 0.2648921310901642, 0.00353650888428092, 0.12690424919128418, 0.14564748108386993, 0.2574688792228699, 0.06553052365779877, 0.30028995871543884, 0.21130535006523132, 0.31865331530570984, -0.11210092157125473, -0.19202947616577148, 0.28055843710899353, 0.12960191071033478, -0.1342126876115799, -0.302422434091568, 0.10026373714208603, 0.0776541531085968, -0.5770207047462463, -0.2875576317310333, -0.10818680375814438, 0.1591963768005371, -0.21189647912979126, -0.19137606024742126, -0.04324133321642876, -0.15449833869934082, 0.1063506007194519, 0.07333295047283173, -0.011297439225018024, -0.053659237921237946, 0.19882185757160187, 0.18406549096107483, -0.00339867384172976, -0.5394593477249146, 0.21521300077438354, -0.314208984375, 0.29421359300613403, -0.30655211210250854, -0.09514345228672028, 0.18695148825645447, 0.4066509008407593, 0.10325285792350769, 0.3489806354045868, 0.11953504383563995, -0.2580714821815491, -0.2762499451637268, 0.08729398250579834, -0.06995879858732224, 0.40451496839523315, 0.19604943692684174, 0.33812716603279114, -0.028442000970244408, 0.069270059466362, 0.19545577466487885, -0.06287293881177902, 0.2469354271888733, -0.4772728681564331, -0.0635710209608078, 0.4770829677581787, -0.0077677140943706036, -0.5834079384803772, -0.458800345659256, -0.028733111917972565, -0.3379412889480591, -0.2079697698354721, 0.15753038227558136, 0.1054937094449997, -0.2072848528623581, 0.3233107924461365, -0.03419431298971176, 0.27940356731414795, -0.32626089453697205, -0.11367327719926834, -0.15017753839492798, 0.2062859833240509, -0.2022613137960434, 0.22359465062618256, 0.15394027531147003, 0.1776585429906845, 0.2562231719493866, 0.2119540125131607, 0.2691638767719269, 0.05984479561448097, 0.2062678188085556, -0.23722749948501587, -0.1175263300538063, -0.14814843237400055, -0.03447932004928589, 0.07741110026836395, 0.188406839966774, -0.27248072624206543, -0.12267471104860306, -0.07505448162555695, -0.30150270462036133, -0.008742393925786018, 0.159201979637146, 0.13326121866703033, 0.1258598268032074, 0.016422266140580177, -0.2135908603668213, 0.3548075556755066, -0.15039624273777008, -0.2694525718688965, 0.2013503462076187, -0.07212842255830765, -0.2583302855491638, -0.1263648122549057, 0.22931286692619324, -0.32643333077430725, -0.17086640000343323, 0.04430997371673584, 0.2678414285182953, -0.08543535321950912, 0.058966487646102905, 0.3666199743747711, 0.0015424714656546712, 0.19801266491413116, -0.15210507810115814, -0.5923179984092712, 0.6651841402053833, -0.7773889899253845, -0.06579883396625519, -0.0005516528617590666, -0.11361798644065857, -0.15772877633571625, 0.13908760249614716, 0.029331594705581665, -0.29635360836982727, -0.20137421786785126, 0.3400665521621704, -0.031103914603590965, -0.15010985732078552, -0.3194093406200409, -0.09263499081134796, 0.33678963780403137, 0.45280492305755615, -0.1628914773464203, -0.04249383136630058, -0.11781314015388489, -0.08548730611801147, -0.041740164160728455, 0.4088154435157776, 0.0696653351187706, -0.1271551102399826, 0.021893499419093132, -0.18314966559410095, 0.16435642540454865, 0.22998687624931335, -0.2909383773803711, 0.25405117869377136, 0.10766607522964478, -0.9638217687606812, -0.2275739163160324, 0.4533103108406067, -0.2547580599784851, -0.6550402045249939, -0.20376750826835632, 0.3090626001358032, -0.2672730088233948, 0.09786123037338257, -0.16166478395462036, 0.36612942814826965, 0.06151772290468216, -0.12694720923900604, 0.096696637570858, 0.0601763017475605, -0.004192337393760681, -0.3270717263221741, -0.1326190084218979, 0.10517575591802597, 0.07881271094083786, 0.25295379757881165, 0.4529840052127838, 0.4550018906593323, 0.26617857813835144, 0.14760783314704895, 0.23407474160194397, 0.22817154228687286, 0.11697331070899963, 0.316596120595932, 0.28234419226646423, 0.1658865064382553, 0.06817023456096649, -0.06692598760128021, 0.4591514468193054, 0.3160402178764343, -0.1627163290977478, -0.1957544982433319, 0.41319331526756287, -0.48462262749671936, 0.12427722662687302, 0.027904076501727104, 0.012188560329377651, -0.20597568154335022, 0.1272016018629074, -0.406619668006897, 0.004663189873099327, 0.23972119390964508, -0.10887482017278671, 0.13831008970737457, -0.26548030972480774, 0.07673665136098862, -0.06142406910657883, 0.1953820437192917, 0.021580172702670097, -0.03668477013707161, -0.13535979390144348, -0.0005476675578393042, -0.043356601148843765, -0.09061430394649506, -0.20829983055591583, 0.18548227846622467, 0.17604278028011322, 0.2232333868741989, 0.12233293801546097, -0.3682175576686859, -0.1841256320476532, -0.053566280752420425, -0.05732811242341995, 0.016051005572080612, 0.02372909151017666, 0.11102531850337982, -0.04540048539638519, 0.1383504718542099, -0.284076452255249, -0.25924456119537354, 0.10575330257415771, -0.35086849331855774, 0.10057388246059418, -0.09940113872289658, -0.11004449427127838, -0.17763246595859528, -0.5282853841781616, -0.13439427316188812, -0.34777596592903137, -0.02615230716764927, 0.0755162388086319, 0.2976171672344208, -0.2034304141998291, 0.21936529874801636, 0.18032537400722504, -0.10814514011144638, 0.3281554579734802, 0.10603364557027817, -0.21060998737812042, -0.2679210603237152, 0.17807550728321075, -0.22155305743217468, 0.007940058596432209, 0.24635525047779083, -0.1296309232711792, -0.007005790714174509, 0.15436023473739624, -0.2549620270729065, -0.09858163446187973, -0.05393831059336662, 0.22666549682617188, 0.062423285096883774, -0.28835755586624146, -0.20405597984790802, 0.3045313358306885, 0.23586611449718475, -0.024508817121386528, -0.261688232421875, 0.03409338742494583, -0.19851404428482056, 0.20603403449058533, 0.052129216492176056, -0.18085193634033203, 0.06330055743455887, -0.0806746557354927, 0.4839155972003937, 0.24622279405593872, -0.0729643851518631, 0.2792002558708191, -0.3118186891078949, 0.2519313097000122, 0.31410765647888184, 0.37752342224121094, -0.3375565707683563, -0.2846820056438446, 0.31501686573028564, -0.41623133420944214, -0.3131351172924042, -0.09678281843662262, 0.06253837049007416, -0.013087146915495396, -0.08516311645507812, 0.06592990458011627, -0.34703490138053894, 0.03881409019231796, -0.041318900883197784, -0.1230342835187912, 0.18541386723518372, 0.24729028344154358, -0.2666778862476349, -0.05914135277271271, -0.10685449838638306, 0.08153596520423889, -0.01344816479831934, 0.10353575646877289, 0.5398581624031067, -0.27826380729675293, 0.11782331019639969, 0.13101071119308472, 0.34448787569999695, 0.37147143483161926, 0.4343835115432739, 0.05482850223779678, 0.1600295752286911, 0.029157012701034546, 0.0896739587187767, -0.17208275198936462, 0.30430707335472107, 0.07725419104099274, -0.15839846432209015, 0.0700274407863617, 0.1549830287694931, 0.42408931255340576, -0.28693440556526184, -0.03230747580528259, -0.10815959423780441, 0.106789730489254, -0.08260015398263931, -0.26620009541511536, -0.039602555334568024, -0.08259399235248566, 0.042218130081892014, -0.12649735808372498, -0.10205915570259094, 0.30802634358406067, 0.036288708448410034, 0.2076338827610016, 0.1973804384469986, -0.37305396795272827, -0.0860799178481102, -0.30504995584487915, 0.4734286069869995, 0.09159287065267563, -0.00303440960124135, 0.12427815049886703, 0.09895510226488113, -0.01705552078783512, -0.19436296820640564, 0.4446815252304077, 0.036086034029722214, 0.19325853884220123, 0.29458633065223694, -0.19206920266151428, -0.6826239824295044, -0.17620934545993805, -0.31174856424331665, 0.2777722477912903, 0.16995839774608612, 0.08621393144130707, -0.17443951964378357, 0.057886719703674316, 0.19328728318214417, 0.0027842901181429625, -0.1897594928741455, 0.027197791263461113, -0.31895625591278076, 0.015281899832189083, 0.32539862394332886, -0.21736939251422882, -0.11527746170759201, 0.5167659521102905, -0.0853353813290596, 0.3249637484550476, 0.09882167726755142, -0.25806915760040283, 0.4550524652004242, -0.002124434569850564, 0.2430940419435501, 0.11786956340074539, -0.18184930086135864, 0.15938535332679749, -0.13895131647586823, -0.5745753049850464, 0.33481156826019287, 0.09035544842481613, -0.5968512296676636, -0.10880061984062195, 0.00811954215168953, 0.48261627554893494, 0.0303758904337883, -0.05051271244883537, -0.08397011458873749, -0.44186103343963623, -0.11237693578004837, -0.5094918608665466, -0.13879302144050598, 0.2701757550239563, -0.0065772500820457935, -0.11409121751785278, -0.3621305227279663, 0.5230094790458679, 0.3907352387905121, -0.331646203994751, 0.35048091411590576, -0.2329188585281372, -0.20785191655158997, 0.22355642914772034, 0.07861439138650894, 0.7750361561775208, 0.23556798696517944, -0.0316280722618103, 0.12106256932020187, -0.256979376077652, 0.5335085988044739, -0.6500670909881592, 0.07334934175014496, -0.2151559591293335, -0.268425315618515, -0.039984554052352905, 0.06009405851364136, 0.048210788518190384, 0.18227246403694153, 0.08014410734176636, 0.5681713819503784, -0.06007906794548035, -0.15259864926338196, -0.03092052787542343, 0.25361138582229614, 0.09966547787189484, -0.12943455576896667, -0.1757189780473709, 0.16745468974113464, -0.1329164206981659, 0.00017677561845630407, -0.27084091305732727, 0.11507463455200195, -0.5599821209907532, 0.06690474599599838, -0.14745046198368073, -0.13702446222305298, -0.18371455371379852, 0.17790383100509644, 0.13588498532772064, -0.42254069447517395, -0.33103421330451965, 0.5675349831581116, 0.725046694278717, 0.06679334491491318, -0.2974846661090851, 0.2634715437889099, -0.17508454620838165, 0.11545851826667786, 0.0364186055958271, -0.09256564825773239, 0.05644403025507927, -0.028396984562277794, -0.02011786587536335, 0.32481008768081665, 0.3469853103160858, -0.002153206616640091, -0.17017854750156403, 0.009425805881619453, -0.14369678497314453, 0.04415067285299301, -0.0592578649520874, 0.24112568795681, 0.34101009368896484, 0.07543107122182846, 0.08714169263839722, 0.2268550843000412, -0.5656334161758423, -0.05264560505747795, 0.20045378804206848, -0.1952410638332367, 0.14268119633197784, 0.2905636727809906, 0.03757636249065399, -0.293536514043808, -0.22153706848621368, 0.0875626653432846, 0.07894787937402725, -0.005835336167365313, -0.04641568288207054, 0.016473950818181038, -0.06334932148456573, 0.01827385276556015, 0.287295937538147, 0.15044505894184113, 0.03996080160140991, 0.506568431854248, 0.3182370364665985, -0.13108158111572266, 0.2522672414779663, -0.45389872789382935, -0.2241072952747345, 0.19057060778141022, 0.08486582338809967, -0.08753083646297455, -0.08937346190214157, -0.08392159640789032, 0.1378520131111145, 0.542251467704773, -0.28925949335098267, -0.23756715655326843, -0.04239954054355621, 0.14428743720054626, -0.16340066492557526, -0.13924908638000488, -0.10749528557062149, 0.14073313772678375, 0.06439004838466644, 0.04496736079454422, -0.18612630665302277, -0.1794566959142685, -0.2494574636220932, 0.1313873827457428, -0.1495238095521927, 0.16889534890651703, -0.06284750998020172, 0.07508085668087006, 0.184678852558136, -0.08171870559453964, -0.05887357518076897, 0.3747638463973999, 0.37351539731025696, 0.3379550874233246, -0.03439314663410187, 0.10218127071857452, -0.056324250996112823, -0.017829785123467445, 0.21329465508460999, 0.2078014463186264, 0.07900924235582352, -0.00853796023875475, 0.14385895431041718, -0.016868287697434425, -0.28753483295440674, -0.08985932171344757, -0.14779630303382874, -0.22689560055732727, 0.14476226270198822, -0.01856963150203228, 0.20657089352607727, -0.21333351731300354, 0.3798781633377075, 0.13268712162971497, -0.07608450204133987, 0.18973107635974884, -0.18290548026561737, 0.1428568810224533, -0.11980114877223969, 0.40618014335632324, 0.1725999414920807, 0.25368550419807434, -0.2150341272354126, 0.258207231760025, -0.033284641802310944, 0.1541895717382431, -0.011891248635947704, -0.008981614373624325, -0.027472520247101784, -0.3529621660709381, 0.21023701131343842, -0.4513002932071686, 0.1374829113483429, -0.07608562707901001, 0.3451693058013916, -0.14173297584056854, 0.08663608133792877, 0.09466511756181717, 0.11317379772663116, 0.20618656277656555, 0.05709479749202728, 0.3819494843482971, 0.05865371227264404, 0.1647115796804428, -0.3350290060043335, 0.2503338158130646, 0.03729349747300148, -0.05603528395295143, 0.06057858094573021, 0.1348668932914734, 0.040382348001003265, -0.19460339844226837, -0.1217290461063385, 0.10459097474813461, 0.1230267733335495, -0.39379435777664185, -0.3598279356956482, -0.25772836804389954, -0.19753225147724152, -0.056885503232479095, -0.004424795508384705, 0.09338676184415817, -0.14682893455028534, 0.05632414668798447, -0.05959947407245636, -0.29237157106399536, -0.398826003074646, 0.15626993775367737, 0.344105988740921, -0.07433120906352997, 0.10791236907243729, 0.20615963637828827, 0.3127463161945343, 0.3157906234264374, 0.3679596483707428, 0.1327415108680725, 0.08966732770204544, -0.5088694095611572, 0.05510054901242256, -0.18205170333385468, -0.07493681460618973, 0.03659472241997719, 0.2758987247943878, 0.19292719662189484, -0.17557115852832794, 0.31664589047431946, 0.15414322912693024, -0.10532353818416595, 0.3391839861869812, -0.14225612580776215, 0.059374481439590454, -0.41194042563438416, 0.5525143146514893, 0.0364963561296463, 0.20788897573947906, -0.20443876087665558, 0.05917315185070038, -0.47369927167892456, -0.18732336163520813, -0.5115629434585571, 0.06559554487466812, 0.06152236834168434, -0.3025525212287903, 0.07796745747327805, 0.08984465897083282, 0.024085137993097305, 0.08686548471450806, 0.42110976576805115, -0.3376670479774475, -0.19173629581928253, -0.38417208194732666, -0.3622974455356598, -0.04848925769329071, 0.11754706501960754, -0.24215500056743622, 0.28721028566360474, -0.06774526089429855, 0.541233479976654, 0.15837720036506653, 0.4834287762641907, 0.05399258807301521, 0.17747762799263, -0.20079343020915985, -0.012256138026714325, -0.09126152843236923, -0.10455956310033798, -0.1025996133685112, 0.22864001989364624, -0.21448960900306702, -0.17608579993247986, 0.036831334233284, -0.15790362656116486, -0.2663091719150543, 0.24678389728069305, 0.6915806531906128, 0.3145757019519806, 0.35585716366767883, 0.4649411141872406, -0.019129948690533638, 0.03926919773221016, -0.05851525440812111, -0.0018536302959546447, -0.05582151934504509, 0.1583450436592102, -0.38414108753204346, 0.3125040829181671, -0.30256569385528564, -0.22766636312007904, 0.10168587416410446, 0.6001901030540466, 0.00941335316747427, -0.20099976658821106, -0.39312344789505005, -0.07098246365785599, 0.018479134887456894, -0.14986266195774078, -0.41239190101623535, 0.33785268664360046, -0.015126701444387436, 0.4427173137664795, -0.009619197808206081, -0.13172683119773865, 0.478371262550354, -0.36559736728668213, 0.06721808761358261, 0.061627019196748734, 0.38429129123687744, 0.2618561089038849, -0.07773318886756897, -0.19687677919864655, -0.03381285071372986, 0.3542476296424866, 0.1926385760307312, -0.0012844557641074061, 0.16451597213745117, -0.1842232495546341, -0.10394877940416336, -0.15975996851921082, -0.01980188488960266, 0.06336183100938797, 0.04165369272232056, 0.10203717648983002, -0.16577677428722382 ]
https://github.com/huggingface/datasets/issues/216
❓ How to get ROUGE-2 with the ROUGE metric ?
Well I just tested with the official script and both rouge1 and rougeL return exactly the same thing for the input you gave, so this is actually fine ^^ I hope it helped :)
I'm trying to use ROUGE metric, but I don't know how to get the ROUGE-2 metric. --- I compute scores with : ```python import nlp rouge = nlp.load_metric('rouge') with open("pred.txt") as p, open("ref.txt") as g: for lp, lg in zip(p, g): rouge.add([lp], [lg]) score = rouge.compute() ``` then : _(print only the F-score for readability)_ ```python for k, s in score.items(): print(k, s.mid.fmeasure) ``` It gives : >rouge1 0.7915168355671788 rougeL 0.7915168355671788 --- **How can I get the ROUGE-2 score ?** Also, it's seems weird that ROUGE-1 and ROUGE-L scores are the same. Did I made a mistake ? @lhoestq
34
❓ How to get ROUGE-2 with the ROUGE metric ? I'm trying to use ROUGE metric, but I don't know how to get the ROUGE-2 metric. --- I compute scores with : ```python import nlp rouge = nlp.load_metric('rouge') with open("pred.txt") as p, open("ref.txt") as g: for lp, lg in zip(p, g): rouge.add([lp], [lg]) score = rouge.compute() ``` then : _(print only the F-score for readability)_ ```python for k, s in score.items(): print(k, s.mid.fmeasure) ``` It gives : >rouge1 0.7915168355671788 rougeL 0.7915168355671788 --- **How can I get the ROUGE-2 score ?** Also, it's seems weird that ROUGE-1 and ROUGE-L scores are the same. Did I made a mistake ? @lhoestq Well I just tested with the official script and both rouge1 and rougeL return exactly the same thing for the input you gave, so this is actually fine ^^ I hope it helped :)
[ 0.12399008870124817, -0.49795085191726685, -0.08898469060659409, 0.3463135063648224, -0.08415459096431732, -0.2206086665391922, -0.3729407787322998, 0.080358125269413, 0.036045148968696594, 0.22104912996292114, -0.2184777706861496, 0.23340599238872528, -0.011108973063528538, -0.0014879658119753003, 0.07686194032430649, -0.28436294198036194, -0.011588569730520248, -0.03263470157980919, 0.08533453941345215, -0.24821960926055908, -0.10775768756866455, 0.40545654296875, -0.15447036921977997, 0.25245705246925354, 0.06542922556400299, 0.3166957497596741, 0.170079305768013, 0.21432356536388397, -0.43286556005477905, -0.18212702870368958, 0.16527074575424194, 0.05036305636167526, 0.14969542622566223, 0.23267754912376404, -0.0001142661421908997, -0.29253774881362915, 0.18494346737861633, -0.1414395123720169, 0.1738610863685608, -0.6005343198776245, 0.058325521647930145, -0.26914945244789124, -0.08119753003120422, -0.06456372141838074, -0.1110232025384903, -0.19320006668567657, -0.28851518034935, -0.12736763060092926, 0.3743640184402466, 0.2322625070810318, 0.1280948966741562, -0.2020396888256073, -0.06131448224186897, -0.05112849920988083, 0.24692222476005554, -0.08830829709768295, 0.08550655841827393, 0.5601392984390259, 0.26604458689689636, -0.03746877983212471, 0.08142247796058655, 0.1509748101234436, 0.23171301186084747, 0.009744318202137947, 0.39035800099372864, 0.17862598598003387, 0.24379891157150269, -0.09382635354995728, -0.22313925623893738, 0.32748842239379883, 0.18336033821105957, -0.1542433500289917, -0.2336723655462265, 0.1524389684200287, 0.027276013046503067, -0.5276967287063599, -0.2502237856388092, -0.09636113792657852, 0.20212006568908691, -0.153611958026886, -0.24245621263980865, -0.030753642320632935, -0.2164042443037033, 0.08503406494855881, 0.00017964419384952635, -0.007078643888235092, -0.06696687638759613, 0.16879858076572418, 0.2031935751438141, 0.011865269392728806, -0.5337009429931641, 0.2507765591144562, -0.3072171211242676, 0.3494212329387665, -0.37095749378204346, -0.16411682963371277, 0.13272228837013245, 0.4145556092262268, 0.10024268180131912, 0.40116146206855774, 0.22469502687454224, -0.1404869109392166, -0.267223596572876, 0.1099192351102829, -0.03308188170194626, 0.4469316601753235, 0.22990891337394714, 0.3148643970489502, -0.027802443131804466, 0.025183141231536865, 0.1482384204864502, -0.02264491841197014, 0.23133720457553864, -0.456952303647995, -0.038009654730558395, 0.39786869287490845, -0.027401261031627655, -0.6806560754776001, -0.4113566279411316, 0.051422227174043655, -0.3216807246208191, -0.14020481705665588, 0.11257915198802948, 0.0980573445558548, -0.1921643763780594, 0.330240398645401, -0.007844320498406887, 0.2373749315738678, -0.331519216299057, -0.1402774155139923, -0.07846434414386749, 0.2287806272506714, -0.25133273005485535, 0.21520040929317474, 0.18882884085178375, 0.21518763899803162, 0.2448730319738388, 0.20560802519321442, 0.3032255172729492, 0.029210476204752922, 0.2829256057739258, -0.2480918914079666, -0.02712087891995907, -0.14943131804466248, -0.07167086750268936, 0.060484472662210464, 0.11274910718202591, -0.2844979763031006, -0.09193672984838486, -0.05520607903599739, -0.2921646237373352, 0.009701142087578773, 0.1461050808429718, 0.13343605399131775, 0.07843813300132751, 0.012796142138540745, -0.11202649772167206, 0.3773162066936493, -0.20596849918365479, -0.26619935035705566, 0.21629922091960907, -0.14620496332645416, -0.28263136744499207, -0.1256779581308365, 0.21286283433437347, -0.3043459355831146, -0.15618181228637695, 0.08477049320936203, 0.25281721353530884, -0.0532623827457428, 0.014621610753238201, 0.4060993194580078, 0.06860829144716263, 0.15228421986103058, -0.09097372740507126, -0.5238790512084961, 0.6502173542976379, -0.7507554888725281, -0.09616758674383163, 0.0344623439013958, -0.0805513858795166, -0.10891702026128769, 0.08223535865545273, 0.021922646090388298, -0.3146570920944214, -0.14471319317817688, 0.28157612681388855, 0.010719622485339642, -0.08700982481241226, -0.31078410148620605, -0.15044115483760834, 0.39708709716796875, 0.4027325510978699, -0.07194726914167404, -0.08274543285369873, -0.17706604301929474, -0.12735801935195923, -0.016922172158956528, 0.4076840579509735, 0.037556473165750504, -0.12502363324165344, 0.11389704793691635, -0.19675959646701813, 0.1464032679796219, 0.27248284220695496, -0.3125994801521301, 0.34934088587760925, 0.07731451839208603, -0.8894869685173035, -0.17240449786186218, 0.4622156620025635, -0.22569240629673004, -0.6651102304458618, -0.2151516079902649, 0.3109324276447296, -0.31003403663635254, 0.10217607766389847, -0.22574616968631744, 0.3516808748245239, 0.1139778345823288, -0.15364307165145874, 0.20693862438201904, 0.061979904770851135, 0.0069306278601288795, -0.31909796595573425, -0.16207414865493774, 0.10615484416484833, 0.05850504711270332, 0.2559981644153595, 0.47424280643463135, 0.4499499201774597, 0.22688065469264984, 0.16882717609405518, 0.21439839899539948, 0.25388383865356445, 0.034971825778484344, 0.25829946994781494, 0.20957620441913605, 0.19598017632961273, -0.01846867799758911, -0.07744082063436508, 0.5199013352394104, 0.30376070737838745, -0.15413634479045868, -0.19438490271568298, 0.46627509593963623, -0.49039340019226074, 0.057442229241132736, 0.030123040080070496, 0.003535803873091936, -0.23440827429294586, 0.04894731566309929, -0.4269282817840576, 0.008210007101297379, 0.3378644287586212, -0.07221459597349167, 0.15080951154232025, -0.21438336372375488, 0.07765533030033112, -0.058721981942653656, 0.1669670194387436, -0.023158853873610497, -0.0700710341334343, -0.09916824102401733, 0.04828375205397606, -0.052478812634944916, -0.08713266253471375, -0.14648179709911346, 0.13671371340751648, 0.14246462285518646, 0.1714954674243927, 0.2192109078168869, -0.4117482900619507, -0.18401721119880676, -0.00791239831596613, -0.15325428545475006, 0.012183813378214836, 0.015409658662974834, 0.06150908023118973, -0.07383936643600464, 0.05183932185173035, -0.3049065172672272, -0.3022271990776062, 0.06790226697921753, -0.32608282566070557, 0.0888364389538765, -0.10181170701980591, -0.1412833333015442, -0.1743474304676056, -0.459698349237442, -0.15071244537830353, -0.3067997694015503, -0.037474554032087326, 0.07321818917989731, 0.2997550666332245, -0.298447847366333, 0.21187330782413483, 0.20868965983390808, -0.09639443457126617, 0.35062742233276367, 0.09900621324777603, -0.2984909117221832, -0.21889297664165497, 0.148515984416008, -0.22972339391708374, 0.010995885357260704, 0.20694595575332642, -0.1500377058982849, -0.0032065054401755333, 0.14732475578784943, -0.2885565459728241, -0.0999545156955719, -0.02525883913040161, 0.16488507390022278, 0.05572749674320221, -0.33034342527389526, -0.21322490274906158, 0.3580799698829651, 0.15705740451812744, -0.043253060430288315, -0.25011029839515686, 0.09181647002696991, -0.12921252846717834, 0.13602477312088013, -0.0053225234150886536, -0.2049800604581833, 0.07565301656723022, -0.016804229468107224, 0.48908236622810364, 0.23215335607528687, -0.025979578495025635, 0.2851453125476837, -0.35402461886405945, 0.2125239223241806, 0.25654518604278564, 0.39148715138435364, -0.32673799991607666, -0.27257564663887024, 0.3942364752292633, -0.43585824966430664, -0.3721911609172821, -0.10294661670923233, 0.10516542196273804, 0.02943924255669117, -0.044423818588256836, 0.05712873861193657, -0.32959774136543274, 0.054108407348394394, -0.0908435806632042, -0.17520231008529663, 0.1330191195011139, 0.22480043768882751, -0.26698818802833557, -0.04010305926203728, -0.12790347635746002, 0.09605003148317337, -0.0023119349498301744, 0.13142900168895721, 0.5086544156074524, -0.2481072098016739, 0.07726944983005524, 0.14046289026737213, 0.27565672993659973, 0.4029235541820526, 0.4460241198539734, 0.04914803057909012, 0.16979755461215973, 0.02271304838359356, 0.0836675837635994, -0.1309284120798111, 0.28202977776527405, 0.04214993119239807, -0.12101132422685623, 0.10032157599925995, 0.184701606631279, 0.33166590332984924, -0.3414808213710785, -0.08046235889196396, -0.09187593311071396, 0.09235560148954391, -0.0638105496764183, -0.24431459605693817, -0.03437425568699837, -0.05484861135482788, 0.05709228664636612, -0.09766294062137604, -0.134943425655365, 0.3102973401546478, 0.022705547511577606, 0.21679846942424774, 0.16918714344501495, -0.3748897314071655, -0.028954531997442245, -0.3219945728778839, 0.42253366112709045, 0.0653609037399292, -0.03763560578227043, 0.13242244720458984, 0.06527949869632721, 0.006213005166500807, -0.22974920272827148, 0.3863010108470917, 0.06365003436803818, 0.21256239712238312, 0.265256404876709, -0.2225475013256073, -0.670853316783905, -0.1808362454175949, -0.31598493456840515, 0.2895922064781189, 0.1978677362203598, 0.023055341094732285, -0.15612204372882843, 0.08311475813388824, 0.18468430638313293, -0.028113199397921562, -0.18701830506324768, 0.04569750651717186, -0.2581057548522949, 0.009998743422329426, 0.3646010458469391, -0.22924284636974335, -0.14277887344360352, 0.47826141119003296, -0.06072985380887985, 0.34218454360961914, 0.14983750879764557, -0.21545206010341644, 0.45361700654029846, 0.04777529090642929, 0.27631431818008423, 0.02659096196293831, -0.20807835459709167, 0.16249145567417145, -0.0944196805357933, -0.6795375347137451, 0.33890724182128906, 0.1927538812160492, -0.5905694365501404, -0.16465133428573608, -0.037636324763298035, 0.4764319956302643, 0.05458896979689598, -0.011679058894515038, -0.09295879304409027, -0.4811422824859619, -0.06214674562215805, -0.5296534895896912, -0.12390255928039551, 0.26546961069107056, -0.008957031182944775, -0.11165093630552292, -0.3016607165336609, 0.48601284623146057, 0.3428497016429901, -0.31324803829193115, 0.3125554621219635, -0.19474424421787262, -0.2029215395450592, 0.25724318623542786, 0.10273773968219757, 0.811381459236145, 0.21375557780265808, 0.002194369677454233, 0.07617899030447006, -0.3513456881046295, 0.47075125575065613, -0.7156761288642883, 0.044999413192272186, -0.26291918754577637, -0.25327253341674805, -0.04182583466172218, 0.04671798273921013, 0.04493146017193794, 0.21422667801380157, 0.06681949645280838, 0.571744441986084, 0.0646706223487854, -0.1611299216747284, -0.0035067081917077303, 0.2737402319908142, 0.139947310090065, -0.05350298807024956, -0.1496005356311798, 0.145536407828331, -0.1289556622505188, -0.027125362306833267, -0.2474435716867447, 0.09093455225229263, -0.5433628559112549, 0.07716041803359985, -0.16859084367752075, -0.12107470631599426, -0.1406952440738678, 0.11651474982500076, 0.13030695915222168, -0.3935346305370331, -0.2577759325504303, 0.6006626486778259, 0.7157086730003357, 0.12420566380023956, -0.37131863832473755, 0.2642662525177002, -0.20645582675933838, 0.15178203582763672, 0.10992224514484406, -0.05415470525622368, 0.04587847366929054, -0.0570327527821064, -0.06457532942295074, 0.2766418159008026, 0.35905739665031433, -0.010358286090195179, -0.22057031095027924, -0.007234475109726191, -0.1653817892074585, 0.0703059583902359, -0.03975014016032219, 0.20711588859558105, 0.3018411099910736, 0.1060420572757721, 0.08863233774900436, 0.21996474266052246, -0.546942949295044, -0.029552036896348, 0.243146613240242, -0.23111717402935028, 0.16570483148097992, 0.3313014805316925, 0.00890005100518465, -0.2933350205421448, -0.254380464553833, 0.07475218921899796, 0.0823364108800888, -0.00511067034676671, -0.035461947321891785, -0.020485196262598038, -0.04117698594927788, 0.020233001559972763, 0.251038134098053, 0.13301078975200653, 0.021784162148833275, 0.43461349606513977, 0.3064218759536743, -0.1111544743180275, 0.23509341478347778, -0.4188058376312256, -0.1954861581325531, 0.1380896419286728, 0.08050411194562912, -0.11472403258085251, -0.031743232160806656, -0.15798793733119965, 0.17435526847839355, 0.5180916786193848, -0.2837797701358795, -0.18762148916721344, -0.028949078172445297, 0.15590406954288483, -0.13453392684459686, -0.20898345112800598, -0.20269271731376648, 0.07882999628782272, 0.046424899250268936, 0.0644054040312767, -0.18087175488471985, -0.1644836664199829, -0.2384835183620453, 0.1403973400592804, -0.13584986329078674, 0.1390451043844223, -0.0938858836889267, 0.05311373248696327, 0.1827075034379959, -0.08895445615053177, -0.07194344699382782, 0.34918099641799927, 0.36998265981674194, 0.45747196674346924, -0.10201887041330338, 0.08420000225305557, -0.11546166986227036, -0.034587204456329346, 0.2173672467470169, 0.22923825681209564, 0.08090738207101822, -0.024483975023031235, 0.17273011803627014, -0.03619283065199852, -0.2474110722541809, -0.054103586822748184, -0.20524181425571442, -0.2769985795021057, 0.20450380444526672, -0.0031600084621459246, 0.2293151617050171, -0.2594712972640991, 0.4298008978366852, 0.16129881143569946, -0.09349506348371506, 0.1403721570968628, -0.18730011582374573, 0.13596105575561523, -0.11948489397764206, 0.357946515083313, 0.20658087730407715, 0.25414010882377625, -0.24037840962409973, 0.2233319729566574, -0.016074163839221, 0.08376356959342957, 0.06477053463459015, -0.031832627952098846, -0.028623254969716072, -0.32644131779670715, 0.1821545660495758, -0.4298461973667145, 0.15969713032245636, -0.10262668877840042, 0.2641345262527466, -0.15839871764183044, 0.06268320232629776, 0.20385798811912537, 0.11040182411670685, 0.22919881343841553, 0.025116965174674988, 0.42615050077438354, 0.04126860946416855, 0.24798661470413208, -0.30988049507141113, 0.2689864933490753, 0.07964710146188736, -0.03128304332494736, 0.10673469305038452, 0.1408994197845459, 0.11347445100545883, -0.23334640264511108, -0.14484363794326782, 0.06972614675760269, 0.12309932708740234, -0.3971673548221588, -0.3614509403705597, -0.2509463131427765, -0.2693318724632263, -0.017957132309675217, -0.051787711679935455, 0.12138763070106506, -0.10855121165513992, 0.1001819372177124, -0.04484071582555771, -0.2422509640455246, -0.3757950961589813, 0.07377666980028152, 0.4461504817008972, -0.10802500694990158, 0.10562525689601898, 0.1896151602268219, 0.38043293356895447, 0.298563688993454, 0.33002015948295593, 0.1283009797334671, 0.08850307762622833, -0.5871767401695251, 0.10413957387208939, -0.18798470497131348, -0.08341566473245621, 0.07671976089477539, 0.32121095061302185, 0.1787654012441635, -0.16076795756816864, 0.32639607787132263, 0.12612158060073853, -0.06357888132333755, 0.348756343126297, -0.11649792641401291, 0.059141360223293304, -0.3998604416847229, 0.43504664301872253, 0.11445923894643784, 0.21158578991889954, -0.2502632737159729, 0.018330538645386696, -0.530182421207428, -0.18016186356544495, -0.5192390084266663, 0.1562502384185791, 0.05731384456157684, -0.33192944526672363, 0.07285428047180176, 0.12624847888946533, 0.07426141202449799, 0.055764611810445786, 0.3771967589855194, -0.30973026156425476, -0.28629428148269653, -0.4125572741031647, -0.3241789937019348, 0.023713093250989914, 0.07984797656536102, -0.20768001675605774, 0.22943340241909027, -0.04843538999557495, 0.5633389949798584, 0.10816986858844757, 0.4685989320278168, 0.05568750947713852, 0.17474479973316193, -0.24061745405197144, 0.05537017434835434, -0.09380460530519485, -0.13297057151794434, -0.14306572079658508, 0.2436823695898056, -0.17661811411380768, -0.1891975998878479, 0.03014885075390339, -0.12438765168190002, -0.22878091037273407, 0.21405847370624542, 0.7476037740707397, 0.26726171374320984, 0.31729429960250854, 0.4162444770336151, -0.002822617767378688, 0.036342307925224304, -0.007561690639704466, 0.06603803485631943, 0.007561011239886284, 0.20577438175678253, -0.30258405208587646, 0.2575331926345825, -0.21721982955932617, -0.25412821769714355, 0.1553339660167694, 0.6273362040519714, -0.040246859192848206, -0.20091032981872559, -0.35345590114593506, -0.15675123035907745, 0.05964111536741257, -0.1647377461194992, -0.5186479687690735, 0.390350341796875, -0.041450221091508865, 0.4093332588672638, -0.0025362339802086353, -0.08588703721761703, 0.4886847734451294, -0.3368973135948181, 0.113932304084301, 0.11974486708641052, 0.3727897107601166, 0.24169683456420898, -0.11075799912214279, -0.23718160390853882, -0.051548656076192856, 0.2927345931529999, 0.1626879721879959, 0.000671992136631161, 0.2067311406135559, -0.15648233890533447, -0.10198116302490234, -0.17159141600131989, 0.06903044879436493, 0.11228419840335846, 0.06057703495025635, 0.06237468495965004, -0.16444095969200134 ]
https://github.com/huggingface/datasets/issues/215
NonMatchingSplitsSizesError when loading blog_authorship_corpus
I just ran it on colab and got this ``` [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=611607465, num_examples=533285, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=35652716, num_examples=30804, dataset_name='blog_authorship_corpus')}] ``` which is different from the `dataset_infos.json` and also different from yours. It looks like the script for generating examples is not consistent
Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on.
53
NonMatchingSplitsSizesError when loading blog_authorship_corpus Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on. I just ran it on colab and got this ``` [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=611607465, num_examples=533285, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=35652716, num_examples=30804, dataset_name='blog_authorship_corpus')}] ``` which is different from the `dataset_infos.json` and also different from yours. It looks like the script for generating examples is not consistent
[ -0.13186021149158478, 0.1403093934059143, 0.059409480541944504, 0.4210388660430908, -0.10276028513908386, 0.06391100585460663, -0.09939002245664597, 0.2403315305709839, -0.16995199024677277, 0.16742543876171112, -0.0025759756099432707, 0.2693621814250946, 0.10592704266309738, -0.029215309768915176, -0.061288680881261826, 0.11617553979158401, -0.07574369758367538, 0.23838597536087036, 0.06643021106719971, 0.009823927655816078, -0.02960858680307865, 0.4090338945388794, -0.2907909154891968, -0.08376237750053406, -0.26350802183151245, -0.20447713136672974, -0.007695795968174934, 0.26649045944213867, -0.028038399294018745, -0.2925412058830261, 0.2720765173435211, 0.011067833751440048, 0.13318189978599548, 0.19191598892211914, -0.00012480425357352942, 0.022843847051262856, 0.5506143569946289, -0.25053179264068604, -0.21130214631557465, -0.2491970956325531, -0.4140855669975281, -0.28161385655403137, 0.03941405192017555, -0.3059527575969696, -0.07530687004327774, 0.13373887538909912, 0.16358354687690735, -0.052976224571466446, 0.2451407015323639, 0.3204345405101776, 0.11070675402879715, 0.07963863015174866, -0.26864486932754517, 0.13457974791526794, 0.12662196159362793, 0.0818084105849266, -0.036990754306316376, -0.0744718387722969, -0.1828918606042862, -0.17931245267391205, -0.0033600626047700644, 0.46177706122398376, -0.2776544392108917, 0.13988235592842102, 0.1918162852525711, 0.17948715388774872, 0.34826284646987915, -0.13864758610725403, 0.1591961830854416, 0.3170473575592041, 0.4259495139122009, 0.15770386159420013, -0.2048095315694809, -0.4242803156375885, -0.0219612754881382, 0.10003824532032013, 0.47994300723075867, 0.32488173246383667, -0.23731021583080292, 0.0331987701356411, -0.4330624043941498, -0.0763508602976799, -0.03681815043091774, 0.009393339045345783, 0.1595861315727234, 0.3218361735343933, 0.0860607773065567, 0.07757145911455154, 0.09996259212493896, 0.013011145405471325, 0.34360364079475403, -0.27978336811065674, -0.21615813672542572, 0.042084552347660065, -0.10203386098146439, -0.1977798193693161, -0.18273696303367615, 0.06244157627224922, 0.15654721856117249, 0.07547769695520401, 0.37092193961143494, -0.08625392615795135, 0.1798444390296936, 0.11776077747344971, 0.371538907289505, -0.013676805421710014, -0.028967464342713356, 0.5226830244064331, 0.10686852037906647, -0.009326500818133354, 0.1176358312368393, 0.04502756521105766, 0.2934999465942383, -0.1499844640493393, -0.19445843994617462, 0.12557059526443481, 0.1616523563861847, -0.6070362329483032, -0.4450768232345581, 0.016515761613845825, -0.5715517401695251, -0.20781314373016357, -0.05524882674217224, 0.24616649746894836, -0.36231106519699097, 0.26535263657569885, -0.18779288232326508, -0.037590693682432175, -0.32581827044487, -0.01764429174363613, -0.18110638856887817, 0.16775581240653992, -0.13677334785461426, 0.2572120726108551, 0.14054368436336517, -0.18570195138454437, 0.4198700785636902, -0.1785367727279663, -0.030709290876984596, -0.15490326285362244, 0.260629802942276, -0.193294495344162, 0.0708690658211708, 0.22536765038967133, 0.003308118088170886, 0.16629444062709808, -0.09200587868690491, -0.29528069496154785, -0.35808485746383667, 0.1588878035545349, -0.25167956948280334, -0.5216350555419922, 0.1034586951136589, 0.12058121711015701, -0.4729642868041992, -0.03648775443434715, -0.03690144792199135, 0.2329142540693283, 0.45762574672698975, -0.3340526521205902, -0.02695961482822895, -0.1752268224954605, -0.21760708093643188, -0.04367863014340401, -0.2373843938112259, 0.43923506140708923, 0.14354382455348969, -0.23715530335903168, 0.13986776769161224, 0.06780404597520828, 0.6044771671295166, 0.36378055810928345, -0.0025904576759785414, 0.07247541099786758, -0.1579885631799698, 0.3344545066356659, 0.13933029770851135, -0.3792094886302948, -0.22019106149673462, 0.4759371876716614, 0.0770471915602684, 0.27838754653930664, 0.2617088258266449, 0.12012124061584473, 0.19283759593963623, -0.1457984894514084, 0.13192351162433624, 0.39678218960762024, -0.016017746180295944, 0.06469859182834625, -0.475728303194046, -0.3340059518814087, 0.45108887553215027, 0.1048203557729721, 0.10609795898199081, 0.016658153384923935, -0.06624093651771545, 0.45338714122772217, 0.2774912118911743, -0.11796852946281433, -0.03857548534870148, 0.17891518771648407, -0.29750537872314453, -0.003019485389813781, -0.00899463426321745, 0.045939672738313675, -0.4310576319694519, -0.06189721077680588, -0.16155444085597992, 0.257390558719635, 0.021547041833400726, -0.33969467878341675, -0.12602579593658447, -0.38537514209747314, -0.012609955854713917, -0.03815142810344696, -0.02499976195394993, 0.32791668176651, 0.3891621530056, 0.020113619044423103, 0.04731030762195587, 0.5165185928344727, -0.07561000436544418, 0.19800028204917908, -0.7434918284416199, 0.2497139573097229, -0.09700039774179459, -0.22091664373874664, 0.4876715838909149, 0.29407206177711487, 0.1883365660905838, -0.20826828479766846, -0.12820002436637878, 0.4203101098537445, -0.09913225471973419, 0.41020575165748596, -0.3028308153152466, 0.009453179314732552, 0.28660091757774353, -0.215903639793396, -0.09509038180112839, 0.3127254545688629, -0.014362218789756298, -0.029680538922548294, -0.0852992981672287, 0.4560611844062805, -0.3552294969558716, 0.19494348764419556, 0.060180529952049255, -0.07444102317094803, 0.11264679580926895, -0.21182729303836823, -0.2256210893392563, -0.46470972895622253, 0.5550789833068848, -0.011282491497695446, -0.33899548649787903, 0.0956975445151329, -0.22145220637321472, 0.09830587357282639, 0.4710584878921509, 0.09217467159032822, -0.13324445486068726, -0.10313460230827332, 0.10757985711097717, -0.13031785190105438, 0.24420160055160522, 0.21329933404922485, 0.4979688227176666, 0.1266189068555832, -0.02547059953212738, 0.05887516215443611, -0.1459996998310089, -0.38097384572029114, 0.1322554051876068, -0.01731020212173462, 0.1433492749929428, 0.2880008816719055, 0.09366187453269958, -0.04037431627511978, -0.2727138102054596, -0.04973475635051727, 0.1679784208536148, 0.2461204081773758, -0.2921511232852936, -0.04091458395123482, -0.38650932908058167, -0.4747580885887146, -0.25838884711265564, -0.15843415260314941, -0.21669384837150574, -0.33608439564704895, 0.16131658852100372, 0.24778391420841217, -0.13261163234710693, 0.06924976408481598, -0.1683560013771057, 0.06610171496868134, -0.3122088313102722, -0.07321539521217346, -0.06265071779489517, -0.286241739988327, -0.243129163980484, -0.043864402920007706, 0.1981612741947174, 0.3729358911514282, 0.2036009579896927, -0.42539361119270325, -0.2250232696533203, 0.14127881824970245, -0.29145270586013794, 0.1450168341398239, -0.02652626857161522, -0.12445605546236038, 0.08017659932374954, -0.2173846960067749, 0.2646995782852173, -0.24174922704696655, 0.04491898790001869, -0.10440641641616821, -0.3310624361038208, -0.10031037777662277, 0.12558455765247345, 0.004529020749032497, -0.21038584411144257, -0.41956716775894165, -0.2590782344341278, -0.34470677375793457, -0.10632038116455078, 0.28229814767837524, 0.30478110909461975, 0.11865726113319397, -0.15986160933971405, -0.11309545487165451, -0.05772758275270462, 0.21963682770729065, -0.3405005633831024, -0.09820704162120819, 0.23921041190624237, -0.08729097992181778, -0.11389204859733582, 0.011299941688776016, -0.056971099227666855, -0.033428654074668884, 0.15223239362239838, -0.5655207633972168, -0.20712706446647644, -0.16636088490486145, -0.10165082663297653, -0.2521018981933594, -0.27892234921455383, 0.5203970074653625, -0.05680689960718155, 0.03545093536376953, 0.06587885320186615, -0.4510561525821686, 0.032256513833999634, 0.16768886148929596, 0.5530170202255249, -0.185030996799469, 0.20177896320819855, -0.05939693748950958, 0.48935624957084656, 0.5270172357559204, 0.0005128696793690324, -0.15433143079280853, 0.15244042873382568, 0.14646854996681213, 0.07625685632228851, -0.03768391162157059, 0.367347776889801, 0.0548575259745121, -0.222249835729599, 0.4481474459171295, -0.005777308251708746, -0.2257876694202423, -0.18706969916820526, -0.109590083360672, -0.26541054248809814, -0.28332987427711487, 0.13767164945602417, -0.1600022315979004, 0.13109610974788666, 0.2893313765525818, 0.32105323672294617, -0.07106932252645493, -0.5795772671699524, -0.07721086591482162, 0.3904191553592682, 0.1502414047718048, 0.12120397388935089, -0.39578595757484436, 0.06771351397037506, -0.5080146789550781, 0.34322699904441833, 0.24453237652778625, 0.2829876244068146, -0.15224570035934448, -0.07150101661682129, 0.10070925205945969, -0.2628660500049591, 0.29469504952430725, -0.4508109390735626, -0.10308386385440826, -0.05666991323232651, -0.03291456028819084, -0.3269297778606415, -0.19002746045589447, -0.16598482429981232, 0.4601229429244995, 0.47317543625831604, 0.46585312485694885, -0.37072116136550903, -0.08754438906908035, 0.20380419492721558, -0.12906329333782196, -0.1338876336812973, -0.17332790791988373, -0.32800185680389404, -0.4676932692527771, -0.16358686983585358, -0.14630037546157837, -0.1255975067615509, 0.33448326587677, -0.013892687857151031, -0.04778596758842468, 0.08378462493419647, 0.09194637089967728, 0.358094185590744, 0.3425993323326111, 0.07846520841121674, 0.11075626313686371, 0.1065606027841568, 0.2138594388961792, 0.5311780571937561, 0.12926580011844635, 0.5807297229766846, 0.1455204039812088, -0.06633463501930237, -0.019350402057170868, -0.2441665530204773, 0.14895132184028625, 0.4908941090106964, 0.13517753779888153, 0.053815800696611404, 0.03748385235667229, -0.08207717537879944, -0.2781616449356079, 0.07523457705974579, -0.007691981736570597, -0.054472241550683975, -0.4696715176105499, -0.3332740068435669, 0.27587801218032837, 0.2949083745479584, -0.14097736775875092, -0.1551346778869629, -0.24002864956855774, -0.3468586206436157, 0.2835208773612976, 0.2437189370393753, 1.0874789953231812, 0.2417481541633606, -0.022105829790234566, -0.001839283388108015, -0.15207724273204803, 0.9227555394172668, -0.276262104511261, 0.4379432499408722, -0.3907802104949951, -0.20731167495250702, -0.0421939417719841, -0.16487109661102295, 0.1221626028418541, 0.022072818130254745, -0.38290080428123474, 0.30941808223724365, 0.3011411726474762, 0.020167408511042595, -0.11544997245073318, 0.38711994886398315, -0.21533241868019104, -0.14622926712036133, -0.2855423390865326, -0.028606224805116653, -0.30701202154159546, 0.43148815631866455, -0.025085868313908577, -0.058468833565711975, -0.21924804151058197, -0.21921685338020325, -0.3474554717540741, 0.2333545982837677, 0.03732415661215782, -0.03378864377737045, 0.2643849849700928, -0.22531411051750183, -0.05423019453883171, 0.05196807160973549, 0.3515729308128357, -0.08515388518571854, -0.10732796043157578, -0.07103370875120163, -0.18470607697963715, -0.14279155433177948, 0.12849268317222595, 0.014371734112501144, 0.23452508449554443, -0.08470714092254639, 0.062451254576444626, 0.06788555532693863, -0.09753060340881348, -0.3518661856651306, -0.1145564541220665, -0.061132777482271194, -0.20195931196212769, -0.15386706590652466, -0.04333454370498657, 0.13102221488952637, -0.08529449254274368, -0.038762181997299194, 0.0029936397913843393, -0.14414578676223755, -0.22105295956134796, 0.27000752091407776, -0.06912102550268173, -0.38187283277511597, -0.15728257596492767, 0.3526187539100647, 0.44283995032310486, 0.03658849745988846, 0.36675068736076355, -0.15284758806228638, -0.0776594802737236, -0.15032599866390228, -0.15030460059642792, -0.3644886314868927, -0.2034793347120285, 0.23492339253425598, -0.23662905395030975, -0.12594136595726013, 0.1424080729484558, 0.4553367495536804, 0.1807447373867035, 0.33318695425987244, -0.061184294521808624, -0.5108866095542908, -0.008190044201910496, 0.17526523768901825, 0.03207867220044136, 0.058280639350414276, -0.05806531757116318, 0.5259189605712891, -0.25158461928367615, 0.21031570434570312, -0.19090032577514648, -0.14888896048069, -0.4438944160938263, 0.33429813385009766, 0.1320311576128006, -0.04377243295311928, -0.009572041220963001, 0.00005427761425380595, -0.16391295194625854, 0.313930869102478, -0.0652245432138443, -0.005955446511507034, -0.06146009638905525, 0.20650018751621246, 0.17853817343711853, -0.10877057909965515, -0.17463715374469757, 0.12480141222476959, 0.07677891105413437, 0.03751707822084427, 0.03463660925626755, 0.2473377138376236, 0.04405060037970543, 0.2818423807621002, -0.3356454372406006, 0.11022841930389404, -0.07830394059419632, 0.22765083611011505, -0.2985173463821411, 0.36041122674942017, -0.0591353103518486, 0.30647528171539307, -0.10010315477848053, 0.015594568103551865, 0.15056967735290527, -0.03606410324573517, 0.04848171025514603, 0.09732724726200104, 0.07677585631608963, -0.17980939149856567, 0.14732098579406738, 0.18416312336921692, -0.08916354179382324, 0.6363034248352051, -0.0754883736371994, -0.19107189774513245, 0.10431266576051712, 0.045415569096803665, -0.08597459644079208, 0.0015476382104679942, 0.2222013771533966, 0.025047797709703445, -0.05612998828291893, 0.5138848423957825, 0.08978325128555298, -0.13231146335601807, -0.13743680715560913, 0.1116626113653183, 0.3375568091869354, -0.0757073312997818, 0.10394518077373505, 0.19666793942451477, -0.20539207756519318, 0.09556011110544205, 0.30461305379867554, -0.28127261996269226, 0.15195663273334503, -0.00005180576772545464, 0.027242561802268028, 0.4766048192977905, 0.16119547188282013, 0.3706957697868347, -0.024317026138305664, -0.1133221760392189, 0.029959702864289284, 0.6413370966911316, 0.2822779715061188, 0.3716787099838257, 0.2842690348625183, 0.38737183809280396, -0.2766489088535309, 0.02878682129085064, -0.3678882420063019, -0.175491601228714, -0.06385727971792221, -0.14051537215709686, -0.46062037348747253, -0.04844372346997261, -0.17240595817565918, 0.20627927780151367, -0.14023523032665253, 0.1260407269001007, 0.4599737226963043, -0.07544803619384766, -0.20757713913917542, -0.37508898973464966, 0.12007753551006317, 0.028892289847135544, 0.166986882686615, -0.23685863614082336, 0.31650403141975403, -0.0766637846827507, -0.0866449773311615, 0.3203834295272827, 0.5079476833343506, 0.4745158553123474, 0.3062897324562073, -0.05767281726002693, -0.20901288092136383, 0.08152608573436737, 0.11034267395734787, 0.22965672612190247, 0.7636841535568237, 0.2975423336029053, -0.25712713599205017, 0.3881789445877075, -0.06476624310016632, -0.07674150913953781, 0.4682408273220062, 0.07057221233844757, -0.08752623945474625, -0.02216578833758831, 0.40950265526771545, -0.07317881286144257, 0.115787073969841, 0.021555010229349136, 0.1373034119606018, -0.25544700026512146, -0.2709899842739105, 0.06821183860301971, -0.2707931101322174, 0.305095911026001, 0.013167817145586014, 0.022203495725989342, -0.3044299781322479, 0.6118450164794922, 0.3897588551044464, -0.12752284109592438, -0.20482787489891052, -0.07230512797832489, -0.43062692880630493, 0.3466022312641144, -0.42263051867485046, -0.23992328345775604, -0.2996871769428253, 0.5030860900878906, -0.22747057676315308, 0.4045395255088806, -0.2877252995967865, 0.3795650601387024, -0.06517530232667923, -0.0316755585372448, -0.22983844578266144, -0.0672013983130455, -0.12271544337272644, -0.2966390550136566, 0.020049039274454117, -0.2617557942867279, 0.22259069979190826, 0.07699619978666306, -0.1308632642030716, 0.001514121307991445, -0.016435878351330757, 0.01601746492087841, 0.40780237317085266, -0.012908640317618847, 0.509090781211853, 0.3904208838939667, -0.14212600886821747, -0.15031278133392334, -0.005525454878807068, -0.3197632133960724, -0.311140239238739, 0.41492024064064026, 0.21395549178123474, 0.3597549498081207, -0.17412494122982025, 0.6665084958076477, -0.06747268885374069, 0.31923648715019226, 0.05682786926627159, 0.022631796076893806, -0.2512764036655426, -0.057016704231500626, -0.2459765374660492, 0.03582853823900223, 0.06754428893327713, 0.45341330766677856, 0.1536416858434677, -0.027758149430155754, -0.2425965815782547, -0.34912997484207153, 0.4797305762767792, -0.3962063193321228, 0.02929672971367836, -0.006678964011371136, 0.18370340764522552, 0.24500298500061035, -0.22634060680866241, -0.7466573715209961, -0.07005781680345535, 0.25677382946014404, -0.21241328120231628, 0.05072454363107681, -0.02079646661877632, -0.05016626417636871, -0.22436054050922394, -0.12880563735961914, -0.23391403257846832, 0.21499444544315338, -0.41724714636802673, 0.2863388955593109, -0.12938763201236725 ]
https://github.com/huggingface/datasets/issues/215
NonMatchingSplitsSizesError when loading blog_authorship_corpus
The files provided by the authors are corrupted and the script seems to ignore the xml files that can't be decoded (it does `try:... except UnicodeDecodeError`). Maybe depending of the environment some files can be opened and some others don't but not sure why
Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on.
44
NonMatchingSplitsSizesError when loading blog_authorship_corpus Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on. The files provided by the authors are corrupted and the script seems to ignore the xml files that can't be decoded (it does `try:... except UnicodeDecodeError`). Maybe depending of the environment some files can be opened and some others don't but not sure why
[ -0.21565401554107666, 0.12347541749477386, 0.02768758311867714, 0.48136061429977417, -0.011549796909093857, 0.09883328527212143, -0.13197045028209686, 0.368587464094162, -0.051696617156267166, 0.18515630066394806, 0.07529912889003754, 0.13688375055789948, 0.0854274183511734, -0.1344742625951767, -0.032237857580184937, 0.025299202650785446, -0.09251761436462402, 0.19295048713684082, 0.038546979427337646, -0.09255405515432358, -0.17508351802825928, 0.4164965748786926, -0.3340629041194916, 0.007659213617444038, -0.21158044040203094, -0.10277792066335678, 0.08266926556825638, 0.42487096786499023, 0.005414026323705912, -0.3878880441188812, 0.017096592113375664, -0.015492144040763378, 0.07717163115739822, 0.20357680320739746, -0.00012377271195873618, 0.0525185763835907, 0.6025202870368958, -0.26091188192367554, -0.1516072303056717, -0.24732530117034912, -0.3983463943004608, -0.34683603048324585, -0.15325473248958588, -0.30981194972991943, 0.04340599849820137, -0.13742025196552277, 0.2702699303627014, -0.21010403335094452, 0.30159056186676025, 0.3589801788330078, 0.10676798969507217, 0.17280273139476776, -0.06732246279716492, 0.13837400078773499, 0.27825382351875305, 0.06059829518198967, -0.0570705346763134, -0.07430363446474075, -0.030548973008990288, -0.2105676829814911, -0.09983694553375244, 0.315449595451355, -0.2957828938961029, -0.0047143627889454365, 0.17416712641716003, 0.09821538627147675, 0.37292414903640747, -0.2041337490081787, 0.17276902496814728, 0.29544174671173096, 0.5296695232391357, 0.13004034757614136, -0.21195365488529205, -0.4479377269744873, -0.15491227805614471, 0.11767438054084778, 0.406570702791214, 0.4266493022441864, -0.20510593056678772, -0.001026144833303988, -0.26339849829673767, -0.03466109558939934, -0.19299134612083435, 0.06627262383699417, 0.2431638240814209, 0.24165230989456177, 0.08146397769451141, 0.06608057022094727, -0.02012023515999317, 0.07188810408115387, 0.09668873250484467, -0.2602754831314087, -0.14188693463802338, 0.09542860090732574, 0.02983928471803665, -0.07551855593919754, -0.1188717633485794, -0.07541241496801376, 0.20608890056610107, 0.20481730997562408, 0.3663545250892639, 0.019850287586450577, 0.20985274016857147, 0.13503974676132202, 0.33779966831207275, -0.07288692891597748, 0.04885457083582878, 0.4268066883087158, 0.20002317428588867, -0.035382770001888275, 0.08910421282052994, 0.013300194405019283, 0.22526982426643372, -0.2736569046974182, -0.15773625671863556, 0.17779506742954254, 0.2460043579339981, -0.49686217308044434, -0.4554830491542816, 0.14754199981689453, -0.4238390028476715, -0.27335819602012634, -0.25034910440444946, 0.2649308145046234, -0.30463868379592896, 0.36598578095436096, -0.13079890608787537, 0.08853164315223694, -0.3846733570098877, -0.10326545685529709, -0.13025517761707306, 0.1645718216896057, -0.1479652374982834, 0.20909711718559265, 0.1741020530462265, -0.145019069314003, 0.3679027557373047, -0.17715829610824585, -0.06438561528921127, -0.2138000726699829, 0.3670298755168915, -0.17966417968273163, 0.04696173220872879, 0.2502399981021881, 0.10622433573007584, 0.17181435227394104, -0.11795493960380554, -0.29181572794914246, -0.28508177399635315, 0.0958578959107399, -0.2795935273170471, -0.3930393159389496, 0.11864642798900604, 0.14444133639335632, -0.3303496837615967, -0.04221566766500473, 0.07145185023546219, 0.04952345788478851, 0.41477522253990173, -0.4803289473056793, -0.09707175940275192, -0.22827795147895813, -0.15888851881027222, -0.12854719161987305, -0.2621927857398987, 0.6329423189163208, 0.1380792111158371, -0.10230616480112076, 0.28261852264404297, -0.029114263132214546, 0.5553827881813049, 0.35930100083351135, 0.06276608258485794, 0.12815912067890167, -0.19397999346256256, 0.4745137095451355, 0.11766566336154938, -0.41009238362312317, -0.15080328285694122, 0.469144731760025, -0.04133430868387222, 0.33927011489868164, 0.4302491545677185, 0.1024659276008606, 0.2844763398170471, -0.17926837503910065, -0.04245772585272789, 0.3146084249019623, 0.06322971731424332, -0.028443235903978348, -0.4521339535713196, -0.14667294919490814, 0.43677598237991333, 0.10489198565483093, 0.18474280834197998, 0.010159454308450222, 0.1415490061044693, 0.43899327516555786, 0.31315895915031433, -0.10660918802022934, 0.1054217591881752, 0.29825156927108765, -0.44768062233924866, 0.16950981318950653, 0.10063289850950241, -0.10743069648742676, -0.34790775179862976, -0.07296061515808105, -0.3559921085834503, 0.14380663633346558, 0.0009278090437874198, -0.22381563484668732, -0.17331509292125702, -0.39921310544013977, -0.07715677469968796, -0.0034570193383842707, -0.022262156009674072, 0.3663993775844574, 0.2873927056789398, 0.15451759099960327, -0.05905602499842644, 0.36677899956703186, -0.03369626775383949, 0.18653330206871033, -0.8968729376792908, 0.23864960670471191, -0.15186764299869537, -0.22171442210674286, 0.4185648560523987, 0.31582868099212646, 0.1912761777639389, -0.19332437217235565, -0.07729288190603256, 0.38104313611984253, -0.1058674305677414, 0.47593438625335693, -0.2341877520084381, 0.07466630637645721, 0.18912188708782196, -0.3120644986629486, -0.020261727273464203, 0.40568801760673523, 0.09887275099754333, -0.11944020539522171, -0.037329453974962234, 0.42273691296577454, -0.3420756161212921, 0.19422751665115356, 0.14112389087677002, -0.06570698320865631, 0.06375353783369064, -0.2677146792411804, -0.24543507397174835, -0.5215456485748291, 0.622148871421814, -0.06730565428733826, -0.2690836191177368, 0.04679831489920616, -0.25621870160102844, 0.025932686403393745, 0.572960615158081, 0.0020748134702444077, -0.0830087661743164, -0.06289917975664139, 0.06719481199979782, -0.0630134642124176, -0.018116872757673264, 0.18459869921207428, 0.45713457465171814, 0.0909477099776268, -0.05557650700211525, 0.06711249053478241, -0.13182002305984497, -0.45842888951301575, 0.12814301252365112, -0.04832801967859268, -0.027972286567091942, 0.5724945068359375, -0.011887671425938606, 0.06637749075889587, -0.34778547286987305, -0.02124849148094654, 0.1145632416009903, 0.13852639496326447, -0.332579106092453, -0.0929337814450264, -0.3299711048603058, -0.40572047233581543, -0.3795692026615143, -0.07616912573575974, -0.1871168315410614, -0.3854968547821045, 0.18350407481193542, 0.21754232048988342, -0.25006765127182007, 0.028112785890698433, -0.19778597354888916, 0.0052078599110245705, -0.3419666290283203, 0.048854511231184006, -0.13127964735031128, -0.1937546581029892, -0.21058011054992676, -0.06514111906290054, 0.2704777121543884, 0.3898126184940338, 0.2202955186367035, -0.5454664826393127, -0.21008841693401337, -0.04476914182305336, -0.20753507316112518, 0.020225567743182182, -0.05990850552916527, -0.17760859429836273, 0.0929604023694992, -0.15598039329051971, 0.5214745402336121, -0.16104117035865784, 0.14776381850242615, -0.1287199705839157, -0.2708491086959839, 0.12490400671958923, 0.2216656357049942, -0.14447808265686035, -0.22062905132770538, -0.5865996479988098, -0.17907948791980743, -0.45689499378204346, -0.09897736459970474, 0.28189438581466675, 0.29488658905029297, 0.042354561388492584, -0.16215108335018158, -0.10634370148181915, -0.21439386904239655, 0.35616421699523926, -0.16313163936138153, 0.0558563768863678, 0.32341763377189636, -0.1839475929737091, -0.27022507786750793, -0.02170742303133011, 0.13760590553283691, -0.05147795006632805, 0.14158597588539124, -0.4621357023715973, -0.05968309938907623, -0.18343868851661682, -0.15344801545143127, -0.2829676568508148, -0.11173224449157715, 0.4699307680130005, -0.034130048006772995, -0.004535318352282047, -0.05661609396338463, -0.2683965265750885, 0.01701795868575573, 0.05479617044329643, 0.30072423815727234, -0.1281159222126007, 0.233890101313591, -0.09957699477672577, 0.416090190410614, 0.5284242033958435, 0.18462400138378143, -0.06775511056184769, 0.07574251294136047, 0.27277079224586487, 0.1663350760936737, -0.01663486286997795, 0.33610016107559204, -0.02377261221408844, -0.09101960808038712, 0.43315309286117554, -0.002345995046198368, -0.09685918688774109, -0.14553633332252502, -0.2459462583065033, -0.34901490807533264, -0.19343797862529755, 0.14629146456718445, -0.1666954904794693, 0.14870765805244446, 0.220889613032341, 0.3066443204879761, -0.038720209151506424, -0.5155572295188904, 0.02454150840640068, 0.4532243609428406, 0.3203480541706085, 0.1850072145462036, -0.47967514395713806, 0.09125714004039764, -0.4247419536113739, 0.29793354868888855, 0.14495442807674408, 0.31214243173599243, -0.18283236026763916, -0.1499032825231552, 0.10623250156641006, -0.25153082609176636, 0.23255686461925507, -0.41618192195892334, 0.019843285903334618, 0.059046920388936996, -0.00655188737437129, -0.3043077290058136, -0.1795511543750763, -0.26142290234565735, 0.4053453505039215, 0.38657477498054504, 0.43123385310173035, -0.38310506939888, -0.16495122015476227, 0.09735114127397537, 0.0051356954500079155, -0.10864117741584778, -0.0825342983007431, -0.2994450032711029, -0.36759141087532043, -0.2296072095632553, -0.29456907510757446, -0.08849712461233139, 0.2692148685455322, -0.09233559668064117, -0.02708537131547928, 0.03468611463904381, 0.10009180009365082, 0.36093422770500183, 0.27080339193344116, 0.08658657222986221, -0.10598348826169968, 0.07422270625829697, 0.1566508561372757, 0.573414146900177, 0.11814223229885101, 0.5895775556564331, 0.0689941942691803, -0.18111085891723633, 0.024747679010033607, -0.43041250109672546, 0.19162815809249878, 0.4658266305923462, 0.08184269815683365, 0.04430010914802551, 0.12616148591041565, -0.02420041151344776, -0.44463998079299927, -0.05971977114677429, 0.16697092354297638, -0.08109491318464279, -0.40408608317375183, -0.42042413353919983, 0.2781202793121338, 0.20545455813407898, -0.13481971621513367, -0.16485299170017242, -0.13094355165958405, -0.2597300112247467, 0.3258017897605896, 0.1342848390340805, 1.0393661260604858, 0.2653285264968872, 0.07107457518577576, -0.06675048172473907, -0.1936637908220291, 0.7017491459846497, -0.23345676064491272, 0.46105659008026123, -0.45272964239120483, -0.18189312517642975, -0.09579308331012726, -0.18577440083026886, 0.28010880947113037, 0.13668407499790192, -0.21743763983249664, 0.33872756361961365, 0.20434294641017914, -0.02720567025244236, -0.1010570377111435, 0.47604668140411377, -0.2647894024848938, -0.03463757783174515, -0.10286534577608109, -0.0529908612370491, -0.1834336519241333, 0.45682984590530396, -0.0012245166581124067, -0.1371757984161377, -0.16188232600688934, -0.2715011239051819, -0.354420006275177, 0.2869674265384674, 0.11578213423490524, -0.07829244434833527, 0.18359088897705078, -0.19995839893817902, 0.05459534749388695, 0.20606929063796997, 0.22772039473056793, -0.08420244604349136, -0.038651060312986374, -0.09670519828796387, -0.18037466704845428, -0.19949793815612793, 0.18710416555404663, 0.125250443816185, 0.34723344445228577, -0.11429061740636826, 0.06230555474758148, 0.18676631152629852, -0.15038475394248962, -0.27849075198173523, -0.3265245854854584, 0.020360851660370827, -0.22256635129451752, -0.18651843070983887, -0.1451810598373413, 0.01597239263355732, -0.1669837385416031, -0.04724081978201866, -0.0031716518569737673, -0.08467763662338257, -0.09633992612361908, 0.19067256152629852, -0.1776459515094757, -0.45024290680885315, -0.12551544606685638, 0.27839425206184387, 0.36409687995910645, -0.03017013520002365, 0.4310382902622223, -0.09358744323253632, -0.1244845911860466, -0.19571980834007263, -0.15213148295879364, -0.2345649003982544, -0.4015832245349884, 0.17225492000579834, -0.20614074170589447, -0.10457778722047806, 0.17576409876346588, 0.2439630925655365, 0.08783239871263504, 0.1960270255804062, -0.12787173688411713, -0.6653311848640442, 0.12647445499897003, 0.1522250920534134, 0.21884332597255707, -0.019053328782320023, -0.04890180751681328, 0.3229840397834778, -0.23107916116714478, 0.16286714375019073, -0.18060222268104553, -0.07511352747678757, -0.3088241517543793, 0.3002742528915405, 0.16600359976291656, 0.04076988995075226, -0.1094680204987526, 0.02919076196849346, -0.13556674122810364, 0.2487291544675827, -0.17746049165725708, 0.03926829993724823, -0.14118988811969757, 0.22275473177433014, 0.15719321370124817, -0.05965344235301018, -0.10897043347358704, 0.019748637452721596, 0.09330209344625473, -0.021798312664031982, 0.03392118215560913, 0.2995970547199249, -0.038626864552497864, 0.34890016913414, -0.3989136219024658, 0.1466968059539795, -0.09037531167268753, 0.2649857699871063, -0.34620028734207153, 0.38974815607070923, -0.20149488747119904, 0.24060320854187012, -0.03934533894062042, 0.022920260205864906, -0.09879720956087112, -0.03177723288536072, 0.12119623273611069, 0.06413549929857254, 0.315273642539978, -0.15873907506465912, 0.21573176980018616, 0.2088230401277542, -0.03733917698264122, 0.7320030331611633, -0.12561577558517456, -0.11623415350914001, 0.062017638236284256, 0.05478929355740547, -0.11067041754722595, -0.017983410507440567, 0.18334680795669556, -0.08880655467510223, -0.08237118273973465, 0.3635847568511963, 0.194892019033432, -0.2530847489833832, -0.18858075141906738, 0.10765539109706879, 0.4120692312717438, -0.0813135877251625, 0.09422293305397034, 0.23868407309055328, -0.011601987294852734, 0.01793307065963745, 0.19760923087596893, -0.1533752977848053, 0.07994996756315231, 0.11493460088968277, 0.09460373222827911, 0.5233206748962402, 0.14145047962665558, 0.43510258197784424, -0.23179736733436584, -0.12785881757736206, 0.018450118601322174, 0.6700659990310669, 0.19276104867458344, 0.48585206270217896, 0.2512132525444031, 0.3217867910861969, -0.29630643129348755, 0.1505877822637558, -0.2669786214828491, -0.17418330907821655, -0.03776175528764725, -0.12803567945957184, -0.46970853209495544, -0.12692983448505402, -0.20043180882930756, 0.06337621062994003, -0.07095912098884583, 0.2730638086795807, 0.31693270802497864, -0.055920131504535675, -0.19864106178283691, -0.29472535848617554, 0.1727561056613922, -0.15478503704071045, 0.18281325697898865, -0.32272830605506897, 0.27617892622947693, 0.10261502861976624, -0.03287234902381897, 0.3759273886680603, 0.5440004467964172, 0.6268024444580078, 0.4050436317920685, -0.10989496111869812, -0.12854284048080444, -0.054338112473487854, 0.047856852412223816, 0.1882709115743637, 0.8659355044364929, 0.21049346029758453, -0.1732780635356903, 0.3700051009654999, -0.03926475718617439, -0.08896417915821075, 0.3375840485095978, 0.1680564433336258, -0.04063244163990021, 0.04704640805721283, 0.38823020458221436, -0.19839505851268768, -0.0021035363897681236, -0.02241436205804348, 0.03416895121335983, -0.41420626640319824, -0.26651227474212646, 0.17878316342830658, -0.14471091330051422, 0.28021082282066345, -0.1523040235042572, 0.019083654507994652, -0.1773102879524231, 0.5529254674911499, 0.47746336460113525, -0.17004342377185822, -0.09665429592132568, -0.022315125912427902, -0.5047853589057922, 0.5072601437568665, -0.4191427528858185, -0.31876516342163086, -0.24322648346424103, 0.37428924441337585, -0.1979198306798935, 0.26287856698036194, -0.09300903975963593, 0.44354283809661865, -0.13743484020233154, 0.10049103200435638, -0.2129071205854416, -0.019861534237861633, 0.014264900237321854, -0.15694878995418549, -0.09848254919052124, -0.2454882711172104, 0.2524292767047882, 0.04092567041516304, -0.07209348678588867, -0.010972030460834503, 0.1572076827287674, -0.0024544803891330957, 0.27439460158348083, -0.026172082871198654, 0.47785013914108276, 0.4452592730522156, -0.2909838557243347, -0.213938370347023, -0.11031963676214218, -0.2303617149591446, -0.28157666325569153, 0.40110015869140625, 0.22582018375396729, 0.3603357970714569, -0.20492379367351532, 0.5031099319458008, -0.09256667643785477, 0.2626861035823822, -0.15710794925689697, 0.1450565904378891, -0.21884819865226746, 0.03149935230612755, -0.23324978351593018, -0.01633157581090927, 0.1572694331407547, 0.45763328671455383, 0.18879956007003784, -0.1253947615623474, -0.23119796812534332, -0.3733322024345398, 0.4127681851387024, -0.3910878300666809, -0.07988296449184418, -0.017233625054359436, 0.193974569439888, 0.1286228597164154, -0.041532523930072784, -0.6443731784820557, -0.059588201344013214, 0.22831957042217255, -0.11385001242160797, 0.029822824522852898, -0.0664222463965416, -0.1520504504442215, -0.35204046964645386, -0.03435492515563965, -0.25079888105392456, 0.2707359790802002, -0.41925013065338135, 0.40055155754089355, -0.10350919514894485 ]
https://github.com/huggingface/datasets/issues/215
NonMatchingSplitsSizesError when loading blog_authorship_corpus
Feel free to do `ignore_verifications=True` for now... The verifications only include a check on the checksums of the downloaded files, and a check on the number of examples in each splits.
Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on.
31
NonMatchingSplitsSizesError when loading blog_authorship_corpus Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on. Feel free to do `ignore_verifications=True` for now... The verifications only include a check on the checksums of the downloaded files, and a check on the number of examples in each splits.
[ -0.15917642414569855, 0.061383508145809174, 0.06137014180421829, 0.398507297039032, -0.09111880511045456, 0.06813999265432358, -0.15415650606155396, 0.3743869960308075, -0.05398913845419884, 0.12912297248840332, 0.09491048753261566, 0.1696774959564209, 0.05089116469025612, -0.0702366828918457, -0.029941456392407417, -0.031388770788908005, -0.046267908066511154, 0.20410314202308655, 0.043887313455343246, 0.0010298172710463405, -0.019637158140540123, 0.4404020607471466, -0.30175015330314636, -0.022966835647821426, -0.24699321389198303, -0.1683318167924881, -0.0421302430331707, 0.32819804549217224, 0.0039549521170556545, -0.3751758933067322, 0.24395844340324402, -0.07514743506908417, 0.13529181480407715, 0.1908106803894043, -0.00012187100219307467, -0.013405069708824158, 0.5785958766937256, -0.2156343162059784, -0.20625559985637665, -0.25492221117019653, -0.40470582246780396, -0.3622804284095764, -0.08360158652067184, -0.23551887273788452, -0.02168651670217514, -0.05503878369927406, 0.1720564216375351, -0.08597991615533829, 0.27252933382987976, 0.2854222059249878, 0.11910389363765717, 0.15230579674243927, -0.11947901546955109, 0.031291037797927856, 0.13146859407424927, 0.11148032546043396, -0.03822566196322441, -0.09602155536413193, -0.08761424571275711, -0.286239892244339, -0.1037815734744072, 0.4399673044681549, -0.19715183973312378, 0.1551973521709442, 0.22587355971336365, 0.11542968451976776, 0.4008733928203583, -0.11456120759248734, 0.03758152201771736, 0.39043542742729187, 0.509495198726654, 0.16033294796943665, -0.2086896300315857, -0.5066434144973755, -0.05045630410313606, 0.19382117688655853, 0.4156515598297119, 0.36488455533981323, -0.23910050094127655, 0.008471859619021416, -0.41110968589782715, -0.07393115758895874, -0.09050969034433365, 0.10174065828323364, 0.266836017370224, 0.198568195104599, 0.11837147921323776, 0.0067394874058663845, 0.12967172265052795, -0.04109881445765495, 0.11463280767202377, -0.2442874163389206, -0.1399921178817749, 0.11282744258642197, -0.18387793004512787, -0.19307401776313782, -0.1448514759540558, -0.01740679144859314, 0.2120877057313919, 0.21555694937705994, 0.35182899236679077, -0.05405338853597641, 0.2256540209054947, 0.11196880787611008, 0.42831718921661377, 0.047906987369060516, -0.01073665451258421, 0.5196455717086792, 0.08321744203567505, 0.0033719190396368504, 0.1151566430926323, 0.0596730038523674, 0.2659810185432434, -0.21909482777118683, -0.145186185836792, 0.11991623789072037, 0.15946318209171295, -0.5667433738708496, -0.43986237049102783, 0.10815704613924026, -0.561587929725647, -0.2434341162443161, -0.09248384833335876, 0.17566236853599548, -0.3614128530025482, 0.33200445771217346, -0.15387791395187378, 0.04779895022511482, -0.36163780093193054, -0.059446096420288086, -0.20667234063148499, 0.17067883908748627, -0.197499617934227, 0.2053249329328537, 0.2014610469341278, -0.1554442048072815, 0.40264183282852173, -0.10782936215400696, -0.05867201089859009, -0.19689762592315674, 0.40113475918769836, -0.10814101248979568, 0.020010706037282944, 0.15724466741085052, -0.029180776327848434, 0.16998538374900818, -0.0008437237702310085, -0.21693286299705505, -0.3886170983314514, 0.07954514026641846, -0.22302435338497162, -0.5368485450744629, 0.10468985885381699, 0.1530231088399887, -0.4432888925075531, -0.061291828751564026, 0.09627249091863632, 0.2109624445438385, 0.36438432335853577, -0.3324517011642456, -0.09742403030395508, -0.2202012687921524, -0.12947231531143188, -0.05456845834851265, -0.2055286467075348, 0.495089590549469, 0.14020076394081116, -0.26605287194252014, 0.13045290112495422, -0.018281934782862663, 0.5335456132888794, 0.34930896759033203, 0.03433426097035408, 0.09644033014774323, -0.19258925318717957, 0.48758551478385925, 0.13571053743362427, -0.35429733991622925, -0.16301192343235016, 0.5454868078231812, 0.028682762756943703, 0.280705064535141, 0.25120460987091064, 0.11916979402303696, 0.194296732544899, -0.09376254677772522, 0.010771316476166248, 0.40808701515197754, -0.02160259708762169, 0.03519224002957344, -0.45149216055870056, -0.30968841910362244, 0.41489920020103455, 0.20669551193714142, 0.11799626052379608, -0.04226292297244072, -0.0019411847461014986, 0.4965001940727234, 0.3046287000179291, -0.13594861328601837, 0.05396599322557449, 0.2919708490371704, -0.35402747988700867, 0.1288684606552124, 0.038535282015800476, -0.11504468321800232, -0.3697335422039032, -0.02302205190062523, -0.2869815230369568, 0.2818145453929901, 0.10698974132537842, -0.258402556180954, -0.1996176838874817, -0.41557472944259644, -0.025225773453712463, -0.029962636530399323, 0.010713798925280571, 0.2815219759941101, 0.3974573612213135, 0.05119401589035988, 0.06851079314947128, 0.4284549951553345, -0.06218862161040306, 0.1739072948694229, -0.7967955470085144, 0.23201744258403778, -0.11876425892114639, -0.2111770510673523, 0.5127648711204529, 0.34133437275886536, 0.22369955480098724, -0.16931086778640747, -0.09573185443878174, 0.43656089901924133, -0.10373696684837341, 0.4019586145877838, -0.17406775057315826, 0.010552458465099335, 0.20788675546646118, -0.2954590320587158, 0.018218237906694412, 0.37473738193511963, -0.028403956443071365, -0.0686894878745079, -0.1335941106081009, 0.41676682233810425, -0.33828768134117126, 0.16545294225215912, 0.06876242160797119, 0.01500984188169241, 0.07816498726606369, -0.23879607021808624, -0.19139553606510162, -0.4554651081562042, 0.6415627598762512, -0.09513697773218155, -0.32226404547691345, 0.04229278862476349, -0.20106367766857147, 0.03303343057632446, 0.4396529197692871, 0.021865177899599075, -0.09082602709531784, -0.026410093531012535, 0.02838948927819729, -0.1608663946390152, 0.09572191536426544, 0.27505430579185486, 0.5138393044471741, 0.12376869469881058, 0.02331429347395897, 0.015096129849553108, -0.1677720993757248, -0.3314586579799652, 0.12296800315380096, -0.06293948739767075, 0.027277177199721336, 0.367068350315094, -0.0033676510211080313, -0.03191020339727402, -0.31080392003059387, -0.012369411997497082, 0.15956208109855652, 0.24881504476070404, -0.32820889353752136, -0.13419097661972046, -0.3583880662918091, -0.4044026732444763, -0.34532350301742554, -0.13506588339805603, -0.2826969623565674, -0.3694407641887665, 0.2020638883113861, 0.19996485114097595, -0.16825969517230988, 0.13886895775794983, -0.16680948436260223, 0.04333563148975372, -0.315635621547699, -0.052826713770627975, -0.01851593516767025, -0.2104286253452301, -0.27342844009399414, -0.026242166757583618, 0.22404174506664276, 0.42671751976013184, 0.24252179265022278, -0.3922346532344818, -0.15776924788951874, 0.046740736812353134, -0.2580871880054474, 0.06250361353158951, -0.05618521198630333, -0.1513657569885254, 0.176835834980011, -0.1386253833770752, 0.3756524324417114, -0.2218140959739685, 0.10164012014865875, -0.18995614349842072, -0.3374042212963104, 0.027316864579916, 0.17911215126514435, 0.035421811044216156, -0.19235478341579437, -0.5341881513595581, -0.2685713768005371, -0.36351513862609863, -0.062366124242544174, 0.28557708859443665, 0.24182595312595367, 0.03172792121767998, -0.12044011056423187, -0.1327032595872879, -0.17049168050289154, 0.2503781318664551, -0.30170774459838867, -0.05888936296105385, 0.24042053520679474, -0.09786761552095413, -0.16751530766487122, -0.03147110342979431, 0.0028390015941113234, -0.050665728747844696, 0.1228749006986618, -0.5468326210975647, -0.10958725214004517, -0.20560301840305328, -0.07945823669433594, -0.32694217562675476, -0.2575344741344452, 0.5263427495956421, -0.07347576320171356, -0.024189162999391556, 0.013537401333451271, -0.3449534475803375, 0.06836891919374466, 0.10190653055906296, 0.5138316750526428, -0.16110685467720032, 0.18403857946395874, -0.01432077493518591, 0.5651076436042786, 0.5774586796760559, 0.07665687799453735, -0.11549961566925049, 0.17679162323474884, 0.09244593977928162, 0.17404836416244507, 0.009828254580497742, 0.37702643871307373, 0.022528626024723053, -0.10745576024055481, 0.41577863693237305, 0.08360524475574493, -0.19980119168758392, -0.11438917368650436, -0.1578511744737625, -0.3104238510131836, -0.18710289895534515, 0.16156257688999176, -0.2299748957157135, 0.2006320059299469, 0.20180648565292358, 0.23883360624313354, -0.18510375916957855, -0.529470682144165, -0.012654370628297329, 0.35910728573799133, 0.22725710272789001, 0.17685940861701965, -0.47010311484336853, 0.1427350640296936, -0.4302377998828888, 0.2958557903766632, 0.21121430397033691, 0.2849103808403015, -0.21085771918296814, -0.15540455281734467, 0.08103758841753006, -0.19880680739879608, 0.2955510914325714, -0.4056420922279358, -0.015020293183624744, -0.04621114209294319, -0.03189621493220329, -0.34224215149879456, -0.20172236859798431, -0.13815321028232574, 0.37310779094696045, 0.5086649060249329, 0.42138606309890747, -0.427228718996048, -0.12459439039230347, 0.1632818728685379, -0.05940469726920128, -0.1309102475643158, -0.16943684220314026, -0.2525317966938019, -0.33338987827301025, -0.17163018882274628, -0.14729276299476624, -0.1254294365644455, 0.3001166582107544, -0.021955857053399086, -0.0020838628988713026, 0.14473280310630798, 0.1252201348543167, 0.3355758488178253, 0.25718241930007935, 0.1685250848531723, -0.03306202590465546, 0.1076679453253746, 0.2788695991039276, 0.5404441356658936, 0.08358272910118103, 0.5131449699401855, 0.10595840215682983, -0.05891048163175583, -0.025091810151934624, -0.30476656556129456, 0.07861349731683731, 0.4301791489124298, 0.17479099333286285, 0.08524948358535767, 0.1424744576215744, -0.02743460424244404, -0.4631119668483734, -0.02574414201080799, 0.11071287095546722, -0.07656677067279816, -0.3716815412044525, -0.39093077182769775, 0.2588399648666382, 0.2693788409233093, -0.14291086792945862, -0.1460685282945633, -0.19557860493659973, -0.25535696744918823, 0.23188114166259766, 0.1594075858592987, 0.9950314164161682, 0.2600896656513214, -0.026020972058176994, -0.027608996257185936, -0.22299738228321075, 0.8679139614105225, -0.2555353343486786, 0.4108973741531372, -0.4282154440879822, -0.25126856565475464, -0.062348391860723495, -0.20891457796096802, 0.15235067903995514, 0.1181776151061058, -0.2996962070465088, 0.33157140016555786, 0.26875633001327515, -0.011666422709822655, -0.10332194715738297, 0.44522324204444885, -0.24535927176475525, -0.07309851050376892, -0.17734752595424652, 0.014801540412008762, -0.2884224057197571, 0.500841498374939, -0.017722036689519882, -0.12158556282520294, -0.21846461296081543, -0.21833407878875732, -0.41547688841819763, 0.2596205770969391, 0.18067559599876404, 0.0051093995571136475, 0.2780002951622009, -0.18391957879066467, -0.0008590386132709682, 0.09157249331474304, 0.3283587694168091, -0.18273353576660156, -0.17481935024261475, 0.00697441678494215, -0.2024378478527069, -0.13968564569950104, 0.19291868805885315, 0.07448702305555344, 0.3397170603275299, -0.12549328804016113, 0.001391140278428793, 0.05643019825220108, -0.12228953838348389, -0.33007746934890747, -0.30359965562820435, -0.11330465972423553, -0.20215852558612823, -0.18769250810146332, -0.08816827088594437, 0.07827674597501755, -0.10347343236207962, -0.06149626523256302, 0.040477387607097626, -0.040233366191387177, -0.21182577311992645, 0.2360779047012329, -0.09049473702907562, -0.3304221034049988, -0.11680646240711212, 0.39779970049858093, 0.36702463030815125, -0.03411328047513962, 0.44290393590927124, -0.10410823673009872, -0.0931103378534317, -0.2093406617641449, -0.12918995320796967, -0.31093543767929077, -0.3431040048599243, 0.1643279790878296, -0.26846233010292053, -0.08667048811912537, 0.19470739364624023, 0.2930179834365845, 0.2384098321199417, 0.3419272005558014, -0.0016043860232457519, -0.5291133522987366, 0.016075296327471733, 0.1371033489704132, -0.024978162720799446, 0.07527948170900345, -0.14363142848014832, 0.42183154821395874, -0.20619715750217438, 0.24041719734668732, -0.21923060715198517, -0.08254670351743698, -0.40393540263175964, 0.3224554657936096, 0.11327142268419266, -0.10176239162683487, -0.0388985238969326, 0.01586868241429329, -0.11925472319126129, 0.21726550161838531, -0.1423013061285019, 0.0013302526203915477, -0.11548487097024918, 0.20137113332748413, 0.14739491045475006, -0.08197441697120667, -0.10277985036373138, 0.11400677263736725, 0.07142945379018784, 0.029016761109232903, -0.05398613214492798, 0.20699386298656464, -0.022113915532827377, 0.32321402430534363, -0.3484605848789215, 0.04232333227992058, -0.06938454508781433, 0.23659799993038177, -0.3490242660045624, 0.293660044670105, -0.09470091760158539, 0.219862163066864, -0.08393599092960358, -0.00015817016537766904, -0.03363383188843727, 0.01333015225827694, 0.08372693508863449, 0.07243602722883224, 0.1884957104921341, -0.17304107546806335, 0.10398256033658981, 0.22139397263526917, 0.021714292466640472, 0.5436682105064392, -0.0795350968837738, -0.10975386947393417, 0.07915380597114563, 0.10070692002773285, -0.12908579409122467, -0.03160998970270157, 0.25471052527427673, -0.03046266734600067, -0.12914413213729858, 0.35050299763679504, 0.1402948945760727, -0.16056570410728455, -0.20050998032093048, 0.10796469449996948, 0.39615774154663086, -0.13299110531806946, 0.12379790097475052, 0.2921995222568512, -0.16179782152175903, -0.01571018621325493, 0.23938395082950592, -0.2638281285762787, 0.13483279943466187, 0.010987630113959312, 0.04418566823005676, 0.46133697032928467, 0.09753625094890594, 0.3920549154281616, -0.14451418817043304, -0.1371356099843979, -0.05386039614677429, 0.7125882506370544, 0.3032742440700531, 0.40754806995391846, 0.23052401840686798, 0.47541218996047974, -0.3398907780647278, -0.0037843806203454733, -0.3541964888572693, -0.18464308977127075, -0.03735315427184105, -0.10190173238515854, -0.5155580639839172, -0.10836496949195862, -0.23224826157093048, 0.12218505144119263, -0.08818651735782623, 0.1432598978281021, 0.40934309363365173, -0.07159369438886642, -0.18918538093566895, -0.47873279452323914, 0.05307810381054878, 0.014808965846896172, 0.1754540503025055, -0.32879117131233215, 0.15796248614788055, 0.03684285283088684, -0.16267044842243195, 0.39498817920684814, 0.5192055106163025, 0.5169352889060974, 0.3021613657474518, -0.14195597171783447, -0.21230392158031464, 0.05457695573568344, 0.15161584317684174, 0.23794211447238922, 0.7681576013565063, 0.27525079250335693, -0.2857534885406494, 0.39371567964553833, -0.051301226019859314, -0.08288413286209106, 0.4031171500682831, 0.05604388937354088, -0.06899291276931763, -0.026875684037804604, 0.4426218271255493, -0.13705334067344666, 0.11488308012485504, -0.00470346212387085, 0.1161603108048439, -0.2987504005432129, -0.3102472722530365, 0.1035744771361351, -0.17181965708732605, 0.28894665837287903, -0.07533513009548187, 0.03425164520740509, -0.31278860569000244, 0.5959124565124512, 0.43539947271347046, -0.21247105300426483, -0.21041014790534973, -0.11609246581792831, -0.49980127811431885, 0.3454269468784332, -0.4612213373184204, -0.23284493386745453, -0.1896781325340271, 0.4772252142429352, -0.2099054753780365, 0.29365774989128113, -0.17123912274837494, 0.4648703634738922, -0.1285225749015808, 0.07755865156650543, -0.2698145806789398, -0.05188245326280594, -0.08174449950456619, -0.19109459221363068, 0.02198435179889202, -0.22614046931266785, 0.19919051229953766, 0.10465013980865479, -0.0841168537735939, 0.03213845193386078, 0.05748956650495529, 0.026997288689017296, 0.4276266396045685, -0.08215106278657913, 0.5023900270462036, 0.4290030002593994, -0.1451624631881714, -0.21732625365257263, -0.08768947422504425, -0.2888009548187256, -0.2974666357040405, 0.37864765524864197, 0.1362176090478897, 0.42173048853874207, -0.17510412633419037, 0.6471746563911438, -0.1287108063697815, 0.32008203864097595, 0.019591232761740685, 0.011847538873553276, -0.3271252512931824, -0.0376971997320652, -0.21452462673187256, 0.02605794370174408, 0.1519400179386139, 0.4710058569908142, 0.17633019387722015, -0.01744754984974861, -0.18155914545059204, -0.3055362105369568, 0.4234530031681061, -0.319659948348999, 0.034292809665203094, -0.02669534645974636, 0.18106958270072937, 0.20320938527584076, -0.11101235449314117, -0.6931036710739136, -0.07680728286504745, 0.29315418004989624, -0.11016219854354858, 0.08574385195970535, -0.035791534930467606, -0.07137186080217361, -0.2900961637496948, -0.022308126091957092, -0.22079923748970032, 0.2341662496328354, -0.3640001118183136, 0.3844623267650604, -0.08561214059591293 ]
https://github.com/huggingface/datasets/issues/215
NonMatchingSplitsSizesError when loading blog_authorship_corpus
I'm getting this same issue when loading the `imdb` corpus via `dataset = load_dataset("imdb")`. When I try `ignore_verifications=True`, no examples are read into the `train` portion of the dataset.
Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on.
29
NonMatchingSplitsSizesError when loading blog_authorship_corpus Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on. I'm getting this same issue when loading the `imdb` corpus via `dataset = load_dataset("imdb")`. When I try `ignore_verifications=True`, no examples are read into the `train` portion of the dataset.
[ -0.1060907244682312, 0.17192859947681427, 0.0861470177769661, 0.37152794003486633, -0.07061786204576492, 0.1975201964378357, -0.08851514011621475, 0.35896268486976624, -0.07059378921985626, 0.13236771523952484, 0.010963981971144676, 0.19261105358600616, 0.019732648506760597, -0.13543331623077393, -0.06135211139917374, 0.0010896628955379128, -0.04968471825122833, 0.151227667927742, 0.08074229955673218, -0.0007665433222427964, -0.019083088263869286, 0.43668171763420105, -0.3507993817329407, -0.06492386758327484, -0.27976810932159424, -0.17144839465618134, 0.00506010465323925, 0.30681219696998596, 0.06778958439826965, -0.3383012115955353, 0.2467605620622635, -0.08211200684309006, 0.1046876311302185, 0.25874486565589905, -0.00012322825205046684, 0.03466244041919708, 0.6713195443153381, -0.23092064261436462, -0.19163449108600616, -0.3499612510204315, -0.3364585340023041, -0.2255396544933319, -0.007184447720646858, -0.21592660248279572, 0.00239368318580091, -0.07125626504421234, 0.19345656037330627, -0.15920580923557281, 0.17167401313781738, 0.3034624457359314, 0.09692032635211945, 0.16826258599758148, -0.17914852499961853, 0.05318061262369156, 0.22526109218597412, 0.1842545121908188, -0.08278989791870117, -0.08363905549049377, -0.09633418172597885, -0.13919824361801147, -0.13139444589614868, 0.4392051696777344, -0.2431434541940689, 0.1521119773387909, 0.27531206607818604, 0.16114865243434906, 0.29860109090805054, -0.2021786868572235, 0.018696404993534088, 0.37583088874816895, 0.48768600821495056, 0.10426004976034164, -0.3134582042694092, -0.47657108306884766, -0.022540299221873283, 0.18318982422351837, 0.36662188172340393, 0.3496962785720825, -0.2403942495584488, 0.03685631603002548, -0.4264366626739502, -0.029453495517373085, -0.13254474103450775, 0.08995367586612701, 0.2136923372745514, 0.10086160153150558, 0.15536978840827942, -0.003271779278293252, 0.044661037623882294, -0.09545880556106567, 0.15903490781784058, -0.23730114102363586, -0.1132025495171547, 0.042656611651182175, -0.13779018819332123, -0.13386541604995728, -0.09080652892589569, 0.015388456173241138, 0.19063471257686615, 0.2048223465681076, 0.3135956823825836, 0.05322405323386192, 0.22500498592853546, 0.21918268501758575, 0.47183486819267273, -0.04583921656012535, -0.08414648473262787, 0.5593457818031311, 0.2446119487285614, 0.03002576157450676, 0.06023688614368439, -0.01003847736865282, 0.27793020009994507, -0.12376896291971207, -0.08744284510612488, 0.0251360721886158, 0.25217586755752563, -0.5562331676483154, -0.46969074010849, 0.10254047065973282, -0.4754388928413391, -0.11085537821054459, -0.04994547367095947, 0.25997743010520935, -0.3506716191768646, 0.3180258572101593, -0.11212296038866043, 0.11180038750171661, -0.33322611451148987, -0.045167647302150726, -0.19081111252307892, 0.15689904987812042, -0.10932169109582901, 0.18961699306964874, 0.3031008541584015, -0.0997934341430664, 0.464413583278656, -0.14836886525154114, 0.0527668222784996, -0.16524390876293182, 0.42668530344963074, -0.17668205499649048, 0.12088891118764877, 0.2269916981458664, 0.0736209899187088, 0.23083314299583435, -0.022406838834285736, -0.14349493384361267, -0.37490126490592957, 0.14235258102416992, -0.18751952052116394, -0.5286052823066711, 0.11082921177148819, 0.12504614889621735, -0.39939799904823303, -0.05795322358608246, -0.010793885216116905, 0.19039051234722137, 0.3466760218143463, -0.385879248380661, -0.08683202415704727, -0.16370506584644318, -0.07959877699613571, 0.004998394288122654, -0.15900030732154846, 0.46130824089050293, 0.18528056144714355, -0.2022436559200287, 0.15319792926311493, 0.05393051728606224, 0.5437247157096863, 0.35257861018180847, 0.02455792762339115, 0.022252580150961876, -0.14071229100227356, 0.45799365639686584, 0.1890597641468048, -0.46631285548210144, -0.1278698444366455, 0.5851625204086304, 0.09912499040365219, 0.317083477973938, 0.2227524220943451, 0.14682307839393616, 0.233754500746727, -0.06082316115498543, 0.04069112613797188, 0.40092599391937256, -0.08982785046100616, 0.07596126198768616, -0.3674882650375366, -0.3140702545642853, 0.40807369351387024, 0.1790602058172226, 0.13126718997955322, 0.010799593292176723, -0.05128692463040352, 0.4866947829723358, 0.2602522075176239, -0.08486145734786987, 0.09025543183088303, 0.2709383964538574, -0.405914843082428, 0.13252176344394684, 0.07757267355918884, -0.09696796536445618, -0.3680000901222229, 0.0032124221324920654, -0.1876494586467743, 0.2800237536430359, 0.0573013499379158, -0.32654303312301636, -0.1517820805311203, -0.4228244423866272, -0.00793265737593174, 0.024887599050998688, -0.0178245659917593, 0.25395479798316956, 0.32474005222320557, 0.09333090484142303, -0.019458703696727753, 0.3759138882160187, -0.02651868388056755, 0.2407078891992569, -0.8501709699630737, 0.13954167068004608, -0.1763739287853241, -0.2839396893978119, 0.4763649106025696, 0.2768568992614746, 0.18114027380943298, -0.14808803796768188, -0.07467442750930786, 0.44800126552581787, -0.07221253961324692, 0.46613064408302307, -0.3489922285079956, 0.022523578256368637, 0.3167484998703003, -0.2829042375087738, -0.01944568380713463, 0.3597464859485626, 0.012246963568031788, -0.061130013316869736, -0.17984256148338318, 0.37260735034942627, -0.3008643090724945, 0.2264278680086136, 0.10579652339220047, 0.009543619118630886, 0.12567362189292908, -0.2296977937221527, -0.2462264448404312, -0.5783407688140869, 0.5604599118232727, -0.10399427264928818, -0.33910125494003296, -0.0010968884453177452, -0.2969706654548645, -0.01972680725157261, 0.428888738155365, 0.11643330752849579, -0.06571304798126221, -0.03648027032613754, -0.005256153643131256, -0.15044257044792175, 0.19341453909873962, 0.18049214780330658, 0.4251810312271118, 0.12823985517024994, 0.08470896631479263, 0.012771978043019772, -0.17027944326400757, -0.35771456360816956, 0.11679737269878387, 0.051716942340135574, 0.003451257012784481, 0.3891894221305847, 0.06336168199777603, 0.032328713685274124, -0.34547853469848633, -0.027957865968346596, 0.20833437144756317, 0.10395707190036774, -0.38145825266838074, -0.1296589970588684, -0.42635393142700195, -0.3968113362789154, -0.23945941030979156, -0.17718367278575897, -0.12151750922203064, -0.3765020966529846, 0.20473913848400116, 0.33553093671798706, -0.2092922478914261, 0.11408428847789764, -0.27202168107032776, -0.012220565229654312, -0.33466193079948425, -0.06908030807971954, -0.07071831077337265, -0.22454892098903656, -0.29378873109817505, -0.03721277415752411, 0.19565996527671814, 0.4540422558784485, 0.22492294013500214, -0.3926708698272705, -0.1600925773382187, 0.054462943226099014, -0.2624691426753998, 0.08240657299757004, -0.08520682156085968, -0.07174309343099594, 0.11912183463573456, -0.1622811108827591, 0.370414674282074, -0.2849781811237335, 0.029572250321507454, -0.161270871758461, -0.29478371143341064, 0.036483049392700195, 0.1533922553062439, 0.03398454934358597, -0.2557275593280792, -0.5162025094032288, -0.2186129242181778, -0.36561378836631775, -0.1080164983868599, 0.2695896327495575, 0.18670199811458588, 0.09474556148052216, -0.15174277126789093, -0.0998639240860939, -0.03366892784833908, 0.3037440776824951, -0.25309568643569946, -0.11181105673313141, 0.3357352316379547, -0.22152751684188843, -0.1377464383840561, 0.054697923362255096, -0.05541593208909035, -0.057857099920511246, 0.12819123268127441, -0.5356532335281372, -0.11282774806022644, -0.22534406185150146, -0.053178589791059494, -0.2187700718641281, -0.30812549591064453, 0.4669429659843445, -0.07963427156209946, -0.02478385902941227, -0.037710223346948624, -0.38316184282302856, 0.07807600498199463, 0.06121470034122467, 0.4086659252643585, -0.11687025427818298, 0.2762601971626282, -0.0010491939028725028, 0.4330560564994812, 0.5189312100410461, 0.08851172029972076, 0.0033364570699632168, 0.13722172379493713, 0.1260216236114502, 0.10288182646036148, 0.028162293136119843, 0.30397528409957886, -0.020976299419999123, -0.12858740985393524, 0.35867926478385925, 0.06537868827581406, -0.28293532133102417, -0.14736051857471466, -0.11749089509248734, -0.3157089948654175, -0.2660253345966339, 0.1234099417924881, -0.34433066844940186, 0.12941215932369232, 0.19345518946647644, 0.34079402685165405, -0.08182303607463837, -0.5395628809928894, -0.018391774967312813, 0.2862551510334015, 0.2730647623538971, 0.13522247970104218, -0.4073793590068817, 0.1556655615568161, -0.46853306889533997, 0.27187836170196533, 0.18034252524375916, 0.3475732207298279, -0.16948001086711884, -0.15780490636825562, 0.055487532168626785, -0.20148690044879913, 0.33941730856895447, -0.4746231436729431, 0.013200522400438786, -0.004179333336651325, -0.07114516943693161, -0.423763245344162, -0.189687117934227, -0.18537156283855438, 0.40302497148513794, 0.49320143461227417, 0.41495776176452637, -0.37820690870285034, -0.07719402015209198, 0.24144481122493744, -0.08462478220462799, -0.05414042994379997, -0.15137523412704468, -0.21289369463920593, -0.3216531574726105, -0.27340802550315857, -0.15363159775733948, 0.0058207763358950615, 0.25832700729370117, -0.10156745463609695, -0.1375628262758255, 0.09534545242786407, 0.14404426515102386, 0.4129197895526886, 0.2945847809314728, 0.1835900843143463, 0.018144845962524414, 0.14950361847877502, 0.3504323959350586, 0.4505656063556671, 0.17933355271816254, 0.4881434142589569, 0.06032538414001465, -0.1553128957748413, 0.06298741698265076, -0.3202919065952301, 0.04379040375351906, 0.4678914546966553, 0.19955559074878693, 0.021901912987232208, 0.12911377847194672, -0.1437588483095169, -0.3940749764442444, -0.005321770440787077, 0.06702899932861328, -0.17417648434638977, -0.33612462878227234, -0.3677903115749359, 0.2681986391544342, 0.3401360809803009, -0.11092506349086761, -0.16600045561790466, -0.1509866863489151, -0.3498975336551666, 0.3399808406829834, 0.1721285581588745, 0.945644736289978, 0.28829699754714966, -0.015686381608247757, -0.0824362263083458, -0.17538733780384064, 0.8163896799087524, -0.23600581288337708, 0.43048304319381714, -0.3811829686164856, -0.24644888937473297, -0.08875585347414017, -0.1746107041835785, 0.13842268288135529, 0.11361934244632721, -0.35978755354881287, 0.3247303068637848, 0.13030101358890533, 0.034615084528923035, -0.191714346408844, 0.4030132293701172, -0.28494536876678467, -0.19131138920783997, -0.15553104877471924, 0.023692715913057327, -0.1681818962097168, 0.384303480386734, 0.055374350398778915, -0.07974913716316223, -0.2306414246559143, -0.21798478066921234, -0.43404167890548706, 0.23839311301708221, 0.15239746868610382, 0.04199132323265076, 0.261804461479187, -0.21192966401576996, -0.02144448831677437, 0.12296968698501587, 0.33395954966545105, -0.17732946574687958, -0.17432692646980286, -0.005739698652178049, -0.10337192565202713, -0.18696163594722748, 0.14108487963676453, 0.05936392769217491, 0.3844887614250183, -0.09873609989881516, -0.011702921241521835, 0.08156541734933853, -0.15946175158023834, -0.265009343624115, -0.17423227429389954, -0.1326058954000473, -0.18644283711910248, -0.18964193761348724, -0.16944268345832825, 0.013096786104142666, -0.06749511510133743, -0.08927696943283081, 0.013088562525808811, -0.13440066576004028, -0.15610583126544952, 0.21438707411289215, -0.03449400141835213, -0.3317166864871979, -0.10979115217924118, 0.4217982292175293, 0.33537736535072327, -0.003499122802168131, 0.4691965878009796, -0.06988602876663208, -0.11542747914791107, -0.15430639684200287, -0.07997774332761765, -0.14899997413158417, -0.33252909779548645, 0.20993627607822418, -0.22150319814682007, -0.08268962800502777, 0.23825141787528992, 0.2852211892604828, 0.133651003241539, 0.2915157973766327, -0.12268858402967453, -0.4712008535861969, -0.009561262093484402, 0.2567784786224365, -0.019193828105926514, 0.11481335014104843, -0.14997422695159912, 0.39572611451148987, -0.18629437685012817, 0.1535692811012268, -0.19016426801681519, -0.07312841713428497, -0.48062992095947266, 0.35391852259635925, 0.0994262769818306, -0.007976687513291836, -0.08883369714021683, -0.027649248018860817, -0.16565969586372375, 0.28192445635795593, -0.01915896311402321, 0.006709287874400616, -0.13281430304050446, 0.19781996309757233, 0.09791260212659836, -0.10110511630773544, -0.13994500041007996, 0.11612612009048462, 0.02054685354232788, -0.00484836008399725, 0.03788478672504425, 0.2670266628265381, -0.008994183503091335, 0.3893875777721405, -0.36771687865257263, 0.09063167124986649, -0.05895492807030678, 0.18782037496566772, -0.3241513967514038, 0.2810365557670593, -0.1203463077545166, 0.21074910461902618, -0.04493272304534912, 0.044363632798194885, 0.03362700343132019, -0.02852052077651024, 0.03280812129378319, 0.04359224811196327, 0.14903274178504944, -0.16278529167175293, 0.14060981571674347, 0.33817780017852783, -0.023935217410326004, 0.5712108016014099, -0.0635785385966301, -0.16158320009708405, 0.03009963408112526, 0.08797784894704819, -0.08461970835924149, -0.0033458888065069914, 0.20257987082004547, -0.060471273958683014, -0.06835423409938812, 0.3554307520389557, 0.1837964653968811, -0.10070601105690002, -0.18464164435863495, 0.1138506531715393, 0.42155104875564575, -0.17783132195472717, 0.19242531061172485, 0.19783838093280792, -0.21465839445590973, -0.015783781185746193, 0.16785579919815063, -0.20996685326099396, 0.19654639065265656, 0.04577749967575073, 0.020209182053804398, 0.47435876727104187, 0.1578318476676941, 0.3953709602355957, -0.07260758429765701, -0.21184633672237396, -0.06808505207300186, 0.601893424987793, 0.3082672357559204, 0.43376463651657104, 0.22709417343139648, 0.325705349445343, -0.2783259153366089, -0.00010940757056232542, -0.35041481256484985, -0.18376082181930542, -0.011417302303016186, -0.16466306149959564, -0.4718987047672272, -0.09388595074415207, -0.23944783210754395, 0.09697277098894119, -0.15479280054569244, 0.1520403027534485, 0.3951610326766968, -0.044419340789318085, -0.20598697662353516, -0.5548536777496338, 0.13787557184696198, -0.13042598962783813, 0.19312043488025665, -0.2679714262485504, 0.18461093306541443, 0.0576748326420784, -0.11533725261688232, 0.43101707100868225, 0.5478071570396423, 0.5472567081451416, 0.32145920395851135, -0.17351742088794708, -0.11064166575670242, 0.03344208747148514, 0.032779715955257416, 0.19032767415046692, 0.7335734367370605, 0.32219818234443665, -0.2711005210876465, 0.37458062171936035, -0.05907514691352844, -0.05481630563735962, 0.413944274187088, 0.12746430933475494, 0.015450875274837017, -0.06342161446809769, 0.5102357268333435, -0.17088817059993744, 0.10250909626483917, 0.006850921083241701, 0.02524922788143158, -0.27466610074043274, -0.3511110842227936, 0.2094677984714508, -0.32757478952407837, 0.22225098311901093, -0.030705513432621956, 0.01895626075565815, -0.2546960711479187, 0.5723082423210144, 0.4384699761867523, -0.14093200862407684, -0.21884936094284058, -0.010653417557477951, -0.4845338463783264, 0.3560204803943634, -0.3736160695552826, -0.16402667760849, -0.13282252848148346, 0.5531365871429443, -0.2862456738948822, 0.23929330706596375, -0.2779919505119324, 0.3856368660926819, -0.17468710243701935, 0.16580167412757874, -0.25717657804489136, -0.013568085618317127, -0.10855469852685928, -0.22183643281459808, 0.006634814664721489, -0.24003109335899353, 0.14324882626533508, 0.04273706674575806, -0.12452125549316406, 0.0315433107316494, -0.024334577843546867, 0.06714487075805664, 0.49722930788993835, -0.0039029463659971952, 0.47871309518814087, 0.4656323194503784, -0.15694749355316162, -0.23094379901885986, -0.1626695990562439, -0.3780439794063568, -0.34398844838142395, 0.3681856095790863, 0.17288751900196075, 0.43301576375961304, -0.17476147413253784, 0.5583013296127319, -0.04618223384022713, 0.2748158574104309, -0.04320492967963219, -0.004775571636855602, -0.320286363363266, 0.012488414533436298, -0.27113020420074463, 0.02506726048886776, 0.12636448442935944, 0.37036898732185364, 0.17302142083644867, -0.020745672285556793, -0.18759030103683472, -0.41432109475135803, 0.4301750957965851, -0.354178786277771, 0.05320926383137703, -0.13932406902313232, 0.1548423320055008, 0.2172611802816391, -0.14222216606140137, -0.753833532333374, -0.08064065873622894, 0.27263835072517395, -0.16095569729804993, 0.050614919513463974, -0.013103938661515713, -0.014781929552555084, -0.2709467113018036, -0.15545177459716797, -0.2552732825279236, 0.24762387573719025, -0.3734522759914398, 0.30728673934936523, -0.09805559366941452 ]
https://github.com/huggingface/datasets/issues/215
NonMatchingSplitsSizesError when loading blog_authorship_corpus
> I'm getting this same issue when loading the `imdb` corpus via `dataset = load_dataset("imdb")`. When I try `ignore_verifications=True`, no examples are read into the `train` portion of the dataset. When the checksums don't match, it may mean that the file you downloaded is corrupted. In this case you can try to load the dataset again `load_dataset("imdb", download_mode="force_redownload")` Also I just checked on my side and it worked fine: ```python from datasets import load_dataset dataset = load_dataset("imdb") print(len(dataset["train"])) # 25000 ``` Let me know if redownloading fixes your issue @EmilyAlsentzer . If not, feel free to open a separate issue.
Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on.
100
NonMatchingSplitsSizesError when loading blog_authorship_corpus Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on. > I'm getting this same issue when loading the `imdb` corpus via `dataset = load_dataset("imdb")`. When I try `ignore_verifications=True`, no examples are read into the `train` portion of the dataset. When the checksums don't match, it may mean that the file you downloaded is corrupted. In this case you can try to load the dataset again `load_dataset("imdb", download_mode="force_redownload")` Also I just checked on my side and it worked fine: ```python from datasets import load_dataset dataset = load_dataset("imdb") print(len(dataset["train"])) # 25000 ``` Let me know if redownloading fixes your issue @EmilyAlsentzer . If not, feel free to open a separate issue.
[ -0.17600978910923004, 0.19232207536697388, 0.06580757349729538, 0.4156298041343689, 0.012083323672413826, 0.19369195401668549, -0.06068258360028267, 0.3878975510597229, -0.05917373672127724, 0.1324821263551712, -0.07451507449150085, 0.17166590690612793, 0.009340972639620304, -0.13655927777290344, -0.07436719536781311, 0.00014040738460607827, -0.061030659824609756, 0.1483679562807083, 0.07987795025110245, 0.035929057747125626, -0.06956396251916885, 0.430107057094574, -0.3462485074996948, -0.08884044736623764, -0.23781442642211914, -0.1127564087510109, 0.029419295489788055, 0.3264782726764679, 0.04625575616955757, -0.3628292977809906, 0.2542952299118042, -0.06689704954624176, 0.1275966912508011, 0.3013105094432831, -0.000123778052511625, 0.018736258149147034, 0.6639289259910583, -0.24313727021217346, -0.19303753972053528, -0.3823125958442688, -0.29839053750038147, -0.27950403094291687, -0.031189963221549988, -0.23511360585689545, 0.00004675639502238482, -0.045577604323625565, 0.13764603435993195, -0.17305052280426025, 0.2249153107404709, 0.34403976798057556, 0.09746165573596954, 0.18349282443523407, -0.09256050735712051, 0.11123760789632797, 0.21578523516654968, 0.12078322470188141, -0.04892580583691597, -0.014452725648880005, -0.017778947949409485, -0.15843066573143005, -0.03852710500359535, 0.3862314820289612, -0.3328864574432373, 0.11739900708198547, 0.257072776556015, 0.14909572899341583, 0.3444834351539612, -0.1964063048362732, 0.031481701880693436, 0.3853248655796051, 0.4552387297153473, 0.02042105235159397, -0.34347349405288696, -0.43859392404556274, -0.003331762505695224, 0.09549654275178909, 0.385768324136734, 0.36564886569976807, -0.25014546513557434, 0.04120177775621414, -0.34488239884376526, 0.020255209878087044, -0.1232108324766159, 0.10563656687736511, 0.16576077044010162, 0.08337178081274033, 0.15274548530578613, -0.013326913118362427, 0.04077646881341934, -0.03611772879958153, 0.14461590349674225, -0.2656337022781372, -0.10599860548973083, 0.09341879189014435, -0.16324135661125183, -0.12530241906642914, -0.18923711776733398, -0.011101770214736462, 0.16719697415828705, 0.19994880259037018, 0.29252442717552185, 0.010139454156160355, 0.1910586655139923, 0.18825729191303253, 0.4692175090312958, -0.05503099411725998, -0.10778764635324478, 0.5959747433662415, 0.2141972780227661, 0.08869105577468872, -0.00538589945062995, -0.031112447381019592, 0.2485346496105194, -0.20115704834461212, 0.02705664373934269, 0.12978431582450867, 0.24780985713005066, -0.5492251515388489, -0.510543167591095, 0.17850784957408905, -0.472481369972229, -0.12943395972251892, -0.03337767720222473, 0.21454067528247833, -0.3933624029159546, 0.3246039152145386, -0.064974844455719, 0.12263613939285278, -0.3191176652908325, -0.004285333678126335, -0.15956458449363708, 0.12740503251552582, -0.13459959626197815, 0.16218243539333344, 0.28917282819747925, -0.1728476881980896, 0.40299683809280396, -0.11030672490596771, 0.06361155956983566, -0.21592584252357483, 0.30708909034729004, -0.206889808177948, 0.06568220257759094, 0.23887181282043457, 0.047185108065605164, 0.2478991597890854, 0.0007343612378463149, -0.20997609198093414, -0.3573240041732788, 0.13655602931976318, -0.20479097962379456, -0.4814675748348236, 0.07134093344211578, 0.12074007838964462, -0.39544326066970825, -0.047843579202890396, -0.07426383346319199, 0.03617919608950615, 0.36518341302871704, -0.42673027515411377, -0.09594748914241791, -0.2285139560699463, -0.07225844264030457, -0.035934146493673325, -0.10529311746358871, 0.5175381302833557, 0.15426714718341827, -0.16895312070846558, 0.13861128687858582, 0.05945953354239464, 0.5408576130867004, 0.39143243432044983, 0.05832086130976677, 0.025509018450975418, -0.19143207371234894, 0.3766971528530121, 0.20354241132736206, -0.530716061592102, -0.2367614358663559, 0.5350238084793091, 0.13700063526630402, 0.3692920207977295, 0.22536338865756989, 0.14809495210647583, 0.2658127248287201, -0.06201300397515297, 0.01656775176525116, 0.4239158630371094, -0.046328574419021606, 0.07779411971569061, -0.3728451430797577, -0.2930842936038971, 0.392194002866745, 0.1419515460729599, 0.15030765533447266, 0.04352617263793945, -0.0213348176330328, 0.44595447182655334, 0.3237892985343933, -0.04585520923137665, 0.09353803843259811, 0.2675274908542633, -0.2806485891342163, 0.25819632411003113, 0.049429528415203094, -0.22007183730602264, -0.3241138458251953, 0.027645595371723175, -0.19926141202449799, 0.2994351387023926, -0.09630519896745682, -0.26976680755615234, -0.17187517881393433, -0.45702919363975525, -0.04208001494407654, -0.00748974597081542, -0.006650733761489391, 0.20860138535499573, 0.28892454504966736, 0.1307031363248825, -0.06835644692182541, 0.42508840560913086, -0.09385278820991516, 0.24119189381599426, -0.8024846911430359, 0.2409888505935669, -0.19558757543563843, -0.31735941767692566, 0.4423465430736542, 0.2632888853549957, 0.19873645901679993, -0.14953100681304932, -0.14108438789844513, 0.44249001145362854, -0.031132962554693222, 0.42057159543037415, -0.3902110457420349, 0.0709906816482544, 0.32512375712394714, -0.3283195495605469, -0.013646182604134083, 0.3792681097984314, 0.0683736652135849, -0.09130802750587463, -0.09971366822719574, 0.3737078309059143, -0.37633273005485535, 0.21936780214309692, 0.13083110749721527, 0.029671713709831238, 0.13359175622463226, -0.21173645555973053, -0.2728489637374878, -0.5291533470153809, 0.5458316206932068, -0.0643116906285286, -0.31283092498779297, 0.004712572321295738, -0.23172347247600555, -0.06337516009807587, 0.41675662994384766, 0.0390753373503685, -0.07154995203018188, -0.01635383628308773, -0.0056105125695466995, -0.09737599641084671, 0.07564883679151535, 0.20833970606327057, 0.4420284926891327, 0.09719030559062958, 0.01375551801174879, 0.07849960029125214, -0.20874330401420593, -0.31791356205940247, 0.16330701112747192, 0.01875685714185238, 0.0033218245953321457, 0.4489206075668335, 0.01393331028521061, 0.0849304124712944, -0.27898821234703064, -0.03021649830043316, 0.11369824409484863, 0.18893900513648987, -0.42711010575294495, -0.09499035775661469, -0.42307066917419434, -0.38129326701164246, -0.32520997524261475, -0.19172699749469757, -0.15214823186397552, -0.3741568922996521, 0.10221884399652481, 0.3356708884239197, -0.09335179626941681, 0.1520269811153412, -0.27478134632110596, -0.07710868120193481, -0.28563815355300903, -0.1671755313873291, -0.040072955191135406, -0.25538569688796997, -0.22221358120441437, -0.06030001491308212, 0.2851202189922333, 0.4162527620792389, 0.2191532701253891, -0.43143609166145325, -0.1876336634159088, -0.003960995469242334, -0.21282848715782166, 0.047138623893260956, -0.0921221449971199, -0.010914796963334084, 0.1246258094906807, -0.09668314456939697, 0.3530615270137787, -0.3084544539451599, 0.13665306568145752, -0.18659105896949768, -0.28652581572532654, 0.08701907098293304, 0.1654970496892929, 0.00590186333283782, -0.16749294102191925, -0.47558751702308655, -0.26043933629989624, -0.3520001173019409, -0.15064622461795807, 0.26975417137145996, 0.18269242346286774, 0.09125323593616486, -0.12194875627756119, -0.1085171177983284, -0.03932345286011696, 0.35023632645606995, -0.21496829390525818, -0.12616309523582458, 0.29936835169792175, -0.15706680715084076, -0.17148812115192413, 0.006665576249361038, -0.019033754244446754, -0.02802501805126667, 0.12993429601192474, -0.525076150894165, -0.1266777515411377, -0.2553535997867584, -0.019448071718215942, -0.22357916831970215, -0.2169739007949829, 0.4017183482646942, -0.01997067593038082, 0.007223912980407476, -0.007094128523021936, -0.3523038327693939, 0.027413614094257355, 0.05319155752658844, 0.38160014152526855, -0.1346924602985382, 0.30944931507110596, -0.03444184735417366, 0.42649316787719727, 0.5960373282432556, 0.03605298697948456, 0.02763943374156952, 0.07134292274713516, 0.14365246891975403, 0.07474680244922638, -0.10544990003108978, 0.26571565866470337, -0.07296619564294815, -0.07890402525663376, 0.2505123019218445, 0.00891830399632454, -0.22383223474025726, -0.1783445179462433, -0.12964576482772827, -0.45140233635902405, -0.27320626378059387, 0.10059618204832077, -0.3199053108692169, 0.17668387293815613, 0.13752329349517822, 0.28149300813674927, -0.05517859756946564, -0.5646659135818481, 0.029398171231150627, 0.41505563259124756, 0.2145816832780838, 0.11504007875919342, -0.3385311961174011, 0.1420573890209198, -0.3411738872528076, 0.26438772678375244, 0.20610591769218445, 0.37408265471458435, -0.1111697107553482, -0.11376488953828812, 0.0390937514603138, -0.19258259236812592, 0.3788107633590698, -0.4746636748313904, 0.0555170476436615, -0.004085300490260124, -0.0704110860824585, -0.41189613938331604, -0.21917076408863068, -0.18260222673416138, 0.36919209361076355, 0.46578559279441833, 0.4691718816757202, -0.35209381580352783, -0.06196654587984085, 0.2556953430175781, 0.0213908851146698, -0.10110480338335037, -0.17457976937294006, -0.22858616709709167, -0.33551129698753357, -0.26188838481903076, -0.1732977032661438, -0.04387543350458145, 0.31121665239334106, -0.07534477114677429, -0.08241888135671616, 0.042251214385032654, 0.1665191799402237, 0.4253271818161011, 0.23124060034751892, 0.20768524706363678, -0.032401252537965775, 0.2240993231534958, 0.25176897644996643, 0.40947484970092773, 0.21285994350910187, 0.5298270583152771, -0.0027637809980660677, -0.19281627237796783, -0.07655362039804459, -0.30114442110061646, 0.005492973607033491, 0.4374135136604309, 0.20441003143787384, -0.03045053221285343, 0.12637093663215637, -0.13738709688186646, -0.42967525124549866, -0.05294442176818848, 0.12660479545593262, -0.10131753981113434, -0.38544780015945435, -0.49731147289276123, 0.29982197284698486, 0.35049936175346375, -0.11287198215723038, -0.1433967798948288, -0.09955043345689774, -0.33479371666908264, 0.34836724400520325, 0.17213237285614014, 0.9992435574531555, 0.22708699107170105, -0.025044381618499756, -0.0683605894446373, -0.20097851753234863, 0.7844552993774414, -0.21298298239707947, 0.3598968982696533, -0.4350256323814392, -0.29632171988487244, -0.10602013766765594, -0.18125852942466736, 0.1910448670387268, 0.10323560982942581, -0.33262643218040466, 0.38838016986846924, 0.10925716906785965, -0.017338231205940247, -0.20386295020580292, 0.40578901767730713, -0.28847745060920715, -0.1145382672548294, -0.1564043164253235, 0.020807486027479172, -0.18778890371322632, 0.4317871034145355, 0.08245346695184708, -0.05726068094372749, -0.26244494318962097, -0.16835258901119232, -0.5494995713233948, 0.221664696931839, 0.14272399246692657, 0.026372166350483894, 0.21179968118667603, -0.2248518466949463, -0.013590328395366669, 0.13423407077789307, 0.3278850317001343, -0.11514629423618317, -0.10681699216365814, -0.0016030421247705817, -0.022616684436798096, -0.1557306945323944, 0.20076487958431244, 0.05771096795797348, 0.4340375065803528, -0.10155361890792847, -0.06176256015896797, 0.10912148654460907, -0.07918635755777359, -0.2611493766307831, -0.13389451801776886, -0.1395936906337738, -0.1802154928445816, -0.21085792779922485, -0.20317086577415466, 0.014056478627026081, -0.029256314039230347, -0.0973198264837265, 0.01700647920370102, -0.05921357870101929, -0.1978055238723755, 0.2540108263492584, -0.09184740483760834, -0.3513317406177521, -0.09548190236091614, 0.4479711353778839, 0.28415730595588684, -0.04742627218365669, 0.4655071794986725, -0.044768743216991425, -0.11638551950454712, -0.14287002384662628, -0.0799197405576706, -0.2282400280237198, -0.36086660623550415, 0.17055682837963104, -0.25035685300827026, 0.0032191630452871323, 0.2146570235490799, 0.2766200602054596, 0.23138098418712616, 0.28160741925239563, -0.0997285321354866, -0.5418651103973389, -0.005885081831365824, 0.1627015918493271, 0.09235384315252304, 0.15330176055431366, -0.19041457772254944, 0.3851993978023529, -0.20321124792099, 0.08855759352445602, -0.18876266479492188, 0.001679087057709694, -0.4216434955596924, 0.3190796971321106, 0.09868451952934265, -0.010334907099604607, -0.10647942125797272, -0.05362092703580856, -0.10954689234495163, 0.23286420106887817, -0.008623616769909859, -0.0021025610622018576, -0.17353090643882751, 0.19913406670093536, 0.07818415015935898, -0.1462138444185257, -0.12778005003929138, 0.029298119246959686, 0.07194796949625015, -0.03990612179040909, -0.032883331179618835, 0.3182606101036072, 0.016946081072092056, 0.4303992688655853, -0.2967918813228607, 0.00705746840685606, -0.05222337320446968, 0.25376269221305847, -0.3740805387496948, 0.3204044997692108, -0.06483282893896103, 0.23397700488567352, -0.0029912416357547045, 0.07607720047235489, 0.005250886082649231, -0.010939202271401882, -0.03330615907907486, 0.044250670820474625, 0.26031816005706787, -0.18230994045734406, 0.15160931646823883, 0.3269649147987366, 0.0630296990275383, 0.6621743440628052, -0.0977918803691864, -0.1435750424861908, 0.1002189889550209, 0.0893855169415474, -0.12129441648721695, 0.017685338854789734, 0.2556923031806946, -0.06329984217882156, -0.080075703561306, 0.37514805793762207, 0.160386860370636, -0.11737366020679474, -0.24033242464065552, 0.09514525532722473, 0.5035057663917542, -0.13529197871685028, 0.18721497058868408, 0.2672550678253174, -0.052081406116485596, -0.035117655992507935, 0.11959274858236313, -0.21357020735740662, 0.11305531859397888, 0.08282117545604706, -0.016397768631577492, 0.4081757664680481, 0.05979673191905022, 0.38107213377952576, -0.11079724878072739, -0.19281350076198578, -0.016883840784430504, 0.5765307545661926, 0.19954814016819, 0.44371160864830017, 0.23778997361660004, 0.3454335331916809, -0.3376036286354065, 0.025902314111590385, -0.31170201301574707, -0.1353369802236557, -0.013288070447742939, -0.1735335886478424, -0.44935303926467896, -0.1327100247144699, -0.22633376717567444, 0.11731559783220291, -0.13277418911457062, 0.16288970410823822, 0.3212735652923584, 0.033501509577035904, -0.2087704986333847, -0.5127847790718079, 0.06974226981401443, -0.11926989257335663, 0.2277921587228775, -0.28510183095932007, 0.11935030668973923, 0.12183278053998947, -0.06800276041030884, 0.3805287182331085, 0.5955411195755005, 0.6343861818313599, 0.39273351430892944, -0.19025813043117523, -0.05602798983454704, 0.036398038268089294, 0.05950111150741577, 0.20654892921447754, 0.7163819074630737, 0.24613206088542938, -0.2202010303735733, 0.3402116596698761, -0.03373246267437935, -0.06545190513134003, 0.3719549775123596, 0.07000888884067535, 0.0255960151553154, -0.08627068996429443, 0.5236680507659912, -0.19162161648273468, 0.11693493276834488, -0.05106043443083763, 0.06794944405555725, -0.32060739398002625, -0.3236783742904663, 0.26137739419937134, -0.26411762833595276, 0.24652712047100067, -0.00307719805277884, 0.01530559454113245, -0.20477698743343353, 0.5182000994682312, 0.5190639495849609, -0.09691651165485382, -0.24339556694030762, -0.011231189593672752, -0.5534386038780212, 0.3506767749786377, -0.3848561644554138, -0.21972613036632538, -0.14044590294361115, 0.5029479265213013, -0.2913239896297455, 0.2718914747238159, -0.1669800728559494, 0.4217170476913452, -0.19998997449874878, 0.23926480114459991, -0.2612937390804291, -0.0482010692358017, -0.08566754311323166, -0.24912485480308533, -0.04120748117566109, -0.23633705079555511, 0.12473712861537933, 0.0006043595494702458, -0.094248466193676, 0.030081115663051605, -0.026399623602628708, 0.06136676296591759, 0.4949744641780853, 0.024867575615644455, 0.4822665750980377, 0.5218863487243652, -0.14398768544197083, -0.26332733035087585, -0.17555679380893707, -0.3367414176464081, -0.30829450488090515, 0.4396428167819977, 0.13124725222587585, 0.3744223713874817, -0.1879594624042511, 0.47989392280578613, -0.13751095533370972, 0.336193323135376, -0.025598766282200813, -0.11676663160324097, -0.37553268671035767, -0.051262471824884415, -0.22220538556575775, -0.031138557940721512, 0.09889984130859375, 0.421176940202713, 0.15882275998592377, 0.08846025913953781, -0.1798771172761917, -0.40906238555908203, 0.43700626492500305, -0.32739973068237305, 0.04309295862913132, -0.11276282370090485, 0.12388177961111069, 0.1499229073524475, -0.039151985198259354, -0.759273111820221, -0.035974305123090744, 0.30051389336586, -0.17703600227832794, -0.027316074818372726, -0.01370744127780199, -0.019451403990387917, -0.22175315022468567, -0.11064101755619049, -0.30953413248062134, 0.22752639651298523, -0.3596229553222656, 0.31355902552604675, -0.08145678788423538 ]
https://github.com/huggingface/datasets/issues/215
NonMatchingSplitsSizesError when loading blog_authorship_corpus
I wasn't aware of the "force_redownload" option and manually removed the '/home/me/.cache/huggingface/datasets/' dir, this worked for me (dataset 'cnn_dailymail')
Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on.
19
NonMatchingSplitsSizesError when loading blog_authorship_corpus Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on. I wasn't aware of the "force_redownload" option and manually removed the '/home/me/.cache/huggingface/datasets/' dir, this worked for me (dataset 'cnn_dailymail')
[ -0.059521086513996124, 0.15316830575466156, 0.07047892361879349, 0.38789090514183044, -0.015147232450544834, 0.14190791547298431, -0.09238570928573608, 0.320081889629364, -0.049356814473867416, 0.00811048224568367, -0.08665341883897781, 0.15768671035766602, 0.03654804453253746, 0.04614752531051636, -0.019716089591383934, 0.08979131281375885, -0.04565126821398735, 0.20413734018802643, 0.04544640704989433, 0.029078885912895203, 0.0009837569668889046, 0.5112382173538208, -0.24448415637016296, -0.06453901529312134, -0.3506292700767517, -0.1498447060585022, 0.054416295140981674, 0.3629712760448456, 0.010663223452866077, -0.3910154402256012, 0.31971681118011475, -0.047284264117479324, 0.11418914794921875, 0.2401961088180542, -0.00012255208275746554, -0.0028430665843188763, 0.4699679911136627, -0.21914024651050568, -0.2050280123949051, -0.27034232020378113, -0.21681641042232513, -0.29916292428970337, -0.046362750232219696, -0.2182561159133911, -0.06674132496118546, -0.06653883308172226, 0.07996447384357452, -0.051992133259773254, 0.37194907665252686, 0.29745885729789734, 0.12554721534252167, 0.09593836218118668, -0.12729528546333313, 0.1026805192232132, 0.08485616743564606, 0.06652500480413437, -0.01188482716679573, -0.04185543954372406, -0.1720016747713089, -0.32199475169181824, -0.12819801270961761, 0.5180575251579285, -0.1915791928768158, 0.1262468695640564, 0.1819387823343277, 0.08008390665054321, 0.28306472301483154, -0.1194165050983429, 0.06791145354509354, 0.3065376877784729, 0.4064789116382599, 0.1085793599486351, -0.22490538656711578, -0.5339215397834778, 0.007109073456376791, 0.05136090889573097, 0.45279964804649353, 0.34143373370170593, -0.19766266644001007, 0.10054169595241547, -0.415865957736969, -0.22632378339767456, -0.054597750306129456, 0.007279455196112394, 0.21390008926391602, 0.2067904770374298, 0.038818325847387314, -0.010092146694660187, 0.13749270141124725, -0.03788347914814949, 0.22440898418426514, -0.09223991632461548, -0.14465226233005524, 0.08490973711013794, -0.21261745691299438, -0.15949290990829468, -0.12389065325260162, 0.037815239280462265, 0.1740894317626953, 0.15352191030979156, 0.23721112310886383, -0.07823175191879272, 0.14463810622692108, 0.04192792624235153, 0.36939334869384766, 0.17495547235012054, -0.0018968230579048395, 0.42824193835258484, 0.1405404508113861, 0.033289384096860886, 0.11029357463121414, 0.0037991213612258434, 0.29398196935653687, -0.15722940862178802, -0.06706444919109344, 0.04519667848944664, 0.20995981991291046, -0.5436294078826904, -0.416538804769516, 0.07986767590045929, -0.4917491674423218, -0.20235198736190796, -0.12295089662075043, 0.13204719126224518, -0.34991776943206787, 0.3523452877998352, -0.07884310185909271, 0.03780760243535042, -0.37327709794044495, -0.055996209383010864, -0.20142947137355804, 0.10900062322616577, -0.19364477694034576, 0.23633521795272827, 0.2714066803455353, -0.28212088346481323, 0.43705159425735474, -0.05147682502865791, -0.09336377680301666, -0.2561865746974945, 0.2646711468696594, -0.030780786648392677, -0.02329210750758648, 0.19040632247924805, -0.13137562572956085, 0.2399364411830902, 0.008580385707318783, -0.2008439153432846, -0.2873157262802124, 0.01637277752161026, -0.27687501907348633, -0.6388425827026367, 0.08123855292797089, 0.1171504482626915, -0.40461665391921997, -0.05964193120598793, 0.017498068511486053, 0.22032175958156586, 0.37056395411491394, -0.2604256570339203, -0.10591307282447815, -0.16876038908958435, -0.10133886337280273, -0.02985287830233574, -0.16532829403877258, 0.5042774677276611, 0.13272725045681, -0.32241442799568176, 0.1607019007205963, -0.03138416260480881, 0.47419580817222595, 0.4351542890071869, -0.006501998286694288, 0.10734177380800247, -0.1725720912218094, 0.35535678267478943, 0.14280261099338531, -0.39499378204345703, -0.2550792694091797, 0.40007564425468445, -0.0010341099696233869, 0.2915744483470917, 0.183280810713768, 0.1310873031616211, 0.09201636910438538, -0.06459524482488632, 0.0591232106089592, 0.4729750156402588, 0.07429248094558716, 0.09922955185174942, -0.45281121134757996, -0.41699761152267456, 0.43306493759155273, 0.16881978511810303, 0.10977091640233994, 0.07702746242284775, -0.019928503781557083, 0.4880300462245941, 0.31506603956222534, -0.14438828825950623, 0.06338523328304291, 0.31531864404678345, -0.3313466012477875, 0.1586516946554184, 0.04939056932926178, -0.08946103602647781, -0.5167132616043091, 0.007717791944742203, -0.2730097472667694, 0.31955817341804504, 0.050448477268218994, -0.19811034202575684, -0.23507963120937347, -0.4627164602279663, -0.07891248166561127, -0.1806631088256836, 0.0007561777601949871, 0.23502124845981598, 0.409032940864563, 0.054551009088754654, -0.018293334171175957, 0.5882978439331055, -0.09648596495389938, 0.24484986066818237, -0.8845165967941284, 0.2083260416984558, -0.12776190042495728, -0.16899597644805908, 0.4122888147830963, 0.2858724296092987, 0.2703772485256195, -0.14165276288986206, -0.13194596767425537, 0.46327075362205505, -0.2258068174123764, 0.4788299798965454, -0.22793690860271454, -0.002263889880850911, 0.2731257379055023, -0.29561731219291687, 0.021120797842741013, 0.21801649034023285, -0.04656475409865379, -0.05527111142873764, -0.0661938264966011, 0.3710584342479706, -0.36957669258117676, 0.1385297328233719, 0.017573844641447067, -0.04477107152342796, 0.10005279630422592, -0.28728124499320984, -0.14379754662513733, -0.4895390272140503, 0.5787158608436584, -0.15024621784687042, -0.2525637447834015, 0.02148449420928955, -0.13383252918720245, 0.01638254150748253, 0.37939906120300293, 0.06490432471036911, -0.09646505117416382, -0.03197634220123291, -0.016053060069680214, -0.20027048885822296, 0.05359133705496788, 0.24537506699562073, 0.5202925801277161, 0.13648900389671326, 0.026375271379947662, 0.09233337640762329, -0.16415871679782867, -0.3306785821914673, 0.1524430364370346, -0.06436366587877274, 0.08235184848308563, 0.36921700835227966, 0.17514647543430328, -0.002143571153283119, -0.5159096121788025, 0.1042153388261795, 0.21197234094142914, 0.2446637898683548, -0.3040996193885803, -0.1348416805267334, -0.36489394307136536, -0.386615514755249, -0.266776978969574, -0.19045580923557281, -0.3162628710269928, -0.3300527036190033, 0.20578348636627197, 0.28559064865112305, -0.11775808036327362, 0.15927386283874512, -0.09243586659431458, 0.06076815351843834, -0.3277832269668579, 0.009110375307500362, -0.10953298956155777, -0.275867760181427, -0.2055220752954483, -0.050362423062324524, 0.10383976250886917, 0.32214972376823425, 0.18946237862110138, -0.356062650680542, -0.19915369153022766, -0.07048135995864868, -0.3379550576210022, 0.04746352136135101, -0.08756743371486664, -0.07165787369012833, 0.08050350844860077, -0.06838131695985794, 0.3928326666355133, -0.15853282809257507, 0.14054273068904877, -0.2609809935092926, -0.30935055017471313, -0.04161081463098526, 0.17137882113456726, 0.19560883939266205, -0.1676418036222458, -0.5419092178344727, -0.2437300682067871, -0.3634907305240631, -0.015157089568674564, 0.2141260802745819, 0.18519483506679535, 0.17180363833904266, -0.13248242437839508, -0.2091616839170456, -0.1113305538892746, 0.25094616413116455, -0.3925289809703827, -0.1390978842973709, 0.19587832689285278, -0.07682496309280396, -0.13032099604606628, -0.07141608744859695, 0.10173924267292023, -0.0858050137758255, 0.07061292231082916, -0.623704195022583, -0.2201807200908661, -0.241982564330101, -0.0938950628042221, -0.30282989144325256, -0.17801013588905334, 0.5007234811782837, -0.045411765575408936, -0.004133107606321573, 0.08094992488622665, -0.45969557762145996, 0.017947647720575333, 0.13317011296749115, 0.4873122572898865, -0.1790710836648941, 0.23810629546642303, -0.019555862993001938, 0.5533110499382019, 0.5295944213867188, 0.05317655950784683, -0.009989600628614426, 0.16884249448776245, 0.22626559436321259, 0.040081750601530075, -0.13294917345046997, 0.33300819993019104, -0.06891188025474548, 0.03726252168416977, 0.3456495404243469, 0.10086365044116974, -0.14929641783237457, -0.14536269009113312, -0.16621951758861542, -0.3622745871543884, -0.18016979098320007, 0.1373780369758606, -0.1392013281583786, 0.12042450904846191, 0.2959349751472473, 0.14354240894317627, -0.17569293081760406, -0.47165849804878235, -0.008708767592906952, 0.24278424680233002, 0.14365431666374207, 0.1978633552789688, -0.457914263010025, 0.17113959789276123, -0.5147837996482849, 0.26574864983558655, 0.1584012657403946, 0.26466092467308044, -0.1318330615758896, -0.14703157544136047, 0.0993957594037056, -0.17282640933990479, 0.394021213054657, -0.4049489498138428, -0.07724948972463608, -0.07967538386583328, -0.0167161226272583, -0.3363563120365143, -0.17841491103172302, -0.03545190766453743, 0.30832505226135254, 0.4751308858394623, 0.49083855748176575, -0.39141038060188293, -0.15796473622322083, 0.1868768334388733, -0.00967000238597393, -0.13670486211776733, -0.1370304822921753, -0.22997349500656128, -0.38902607560157776, -0.27632537484169006, -0.11392250657081604, -0.13009323179721832, 0.3114844858646393, 0.02537461556494236, 0.016825763508677483, 0.12226467579603195, 0.09885254502296448, 0.2510804235935211, 0.18979115784168243, 0.1321551501750946, 0.025374621152877808, 0.07778124511241913, 0.383108526468277, 0.5103029608726501, 0.036919619888067245, 0.6296113133430481, -0.010879415087401867, -0.037916794419288635, -0.08799190074205399, -0.08143600076436996, 0.027406811714172363, 0.550567090511322, 0.20019087195396423, 0.0538073368370533, 0.1374892145395279, 0.010431629605591297, -0.414275199174881, -0.040837857872247696, 0.11529874801635742, -0.08835182338953018, -0.4278702735900879, -0.29411065578460693, 0.31697896122932434, 0.23538480699062347, -0.13394224643707275, -0.13132084906101227, -0.13563509285449982, -0.2719821631908417, 0.24533551931381226, 0.055233705788850784, 1.0419281721115112, 0.20728415250778198, 0.04186470806598663, -0.05581317096948624, -0.14920100569725037, 0.9380418062210083, -0.300078421831131, 0.4176374673843384, -0.35171064734458923, -0.1780041605234146, -0.057503558695316315, -0.14746706187725067, 0.03603379800915718, 0.09087706357240677, -0.15406063199043274, 0.37495219707489014, 0.2888358533382416, 0.05044002830982208, -0.12973564863204956, 0.42349565029144287, -0.24959734082221985, -0.15231451392173767, -0.3030512034893036, 0.017057307064533234, -0.30228376388549805, 0.5400665402412415, -0.03814329579472542, -0.07119659334421158, -0.08948380500078201, -0.20423318445682526, -0.37206536531448364, 0.252585768699646, 0.024172693490982056, 0.12936724722385406, 0.23180124163627625, -0.1961919367313385, -0.07764165103435516, 0.10560780763626099, 0.4005056917667389, -0.16487476229667664, -0.1290959268808365, -0.006211291998624802, -0.14600776135921478, -0.12728780508041382, 0.11892896890640259, 0.09156037122011185, 0.28809744119644165, -0.09807223826646805, -0.030079998075962067, -0.01380288414657116, -0.14492270350456238, -0.3308664858341217, -0.2484138458967209, -0.05192253738641739, -0.16080942749977112, -0.22444908320903778, -0.014996140263974667, 0.1725158989429474, -0.028157642111182213, -0.13444571197032928, 0.020831534639000893, -0.047644905745983124, -0.16959379613399506, 0.11480879038572311, -0.09480107575654984, -0.3547864258289337, -0.18660888075828552, 0.4653010964393616, 0.41192227602005005, -0.13000524044036865, 0.48891645669937134, -0.10779165476560593, -0.047152452170848846, -0.16125038266181946, -0.18598589301109314, -0.262263685464859, -0.4254761338233948, 0.2058078646659851, -0.21055278182029724, -0.17630617320537567, 0.20774608850479126, 0.2605409026145935, 0.23746633529663086, 0.3127461373806, -0.013969340361654758, -0.48175349831581116, -0.03925470635294914, 0.13603469729423523, -0.011188521049916744, 0.09119043499231339, -0.02453990839421749, 0.3797798752784729, -0.24098186194896698, 0.3181023895740509, -0.21519352495670319, 0.0392879843711853, -0.3406299948692322, 0.3029211759567261, 0.19331961870193481, -0.13923437893390656, 0.06780225038528442, 0.012920859269797802, -0.14105574786663055, 0.26413580775260925, -0.10220276564359665, -0.017113177105784416, -0.14058798551559448, 0.19693763554096222, 0.15148590505123138, -0.11185362935066223, -0.15570734441280365, 0.079828642308712, 0.06451985239982605, 0.0160156711935997, 0.06329832226037979, 0.28377798199653625, 0.06752466410398483, 0.30502891540527344, -0.28709176182746887, -0.039535023272037506, 0.000517253065481782, 0.19909077882766724, -0.24377425014972687, 0.2209278643131256, -0.03934994339942932, 0.2127683013677597, -0.16374027729034424, 0.05937040224671364, -0.09963280707597733, 0.0008997877594083548, 0.08639179170131683, 0.08966068923473358, 0.08942961692810059, -0.25115707516670227, 0.03479785844683647, 0.2602267563343048, 0.053794976323843, 0.5116798877716064, -0.08315594494342804, -0.15430867671966553, 0.12484146654605865, 0.08711441606283188, -0.141017347574234, -0.06204913556575775, 0.3102421760559082, -0.0773712769150734, -0.09638036042451859, 0.3810332417488098, 0.173047736287117, -0.12971511483192444, -0.14074619114398956, 0.10686618089675903, 0.45320653915405273, -0.19162830710411072, 0.15098369121551514, 0.12681451439857483, -0.10557380318641663, -0.08674190193414688, 0.37435027956962585, -0.20107775926589966, 0.12214735895395279, 0.11047101765871048, 0.04505368322134018, 0.3720737397670746, 0.054727960377931595, 0.45312878489494324, -0.1651650220155716, -0.15896669030189514, -0.03961080685257912, 0.6585911512374878, 0.1879720687866211, 0.4303010404109955, 0.1421816647052765, 0.5052540898323059, -0.3357347249984741, -0.015216338448226452, -0.3209923505783081, 0.021359430626034737, -0.09274664521217346, -0.1665070652961731, -0.30164197087287903, -0.11755875498056412, -0.10489492863416672, 0.15447787940502167, -0.15096986293792725, 0.05504477396607399, 0.4208003580570221, -0.06759705394506454, -0.29463058710098267, -0.5726765394210815, -0.0037669860757887363, 0.055578216910362244, 0.2770497798919678, -0.22729730606079102, 0.2844376266002655, 0.07585018873214722, -0.20439910888671875, 0.3192937970161438, 0.6775819063186646, 0.5208286046981812, 0.23217296600341797, -0.03278952091932297, -0.07751891016960144, 0.07033353298902512, 0.12693901360034943, 0.2107742577791214, 0.7450392842292786, 0.29596900939941406, -0.3133506178855896, 0.27542489767074585, -0.03803376108407974, -0.11196880042552948, 0.3472534418106079, 0.025003891438245773, 0.07327745109796524, -0.03658033162355423, 0.38708585500717163, -0.1618288904428482, 0.0833641067147255, -0.06547283381223679, 0.09793011099100113, -0.32492464780807495, -0.21097297966480255, 0.16343837976455688, -0.172527015209198, 0.30173975229263306, -0.08095704764127731, 0.04415689408779144, -0.24738383293151855, 0.5227414965629578, 0.5025145411491394, -0.31866154074668884, -0.150527223944664, -0.026378676295280457, -0.5267682671546936, 0.29629501700401306, -0.3916662931442261, -0.08136048167943954, -0.20274877548217773, 0.44821131229400635, -0.19692012667655945, 0.2801702916622162, -0.17557309567928314, 0.41491878032684326, -0.07420191913843155, -0.018937231972813606, -0.23684760928153992, 0.07650040835142136, -0.031340643763542175, -0.19340305030345917, 0.07157577574253082, -0.24616380035877228, 0.2173757255077362, 0.15779562294483185, -0.10438935458660126, 0.03160826861858368, 0.10152042657136917, -0.01613362319767475, 0.390072226524353, -0.07530275732278824, 0.5313581228256226, 0.2922847270965576, -0.03974778950214386, -0.2839519679546356, -0.14522980153560638, -0.4025913178920746, -0.31096357107162476, 0.34221428632736206, 0.20279109477996826, 0.4863949120044708, -0.22397997975349426, 0.4722752273082733, -0.16622591018676758, 0.3110296428203583, 0.02751230262219906, -0.06190120056271553, -0.31948965787887573, -0.02634184993803501, -0.17405222356319427, 0.044957783073186874, 0.20531389117240906, 0.5501673221588135, 0.14749214053153992, 0.05282975733280182, -0.26446643471717834, -0.4141327738761902, 0.3889920115470886, -0.38547855615615845, 0.00392143102362752, 0.07363033294677734, 0.23610730469226837, 0.274517685174942, -0.20531943440437317, -0.7636080384254456, 0.04344790428876877, 0.22158879041671753, -0.13191282749176025, -0.00814608484506607, 0.06734839081764221, -0.188516765832901, -0.27262768149375916, -0.014860769733786583, -0.177165225148201, 0.36956942081451416, -0.40832605957984924, 0.4173520803451538, -0.019796645268797874 ]
https://github.com/huggingface/datasets/issues/215
NonMatchingSplitsSizesError when loading blog_authorship_corpus
Yes I think this might not be documented well enough. Let’s add it to the doc @lhoestq @SBrandeis. And everything on how to control the cache behavior better (removing, overriding, changing the path, etc)
Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on.
34
NonMatchingSplitsSizesError when loading blog_authorship_corpus Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on. Yes I think this might not be documented well enough. Let’s add it to the doc @lhoestq @SBrandeis. And everything on how to control the cache behavior better (removing, overriding, changing the path, etc)
[ -0.047863151878118515, 0.27813613414764404, 0.0776575431227684, 0.3915250599384308, -0.10355405509471893, 0.03640417009592056, -0.08268307894468307, 0.2459602653980255, -0.0835331454873085, 0.04618614912033081, 0.08917726576328278, 0.1745031774044037, 0.04060787707567215, -0.15940971672534943, 0.017909089103341103, 0.05279991030693054, -0.0644669234752655, 0.12483152002096176, 0.1059795394539833, 0.04289756342768669, -0.010563766583800316, 0.3961234986782074, -0.19377809762954712, -0.04041977971792221, -0.30533409118652344, -0.233677938580513, 0.029021982103586197, 0.23348209261894226, 0.04022783786058426, -0.4769922196865082, 0.24664737284183502, 0.056629035621881485, 0.10053170472383499, 0.1670154184103012, -0.00012176284508313984, -0.0632687583565712, 0.5389476418495178, -0.19484315812587738, -0.2518872916698456, -0.02978958934545517, -0.38360336422920227, -0.2797573506832123, 0.0038725154008716345, -0.2842758297920227, 0.01517886109650135, 0.023536739870905876, 0.17921555042266846, -0.16523925960063934, 0.2291368693113327, 0.30275893211364746, 0.12136092782020569, 0.10980737954378128, -0.18810977041721344, 0.16789405047893524, 0.17320458590984344, -0.027630388736724854, -0.06475881487131119, -0.15657763183116913, -0.11170952022075653, -0.28193074464797974, -0.24037209153175354, 0.45989346504211426, -0.23572349548339844, 0.15005958080291748, 0.26409223675727844, 0.08830901980400085, 0.32167693972587585, -0.08726507425308228, 0.013094069436192513, 0.39029133319854736, 0.586207926273346, -0.046833451837301254, -0.2849667966365814, -0.49856653809547424, -0.1296737641096115, 0.028640126809477806, 0.5057656764984131, 0.2795499861240387, -0.15703818202018738, 0.0537014864385128, -0.4670754075050354, -0.1777775138616562, -0.05261118710041046, -0.0008108822512440383, 0.3168705403804779, 0.19807390868663788, 0.12562760710716248, -0.05163551867008209, 0.030756013467907906, -0.010176761075854301, 0.2684493362903595, -0.18168959021568298, -0.12499304860830307, 0.08544939756393433, -0.2321663796901703, -0.20516344904899597, -0.02672516368329525, 0.09116481989622116, 0.2567504048347473, 0.21745680272579193, 0.4005259871482849, 0.005886316765099764, 0.2139706015586853, 0.1242160052061081, 0.33585792779922485, 0.20841912925243378, -0.014871146529912949, 0.39324358105659485, 0.15307891368865967, -0.002486474346369505, 0.13733002543449402, 0.05757516995072365, 0.35157960653305054, -0.20486567914485931, -0.011022728867828846, 0.20190498232841492, 0.15973465144634247, -0.48777610063552856, -0.3507693409919739, 0.1007281094789505, -0.3997795879840851, -0.2605063319206238, -0.05867178365588188, 0.24248559772968292, -0.37205642461776733, 0.3226315975189209, -0.269064337015152, -0.03352263197302818, -0.4499548077583313, -0.12790562212467194, -0.2333000749349594, 0.17981594800949097, -0.14685606956481934, 0.3382389545440674, 0.2391369640827179, -0.07944665104150772, 0.3943142294883728, -0.16760492324829102, -0.09943599253892899, -0.1319217085838318, 0.3565538823604584, -0.10957316309213638, 0.08933702111244202, 0.20857876539230347, -0.0928037017583847, 0.15524999797344208, -0.04847068339586258, -0.09950092434883118, -0.48516032099723816, 0.041017673909664154, -0.32748943567276, -0.6717565655708313, 0.09786254167556763, 0.11474655568599701, -0.3720287084579468, -0.049001552164554596, 0.1598808914422989, 0.30459141731262207, 0.5192115306854248, -0.379810631275177, 0.0015791594050824642, -0.08771204948425293, -0.19829106330871582, -0.15863658487796783, -0.1705338954925537, 0.5231297016143799, 0.07261832058429718, -0.306856244802475, 0.12426751106977463, 0.01171808410435915, 0.48262178897857666, 0.36976712942123413, -0.035665132105350494, 0.1533680409193039, -0.16418732702732086, 0.35587453842163086, 0.17119823396205902, -0.41636425256729126, -0.2768099308013916, 0.380389541387558, 0.10637632012367249, 0.27490147948265076, 0.24967393279075623, 0.12377031147480011, 0.2442072480916977, -0.1929769515991211, 0.0163564570248127, 0.42986124753952026, -0.06306622922420502, 0.09338262677192688, -0.5093788504600525, -0.41878798604011536, 0.43633052706718445, 0.11583733558654785, 0.06944510340690613, 0.03629525005817413, -0.0036747667472809553, 0.4742714464664459, 0.1634601205587387, -0.08560086786746979, 0.09851223975419998, 0.25446817278862, -0.34776726365089417, 0.13968198001384735, 0.08394356071949005, -0.07341723889112473, -0.49339497089385986, -0.029928816482424736, -0.27951493859291077, 0.17815463244915009, 0.1481744945049286, -0.3201978802680969, -0.0982881709933281, -0.47745218873023987, -0.09373929351568222, -0.1365208774805069, 0.029480576515197754, 0.20342375338077545, 0.5127826929092407, 0.08227599412202835, 0.019413966685533524, 0.5455960035324097, -0.006967963185161352, 0.1613590568304062, -0.8287144303321838, 0.06388108432292938, -0.13689163327217102, -0.1838446706533432, 0.3626120984554291, 0.35950547456741333, 0.21638256311416626, -0.05094444751739502, -0.0946408361196518, 0.5340723395347595, -0.14141760766506195, 0.4370483160018921, -0.22054420411586761, 0.1521023064851761, 0.24981950223445892, -0.20668917894363403, 0.11677256971597672, 0.21492594480514526, -0.1372481882572174, -0.0022527535911649466, -0.09489598125219345, 0.3331184685230255, -0.3746820092201233, 0.08620306104421616, 0.06745541095733643, -0.10661504417657852, 0.05747590586543083, -0.2557363510131836, -0.27272456884384155, -0.45996853709220886, 0.5414720177650452, -0.05310400575399399, -0.23666363954544067, 0.013625831343233585, -0.11991278827190399, 0.10137099772691727, 0.4620297849178314, 0.10436466336250305, -0.015319294296205044, -0.06299596279859543, 0.09480494260787964, -0.25104647874832153, 0.20877361297607422, 0.21671496331691742, 0.46867281198501587, 0.15137308835983276, 0.07372031360864639, 0.019000492990016937, -0.18847407400608063, -0.33689555525779724, 0.10088247805833817, -0.016211437061429024, -0.02424059621989727, 0.2759978473186493, 0.011330530978739262, -0.001432315562851727, -0.3033413589000702, 0.022108303382992744, 0.1912079006433487, 0.17614959180355072, -0.33251315355300903, -0.06196743994951248, -0.39287176728248596, -0.32350051403045654, -0.2495741844177246, -0.13010668754577637, -0.20251081883907318, -0.35886093974113464, 0.302448034286499, 0.16726437211036682, -0.2591959238052368, 0.1308683156967163, -0.25160038471221924, 0.18590442836284637, -0.276306688785553, -0.10526437312364578, -0.07306012511253357, -0.277865469455719, -0.27610263228416443, -0.06377263367176056, 0.11679238826036453, 0.31539809703826904, 0.31786832213401794, -0.41521814465522766, -0.21306519210338593, 0.04067404195666313, -0.3145895302295685, 0.0671834722161293, 0.059937406331300735, -0.10611604154109955, 0.10012239217758179, -0.21414683759212494, 0.37403467297554016, -0.22101689875125885, -0.05585498735308647, -0.2211209237575531, -0.252110093832016, -0.06747788190841675, 0.2016219198703766, 0.245502769947052, -0.19950589537620544, -0.5588034987449646, -0.3258545398712158, -0.36632636189460754, -0.05701996386051178, 0.10900750011205673, 0.29956376552581787, 0.08219227194786072, -0.21077017486095428, -0.21016067266464233, -0.18786467611789703, 0.31254035234451294, -0.3795470595359802, -0.16327343881130219, 0.3584213852882385, -0.1916775405406952, -0.07630955427885056, 0.044110219925642014, -0.04246676340699196, -0.07117635756731033, 0.12719465792179108, -0.5488823652267456, -0.17458409070968628, -0.11559607833623886, -0.04159959405660629, -0.21918228268623352, -0.22065365314483643, 0.6390054821968079, -0.044647011905908585, 0.04780681058764458, 0.0388735793530941, -0.361352801322937, 0.07012827694416046, 0.0347481444478035, 0.5672551393508911, -0.04162050411105156, 0.23606432974338531, 0.0665677934885025, 0.5585219264030457, 0.47468918561935425, 0.14414550364017487, -0.024808797985315323, 0.30200445652008057, 0.17004087567329407, 0.08520074933767319, 0.09485143423080444, 0.3581473231315613, -0.07355312258005142, -0.20108556747436523, 0.357281893491745, 0.1408153921365738, -0.14786845445632935, -0.08919988572597504, -0.04618241637945175, -0.2848125100135803, -0.26708975434303284, 0.1735619306564331, -0.3263130784034729, 0.20595712959766388, 0.2735859751701355, 0.18179230391979218, -0.16912174224853516, -0.47976386547088623, 0.05117829144001007, 0.32504335045814514, 0.25479111075401306, 0.2598884105682373, -0.4101261496543884, 0.18531866371631622, -0.5481981635093689, 0.16737167537212372, 0.12330997735261917, 0.2695996165275574, -0.1897914558649063, -0.1771593540906906, 0.04420514404773712, -0.151192307472229, 0.2935729920864105, -0.5013002157211304, -0.010341248475015163, -0.11806618422269821, -0.0182663481682539, -0.19106526672840118, -0.175484299659729, 0.017090367153286934, 0.30750223994255066, 0.4353831708431244, 0.31083598732948303, -0.4779394567012787, -0.07290174067020416, 0.11803160607814789, -0.08828973770141602, -0.14250555634498596, -0.22670301795005798, -0.17830847203731537, -0.2467595338821411, -0.17376427352428436, -0.14634308218955994, -0.06844274699687958, 0.20476171374320984, -0.0036699664779007435, 0.05763475224375725, 0.07807642221450806, 0.1945190280675888, 0.2609218657016754, 0.23615556955337524, 0.24638903141021729, -0.07365652918815613, 0.1055096983909607, 0.3441770672798157, 0.3810717463493347, 0.04580065608024597, 0.4809649586677551, -0.04737796261906624, -0.03335319831967354, 0.03335155174136162, -0.1582031548023224, -0.055853940546512604, 0.4643758237361908, 0.14981043338775635, 0.0582268089056015, 0.04612589254975319, -0.11693387478590012, -0.3775768578052521, 0.0007062402437441051, -0.05507654324173927, -0.1400311291217804, -0.2999066114425659, -0.4009567201137543, 0.2616017162799835, 0.34820184111595154, -0.1439933031797409, -0.1402902603149414, -0.17629551887512207, -0.3403507471084595, 0.2590385675430298, 0.2319076657295227, 1.0419527292251587, 0.2871277630329132, 0.07867734879255295, -0.1415203958749771, -0.19598744809627533, 0.8950438499450684, -0.4584694802761078, 0.38842901587486267, -0.3797580301761627, -0.16995297372341156, -0.049717847257852554, -0.13921774923801422, 0.06719143688678741, 0.07828176021575928, -0.3491450548171997, 0.44244804978370667, 0.2311290055513382, 0.06170284003019333, -0.14359508454799652, 0.3393721580505371, -0.24981428682804108, -0.18752306699752808, -0.19505459070205688, -0.00536996079608798, -0.23499362170696259, 0.34888187050819397, -0.009585735388100147, -0.08522585034370422, -0.059862349182367325, -0.18270397186279297, -0.3571866452693939, 0.11464311927556992, 0.09397128969430923, 0.012489796616137028, 0.23694106936454773, -0.20036770403385162, -0.11602877825498581, 0.09369528293609619, 0.366013765335083, -0.18719890713691711, -0.16472187638282776, -0.050345148891210556, -0.13756614923477173, -0.07333600521087646, 0.1151702031493187, 0.06339465081691742, 0.34322863817214966, -0.08296223729848862, 0.06117700785398483, 0.11704806983470917, -0.19855347275733948, -0.44435107707977295, -0.21357275545597076, -0.13070237636566162, -0.14821848273277283, -0.11806513369083405, -0.023201126605272293, 0.08216463774442673, -0.013900220394134521, -0.05310828611254692, 0.042126111686229706, 0.0036887519527226686, -0.15956872701644897, 0.3063158094882965, -0.2527935802936554, -0.2312992811203003, -0.16731923818588257, 0.38638824224472046, 0.3539518415927887, -0.022716103121638298, 0.383533775806427, -0.1494126170873642, -0.16333311796188354, -0.14232735335826874, -0.2202540934085846, -0.24535754323005676, -0.3556597828865051, 0.11342185735702515, -0.3243698477745056, -0.23751333355903625, 0.09655291587114334, 0.19885803759098053, 0.17593833804130554, 0.39818522334098816, -0.059355150908231735, -0.4329325258731842, -0.08512701839208603, 0.1618925780057907, -0.08562657237052917, 0.10745802521705627, -0.034591469913721085, 0.2953036427497864, -0.23416513204574585, 0.30818501114845276, -0.2241593450307846, -0.21442461013793945, -0.39406847953796387, 0.2727064788341522, 0.08486243337392807, -0.038590945303440094, 0.014369780197739601, 0.04511474072933197, -0.21619153022766113, 0.2673089802265167, -0.16531935334205627, 0.005659222602844238, -0.05170426517724991, 0.20624296367168427, 0.1315203458070755, -0.03179614618420601, -0.185196191072464, 0.1290428191423416, 0.0924130529165268, 0.04547353461384773, 0.028972050175070763, 0.2819375693798065, 0.040343593806028366, 0.40376541018486023, -0.25698044896125793, 0.084502674639225, -0.019484397023916245, 0.20341485738754272, -0.27178552746772766, 0.3247484862804413, -0.08645539730787277, 0.3065057396888733, -0.20353463292121887, 0.029181046411395073, -0.02076403982937336, 0.07677129656076431, 0.1272362470626831, 0.011338782496750355, 0.15322579443454742, -0.14263609051704407, 0.02795151062309742, 0.27671828866004944, -0.07625966519117355, 0.5579020977020264, -0.11840254813432693, -0.2716970145702362, 0.13532336056232452, 0.0732981488108635, -0.02431291714310646, -0.0656943991780281, 0.09087354689836502, 0.027588944882154465, -0.07582689821720123, 0.42469286918640137, 0.22875241935253143, -0.13601647317409515, -0.012512536719441414, 0.23883190751075745, 0.4896007776260376, -0.26902318000793457, 0.0784410759806633, 0.1427549123764038, -0.18673263490200043, 0.05787567049264908, 0.39533472061157227, -0.23366142809391022, 0.14236822724342346, 0.06332098692655563, 0.11303863674402237, 0.4625704884529114, 0.20257821679115295, 0.3750728964805603, -0.18873663246631622, -0.10802429914474487, -0.05696152150630951, 0.6232272982597351, 0.27606910467147827, 0.33051958680152893, 0.19891861081123352, 0.3161478638648987, -0.14343459904193878, 0.010834557935595512, -0.33690378069877625, -0.18920555710792542, -0.037813059985637665, -0.1468658298254013, -0.3757616877555847, -0.09626253694295883, -0.08604671061038971, 0.0835808515548706, -0.13657604157924652, 0.1039004996418953, 0.5570073127746582, -0.17096593976020813, -0.1825265735387802, -0.5047014951705933, 0.2223152369260788, 0.003293423680588603, 0.2047148048877716, -0.22498634457588196, 0.3006284534931183, -0.04419228434562683, -0.20166677236557007, 0.2878987789154053, 0.523901641368866, 0.4250335991382599, 0.29359620809555054, -0.04037944972515106, -0.17957104742527008, 0.07012111693620682, 0.1715543121099472, 0.1945973038673401, 0.7336784601211548, 0.22640864551067352, -0.36697831749916077, 0.27984827756881714, -0.07750452309846878, -0.04813354089856148, 0.4749352037906647, 0.010236678645014763, 0.032687585800886154, -0.027301542460918427, 0.4250287711620331, 0.01057647168636322, 0.11824992299079895, 0.1175379827618599, 0.1492505818605423, -0.22901619970798492, -0.28582504391670227, 0.13254649937152863, -0.17832864820957184, 0.287056565284729, -0.1423877775669098, 0.04563356563448906, -0.20899400115013123, 0.6096136569976807, 0.457562118768692, -0.23569628596305847, -0.22168295085430145, 0.0019245706498622894, -0.5511800050735474, 0.33748605847358704, -0.361329585313797, -0.20019222795963287, -0.17516082525253296, 0.5507525205612183, -0.34127411246299744, 0.3287827670574188, -0.3048568665981293, 0.31894829869270325, -0.03687421604990959, 0.03029712848365307, -0.2732526361942291, -0.09151332080364227, -0.08998378366231918, -0.14913113415241241, 0.0885564312338829, -0.2684875428676605, 0.25103873014450073, 0.23709242045879364, -0.10101079195737839, -0.020665530115365982, -0.02704506926238537, 0.17944811284542084, 0.5358684062957764, -0.02898741327226162, 0.5525880455970764, 0.21866686642169952, -0.11766799539327621, -0.17617957293987274, -0.215385302901268, -0.24386994540691376, -0.30985110998153687, 0.15333831310272217, 0.13062886893749237, 0.30935123562812805, -0.27565696835517883, 0.49835988879203796, -0.14632421731948853, 0.41652798652648926, -0.01682974584400654, 0.008334715850651264, -0.23042751848697662, -0.0479775108397007, -0.09355030208826065, 0.12309963256120682, 0.18964484333992004, 0.4126790761947632, 0.17846421897411346, 0.1410260945558548, -0.1901102215051651, -0.3322606086730957, 0.3096320927143097, -0.3596901297569275, 0.08010423183441162, -0.035427115857601166, 0.26248687505722046, 0.27984482049942017, -0.24873392283916473, -0.7318597435951233, -0.0688432827591896, 0.2349819391965866, -0.12555186450481415, 0.06184467673301697, 0.04573842138051987, -0.11327055841684341, -0.20846401154994965, -0.09276966750621796, -0.1714310348033905, 0.28795892000198364, -0.3046324551105499, 0.34687158465385437, -0.15724092721939087 ]
https://github.com/huggingface/datasets/issues/211
[Arrow writer, Trivia_qa] Could not convert TagMe with type str: converting to null type
Here the full error trace: ``` ArrowInvalid Traceback (most recent call last) <ipython-input-1-7aaf3f011358> in <module> 1 import nlp 2 ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ----> 3 ds.map(lambda x: x, load_from_cache_file=False) ~/python_bin/nlp/arrow_dataset.py in map(self, function, with_indices, batched, batch_size, remove_columns, keep_in_memory, load_from_cache_file, cache_file_name, writer_batch_size, arrow_schema, disable_nullable) 549 550 if update_data: --> 551 writer.finalize() # close_stream=bool(buf_writer is None)) # We only close if we are writing in a file 552 553 # Create new Dataset from buffer or file ~/python_bin/nlp/arrow_writer.py in finalize(self, close_stream) 182 def finalize(self, close_stream=True): 183 if self.pa_writer is not None: --> 184 self.write_on_file() 185 self.pa_writer.close() 186 if close_stream: ~/python_bin/nlp/arrow_writer.py in write_on_file(self) 104 """ 105 if self.current_rows: --> 106 pa_array = pa.array(self.current_rows, type=self._type) 107 first_example = pa.array(self.current_rows[0:1], type=self._type)[0] 108 # Sanity check ~/hugging_face/venv_3.7/lib/python3.7/site-packages/pyarrow/array.pxi in pyarrow.lib.array() ~/hugging_face/venv_3.7/lib/python3.7/site-packages/pyarrow/array.pxi in pyarrow.lib._sequence_to_array() ~/hugging_face/venv_3.7/lib/python3.7/site-packages/pyarrow/error.pxi in pyarrow.lib.check_status() ArrowInvalid: Could not convert TagMe with type str: converting to null type ```
Running the following code ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, load_from_cache_file=False) ``` triggers a `ArrowInvalid: Could not convert TagMe with type str: converting to null type` error. On the other hand if we remove a certain column of `trivia_qa` which seems responsible for the bug, it works: ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, remove_columns=["entity_pages"], load_from_cache_file=False) ``` . Seems quite hard to debug what's going on here... @lhoestq @thomwolf - do you have a good first guess what the problem could be? **Note** BTW: I think this could be a good test to check that the datasets work correctly: Take a tiny portion of the dataset and check that it can be written correctly.
155
[Arrow writer, Trivia_qa] Could not convert TagMe with type str: converting to null type Running the following code ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, load_from_cache_file=False) ``` triggers a `ArrowInvalid: Could not convert TagMe with type str: converting to null type` error. On the other hand if we remove a certain column of `trivia_qa` which seems responsible for the bug, it works: ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, remove_columns=["entity_pages"], load_from_cache_file=False) ``` . Seems quite hard to debug what's going on here... @lhoestq @thomwolf - do you have a good first guess what the problem could be? **Note** BTW: I think this could be a good test to check that the datasets work correctly: Take a tiny portion of the dataset and check that it can be written correctly. Here the full error trace: ``` ArrowInvalid Traceback (most recent call last) <ipython-input-1-7aaf3f011358> in <module> 1 import nlp 2 ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ----> 3 ds.map(lambda x: x, load_from_cache_file=False) ~/python_bin/nlp/arrow_dataset.py in map(self, function, with_indices, batched, batch_size, remove_columns, keep_in_memory, load_from_cache_file, cache_file_name, writer_batch_size, arrow_schema, disable_nullable) 549 550 if update_data: --> 551 writer.finalize() # close_stream=bool(buf_writer is None)) # We only close if we are writing in a file 552 553 # Create new Dataset from buffer or file ~/python_bin/nlp/arrow_writer.py in finalize(self, close_stream) 182 def finalize(self, close_stream=True): 183 if self.pa_writer is not None: --> 184 self.write_on_file() 185 self.pa_writer.close() 186 if close_stream: ~/python_bin/nlp/arrow_writer.py in write_on_file(self) 104 """ 105 if self.current_rows: --> 106 pa_array = pa.array(self.current_rows, type=self._type) 107 first_example = pa.array(self.current_rows[0:1], type=self._type)[0] 108 # Sanity check ~/hugging_face/venv_3.7/lib/python3.7/site-packages/pyarrow/array.pxi in pyarrow.lib.array() ~/hugging_face/venv_3.7/lib/python3.7/site-packages/pyarrow/array.pxi in pyarrow.lib._sequence_to_array() ~/hugging_face/venv_3.7/lib/python3.7/site-packages/pyarrow/error.pxi in pyarrow.lib.check_status() ArrowInvalid: Could not convert TagMe with type str: converting to null type ```
[ 0.04782375693321228, -0.014994163997471333, 0.09996074438095093, 0.4827658534049988, 0.25262361764907837, 0.11827056854963303, 0.12425896525382996, 0.46299809217453003, 0.20459362864494324, -0.11430183053016663, 0.19549207389354706, 0.5628964900970459, -0.11229002475738525, -0.21451012790203094, -0.0685441866517067, -0.18655332922935486, 0.11651556193828583, 0.2422814518213272, -0.003950863610953093, -0.019452542066574097, -0.12144756317138672, 0.12436140328645706, -0.264439195394516, 0.1609533578157425, -0.4509817659854889, -0.13436345756053925, 0.10060359537601471, 0.0705767348408699, -0.16652897000312805, -0.45291194319725037, 0.13311581313610077, -0.38631612062454224, -0.014493754133582115, 0.26769036054611206, -0.0001202561252284795, -0.10239160060882568, 0.3526800274848938, 0.048439547419548035, -0.2610155940055847, -0.24791744351387024, -0.4819882810115814, 0.09865797311067581, 0.30856263637542725, -0.05106065049767494, 0.2251928597688675, 0.02437903918325901, 0.02387930639088154, -0.597424328327179, 0.18228361010551453, 0.5890494585037231, 0.18119469285011292, 0.31175026297569275, -0.05861680209636688, 0.08240789920091629, 0.227017343044281, 0.08643096685409546, -0.3474985957145691, 0.15860100090503693, -0.041864488273859024, -0.16374993324279785, -0.09725948423147202, 0.42489689588546753, -0.17378707230091095, 0.12876737117767334, 0.6042652130126953, 0.10137573629617691, 0.10312166064977646, -0.30099624395370483, 0.18946464359760284, 0.00226260325871408, 0.27389949560165405, -0.1859186589717865, -0.08067543059587479, 0.030914420261979103, -0.14413070678710938, -0.3694719970226288, 0.20961213111877441, 0.17653004825115204, -0.3138512074947357, 0.01766207069158554, 0.011314960196614265, 0.08406312763690948, -0.0632316842675209, -0.07367515563964844, -0.05972418561577797, 0.06742032617330551, 0.013082322664558887, 0.11426404118537903, -0.05431681498885155, -0.18600478768348694, -0.1897279918193817, -0.22003136575222015, -0.20488370954990387, 0.24149301648139954, -0.3378821909427643, 0.048407748341560364, -0.05792529135942459, -0.20891311764717102, -0.16488344967365265, -0.11893334239721298, 0.2784719169139862, -0.00025343341985717416, 0.14469842612743378, 0.129641592502594, 0.33332526683807373, 0.28613632917404175, 0.32042112946510315, 0.24878492951393127, 0.0986194834113121, -0.30730584263801575, -0.18543514609336853, -0.01363749336451292, -0.011412332765758038, -0.009702254086732864, 0.5141094923019409, 0.016387145966291428, 0.3002519905567169, -0.36010658740997314, -0.354993999004364, -0.011048904620110989, -0.7358065843582153, 0.15996505320072174, -0.1690237671136856, 0.10665487498044968, -0.008252542465925217, 0.4835475981235504, 0.24768398702144623, 0.2814030349254608, -0.3391993045806885, -0.19262802600860596, -0.1142076849937439, -0.004297515377402306, -0.2985196113586426, 0.015622125007212162, 0.3052986264228821, -0.036364950239658356, 0.05392255634069443, 0.0657089501619339, -0.4077068865299225, -0.20890912413597107, 0.37390899658203125, -0.42265230417251587, 0.17886844277381897, -0.008708421140909195, -0.32873791456222534, 0.41525107622146606, 0.2411910444498062, -0.40890243649482727, -0.21291378140449524, 0.1715904176235199, -0.1304960399866104, 0.03689861297607422, -0.19280901551246643, 0.12478355318307877, 0.15921320021152496, -0.2186262458562851, -0.1702643185853958, -0.06532536447048187, 0.3925430178642273, -0.10832763463258743, 0.1037176251411438, -0.29295575618743896, -0.040322497487068176, -0.24107493460178375, -0.08629407733678818, 0.4788670837879181, -0.2231428623199463, 0.06661075353622437, -0.055577926337718964, 0.18355421721935272, 0.36413973569869995, 0.37810957431793213, -0.02037329226732254, 0.17711471021175385, 0.10718746483325958, 0.28203466534614563, 0.4739590585231781, -0.30076491832733154, -0.1000545471906662, 0.09920648485422134, -0.028821958228945732, -0.007086336147040129, -0.041806660592556, 0.12140262871980667, 0.07706907391548157, -0.12026631087064743, 0.16744688153266907, 0.19477051496505737, -0.06231988966464996, 0.2353254109621048, -0.4206894636154175, -0.10153141617774963, 0.5093362927436829, -0.17420139908790588, -0.075660839676857, 0.026094915345311165, -0.07687550783157349, 0.050202518701553345, -0.06762714684009552, -0.013059867545962334, 0.4281768202781677, 0.2600766122341156, -0.09044022113084793, -0.08258424699306488, 0.08325596153736115, -0.1882932335138321, -0.3873162865638733, -0.05201730132102966, -0.10536365211009979, 0.13355955481529236, -0.1724088490009308, -0.040386419743299484, -0.34554463624954224, 0.07457039505243301, -0.027131395414471626, -0.19891300797462463, 0.12534494698047638, 0.108247771859169, 0.15891793370246887, 0.05802774429321289, -0.22786886990070343, 0.12510867416858673, 0.004824083298444748, 0.12739749252796173, -0.39792555570602417, -0.14816635847091675, -0.08023658394813538, -0.35545381903648376, -0.0991663783788681, 0.17995978891849518, 0.23762942850589752, 0.16755422949790955, -0.26986369490623474, 0.18731644749641418, 0.018680358305573463, 0.21714353561401367, -0.25498166680336, 0.02990170754492283, 0.018767550587654114, -0.40214958786964417, 0.22020117938518524, 0.3269735872745514, 0.3616490960121155, -0.24359045922756195, -0.3964942395687103, -0.0214628167450428, 0.014600375667214394, 0.5858559608459473, -0.13255614042282104, 0.10853394865989685, -0.07677560299634933, -0.036233074963092804, 0.09410513192415237, -0.07976371049880981, 0.09601082652807236, -0.1133069321513176, 0.19795307517051697, 0.2924371063709259, -0.30972379446029663, 0.27264919877052307, 0.4216947555541992, 0.30024999380111694, 0.1342768520116806, 0.15981072187423706, -0.33452165126800537, -0.08566450327634811, 0.26441681385040283, -0.012384661473333836, 0.4610905051231384, -0.043286118656396866, 0.2145611047744751, -0.08492746949195862, -0.15070249140262604, -0.14859634637832642, 0.35783982276916504, -0.09661856293678284, 0.017265560105443, -0.009265835396945477, 0.4907107651233673, 0.010625896044075489, -0.17871424555778503, 0.2008916735649109, 0.005513680167496204, 0.23830631375312805, -0.35528361797332764, 0.32903772592544556, -0.3391128182411194, -0.16789540648460388, -0.3810410797595978, -0.2544398605823517, -0.01850913278758526, -0.5028195381164551, -0.1405358612537384, 0.07923328876495361, 0.006111367139965296, 0.1203894391655922, -0.25411662459373474, -0.1418566107749939, 0.018085014075040817, -0.11071254312992096, -0.3347357511520386, -0.35406163334846497, -0.37050187587738037, -0.06328770518302917, 0.46462786197662354, 0.3231956362724304, 0.18738147616386414, 0.05303458124399185, -0.06295030564069748, -0.06901048123836517, -0.360387921333313, 0.030563365668058395, -0.19892269372940063, 0.057185832411050797, 0.1203935295343399, 0.07613984495401382, -0.13603056967258453, -0.12729652225971222, 0.42562049627304077, 0.003183178137987852, -0.24060207605361938, 0.32919102907180786, -0.02077125944197178, 0.04902206361293793, -0.2927950322628021, -0.48070743680000305, 0.013537724502384663, -0.46828681230545044, 0.1370299905538559, -0.11211814731359482, 0.20931914448738098, 0.1776324361562729, 0.12377558648586273, 0.3106846511363983, -0.05574415624141693, 0.18512533605098724, 0.20702756941318512, 0.36536705493927, 0.6162765026092529, -0.1331360787153244, -0.23794686794281006, 0.208466038107872, -0.07696899026632309, 0.3138103783130646, 0.20748859643936157, -0.17882099747657776, 0.33277013897895813, -0.42187827825546265, 0.3207610547542572, -0.1633075773715973, -0.13050557672977448, 0.22173579037189484, 0.22102083265781403, -0.04805256426334381, -0.06043842434883118, -0.00751712080091238, 0.2183125764131546, -0.21776562929153442, 0.15179342031478882, 0.12625296413898468, 0.2142091542482376, -0.053450003266334534, 0.3817191421985626, 0.14399771392345428, -0.22840829193592072, 0.2830311954021454, 0.016808755695819855, -0.04759531468153, -0.3022206127643585, -0.2784264385700226, -0.08813589811325073, -0.211570143699646, 0.12770849466323853, -0.17738918960094452, 0.03394472971558571, -0.4855807423591614, -0.02226489968597889, 0.2889745533466339, -0.18293249607086182, -0.13431291282176971, 0.3022048771381378, -0.21380212903022766, 0.29014042019844055, 0.13460984826087952, 0.12581974267959595, -0.1979224532842636, -0.4124388098716736, 0.0329362228512764, -0.024656573310494423, 0.15686160326004028, 0.024755064398050308, -0.10193711519241333, -0.052897583693265915, -0.4739769995212555, 0.2884024381637573, 0.1751553863286972, 0.43267539143562317, -0.07643677294254303, -0.14325213432312012, 0.19170716404914856, -0.11351989209651947, 0.298640638589859, -0.14612405002117157, -0.027595676481723785, 0.3758090138435364, 0.2723758816719055, -0.4777472913265228, 0.2693086266517639, -0.23111553490161896, 0.3493029475212097, 0.26166707277297974, 0.28345468640327454, -0.3896854817867279, -0.31311285495758057, 0.25790244340896606, 0.1610284447669983, 0.016652440652251244, 0.08997327089309692, -0.3169792890548706, -0.0035179525148123503, -0.15553443133831024, -0.08673945814371109, 0.2757052481174469, 0.31060150265693665, 0.08262249082326889, 0.054902222007513046, -0.33018678426742554, 0.17894931137561798, 0.18064317107200623, 0.026372622698545456, 0.42826685309410095, -0.3036711812019348, 0.13769066333770752, -0.15683092176914215, 0.208980530500412, -0.08677136152982712, 0.32404008507728577, -0.06348980218172073, -0.19840160012245178, 0.2668812572956085, -0.057536520063877106, 0.3644375801086426, 0.32926705479621887, 0.036092936992645264, 0.3470534086227417, -0.12237449735403061, -0.09592227637767792, -0.20330126583576202, 0.02960449457168579, 0.1675887256860733, -0.0528097040951252, 0.04342694953083992, -0.21204352378845215, 0.48226696252822876, 0.29814910888671875, -0.0335705392062664, 0.6272779107093811, 0.13703176379203796, -0.2833764851093292, 0.23740047216415405, 0.18653222918510437, 1.18215811252594, 0.15122564136981964, 0.14966821670532227, 0.13378271460533142, -0.43013617396354675, 0.3339077830314636, 0.06193780526518822, 0.32545989751815796, -0.4628361165523529, 0.32495805621147156, 0.02449791505932808, -0.16893230378627777, 0.06731438636779785, -0.033896107226610184, -0.3024044334888458, 0.08304446190595627, -0.19602952897548676, 0.3092469871044159, 0.11238639056682587, 0.058070097118616104, -0.3026755750179291, -0.31657904386520386, -0.37491902709007263, 0.080600306391716, -0.250549852848053, -0.0766257643699646, 0.11036068946123123, -0.21556638181209564, -0.49149376153945923, -0.04550549015402794, -0.06853924691677094, 0.11479679495096207, -0.17067201435565948, 0.046522412449121475, 0.23672977089881897, -0.1576872318983078, 0.18111805617809296, 0.07256621867418289, -0.10370845347642899, -0.06586966663599014, -0.13383430242538452, -0.08608843386173248, 0.1967156082391739, 0.10096566379070282, 0.12061937153339386, -0.07572042942047119, 0.011201257817447186, 0.07623374462127686, 0.024822168052196503, 0.22761733829975128, -0.36800864338874817, -0.013540858402848244, -0.06473197042942047, 0.16641655564308167, 0.1192656010389328, -0.09275105595588684, -0.08846437931060791, 0.09907583147287369, -0.0605180561542511, 0.0031135764438658953, 0.0932030975818634, 0.025668328627943993, 0.15167877078056335, 0.08057030290365219, -0.15691064298152924, -0.26000797748565674, 0.2690161168575287, 0.3413315713405609, 0.04775894805788994, 0.24060384929180145, 0.7169964909553528, 0.059508662670850754, -0.18554100394248962, -0.17794471979141235, 0.07761194556951523, -0.052394162863492966, -0.7310895919799805, 0.32937875390052795, -0.331417977809906, -0.22107844054698944, 0.034033603966236115, 0.45446813106536865, -0.2029375433921814, 0.08033878356218338, -0.15941329300403595, -0.2330784648656845, -0.2189442664384842, 0.17904920876026154, -0.03060038760304451, 0.034671567380428314, -0.45244935154914856, -0.05752820149064064, 0.09664421528577805, 0.1214938759803772, -0.22719307243824005, -0.010041228495538235, -0.17429250478744507, 0.3713478446006775, 0.09213253110647202, 0.032965123653411865, -0.05073273554444313, -0.18051062524318695, -0.01144365593791008, 0.5151128172874451, -0.09715480357408524, -0.1443876028060913, -0.2621203064918518, 0.1471552550792694, 0.07113328576087952, -0.16618278622627258, 0.07055386155843735, 0.024172445759177208, 0.07843555510044098, -0.2128690928220749, -0.151078462600708, -0.10288633406162262, -0.1862981617450714, 0.056038763374090195, -0.08770083636045456, 0.20297840237617493, -0.17871826887130737, 0.09925492852926254, -0.14713293313980103, 0.056820690631866455, -0.2434588074684143, 0.10793520510196686, 0.21070456504821777, 0.05712320655584335, -0.20819628238677979, 0.048778582364320755, -0.09463757276535034, -0.1588643491268158, 0.30473512411117554, -0.37726378440856934, -0.1661875993013382, 0.0565507747232914, -0.19067248702049255, 0.4166315495967865, -0.09205659478902817, 0.0020526836160570383, -0.08149801939725876, 0.0977553203701973, 0.07033456116914749, 0.31662604212760925, -0.17869526147842407, -0.16311267018318176, -0.11346375197172165, 0.19117142260074615, -0.023984557017683983, -0.1549673080444336, -0.03029157593846321, 0.09488939493894577, 0.03304680809378624, -0.4421163499355316, -0.1640719175338745, 0.1360546499490738, -0.025478294119238853, -0.07160701602697372, 0.3569360375404358, -0.1045391857624054, 0.021537553519010544, 0.020303167402744293, 0.3664785325527191, 0.33404630422592163, 0.4364636242389679, 0.2034333050251007, -0.06764207035303116, 0.11950613558292389, -0.02497176267206669, 0.38919946551322937, 0.2442488670349121, 0.15110105276107788, 0.17533648014068604, 0.5244448781013489, -0.29807913303375244, -0.3281247317790985, 0.06573379039764404, 0.10832750797271729, -0.27272358536720276, -0.08761119842529297, -0.5703163743019104, -0.11522719264030457, -0.44190776348114014, -0.08809895813465118, -0.19053824245929718, 0.25448861718177795, 0.6928650140762329, 0.2337900847196579, 0.016697721555829048, -0.5845620632171631, -0.12988415360450745, 0.13295428454875946, 0.3920842409133911, -0.24886007606983185, 0.4035942554473877, 0.10868507623672485, -0.07600261270999908, -0.13055169582366943, 0.49973493814468384, 0.46834543347358704, 0.5678131580352783, -0.288379043340683, -0.06765243411064148, 0.05808202549815178, -0.18258124589920044, 0.23777781426906586, 0.5766584873199463, 0.33824190497398376, 0.14018091559410095, 0.03504405543208122, -0.037965841591358185, -0.10737576335668564, -0.09008068591356277, -0.029721863567829132, 0.2774466574192047, -0.16376079618930817, 0.2949461340904236, -0.23841078579425812, -0.2226382941007614, -0.04625173285603523, -0.2109307497739792, -0.3650144040584564, -0.0462195910513401, 0.2055170238018036, 0.04957457631826401, 0.015668893232941628, -0.33418136835098267, 0.0574212446808815, 0.2021239995956421, 0.5171065330505371, 0.27114227414131165, 0.1386128067970276, -0.4753839075565338, -0.11695356667041779, -0.7315419912338257, 0.17639389634132385, -0.31489959359169006, -0.15671460330486298, -0.01451841089874506, 0.36327826976776123, 0.04523323103785515, 0.31469908356666565, -0.034921567887067795, 0.22894978523254395, 0.25130078196525574, 0.3053605258464813, -0.3416893184185028, -0.11979447305202484, 0.3530436158180237, -0.3037296533584595, 0.023086247965693474, -0.5888960361480713, 0.26860204339027405, 0.051242098212242126, 0.06511188298463821, -0.008692285977303982, -0.2844811677932739, 0.2985386252403259, 0.4892457127571106, 0.2755069136619568, 0.263793021440506, 0.28560444712638855, 0.24758198857307434, -0.5557771325111389, -0.2176065444946289, -0.17345596849918365, -0.20772762596607208, -0.02030346915125847, 0.0898684486746788, 0.5805656909942627, -0.14044833183288574, 0.062437016516923904, -0.3035687804222107, 0.32777053117752075, -0.22651752829551697, 0.034378815442323685, -0.17557349801063538, 0.06700634211301804, -0.3490563929080963, 0.16302484273910522, 0.07218926399946213, 0.12040093541145325, 0.1978321224451065, 0.24254368245601654, -0.534783661365509, -0.387428879737854, 0.3043855130672455, -0.519408106803894, -0.44668930768966675, 0.29319116473197937, 0.14298847317695618, 0.44685235619544983, -0.012415866367518902, -0.7071860432624817, -0.26888132095336914, 0.21321970224380493, -0.09643765538930893, 0.025331169366836548, 0.023004839196801186, 0.026594702154397964, 0.07402858883142471, -0.19160300493240356, 0.3619839549064636, 0.008301920257508755, 0.049280595034360886, 0.048330262303352356, -0.2284112572669983 ]
https://github.com/huggingface/datasets/issues/211
[Arrow writer, Trivia_qa] Could not convert TagMe with type str: converting to null type
Actually thinking a bit more about it, it's probably a data sample that is not correct in `trivia_qa`. But I'm a bit surprised though that we managed to write it in .arrow format and now cannot write it anymore after an "identity" mapping.
Running the following code ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, load_from_cache_file=False) ``` triggers a `ArrowInvalid: Could not convert TagMe with type str: converting to null type` error. On the other hand if we remove a certain column of `trivia_qa` which seems responsible for the bug, it works: ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, remove_columns=["entity_pages"], load_from_cache_file=False) ``` . Seems quite hard to debug what's going on here... @lhoestq @thomwolf - do you have a good first guess what the problem could be? **Note** BTW: I think this could be a good test to check that the datasets work correctly: Take a tiny portion of the dataset and check that it can be written correctly.
43
[Arrow writer, Trivia_qa] Could not convert TagMe with type str: converting to null type Running the following code ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, load_from_cache_file=False) ``` triggers a `ArrowInvalid: Could not convert TagMe with type str: converting to null type` error. On the other hand if we remove a certain column of `trivia_qa` which seems responsible for the bug, it works: ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, remove_columns=["entity_pages"], load_from_cache_file=False) ``` . Seems quite hard to debug what's going on here... @lhoestq @thomwolf - do you have a good first guess what the problem could be? **Note** BTW: I think this could be a good test to check that the datasets work correctly: Take a tiny portion of the dataset and check that it can be written correctly. Actually thinking a bit more about it, it's probably a data sample that is not correct in `trivia_qa`. But I'm a bit surprised though that we managed to write it in .arrow format and now cannot write it anymore after an "identity" mapping.
[ 0.12082352489233017, 0.08130035549402237, 0.08892334997653961, 0.4082977771759033, 0.28144127130508423, 0.0546015202999115, 0.11376441270112991, 0.33654868602752686, 0.1860097348690033, -0.18694984912872314, 0.2484581172466278, 0.5634831786155701, 0.0009859593119472265, -0.07831809669733047, -0.02355879545211792, -0.11811408400535583, 0.07306340336799622, 0.2508920133113861, 0.05063005909323692, -0.1244533509016037, -0.09246131777763367, 0.16164346039295197, -0.18193930387496948, 0.22253011167049408, -0.513098418712616, -0.15872855484485626, 0.14412806928157806, 0.08026697486639023, -0.11963732540607452, -0.499011754989624, 0.0155159467831254, -0.3473427891731262, 0.030818453058600426, 0.2083606868982315, -0.0001191395494970493, -0.1659441441297531, 0.3734928369522095, -0.05221517011523247, -0.19603218138217926, -0.2107965648174286, -0.5531385540962219, 0.1664821058511734, 0.30328214168548584, -0.06724139302968979, 0.23992852866649628, -0.10213465243577957, 0.06580609083175659, -0.5268747210502625, 0.09618153423070908, 0.49537715315818787, 0.19496838748455048, 0.15002742409706116, 0.11893009394407272, 0.08554011583328247, 0.2670641541481018, 0.11917208135128021, -0.28745394945144653, 0.04326258972287178, -0.1304044872522354, -0.2241954356431961, -0.06938862800598145, 0.3475061058998108, -0.049195729196071625, -0.004544713068753481, 0.5662901401519775, 0.0445675402879715, 0.19759446382522583, -0.18642768263816833, 0.2564820945262909, 0.058506473898887634, 0.35495641827583313, -0.061708129942417145, -0.10864070057868958, 0.09885405004024506, -0.1233237162232399, -0.36338183283805847, 0.28847530484199524, 0.22252129018306732, -0.2659066319465637, -0.012292015366256237, 0.11637628823518753, 0.06312450766563416, 0.009714615531265736, -0.12463594973087311, -0.15653839707374573, 0.14945432543754578, 0.018611319363117218, 0.11266766488552094, -0.050113495439291, -0.1976822018623352, -0.18047213554382324, -0.07938511669635773, -0.30318939685821533, 0.21402011811733246, -0.2830837368965149, 0.011638865806162357, -0.05160348117351532, -0.16387344896793365, -0.20187877118587494, -0.08790993690490723, 0.41374650597572327, -0.00014435498451348394, -0.005353535059839487, 0.0675487369298935, 0.31440818309783936, 0.4288605749607086, 0.4697207510471344, 0.09827294945716858, 0.13660788536071777, -0.3048892915248871, -0.18265220522880554, -0.034475892782211304, -0.025242183357477188, -0.009739777073264122, 0.43271559476852417, 0.024083439260721207, 0.07819689065217972, -0.3047504127025604, -0.2590416967868805, 0.02684135176241398, -0.711564302444458, 0.22196853160858154, -0.21643337607383728, 0.13746269047260284, 0.06488624960184097, 0.4073728322982788, 0.29524320363998413, 0.2960967719554901, -0.20428569614887238, -0.20120763778686523, -0.11234593391418457, -0.09873455762863159, -0.3487420082092285, -0.15234257280826569, 0.19270986318588257, -0.08550103008747101, 0.08561725169420242, 0.021654454991221428, -0.5190787315368652, -0.22347389161586761, 0.29673415422439575, -0.28556084632873535, 0.29618263244628906, 0.0604485385119915, -0.2745223939418793, 0.4591625928878784, 0.21227796375751495, -0.5164347290992737, -0.09735027700662613, 0.09279680997133255, -0.20460042357444763, 0.07067390531301498, -0.18194662034511566, 0.13979122042655945, 0.03730914369225502, -0.27449139952659607, -0.20205269753932953, -0.06351874023675919, 0.38846516609191895, -0.16859513521194458, 0.09392968565225601, -0.20465373992919922, -0.012687630951404572, -0.28410252928733826, -0.1489119529724121, 0.3834082782268524, -0.18657854199409485, 0.10843648761510849, -0.07518047839403152, 0.1669817715883255, 0.3044075667858124, 0.3455914556980133, 0.022883182391524315, 0.19979698956012726, 0.10551311075687408, 0.35004332661628723, 0.39273667335510254, -0.22957998514175415, -0.009729161858558655, -0.0022538744378834963, -0.017602117732167244, -0.00765859754756093, -0.05828201398253441, -0.0028021743055433035, 0.04202081263065338, -0.11749909073114395, 0.08269582688808441, 0.12329769879579544, 0.012098685838282108, 0.2591467499732971, -0.3936404287815094, -0.062193334102630615, 0.4800756871700287, -0.31273186206817627, -0.2060410976409912, -0.030394699424505234, 0.016499483957886696, 0.10142763704061508, -0.026429852470755577, -0.09374630451202393, 0.3080482482910156, 0.1747930645942688, -0.07066237181425095, -0.1480058878660202, 0.15718434751033783, -0.04129156842827797, -0.39961451292037964, -0.05961420759558678, -0.11000032722949982, 0.1341492235660553, -0.05115344375371933, 0.002953039947897196, -0.2747969627380371, 0.05228047072887421, -0.00779563607648015, -0.2527810335159302, 0.1424804925918579, 0.06316851824522018, 0.16464950144290924, 0.12633100152015686, -0.2156062126159668, 0.06162474304437637, 0.021692680194973946, 0.12927861511707306, -0.3672279715538025, -0.2625301480293274, 0.010546890087425709, -0.3314332365989685, -0.19128243625164032, 0.23127081990242004, 0.22903786599636078, 0.1343214064836502, -0.2543696463108063, 0.192537322640419, -0.009426331147551537, 0.3920420706272125, -0.21428358554840088, 0.07450968772172928, -0.013634569011628628, -0.504682183265686, 0.1931382119655609, 0.4102118909358978, 0.299521803855896, -0.2957395017147064, -0.4034365117549896, -0.053517505526542664, -0.03167328983545303, 0.590488076210022, -0.09309849143028259, 0.25580862164497375, -0.02029447816312313, -0.06831389665603638, 0.07016165554523468, -0.12563391029834747, 0.19013677537441254, -0.169069305062294, 0.17845679819583893, 0.23488174378871918, -0.23076699674129486, 0.3158995509147644, 0.3804570138454437, 0.3273269832134247, 0.21452637016773224, 0.038707707077264786, -0.3840380012989044, -0.06015699729323387, 0.2709561586380005, 0.06697278469800949, 0.3688613474369049, -0.05914023518562317, 0.22426332533359528, -0.13389992713928223, -0.15443681180477142, -0.17635619640350342, 0.3210904598236084, -0.20176923274993896, 0.07100435346364975, 0.06617741286754608, 0.5125982165336609, 0.01798960380256176, -0.18942147493362427, 0.30087465047836304, -0.016682641580700874, 0.23172098398208618, -0.41760414838790894, 0.23898622393608093, -0.31163734197616577, -0.0587846077978611, -0.34087440371513367, -0.21178214251995087, -0.0825011357665062, -0.5657790303230286, -0.03033321164548397, -0.03345087543129921, -0.07007171213626862, 0.03609125316143036, -0.11940274387598038, -0.14799173176288605, -0.10889928787946701, -0.0712956041097641, -0.37340956926345825, -0.4300119876861572, -0.2597953975200653, -0.04753952845931053, 0.3729808032512665, 0.23948852717876434, 0.10474298149347305, -0.012624762952327728, -0.07535742223262787, 0.0003270961169619113, -0.47142863273620605, -0.004124365746974945, -0.22297115623950958, 0.0679432600736618, 0.18282322585582733, 0.021070389077067375, -0.030382581055164337, -0.014436699450016022, 0.36036160588264465, -0.06673930585384369, -0.27643445134162903, 0.35011184215545654, 0.0663587749004364, 0.05440806597471237, -0.29347938299179077, -0.3115214407444, 0.08216596394777298, -0.4667847752571106, 0.300645112991333, -0.09647846966981888, 0.24163766205310822, 0.21147532761096954, -0.017977360635995865, 0.22887639701366425, -0.06533212959766388, 0.16171418130397797, 0.18838968873023987, 0.4239374101161957, 0.540738582611084, -0.2293756753206253, -0.3174418807029724, 0.2958123981952667, -0.037645719945430756, 0.38819605112075806, 0.14681152999401093, -0.048438142985105515, 0.4057910740375519, -0.3843775987625122, 0.13194125890731812, -0.14741848409175873, -0.07091381400823593, 0.1809765249490738, 0.23529167473316193, -0.05996207147836685, -0.1455635279417038, 0.033156536519527435, 0.21858540177345276, -0.030756212770938873, 0.24850602447986603, 0.09110970795154572, 0.0045342035591602325, -0.113542839884758, 0.3820222318172455, 0.30407631397247314, -0.15708640217781067, 0.2629910707473755, 0.006090229842811823, -0.0635354146361351, -0.230088472366333, -0.27072814106941223, -0.12888337671756744, -0.17353898286819458, 0.1558312028646469, -0.15172052383422852, 0.0335836224257946, -0.3819095492362976, -0.031434036791324615, 0.3377803564071655, -0.18510060012340546, -0.1455889642238617, 0.29326602816581726, -0.27848073840141296, 0.23703905940055847, 0.08548372238874435, 0.1900608092546463, -0.1247149333357811, -0.4073355197906494, 0.1250004768371582, -0.0006938294391147792, 0.11784909665584564, 0.036328744143247604, -0.1986617147922516, -0.1296972632408142, -0.41023313999176025, 0.3197307288646698, 0.039933692663908005, 0.360186904668808, -0.16127030551433563, -0.15637345612049103, 0.18404246866703033, -0.1521962583065033, 0.4350278675556183, -0.36648014187812805, -0.05906342715024948, 0.30513739585876465, 0.3421054184436798, -0.4739360213279724, 0.34192028641700745, -0.22176504135131836, 0.24682144820690155, 0.16475537419319153, 0.14953535795211792, -0.3234817385673523, -0.23489433526992798, 0.1747419387102127, 0.020166005939245224, 0.023735633119940758, 0.1704343557357788, -0.3350856602191925, 0.026662440970540047, -0.07697520405054092, -0.1739080846309662, 0.2077370136976242, 0.24202120304107666, 0.018690332770347595, 0.07589264214038849, -0.4179532825946808, 0.19375494122505188, 0.2174454629421234, 0.08719145506620407, 0.3997781574726105, -0.24007274210453033, 0.011526733636856079, -0.020228853449225426, 0.19943805038928986, -0.07045351713895798, 0.4529924690723419, -0.04165403172373772, -0.1785091757774353, 0.25579404830932617, -0.08369319140911102, 0.4723670482635498, 0.3834904134273529, 0.08568950742483139, 0.32462552189826965, -0.27232176065444946, -0.22830967605113983, -0.3209860920906067, 0.00694937352091074, 0.041178762912750244, 0.00267806276679039, 0.14303724467754364, -0.10382387787103653, 0.5234208106994629, 0.2486489862203598, -0.07634872198104858, 0.5133993029594421, 0.16703180968761444, -0.3357830345630646, 0.32315903902053833, 0.22793793678283691, 1.157379388809204, 0.05849060043692589, 0.10378333926200867, -0.05825208127498627, -0.46720442175865173, 0.44061094522476196, 0.06488718092441559, 0.2514752745628357, -0.47175082564353943, 0.35401567816734314, 0.004887866787612438, -0.2005733847618103, 0.04434827342629433, -0.05194282531738281, -0.30089521408081055, 0.11113577336072922, -0.21602806448936462, 0.2448735386133194, 0.09667272865772247, 0.09772028028964996, -0.3100966215133667, -0.30526241660118103, -0.2676105797290802, 0.09327080100774765, -0.18762537837028503, -0.00799136608839035, 0.11706539243459702, -0.29040437936782837, -0.5326505899429321, -0.0551835335791111, -0.08857368677854538, 0.1917651742696762, -0.19879907369613647, -0.07285914570093155, 0.25250282883644104, -0.20954354107379913, 0.2399556189775467, 0.23208840191364288, 0.08816596865653992, -0.06347980350255966, -0.17391429841518402, -0.047285061329603195, 0.29203933477401733, 0.1568920910358429, 0.12021462619304657, -0.0331294983625412, 0.014185520820319653, 0.10079514980316162, 0.04560549929738045, 0.25179681181907654, -0.4120059907436371, 0.04076381400227547, -0.08722461014986038, 0.2069534808397293, 0.09257873147726059, -0.05611662194132805, 0.05833389610052109, 0.05692851543426514, -0.05763848125934601, -0.024936117231845856, 0.10847318172454834, 0.013143637217581272, 0.16904456913471222, 0.03570549562573433, -0.11537537723779678, -0.2564200758934021, 0.3108493983745575, 0.3111068904399872, -0.011716415174305439, 0.2638019919395447, 0.7341852188110352, 0.027994580566883087, -0.1527639478445053, -0.19830748438835144, 0.028630534186959267, -0.024160174652934074, -0.731462299823761, 0.31250956654548645, -0.31606706976890564, -0.36789244413375854, 0.020080441609025, 0.5516511797904968, -0.23535038530826569, 0.18464386463165283, -0.18532207608222961, -0.26003187894821167, -0.19181127846240997, 0.29425421357154846, -0.06804889440536499, 0.006812755949795246, -0.4693596661090851, -0.10694549232721329, 0.029502982273697853, 0.1694559007883072, -0.24777010083198547, 0.035324256867170334, -0.1177469864487648, 0.3626191020011902, -0.04668654128909111, -0.005117875523865223, -0.0046821958385407925, -0.2574218213558197, 0.01005176268517971, 0.5666001439094543, -0.12076998502016068, -0.14190907776355743, -0.28333723545074463, 0.14587780833244324, 0.08442272990942001, -0.17020735144615173, 0.16537907719612122, -0.01781695894896984, 0.06224733218550682, -0.21514610946178436, -0.07266789674758911, -0.20808859169483185, -0.18448610603809357, 0.07485037297010422, -0.02110034041106701, 0.24461881816387177, -0.23421530425548553, 0.08050135523080826, -0.07008925825357437, 0.06816471368074417, -0.3697080910205841, 0.00714884651824832, -0.014692851342260838, 0.002423536963760853, -0.035316284745931625, 0.15229454636573792, -0.18646304309368134, -0.11169900000095367, 0.2823202311992645, -0.29534220695495605, -0.12911154329776764, -0.0050050606951117516, -0.1942414939403534, 0.337995707988739, -0.15246935188770294, 0.048098124563694, -0.04230814427137375, 0.11493926495313644, 0.13665582239627838, 0.3571574091911316, -0.26209986209869385, 0.021108334884047508, -0.08318186551332474, 0.2757546901702881, 0.12450657039880753, -0.22824057936668396, -0.10933318734169006, 0.06607474386692047, -0.06436789035797119, -0.369974285364151, -0.13122640550136566, 0.1212935820221901, -0.14155219495296478, -0.002717074006795883, 0.31574881076812744, -0.15956202149391174, -0.03928850591182709, -0.0020723100751638412, 0.39843764901161194, 0.31034424901008606, 0.5081049203872681, 0.20689082145690918, 0.003498430596664548, 0.09589101374149323, 0.00870201550424099, 0.24191176891326904, 0.3457309603691101, 0.3615863025188446, 0.2970237731933594, 0.6127427816390991, -0.3836425542831421, -0.26814690232276917, -0.02600928023457527, 0.10279225558042526, -0.2622460126876831, -0.14170002937316895, -0.5835604071617126, -0.12105085700750351, -0.44204121828079224, -0.08055131882429123, -0.21707743406295776, 0.19709809124469757, 0.5802695155143738, 0.21965718269348145, -0.045474473387002945, -0.629024863243103, -0.20984308421611786, 0.05817921832203865, 0.5017370581626892, -0.22398348152637482, 0.45599639415740967, 0.016069067642092705, -0.05682409182190895, -0.018256286159157753, 0.5782517790794373, 0.43931037187576294, 0.5464792847633362, -0.3213455379009247, -0.013242781162261963, 0.0590827502310276, -0.1702311784029007, 0.2828426957130432, 0.6259992122650146, 0.37193578481674194, 0.0720667913556099, 0.049773309379816055, -0.02863697148859501, -0.09993834048509598, -0.26097285747528076, 0.02182992547750473, 0.456132709980011, -0.009454096667468548, 0.2189071625471115, -0.29675373435020447, -0.13070376217365265, -0.07994164526462555, -0.2456393986940384, -0.301233172416687, -0.07740378379821777, 0.13583458960056305, 0.016438117250800133, 0.0205108392983675, -0.3425390124320984, 0.07126209139823914, 0.160375714302063, 0.5518667101860046, 0.2615603804588318, 0.1266721934080124, -0.41710996627807617, -0.11879701167345047, -0.5855775475502014, 0.2008872777223587, -0.2385326325893402, -0.1507570743560791, -0.09952836483716965, 0.3261133134365082, 0.10537153482437134, 0.24450449645519257, -0.02768172137439251, 0.20349782705307007, 0.3321015238761902, 0.3179222643375397, -0.3015892207622528, -0.062345899641513824, 0.3174634277820587, -0.32736799120903015, 0.008463076315820217, -0.575021505355835, 0.29662367701530457, 0.067864328622818, 0.06840582937002182, 0.1313173472881317, -0.30147168040275574, 0.20186811685562134, 0.44387027621269226, 0.37146812677383423, 0.21601693332195282, 0.25399497151374817, 0.25643858313560486, -0.45232388377189636, -0.2691612243652344, -0.2093794345855713, -0.2386341243982315, 0.10153771191835403, 0.10521799325942993, 0.5344883799552917, -0.08086124807596207, 0.1265619397163391, -0.2559089958667755, 0.15339700877666473, -0.17736394703388214, 0.10928677767515182, -0.2637519836425781, 0.14666658639907837, -0.31743040680885315, 0.08098729699850082, 0.12022680789232254, 0.12272747606039047, 0.15131613612174988, 0.14416058361530304, -0.4850176274776459, -0.45378050208091736, 0.3317623436450958, -0.3799087703227997, -0.3346240520477295, 0.3907823860645294, 0.21516910195350647, 0.4396253824234009, -0.10427311062812805, -0.7881220579147339, -0.35706278681755066, 0.14721404016017914, -0.1391952633857727, -0.050900012254714966, -0.11611464619636536, -0.056644365191459656, -0.00523604080080986, -0.21112196147441864, 0.3439011573791504, -0.008731664158403873, 0.08372963964939117, 0.09028727561235428, -0.16827607154846191 ]
https://github.com/huggingface/datasets/issues/211
[Arrow writer, Trivia_qa] Could not convert TagMe with type str: converting to null type
Actually, I know what the problem is...I'm quite sure it's a bug. Here we take some test inputs: https://github.com/huggingface/nlp/blob/0e0ef12c14d2175e0b0bd7d8aa814b09e2cd7e1f/src/nlp/arrow_dataset.py#L472 It might be that in the test inputs, a `Sequence` type value is an emtpy list. So in my case I have `ds[0]["entity_pages']["wiki_context"] = []`. => this leads to an `arrow_schema` equal to `null` for `["entity_pages']["wiki_context"]` => see line: https://github.com/huggingface/nlp/blob/0e0ef12c14d2175e0b0bd7d8aa814b09e2cd7e1f/src/nlp/arrow_dataset.py#L501 instead of list of string which it should for other examples. Guess it's an edge case, but it can happen.
Running the following code ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, load_from_cache_file=False) ``` triggers a `ArrowInvalid: Could not convert TagMe with type str: converting to null type` error. On the other hand if we remove a certain column of `trivia_qa` which seems responsible for the bug, it works: ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, remove_columns=["entity_pages"], load_from_cache_file=False) ``` . Seems quite hard to debug what's going on here... @lhoestq @thomwolf - do you have a good first guess what the problem could be? **Note** BTW: I think this could be a good test to check that the datasets work correctly: Take a tiny portion of the dataset and check that it can be written correctly.
79
[Arrow writer, Trivia_qa] Could not convert TagMe with type str: converting to null type Running the following code ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, load_from_cache_file=False) ``` triggers a `ArrowInvalid: Could not convert TagMe with type str: converting to null type` error. On the other hand if we remove a certain column of `trivia_qa` which seems responsible for the bug, it works: ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, remove_columns=["entity_pages"], load_from_cache_file=False) ``` . Seems quite hard to debug what's going on here... @lhoestq @thomwolf - do you have a good first guess what the problem could be? **Note** BTW: I think this could be a good test to check that the datasets work correctly: Take a tiny portion of the dataset and check that it can be written correctly. Actually, I know what the problem is...I'm quite sure it's a bug. Here we take some test inputs: https://github.com/huggingface/nlp/blob/0e0ef12c14d2175e0b0bd7d8aa814b09e2cd7e1f/src/nlp/arrow_dataset.py#L472 It might be that in the test inputs, a `Sequence` type value is an emtpy list. So in my case I have `ds[0]["entity_pages']["wiki_context"] = []`. => this leads to an `arrow_schema` equal to `null` for `["entity_pages']["wiki_context"]` => see line: https://github.com/huggingface/nlp/blob/0e0ef12c14d2175e0b0bd7d8aa814b09e2cd7e1f/src/nlp/arrow_dataset.py#L501 instead of list of string which it should for other examples. Guess it's an edge case, but it can happen.
[ 0.09355766326189041, -0.05218100547790527, 0.09173014760017395, 0.49227437376976013, 0.25251901149749756, 0.1492108851671219, 0.10871338099241257, 0.4077002704143524, 0.1446874439716339, -0.09357922524213791, 0.345406711101532, 0.5846510529518127, -0.12629729509353638, -0.08148028701543808, -0.03906698152422905, -0.18967218697071075, 0.1572442501783371, 0.2686879336833954, 0.0862920731306076, -0.06003141030669212, -0.14310400187969208, 0.1505410373210907, -0.27892494201660156, 0.13217096030712128, -0.42001497745513916, -0.10315194725990295, 0.07676168531179428, 0.05711652711033821, -0.08841797709465027, -0.4528536796569824, 0.07614684104919434, -0.32569846510887146, -0.006574613973498344, 0.25914573669433594, -0.00012003634037682787, -0.12074823677539825, 0.4130159914493561, -0.04033127799630165, -0.2655787467956543, -0.2495540827512741, -0.4358331859111786, 0.031889259815216064, 0.3316240906715393, -0.042490363121032715, 0.18057368695735931, -0.06374640762805939, 0.023802658542990685, -0.5875160098075867, 0.20610851049423218, 0.5833629369735718, 0.17648322880268097, 0.2174147665500641, -0.06840876489877701, 0.13026194274425507, 0.29421254992485046, 0.053159862756729126, -0.3383094072341919, -0.08270297944545746, 0.02263263799250126, -0.12178908288478851, -0.13316574692726135, 0.3214308023452759, -0.06175127252936363, 0.05857021361589432, 0.6247365474700928, 0.09327878803014755, 0.1312829554080963, -0.3116703927516937, 0.14749625325202942, 0.1445361226797104, 0.39628154039382935, -0.08866601437330246, -0.08991008251905441, -0.017770590260624886, -0.16316336393356323, -0.3528529107570648, 0.2785196900367737, 0.18064726889133453, -0.29843664169311523, -0.003344345837831497, 0.07484898716211319, 0.14837564527988434, -0.08723309636116028, -0.06426207721233368, -0.11870991438627243, 0.15310683846473694, 0.03931651636958122, 0.06686494499444962, -0.104098841547966, -0.20409707725048065, -0.10837506502866745, -0.23340703547000885, -0.2524460554122925, 0.31610387563705444, -0.4382787048816681, 0.03797251731157303, -0.051442280411720276, -0.13796330988407135, -0.1476161926984787, 0.06295989453792572, 0.38411587476730347, 0.09981704503297806, 0.027437696233391762, 0.07166105508804321, 0.35871410369873047, 0.3659045994281769, 0.31820401549339294, 0.13503821194171906, 0.0632362961769104, -0.22235779464244843, -0.0816001445055008, 0.017592081800103188, 0.01669975183904171, -0.03930918872356415, 0.40901151299476624, 0.009138377383351326, 0.1497366577386856, -0.3318137228488922, -0.24876804649829865, 0.05874061957001686, -0.7114659547805786, 0.2303275465965271, -0.19834335148334503, 0.09589500725269318, -0.024716250598430634, 0.522648811340332, 0.2696354389190674, 0.3240777254104614, -0.22753983736038208, -0.1015937551856041, -0.1469992995262146, -0.13756315410137177, -0.2435154914855957, -0.04535217210650444, 0.2112690508365631, -0.07352469861507416, 0.16572313010692596, 0.024609845131635666, -0.4914281666278839, -0.20295187830924988, 0.266869455575943, -0.3210097849369049, 0.17625240981578827, 0.06481609493494034, -0.3296072483062744, 0.4532787799835205, 0.18611638247966766, -0.39012324810028076, -0.18190471827983856, 0.10338719934225082, -0.19294776022434235, 0.09501991420984268, -0.15740551054477692, 0.12872961163520813, 0.02487931400537491, -0.14852629601955414, -0.302363783121109, -0.032280128449201584, 0.3422367572784424, -0.11202200502157211, 0.15281814336776733, -0.24676993489265442, -0.10657190531492233, -0.2501494288444519, -0.06317627429962158, 0.40901312232017517, -0.1929149478673935, 0.08677531033754349, -0.018420295789837837, 0.1405181735754013, 0.2578374743461609, 0.2855258285999298, 0.016144372522830963, 0.12157026678323746, 0.0634678304195404, 0.3946039378643036, 0.431523859500885, -0.2790813148021698, 0.007540145423263311, 0.13285373151302338, -0.03557681292295456, -0.011855741031467915, 0.0022207186557352543, 0.10645212978124619, 0.03309134766459465, -0.2236136496067047, 0.10748156160116196, 0.19908714294433594, -0.0646124854683876, 0.2510935962200165, -0.4027024805545807, -0.06630803644657135, 0.49514520168304443, -0.201971173286438, -0.18115805089473724, -0.05476008728146553, -0.053745515644550323, 0.15148408710956573, -0.014162950217723846, -0.028818197548389435, 0.38277581334114075, 0.3288469612598419, -0.14390338957309723, -0.16071735322475433, 0.10063563287258148, -0.10738221555948257, -0.46785426139831543, 0.01179387979209423, -0.1890222728252411, 0.20261098444461823, -0.08662533760070801, 0.028060849756002426, -0.34109804034233093, 0.12534889578819275, -0.1016518771648407, -0.22731013596057892, 0.14942359924316406, 0.05417415127158165, 0.16102512180805206, 0.1465352475643158, -0.12235677242279053, 0.11216121166944504, 0.013423270545899868, 0.08212599903345108, -0.4639512002468109, -0.08359189331531525, 0.0000042855485844484065, -0.28666579723358154, -0.08776101469993591, 0.2652682363986969, 0.27048373222351074, 0.11230027675628662, -0.2972470223903656, 0.21522992849349976, 0.11378917098045349, 0.27887773513793945, -0.402119517326355, 0.07179590314626694, 0.06320440024137497, -0.4297589063644409, 0.1421477496623993, 0.43532252311706543, 0.3686196804046631, -0.3424835801124573, -0.3888028562068939, -0.04157598316669464, 0.02234947495162487, 0.6803267598152161, -0.1308952420949936, 0.14410792291164398, -0.0329953096807003, -0.03734302893280983, 0.07075387984514236, -0.054917752742767334, 0.16543681919574738, 0.017986038699746132, 0.12332194298505783, 0.30982547998428345, -0.24867281317710876, 0.2911902666091919, 0.3783717155456543, 0.2824704945087433, 0.11098290234804153, 0.12216751277446747, -0.38529273867607117, -0.21817083656787872, 0.3125988841056824, 0.025050168856978416, 0.3757631480693817, -0.05269020050764084, 0.18641436100006104, -0.05416795238852501, -0.18275856971740723, -0.17576919496059418, 0.3537931740283966, -0.05842425674200058, 0.19384706020355225, 0.036412253975868225, 0.5011972188949585, -0.0158939640969038, -0.131381094455719, 0.24980710446834564, -0.00886718463152647, 0.2077244073152542, -0.41092801094055176, 0.3374968469142914, -0.4077058732509613, -0.13143102824687958, -0.37055832147598267, -0.15277187526226044, -0.02439740113914013, -0.6165552139282227, -0.04489520192146301, -0.018378380686044693, -0.030628075823187828, 0.12242794036865234, -0.28436094522476196, -0.13166160881519318, 0.013293184340000153, -0.07765060663223267, -0.40242815017700195, -0.42999595403671265, -0.4094296097755432, -0.04794970154762268, 0.43011748790740967, 0.3920780122280121, 0.10363422334194183, 0.032187048345804214, -0.10713531821966171, -0.036247964948415756, -0.41251081228256226, 0.034375015646219254, -0.2577628791332245, 0.09031189233064651, 0.15104347467422485, 0.06090349704027176, 0.01968619041144848, -0.08685033768415451, 0.324641615152359, 0.0006720252567902207, -0.3420938551425934, 0.39525091648101807, 0.053129855543375015, 0.14244604110717773, -0.3051876425743103, -0.4616941809654236, 0.03970060124993324, -0.46554744243621826, 0.18202948570251465, 0.006348861381411552, 0.18158811330795288, 0.23612694442272186, 0.05442655459046364, 0.28018149733543396, -0.043023254722356796, 0.16916722059249878, 0.19449365139007568, 0.47960445284843445, 0.5730380415916443, -0.14828301966190338, -0.36124691367149353, 0.2676405608654022, -0.13377965986728668, 0.29561081528663635, 0.1793285310268402, -0.15094240009784698, 0.3191583752632141, -0.3194126784801483, 0.15346930921077728, -0.13758321106433868, -0.12629735469818115, 0.23981399834156036, 0.18344949185848236, -0.04229230806231499, -0.10745716094970703, 0.03156398609280586, 0.25976717472076416, -0.18653292953968048, 0.19254785776138306, 0.11958233267068863, 0.20338129997253418, -0.04648185521364212, 0.3153972327709198, 0.18806946277618408, -0.2586815357208252, 0.33788928389549255, -0.0351509265601635, -0.008244981989264488, -0.28940466046333313, -0.1092589870095253, -0.08718905597925186, -0.14598052203655243, 0.15237271785736084, -0.08302395045757294, 0.0713333860039711, -0.45441335439682007, -0.028706520795822144, 0.29916152358055115, -0.10596811026334763, -0.14508476853370667, 0.2503581941127777, -0.27159324288368225, 0.2326231449842453, 0.07304343581199646, 0.15341870486736298, -0.11898381263017654, -0.5184163451194763, 0.044182710349559784, -0.03199446201324463, 0.09283964335918427, -0.02092238888144493, -0.20190250873565674, 0.01781037263572216, -0.4196453094482422, 0.28542956709861755, 0.16831742227077484, 0.4798281192779541, -0.04797329753637314, -0.19165867567062378, 0.17190347611904144, -0.15334923565387726, 0.38575008511543274, -0.27435535192489624, -0.013487877324223518, 0.36245670914649963, 0.3600074052810669, -0.48109734058380127, 0.2678271234035492, -0.25803855061531067, 0.2621382772922516, 0.17987297475337982, 0.19627130031585693, -0.4151628911495209, -0.3015872836112976, 0.1404121071100235, 0.1914491057395935, -0.021241120994091034, 0.16336673498153687, -0.2792501449584961, -0.023449843749403954, -0.06628536432981491, -0.17774102091789246, 0.2659941613674164, 0.3581659495830536, 0.03981360048055649, -0.0013975698966532946, -0.43300962448120117, 0.11331041902303696, 0.12204871326684952, 0.02093997225165367, 0.45545026659965515, -0.29696565866470337, 0.12399213761091232, -0.20001664757728577, 0.1642337590456009, -0.06789131462574005, 0.3436698913574219, -0.06636131554841995, -0.09572024643421173, 0.3054822087287903, -0.12876875698566437, 0.3905344605445862, 0.33944132924079895, -0.02647395245730877, 0.3637794852256775, -0.17648418247699738, -0.15365412831306458, -0.24443653225898743, -0.02531627006828785, 0.14097005128860474, -0.08740124106407166, 0.10841009765863419, -0.19618138670921326, 0.541181206703186, 0.3007127642631531, -0.03165442496538162, 0.44397681951522827, 0.18964138627052307, -0.32779377698898315, 0.3507964015007019, 0.21915358304977417, 1.112839698791504, 0.11890758574008942, 0.1753784418106079, 0.08425147831439972, -0.34730711579322815, 0.4513011872768402, 0.022426730021834373, 0.2541965842247009, -0.48580217361450195, 0.2573942542076111, -0.03554905205965042, -0.1875907927751541, 0.040305085480213165, 0.024343961849808693, -0.28638988733291626, 0.19320900738239288, -0.1737767457962036, 0.26287317276000977, 0.12903361022472382, 0.019911402836441994, -0.2860625088214874, -0.3354416489601135, -0.4839943051338196, 0.07514113932847977, -0.15866246819496155, -0.06433840841054916, 0.10971218347549438, -0.2614796459674835, -0.4933299124240875, -0.09190774708986282, -0.12715232372283936, 0.14361156523227692, -0.2413683980703354, -0.06229012459516525, 0.2582911550998688, -0.17976689338684082, 0.22817149758338928, 0.16256064176559448, -0.04886376112699509, -0.05741990730166435, -0.18875886499881744, -0.11977733671665192, 0.20588086545467377, 0.09072998911142349, 0.1580936312675476, -0.019870692864060402, 0.14279165863990784, 0.05803072452545166, 0.0911153256893158, 0.26428547501564026, -0.4077533781528473, -0.0016339104622602463, 0.06845354288816452, 0.16407932341098785, 0.035154107958078384, -0.021139148622751236, -0.12933379411697388, 0.10254568606615067, -0.016364311799407005, -0.03418673202395439, 0.08982974290847778, 0.04698130488395691, 0.08257539570331573, 0.05820777267217636, -0.10640516132116318, -0.22666843235492706, 0.26345932483673096, 0.353193461894989, 0.057198356837034225, 0.14272695779800415, 0.748661458492279, 0.16375793516635895, -0.13262833654880524, -0.19221459329128265, 0.0859968289732933, 0.01857256144285202, -0.7875836491584778, 0.28427791595458984, -0.27143025398254395, -0.3222106397151947, 0.0358475036919117, 0.6465660333633423, -0.2318478226661682, 0.08951954543590546, -0.1663874089717865, -0.26115164160728455, -0.1747756004333496, 0.17243994772434235, -0.037130579352378845, 0.06116075813770294, -0.5098102688789368, -0.18457739055156708, 0.10252352058887482, 0.1319335699081421, -0.21777282655239105, 0.00017779628979042172, -0.2947470545768738, 0.3535670340061188, -0.014866601675748825, 0.1016855463385582, 0.00531828124076128, -0.17168766260147095, -0.004655370954424143, 0.4534411132335663, -0.16379664838314056, -0.14825889468193054, -0.23846152424812317, 0.15354657173156738, 0.11453521996736526, -0.11689381301403046, 0.06979844719171524, -0.021558290347456932, 0.030570514500141144, -0.21219263970851898, -0.10736989974975586, -0.21035702526569366, -0.17519158124923706, -0.01351905707269907, 0.07630708068609238, 0.2623259425163269, -0.27176329493522644, 0.12272558361291885, -0.172366201877594, 0.07838195562362671, -0.3097916543483734, 0.02072833850979805, 0.11779620498418808, 0.021430138498544693, -0.1464223712682724, 0.10597628355026245, -0.16352549195289612, -0.21059295535087585, 0.32429587841033936, -0.34931859374046326, -0.20742009580135345, 0.04359722137451172, -0.1591494083404541, 0.3959813714027405, -0.10549292713403702, 0.0026339900214225054, -0.07950644195079803, 0.08346452564001083, 0.054279930889606476, 0.32216352224349976, -0.1934608370065689, -0.11521857976913452, -0.04124770686030388, 0.28510916233062744, 0.17262502014636993, -0.2153887152671814, -0.026877859607338905, 0.11490268260240555, 0.02829141728579998, -0.4184589982032776, -0.08133561909198761, 0.18888068199157715, -0.048989348113536835, 0.0191485658288002, 0.26837390661239624, -0.1958974450826645, 0.028500817716121674, -0.0704050287604332, 0.43938660621643066, 0.41986083984375, 0.4042411744594574, 0.16703341901302338, -0.024890005588531494, 0.08554977923631668, 0.05491384491324425, 0.3567959666252136, 0.34315255284309387, 0.2155420035123825, 0.2542288899421692, 0.5395470857620239, -0.17617860436439514, -0.2687586843967438, -0.056569285690784454, 0.12042010575532913, -0.22145405411720276, -0.13927051424980164, -0.4548981189727783, -0.15396693348884583, -0.4800412058830261, -0.20011095702648163, -0.16494038701057434, 0.138153076171875, 0.5681524276733398, 0.22027188539505005, 0.04264887422323227, -0.5250192284584045, -0.1378670036792755, 0.10644234716892242, 0.3856295943260193, -0.17447805404663086, 0.4604705274105072, 0.0065398914739489555, -0.07063817977905273, -0.03053625300526619, 0.4228677451610565, 0.42817649245262146, 0.5327763557434082, -0.3272393047809601, -0.0688585489988327, 0.12607942521572113, -0.16419455409049988, 0.24071285128593445, 0.5911155939102173, 0.24773690104484558, 0.0660373643040657, 0.06582170724868774, -0.05595829710364342, -0.08274539560079575, -0.10225710272789001, -0.022090107202529907, 0.3411233723163605, -0.12443617731332779, 0.4214782118797302, -0.277730792760849, -0.14602114260196686, -0.1004435271024704, -0.18617010116577148, -0.2864459753036499, 0.03204261511564255, 0.10367326438426971, 0.03099736012518406, 0.07790971547365189, -0.3275085687637329, 0.0678439512848854, 0.18376100063323975, 0.5826790928840637, 0.224173441529274, 0.1032862439751625, -0.39242225885391235, -0.09169916063547134, -0.7618368268013, 0.17532789707183838, -0.24168361723423004, -0.21087729930877686, 0.10276138037443161, 0.3020719587802887, 0.0831141471862793, 0.3048587739467621, -0.04290934279561043, 0.2784472703933716, 0.2520654797554016, 0.2867039144039154, -0.38265106081962585, -0.1970779448747635, 0.3445274233818054, -0.28109580278396606, -0.052883487194776535, -0.5603258013725281, 0.23074781894683838, 0.01634921319782734, 0.0587674081325531, 0.03245606645941734, -0.21475018560886383, 0.18569064140319824, 0.4447222352027893, 0.20759929716587067, 0.18357287347316742, 0.17585748434066772, 0.27285662293434143, -0.4320274293422699, -0.3502245843410492, -0.21792975068092346, -0.19331420958042145, 0.03421096131205559, 0.03560389205813408, 0.4834496080875397, -0.16845664381980896, 0.03575218841433525, -0.27571362257003784, 0.1209489181637764, -0.1355479210615158, 0.1336914747953415, -0.2686525583267212, 0.1832190454006195, -0.4139765202999115, 0.19967924058437347, 0.22928112745285034, 0.022406091913580894, 0.12845252454280853, 0.14924822747707367, -0.4297786355018616, -0.49896353483200073, 0.2569376826286316, -0.47044065594673157, -0.3264242112636566, 0.2818511724472046, 0.14563846588134766, 0.35117805004119873, -0.06056055426597595, -0.740279495716095, -0.2556822896003723, 0.23546096682548523, -0.12984558939933777, 0.017962753772735596, -0.04342944175004959, -0.04658288136124611, 0.026109222322702408, -0.23466290533542633, 0.23958688974380493, 0.0006647949921898544, 0.1402984857559204, 0.033132340759038925, -0.25771474838256836 ]
https://github.com/huggingface/datasets/issues/211
[Arrow writer, Trivia_qa] Could not convert TagMe with type str: converting to null type
Good point, I think the schema should be infered at the writing stage where we have a `writer_batch_size` number of examples (typically 10k) so it's even less likely to run into this scenario.
Running the following code ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, load_from_cache_file=False) ``` triggers a `ArrowInvalid: Could not convert TagMe with type str: converting to null type` error. On the other hand if we remove a certain column of `trivia_qa` which seems responsible for the bug, it works: ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, remove_columns=["entity_pages"], load_from_cache_file=False) ``` . Seems quite hard to debug what's going on here... @lhoestq @thomwolf - do you have a good first guess what the problem could be? **Note** BTW: I think this could be a good test to check that the datasets work correctly: Take a tiny portion of the dataset and check that it can be written correctly.
33
[Arrow writer, Trivia_qa] Could not convert TagMe with type str: converting to null type Running the following code ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, load_from_cache_file=False) ``` triggers a `ArrowInvalid: Could not convert TagMe with type str: converting to null type` error. On the other hand if we remove a certain column of `trivia_qa` which seems responsible for the bug, it works: ``` import nlp ds = nlp.load_dataset("trivia_qa", "rc", split="validation[:1%]") # this might take 2.3 min to download but it's cached afterwards... ds.map(lambda x: x, remove_columns=["entity_pages"], load_from_cache_file=False) ``` . Seems quite hard to debug what's going on here... @lhoestq @thomwolf - do you have a good first guess what the problem could be? **Note** BTW: I think this could be a good test to check that the datasets work correctly: Take a tiny portion of the dataset and check that it can be written correctly. Good point, I think the schema should be infered at the writing stage where we have a `writer_batch_size` number of examples (typically 10k) so it's even less likely to run into this scenario.
[ -0.06821571290493011, 0.06093074008822441, 0.10538256913423538, 0.47370103001594543, 0.25208044052124023, 0.10438002645969391, 0.0982680395245552, 0.38123026490211487, 0.22328874468803406, -0.0625070109963417, 0.3270520269870758, 0.47998303174972534, -0.08507680147886276, -0.02218559943139553, -0.045739635825157166, -0.13106898963451385, 0.1612941324710846, 0.32369357347488403, 0.0635414570569992, -0.12184275686740875, -0.143835186958313, 0.08432035893201828, -0.23519447445869446, 0.07620338350534439, -0.5430988669395447, -0.16950716078281403, 0.05933636799454689, 0.10318562388420105, -0.15469597280025482, -0.444821834564209, 0.014603621326386929, -0.2505456805229187, 0.1371711939573288, 0.3100917339324951, -0.00011774223821703345, -0.13116763532161713, 0.3823634088039398, -0.045255277305841446, -0.21709266304969788, -0.15531252324581146, -0.4626953601837158, 0.06080995500087738, 0.26331013441085815, -0.06844861060380936, 0.19009912014007568, -0.002296833321452141, 0.06410693377256393, -0.49992018938064575, 0.10783161967992783, 0.5157186388969421, 0.20820803940296173, 0.2796899974346161, -0.018799923360347748, 0.03328898921608925, 0.32798701524734497, 0.02105787955224514, -0.36564770340919495, 0.0005635718698613346, -0.0283157080411911, -0.15549702942371368, -0.16190877556800842, 0.3987061679363251, -0.08816108107566833, 0.05234824866056442, 0.6084602475166321, 0.09695235639810562, 0.11834535747766495, -0.29125991463661194, 0.19596010446548462, 0.12539058923721313, 0.4451824724674225, -0.10498611629009247, -0.11374671757221222, -0.003936838358640671, -0.13454768061637878, -0.39042434096336365, 0.23909571766853333, 0.19062256813049316, -0.2509995996952057, -0.003180446568876505, 0.05959679186344147, 0.10651864111423492, -0.035818811506032944, -0.19298069179058075, -0.06625010073184967, 0.20389649271965027, 0.06159031763672829, 0.04267077147960663, -0.10467099398374557, -0.1581258326768875, -0.012559236958622932, -0.2313852161169052, -0.3124712407588959, 0.2501974403858185, -0.33636394143104553, 0.042031411081552505, 0.00923119392246008, -0.10786877572536469, -0.1433400809764862, -0.07465704530477524, 0.3727557063102722, 0.023687245324254036, 0.13817273080348969, 0.11149328947067261, 0.3602561950683594, 0.3803674578666687, 0.3734176456928253, 0.13893984258174896, 0.1727735698223114, -0.25541576743125916, -0.20995932817459106, -0.018516505137085915, -0.02588663622736931, -0.02692083828151226, 0.48337507247924805, -0.018195142969489098, 0.12692323327064514, -0.34274446964263916, -0.28641602396965027, 0.03796941041946411, -0.6072560548782349, 0.17981718480587006, -0.21649353206157684, 0.18091121315956116, 0.014958344399929047, 0.4404577612876892, 0.14115053415298462, 0.3197545111179352, -0.2866175174713135, -0.2129356861114502, -0.16906729340553284, -0.06719358265399933, -0.3529975414276123, -0.04740668833255768, 0.23163479566574097, 0.025345223024487495, 0.17099273204803467, 0.04575207084417343, -0.5268826484680176, -0.19393038749694824, 0.2752547562122345, -0.33649227023124695, 0.1010017991065979, 0.09799926728010178, -0.18175622820854187, 0.38011592626571655, 0.15239612758159637, -0.3950263261795044, -0.2050779163837433, 0.07577133923768997, -0.14228346943855286, -0.014657155610620975, -0.14263610541820526, 0.16258662939071655, 0.019487546756863594, -0.2284373939037323, -0.2854868173599243, -0.04417099431157112, 0.37683507800102234, -0.08247721940279007, 0.033103324472904205, -0.19425275921821594, -0.07683047652244568, -0.25467318296432495, -0.14971138536930084, 0.49143555760383606, -0.17050717771053314, 0.16371405124664307, 0.05206683650612831, 0.15883444249629974, 0.3200910687446594, 0.346475213766098, -0.05128033086657524, 0.23605962097644806, 0.0744529590010643, 0.3285943567752838, 0.47833213210105896, -0.27670934796333313, 0.05458498001098633, 0.14386841654777527, -0.04974459484219551, -0.01411892008036375, 0.055780213326215744, 0.00474648829549551, 0.11313492059707642, -0.18608157336711884, 0.11334670335054398, 0.174922376871109, -0.10872144997119904, 0.2992881238460541, -0.4459868371486664, -0.04967779666185379, 0.3986055850982666, -0.20753492414951324, -0.08162340521812439, 0.05908762663602829, 0.010010628961026669, 0.05539286881685257, 0.01792507991194725, -0.08591949194669724, 0.3454482853412628, 0.29823562502861023, -0.12419234216213226, -0.10706939548254013, 0.08703389763832092, -0.09944067895412445, -0.44471409916877747, 0.03243102878332138, -0.1577640026807785, 0.12304985523223877, 0.050344910472631454, -0.0975474938750267, -0.2944735884666443, 0.09615019708871841, -0.03822631016373634, -0.25051963329315186, 0.15453067421913147, 0.023196587339043617, 0.17318350076675415, 0.05925016105175018, -0.2021060585975647, 0.06613398343324661, -0.015743911266326904, 0.10741586983203888, -0.351253479719162, -0.17887376248836517, -0.044257212430238724, -0.3509327173233032, -0.03988877683877945, 0.23988406360149384, 0.2370176613330841, 0.1002112627029419, -0.25856703519821167, 0.23891878128051758, -0.0016579184448346496, 0.3618727922439575, -0.2429504543542862, 0.087574802339077, -0.04195716977119446, -0.26053425669670105, 0.1956302374601364, 0.2751902639865875, 0.35893139243125916, -0.29758116602897644, -0.4221459627151489, -0.006539704278111458, 0.0379456952214241, 0.5811734795570374, -0.044700413942337036, 0.13578718900680542, 0.033680785447359085, -0.07959777861833572, 0.13727375864982605, -0.10519855469465256, 0.1780073344707489, -0.0735727995634079, 0.06387805938720703, 0.3371363878250122, -0.2674143612384796, 0.31074702739715576, 0.42854443192481995, 0.2990174889564514, 0.11164850741624832, 0.05712142586708069, -0.29339683055877686, -0.13686704635620117, 0.3519337773323059, 0.13247939944267273, 0.45555809140205383, -0.015704968944191933, 0.20980578660964966, -0.07270622253417969, -0.1649341732263565, -0.20099230110645294, 0.3175549805164337, -0.1493179053068161, 0.19674518704414368, 0.06621222198009491, 0.45941340923309326, 0.020023269578814507, -0.21954306960105896, 0.21167410910129547, 0.036517925560474396, 0.2187831699848175, -0.378315806388855, 0.2838192582130432, -0.28597819805145264, -0.07724259793758392, -0.43136969208717346, -0.04244379699230194, -0.04646889492869377, -0.565082848072052, -0.03836679086089134, 0.007924099452793598, -0.09303852915763855, 0.036544762551784515, -0.3283859193325043, -0.11925545334815979, -0.12170161306858063, -0.06536497175693512, -0.2936859130859375, -0.4041464030742645, -0.35029909014701843, -0.05848395451903343, 0.4213874638080597, 0.3856319189071655, 0.19977723062038422, 0.01642858237028122, -0.005452368408441544, -0.02815060503780842, -0.4237631559371948, 0.047018684446811676, -0.22188636660575867, 0.043297119438648224, 0.16059298813343048, -0.005856915842741728, 0.05179949477314949, -0.10299549251794815, 0.36970099806785583, -0.05113082379102707, -0.2518925070762634, 0.3521766662597656, 0.02231474034488201, -0.03763404116034508, -0.2629382312297821, -0.4058383107185364, 0.1166725605726242, -0.5667975544929504, 0.10078790783882141, -0.0469367615878582, 0.20174561440944672, 0.17167185246944427, 0.08207482099533081, 0.22550299763679504, -0.06525611132383347, 0.19515204429626465, 0.28039228916168213, 0.37839585542678833, 0.5403062105178833, -0.22598020732402802, -0.32307562232017517, 0.2950862646102905, -0.11089568585157394, 0.3608185648918152, 0.23071341216564178, -0.15202011168003082, 0.40528711676597595, -0.41035395860671997, 0.17004750669002533, -0.08374345302581787, -0.11723299324512482, 0.31236788630485535, 0.190455362200737, -0.0683734342455864, -0.09682302922010422, 0.017994465306401253, 0.21361719071865082, -0.2014838308095932, 0.16506949067115784, 0.008490102365612984, 0.19980493187904358, -0.10333242267370224, 0.3400722146034241, 0.12450022250413895, -0.16565261781215668, 0.36143723130226135, 0.011600097641348839, 0.012079466134309769, -0.2510288655757904, -0.1440342366695404, 0.019972167909145355, -0.21486856043338776, 0.061621807515621185, -0.04792586714029312, 0.056544143706560135, -0.4713306725025177, -0.0346459336578846, 0.23681169748306274, -0.147640198469162, -0.16390952467918396, 0.3225655257701874, -0.3387588858604431, 0.28037068247795105, 0.09475880861282349, 0.13657431304454803, -0.11946523934602737, -0.40645861625671387, 0.023601563647389412, -0.05851231887936592, 0.10876714438199997, -0.016414441168308258, -0.2646564841270447, 0.002610077615827322, -0.5347053408622742, 0.2495778352022171, 0.05217316374182701, 0.3960880637168884, -0.04335794225335121, -0.18558385968208313, 0.18215163052082062, -0.2292850762605667, 0.4075528681278229, -0.31268811225891113, 0.01877390407025814, 0.285961389541626, 0.29861000180244446, -0.4982850253582001, 0.18488070368766785, -0.2240576148033142, 0.33020952343940735, 0.22574815154075623, 0.18554334342479706, -0.3720526397228241, -0.3285733461380005, 0.2298184037208557, 0.04473932459950447, -0.0006503329495899379, 0.10367891192436218, -0.35246676206588745, -0.05329129099845886, -0.06974101811647415, -0.14215707778930664, 0.2665815055370331, 0.3564808666706085, -0.07092572003602982, 0.007949869148433208, -0.43612489104270935, 0.1583183854818344, 0.15794974565505981, 0.1309376060962677, 0.3880293667316437, -0.3398948013782501, 0.12485559284687042, -0.1658342182636261, 0.19165848195552826, -0.04550092667341232, 0.42883384227752686, -0.2038116604089737, -0.19006265699863434, 0.3096033036708832, -0.1745517998933792, 0.4862302243709564, 0.3790242373943329, 0.04215044155716896, 0.3114556670188904, -0.2637413740158081, -0.17829710245132446, -0.27737149596214294, -0.04351421818137169, 0.1307753175497055, -0.08891061693429947, 0.0912306010723114, -0.2324364334344864, 0.5531402230262756, 0.3016226589679718, -0.05554787442088127, 0.402709037065506, 0.18882033228874207, -0.35648295283317566, 0.46582669019699097, 0.2532045543193817, 1.1186988353729248, 0.10575833171606064, 0.22237159311771393, 0.025403855368494987, -0.4089376926422119, 0.35305437445640564, -0.03306611254811287, 0.27251768112182617, -0.49555546045303345, 0.22722499072551727, 0.02778870053589344, -0.12293685227632523, 0.07964378595352173, 0.05283230543136597, -0.2791069746017456, 0.15774792432785034, -0.19735132157802582, 0.33613041043281555, 0.07588723301887512, 0.12650927901268005, -0.4026111960411072, -0.3452477753162384, -0.48598966002464294, 0.08268708735704422, -0.0839156061410904, -0.08008492738008499, 0.026806294918060303, -0.22409138083457947, -0.4069483280181885, -0.13697312772274017, -0.13860879838466644, 0.1344418078660965, -0.15800288319587708, -0.11935333162546158, 0.2568901777267456, -0.15448161959648132, 0.1984223872423172, 0.22811508178710938, 0.044915758073329926, -0.05061230808496475, -0.16760925948619843, -0.014997882768511772, 0.27013689279556274, 0.119290791451931, 0.14295729994773865, -0.02716231346130371, 0.07144752889871597, 0.04230803996324539, 0.08588617295026779, 0.23091283440589905, -0.41005608439445496, 0.011083422228693962, -0.12113763391971588, 0.2061464637517929, -0.02342335507273674, -0.13388706743717194, -0.08782460540533066, 0.1156163364648819, -0.034163787961006165, -0.08314420282840729, 0.11682289093732834, 0.02967080846428871, 0.11227322369813919, 0.09993942826986313, -0.15056654810905457, -0.2468956857919693, 0.22061583399772644, 0.31885451078414917, 0.05648946389555931, 0.23681750893592834, 0.6745467782020569, 0.08035581558942795, -0.1407400518655777, -0.23257209360599518, 0.04807364195585251, 0.05120456591248512, -0.7669649124145508, 0.3388831615447998, -0.26796698570251465, -0.4024945795536041, 0.13678430020809174, 0.5917766690254211, -0.18555505573749542, 0.013741085305809975, -0.09446477144956589, -0.26656293869018555, -0.0999530628323555, 0.19291791319847107, 0.031858328729867935, 0.07356467843055725, -0.47675418853759766, -0.12634235620498657, 0.077013298869133, 0.2244807481765747, -0.26808497309684753, -0.04891272261738777, -0.23274007439613342, 0.37962934374809265, 0.014470371417701244, 0.08027582615613937, -0.029821699485182762, -0.15083588659763336, -0.006759923417121172, 0.4313047528266907, -0.1437537670135498, -0.15116289258003235, -0.27181383967399597, 0.13702808320522308, 0.07310252636671066, -0.02484019286930561, 0.15791796147823334, 0.029813986271619797, -0.012405332177877426, -0.25564220547676086, -0.0019760739523917437, -0.18603654205799103, -0.22291263937950134, -0.028439858928322792, 0.058520592749118805, 0.28256914019584656, -0.20867182314395905, 0.12727826833724976, -0.1877673864364624, 0.07987657934427261, -0.33941298723220825, 0.07165087014436722, 0.0795845165848732, 0.01573120430111885, 0.03195768594741821, 0.09080973267555237, -0.1425258219242096, -0.0751568153500557, 0.3140852451324463, -0.34542152285575867, -0.10453858226537704, 0.11357194930315018, -0.11710692942142487, 0.3987213969230652, -0.08429963141679764, 0.0142188910394907, -0.02621438354253769, 0.132892906665802, 0.07714112102985382, 0.2576908767223358, -0.27105873823165894, -0.12639227509498596, -0.051077645272016525, 0.30452293157577515, 0.17767731845378876, -0.21350979804992676, -0.040585022419691086, 0.12846365571022034, 0.05426356941461563, -0.39167746901512146, -0.14858029782772064, 0.08976992964744568, -0.09947672486305237, 0.007058172952383757, 0.3367122709751129, -0.11197759211063385, 0.02153922989964485, 0.08279450237751007, 0.4989663362503052, 0.45821982622146606, 0.3991211950778961, 0.1686481237411499, -0.1109297052025795, -0.0017289536772295833, 0.021545443683862686, 0.3171042203903198, 0.3153477907180786, 0.3045444190502167, 0.15489110350608826, 0.5104096531867981, -0.3058183193206787, -0.26720649003982544, -0.09200482815504074, 0.06214466318488121, -0.2543641924858093, -0.14574840664863586, -0.5064582228660583, -0.14190508425235748, -0.4231628477573395, -0.10091531276702881, -0.18652161955833435, 0.2660883069038391, 0.5050708651542664, 0.17831450700759888, 0.002986667212098837, -0.5828505158424377, -0.05613236129283905, 0.031109195202589035, 0.39658311009407043, -0.24684008955955505, 0.5015378594398499, -0.034852705895900726, -0.09850364178419113, -0.0982264056801796, 0.5166689157485962, 0.4602739214897156, 0.658098042011261, -0.2757396399974823, -0.07016029953956604, 0.0755322054028511, -0.1338510811328888, 0.15214860439300537, 0.560676097869873, 0.3276122510433197, 0.1597321629524231, 0.119194395840168, -0.034418873488903046, -0.09463370591402054, -0.11874871701002121, -0.03437414765357971, 0.2678045332431793, -0.09244730323553085, 0.3559325039386749, -0.15395519137382507, -0.2433711439371109, -0.08671598136425018, -0.15515699982643127, -0.2883930504322052, -0.05635581165552139, 0.131995290517807, -0.07147859781980515, 0.054644398391246796, -0.4272514879703522, 0.07915181666612625, 0.20936912298202515, 0.6322180032730103, 0.2378585934638977, 0.16091451048851013, -0.4584113657474518, -0.0798666924238205, -0.7588366270065308, 0.1408187597990036, -0.3934159278869629, -0.16126668453216553, 0.026848670095205307, 0.3510177731513977, -0.014346035197377205, 0.22938190400600433, -0.05591055750846863, 0.32126232981681824, 0.17825938761234283, 0.29825806617736816, -0.3454141616821289, -0.17916670441627502, 0.2205495834350586, -0.34437111020088196, -0.04905865341424942, -0.6017597317695618, 0.3270026743412018, 0.05008573457598686, 0.08173463493585587, -0.0072829569689929485, -0.22223277390003204, 0.33629393577575684, 0.28822779655456543, 0.32577577233314514, 0.16702665388584137, 0.28842398524284363, 0.19261880218982697, -0.4587063193321228, -0.39807507395744324, -0.19980397820472717, -0.30666327476501465, -0.09984147548675537, 0.13190890848636627, 0.469083309173584, -0.17963816225528717, 0.11805962771177292, -0.2406579852104187, 0.25994083285331726, -0.10598070174455643, 0.1154150664806366, -0.15053409337997437, 0.1351410150527954, -0.32362595200538635, 0.11830664426088333, 0.2612651586532593, 0.16057147085666656, 0.13314561545848846, 0.12775252759456635, -0.4737321138381958, -0.44358330965042114, 0.28088903427124023, -0.382368266582489, -0.38338354229927063, 0.2526189386844635, 0.22172559797763824, 0.39790093898773193, -0.09128426015377045, -0.7310709953308105, -0.30311861634254456, 0.13606230914592743, -0.08944423496723175, 0.004416736774146557, 0.06323638558387756, 0.0770341157913208, -0.07732492685317993, -0.21661345660686493, 0.28396815061569214, -0.055229928344488144, 0.026483088731765747, 0.05415172129869461, -0.25348347425460815 ]
https://github.com/huggingface/datasets/issues/207
Remove test set from NLP viewer
Appears that [two thirds of those polled on Twitter](https://twitter.com/srush_nlp/status/1265734497632477185) are in favor of _some_ mechanism for averting eyeballs from the test data.
While the new [NLP viewer](https://huggingface.co/nlp/viewer/) is a great tool, I think it would be best to outright remove the option of looking at the test sets. At the very least, a warning should be displayed to users before showing the test set. Newcomers to the field might not be aware of best practices, and small things like this can help increase awareness.
22
Remove test set from NLP viewer While the new [NLP viewer](https://huggingface.co/nlp/viewer/) is a great tool, I think it would be best to outright remove the option of looking at the test sets. At the very least, a warning should be displayed to users before showing the test set. Newcomers to the field might not be aware of best practices, and small things like this can help increase awareness. Appears that [two thirds of those polled on Twitter](https://twitter.com/srush_nlp/status/1265734497632477185) are in favor of _some_ mechanism for averting eyeballs from the test data.
[ -0.13736309111118317, 0.17048241198062897, -0.19007155299186707, -0.2922055423259735, -0.17911377549171448, 0.04491616040468216, 0.2525574564933777, 0.5567144751548767, 0.05526884272694588, 0.05124494805932045, 0.014377610757946968, 0.1753867268562317, -0.36144739389419556, 0.08514582365751266, 0.07604800164699554, 0.1644965261220932, -0.3026103973388672, 0.5070192813873291, -0.1310981810092926, 0.12752597033977509, -0.22304971516132355, -0.20796167850494385, 0.0001338271686108783, 0.24395659565925598, 0.18622946739196777, -0.3947819769382477, 0.187974214553833, -0.16565647721290588, -0.01923190802335739, -0.5875658392906189, -0.1071891337633133, 0.20037777721881866, -0.17094111442565918, -0.11505236476659775, -0.00009701008821139112, 0.0429435595870018, 0.16814827919006348, 0.12773188948631287, -0.2890508472919464, 0.10696448385715485, -0.12031278759241104, -0.058830853551626205, 0.2786581516265869, -0.13759101927280426, -0.08296416699886322, -0.22412878274917603, 0.14215503633022308, -0.22430960834026337, 0.1030537337064743, 0.35247379541397095, 0.3387938141822815, 0.35248619318008423, -0.30852431058883667, 0.4340731203556061, 0.13730216026306152, -0.011704793199896812, -0.2752701938152313, 0.04518892616033554, 0.03301114961504936, 0.015701161697506905, -0.3985557556152344, 0.5405017733573914, -0.06409753859043121, -0.007953503169119358, -0.038286928087472916, -0.17734363675117493, 0.39336293935775757, -0.2694500684738159, -0.30618682503700256, 0.08690247684717178, 0.19845092296600342, -0.1947709619998932, -0.062350787222385406, -0.41920173168182373, 0.23830950260162354, 0.04672984033823013, 0.09231418371200562, 0.2671627104282379, -0.058676376938819885, 0.2318631410598755, -0.5527656674385071, -0.41144225001335144, -0.11093438416719437, 0.08389531075954437, 0.14198194444179535, -0.10454492270946503, 0.08363674581050873, -0.056577373296022415, 0.35432329773902893, -0.08692149817943573, 0.1727447509765625, -0.05581134185194969, -0.25806012749671936, 0.014445465989410877, -0.005055202171206474, -0.5017731785774231, 0.0999644473195076, -0.0190570168197155, -0.0335976667702198, 0.2049787938594818, 0.08029267936944962, 0.10392972081899643, -0.20464003086090088, 0.06354318559169769, 0.10912445932626724, 0.12340128421783447, 0.35100439190864563, 0.06238699331879616, 0.2597541809082031, 0.2474016696214676, 0.24320942163467407, 0.30436408519744873, 0.3376288115978241, -0.25187209248542786, -0.2752174139022827, 0.1266622394323349, 0.3540286719799042, -0.23466217517852783, -0.34609290957450867, 0.0034455526620149612, -0.5296182036399841, 0.014235141687095165, 0.014480259269475937, 0.10928118228912354, -0.0006177048780955374, 0.12058725208044052, 0.035311464220285416, 0.03173805773258209, -0.36195671558380127, -0.35781925916671753, -0.054423511028289795, 0.10798407346010208, -0.09068254381418228, 0.1193414106965065, 0.4057035744190216, 0.043408170342445374, 0.16046151518821716, -0.03420179337263107, 0.09332382678985596, 0.040589589625597, 0.3698309361934662, 0.02462463639676571, 0.3463379442691803, -0.057568036019802094, -0.11738015711307526, -0.21008720993995667, -0.2158709466457367, -0.14201435446739197, -0.18773360550403595, 0.15288551151752472, 0.09354343265295029, -0.20848298072814941, 0.052669133991003036, 0.35050898790359497, -0.4395666718482971, -0.014217783696949482, 0.0917811170220375, 0.23471501469612122, -0.3213297128677368, -0.04506218805909157, 0.14012688398361206, -0.17458830773830414, -0.13457627594470978, 0.073859803378582, -0.17334561049938202, 0.2905425429344177, -0.09705211222171783, -0.08341508358716965, -0.03997648507356644, -0.09570499509572983, -0.20757614076137543, -0.024172253906726837, -0.4004933834075928, 0.1751304715871811, -0.2039685845375061, -0.06718969345092773, 0.3619678318500519, -0.2131473571062088, -0.10236624628305435, 0.2003476470708847, -0.23160776495933533, 0.016959914937615395, -0.14347781240940094, 0.07153677940368652, -0.16221840679645538, -0.4502270519733429, -0.47924813628196716, 0.0716373473405838, 0.06016077846288681, 0.14337126910686493, -0.44935888051986694, -0.42560943961143494, 0.4917643666267395, 0.04578416794538498, -0.05306456983089447, -0.1359756737947464, -0.1581977754831314, 0.25878414511680603, 0.4374868869781494, 0.028907835483551025, -0.03932274505496025, -0.23834086954593658, 0.2083147168159485, -0.3265896439552307, -0.2815675437450409, 0.00792440865188837, -0.13267399370670319, -0.16728749871253967, -0.08347085118293762, 0.18698106706142426, 0.4207231104373932, -0.2829241156578064, -0.09006613492965698, -0.09678644686937332, -0.15172308683395386, -0.09776090085506439, 0.29000481963157654, -0.14137011766433716, -0.09831620007753372, 0.13745425641536713, 0.09653115272521973, -0.1579170823097229, -0.17216801643371582, 0.038154978305101395, 0.020720187574625015, -0.17654553055763245, -0.05305396392941475, 0.10111177712678909, 0.31183338165283203, 0.3026479482650757, 0.07121872901916504, 0.06072955206036568, 0.07649943977594376, 0.17240816354751587, 0.06634186953306198, 0.10490531474351883, 0.23246608674526215, 0.27824151515960693, 0.07447241991758347, -0.44166043400764465, -0.01435122825205326, 0.13553407788276672, -0.15129214525222778, 0.23665398359298706, -0.3167857527732849, 0.25637221336364746, 0.1614445000886917, -0.07265815883874893, 0.003108428791165352, -0.029503514990210533, 0.048931822180747986, -0.33980298042297363, -0.13817612826824188, -0.3378734886646271, -0.02647312730550766, 0.2268938273191452, -0.18623073399066925, 0.23691892623901367, -0.33784252405166626, 0.17116950452327728, 0.09643427282571793, -0.015973854809999466, -0.026087403297424316, 0.10548385232686996, 0.2049354612827301, -0.35236161947250366, 0.23605777323246002, 0.0762559249997139, 0.10919374972581863, 0.3495018184185028, 0.22472383081912994, 0.17187008261680603, -0.21009552478790283, -0.2726421356201172, 0.31514957547187805, 0.3302740454673767, 0.2265135496854782, -0.024842604994773865, 0.24631288647651672, -0.16664554178714752, -0.24240075051784515, 0.07095260918140411, -0.10728862881660461, -0.05060568451881409, -0.37377095222473145, 0.024834895506501198, -0.14935225248336792, -0.5540047287940979, -0.1692960411310196, 0.19525079429149628, -0.09234829246997833, -0.2506902515888214, 0.5380778312683105, -0.028284285217523575, -0.3674876391887665, 0.22021853923797607, -0.29027318954467773, 0.6584699153900146, -0.20973874628543854, 0.2230774164199829, -0.06582248210906982, -0.020816441625356674, -0.36914482712745667, 0.3089110851287842, 0.17983654141426086, 0.06083407998085022, 0.5227044820785522, 0.10138288140296936, -0.17568477988243103, -0.17620231211185455, -0.272009938955307, 0.08650889247655869, -0.03296928107738495, 0.3253639042377472, -0.0012538830051198602, -0.16571028530597687, -0.04760980233550072, -0.14837419986724854, -0.4186190962791443, -0.22404420375823975, -0.00045906705781817436, -0.1523250937461853, 0.1457800269126892, 0.3848962187767029, -0.24102413654327393, -0.46456432342529297, 0.0017176615074276924, -0.21992281079292297, 0.020160358399152756, -0.014971165917813778, 0.20743195712566376, 0.20418451726436615, -0.2656363844871521, 0.005003551486879587, -0.14018693566322327, -0.05355378985404968, -0.11937110871076584, 0.04450145736336708, 0.012334302999079227, -0.270201712846756, -0.28076621890068054, 0.019245941191911697, -0.2571012079715729, 0.1463608294725418, -0.17370162904262543, -0.6786444783210754, -0.32688578963279724, 0.18293803930282593, -0.09719084203243256, -0.04581768810749054, -0.5455363392829895, -0.11407420039176941, 0.1751355528831482, -0.262187659740448, 0.12555097043514252, 0.08653390407562256, 0.2562341094017029, 0.0054036714136600494, 0.14535996317863464, -0.10717689245939255, -0.10614803433418274, 0.26796677708625793, 0.1563168168067932, 0.2747742235660553, 0.11752189695835114, -0.007231594063341618, 0.08923893421888351, 0.260608434677124, -0.008756565861403942, -0.03511315584182739, 0.04626874625682831, 0.1289527416229248, 0.01569892093539238, 0.33214277029037476, -0.11922289431095123, -0.31336134672164917, 0.05188710615038872, 0.08164607733488083, -0.21345077455043793, -0.40376412868499756, 0.14978037774562836, -0.4017370641231537, 0.1390129029750824, 0.12182235717773438, 0.04311315715312958, 0.06124291941523552, -0.5748981237411499, 0.058783967047929764, -0.19713015854358673, 0.22688190639019012, -0.06648273766040802, -0.2071051299571991, -0.013696732930839062, -0.27502650022506714, -0.011230655945837498, 0.1462322175502777, 0.384588360786438, 0.01064359862357378, -0.2336488962173462, 0.24429193139076233, 0.16834469139575958, 0.21813783049583435, 0.06173032894730568, 0.20189164578914642, 0.13025884330272675, -0.12362939119338989, -0.30144646763801575, 0.18793058395385742, -0.21928907930850983, -0.09623338282108307, 0.15762045979499817, -0.02711099572479725, 0.029720239341259003, -0.24706539511680603, 0.3698942959308624, 0.12154686450958252, -0.2316068857908249, 0.02466561459004879, -0.013284413143992424, 0.010629900731146336, -0.03718144819140434, 0.24638673663139343, -0.1701391637325287, -0.02812204509973526, -0.3634011149406433, -0.1015850156545639, -0.024530071765184402, -0.20282815396785736, 0.3764398396015167, 0.28827857971191406, -0.01591493934392929, 0.27267417311668396, 0.09733467549085617, 0.2122037261724472, 0.03915722668170929, -0.1911204308271408, 0.18984751403331757, 0.2478581815958023, 0.3424510955810547, -0.044840577989816666, -0.20925548672676086, 0.34741106629371643, 0.4122863709926605, 0.13903388381004333, 0.0013977538328617811, -0.38751837611198425, 0.05723695829510689, -0.2808184027671814, 0.2900630831718445, 0.07467538118362427, -0.17464786767959595, 0.05383197218179703, -0.1256577968597412, 0.24363839626312256, -0.12188340723514557, -0.36395180225372314, 0.0006386784370988607, 0.11594981700181961, -0.18264415860176086, 0.2092994600534439, 0.4568054974079132, 0.7854481339454651, 0.04531301558017731, 0.19970214366912842, 0.47228071093559265, -0.1578017771244049, 0.43638351559638977, -0.1543489247560501, 0.2508733868598938, -0.1341296136379242, 0.09534647315740585, 0.05953192710876465, -0.1178390309214592, 0.11199314147233963, 0.2078068107366562, 0.002055149059742689, 0.0006367655005306005, 0.26226556301116943, -0.3583506643772125, -0.11938635259866714, 0.3215343654155731, 0.04634006693959236, -0.021234925836324692, -0.30379751324653625, 0.2726382315158844, -0.16115760803222656, -0.20504239201545715, 0.10787377506494522, 0.028086073696613312, 0.3095529079437256, 0.04540582373738289, 0.18522466719150543, 0.1358148753643036, 0.11219006031751633, 0.1330634206533432, -0.13726204633712769, -0.18840186297893524, 0.3965901732444763, 0.23549354076385498, -0.1709231585264206, 0.12436362355947495, -0.08929352462291718, -0.012548196129500866, -0.1636730581521988, 0.2965008616447449, 0.21388866007328033, 0.28330379724502563, 0.3150733411312103, -0.1993122100830078, -0.17023946344852448, 0.264915406703949, 0.19701603055000305, 0.059723980724811554, -0.08262798190116882, -0.005859395023435354, 0.21425308287143707, -0.47217732667922974, 0.03237330541014671, 0.03920373693108559, 0.11395228654146194, -0.15051233768463135, 0.28724604845046997, 0.04401303082704544, -0.038295138627290726, 0.04322884604334831, 0.3496761620044708, -0.23045501112937927, 0.10065090656280518, 0.05798110365867615, 0.09893631935119629, 0.10475975275039673, 0.4397662878036499, -0.16813936829566956, -0.24870665371418, -0.4134775996208191, 0.259731650352478, -0.12139647454023361, -0.2759037911891937, 0.10330574214458466, 0.12412769347429276, -0.0599198192358017, -0.036588091403245926, 0.21929119527339935, -0.10416185855865479, -0.00694270571693778, -0.09856217354536057, 0.004728649277240038, 0.0028258066158741713, 0.1621440052986145, -0.13531611859798431, 0.26489511132240295, 0.23671163618564606, 0.17586511373519897, -0.07974201440811157, 0.16820965707302094, -0.4586579501628876, -0.23570460081100464, -0.36296021938323975, 0.16979293525218964, -0.3387686014175415, 0.1633303463459015, 0.025976600125432014, 0.0013713031075894833, 0.13112425804138184, -0.2586059272289276, -0.23900224268436432, -0.3924417197704315, -0.11555536836385727, 0.05777059122920036, -0.014545128680765629, -0.1714857667684555, 0.0207369402050972, 0.011869668960571289, 0.10929109901189804, 0.05694764852523804, -0.11095643043518066, -0.3612697124481201, 0.1871645748615265, 0.6549925208091736, -0.22639766335487366, -0.37684395909309387, 0.04805314168334007, 0.0021647317335009575, -0.33640173077583313, 0.18045571446418762, 0.19234509766101837, 0.25698819756507874, 0.07886822521686554, -0.07736138999462128, 0.09211722016334534, 0.14207778871059418, -0.0038997551891952753, 0.24301698803901672, -0.22079457342624664, 0.056131791323423386, 0.016851594671607018, -0.029933232814073563, -0.025893164798617363, -0.027973689138889313, -0.3015800714492798, -0.29457855224609375, 0.2904885411262512, 0.3381890654563904, -0.31036195158958435, 0.027729492634534836, 0.3251906931400299, 0.04844783991575241, 0.07958592474460602, 0.21352927386760712, 0.08040444552898407, 0.02968447655439377, 0.295149564743042, 0.030976224690675735, -0.00004258426997694187, -0.32472988963127136, 0.08860588818788528, 0.16464734077453613, -0.41913872957229614, -0.2750464677810669, 0.12339724600315094, 0.1220627948641777, 0.17781174182891846, -0.019305448979139328, 0.19298267364501953, 0.601468563079834, -0.04290498420596123, 0.3327937424182892, -0.029134884476661682, -0.04987712949514389, -0.06779879331588745, 0.006361773703247309, -0.17296257615089417, 0.08005885034799576, 0.3721136152744293, 0.3768928349018097, 0.2283656895160675, -0.12036123871803284, -0.06467608362436295, 0.04855969920754433, 0.017171278595924377, 0.17235636711120605, -0.2735922336578369, 0.11902850866317749, -0.09038832783699036, 0.1302608847618103, 0.07079992443323135, -0.209639772772789, 0.10960591584444046, 0.11861130595207214, -0.10826563090085983, -0.10142280161380768, 0.1704082190990448, -0.03653175011277199, 0.25225892663002014, -0.09298396855592728, -0.19044877588748932, -0.0768478587269783, 0.34443750977516174, 0.16699548065662384, 0.3699306845664978, -0.14906471967697144, -0.18440505862236023, -0.02608889527618885, -0.18871255218982697, -0.09965503960847855, -0.009150553494691849, -0.3066663444042206, 0.26656508445739746, 0.07575138658285141, 0.14168745279312134, 0.25459790229797363, 0.21736067533493042, -0.09045793116092682, 0.059996020048856735, 0.3468180298805237, -0.22026768326759338, -0.2666853964328766, 0.2640351951122284, 0.5363552570343018, -0.33223021030426025, -0.06163092702627182, -0.1159563958644867, 0.059843096882104874, 0.10731934010982513, 0.2752535343170166, -0.24124665558338165, 0.09731150418519974, -0.2812126874923706, 0.15977290272712708, 0.2776830494403839, 0.43047937750816345, 0.005765522830188274, 0.17219898104667664, -0.09262137115001678, -0.07269597798585892, -0.5211899280548096, 0.18362869322299957, 0.22089840471744537, -0.14337532222270966, 0.2590456008911133, -0.026720397174358368, -0.18254782259464264, 0.2621278464794159, -0.3430917263031006, 0.19634002447128296, -0.03144639730453491, -0.2231312245130539, -0.2514566481113434, 0.035639677196741104, -0.03186836838722229, 0.2519476115703583, -0.17884774506092072, -0.10400588810443878, 0.20293278992176056, -0.18828240036964417, 0.158780038356781, 0.06955959647893906, 0.0849524736404419, 0.14256049692630768, 0.2857637107372284, 0.2812420129776001, 0.16929619014263153, 0.12046943604946136, -0.03955947235226631, 0.09596670418977737, 0.0005992946098558605, -0.018906015902757645, -0.18427997827529907, -0.20551955699920654, 0.09908095747232437, 0.19717472791671753, -0.09581457823514938, 0.2707025706768036, 0.12599599361419678, 0.16755004227161407, -0.027381159365177155, 0.22046241164207458, -0.11364404857158661, 0.1647745817899704, 0.03281619772315025, 0.29282325506210327, -0.12845435738563538, 0.22897577285766602, -0.09584255516529083, -0.1429789513349533, -0.07893220335245132, -0.19231905043125153, 0.2561526298522949, -0.29874616861343384, -0.04730799421668053, 0.02888866700232029, 0.37274402379989624, 0.009692187421023846, -0.31185078620910645, -0.45354220271110535, -0.28493228554725647, 0.2458088994026184, -0.08283800631761551, 0.24263308942317963, 0.44049498438835144, -0.11160431802272797, -0.1418987214565277, -0.08976597338914871, 0.31003081798553467, 0.07364293932914734, 0.13642029464244843, 0.07103078067302704, -0.34688350558280945 ]
https://github.com/huggingface/datasets/issues/206
[Question] Combine 2 datasets which have the same columns
We are thinking about ways to combine datasets for T5 in #217, feel free to share your thoughts about this.
Hi, I am using ``nlp`` to load personal datasets. I created summarization datasets in multi-languages based on wikinews. I have one dataset for english and one for german (french is getting to be ready as well). I want to keep these datasets independent because they need different pre-processing (add different task-specific prefixes for T5 : *summarize:* for english and *zusammenfassen:* for german) My issue is that I want to train T5 on the combined english and german datasets to see if it improves results. So I would like to combine 2 datasets (which have the same columns) to make one and train T5 on it. I was wondering if there is a proper way to do it? I assume that it can be done by combining all examples of each dataset but maybe you have a better solution. Hoping this is clear enough, Thanks a lot 😊 Best
20
[Question] Combine 2 datasets which have the same columns Hi, I am using ``nlp`` to load personal datasets. I created summarization datasets in multi-languages based on wikinews. I have one dataset for english and one for german (french is getting to be ready as well). I want to keep these datasets independent because they need different pre-processing (add different task-specific prefixes for T5 : *summarize:* for english and *zusammenfassen:* for german) My issue is that I want to train T5 on the combined english and german datasets to see if it improves results. So I would like to combine 2 datasets (which have the same columns) to make one and train T5 on it. I was wondering if there is a proper way to do it? I assume that it can be done by combining all examples of each dataset but maybe you have a better solution. Hoping this is clear enough, Thanks a lot 😊 Best We are thinking about ways to combine datasets for T5 in #217, feel free to share your thoughts about this.
[ -0.24451382458209991, 0.06389869004487991, 0.024110980331897736, 0.362822949886322, 0.10962217301130295, 0.43875232338905334, 0.26998773217201233, 0.3603741228580475, 0.044978752732276917, -0.11644750833511353, -0.3814811110496521, 0.12374284118413925, 0.003628325881436467, 0.27198466658592224, 0.2713170647621155, -0.5642142295837402, 0.058044884353876114, 0.21875324845314026, -0.5200102925300598, 0.18503721058368683, 0.007575650233775377, 0.005126405972987413, -0.1654464304447174, 0.012677017599344254, -0.2030433565378189, 0.23494407534599304, -0.34226280450820923, -0.22793914377689362, 0.046384040266275406, 0.15664565563201904, 0.3019717037677765, 0.16707953810691833, 0.2146206796169281, 0.31491562724113464, -0.000119927448395174, -0.07949525862932205, -0.01648912951350212, -0.09052349627017975, 0.14189749956130981, -0.19872885942459106, 0.09682527929544449, -0.3619499206542969, -0.06335436552762985, -0.15658177435398102, 0.08227751404047012, -0.2395307719707489, -0.233183354139328, 0.03752850368618965, 0.2680857181549072, -0.16603007912635803, 0.05411143600940704, -0.18832089006900787, -0.2511498034000397, 0.34095093607902527, -0.07127469778060913, -0.1605684906244278, 0.061465684324502945, -0.06943436712026596, 0.08189096301794052, -0.31951552629470825, 0.5779259204864502, 0.21333351731300354, -0.035110339522361755, -0.11987728625535965, 0.08233312517404556, 0.14248372614383698, 0.14339230954647064, -0.1095089539885521, 0.0553346686065197, 0.19000254571437836, 0.7741909623146057, -0.0596495047211647, 0.008375687524676323, -0.291532963514328, 0.40387386083602905, -0.13154691457748413, -0.19321149587631226, 0.1430448591709137, 0.01704300008714199, 0.025498047471046448, -0.2548201084136963, -0.15631981194019318, -0.016406523063778877, 0.3023144006729126, -0.04316458851099014, 0.33158162236213684, 0.1102491095662117, 0.27707937359809875, -0.06372163444757462, -0.1514381468296051, 0.16603219509124756, -0.33642616868019104, 0.0009917286224663258, 0.1547575443983078, -0.30736345052719116, -0.4046191871166229, -0.3330448865890503, -0.18422916531562805, 0.2726253271102905, -0.15348786115646362, -0.1543884575366974, -0.006971367169171572, -0.10903947800397873, 0.13872353732585907, 0.3026828467845917, 0.15827211737632751, 0.13860982656478882, 0.21253810822963715, -0.18847210705280304, -0.31628310680389404, -0.3603149354457855, 0.16161008179187775, 0.07011718302965164, -0.07282983511686325, -0.12339496612548828, -0.2811371386051178, -0.05845237150788307, -0.3414455056190491, 0.14554014801979065, -0.19353166222572327, -0.2755388021469116, -0.23108279705047607, -0.102567158639431, -0.15161018073558807, -0.01819591224193573, 0.3452565670013428, -0.14899329841136932, 0.30911681056022644, -0.10607443749904633, -0.013937405310571194, -0.02503010630607605, 0.07381518930196762, -0.17078687250614166, 0.12826980650424957, 0.09456341713666916, 0.17032967507839203, -0.13089551031589508, 0.41861793398857117, -0.18310436606407166, 0.1015595942735672, 0.32089918851852417, -0.19957438111305237, -0.07107508927583694, -0.24224181473255157, 0.3129125237464905, 0.16381829977035522, 0.16308432817459106, 0.1007978543639183, -0.30066707730293274, 0.16731731593608856, -0.17366516590118408, -0.08247588574886322, 0.0157319363206625, 0.05949970707297325, -0.08394355326890945, -0.3012196719646454, -0.12381000816822052, 0.6368829607963562, 0.23763449490070343, -0.3169041574001312, -0.14158740639686584, 0.09555366635322571, -0.4641839861869812, -0.04789231717586517, 0.20856651663780212, 0.025119882076978683, -0.3501131534576416, -0.13585294783115387, 0.0010473888833075762, -0.00636476743966341, -0.050108857452869415, 0.28892144560813904, -0.26644930243492126, 0.20656637847423553, 0.09581867605447769, 0.06720889359712601, 0.6602528691291809, 0.017082922160625458, -0.21449244022369385, 0.22153355181217194, -0.15071648359298706, 0.13258783519268036, -0.10475068539381027, 0.11741260439157486, -0.37669840455055237, -0.0470036081969738, 0.1844799816608429, 0.838509738445282, -0.2917751669883728, -0.15610268712043762, -0.06377508491277695, -0.3239148259162903, 0.33796387910842896, 0.15537384152412415, 0.1021534651517868, -0.24988193809986115, -0.1425955593585968, 0.16948100924491882, 0.16629908978939056, -0.35606640577316284, 0.16821642220020294, 0.18711203336715698, -0.029124926775693893, 0.11734015494585037, -0.22763417661190033, -0.2857428789138794, -0.2659728527069092, -0.13448543846607208, 0.11240158975124359, 0.04761658236384392, 0.44605568051338196, -0.45091038942337036, 0.06634581834077835, -0.6603158712387085, -0.020744120702147484, 0.09107719361782074, -0.0005951409111730754, 0.31140586733818054, 0.10588061809539795, -0.2892330586910248, -0.17570403218269348, 0.3350968658924103, -0.052014146000146866, -0.16662946343421936, -0.1528075784444809, 0.4420304596424103, 0.03853129222989082, 0.0737515464425087, 0.00935167632997036, 0.5092475414276123, 0.04848816990852356, 0.03464546799659729, 0.14941996335983276, -0.12538082897663116, -0.5496702790260315, 0.18980063498020172, 0.18357008695602417, 0.028573859483003616, -0.15538720786571503, -0.06806933134794235, 0.3543195426464081, -0.33441996574401855, -0.005318505689501762, -0.16024212539196014, -0.07414712011814117, 0.4110817313194275, -0.24898134171962738, 0.36992505192756653, -0.02039080485701561, -0.11749543994665146, 0.01692468486726284, -0.15341854095458984, 0.021901628002524376, -0.10193340480327606, 0.2140548676252365, -0.007762819994240999, 0.013551082462072372, 0.34956732392311096, -0.38780564069747925, 0.24616017937660217, 0.28392359614372253, 0.1036929190158844, -0.04670192673802376, -0.01428297720849514, -0.14048494398593903, -0.047391895204782486, 0.03492921218276024, 0.17373283207416534, 0.5730594396591187, 0.19658632576465607, 0.20182695984840393, 0.23324693739414215, -0.15427322685718536, 0.12579281628131866, 0.014814486727118492, 0.1872757226228714, 0.06608539074659348, 0.30791226029396057, 0.09058137238025665, 0.2213161289691925, 0.18971006572246552, 0.3987441658973694, 0.20331713557243347, -0.09153178334236145, -0.19266954064369202, -0.005747878458350897, -0.2850170135498047, -0.6398733854293823, -0.5628807544708252, -0.12821359932422638, -0.22112107276916504, -0.05307042971253395, -0.09376536309719086, -0.0052851964719593525, 0.008729212917387486, 0.18447570502758026, -0.11454137414693832, 0.4212615191936493, -0.4750217795372009, -0.27784043550491333, 0.2787615954875946, 0.04162326082587242, -0.1818006932735443, 0.07090511173009872, 0.40793895721435547, 0.21213595569133759, 0.25454768538475037, -0.05741814151406288, -0.23279772698879242, -0.060105349868535995, -0.5316919684410095, 0.050194915384054184, -0.3654114007949829, -0.1488938182592392, -0.25188788771629333, -0.10150112211704254, -0.21497665345668793, -0.3766777515411377, 0.051559094339609146, 0.31181788444519043, 0.0527416355907917, -0.12870056927204132, 0.05726407840847969, 0.28694143891334534, -0.10500973463058472, -0.29453176259994507, -0.4214807152748108, -0.05987599864602089, 0.04238336533308029, -0.16191205382347107, -0.02081904374063015, -0.36425936222076416, -0.05395563319325447, -0.03650941699743271, -0.26720666885375977, -0.011379940435290337, -0.20162345468997955, 0.0008079031249508262, 0.1787797510623932, -0.13474680483341217, -0.12834617495536804, -0.08461173623800278, -0.18979676067829132, 0.299744576215744, -0.030815821141004562, -0.10249166935682297, -0.15170371532440186, -0.13994339108467102, -0.002392875263467431, -0.14289139211177826, 0.3774224817752838, 0.17890025675296783, 0.017671838402748108, 0.14856775104999542, 0.07189647108316422, -0.22946876287460327, 0.021115802228450775, 0.09855344146490097, 0.4574986398220062, -0.19675227999687195, 0.25292709469795227, -0.14573264122009277, 0.6077178716659546, 0.5213463306427002, 0.16695119440555573, 0.06605686247348785, 0.2555118799209595, 0.1424030214548111, 0.063320092856884, -0.20977582037448883, 0.20084504783153534, 0.19462954998016357, 0.057669442147016525, 0.26992395520210266, 0.10378686338663101, -0.053352240473032, -0.33397626876831055, 0.09006959199905396, -0.3042639195919037, -0.11738710850477219, 0.37775230407714844, -0.27257445454597473, 0.11911506205797195, 0.011027183383703232, -0.46415653824806213, -0.3556425869464874, -0.13096462190151215, -0.21647541224956512, 0.11839047819375992, -0.25855615735054016, 0.13618578016757965, -0.5622231960296631, -0.10364466160535812, -0.14068950712680817, 0.3238656520843506, 0.15147565305233002, 0.308350533246994, 0.04973016679286957, 0.055047407746315, -0.24609525501728058, -0.04382020980119705, 0.06486961245536804, -0.49241185188293457, -0.13473418354988098, 0.06497935950756073, 0.014051536098122597, -0.23107853531837463, -0.1758926957845688, 0.017376063391566277, -0.0913957804441452, -0.182997927069664, 0.3364604711532593, -0.31847742199897766, -0.15623876452445984, 0.6944689154624939, 0.43872082233428955, -0.07578422874212265, -0.3850884735584259, 0.04643484950065613, -0.1232316717505455, -0.1433953493833542, -0.26000750064849854, -0.04751051217317581, 0.16320332884788513, -0.36894550919532776, 0.1504460722208023, 0.01641148142516613, 0.08038848638534546, 0.45044881105422974, 0.10332699865102768, 0.0721360445022583, 0.352663516998291, 0.3876458406448364, -0.24028676748275757, -0.0815339908003807, -0.252618670463562, 0.4975561201572418, -0.07558069378137589, -0.3227156400680542, -0.17115619778633118, -0.14382457733154297, 0.7155781984329224, -0.06854794919490814, 0.15113170444965363, 0.23169198632240295, -0.39959779381752014, -0.3615082800388336, 0.0019529948476701975, 0.20284651219844818, 0.28261324763298035, 0.3221149146556854, -0.39712855219841003, -0.44975560903549194, 0.6233005523681641, 0.41063886880874634, -0.020718872547149658, 0.2541218400001526, -0.2573799788951874, -0.17415769398212433, 0.09945603460073471, -0.10148340463638306, 0.8141270875930786, 0.02181338146328926, 0.3815935254096985, -0.0004790184902958572, -0.1681196689605713, 0.7002719044685364, -0.09663413465023041, 0.019922100007534027, 0.02017255686223507, -0.14376352727413177, -0.035385943949222565, -0.03009582869708538, -0.12921667098999023, 0.37022170424461365, -0.20778588950634003, 0.317170649766922, -0.0004514182801358402, -0.005243171472102404, 0.0035990418400615454, 0.4091915786266327, 0.06669445335865021, -0.258772611618042, -0.3931369483470917, -0.03033839724957943, -0.302725613117218, -0.05387965217232704, -0.1444963961839676, 0.07147295027971268, -0.1313789039850235, 0.07047813385725021, -0.14826825261116028, 0.29996412992477417, 0.5066028833389282, 0.04458443075418472, -0.4117031395435333, -0.4415704309940338, 0.2938781976699829, 0.14138098061084747, 0.355612576007843, 0.029089447110891342, 0.029953358694911003, 0.014914832077920437, -0.35656362771987915, -0.01575484871864319, -0.11483701318502426, -0.2938969135284424, 0.2594216465950012, 0.020187411457300186, -0.25334009528160095, 0.06901312619447708, 0.4127650260925293, -0.08236540853977203, -0.254402756690979, -0.17855194211006165, 0.3314894139766693, -0.05946112424135208, -0.5873773097991943, 0.657164990901947, 0.15120774507522583, -0.07791898399591446, 0.010054511949419975, -0.11371666938066483, -0.2892288267612457, 0.38451629877090454, -0.2728395164012909, -0.2554662525653839, 0.06018338352441788, 0.18039080500602722, 0.30600014328956604, -0.3658265471458435, 0.4701858162879944, 0.1839188188314438, 0.27440986037254333, 0.022973431274294853, -0.06155139580368996, 0.2622482180595398, -0.07377304881811142, -0.35813164710998535, 0.10253539681434631, 0.20453286170959473, 0.11256226152181625, 0.15273042023181915, 0.16331599652767181, 0.1222742423415184, 0.2067319005727768, -0.06777968257665634, -0.3782423734664917, 0.2728556990623474, 0.20024091005325317, 0.2617899477481842, 0.25701791048049927, 0.3128030598163605, 0.06446409970521927, 0.6031779646873474, -0.16969521343708038, -0.019097592681646347, 0.13622424006462097, 0.285241961479187, 0.3465155363082886, 0.08450649678707123, -0.4070691764354706, 0.014368128031492233, -0.04422110319137573, 0.0018222432117909193, 0.11410626024007797, -0.14427243173122406, -0.07686248421669006, 0.1759718656539917, 0.34493157267570496, 0.0011230944655835629, 0.060232967138290405, -0.28074121475219727, -0.011822552420198917, -0.022508984431624413, 0.06651231646537781, -0.09257759898900986, -0.0002627990616019815, 0.3806403875350952, -0.07365352660417557, 0.03403816744685173, -0.2856472432613373, 0.26426956057548523, -0.13234993815422058, -0.03567854315042496, 0.259003609418869, 0.10173752903938293, 0.39166635274887085, -0.04448321461677551, -0.2572685778141022, -0.26786404848098755, 0.2171396166086197, 0.06505493819713593, -0.3455160856246948, -0.03496995195746422, 0.04609052091836929, -0.061143580824136734, 0.15212325751781464, 0.3250304162502289, -0.10721519589424133, -0.02506868727505207, 0.23458248376846313, 0.0007481424254365265, -0.19700078666210175, 0.0721050575375557, 0.23050765693187714, -0.06402736902236938, -0.14158546924591064, 0.21958497166633606, 0.09061215072870255, 0.37726137042045593, -0.311191588640213, -0.012799124233424664, -0.00748078990727663, -0.03604850545525551, 0.3243546783924103, 0.5464569926261902, -0.032420992851257324, -0.10312557220458984, -0.01200280524790287, 0.0809040293097496, 0.05886481702327728, -0.05017796531319618, -0.1465628743171692, -0.06746146082878113, -0.08968918770551682, -0.07963176816701889, 0.23124255239963531, -0.08268369734287262, 0.06398651748895645, 0.29144901037216187, 0.2242591679096222, 0.13219574093818665, 0.05987716093659401, 0.27598142623901367, 0.0057379514910280704, -0.46567654609680176, 0.027942780405282974, 0.41669222712516785, -0.07972192019224167, 0.0029163311701267958, -0.3966142237186432, -0.0038866051472723484, -0.1403137594461441, 0.2057415246963501, -0.1246994212269783, -0.3338139057159424, 0.5012155175209045, 0.09291654080152512, -0.14959865808486938, -0.2275860458612442, 0.12669208645820618, 0.5161341428756714, -0.035095371305942535, -0.04846221208572388, 0.052022967487573624, 0.16546906530857086, -0.12800754606723785, 0.03769982233643532, 0.25175711512565613, 0.110866978764534, 0.029335685074329376, 0.0011412217281758785, 0.024303121492266655, 0.14044035971164703, 0.1221235990524292, 0.10343042761087418, 0.019177811220288277, 0.08414570987224579, -0.016204580664634705, 0.2771758437156677, 0.10079753398895264, -0.061256252229213715, 0.07313242554664612, -0.12367559969425201, 0.04407620057463646, -0.5888593792915344, 0.2657963037490845, 0.3155157268047333, -0.023779546841979027, -0.05252768471837044, 0.17175747454166412, -0.3293575048446655, -0.18782657384872437, 0.5251405835151672, 0.2926931381225586, -0.15609630942344666, -0.1533709168434143, 0.035338155925273895, -0.029740452766418457, 0.3186293840408325, 0.35710588097572327, 0.08333612978458405, -0.42110368609428406, -0.18525874614715576, -0.7274482250213623, 0.11188368499279022, -0.1998966634273529, -0.11876030266284943, -0.2600691020488739, 0.04048542305827141, 0.2637617290019989, 0.2332361936569214, 0.06438446044921875, 0.43237563967704773, 0.141852468252182, 0.155785471200943, -0.43412336707115173, -0.1574392020702362, 0.08547874540090561, 0.18689297139644623, 0.05970935523509979, -0.11151757836341858, 0.024231454357504845, 0.35462549328804016, -0.18669790029525757, -0.0636829063296318, 0.019787950441241264, 0.13367408514022827, -0.0402185320854187, 0.23977096378803253, 0.03683681786060333, 0.6354691982269287, -0.16606871783733368, 0.09476187825202942, 0.11244634538888931, -0.05096951872110367, -0.5395124554634094, 0.3559368848800659, -0.15957602858543396, 0.01929345540702343, -0.6011698246002197, 0.6696314215660095, 0.12462140619754791, -0.2897532284259796, -0.08932236582040787, 0.12198415398597717, -0.05398513749241829, -0.1360836774110794, 0.28729933500289917, 0.06956455111503601, 0.2895840108394623, 0.6231245398521423, 0.04480636119842529, 0.08362899720668793, -0.18011678755283356, -0.09812988340854645, 0.23890705406665802, 0.10177021473646164, -0.43847382068634033, -0.09930111467838287, -0.14744070172309875, 0.020528171211481094, -0.08745360374450684, -0.5327375531196594, -0.036130640655756, 0.3234040141105652, -0.17692336440086365, 0.3889913856983185, 0.5019958019256592, 0.01669052056968212, 0.06571757048368454, -0.2933850586414337, -0.06997055560350418, -0.07178251445293427, -0.0772053599357605, -0.12362546473741531, -0.26133376359939575 ]
https://github.com/huggingface/datasets/issues/197
Scientific Papers only downloading Pubmed
Hi so there are indeed two configurations in the datasets as you can see [here](https://github.com/huggingface/nlp/blob/master/datasets/scientific_papers/scientific_papers.py#L81-L82). You can load either one with: ```python dataset = nlp.load_dataset('scientific_papers', 'pubmed') dataset = nlp.load_dataset('scientific_papers', 'arxiv') ``` This issues is actually related to a similar user-experience issue with GLUE. When several configurations are available and the first configuration is loaded by default (see issue #152 and #130), it seems to be unexpected for users. I think we should maybe raise a (very explicit) error when there are several configurations available and the user doesn't specify one. What do you think @lhoestq @patrickvonplaten @mariamabarham ?
Hi! I have been playing around with this module, and I am a bit confused about the `scientific_papers` dataset. I thought that it would download two separate datasets, arxiv and pubmed. But when I run the following: ``` dataset = nlp.load_dataset('scientific_papers', data_dir='.', cache_dir='.') Downloading: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 5.05k/5.05k [00:00<00:00, 2.66MB/s] Downloading: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4.90k/4.90k [00:00<00:00, 2.42MB/s] Downloading and preparing dataset scientific_papers/pubmed (download: 4.20 GiB, generated: 2.33 GiB, total: 6.53 GiB) to ./scientific_papers/pubmed/1.1.1... Downloading: 3.62GB [00:40, 90.5MB/s] Downloading: 880MB [00:08, 101MB/s] Dataset scientific_papers downloaded and prepared to ./scientific_papers/pubmed/1.1.1. Subsequent calls will reuse this data. ``` only a pubmed folder is created. There doesn't seem to be something for arxiv. Are these two datasets merged? Or have I misunderstood something? Thanks!
98
Scientific Papers only downloading Pubmed Hi! I have been playing around with this module, and I am a bit confused about the `scientific_papers` dataset. I thought that it would download two separate datasets, arxiv and pubmed. But when I run the following: ``` dataset = nlp.load_dataset('scientific_papers', data_dir='.', cache_dir='.') Downloading: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 5.05k/5.05k [00:00<00:00, 2.66MB/s] Downloading: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4.90k/4.90k [00:00<00:00, 2.42MB/s] Downloading and preparing dataset scientific_papers/pubmed (download: 4.20 GiB, generated: 2.33 GiB, total: 6.53 GiB) to ./scientific_papers/pubmed/1.1.1... Downloading: 3.62GB [00:40, 90.5MB/s] Downloading: 880MB [00:08, 101MB/s] Dataset scientific_papers downloaded and prepared to ./scientific_papers/pubmed/1.1.1. Subsequent calls will reuse this data. ``` only a pubmed folder is created. There doesn't seem to be something for arxiv. Are these two datasets merged? Or have I misunderstood something? Thanks! Hi so there are indeed two configurations in the datasets as you can see [here](https://github.com/huggingface/nlp/blob/master/datasets/scientific_papers/scientific_papers.py#L81-L82). You can load either one with: ```python dataset = nlp.load_dataset('scientific_papers', 'pubmed') dataset = nlp.load_dataset('scientific_papers', 'arxiv') ``` This issues is actually related to a similar user-experience issue with GLUE. When several configurations are available and the first configuration is loaded by default (see issue #152 and #130), it seems to be unexpected for users. I think we should maybe raise a (very explicit) error when there are several configurations available and the user doesn't specify one. What do you think @lhoestq @patrickvonplaten @mariamabarham ?
[ 0.3410275876522064, -0.05078306421637535, 0.04983749985694885, 0.20578421652317047, -0.012493719346821308, -0.09158283472061157, 0.1314340978860855, 0.24781416356563568, 0.07514096051454544, -0.2193240374326706, -0.16563859581947327, 0.10058081895112991, 0.06382755190134048, -0.29008451104164124, 0.24017539620399475, 0.07839687913656235, 0.09766237437725067, 0.11659561842679977, 0.14824886620044708, -0.19698093831539154, -0.11032894998788834, 0.4662143290042877, -0.11777038872241974, -0.1241876482963562, 0.026482487097382545, 0.05782882124185562, -0.25497695803642273, 0.22920358180999756, -0.17581495642662048, -0.269314169883728, 0.22969616949558258, 0.46074461936950684, 0.010647257789969444, 0.24424856901168823, -0.00012602117203641683, -0.026637371629476547, 0.4085165560245514, 0.037491992115974426, -0.07534440606832504, -0.12848317623138428, 0.1525416225194931, -0.31279516220092773, 0.3978445529937744, -0.1065416932106018, 0.04228195175528526, -0.39309123158454895, 0.17608559131622314, -0.04638897255063057, 0.17987219989299774, 0.0012881001457571983, 0.08718008548021317, 0.12423481047153473, 0.06380835920572281, 0.06727642565965652, 0.49336209893226624, 0.01011543720960617, 0.06612035632133484, 0.4198920726776123, 0.2697267532348633, -0.16339482367038727, 0.1365492045879364, 0.24459674954414368, -0.11495199799537659, 0.15639469027519226, 0.1796971708536148, 0.4194621741771698, 0.05964412912726402, -0.46717801690101624, -0.20219352841377258, 0.3455395996570587, 0.014853717759251595, 0.013018778525292873, -0.046098630875349045, -0.46777966618537903, -0.012200591154396534, 0.0002345803804928437, 0.20772826671600342, 0.4563731849193573, -0.16081488132476807, -0.06489183753728867, 0.08118154853582382, -0.25007063150405884, -0.060462597757577896, 0.5617241859436035, 0.16432136297225952, 0.048341624438762665, -0.19733865559101105, 0.3637365996837616, 0.4003928005695343, 0.1685730516910553, -0.2624979317188263, -0.287566214799881, 0.19754008948802948, 0.1875084936618805, -0.30293360352516174, 0.008081098087131977, 0.08924124389886856, -0.06685415655374527, 0.24016189575195312, 0.34288355708122253, 0.08698314428329468, -0.04315409064292908, 0.19389522075653076, 0.12176142632961273, 0.5296932458877563, -0.06093272566795349, -0.020973946899175644, 0.1862226128578186, 0.3687579333782196, -0.016591539606451988, 0.08867759257555008, 0.09830314666032791, 0.0916954055428505, -0.02542254701256752, -0.23397693037986755, -0.5981557369232178, 0.2879771292209625, -0.2756764590740204, 0.24873140454292297, -0.023608239367604256, -0.01119091734290123, -0.05599595978856087, -0.2106601893901825, 0.004625622648745775, 0.018987420946359634, 0.22263126075267792, 0.11903005093336105, 0.2808915376663208, -0.20559199154376984, 0.2087717205286026, -0.07826881855726242, 0.005981168709695339, -0.3656753897666931, 0.039704520255327225, 0.2711537480354309, -0.2781742513179779, 0.1672440469264984, 0.09166543930768967, -0.11322520673274994, -0.18400992453098297, 0.17918287217617035, -0.1315312534570694, 0.0667218267917633, 0.3698318600654602, 0.2778124511241913, 0.5080723166465759, -0.0019687083549797535, -0.2545323967933655, -0.27548766136169434, -0.06114421784877777, -0.2797977328300476, -0.16831760108470917, -0.18681000173091888, 0.0298581812530756, -0.3902367651462555, 0.16661040484905243, 0.00011825386172858998, 0.28824153542518616, 0.1552383303642273, -0.17051126062870026, -0.11550717800855637, -0.0045685237273573875, -0.08512412011623383, -0.3703429102897644, 0.1219870001077652, 0.3437544107437134, 0.19060131907463074, -0.10569165647029877, -0.1270906925201416, -0.10054012387990952, 0.08209928870201111, 0.3111582398414612, -0.2761954069137573, 0.14208722114562988, -0.241653710603714, 0.23108156025409698, 0.590416669845581, -0.2710135281085968, -0.45266878604888916, 0.3115585148334503, -0.024965541437268257, 0.07489562779664993, 0.15223008394241333, 0.2749255299568176, -0.1387358158826828, -0.02256168983876705, 0.16591571271419525, 0.3668285310268402, 0.09084150940179825, -0.14627526700496674, -0.3049764633178711, -0.19426599144935608, 0.302320271730423, -0.09074021130800247, 0.1159532368183136, 0.0059385658241808414, -0.008975031785666943, 0.06970755010843277, 0.3066049814224243, 0.229495570063591, 0.1860748827457428, 0.24630200862884521, -0.07493343949317932, -0.009010808542370796, 0.07086269557476044, -0.2000003606081009, -0.33421558141708374, 0.05639487877488136, -0.355522096157074, 0.24316798150539398, 0.2622116804122925, -0.04236579313874245, -0.23279714584350586, -0.49318912625312805, -0.17978273332118988, -0.2683064639568329, 0.004727189429104328, 0.21896886825561523, 0.12241169065237045, 0.16072703897953033, -0.2678840756416321, 0.17435498535633087, 0.009839650243520737, 0.08805974572896957, -0.8079341650009155, 0.3179456293582916, -0.017845602706074715, -0.07045108824968338, 0.06313509494066238, 0.1433010846376419, 0.0012620689813047647, -0.16912350058555603, 0.32444778084754944, 0.43658125400543213, -0.13749364018440247, 0.22820690274238586, 0.22558556497097015, 0.05751818045973778, 0.25377923250198364, -0.1108492836356163, 0.15133513510227203, -0.12579110264778137, 0.11191105842590332, -0.23113113641738892, -0.30425330996513367, 0.11682773381471634, -0.07708699256181717, 0.3747083842754364, 0.1469537913799286, -0.01149560883641243, 0.010348991490900517, -0.23012582957744598, -0.47462499141693115, -0.2160554677248001, 0.20148776471614838, 0.2280125468969345, -0.11197280138731003, 0.1459309309720993, -0.29982665181159973, 0.07807818055152893, 0.07013934850692749, 0.017231879755854607, -0.020467661321163177, 0.1736775040626526, -0.27142319083213806, -0.2440957874059677, -0.18221400678157806, 0.5162524580955505, 0.8718114495277405, 0.1978529840707779, 0.32297390699386597, 0.10880778729915619, -0.17633940279483795, -0.19697558879852295, 0.08697959780693054, 0.03373314440250397, -0.12805910408496857, 0.21403977274894714, 0.17132674157619476, -0.008642409928143024, -0.42317691445350647, 0.38059061765670776, 0.16046485304832458, 0.02128945104777813, -0.37637507915496826, -0.19031348824501038, -0.40123823285102844, -0.3012849986553192, -0.6075857877731323, -0.1843884438276291, -0.3624565303325653, -0.3466339111328125, -0.06475497037172318, 0.13994979858398438, -0.050487276166677475, 0.024638913571834564, -0.20654024183750153, 0.1555749773979187, -0.239600270986557, 0.06841805577278137, 0.1411573439836502, 0.12482736259698868, -0.3013817369937897, -0.0807427242398262, 0.03651624917984009, 0.3887549340724945, 0.5186214447021484, 0.054689545184373856, -0.20739388465881348, -0.08984798938035965, -0.45796722173690796, 0.2533678412437439, -0.20390884578227997, 0.4136424660682678, 0.28525587916374207, 0.09833773970603943, 0.3634146749973297, -0.08333981037139893, 0.2817312180995941, -0.14543601870536804, 0.03042098879814148, 0.008215513080358505, 0.08342500776052475, -0.02024722658097744, -0.2850772440433502, -0.569351077079773, -0.30609121918678284, -0.2529951333999634, -0.1100711077451706, -0.05502288416028023, 0.12144801765680313, 0.24505040049552917, -0.08112381398677826, 0.1239975243806839, -0.464856892824173, -0.04552529379725456, -0.36400437355041504, -0.08933532238006592, 0.2416916936635971, -0.2270822525024414, -0.40880250930786133, -0.043363090604543686, -0.024774005636572838, -0.09826160222291946, 0.15945154428482056, -0.3808136284351349, -0.41890573501586914, -0.21503759920597076, -0.50697922706604, 0.17532074451446533, 0.41701334714889526, 0.3016418516635895, -0.33650583028793335, -0.10329844802618027, -0.08495737612247467, -0.16373059153556824, 0.18865858018398285, 0.21752363443374634, 0.36106887459754944, 0.20806339383125305, -0.09732377529144287, -0.015406563878059387, 0.5112139582633972, 0.4535945653915405, 0.18884378671646118, 0.2588811218738556, -0.030591106042265892, 0.6296037435531616, 0.08015425503253937, -0.35266467928886414, 0.0025887300726026297, -0.023501744493842125, 0.2995643615722656, 0.32156434655189514, 0.24641655385494232, -0.17453691363334656, -0.09905512630939484, 0.1481449455022812, -0.35031548142433167, -0.3976813554763794, 0.220642551779747, -0.11810324341058731, 0.03348483517765999, 0.21164017915725708, -0.37359631061553955, 0.05672029033303261, -0.05050119385123253, -0.35609951615333557, 0.39024823904037476, -0.13153022527694702, 0.12844467163085938, -0.5994893312454224, 0.11785123497247696, -0.43639689683914185, 0.13917535543441772, -0.12238940596580505, 0.16424332559108734, 0.1163775622844696, -0.06740368902683258, 0.39220017194747925, 0.028027435764670372, 0.5079413652420044, -0.13392317295074463, -0.21793530881404877, 0.18318815529346466, -0.09627906233072281, -0.8417407274246216, -0.16880550980567932, -0.0777038112282753, -0.1152687817811966, 0.2761748731136322, 0.302824467420578, -0.5176627039909363, -0.25081342458724976, 0.20353595912456512, 0.1481359601020813, -0.17218631505966187, -0.24935102462768555, -0.26382681727409363, -0.44855770468711853, -0.31027671694755554, -0.39140981435775757, -0.12274365872144699, -0.05234192684292793, -0.11758974194526672, -0.12879563868045807, 0.15744562447071075, -0.1702955812215805, 0.14247334003448486, -0.160899817943573, 0.26104751229286194, -0.02096606232225895, 0.03182672709226608, 0.4652025103569031, 0.1336738020181656, -0.5750660300254822, 0.5631842017173767, -0.1878594309091568, 0.2896808683872223, 0.13970738649368286, -0.17257258296012878, 0.36342331767082214, 0.5340468287467957, -0.06888512521982193, 0.06709103286266327, 0.03953137993812561, -0.07136215269565582, -0.11491364985704422, 0.25564417243003845, 0.5111371874809265, -0.34914904832839966, -0.1387997269630432, -0.33767688274383545, 0.5596755743026733, 0.0816647931933403, -0.40842482447624207, 0.19925887882709503, 0.2666527330875397, -0.2441544532775879, 0.28216037154197693, -0.3538728654384613, 1.0832157135009766, 0.04487157613039017, 0.31874603033065796, 0.08139228075742722, -0.2079668790102005, 0.8912694454193115, -0.019847044721245766, 0.02509402669966221, -0.2386830598115921, -0.040029287338256836, 0.016051165759563446, -0.18522730469703674, 0.05906468257308006, 0.08315709978342056, -0.026411214843392372, 0.2931695282459259, 0.2207300364971161, 0.09865697473287582, -0.10229551047086716, 0.5956615805625916, 0.04097234085202217, -0.08230273425579071, -0.3292582333087921, 0.04090878739953041, -0.013093141838908195, 0.3151538074016571, 0.06848553568124771, -0.14593514800071716, -0.2229432761669159, -0.4063076376914978, -0.267161101102829, 0.04904688149690628, 0.05897744372487068, 0.1323080211877823, 0.10286843776702881, -0.11787678301334381, -0.29500025510787964, 0.029080664739012718, -0.04239339381456375, 0.0197905246168375, -0.25581473112106323, -0.08757369220256805, 0.0039717354811728, -0.30738794803619385, -0.13085705041885376, -0.2202378660440445, 0.2706197202205658, -0.12889164686203003, -0.182833731174469, -0.2363050878047943, 0.040532924234867096, -0.1591941863298416, -0.46211016178131104, 0.31663963198661804, 0.2644227147102356, -0.08291423320770264, -0.296355277299881, 0.36175721883773804, 0.1117045134305954, -0.002363532315939665, 0.02936004102230072, 0.35258588194847107, 0.07839546352624893, 0.36199304461479187, 0.13534189760684967, -0.19166676700115204, -0.06823715567588806, 0.3341461718082428, 0.5289815068244934, -0.09188313782215118, 0.2554856240749359, -0.11541934311389923, 0.05190782994031906, -0.05099848285317421, -0.16937880218029022, -0.08383090049028397, -0.33998486399650574, 0.12292981892824173, -0.012368220835924149, 0.0692097395658493, 0.13143163919448853, 0.043052323162555695, 0.09093374013900757, -0.21702678501605988, -0.10895229130983353, -0.4528340995311737, 0.06527508050203323, 0.2933409810066223, -0.1723700612783432, 0.4097401797771454, 0.04056276008486748, -0.12469819188117981, 0.25025734305381775, 0.31264063715934753, -0.16058339178562164, 0.2575589716434479, 0.06299566477537155, 0.24464063346385956, 0.6299234628677368, 0.01803257316350937, 0.08861207962036133, -0.11040453612804413, -0.08225184679031372, -0.057103194296360016, -0.177371546626091, -0.09637568891048431, -0.2691379487514496, 0.2264515906572342, 0.14458264410495758, 0.0587296225130558, 0.09794025123119354, -0.23562240600585938, 0.03194234147667885, 0.008571567945182323, 0.20746628940105438, -0.06657211482524872, -0.03942136839032173, -0.025018131360411644, -0.16160526871681213, 0.03848303481936455, 0.11174486577510834, 0.3407118022441864, -0.36419451236724854, -0.042735327035188675, 0.34211671352386475, 0.008133234456181526, -0.2114960104227066, -0.0978815108537674, -0.2852175235748291, -0.030537322163581848, 0.10420649498701096, -0.04821806401014328, 0.11100742220878601, 0.00643351674079895, -0.029975824058055878, -0.1822148561477661, 0.3628310263156891, 0.5864513516426086, -0.2662340998649597, 0.087650366127491, -0.05240936949849129, 0.02447769045829773, -0.17970028519630432, -0.03848183900117874, 0.2230919450521469, 0.26938551664352417, -0.08858197182416916, 0.26846641302108765, 0.32481205463409424, 0.025377919897437096, -0.1227574422955513, -0.019717764109373093, 0.6393304467201233, -0.19335374236106873, 0.08591988682746887, -0.2401425540447235, -0.04187896102666855, -0.11093265563249588, -0.017788173630833626, -0.03344742953777313, 0.007564181927591562, -0.09867077320814133, 0.20377254486083984, 0.2015051394701004, -0.035435013473033905, 0.052560098469257355, 0.23083533346652985, -0.18457351624965668, 0.016510071232914925, 0.7022255063056946, 0.2064048945903778, 0.30672305822372437, 0.14233793318271637, 0.7357043027877808, -0.11739101260900497, -0.050035543739795685, -0.13992109894752502, 0.5564054250717163, 0.2685464322566986, -0.23760439455509186, 0.06799263507127762, -0.11369511485099792, -0.14237579703330994, 0.04823797568678856, -0.044042449444532394, -0.328214555978775, 0.27146121859550476, -0.02766067534685135, -0.19461390376091003, -0.7354897856712341, 0.16255070269107819, 0.07861088961362839, 0.10736674070358276, -0.15201421082019806, 0.18625995516777039, -0.07005401700735092, -0.0500728115439415, 0.2191653996706009, 0.5182142853736877, 0.29087725281715393, 0.3593863844871521, -0.006482074502855539, -0.16204600036144257, 0.24868369102478027, 0.17708556354045868, -0.2095000445842743, 0.15459488332271576, 0.34004920721054077, 0.07801402360200882, 0.32010093331336975, 0.0420922227203846, -0.10348379611968994, -0.06722740083932877, 0.07740665972232819, 0.029380615800619125, -0.32071951031684875, 0.5537850856781006, -0.28373441100120544, 0.08010026067495346, -0.01782439462840557, 0.06254695355892181, -0.616965115070343, -0.19064947962760925, 0.2139841467142105, 0.005209835711866617, 0.16936847567558289, -0.46065622568130493, 0.014075754210352898, 0.17706634104251862, 0.5300930738449097, 0.0472886860370636, 0.06634998321533203, -0.15457844734191895, -0.244127556681633, -0.789830207824707, 0.13781431317329407, -0.22904323041439056, 0.14415177702903748, -0.04818122833967209, 0.39347124099731445, 0.03287484869360924, 0.31124669313430786, -0.022219939157366753, 0.130643829703331, -0.5399712324142456, -0.13637740910053253, -0.17598295211791992, -0.04879850521683693, -0.21684108674526215, -0.058267902582883835, 0.005119413137435913, -0.11321878433227539, 0.10097906738519669, -0.2995997369289398, -0.07709407061338425, 0.09741416573524475, 0.07090751826763153, 0.07483290135860443, 0.05449523776769638, 0.4069506525993347, 0.23451301455497742, 0.4126705229282379, -0.0002771328727249056, -0.2844560146331787, -0.5316900014877319, 0.045554764568805695, -0.4169573187828064, 0.09936445951461792, 0.02154848910868168, 0.20746056735515594, -0.4415889382362366, 0.1541765332221985, -0.11783277243375778, -0.07033431529998779, 0.13117718696594238, 0.0565231628715992, 0.036482810974121094, -0.008407062850892544, 0.14452411234378815, 0.1382405012845993, 0.23347680270671844, 0.41835594177246094, -0.225312739610672, 0.13117705285549164, -0.19558179378509521, -0.27393946051597595, 0.2897917628288269, -0.36235687136650085, -0.13488726317882538, 0.12069623917341232, 0.24682049453258514, 0.07023007422685623, 0.06666920334100723, -0.6980177164077759, -0.263947993516922, 0.3107894957065582, -0.0424811951816082, -0.18357011675834656, 0.2500549852848053, -0.13164988160133362, -0.47837772965431213, -0.022704048082232475, 0.04410466551780701, 0.15658123791217804, -0.38131535053253174, 0.01163423154503107, -0.23016810417175293 ]
https://github.com/huggingface/datasets/issues/197
Scientific Papers only downloading Pubmed
Now if you don't specify which part you want, it raises an error: ``` ValueError: Config name is missing. Please pick one among the available configs: ['pubmed', 'arxiv'] Example of usage: `load_dataset('scientific_papers', 'pubmed')` ```
Hi! I have been playing around with this module, and I am a bit confused about the `scientific_papers` dataset. I thought that it would download two separate datasets, arxiv and pubmed. But when I run the following: ``` dataset = nlp.load_dataset('scientific_papers', data_dir='.', cache_dir='.') Downloading: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 5.05k/5.05k [00:00<00:00, 2.66MB/s] Downloading: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4.90k/4.90k [00:00<00:00, 2.42MB/s] Downloading and preparing dataset scientific_papers/pubmed (download: 4.20 GiB, generated: 2.33 GiB, total: 6.53 GiB) to ./scientific_papers/pubmed/1.1.1... Downloading: 3.62GB [00:40, 90.5MB/s] Downloading: 880MB [00:08, 101MB/s] Dataset scientific_papers downloaded and prepared to ./scientific_papers/pubmed/1.1.1. Subsequent calls will reuse this data. ``` only a pubmed folder is created. There doesn't seem to be something for arxiv. Are these two datasets merged? Or have I misunderstood something? Thanks!
34
Scientific Papers only downloading Pubmed Hi! I have been playing around with this module, and I am a bit confused about the `scientific_papers` dataset. I thought that it would download two separate datasets, arxiv and pubmed. But when I run the following: ``` dataset = nlp.load_dataset('scientific_papers', data_dir='.', cache_dir='.') Downloading: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 5.05k/5.05k [00:00<00:00, 2.66MB/s] Downloading: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4.90k/4.90k [00:00<00:00, 2.42MB/s] Downloading and preparing dataset scientific_papers/pubmed (download: 4.20 GiB, generated: 2.33 GiB, total: 6.53 GiB) to ./scientific_papers/pubmed/1.1.1... Downloading: 3.62GB [00:40, 90.5MB/s] Downloading: 880MB [00:08, 101MB/s] Dataset scientific_papers downloaded and prepared to ./scientific_papers/pubmed/1.1.1. Subsequent calls will reuse this data. ``` only a pubmed folder is created. There doesn't seem to be something for arxiv. Are these two datasets merged? Or have I misunderstood something? Thanks! Now if you don't specify which part you want, it raises an error: ``` ValueError: Config name is missing. Please pick one among the available configs: ['pubmed', 'arxiv'] Example of usage: `load_dataset('scientific_papers', 'pubmed')` ```
[ 0.33375269174575806, 0.15182602405548096, 0.0403822623193264, 0.12209053337574005, -0.011581690050661564, 0.0701768547296524, 0.10876773297786713, 0.3869028091430664, -0.11695440858602524, -0.22115777432918549, -0.1246669590473175, 0.33596864342689514, 0.04204653576016426, -0.32191410660743713, 0.33534082770347595, 0.059789594262838364, 0.17052827775478363, 0.11987831443548203, 0.24696096777915955, -0.32216981053352356, -0.15091967582702637, 0.36268073320388794, -0.20078347623348236, -0.07502947002649307, 0.17173191905021667, 0.1704220175743103, -0.13634683191776276, 0.23558485507965088, -0.2803621292114258, -0.28337788581848145, 0.3049049377441406, 0.41182878613471985, 0.12482397258281708, 0.3285548985004425, -0.0001224009320139885, -0.051075007766485214, 0.399665892124176, -0.11453292518854141, -0.12407606840133667, -0.24390456080436707, 0.025619829073548317, -0.2772785723209381, 0.36111268401145935, -0.1576250046491623, 0.10323184728622437, -0.5018041729927063, 0.22668394446372986, -0.10192909091711044, 0.002705290215089917, 0.2615562975406647, 0.12145330756902695, -0.08232848346233368, 0.13635104894638062, 0.003647337667644024, 0.3264041543006897, 0.018630648031830788, 0.09062385559082031, 0.30791038274765015, 0.17882320284843445, 0.08364755660295486, 0.19623710215091705, 0.21780815720558167, -0.12764742970466614, 0.0741182342171669, 0.33376508951187134, 0.37732863426208496, 0.0491853691637516, -0.45332983136177063, -0.09537716954946518, 0.25129467248916626, 0.4695167541503906, -0.08259274065494537, 0.03704746067523956, -0.20686979591846466, -0.040106069296598434, -0.08870575577020645, 0.10811006277799606, 0.39001035690307617, -0.22332635521888733, -0.0482567623257637, 0.021009119227528572, -0.38639724254608154, -0.19638125598430634, 0.5864763259887695, 0.1692328006029129, 0.15303434431552887, -0.08709662407636642, 0.363867849111557, 0.21774160861968994, 0.1375075727701187, 0.033936284482479095, -0.3225848972797394, 0.2554103136062622, 0.2819865942001343, -0.2922055125236511, -0.007908659987151623, -0.03503989800810814, -0.20915596187114716, 0.18896634876728058, 0.3261251151561737, 0.07502642273902893, -0.0670475885272026, 0.20057806372642517, 0.12096546590328217, 0.4643375873565674, -0.15208926796913147, 0.08678413927555084, 0.3539922833442688, 0.21561063826084137, -0.1866047978401184, 0.03402045741677284, 0.018245108425617218, -0.07721942663192749, 0.07174455374479294, -0.22967815399169922, -0.6142796874046326, 0.1409287303686142, -0.4055894911289215, 0.4478227496147156, -0.025239208713173866, 0.21384789049625397, -0.10629180818796158, -0.23445355892181396, 0.08172966539859772, 0.1051739901304245, 0.2092120200395584, 0.08533553779125214, 0.17645752429962158, -0.11764824390411377, 0.09517494589090347, -0.08684959262609482, 0.016504548490047455, -0.3610500991344452, -0.0581570640206337, 0.23637673258781433, -0.013213790021836758, 0.261093407869339, -0.09314685314893723, 0.07390709966421127, -0.09015914797782898, 0.24405497312545776, -0.0596812404692173, 0.14490318298339844, 0.496426522731781, 0.3115924596786499, 0.5195350646972656, -0.00015620427438989282, -0.1992076337337494, -0.27722230553627014, 0.1592061072587967, -0.3895837068557739, -0.12449304014444351, -0.11835875362157822, 0.0739862471818924, -0.3111491799354553, -0.06145278364419937, 0.10784744471311569, 0.2107422649860382, 0.18148605525493622, -0.35827383399009705, -0.10828685760498047, 0.0256291963160038, 0.040218666195869446, -0.3753199279308319, 0.17049895226955414, 0.13123658299446106, 0.06589129567146301, -0.0389045812189579, -0.28599441051483154, -0.06429087370634079, 0.22669680416584015, 0.07826316356658936, -0.3078140616416931, 0.1558767408132553, -0.1210063174366951, 0.0539570152759552, 0.8645715713500977, -0.2199849784374237, -0.40715402364730835, 0.2794370949268341, 0.05443645641207695, -0.1435593068599701, 0.09176114946603775, 0.3109303116798401, -0.2846706807613373, 0.05149829015135765, 0.09131016582250595, 0.28850024938583374, -0.012590521946549416, -0.2461201250553131, -0.14480462670326233, -0.11279961466789246, 0.308638334274292, -0.07593276351690292, 0.2943565547466278, 0.1680898219347, -0.015603529289364815, 0.14979968965053558, 0.4100162088871002, 0.18755614757537842, 0.2137865126132965, 0.22659888863563538, -0.11479020863771439, -0.10925071686506271, 0.20860159397125244, -0.17600058019161224, -0.16421101987361908, 0.06642244756221771, -0.4373753070831299, 0.2642482817173004, 0.32023951411247253, 0.06156071275472641, -0.2903924584388733, -0.5201041102409363, -0.06402678042650223, -0.3106153607368469, -0.012994172982871532, 0.29097938537597656, -0.0014562280848622322, 0.07865040749311447, -0.3556837737560272, 0.04744693264365196, -0.12707865238189697, 0.014778337441384792, -0.7680269479751587, 0.1568286269903183, -0.0025856310967355967, -0.014122090302407742, 0.0891619399189949, -0.030355697497725487, -0.0414162203669548, -0.018181832507252693, 0.2261609584093094, 0.41557732224464417, -0.031880225986242294, 0.3567318320274353, 0.3559609651565552, -0.10024514049291611, 0.1163012832403183, -0.2353430539369583, 0.15720075368881226, -0.10833925008773804, 0.2621350586414337, -0.20873011648654938, -0.34462618827819824, 0.1056322231888771, -0.15303196012973785, 0.3154359459877014, 0.15342335402965546, -0.002796567277982831, 0.06230355426669121, -0.12727266550064087, -0.40658992528915405, -0.14812879264354706, 0.17097918689250946, 0.2773890197277069, -0.3150269687175751, 0.09033016115427017, -0.26422277092933655, 0.03734680265188217, 0.12562455236911774, 0.006544035859405994, -0.17453725636005402, 0.18298488855361938, -0.311465859413147, -0.21895015239715576, -0.13686028122901917, 0.46691983938217163, 0.9032635688781738, 0.25674036145210266, 0.38866618275642395, 0.09151812642812729, -0.05875339359045029, -0.16813647747039795, 0.1890704333782196, 0.1316784769296646, 0.012559083290398121, 0.16735802590847015, -0.06914068013429642, -0.04874551668763161, -0.4017602801322937, 0.27937430143356323, 0.16855388879776, 0.06223781034350395, -0.30279800295829773, -0.03771255910396576, -0.37835732102394104, -0.19744883477687836, -0.511610746383667, -0.25710561871528625, -0.2551214396953583, -0.3293745517730713, -0.16789761185646057, 0.010061752051115036, 0.05495456978678703, 0.01970987394452095, -0.531943678855896, 0.045221224427223206, -0.1329425871372223, -0.15076598525047302, 0.122951939702034, -0.02294759266078472, -0.24956677854061127, 0.007366005796939135, -0.06586308777332306, 0.3898339867591858, 0.4878615438938141, -0.07025731354951859, -0.042926739901304245, 0.052867595106363297, -0.4029850959777832, 0.2302953153848648, -0.2439562976360321, 0.23640774190425873, 0.241521418094635, 0.129068061709404, 0.2934620976448059, -0.0029637583065778017, 0.15736135840415955, -0.18025967478752136, 0.02490473911166191, -0.04032791033387184, 0.10109694302082062, -0.10887978971004486, -0.275493860244751, -0.8172603845596313, -0.3088199496269226, -0.18090619146823883, -0.23882023990154266, 0.18941572308540344, 0.16927319765090942, 0.3574637174606323, 0.020866474136710167, 0.12806415557861328, -0.32717832922935486, 0.05769855156540871, -0.3687760531902313, -0.15246835350990295, 0.4007764160633087, -0.2736346125602722, -0.379485160112381, 0.10961031913757324, 0.14097337424755096, 0.038805894553661346, 0.14969463646411896, -0.27083054184913635, -0.05013080686330795, -0.007899540476500988, -0.5291986465454102, 0.15095311403274536, 0.33623066544532776, 0.2992245852947235, -0.33691930770874023, -0.11374466121196747, -0.139827698469162, -0.06552594155073166, 0.11963124573230743, 0.24273723363876343, 0.5703092217445374, 0.13668431341648102, -0.061146173626184464, -0.18793024122714996, 0.4702356159687042, 0.33278051018714905, 0.21936634182929993, 0.13947275280952454, -0.08077400922775269, 0.5153941512107849, 0.07098858803510666, -0.25418820977211, 0.17766205966472626, 0.0340990275144577, 0.11521948873996735, 0.4011802673339844, 0.14756087958812714, -0.31791752576828003, -0.22711944580078125, 0.029213692992925644, -0.36350512504577637, -0.291766494512558, 0.22638468444347382, -0.16411012411117554, 0.1019238606095314, 0.26066893339157104, -0.4112977981567383, -0.10925701260566711, -0.0217902809381485, -0.42039135098457336, 0.43069228529930115, -0.15378323197364807, 0.153130903840065, -0.654884397983551, 0.16735504567623138, -0.44770392775535583, 0.11611570417881012, -0.2242058962583542, 0.1797572821378708, 0.013182658702135086, -0.10940807312726974, 0.20083549618721008, 0.07203856110572815, 0.3785324990749359, -0.09239040315151215, -0.2777597904205322, 0.16358697414398193, 0.11103236675262451, -0.6928820610046387, -0.054513268172740936, 0.0354834608733654, 0.06715422868728638, 0.07046312838792801, 0.1560639590024948, -0.32216596603393555, -0.21822907030582428, 0.28061333298683167, 0.12941697239875793, -0.20678293704986572, -0.2330891340970993, -0.3231196403503418, -0.5719004273414612, -0.23541899025440216, -0.4505123198032379, -0.14295238256454468, -0.045315809547901154, -0.13631442189216614, -0.0938030257821083, 0.2150987833738327, -0.12031532824039459, 0.2540312707424164, -0.06086517870426178, 0.26602479815483093, -0.03785590082406998, -0.033680662512779236, 0.4772065579891205, 0.0953952968120575, -0.2791571319103241, 0.42964252829551697, -0.14281505346298218, 0.20302093029022217, 0.13304799795150757, -0.3720070421695709, 0.47840917110443115, 0.3028565049171448, -0.10998030006885529, -0.00590578792616725, 0.19600790739059448, -0.2528352737426758, -0.008956502191722393, 0.14065779745578766, 0.40209999680519104, -0.346855103969574, -0.32030200958251953, -0.58281010389328, 0.5492749214172363, 0.003125912044197321, -0.36882954835891724, 0.2551998794078827, -0.010727684013545513, -0.3315707743167877, 0.23260775208473206, -0.22382239997386932, 1.1526607275009155, 0.1274232268333435, 0.3265951871871948, -0.0009181098430417478, -0.10209719836711884, 0.8224453330039978, -0.054458919912576675, 0.08067682385444641, -0.10956143587827682, -0.04315533488988876, -0.13305158913135529, -0.20595571398735046, 0.1671733409166336, 0.1299719512462616, -0.09159757941961288, 0.3078961968421936, 0.17555078864097595, 0.0021265523973852396, 0.07282976806163788, 0.4751613438129425, 0.1268330216407776, 0.004293225239962339, -0.35017284750938416, 0.06220218539237976, 0.031313151121139526, 0.1564972698688507, 0.0473199300467968, -0.23766227066516876, -0.4849186837673187, -0.41060158610343933, -0.09666313976049423, 0.09266935288906097, 0.3221285939216614, 0.07424373179674149, -0.07986145466566086, -0.19807744026184082, -0.2268858700990677, 0.20781296491622925, -0.008843359537422657, -0.05848034843802452, -0.33503973484039307, 0.1760789006948471, -0.14393234252929688, -0.2794753611087799, -0.20071320235729218, -0.1652781218290329, 0.2625408172607422, -0.25820496678352356, -0.26511821150779724, -0.010710452683269978, -0.0790310874581337, -0.28006163239479065, -0.3210088908672333, 0.2931114733219147, 0.4260565936565399, 0.010683136992156506, -0.2837386429309845, 0.4091005325317383, -0.024119729176163673, -0.004242885857820511, 0.05624654144048691, 0.4037863612174988, 0.02478090114891529, 0.4783609211444855, 0.19976341724395752, -0.3077811896800995, -0.11372805386781693, 0.3047425150871277, 0.7428829669952393, -0.07022089511156082, 0.3021062910556793, -0.09806854277849197, -0.011664054356515408, -0.10569610446691513, -0.10398538410663605, -0.20967814326286316, -0.1769358366727829, 0.0328197255730629, 0.03127063065767288, 0.251249760389328, 0.21842822432518005, 0.011708667501807213, -0.011572939343750477, -0.18024489283561707, -0.1883937120437622, -0.461404025554657, -0.13243554532527924, 0.08870572596788406, -0.3724980056285858, 0.40051084756851196, 0.0721593126654625, -0.1330241858959198, 0.3243182897567749, 0.1654839962720871, -0.2065475583076477, 0.15494373440742493, -0.035977840423583984, 0.2704542875289917, 0.4610452353954315, 0.11127416044473648, 0.05570828914642334, -0.20335392653942108, -0.06352409720420837, -0.1928076595067978, -0.11600193381309509, -0.10259056091308594, -0.2786969244480133, 0.22386667132377625, 0.12651003897190094, 0.048987772315740585, 0.01391167938709259, -0.2325642704963684, -0.19723495841026306, 0.10693574696779251, 0.012789835222065449, -0.0910222977399826, 0.04635284096002579, 0.1213919073343277, -0.13883931934833527, 0.014664708636701107, 0.045174285769462585, 0.3172227144241333, -0.28537997603416443, 0.11752595752477646, 0.3289768099784851, 0.11331642419099808, -0.2091720849275589, -0.13677506148815155, -0.26757118105888367, -0.01486930064857006, 0.12046156078577042, -0.1788608878850937, 0.3095262050628662, 0.02365964464843273, 0.06650518625974655, -0.2103354036808014, 0.2348409742116928, 0.737738311290741, -0.26888778805732727, 0.1399053931236267, -0.07228129357099533, 0.026590954512357712, -0.26007387042045593, -0.03565452992916107, 0.14888793230056763, 0.22216255962848663, -0.16084378957748413, 0.1575271189212799, 0.1493341326713562, -0.07433734089136124, -0.18297052383422852, 0.01084597222507, 0.7443416714668274, -0.1675221472978592, 0.10861516743898392, -0.13933466374874115, -0.0065565574914216995, 0.04117714613676071, -0.015504982322454453, 0.17375369369983673, 0.021767066791653633, 0.10403049737215042, 0.0990992859005928, 0.29588139057159424, 0.04692915454506874, -0.08133488893508911, 0.1565871685743332, -0.1968528777360916, -0.023893918842077255, 0.7082404494285583, 0.10019394755363464, 0.3292534351348877, 0.21454094350337982, 0.5007227659225464, 0.04617145285010338, 0.08842502534389496, -0.15246091783046722, 0.4521012604236603, 0.30750808119773865, -0.12500467896461487, -0.1658215969800949, -0.276661217212677, -0.0663161352276802, -0.1019153892993927, 0.004147159401327372, -0.27595198154449463, 0.1621083915233612, 0.013707396574318409, -0.2040804773569107, -0.6190455555915833, 0.3473253846168518, 0.028992000967264175, -0.08817125111818314, -0.15126630663871765, 0.05593785271048546, -0.05157645791769028, -0.106229767203331, 0.3673674166202545, 0.33350178599357605, 0.24257710576057434, 0.29391366243362427, 0.0698905736207962, -0.16431932151317596, 0.3510328531265259, 0.030839579179883003, -0.1833093762397766, 0.19911254942417145, 0.31850293278694153, 0.09088330715894699, 0.3007444143295288, 0.1006975769996643, -0.03643719479441643, 0.03841126337647438, 0.2743447422981262, 0.0016552555607631803, -0.2547135651111603, 0.4916636645793915, -0.1473430097103119, 0.06510000675916672, -0.011430514976382256, 0.11638002097606659, -0.6328662037849426, -0.2241937518119812, 0.29300516843795776, 0.09583374112844467, 0.11507967859506607, -0.3418157994747162, 0.014299304224550724, 0.09833154827356339, 0.6334850192070007, -0.07816998660564423, 0.12740185856819153, -0.12272841483354568, -0.3408302962779999, -0.6841955780982971, 0.22655120491981506, -0.1193961352109909, 0.13815122842788696, 0.0011205475311726332, 0.2716943919658661, 0.28507909178733826, 0.24463270604610443, 0.08168870955705643, 0.05564732849597931, -0.6260475516319275, 0.018999774008989334, -0.043697379529476166, -0.22733014822006226, -0.2766311466693878, 0.2152404934167862, -0.08741699159145355, -0.03948570415377617, 0.10128520429134369, -0.35122114419937134, -0.02895716205239296, -0.07868736982345581, -0.08384790271520615, 0.2922545075416565, 0.16823986172676086, 0.5404801368713379, 0.11159107834100723, 0.5250933170318604, -0.1857175976037979, -0.2845461964607239, -0.3882063031196594, 0.11773788183927536, -0.3753584325313568, 0.10256795585155487, -0.0020776621531695127, 0.06304620206356049, -0.46579068899154663, 0.3216524124145508, -0.07286219298839569, 0.013991940766572952, -0.04858587309718132, -0.22922463715076447, -0.0022878057789057493, -0.05107443407177925, 0.0779462680220604, 0.10831804573535919, 0.18994592130184174, 0.29205793142318726, -0.2019282728433609, 0.0743769183754921, -0.18357212841510773, -0.13972774147987366, 0.38617274165153503, -0.20822648704051971, -0.19402609765529633, 0.0987420380115509, 0.018720369786024094, -0.1719147264957428, 0.16188129782676697, -0.7090774774551392, -0.0625184029340744, 0.37038546800613403, -0.08114296197891235, -0.1896418035030365, 0.10166256129741669, -0.0853830948472023, -0.3409075438976288, -0.10702230036258698, 0.2071159929037094, 0.011787394061684608, -0.27496498823165894, -0.19342850148677826, -0.18986165523529053 ]
https://github.com/huggingface/datasets/issues/193
[Tensorflow] Use something else than `from_tensor_slices()`
`from_generator` is not working on TPU, I met the following error : ``` File "/usr/local/lib/python3.6/contextlib.py", line 88, in __exit__ next(self.gen) File "/home/usr/.venv/bart/lib/python3.6/site-packages/tensorflow_core/python/eager/context.py", line 1900, in execution_mode executor_new.wait() File "/home/usr/.venv/bart/lib/python3.6/site-packages/tensorflow_core/python/eager/executor.py", line 67, in wait pywrap_tensorflow.TFE_ExecutorWaitForAllPendingNodes(self._handle) tensorflow.python.framework.errors_impl.NotFoundError: No registered 'PyFunc' OpKernel for 'CPU' devices compatible with node {{node PyFunc}} . Registered: <no registered kernels> [[PyFunc]] ``` --- @lhoestq It seems you merged some changes that allow lazy-loading. **Can you give an example of how to use ?** Maybe the Colab notebook should be updated with this method as well.
In the example notebook, the TF Dataset is built using `from_tensor_slices()` : ```python columns = ['input_ids', 'token_type_ids', 'attention_mask', 'start_positions', 'end_positions'] train_tf_dataset.set_format(type='tensorflow', columns=columns) features = {x: train_tf_dataset[x] for x in columns[:3]} labels = {"output_1": train_tf_dataset["start_positions"]} labels["output_2"] = train_tf_dataset["end_positions"] tfdataset = tf.data.Dataset.from_tensor_slices((features, labels)).batch(8) ``` But according to [official tensorflow documentation](https://www.tensorflow.org/guide/data#consuming_numpy_arrays), this will load the entire dataset to memory. **This defeats one purpose of this library, which is lazy loading.** Is there any other way to load the `nlp` dataset into TF dataset lazily ? --- For example, is it possible to use [Arrow dataset](https://www.tensorflow.org/io/api_docs/python/tfio/arrow/ArrowDataset) ? If yes, is there any code example ?
87
[Tensorflow] Use something else than `from_tensor_slices()` In the example notebook, the TF Dataset is built using `from_tensor_slices()` : ```python columns = ['input_ids', 'token_type_ids', 'attention_mask', 'start_positions', 'end_positions'] train_tf_dataset.set_format(type='tensorflow', columns=columns) features = {x: train_tf_dataset[x] for x in columns[:3]} labels = {"output_1": train_tf_dataset["start_positions"]} labels["output_2"] = train_tf_dataset["end_positions"] tfdataset = tf.data.Dataset.from_tensor_slices((features, labels)).batch(8) ``` But according to [official tensorflow documentation](https://www.tensorflow.org/guide/data#consuming_numpy_arrays), this will load the entire dataset to memory. **This defeats one purpose of this library, which is lazy loading.** Is there any other way to load the `nlp` dataset into TF dataset lazily ? --- For example, is it possible to use [Arrow dataset](https://www.tensorflow.org/io/api_docs/python/tfio/arrow/ArrowDataset) ? If yes, is there any code example ? `from_generator` is not working on TPU, I met the following error : ``` File "/usr/local/lib/python3.6/contextlib.py", line 88, in __exit__ next(self.gen) File "/home/usr/.venv/bart/lib/python3.6/site-packages/tensorflow_core/python/eager/context.py", line 1900, in execution_mode executor_new.wait() File "/home/usr/.venv/bart/lib/python3.6/site-packages/tensorflow_core/python/eager/executor.py", line 67, in wait pywrap_tensorflow.TFE_ExecutorWaitForAllPendingNodes(self._handle) tensorflow.python.framework.errors_impl.NotFoundError: No registered 'PyFunc' OpKernel for 'CPU' devices compatible with node {{node PyFunc}} . Registered: <no registered kernels> [[PyFunc]] ``` --- @lhoestq It seems you merged some changes that allow lazy-loading. **Can you give an example of how to use ?** Maybe the Colab notebook should be updated with this method as well.
[ -0.1073189303278923, -0.0873601958155632, 0.05800769850611687, 0.19675105810165405, 0.21143007278442383, 0.09753461182117462, 0.28022217750549316, 0.37640878558158875, -0.15019261837005615, 0.0037052403204143047, -0.06848964095115662, 0.5348147749900818, -0.1144312173128128, 0.003604027908295393, 0.3254980444908142, -0.17846260964870453, -0.034475427120923996, 0.25624608993530273, -0.0664241835474968, -0.06997563689947128, -0.04355105012655258, -0.024568818509578705, -0.09649649262428284, -0.026775924488902092, 0.056818120181560516, 0.018074840307235718, 0.08552568405866623, -0.14649777114391327, 0.4081125855445862, 0.04641757532954216, 0.38986900448799133, 0.023344988003373146, -0.03362147882580757, 0.1560436636209488, -0.000114742862933781, 0.4061403274536133, 0.03541801869869232, -0.19170315563678741, 0.07188346982002258, -0.1159355565905571, 0.16804364323616028, -0.16358785331249237, 0.16523249447345734, -0.21809466183185577, -0.05620698258280754, 0.02299570105969906, 0.2129848450422287, -0.060164835304021835, 0.2803555130958557, 0.13797186315059662, 0.09116267412900925, 0.5434454679489136, -0.17063161730766296, 0.2633912265300751, 0.1432105004787445, -0.15265363454818726, -0.23502609133720398, -0.254381000995636, -0.1945095658302307, 0.020991520956158638, 0.11463548243045807, 0.08944666385650635, -0.23712687194347382, 0.12987102568149567, 0.46162736415863037, 0.16119170188903809, -0.06941181421279907, -0.36647117137908936, -0.20929767191410065, 0.4291148781776428, 0.2616470456123352, -0.06549938768148422, -0.17502063512802124, -0.4465813636779785, -0.23368868231773376, -0.3110019266605377, -0.10490377992391586, 0.12665270268917084, -0.2806473672389984, 0.22548286616802216, -0.09822240471839905, -0.08737309277057648, -0.2566457688808441, 0.1765187829732895, 0.1964697390794754, -0.044815246015787125, 0.12743735313415527, 0.11077558994293213, 0.20774735510349274, 0.154510959982872, 0.4083366394042969, 0.02135806530714035, 0.3032935857772827, 0.40228039026260376, -0.2894178628921509, 0.02911575511097908, 0.17799705266952515, -0.5676334500312805, -0.0688973069190979, -0.2302417904138565, 0.39170101284980774, 0.16744378209114075, -0.3058209717273712, 0.2517024874687195, -0.010725426487624645, -0.09433666616678238, -0.4143877625465393, 0.17374259233474731, -0.15313921868801117, -0.5154545903205872, 0.05342891812324524, 0.0449388362467289, -0.2540411949157715, -0.010109999217092991, 0.02233605645596981, -0.5388684868812561, -0.1700577437877655, 0.22495514154434204, -0.3130659759044647, -0.42528679966926575, -0.25238993763923645, -0.04954863712191582, 0.0026704049669206142, 0.2106461524963379, -0.12489917129278183, 0.5658951997756958, 0.14857371151447296, -0.13732144236564636, -0.30453813076019287, -0.048636503517627716, -0.19958890974521637, 0.08693216741085052, -0.06759520620107651, -0.08612387627363205, 0.13496723771095276, -0.17564843595027924, 0.1778351068496704, -0.15483301877975464, -0.16657774150371552, 0.13682706654071808, 0.299893856048584, -0.10216767340898514, 0.0943407341837883, 0.2443118691444397, 0.046297308057546616, -0.05606013908982277, 0.047438692301511765, 0.07257132232189178, -0.2107289731502533, 0.3622112274169922, -0.5235630869865417, -0.30007046461105347, -0.041467197239398956, 0.0968170315027237, 0.02677207626402378, -0.19203898310661316, -0.32315269112586975, 0.2882845401763916, 0.1580440253019333, -0.13978342711925507, -0.17260891199111938, -0.17757491767406464, -0.43399539589881897, -0.2991338074207306, 0.3465133607387543, 0.0750122144818306, -0.5027426481246948, -0.12278886884450912, -0.012916401028633118, -0.05792608857154846, 0.2622862160205841, 0.427792489528656, -0.22205686569213867, 0.538047730922699, -0.08767692744731903, -0.027470698580145836, 0.8010424971580505, -0.0747700110077858, -0.17104381322860718, 0.24721165001392365, -0.11224613338708878, 0.1672293245792389, 0.028713300824165344, 0.387594074010849, -0.0019182333489879966, -0.03081885166466236, 0.38124847412109375, 0.637941837310791, -0.15168552100658417, 0.02548350766301155, -0.10614724457263947, -0.34956666827201843, 0.2707490622997284, 0.24196387827396393, 0.1044936403632164, -0.030045440420508385, -0.23651081323623657, 0.3986968696117401, 0.13126157224178314, -0.024340849369764328, -0.19287601113319397, 0.14614519476890564, -0.14589954912662506, 0.025187570601701736, -0.18390023708343506, 0.1019967719912529, -0.6190455555915833, 0.2335280328989029, 0.1444830745458603, 0.08388926833868027, 0.23681828379631042, -0.04436876252293587, 0.3555767238140106, 0.12261123955249786, 0.0514349602162838, 0.15910974144935608, -0.04666570946574211, 0.12642869353294373, -0.05838850140571594, -0.0436280257999897, -0.3507686257362366, 0.1307992786169052, -0.350517213344574, -0.0940878838300705, -0.3933817744255066, 0.29612085223197937, 0.3782665729522705, -0.03691062703728676, -0.11612729728221893, 0.25752630829811096, -0.3368946611881256, -0.10079947859048843, 0.059604745358228683, 0.20393307507038116, 0.008491909131407738, 0.3252091407775879, -0.7214282155036926, 0.4762922525405884, 0.05750986188650131, -0.017924945801496506, -0.08479440212249756, 0.13055141270160675, 0.009945041500031948, -0.22096234560012817, -0.01816907525062561, 0.27505168318748474, -0.23169873654842377, 0.1988259255886078, 0.28876686096191406, -0.28769582509994507, 0.14816173911094666, 0.07204973697662354, 0.18649309873580933, -0.11101426184177399, 0.47624003887176514, 0.17128786444664001, 0.17858915030956268, 0.3287809193134308, -0.532170832157135, 0.20310787856578827, 0.6362364888191223, -0.01515463087707758, -0.1437106430530548, 0.1793668419122696, 0.01802227273583412, -0.2019820362329483, -0.036248497664928436, -0.07885102927684784, 0.2486160397529602, 0.10557577759027481, 0.04468768090009689, -0.018214115872979164, -0.09769808501005173, -0.28492337465286255, 0.21820412576198578, 0.10028044879436493, 0.40321019291877747, -0.09976126253604889, 0.026070235297083855, 0.2311626374721527, -0.17332805693149567, 0.0878036767244339, 0.24988047778606415, 0.17551663517951965, -0.24388694763183594, 0.017860880121588707, -0.2885762155056, -0.27924275398254395, -0.055799324065446854, 0.08746649324893951, 0.139637291431427, -0.16931822896003723, -0.0913475751876831, 0.3872901499271393, -0.009227724745869637, -0.08408679813146591, -0.14456342160701752, 0.2974521517753601, 0.14736445248126984, -0.4759063124656677, -0.1068064272403717, -0.06507272273302078, -0.21300120651721954, 0.03152335435152054, 0.37497058510780334, 0.1918809562921524, 0.22233009338378906, 0.008134047500789165, -0.13919593393802643, 0.18115556240081787, -0.08397350460290909, 0.2416810691356659, -0.03193896636366844, 0.13781872391700745, -0.046962346881628036, 0.11223521083593369, -0.07272319495677948, -0.2634592056274414, -0.06183754280209541, -0.42909643054008484, 0.039671752601861954, 0.18683350086212158, 0.017481744289398193, 0.16025826334953308, -0.23621904850006104, -0.1386367678642273, -0.28904715180397034, -0.39600300788879395, -0.018298322334885597, 0.1862497627735138, 0.32266610860824585, 0.4042643904685974, 0.0039605265483260155, 0.2540593147277832, 0.393064022064209, 0.06724508851766586, 0.08527900278568268, -0.15549074113368988, 0.4933820962905884, -0.27669811248779297, -0.2668294310569763, -0.14344799518585205, -0.14479824900627136, 0.24469903111457825, 0.368446946144104, -0.6950927376747131, 0.2113615721464157, -0.0967334508895874, 0.1460212916135788, -0.09841293096542358, -0.015865687280893326, 0.22807812690734863, -0.2083778977394104, 0.06874624639749527, -0.01793673448264599, 0.09904482215642929, 0.13033054769039154, 0.02588055096566677, -0.1436576098203659, 0.4220215380191803, 0.4722149074077606, 0.31678804755210876, 0.7827595472335815, -0.12301839888095856, -0.24631240963935852, 0.28384584188461304, -0.22802145779132843, 0.25997820496559143, -0.19186244904994965, -0.016254866495728493, 0.0662873163819313, 0.04867066442966461, -0.06577876955270767, 0.23008982837200165, -0.10858732461929321, -0.13981179893016815, -0.15558433532714844, 0.2244047224521637, 0.09555874019861221, -0.22741520404815674, 0.22539600729942322, -0.2080325037240982, 0.21618549525737762, 0.09425517171621323, 0.22961772978305817, -0.3376867473125458, -0.42883190512657166, -0.02579827606678009, -0.177210733294487, 0.4478471577167511, -0.05506278574466705, -0.2819492518901825, 0.13209529221057892, -0.6937564611434937, 0.0947369858622551, 0.04768572002649307, 0.3188587725162506, 0.07766298204660416, -0.2423044890165329, 0.004207187797874212, 0.06520787626504898, 0.6028187870979309, 0.04434771090745926, -0.08510803431272507, 0.2141726315021515, -0.19752942025661469, -0.5001956224441528, 0.022524576634168625, -0.031511418521404266, -0.08884205669164658, 0.2648511826992035, 0.12026425451040268, -0.29280272126197815, -0.5042571425437927, -0.07943680137395859, -0.00020346706151030958, -0.3595183193683624, -0.09357253462076187, -0.22388868033885956, -0.26167425513267517, -0.028192788362503052, -0.38269078731536865, 0.021887527778744698, 0.30103784799575806, -0.1224641501903534, 0.02017262764275074, -0.1732185035943985, 0.001603724085725844, 0.2339189052581787, 0.011684953235089779, 0.1428959220647812, 0.4014412462711334, 0.040681783109903336, -0.061452481895685196, -0.034300047904253006, -0.09876597672700882, 0.20590762794017792, -0.09806337207555771, -0.1612318754196167, 0.036569517105817795, 0.15343141555786133, 0.36917638778686523, 0.03178919851779938, 0.02326355315744877, 0.12791909277439117, -0.14821946620941162, 0.038300905376672745, -0.14445938169956207, 0.3786458969116211, 0.09082195162773132, -0.3453334867954254, -0.07255542278289795, -0.5526953935623169, 0.5314414501190186, 0.29880064725875854, 0.15192659199237823, 0.4885193407535553, -0.19536982476711273, -0.35578614473342896, 0.49162474274635315, 0.396874338388443, 0.5634971261024475, -0.3789258599281311, 0.48827844858169556, 0.7028645873069763, 0.5219011902809143, 0.9219173789024353, -0.38402482867240906, 0.28374183177948, -0.11130274832248688, -0.385222464799881, -0.07143624871969223, -0.2753489017486572, -0.32653138041496277, 0.13125759363174438, 0.13309505581855774, 0.24553123116493225, 0.1267232447862625, 0.2848155200481415, 0.2412850260734558, 0.3513437807559967, 0.3585636615753174, -0.4072933495044708, -0.21223026514053345, 0.013731996528804302, 0.051649484783411026, -0.14456020295619965, -0.11486571282148361, -0.20799125730991364, 0.21388408541679382, 0.05818139761686325, -0.014009100385010242, 0.15659713745117188, -0.26246488094329834, 0.21527208387851715, 0.13642588257789612, -0.13545134663581848, 0.06786584854125977, 0.2597348392009735, 0.12310487031936646, -0.18345627188682556, -0.07790299504995346, -0.07857682555913925, 0.061042603105306625, -0.011666083708405495, 0.14572300016880035, -0.04938458278775215, 0.6279590725898743, -0.05613862723112106, 0.19453710317611694, 0.1661253124475479, -0.08871261775493622, -0.571548581123352, -0.1340983510017395, -0.14428333938121796, 0.07634370028972626, 0.18374967575073242, -0.5355886816978455, 0.28123295307159424, -0.2824200689792633, 0.2070796638727188, 0.05103131756186485, 0.13851426541805267, -0.22972017526626587, 0.039809878915548325, 0.016621697694063187, -0.192434623837471, 0.20765309035778046, 0.035004641860723495, 0.09301959723234177, 0.03120243363082409, 0.373372346162796, 0.22355785965919495, 0.02071729674935341, -0.017776818946003914, 0.08544106781482697, 0.28490737080574036, -0.07938233762979507, 0.0800856202840805, 0.026713965460658073, 0.20169711112976074, 0.07773012667894363, 0.3733218014240265, -0.18203824758529663, 0.0651799887418747, -0.2921448349952698, -0.16977271437644958, -0.2766074538230896, 0.3394337296485901, -0.08160224556922913, 0.1277526617050171, -0.04946049675345421, 0.21621233224868774, -0.10642240941524506, 0.24101051688194275, -0.18489331007003784, -0.05398191139101982, 0.002245950046926737, 0.04347416013479233, -0.14907680451869965, -0.017296282574534416, 0.04800492897629738, 0.11890821158885956, 0.027415882796049118, 0.12441655248403549, -0.32206106185913086, -0.14280183613300323, -0.027802275493741035, 0.20201465487480164, 0.051952581852674484, -0.2506323754787445, -0.023800848051905632, -0.3716358542442322, -0.3312898576259613, -0.20353351533412933, 0.11228173226118088, -0.11092831194400787, -0.2844444811344147, 0.02243868261575699, -0.078731469810009, 0.36041149497032166, -0.2519085109233856, 0.061955247074365616, -0.0862828865647316, 0.3406100571155548, -0.09359583258628845, 0.25554782152175903, 0.15018928050994873, -0.022436855360865593, -0.37363654375076294, -0.12177389860153198, 0.17380164563655853, -0.15421320497989655, 0.12140530347824097, -0.328571081161499, -0.07203742116689682, -0.257597953081131, 0.024638677015900612, 0.3140626549720764, -0.19655245542526245, 0.12919358909130096, 0.22686819732189178, 0.06306266039609909, -0.1053430587053299, -0.11044999212026596, 0.21441708505153656, 0.026152458041906357, -0.03034953773021698, 0.2293282449245453, 0.07449854910373688, 0.15259398519992828, -0.10686349123716354, -0.03071664273738861, 0.0404677577316761, -0.572433590888977, -0.18100404739379883, 0.4083302319049835, -0.07960913330316544, 0.0684163048863411, -0.17246757447719574, 0.10472401231527328, 0.13364669680595398, 0.13838393986225128, 0.30736637115478516, 0.17160165309906006, -0.1019175574183464, -0.08759202808141708, -0.04754037410020828, -0.37738367915153503, -0.4539463520050049, 0.48343321681022644, -0.04196338355541229, 0.5714566707611084, -0.26525574922561646, 0.0016516963951289654, 0.11766763776540756, -0.16160432994365692, -0.17609821259975433, 0.28459861874580383, -0.10150345414876938, -0.018113192170858383, -0.2557772099971771, 0.05827958509325981, -0.16195832192897797, 0.273561954498291, -0.09264330565929413, -0.2304716408252716, -0.21142855286598206, 0.3268534243106842, 0.12374904006719589, 0.16980482637882233, 0.0880313366651535, 0.08871407806873322, 0.20860172808170319, -0.1412576586008072, 0.24054566025733948, -0.28913503885269165, -0.09723278880119324, -0.3500281870365143, 0.015347659587860107, 0.08870036154985428, 0.3145703673362732, -0.30286291241645813, 0.12886179983615875, 0.10813026875257492, -0.06667692959308624, -0.015008625574409962, 0.260098934173584, 0.23029519617557526, 0.11595559865236282, 0.2454717606306076, -0.045211311429739, -0.025874655693769455, 0.14597779512405396, 0.18008217215538025, 0.16942836344242096, -0.5887805223464966, 0.6027830839157104, 0.18034861981868744, -0.09682058542966843, -0.00834179762750864, 0.1899622529745102, -0.05924702063202858, -0.03979833424091339, 0.4897260069847107, -0.2684251368045807, 0.039919592440128326, 0.056890569627285004, 0.013276300393044949, -0.18620267510414124, 0.6324132680892944, 0.08844910562038422, 0.491703063249588, -0.10871725529432297, 0.026488665491342545, -0.46973177790641785, 0.1212523877620697, -0.11352182924747467, 0.13867077231407166, 0.25990134477615356, 0.2758074104785919, 0.15131770074367523, 0.17945979535579681, -0.43608278036117554, -0.18453675508499146, 0.0054649175144732, 0.09761244058609009, -0.12733004987239838, -0.15381582081317902, -0.49346449971199036, 0.1202746108174324, -0.15698663890361786, -0.4985116422176361, 0.0545867383480072, -0.413902223110199, -0.10423564165830612, -0.09382867813110352, 0.25318029522895813, 0.3283686935901642, 0.4639805257320404, 0.2814912497997284, -0.033510785549879074, -0.04955597221851349, 0.0429597944021225, -0.08725900202989578, -0.1108313724398613, -0.02888859063386917, -0.35535362362861633, -0.049638502299785614, -0.17677170038223267, 0.07457433640956879, -0.2796945571899414, -0.025637291371822357, -0.3028123378753662, -0.24400842189788818, 0.017998358234763145, -0.08510861545801163, -0.1251232624053955, 0.0793658122420311, -0.2918234169483185, 0.18006528913974762, 0.32479456067085266, 0.3962629437446594, -0.129053995013237, -0.13008856773376465, -0.29265573620796204, 0.14886410534381866, 0.36815690994262695, -0.6422930955886841, -0.17142201960086823, 0.04930349811911583, -0.15187107026576996, 0.24562609195709229, 0.0641321912407875, -0.279304176568985, -0.20046310126781464, 0.30683737993240356, -0.12597031891345978, -0.19656646251678467, 0.19951149821281433, 0.3359585702419281, -0.05168238654732704, -0.11604385823011398, 0.27684783935546875, -0.0313042588531971, -0.4617125988006592, -0.2316129207611084, -0.4398326277732849 ]
https://github.com/huggingface/datasets/issues/193
[Tensorflow] Use something else than `from_tensor_slices()`
Could you send me the code you used to run create the dataset using `.from_generator` ? What version of tensorflow are you using ?
In the example notebook, the TF Dataset is built using `from_tensor_slices()` : ```python columns = ['input_ids', 'token_type_ids', 'attention_mask', 'start_positions', 'end_positions'] train_tf_dataset.set_format(type='tensorflow', columns=columns) features = {x: train_tf_dataset[x] for x in columns[:3]} labels = {"output_1": train_tf_dataset["start_positions"]} labels["output_2"] = train_tf_dataset["end_positions"] tfdataset = tf.data.Dataset.from_tensor_slices((features, labels)).batch(8) ``` But according to [official tensorflow documentation](https://www.tensorflow.org/guide/data#consuming_numpy_arrays), this will load the entire dataset to memory. **This defeats one purpose of this library, which is lazy loading.** Is there any other way to load the `nlp` dataset into TF dataset lazily ? --- For example, is it possible to use [Arrow dataset](https://www.tensorflow.org/io/api_docs/python/tfio/arrow/ArrowDataset) ? If yes, is there any code example ?
24
[Tensorflow] Use something else than `from_tensor_slices()` In the example notebook, the TF Dataset is built using `from_tensor_slices()` : ```python columns = ['input_ids', 'token_type_ids', 'attention_mask', 'start_positions', 'end_positions'] train_tf_dataset.set_format(type='tensorflow', columns=columns) features = {x: train_tf_dataset[x] for x in columns[:3]} labels = {"output_1": train_tf_dataset["start_positions"]} labels["output_2"] = train_tf_dataset["end_positions"] tfdataset = tf.data.Dataset.from_tensor_slices((features, labels)).batch(8) ``` But according to [official tensorflow documentation](https://www.tensorflow.org/guide/data#consuming_numpy_arrays), this will load the entire dataset to memory. **This defeats one purpose of this library, which is lazy loading.** Is there any other way to load the `nlp` dataset into TF dataset lazily ? --- For example, is it possible to use [Arrow dataset](https://www.tensorflow.org/io/api_docs/python/tfio/arrow/ArrowDataset) ? If yes, is there any code example ? Could you send me the code you used to run create the dataset using `.from_generator` ? What version of tensorflow are you using ?
[ -0.16860908269882202, -0.0939883142709732, 0.011844159103929996, 0.16997626423835754, 0.27237075567245483, 0.08235664665699005, 0.22564618289470673, 0.38973090052604675, -0.05037693306803703, 0.005148948170244694, -0.08119332045316696, 0.6080825328826904, -0.1429208666086197, -0.014412811025977135, 0.2508257031440735, -0.2165248841047287, -0.10149414092302322, 0.25961896777153015, -0.14962494373321533, -0.09359471499919891, -0.014412195421755314, -0.03481737896800041, -0.14135682582855225, -0.027524571865797043, -0.060128532350063324, -0.0839260071516037, 0.07913295179605484, -0.09566803276538849, 0.3711792528629303, 0.039042599499225616, 0.2916140556335449, 0.0025851076934486628, -0.024421097710728645, 0.16511112451553345, -0.000103201404272113, 0.321574330329895, 0.03437275066971779, -0.14773273468017578, 0.0488395020365715, -0.07913176715373993, 0.16531766951084137, -0.2739708423614502, 0.1919616311788559, -0.302823007106781, -0.054379191249608994, 0.00816937256604433, 0.19384154677391052, -0.0288156196475029, 0.3141273260116577, 0.15881867706775665, 0.2169724553823471, 0.4294990301132202, -0.11771481484174728, 0.30170243978500366, 0.11722374707460403, -0.17708224058151245, -0.1900474727153778, -0.22906723618507385, -0.21755370497703552, -0.014430584385991096, 0.03290586546063423, 0.26630154252052307, -0.2041039913892746, 0.1305106282234192, 0.4554981589317322, 0.14307770133018494, -0.11107146739959717, -0.37035706639289856, -0.26494020223617554, 0.37896573543548584, 0.2244817018508911, -0.08395063877105713, -0.15154649317264557, -0.4547233283519745, -0.17095236480236053, -0.3064981698989868, -0.1505763679742813, 0.11813352257013321, -0.2600310444831848, 0.12261244654655457, -0.10149893909692764, -0.0880126953125, -0.2320559322834015, 0.14483609795570374, 0.1663920134305954, -0.04291354492306709, 0.08226271718740463, 0.10815596580505371, 0.15611250698566437, 0.04901333898305893, 0.24615341424942017, 0.05889127403497696, 0.35808202624320984, 0.36975085735321045, -0.34041672945022583, 0.053697697818279266, 0.17792171239852905, -0.5166307687759399, -0.0021291824523359537, -0.25052571296691895, 0.323152095079422, 0.17214292287826538, -0.2449934482574463, 0.1835702806711197, -0.0725676640868187, 0.012311302125453949, -0.417817622423172, 0.11400589346885681, -0.11787019670009613, -0.4833918511867523, -0.023928627371788025, 0.02170192077755928, -0.2394130676984787, -0.0485653281211853, 0.1193627119064331, -0.4675557315349579, -0.03389643877744675, 0.2352779507637024, -0.375856876373291, -0.37095242738723755, -0.1966252624988556, 0.056531332433223724, 0.013881736434996128, 0.27162685990333557, -0.18506430089473724, 0.4295896291732788, 0.15240775048732758, -0.10676467418670654, -0.33988818526268005, 0.004123578313738108, -0.2882518768310547, 0.12291725724935532, -0.12843964993953705, -0.11231785267591476, 0.14382225275039673, -0.07857140153646469, 0.20242609083652496, -0.08082997053861618, -0.17189054191112518, 0.1340389847755432, 0.2805541455745697, -0.101080022752285, 0.08383265882730484, 0.21504884958267212, 0.05136311054229736, -0.013148772530257702, 0.04464416950941086, 0.09676892310380936, -0.16510321199893951, 0.3000503182411194, -0.514404833316803, -0.24342842400074005, -0.034085553139448166, 0.20138071477413177, 0.11621610075235367, -0.14932744204998016, -0.25425800681114197, 0.3058810830116272, 0.13064271211624146, -0.0581669844686985, -0.17962691187858582, -0.1691521853208542, -0.41828593611717224, -0.327901691198349, 0.31852224469184875, 0.058431558310985565, -0.44932523369789124, -0.07594025135040283, -0.02289506234228611, -0.00007904658559709787, 0.2840334475040436, 0.41180580854415894, -0.20259393751621246, 0.47592002153396606, -0.06604285538196564, 0.06232558935880661, 0.727269172668457, -0.0797073096036911, -0.20461837947368622, 0.2472715526819229, -0.18754488229751587, 0.16651448607444763, -0.05649450048804283, 0.3881361186504364, 0.01457935106009245, 0.006611118093132973, 0.38259825110435486, 0.6268609762191772, -0.13027796149253845, 0.12627558410167694, -0.16898010671138763, -0.2685699462890625, 0.3044627904891968, 0.2909621000289917, 0.05505381152033806, -0.09941801428794861, -0.26903194189071655, 0.353055477142334, 0.140915647149086, -0.04823559895157814, -0.17051942646503448, 0.174309641122818, -0.10140753537416458, 0.03389362990856171, -0.19647015631198883, 0.023536941036581993, -0.6025983691215515, 0.2567141354084015, 0.2188519388437271, 0.11899318546056747, 0.148202583193779, -0.09703924506902695, 0.2708936631679535, 0.11006123572587967, 0.03228212893009186, 0.15630656480789185, 0.11245046555995941, 0.11767584085464478, -0.017310423776507378, -0.010448507964611053, -0.25076282024383545, 0.21397973597049713, -0.37623199820518494, -0.16274890303611755, -0.357413113117218, 0.20851120352745056, 0.30177509784698486, -0.035390257835388184, 0.004790897481143475, 0.2599492073059082, -0.2557675838470459, -0.11474409699440002, 0.05725532025098801, 0.275302529335022, -0.028167294338345528, 0.23908819258213043, -0.626107931137085, 0.3982262909412384, -0.016409412026405334, -0.06816964596509933, -0.05593867972493172, 0.145326167345047, 0.04535258933901787, -0.1716870367527008, 0.01637233793735504, 0.2203737497329712, -0.19826100766658783, 0.0816417783498764, 0.212279811501503, -0.250498503446579, 0.17031976580619812, 0.02143077738583088, 0.17780974507331848, -0.09024905413389206, 0.39968007802963257, 0.0517190545797348, 0.12820971012115479, 0.2929064631462097, -0.5027151703834534, 0.24715454876422882, 0.6455686092376709, 0.0017065012361854315, -0.060610488057136536, 0.21915079653263092, -0.004607416223734617, -0.18649259209632874, -0.11462320387363434, -0.07552824914455414, 0.22046244144439697, 0.23334971070289612, 0.004561163019388914, -0.049156319350004196, -0.1739036738872528, -0.23643594980239868, 0.22989444434642792, 0.10348106920719147, 0.39502036571502686, -0.09640909731388092, 0.046553224325180054, 0.15590418875217438, -0.21102270483970642, 0.0506056472659111, 0.22332805395126343, 0.28673669695854187, -0.20513209700584412, -0.03855074197053909, -0.24321706593036652, -0.262923002243042, -0.07050086557865143, 0.018393103033304214, 0.10047055035829544, -0.1776508241891861, -0.06376740336418152, 0.3268815577030182, -0.005278523545712233, -0.0022929818369448185, -0.10702654719352722, 0.2660946249961853, 0.2539460361003876, -0.36759209632873535, -0.04812665283679962, -0.09654681384563446, -0.2597520053386688, 0.12326311320066452, 0.34578046202659607, 0.26507583260536194, 0.3060995042324066, 0.042402755469083786, -0.029359307140111923, 0.1571473479270935, -0.029309317469596863, 0.16931883990764618, -0.18968982994556427, 0.12898238003253937, -0.07206517457962036, 0.2025318294763565, -0.10920284688472748, -0.21997714042663574, -0.0401364304125309, -0.42298489809036255, 0.045517344027757645, 0.19250079989433289, 0.016064422205090523, 0.05312606319785118, -0.2264474630355835, -0.22006995975971222, -0.2942710816860199, -0.42354512214660645, -0.011317514814436436, 0.11504660546779633, 0.28670719265937805, 0.46687278151512146, -0.03242749720811844, 0.27440401911735535, 0.35955312848091125, 0.1347765177488327, 0.02100195549428463, -0.13184277713298798, 0.5142490863800049, -0.29144060611724854, -0.2923905849456787, -0.1259395033121109, -0.15778841078281403, 0.21387909352779388, 0.26455411314964294, -0.6998732686042786, 0.1633124202489853, -0.18016670644283295, 0.09063251316547394, -0.17192885279655457, -0.023034512996673584, 0.2796552777290344, -0.18283778429031372, -0.037740156054496765, -0.06421414017677307, 0.1282201111316681, 0.15930955111980438, -0.0026163591537624598, -0.1744402050971985, 0.29112568497657776, 0.43631571531295776, 0.293207585811615, 0.6823471188545227, -0.04952210187911987, -0.2472585290670395, 0.24511626362800598, -0.20342116057872772, 0.1798972189426422, -0.15880434215068817, -0.12973561882972717, 0.0691831111907959, 0.0468609593808651, -0.0193074531853199, 0.26274287700653076, -0.07912977039813995, -0.19225268065929413, -0.085463747382164, 0.19334043562412262, 0.023228144273161888, -0.2450515329837799, 0.2765369117259979, -0.19784517586231232, 0.17820201814174652, 0.10248145461082458, 0.17246198654174805, -0.38271084427833557, -0.43289661407470703, -0.02819054014980793, -0.21834143996238708, 0.36756935715675354, -0.05496933311223984, -0.3525265157222748, 0.12331748008728027, -0.6885600686073303, 0.0714336633682251, 0.09516609460115433, 0.29508161544799805, 0.044006720185279846, -0.26446864008903503, 0.027434788644313812, 0.06361322849988937, 0.5770878791809082, 0.08534533530473709, -0.03744007274508476, 0.1942191869020462, -0.060622137039899826, -0.5440307855606079, 0.02768373303115368, -0.08388595283031464, -0.02656622976064682, 0.25266334414482117, 0.13720068335533142, -0.3163086473941803, -0.5172992944717407, -0.05490103363990784, 0.011884755454957485, -0.3238765299320221, -0.1030321791768074, -0.2559071481227875, -0.24362227320671082, -0.012445760890841484, -0.29287317395210266, 0.06926226615905762, 0.2637309432029724, -0.1118287444114685, -0.00764614250510931, -0.1052023321390152, -0.0017561325803399086, 0.2264772653579712, 0.021024586632847786, 0.21798676252365112, 0.4974190592765808, 0.09195953607559204, -0.006784370634704828, 0.044760625809431076, -0.17461474239826202, 0.2100192904472351, -0.10321379452943802, -0.12122233957052231, -0.0029219656717032194, 0.11823977530002594, 0.3274330794811249, 0.0437074676156044, 0.04576598480343819, 0.08590217679738998, -0.21129585802555084, 0.11721077561378479, -0.2157900184392929, 0.41077524423599243, 0.1617109179496765, -0.2892138361930847, -0.1019580140709877, -0.45165881514549255, 0.4895877242088318, 0.22257544100284576, 0.10899365693330765, 0.4912775456905365, -0.17228466272354126, -0.3628089427947998, 0.4391491115093231, 0.3384194076061249, 0.540488600730896, -0.3218809962272644, 0.4336532652378082, 0.7526196837425232, 0.48587971925735474, 0.8558765649795532, -0.35158705711364746, 0.28582462668418884, -0.15107682347297668, -0.3399319648742676, 0.006260357331484556, -0.2254667729139328, -0.4014180898666382, 0.06774912029504776, 0.06809141486883163, 0.18563011288642883, 0.08648081123828888, 0.22489535808563232, 0.17575803399085999, 0.39436402916908264, 0.36009663343429565, -0.4310417175292969, -0.2145262360572815, 0.14980219304561615, -0.036545854061841965, -0.15512986481189728, -0.05978782847523689, -0.20871643722057343, 0.1818535476922989, 0.06399659067392349, 0.02201472967863083, 0.15056204795837402, -0.1861504167318344, 0.28352025151252747, 0.0552060604095459, -0.06650636345148087, -0.013074704445898533, 0.2632390558719635, 0.12207864969968796, -0.17733293771743774, -0.09848332405090332, -0.057642027735710144, 0.02004092186689377, 0.003613065928220749, 0.08650219440460205, -0.05334313586354256, 0.6400732398033142, -0.049178361892700195, 0.13697384297847748, 0.1090063527226448, -0.07215574383735657, -0.48971840739250183, -0.19590513408184052, -0.11551954597234726, 0.08588781952857971, 0.08319542557001114, -0.5479530096054077, 0.23642663657665253, -0.20111961662769318, 0.18663033843040466, 0.16215384006500244, 0.12269847095012665, -0.1949729025363922, 0.09241042286157608, 0.09988853335380554, -0.2133835256099701, 0.22173047065734863, 0.035102471709251404, 0.1007731482386589, 0.09124832600355148, 0.3558175265789032, 0.22108863294124603, -0.02595352567732334, -0.07953140884637833, 0.08418277651071548, 0.303434282541275, -0.06420739740133286, 0.035457175225019455, 0.04730672389268875, 0.2883150577545166, 0.0701756626367569, 0.34780314564704895, -0.2279077172279358, 0.014586657285690308, -0.19920094311237335, -0.1821100115776062, -0.3068374693393707, 0.33475157618522644, -0.06750871241092682, 0.12765027582645416, -0.025997038930654526, 0.21559080481529236, -0.07443264871835709, 0.2593861520290375, -0.3228650987148285, -0.04293421283364296, -0.02249479852616787, 0.07561125606298447, -0.10007336735725403, -0.12442184239625931, 0.09797804802656174, 0.09842175245285034, 0.14725002646446228, 0.12966051697731018, -0.3589875102043152, -0.2738024890422821, -0.02206369861960411, 0.11938424408435822, 0.02181496098637581, -0.23054175078868866, -0.024568961933255196, -0.3880217671394348, -0.3245268762111664, -0.17464731633663177, 0.10038861632347107, -0.1464446485042572, -0.291619211435318, 0.059871856123209, -0.11217328161001205, 0.27732640504837036, -0.09360291063785553, 0.008463822305202484, -0.17180182039737701, 0.2840944528579712, -0.06256347894668579, 0.1911996603012085, 0.1374155431985855, -0.07968724519014359, -0.38825055956840515, -0.13485835492610931, 0.12812569737434387, -0.10049517452716827, -0.0008417729986831546, -0.3645169138908386, -0.08045923709869385, -0.20343394577503204, 0.09064158797264099, 0.24003998935222626, -0.2507553696632385, 0.03492068126797676, 0.2132696509361267, 0.20128333568572998, -0.12772296369075775, -0.10396546870470047, 0.19983313977718353, 0.023196766152977943, 0.061839260160923004, 0.12107637524604797, -0.021087413653731346, 0.16453807055950165, -0.10702791064977646, 0.015114977955818176, 0.051866959780454636, -0.5421461462974548, -0.2646961808204651, 0.32220467925071716, -0.15680605173110962, -0.03405848145484924, -0.0866410955786705, 0.13565951585769653, 0.12472379952669144, 0.2581654489040375, 0.318068265914917, 0.23370763659477234, -0.07991089671850204, -0.07542397826910019, -0.1047290563583374, -0.36424413323402405, -0.40943652391433716, 0.41810232400894165, -0.13466966152191162, 0.5021535754203796, -0.2514804005622864, 0.02721542678773403, 0.119999960064888, -0.2684732973575592, -0.17018377780914307, 0.305292010307312, -0.21323716640472412, -0.011488892138004303, -0.24748295545578003, 0.0629415437579155, -0.11182516068220139, 0.2272561490535736, -0.06825169175863266, -0.19298410415649414, -0.18089869618415833, 0.29027432203292847, 0.0493139773607254, 0.1948872059583664, 0.07147729396820068, 0.1825227588415146, 0.2388373613357544, -0.12766538560390472, 0.23899801075458527, -0.227150097489357, -0.09096414595842361, -0.4048541188240051, 0.02497979998588562, 0.10322364419698715, 0.3352166712284088, -0.3399098515510559, 0.12908825278282166, 0.09992969036102295, -0.10254424810409546, -0.06795168668031693, 0.23784439265727997, 0.2075456827878952, 0.1221882775425911, 0.29464420676231384, 0.0724024698138237, -0.12988561391830444, 0.14953717589378357, 0.19643688201904297, 0.1768037974834442, -0.46683019399642944, 0.49007338285446167, 0.1949758380651474, -0.020242204889655113, -0.1069500520825386, 0.14637622237205505, -0.18310898542404175, -0.0890938863158226, 0.5219664573669434, -0.24887971580028534, 0.04972735419869423, -0.08820942044258118, 0.07930823415517807, -0.22433096170425415, 0.693946123123169, 0.033031169325113297, 0.4624565839767456, -0.15681688487529755, 0.010074657388031483, -0.48593637347221375, 0.09705610573291779, -0.19249588251113892, 0.1519232839345932, 0.24726253747940063, 0.33283060789108276, 0.10673568397760391, 0.18085801601409912, -0.43699467182159424, -0.1848956048488617, -0.05393637716770172, 0.11275415122509003, -0.20269343256950378, -0.0650595873594284, -0.5096164345741272, 0.11405090987682343, -0.07576990127563477, -0.4709778428077698, 0.11253862082958221, -0.43344566226005554, 0.056311171501874924, -0.09103833138942719, 0.27836525440216064, 0.28070709109306335, 0.41367268562316895, 0.33369943499565125, -0.041489873081445694, -0.04722544550895691, -0.019792117178440094, -0.10569372028112411, -0.12665586173534393, 0.023456741124391556, -0.34460657835006714, -0.07571838051080704, -0.05020656809210777, 0.21800051629543304, -0.23613139986991882, -0.03747343271970749, -0.23950158059597015, -0.2059686928987503, 0.09328244626522064, -0.11421532928943634, -0.09886130690574646, 0.11413324624300003, -0.1716417670249939, 0.18321624398231506, 0.1843811273574829, 0.4326111376285553, -0.09274666756391525, -0.056560780853033066, -0.27200257778167725, -0.037574879825115204, 0.46004125475883484, -0.5970892310142517, -0.2347230464220047, 0.09746205806732178, -0.06007571518421173, 0.22572462260723114, 0.04289274662733078, -0.29877734184265137, -0.15917865931987762, 0.2910028398036957, -0.13124839961528778, -0.24028590321540833, 0.2773433327674866, 0.3229186534881592, -0.08262793719768524, -0.08083834499120712, 0.2904898226261139, -0.06068572402000427, -0.39346352219581604, -0.1542896330356598, -0.4548880159854889 ]
https://github.com/huggingface/datasets/issues/193
[Tensorflow] Use something else than `from_tensor_slices()`
I'm using TF2.2 Here is my code : ``` import nlp from transformers import BartTokenizer tokenizer = BartTokenizer.from_pretrained('bart-large') def encode(sample): article_inputs = tokenizer.encode_plus(sample["article"], max_length=tokenizer.model_max_length, pad_to_max_length=True) summary_inputs = tokenizer.encode_plus(sample["highlights"], max_length=tokenizer.model_max_length, pad_to_max_length=True) article_inputs.update({"lm_labels": summary_inputs['input_ids']}) return article_inputs cnn_dm = nlp.load_dataset('cnn_dailymail', '3.0.0', split='test') cnn_dm = cnn_dm.map(encode) def gen(): for sample in cnn_dm: s = {} s['input_ids'] = sample['input_ids'] s['attention_mask'] = sample['attention_mask'] s['lm_labels'] = sample['lm_labels'] yield s dataset = tf.data.Dataset.from_generator(gen, output_types={k: tf.int32 for k in ['input_ids', 'attention_mask', 'lm_labels']}, output_shapes={k: tf.TensorShape([tokenizer.model_max_length]) for k in ['input_ids', 'attention_mask', 'lm_labels']} ```
In the example notebook, the TF Dataset is built using `from_tensor_slices()` : ```python columns = ['input_ids', 'token_type_ids', 'attention_mask', 'start_positions', 'end_positions'] train_tf_dataset.set_format(type='tensorflow', columns=columns) features = {x: train_tf_dataset[x] for x in columns[:3]} labels = {"output_1": train_tf_dataset["start_positions"]} labels["output_2"] = train_tf_dataset["end_positions"] tfdataset = tf.data.Dataset.from_tensor_slices((features, labels)).batch(8) ``` But according to [official tensorflow documentation](https://www.tensorflow.org/guide/data#consuming_numpy_arrays), this will load the entire dataset to memory. **This defeats one purpose of this library, which is lazy loading.** Is there any other way to load the `nlp` dataset into TF dataset lazily ? --- For example, is it possible to use [Arrow dataset](https://www.tensorflow.org/io/api_docs/python/tfio/arrow/ArrowDataset) ? If yes, is there any code example ?
82
[Tensorflow] Use something else than `from_tensor_slices()` In the example notebook, the TF Dataset is built using `from_tensor_slices()` : ```python columns = ['input_ids', 'token_type_ids', 'attention_mask', 'start_positions', 'end_positions'] train_tf_dataset.set_format(type='tensorflow', columns=columns) features = {x: train_tf_dataset[x] for x in columns[:3]} labels = {"output_1": train_tf_dataset["start_positions"]} labels["output_2"] = train_tf_dataset["end_positions"] tfdataset = tf.data.Dataset.from_tensor_slices((features, labels)).batch(8) ``` But according to [official tensorflow documentation](https://www.tensorflow.org/guide/data#consuming_numpy_arrays), this will load the entire dataset to memory. **This defeats one purpose of this library, which is lazy loading.** Is there any other way to load the `nlp` dataset into TF dataset lazily ? --- For example, is it possible to use [Arrow dataset](https://www.tensorflow.org/io/api_docs/python/tfio/arrow/ArrowDataset) ? If yes, is there any code example ? I'm using TF2.2 Here is my code : ``` import nlp from transformers import BartTokenizer tokenizer = BartTokenizer.from_pretrained('bart-large') def encode(sample): article_inputs = tokenizer.encode_plus(sample["article"], max_length=tokenizer.model_max_length, pad_to_max_length=True) summary_inputs = tokenizer.encode_plus(sample["highlights"], max_length=tokenizer.model_max_length, pad_to_max_length=True) article_inputs.update({"lm_labels": summary_inputs['input_ids']}) return article_inputs cnn_dm = nlp.load_dataset('cnn_dailymail', '3.0.0', split='test') cnn_dm = cnn_dm.map(encode) def gen(): for sample in cnn_dm: s = {} s['input_ids'] = sample['input_ids'] s['attention_mask'] = sample['attention_mask'] s['lm_labels'] = sample['lm_labels'] yield s dataset = tf.data.Dataset.from_generator(gen, output_types={k: tf.int32 for k in ['input_ids', 'attention_mask', 'lm_labels']}, output_shapes={k: tf.TensorShape([tokenizer.model_max_length]) for k in ['input_ids', 'attention_mask', 'lm_labels']} ```
[ -0.1217658668756485, -0.025931330397725105, 0.04709159582853317, 0.1898200362920761, 0.198004350066185, 0.15442733466625214, 0.24837516248226166, 0.45679765939712524, -0.0371507853269577, -0.05661466717720032, -0.14168518781661987, 0.6093762516975403, -0.18237659335136414, -0.03136470913887024, 0.2451843023300171, -0.1602427065372467, -0.05153365805745125, 0.21017992496490479, -0.08378042280673981, -0.0511188842356205, 0.041434887796640396, -0.10304678976535797, -0.11337538063526154, -0.08961909264326096, -0.08411501348018646, 0.03781047835946083, 0.02680441364645958, -0.15933330357074738, 0.4063454568386078, -0.015847502276301384, 0.3480536937713623, 0.035481035709381104, -0.061999645084142685, 0.10191045701503754, -0.00011394677858334035, 0.38300827145576477, 0.03163405880331993, -0.16971956193447113, 0.0051156398840248585, -0.04868374764919281, 0.1872231513261795, -0.22094693779945374, 0.24761827290058136, -0.2797366976737976, 0.022960418835282326, -0.043305639177560806, 0.22824761271476746, 0.021492211148142815, 0.2741084694862366, 0.056343816220760345, 0.11358030885457993, 0.46258437633514404, -0.22616249322891235, 0.36655694246292114, 0.15347038209438324, -0.10670565068721771, -0.13823117315769196, -0.27631357312202454, -0.21094068884849548, 0.04131274297833443, 0.046295154839754105, 0.2148602306842804, -0.2967180907726288, 0.10010848194360733, 0.4883717894554138, 0.134610116481781, -0.11587438732385635, -0.29013505578041077, -0.20416900515556335, 0.35375380516052246, 0.19012628495693207, -0.07160356640815735, -0.17686021327972412, -0.5691714882850647, -0.23419807851314545, -0.2198171466588974, -0.21410535275936127, 0.04874308779835701, -0.18753649294376373, 0.1386621594429016, -0.21671608090400696, -0.0914962962269783, -0.19212760031223297, 0.16961531341075897, 0.17300373315811157, 0.08846854418516159, 0.1593903750181198, 0.15273478627204895, 0.1636829674243927, 0.040270108729600906, 0.276839017868042, -0.011179623194038868, 0.2941535711288452, 0.4132716655731201, -0.3397204279899597, -0.0049698976799845695, 0.14997687935829163, -0.5529500246047974, -0.002599742729216814, -0.21910324692726135, 0.31478551030158997, 0.21479780972003937, -0.25013092160224915, 0.17884483933448792, -0.006410614121705294, -0.06401248276233673, -0.3965282440185547, 0.14390507340431213, -0.17764891684055328, -0.5946305394172668, 0.0709008276462555, 0.025447657331824303, -0.29087886214256287, 0.09226302802562714, 0.05414406210184097, -0.5405900478363037, -0.09034091234207153, 0.22742724418640137, -0.3082316219806671, -0.45655444264411926, -0.24953961372375488, 0.03222231566905975, 0.028258729726076126, 0.26590296626091003, -0.17248490452766418, 0.49490591883659363, 0.046801015734672546, -0.12600670754909515, -0.235796719789505, 0.07146377861499786, -0.20548029243946075, 0.08548394590616226, -0.05327322706580162, -0.09084247052669525, 0.18941907584667206, -0.04464948922395706, 0.23619937896728516, -0.18066535890102386, -0.15809693932533264, 0.16095097362995148, 0.25847092270851135, -0.09254162758588791, 0.07164032757282257, 0.2176818549633026, -0.014463166706264019, 0.061746373772621155, 0.047353923320770264, 0.1133970394730568, -0.23381264507770538, 0.3567086458206177, -0.5510231852531433, -0.22753320634365082, -0.004840495530515909, 0.08490922302007675, 0.10352467745542526, -0.23398105800151825, -0.3640972673892975, 0.3549087941646576, 0.10241913050413132, -0.12468891590833664, -0.14371740818023682, -0.14855578541755676, -0.4110410511493683, -0.2633945941925049, 0.28865382075309753, 0.11818340420722961, -0.3723021447658539, -0.1996259093284607, 0.013991568237543106, 0.039289262145757675, 0.21835826337337494, 0.4440423250198364, -0.2287718504667282, 0.4910319149494171, -0.06854376941919327, 0.17658226191997528, 0.7466111183166504, -0.16834159195423126, -0.21639315783977509, 0.2665836811065674, -0.10920905321836472, 0.20277419686317444, 0.007078719325363636, 0.48750039935112, 0.055612411350011826, -0.028245238587260246, 0.3278442919254303, 0.6070600152015686, -0.11004310101270676, 0.09091220796108246, -0.06263861805200577, -0.3540174663066864, 0.38336822390556335, 0.22360169887542725, 0.049433767795562744, -0.08908886462450027, -0.24513199925422668, 0.4838311970233917, 0.15792743861675262, 0.00607340456917882, -0.182203009724617, 0.13027849793434143, -0.18597283959388733, 0.005775886122137308, -0.22100558876991272, 0.04396410286426544, -0.6262143850326538, 0.29450780153274536, 0.2898073196411133, 0.019801784306764603, 0.22485041618347168, -0.1320509910583496, 0.30463334918022156, 0.08697132766246796, 0.05847438797354698, 0.1558161824941635, -0.036375753581523895, 0.15324489772319794, -0.008287030272185802, -0.05614372715353966, -0.33982768654823303, 0.2135782092809677, -0.35561296343803406, -0.0859224796295166, -0.28733837604522705, 0.2686130106449127, 0.44628220796585083, 0.006337897852063179, -0.13960523903369904, 0.2558361887931824, -0.23948167264461517, -0.0909075066447258, 0.09431453049182892, 0.22145701944828033, -0.037428561598062515, 0.31023043394088745, -0.6426630616188049, 0.517551839351654, -0.00729947118088603, -0.05198514461517334, -0.03803340718150139, 0.15125638246536255, 0.04083877429366112, -0.22987142205238342, -0.04504171386361122, 0.20381353795528412, -0.1811840683221817, 0.25211265683174133, 0.22757047414779663, -0.2973989248275757, 0.0864400863647461, 0.09497027099132538, 0.14749138057231903, -0.10604090988636017, 0.3818446695804596, 0.08208683133125305, 0.25895917415618896, 0.29437336325645447, -0.4201221764087677, 0.20113877952098846, 0.5873510837554932, -0.03383931890130043, -0.08823513984680176, 0.24291597306728363, -0.08980366587638855, -0.21951782703399658, -0.04968223720788956, -0.15792830288410187, 0.23874761164188385, 0.15569519996643066, 0.04756329581141472, -0.018360352143645287, -0.08513443917036057, -0.18957826495170593, 0.23301106691360474, 0.13817870616912842, 0.34429123997688293, -0.10489462316036224, 0.07945554703474045, 0.24394972622394562, -0.2006930410861969, 0.12063613533973694, 0.2041323184967041, 0.16148746013641357, -0.18216603994369507, 0.02087452821433544, -0.274058997631073, -0.323343962430954, -0.06881626695394516, -0.015634533017873764, 0.11983688175678253, -0.1766929179430008, -0.08395581692457199, 0.35642510652542114, 0.03898758068680763, -0.026909351348876953, -0.1203833818435669, 0.30149045586586, 0.1829601377248764, -0.4852122366428375, -0.14093254506587982, -0.07228980213403702, -0.18762874603271484, 0.012351211160421371, 0.34143123030662537, 0.24763703346252441, 0.21397224068641663, 0.11561937630176544, -0.10640077292919159, 0.09337663650512695, -0.13714541494846344, 0.22248069941997528, -0.1411900520324707, 0.07441695034503937, -0.09138822555541992, 0.08587653189897537, -0.13170845806598663, -0.23741286993026733, -0.13842542469501495, -0.4001227915287018, 0.015284137800335884, 0.17810235917568207, 0.04398329183459282, 0.1407824456691742, -0.17305488884449005, -0.0541105642914772, -0.31138163805007935, -0.36805543303489685, 0.025136662647128105, 0.16438336670398712, 0.26305413246154785, 0.426656574010849, -0.09771568328142166, 0.28752386569976807, 0.2841942608356476, 0.05660080909729004, 0.08437559753656387, -0.0591103695333004, 0.5510885119438171, -0.21195611357688904, -0.20756588876247406, -0.19724544882774353, -0.16827771067619324, 0.12386772036552429, 0.287994921207428, -0.7350274920463562, 0.25820639729499817, -0.1476229876279831, 0.05228998884558678, -0.13607294857501984, -0.04323436692357063, 0.30103906989097595, -0.24622775614261627, 0.057981930673122406, -0.019213825464248657, 0.04298059269785881, 0.0740940049290657, 0.01504224818199873, -0.21246114373207092, 0.28354498744010925, 0.4253312647342682, 0.3563779890537262, 0.8198500275611877, -0.08152435719966888, -0.19571244716644287, 0.2159968614578247, -0.24318481981754303, 0.204217329621315, -0.21248508989810944, -0.07977601885795593, 0.07321752607822418, 0.014921271242201328, -0.019207274541258812, 0.30555540323257446, -0.12085390836000443, -0.23511408269405365, -0.06570623070001602, 0.16245514154434204, 0.07112377882003784, -0.18955527245998383, 0.25377798080444336, -0.23828238248825073, 0.12286172062158585, 0.14369402825832367, 0.22716517746448517, -0.39819401502609253, -0.5145754218101501, -0.05292618274688721, -0.19155701994895935, 0.3439481854438782, -0.038793157786130905, -0.4047642946243286, 0.18230682611465454, -0.7144542932510376, 0.13730929791927338, 0.056387197226285934, 0.33463820815086365, 0.10766397416591644, -0.2996101677417755, 0.02328980714082718, 0.07969038188457489, 0.6119206547737122, 0.16484808921813965, -0.07098272442817688, 0.17361602187156677, -0.15104812383651733, -0.5380699038505554, 0.012871168553829193, -0.051564816385507584, 0.037138599902391434, 0.274997353553772, 0.159707173705101, -0.3015442192554474, -0.5077012777328491, -0.04973502457141876, 0.006743426900357008, -0.352834552526474, -0.06369978934526443, -0.22928285598754883, -0.15957185626029968, -0.03656740114092827, -0.3108157813549042, 0.07019105553627014, 0.23305240273475647, -0.05486323684453964, 0.03534510359168053, -0.2360270768404007, -0.07313350588083267, 0.2959994375705719, -0.007531976792961359, 0.17190705239772797, 0.5554139614105225, 0.068697988986969, -0.05782302841544151, 0.02807164378464222, -0.22288253903388977, 0.22437617182731628, -0.0465519018471241, -0.1779031902551651, 0.020541228353977203, 0.139253631234169, 0.42129504680633545, 0.09153509140014648, 0.0426737442612648, 0.17796486616134644, -0.1819620579481125, 0.026481661945581436, -0.2325233370065689, 0.4516885578632355, 0.10564064234495163, -0.28788378834724426, -0.1127607673406601, -0.4171384274959564, 0.5051946043968201, 0.26663675904273987, 0.09778330475091934, 0.4054187536239624, -0.1845550537109375, -0.37959757447242737, 0.47471266984939575, 0.36818748712539673, 0.5930817127227783, -0.3568761944770813, 0.5228302478790283, 0.6306702494621277, 0.5135160088539124, 0.9229065775871277, -0.36990782618522644, 0.25785931944847107, -0.10761364549398422, -0.2891256809234619, -0.034149255603551865, -0.27011382579803467, -0.36774131655693054, 0.09762505441904068, 0.1489039957523346, 0.2554051876068115, 0.05960842967033386, 0.32816022634506226, 0.22872576117515564, 0.4213234484195709, 0.4004007577896118, -0.41907864809036255, -0.13934682309627533, 0.04883745685219765, -0.00046828846097923815, -0.19989678263664246, -0.08649653941392899, -0.1306753158569336, 0.22034864127635956, -0.03784703090786934, -0.06312455981969833, 0.09271613508462906, -0.19131869077682495, 0.273310124874115, 0.0886058434844017, -0.022878464311361313, -0.030651910230517387, 0.23595966398715973, 0.0516061969101429, -0.16254620254039764, -0.07398103922605515, -0.07821504026651382, 0.020682044327259064, 0.012832365930080414, 0.13242118060588837, -0.07321570068597794, 0.5777004957199097, -0.003746628761291504, 0.23270945250988007, 0.20195771753787994, -0.07018738240003586, -0.6475855112075806, -0.2420331984758377, -0.0774635523557663, 0.1646004617214203, 0.12083633244037628, -0.574008047580719, 0.2775152921676636, -0.21643990278244019, 0.20163758099079132, 0.04024408757686615, 0.05791177973151207, -0.18905116617679596, 0.05720091238617897, 0.08815090358257294, -0.2242264300584793, 0.23804381489753723, 0.07987481355667114, 0.0802137479186058, 0.00434208195656538, 0.40031349658966064, 0.09414923191070557, 0.026445019990205765, 0.0004958930076099932, 0.14066679775714874, 0.31898966431617737, -0.053173381835222244, -0.0355193130671978, -0.032099220901727676, 0.22842785716056824, -0.002215232467278838, 0.31702837347984314, -0.22109848260879517, 0.04519467055797577, -0.32712388038635254, -0.22691558301448822, -0.27978140115737915, 0.40974029898643494, -0.03227900713682175, 0.17035353183746338, -0.02966219186782837, 0.23653832077980042, -0.2120736688375473, 0.25990116596221924, -0.19452986121177673, -0.0003527412482071668, 0.05574426054954529, 0.0645964965224266, -0.20630165934562683, -0.03718273714184761, 0.08591114729642868, 0.14745460450649261, 0.029787026345729828, 0.145842045545578, -0.2863140404224396, -0.16016770899295807, -0.05519567057490349, 0.21042372286319733, 0.06890018284320831, -0.18370740115642548, -0.07681222259998322, -0.3510492146015167, -0.3054195046424866, -0.21098506450653076, 0.07133498787879944, -0.11096139997243881, -0.25265032052993774, -0.017708519473671913, -0.016698475927114487, 0.30331170558929443, -0.18731622397899628, 0.05895800143480301, -0.03881028667092323, 0.32052183151245117, -0.06789502501487732, 0.2631840109825134, 0.17212112247943878, -0.016810934990644455, -0.3609752655029297, -0.11104509234428406, 0.2617635130882263, -0.12928369641304016, -0.019321288913488388, -0.35425958037376404, -0.028004668653011322, -0.24651026725769043, 0.019433265551924706, 0.381717324256897, -0.2598596513271332, 0.0119019141420722, 0.2552919387817383, 0.07606125622987747, -0.06930886954069138, -0.1370902806520462, 0.28984591364860535, 0.01564158871769905, -0.01702932082116604, 0.21574896574020386, 0.019055670127272606, 0.23018820583820343, -0.07639665901660919, -0.059450648725032806, 0.050926804542541504, -0.45812997221946716, -0.2483103722333908, 0.34422567486763, -0.14797449111938477, -0.044531796127557755, -0.15660017728805542, 0.15954932570457458, 0.15881988406181335, 0.2009689062833786, 0.2688860595226288, 0.2720526456832886, -0.1309787482023239, -0.08041998744010925, -0.10907331109046936, -0.2988508343696594, -0.4159930646419525, 0.4332540035247803, -0.03094954416155815, 0.5543481111526489, -0.30145910382270813, -0.04428989067673683, 0.21874752640724182, -0.28898385167121887, -0.14263573288917542, 0.2879599630832672, -0.12157328426837921, -0.09826479852199554, -0.184583380818367, 0.06264065951108932, -0.21620088815689087, 0.2093064934015274, -0.1255277395248413, -0.24214628338813782, -0.18847402930259705, 0.28286218643188477, 0.1213042214512825, 0.14187593758106232, 0.03304439038038254, 0.139256089925766, 0.26242390275001526, -0.16843979060649872, 0.28532472252845764, -0.303865909576416, -0.02345472201704979, -0.30802327394485474, 0.05243302136659622, 0.030013376846909523, 0.32330068945884705, -0.3477114737033844, 0.13549581170082092, 0.07323861122131348, -0.05230730399489403, -0.02184964530169964, 0.2548745274543762, 0.21410450339317322, 0.11187884211540222, 0.2443319857120514, -0.047419473528862, -0.0380629263818264, 0.15545962750911713, 0.13566064834594727, 0.10434849560260773, -0.6335430145263672, 0.5223556756973267, 0.2589552104473114, -0.034853219985961914, 0.0482245534658432, 0.23610079288482666, -0.17642280459403992, -0.1175670474767685, 0.49291667342185974, -0.26624637842178345, 0.07922790199518204, 0.08962046355009079, 0.023243896663188934, -0.23603728413581848, 0.6891624331474304, 0.0156595166772604, 0.5115364789962769, -0.1294793039560318, 0.08085687458515167, -0.48336246609687805, 0.04604009538888931, -0.13137975335121155, 0.1344795525074005, 0.24390113353729248, 0.28714635968208313, 0.15486234426498413, 0.21748192608356476, -0.37715452909469604, -0.29307103157043457, -0.07903575897216797, 0.005898837931454182, -0.12049239128828049, -0.09162572026252747, -0.4695660471916199, 0.20530156791210175, -0.06458131968975067, -0.48753663897514343, 0.053547147661447525, -0.3142370581626892, -0.10486918687820435, -0.06824883818626404, 0.20545029640197754, 0.27082696557044983, 0.3716530501842499, 0.32452285289764404, -0.05926663056015968, 0.0022980268113315105, 0.062266673892736435, -0.10009386390447617, -0.13290560245513916, -0.02840886078774929, -0.3497494161128998, -0.08344747871160507, -0.16258354485034943, 0.1681801974773407, -0.33418476581573486, 0.01335954014211893, -0.2161337435245514, -0.32348838448524475, 0.05041749030351639, -0.20108459889888763, -0.14042317867279053, 0.09030690044164658, -0.30964699387550354, 0.31978514790534973, 0.13747571408748627, 0.4710557162761688, -0.13151848316192627, -0.03896171972155571, -0.19754283130168915, 0.1254427582025528, 0.4762638509273529, -0.6969993114471436, -0.1845293641090393, 0.08059677481651306, -0.13366742432117462, 0.28647273778915405, 0.08193077147006989, -0.3468104600906372, -0.2159188687801361, 0.3673853278160095, -0.06303614377975464, -0.21892155706882477, 0.23914073407649994, 0.3046945631504059, -0.10874563455581665, -0.11808701604604721, 0.3073832392692566, -0.08267579972743988, -0.3657679855823517, -0.19985389709472656, -0.5045337677001953 ]
https://github.com/huggingface/datasets/issues/193
[Tensorflow] Use something else than `from_tensor_slices()`
Apparently we'll have to wait for the next tensorflow release to use `.from_generator` and TPU. See https://github.com/tensorflow/tensorflow/issues/34346#issuecomment-598262489
In the example notebook, the TF Dataset is built using `from_tensor_slices()` : ```python columns = ['input_ids', 'token_type_ids', 'attention_mask', 'start_positions', 'end_positions'] train_tf_dataset.set_format(type='tensorflow', columns=columns) features = {x: train_tf_dataset[x] for x in columns[:3]} labels = {"output_1": train_tf_dataset["start_positions"]} labels["output_2"] = train_tf_dataset["end_positions"] tfdataset = tf.data.Dataset.from_tensor_slices((features, labels)).batch(8) ``` But according to [official tensorflow documentation](https://www.tensorflow.org/guide/data#consuming_numpy_arrays), this will load the entire dataset to memory. **This defeats one purpose of this library, which is lazy loading.** Is there any other way to load the `nlp` dataset into TF dataset lazily ? --- For example, is it possible to use [Arrow dataset](https://www.tensorflow.org/io/api_docs/python/tfio/arrow/ArrowDataset) ? If yes, is there any code example ?
17
[Tensorflow] Use something else than `from_tensor_slices()` In the example notebook, the TF Dataset is built using `from_tensor_slices()` : ```python columns = ['input_ids', 'token_type_ids', 'attention_mask', 'start_positions', 'end_positions'] train_tf_dataset.set_format(type='tensorflow', columns=columns) features = {x: train_tf_dataset[x] for x in columns[:3]} labels = {"output_1": train_tf_dataset["start_positions"]} labels["output_2"] = train_tf_dataset["end_positions"] tfdataset = tf.data.Dataset.from_tensor_slices((features, labels)).batch(8) ``` But according to [official tensorflow documentation](https://www.tensorflow.org/guide/data#consuming_numpy_arrays), this will load the entire dataset to memory. **This defeats one purpose of this library, which is lazy loading.** Is there any other way to load the `nlp` dataset into TF dataset lazily ? --- For example, is it possible to use [Arrow dataset](https://www.tensorflow.org/io/api_docs/python/tfio/arrow/ArrowDataset) ? If yes, is there any code example ? Apparently we'll have to wait for the next tensorflow release to use `.from_generator` and TPU. See https://github.com/tensorflow/tensorflow/issues/34346#issuecomment-598262489
[ -0.1305493712425232, -0.03085951879620552, 0.036304667592048645, 0.12274178862571716, 0.22812119126319885, 0.09177538007497787, 0.27305594086647034, 0.40658968687057495, -0.11111819744110107, 0.002790682017803192, -0.08880793303251266, 0.5513768792152405, -0.15211956202983856, 0.04721105843782425, 0.31960758566856384, -0.26276513934135437, -0.06650439649820328, 0.24885842204093933, -0.12980660796165466, -0.01431279070675373, -0.04055076092481613, -0.05323105305433273, -0.0955614447593689, -0.016440225765109062, 0.01506031770259142, -0.033535413444042206, 0.05115019530057907, -0.15202926099300385, 0.4169294536113739, 0.021076835691928864, 0.35533636808395386, 0.0021881412249058485, -0.03409401699900627, 0.11226992309093475, -0.00010922730143647641, 0.3845246136188507, 0.03214544430375099, -0.149605855345726, 0.06209913268685341, -0.0740661695599556, 0.20812687277793884, -0.22736473381519318, 0.1988501250743866, -0.2661622166633606, -0.04680860787630081, -0.005251426249742508, 0.23685558140277863, -0.03428541496396065, 0.2574252188205719, 0.11936879903078079, 0.15609553456306458, 0.5252823233604431, -0.1733795702457428, 0.27862754464149475, 0.1325305998325348, -0.15955999493598938, -0.21689757704734802, -0.27933114767074585, -0.13236476480960846, -0.010954819619655609, 0.0657457709312439, 0.14796733856201172, -0.22829338908195496, 0.09629976749420166, 0.43259796500205994, 0.15464533865451813, -0.11124931275844574, -0.3790851831436157, -0.2610156536102295, 0.4085840582847595, 0.20153601467609406, -0.07298071682453156, -0.1770700216293335, -0.48519688844680786, -0.1757752150297165, -0.2934434115886688, -0.1783647984266281, 0.08686342090368271, -0.26259544491767883, 0.18190962076187134, -0.10118100792169571, -0.12041499465703964, -0.21891914308071136, 0.12010456621646881, 0.2120119035243988, -0.035575803369283676, 0.1604188233613968, 0.11686224490404129, 0.22441333532333374, 0.10827670246362686, 0.34945395588874817, 0.08066081255674362, 0.3548700213432312, 0.36756521463394165, -0.2840500473976135, -0.011490941047668457, 0.16927216947078705, -0.5806720852851868, -0.04443413019180298, -0.25259608030319214, 0.35710760951042175, 0.1853385865688324, -0.29192689061164856, 0.2165091633796692, -0.013307205401360989, -0.07356627285480499, -0.4224953353404999, 0.13194559514522552, -0.13990896940231323, -0.5050939917564392, 0.047027718275785446, 0.046306464821100235, -0.22082176804542542, -0.01992824859917164, 0.060598503798246384, -0.5112510919570923, -0.14770346879959106, 0.26907333731651306, -0.2849743664264679, -0.44184058904647827, -0.23016850650310516, 0.028392812237143517, -0.014332825317978859, 0.17374660074710846, -0.1279168277978897, 0.5364347696304321, 0.13731777667999268, -0.12483660876750946, -0.28134119510650635, -0.02957746386528015, -0.2323073148727417, 0.11308934539556503, -0.061465196311473846, -0.09602817893028259, 0.15568828582763672, -0.16067323088645935, 0.17979362607002258, -0.12268897891044617, -0.12777145206928253, 0.18162094056606293, 0.2595539391040802, -0.08195042610168457, 0.10017484426498413, 0.2579144537448883, 0.014886037446558475, -0.06456991285085678, 0.038228459656238556, 0.074913889169693, -0.21613110601902008, 0.35391765832901, -0.5225846767425537, -0.28902262449264526, -0.04879627376794815, 0.13496723771095276, 0.0877804383635521, -0.21369042992591858, -0.33316704630851746, 0.32619601488113403, 0.12096328288316727, -0.11928132176399231, -0.1852191835641861, -0.207504540681839, -0.42386922240257263, -0.30116698145866394, 0.3041066527366638, 0.040642816573381424, -0.46133407950401306, -0.11523377895355225, 0.006507989019155502, -0.01659257896244526, 0.2493128776550293, 0.44560661911964417, -0.23294784128665924, 0.4801921546459198, -0.05343633517622948, 0.0012754063354805112, 0.7598662972450256, -0.08542132377624512, -0.16154193878173828, 0.22563143074512482, -0.12333789467811584, 0.20281286537647247, -0.004924764856696129, 0.40537551045417786, 0.0003449503274168819, -0.029118746519088745, 0.38686397671699524, 0.6318569183349609, -0.15841363370418549, 0.07676459103822708, -0.10929101705551147, -0.3459326922893524, 0.30011430382728577, 0.279721200466156, 0.09528962522745132, -0.05658266693353653, -0.26506099104881287, 0.3954327702522278, 0.13062460720539093, -0.03432202339172363, -0.18207469582557678, 0.06700441241264343, -0.12869009375572205, 0.0004337819991633296, -0.1707858443260193, 0.06576748192310333, -0.6022963523864746, 0.2533167004585266, 0.19127778708934784, 0.09608181565999985, 0.18939775228500366, -0.11190406233072281, 0.36263182759284973, 0.10908019542694092, 0.11826259642839432, 0.15904958546161652, 0.019095033407211304, 0.07548099756240845, -0.028345534577965736, -0.011725767515599728, -0.3443223834037781, 0.15897108614444733, -0.375440776348114, -0.12695854902267456, -0.34896767139434814, 0.26449716091156006, 0.37324249744415283, 0.017966564744710922, -0.08456011861562729, 0.3013327419757843, -0.3664577007293701, -0.12099885940551758, 0.11764302849769592, 0.21522849798202515, -0.030569637194275856, 0.29321420192718506, -0.6672236323356628, 0.44242262840270996, 0.051598936319351196, -0.02275564894080162, -0.10921371728181839, 0.12416078895330429, 0.002570967422798276, -0.19744892418384552, 0.010840136557817459, 0.28054279088974, -0.22586120665073395, 0.14711356163024902, 0.2610127031803131, -0.24422100186347961, 0.15271635353565216, 0.06381198763847351, 0.1342630535364151, -0.11270689219236374, 0.3706357181072235, 0.09925197064876556, 0.13777044415473938, 0.2951820492744446, -0.5292934775352478, 0.2312246859073639, 0.6706401705741882, -0.02219812013208866, -0.11366341263055801, 0.21687667071819305, 0.0373346209526062, -0.21183188259601593, -0.08162252604961395, -0.0811692625284195, 0.2019302397966385, 0.1457827091217041, 0.04163608327507973, 0.0004892915021628141, -0.16413991153240204, -0.25839459896087646, 0.2476223260164261, 0.13116778433322906, 0.3808506429195404, -0.13959947228431702, 0.030044950544834137, 0.2035277783870697, -0.1915542334318161, 0.08323947340250015, 0.21064752340316772, 0.17547312378883362, -0.2056150883436203, 0.03021683543920517, -0.2745702862739563, -0.3066863715648651, -0.08983314782381058, 0.03501923382282257, 0.1440558135509491, -0.18368211388587952, -0.04992381110787392, 0.3294695317745209, -0.049427103251218796, -0.053616248071193695, -0.11888307332992554, 0.31492191553115845, 0.16494853794574738, -0.44491884112358093, -0.06582368910312653, -0.10550256073474884, -0.18507146835327148, 0.09022989869117737, 0.36076778173446655, 0.17744138836860657, 0.24297870695590973, 0.02639407478272915, -0.08753244578838348, 0.1681308001279831, -0.07798183709383011, 0.22266918420791626, -0.11164238303899765, 0.11009341478347778, -0.04960857704281807, 0.07181588560342789, -0.12877149879932404, -0.25889256596565247, -0.11553186923265457, -0.43128228187561035, 0.05385304614901543, 0.15509119629859924, -0.0024250673595815897, 0.14857907593250275, -0.24266724288463593, -0.17587347328662872, -0.3055383563041687, -0.3943732976913452, 0.018097637221217155, 0.16430525481700897, 0.3020913600921631, 0.42609965801239014, -0.029929103329777718, 0.2822215259075165, 0.38556861877441406, 0.05240771919488907, 0.0802658274769783, -0.1383150815963745, 0.4784575402736664, -0.27619507908821106, -0.25517183542251587, -0.15608875453472137, -0.17423570156097412, 0.1930476427078247, 0.3186536431312561, -0.7124728560447693, 0.14556314051151276, -0.08327411860227585, 0.08658672124147415, -0.08990921825170517, -0.041092775762081146, 0.23172467947006226, -0.20331734418869019, 0.020510729402303696, 0.004608663264662027, 0.16507427394390106, 0.14517663419246674, 0.024806609377264977, -0.20415571331977844, 0.37039095163345337, 0.42371463775634766, 0.33127185702323914, 0.7575734853744507, -0.08429126441478729, -0.2142927199602127, 0.23427128791809082, -0.18224920332431793, 0.19702790677547455, -0.2328154444694519, -0.06427564471960068, 0.04293925687670708, -0.013692923821508884, -0.005353451706469059, 0.22368411719799042, -0.100008025765419, -0.19537709653377533, -0.11733975261449814, 0.2207116186618805, 0.11836251616477966, -0.23836800456047058, 0.25822997093200684, -0.21444033086299896, 0.18313220143318176, 0.11112041026353836, 0.21261031925678253, -0.3299855589866638, -0.3997199833393097, -0.01931050606071949, -0.23272506892681122, 0.41783609986305237, -0.04035969451069832, -0.32866764068603516, 0.10396529734134674, -0.7131767868995667, 0.06532419472932816, 0.11960644274950027, 0.30858200788497925, 0.08994624763727188, -0.29901519417762756, -0.006997276097536087, 0.06775613129138947, 0.54835045337677, 0.09833641350269318, -0.08745607733726501, 0.20974920690059662, -0.1773279905319214, -0.4892937242984772, 0.03407872095704079, -0.05583788454532623, -0.07893317937850952, 0.24943692982196808, 0.1401747614145279, -0.26192647218704224, -0.491312175989151, -0.02455650456249714, -0.0164188202470541, -0.33952680230140686, -0.04881987348198891, -0.18680141866207123, -0.24344752728939056, 0.001078475615940988, -0.31565260887145996, 0.05761346220970154, 0.27913495898246765, -0.1710856556892395, 0.020696090534329414, -0.15080328285694122, -0.0014764488441869617, 0.21223410964012146, 0.0035507420543581247, 0.1631869524717331, 0.4632871150970459, 0.047394562512636185, -0.048913486301898956, -0.008088097907602787, -0.1308065801858902, 0.224055215716362, -0.09222051501274109, -0.1522686779499054, 0.018758229911327362, 0.15853554010391235, 0.37199580669403076, 0.04845492169260979, 0.03609539568424225, 0.1285245418548584, -0.17415070533752441, 0.03312743827700615, -0.19866769015789032, 0.3670531213283539, 0.08676262944936752, -0.3417379558086395, -0.08246522396802902, -0.510940670967102, 0.5615842938423157, 0.2959672212600708, 0.1288377195596695, 0.4830288290977478, -0.1970454603433609, -0.36169517040252686, 0.5212911367416382, 0.38263049721717834, 0.5099004507064819, -0.37036868929862976, 0.5261974930763245, 0.72294020652771, 0.47835981845855713, 0.9239939451217651, -0.3590678572654724, 0.27109068632125854, -0.12388866394758224, -0.35866406559944153, -0.028797559440135956, -0.255656361579895, -0.3750627040863037, 0.07888790965080261, 0.10311441868543625, 0.2167326658964157, 0.13108684122562408, 0.27752771973609924, 0.217911034822464, 0.3121577799320221, 0.3466190695762634, -0.42222920060157776, -0.19411925971508026, 0.06759988516569138, 0.02967003546655178, -0.2340952306985855, -0.08185143023729324, -0.19640444219112396, 0.2307492196559906, 0.06485644727945328, 0.020356886088848114, 0.12006030976772308, -0.18482451140880585, 0.28192639350891113, 0.13958978652954102, -0.09903112053871155, 0.04047853499650955, 0.24172726273536682, 0.12007810175418854, -0.20490717887878418, -0.10917849838733673, -0.0896356850862503, 0.06531897932291031, 0.010170694440603256, 0.1468190997838974, -0.0447651632130146, 0.6814290285110474, -0.021972330287098885, 0.19641423225402832, 0.16399019956588745, -0.07861150801181793, -0.5933405756950378, -0.18238632380962372, -0.11446291953325272, 0.08533591777086258, 0.1685636043548584, -0.5447575449943542, 0.2875628173351288, -0.19816631078720093, 0.18102902173995972, 0.09797942638397217, 0.11959993094205856, -0.20460988581180573, 0.05496605113148689, 0.05523880943655968, -0.18702711164951324, 0.22112669050693512, 0.02921549417078495, 0.12724588811397552, 0.03463367000222206, 0.35292479395866394, 0.22735242545604706, 0.004736479837447405, -0.025911424309015274, 0.10443626344203949, 0.28496333956718445, -0.04618600010871887, 0.06443624198436737, 0.036644816398620605, 0.22891248762607574, 0.0641891360282898, 0.3598417341709137, -0.22694441676139832, 0.031078271567821503, -0.2919076681137085, -0.17317961156368256, -0.30188632011413574, 0.3662432134151459, -0.059796568006277084, 0.12840881943702698, -0.003767472691833973, 0.21896663308143616, -0.1237194612622261, 0.22889849543571472, -0.2416870892047882, -0.02274288609623909, 0.00589331891387701, 0.06760502606630325, -0.12601563334465027, -0.058639928698539734, 0.0696457251906395, 0.13089972734451294, 0.06438548862934113, 0.14221805334091187, -0.3747170567512512, -0.20188719034194946, -0.04005352035164833, 0.16559559106826782, 0.03266608342528343, -0.2386946827173233, -0.004422334022819996, -0.3472805321216583, -0.33256152272224426, -0.16482409834861755, 0.11657588183879852, -0.12507429718971252, -0.2715754210948944, 0.06031174212694168, -0.059460997581481934, 0.2978954613208771, -0.13933652639389038, 0.03453801944851875, -0.10700326412916183, 0.3132719099521637, -0.08346689492464066, 0.2498142272233963, 0.0991898700594902, -0.050093457102775574, -0.31366607546806335, -0.11679309606552124, 0.1376473307609558, -0.1159994900226593, 0.06822949647903442, -0.3133595585823059, -0.07454583793878555, -0.2488778531551361, 0.042873311787843704, 0.2715884745121002, -0.2371542900800705, 0.09041208773851395, 0.24579152464866638, 0.11498232185840607, -0.09760551899671555, -0.15072134137153625, 0.20687147974967957, 0.08234630525112152, -0.024316847324371338, 0.20025362074375153, 0.08809409290552139, 0.16034257411956787, -0.06396044045686722, -0.04028838500380516, 0.010186923667788506, -0.5695465803146362, -0.19386830925941467, 0.40815216302871704, -0.1363222450017929, -0.014250025153160095, -0.16127115488052368, 0.0954628437757492, 0.16480153799057007, 0.21101626753807068, 0.2794184684753418, 0.2288534790277481, -0.07180015742778778, -0.09621987491846085, -0.061399538069963455, -0.3514275848865509, -0.4288904368877411, 0.44052088260650635, -0.06125330924987793, 0.5847460031509399, -0.22746171057224274, 0.035516392439603806, 0.1482977718114853, -0.21973031759262085, -0.16932731866836548, 0.2957041263580322, -0.1487514227628708, -0.06481285393238068, -0.23912519216537476, 0.06481707096099854, -0.1506691426038742, 0.2991456091403961, -0.10230962187051773, -0.2143225222826004, -0.16765934228897095, 0.2694041430950165, 0.09658190608024597, 0.2263260930776596, 0.11800194531679153, 0.07471519708633423, 0.24854080379009247, -0.1172582283616066, 0.19789005815982819, -0.2634515166282654, -0.1078861802816391, -0.3696131408214569, 0.011613166891038418, 0.0912148728966713, 0.2826961874961853, -0.3337540924549103, 0.12661010026931763, 0.07012628018856049, -0.04584909602999687, -0.020807096734642982, 0.24407058954238892, 0.2368534654378891, 0.09861940145492554, 0.304680734872818, -0.0035822356585413218, -0.05681219696998596, 0.15154998004436493, 0.15705399215221405, 0.17538706958293915, -0.578890860080719, 0.5717278718948364, 0.2261795848608017, -0.055576253682374954, -0.0059855785220861435, 0.2032666802406311, -0.08927438408136368, -0.1207638531923294, 0.5214146971702576, -0.2620497941970825, 0.03671780601143837, -0.0025084514636546373, 0.046714525669813156, -0.19307412207126617, 0.6902197003364563, 0.057421427220106125, 0.45538556575775146, -0.08694536983966827, 0.04564293473958969, -0.44194990396499634, 0.09158571064472198, -0.10719571262598038, 0.21197332441806793, 0.26062825322151184, 0.29831385612487793, 0.14649181067943573, 0.2051636278629303, -0.4557717442512512, -0.21373851597309113, -0.01620802842080593, 0.048399537801742554, -0.14886347949504852, -0.10379461199045181, -0.5044934153556824, 0.12454109638929367, -0.13039030134677887, -0.4423992335796356, 0.0877692773938179, -0.4213770031929016, -0.03412565588951111, -0.0869128480553627, 0.22290639579296112, 0.3233662545681, 0.4466140568256378, 0.3336487114429474, -0.03963187336921692, -0.10492025315761566, 0.04959724843502045, -0.07954354584217072, -0.11297428607940674, -0.027450961992144585, -0.3176702857017517, -0.07864496111869812, -0.14929966628551483, 0.12864720821380615, -0.2904021739959717, -0.0474701002240181, -0.2210419923067093, -0.2627435028553009, 0.02521299012005329, -0.10299514979124069, -0.13584543764591217, 0.09924563020467758, -0.2605098485946655, 0.17538076639175415, 0.23030783236026764, 0.44776806235313416, -0.14069710671901703, -0.12991295754909515, -0.21534103155136108, 0.08180026710033417, 0.38260072469711304, -0.6043995022773743, -0.2027837485074997, 0.06831548362970352, -0.16580991446971893, 0.2554984390735626, 0.06379826366901398, -0.2777387797832489, -0.21997898817062378, 0.2798735797405243, -0.12841488420963287, -0.21454496681690216, 0.25627371668815613, 0.32339802384376526, -0.10076190531253815, -0.12213389575481415, 0.2780945599079132, -0.06114077940583229, -0.4330613613128662, -0.2135881632566452, -0.4865180552005768 ]
https://github.com/huggingface/datasets/issues/192
[Question] Create Apache Arrow dataset from raw text file
We store every dataset in the Arrow format. This is convenient as it supports nested types and memory mapping. If you are curious feel free to check the [pyarrow documentation](https://arrow.apache.org/docs/python/) You can use this library to load your covid papers by creating a dataset script. You can find inspiration from the ones we've already written in `/datasets`. Here is a link to the steps to [add a dataset](https://github.com/huggingface/nlp/blob/master/CONTRIBUTING.md#how-to-add-a-dataset)
Hi guys, I have gathered and preprocessed about 2GB of COVID papers from CORD dataset @ Kggle. I have seen you have a text dataset as "Crime and punishment" in Apache arrow format. Do you have any script to do it from a raw txt file (preprocessed as for BERT like) or any guide? Is the worth of send it to you and add it to the NLP library? Thanks, Manu
68
[Question] Create Apache Arrow dataset from raw text file Hi guys, I have gathered and preprocessed about 2GB of COVID papers from CORD dataset @ Kggle. I have seen you have a text dataset as "Crime and punishment" in Apache arrow format. Do you have any script to do it from a raw txt file (preprocessed as for BERT like) or any guide? Is the worth of send it to you and add it to the NLP library? Thanks, Manu We store every dataset in the Arrow format. This is convenient as it supports nested types and memory mapping. If you are curious feel free to check the [pyarrow documentation](https://arrow.apache.org/docs/python/) You can use this library to load your covid papers by creating a dataset script. You can find inspiration from the ones we've already written in `/datasets`. Here is a link to the steps to [add a dataset](https://github.com/huggingface/nlp/blob/master/CONTRIBUTING.md#how-to-add-a-dataset)
[ -0.17388372123241425, -0.06410497426986694, -0.04578731954097748, 0.06958875805139542, -0.3965008854866028, 0.13637591898441315, 0.06761064380407333, 0.4003838896751404, -0.06824392825365067, -0.32224202156066895, 0.13134367763996124, 0.18338918685913086, -0.08250487595796585, 0.07896027714014053, 0.39143106341362, -0.04079053923487663, -0.058886345475912094, 0.28287461400032043, -0.032862234860658646, -0.03655685484409332, 0.16092778742313385, -0.021286694332957268, -0.1044444814324379, 0.007256089709699154, -0.19171859323978424, -0.34966641664505005, -0.16615353524684906, 0.15690922737121582, -0.14719252288341522, -0.4131219685077667, 0.1118624210357666, 0.08591306954622269, 0.5545088648796082, 0.152288019657135, -0.00012087303912267089, -0.4648415148258209, 0.14617322385311127, -0.07211573421955109, -0.1525297611951828, -0.20889556407928467, -0.23260009288787842, -0.09182356297969818, 0.1721591055393219, -0.33904919028282166, 0.07731235772371292, -0.4477840065956116, 0.17121949791908264, -0.2712027430534363, 0.665780782699585, 0.3493712544441223, 0.07335954159498215, -0.10189543664455414, 0.1893395036458969, 0.22857488691806793, 0.4799532890319824, 0.5429453253746033, -0.3052918314933777, 0.15295414626598358, 0.2607787847518921, -0.20762257277965546, -0.2825189530849457, 0.1587996929883957, -0.09765501320362091, -0.23909755051136017, 0.33057841658592224, 0.11803030967712402, -0.27039942145347595, -0.3330770432949066, 0.04893825203180313, 0.3964376747608185, 0.36962565779685974, -0.4938063323497772, 0.048538919538259506, -0.05016384273767471, 0.18940627574920654, -0.48908141255378723, -0.14875246584415436, 0.6180793046951294, -0.5020700097084045, 0.35973915457725525, 0.33444252610206604, -0.38164618611335754, -0.43398362398147583, 0.12841829657554626, 0.4575880765914917, -0.07157063484191895, -0.047139547765254974, -0.07582789659500122, -0.05029016360640526, 0.1744941622018814, 0.21123680472373962, -0.24556170403957367, -0.14875802397727966, 0.34700530767440796, -0.1437055766582489, -0.16275876760482788, -0.4877946376800537, -0.20781882107257843, 0.2758220434188843, -0.02975361980497837, 0.3700712025165558, -0.13018089532852173, -0.03909542039036751, -0.09167711436748505, 0.03946316987276077, -0.09501903504133224, 0.45105770230293274, 0.2062091827392578, -0.00900284294039011, -0.2627827227115631, 0.16223561763763428, -0.08094597607851028, -0.20558711886405945, 0.0966670960187912, -0.15640978515148163, 0.03321874514222145, 0.049624133855104446, -0.19279055297374725, 0.01111215353012085, 0.006018821150064468, -0.43664783239364624, -0.024569638073444366, -0.2640625834465027, 0.23352771997451782, 0.18255937099456787, -0.09135601669549942, 0.22798126935958862, 0.39805835485458374, 0.002583087421953678, -0.191006600856781, 0.1229754164814949, 0.07351323217153549, -0.2823691666126251, 0.28089281916618347, 0.24723444879055023, 0.048766616731882095, -0.0020072264596819878, -0.09525113552808762, 0.047782428562641144, -0.30196231603622437, 0.1904061734676361, -0.011525362730026245, 0.16694170236587524, -0.11928243935108185, 0.26791253685951233, 0.04077565670013428, -0.14994056522846222, -0.12755240499973297, -0.10135633498430252, 0.3692101538181305, -0.3687524199485779, -0.12114272266626358, -0.13666395843029022, -0.0014924067072570324, 0.05503535643219948, -0.14150689542293549, -0.27779650688171387, 0.25129222869873047, 0.13343480229377747, 0.1928407847881317, 0.055643945932388306, 0.2801021337509155, -0.013973555527627468, -0.2512497305870056, 0.0342513732612133, 0.4760323464870453, -0.23271925747394562, 0.1828199028968811, -0.1303553581237793, 0.42202699184417725, 0.04189235344529152, 0.19998104870319366, -0.25090524554252625, 0.33731165528297424, 0.028533168137073517, 0.3784562945365906, 0.7308465242385864, -0.07829371094703674, 0.2264496237039566, 0.4033631682395935, 0.18155382573604584, -0.3522210419178009, 0.14907316863536835, -0.043225377798080444, 0.10294368118047714, -0.09913045167922974, -0.30004048347473145, 0.49250900745391846, 0.11881684511899948, 0.03809281438589096, -0.07898488640785217, -0.2024592012166977, -0.18061652779579163, 0.013828638941049576, -0.14248567819595337, -0.15212230384349823, 0.09050896763801575, -0.45889052748680115, 0.4787384569644928, -0.2984764873981476, 0.4021228551864624, 0.29039740562438965, 0.04152541980147362, -0.014198892749845982, 0.24639356136322021, -0.11176865547895432, -0.18240483105182648, -0.31211790442466736, -0.4065294861793518, 0.2906285226345062, -0.25427594780921936, -0.2743140459060669, -0.25395286083221436, -0.11944711208343506, 0.07688603550195694, -0.03059508092701435, -0.03145672753453255, -0.24361303448677063, -0.2193363606929779, 0.01776365377008915, -0.20738880336284637, 0.2594849765300751, -0.27977171540260315, 0.13498534262180328, -0.2302948385477066, 0.21568022668361664, 0.08199535310268402, -0.19584135711193085, 0.24660922586917877, 0.2380744367837906, -0.27999481558799744, 0.1489575207233429, 0.06565996259450912, 0.1418323814868927, -0.053554803133010864, 0.20256252586841583, 0.427729070186615, -0.28081583976745605, 0.0847725197672844, -0.6659767627716064, 0.30371013283729553, -0.04134857654571533, 0.2257370501756668, -0.11210645735263824, -0.3082045614719391, 0.44492846727371216, -0.062422286719083786, 0.1924302875995636, -0.1610330492258072, 0.018284078687429428, -0.07412037998437881, -0.030854322016239166, -0.08382134884595871, 0.1798776388168335, 0.055171966552734375, 0.43464675545692444, 0.15664012730121613, -0.151343435049057, 0.027362480759620667, 0.08514956384897232, -0.12751811742782593, -0.1711125373840332, 0.2346556931734085, 0.23757675290107727, -0.3029528260231018, -0.1035984680056572, -0.16728371381759644, 0.04839351400732994, 0.0925913155078888, 0.1848362535238266, -0.15063898265361786, -0.040357958525419235, 0.08050628006458282, -0.09687495976686478, 0.3708857595920563, -0.03018072247505188, 0.2931440770626068, -0.1610877811908722, -0.0907881110906601, -0.023355970159173012, -0.17951416969299316, -0.2806982696056366, -0.04755263775587082, 0.13846668601036072, -0.2195432484149933, 0.028408704325556755, -0.06600562483072281, -0.37388572096824646, -0.33117735385894775, 0.1296108067035675, -0.041425589472055435, 0.252734512090683, 0.012834295630455017, -0.24280226230621338, 0.160663440823555, 0.0032890066504478455, -0.015388344414532185, 0.5403866767883301, 0.09072723984718323, 0.07187308371067047, -0.18865971267223358, -0.5223942995071411, -0.18206334114074707, 0.056400299072265625, 0.2462330460548401, 0.4204172194004059, 0.36549460887908936, -0.18795715272426605, 0.1201983392238617, -0.037211425602436066, -0.015586599707603455, 0.039233818650245667, -0.1758021116256714, -0.27563250064849854, -0.18567432463169098, 0.12907709181308746, -0.5696606636047363, -0.10519001632928848, -0.21616171300411224, 0.04164513200521469, -0.1757652461528778, -0.2118133306503296, -0.07265700399875641, -0.040174148976802826, 0.1962091475725174, -0.5246597528457642, -0.15463583171367645, -0.12683838605880737, 0.3574325442314148, 0.30721578001976013, 0.2926773130893707, -0.024483775720000267, 0.2397528737783432, 0.11662006378173828, -0.3075912296772003, -0.05353710427880287, 0.19105178117752075, -0.06540194153785706, 0.37870243191719055, -0.3172716200351715, -0.5370227694511414, 0.011981671676039696, -0.05331210419535637, -0.07275471836328506, 0.3109486401081085, -0.35481566190719604, 0.419631689786911, 0.08230754733085632, 0.029532812535762787, -0.1032712459564209, -0.012858808040618896, 0.24516241252422333, -0.00042280126945115626, 0.06135645508766174, -0.058393631130456924, 0.09048828482627869, 0.35686323046684265, -0.12842397391796112, 0.3276354968547821, 0.3448914587497711, 0.14565689861774445, -0.046614598482847214, 0.5362746715545654, 0.25174808502197266, 0.060705795884132385, 0.17636597156524658, -0.11871405690908432, 0.07755386829376221, 0.10546091943979263, 0.03187001496553421, -0.07428260892629623, 0.12698692083358765, 0.037481699138879776, 0.18022391200065613, 0.3182491362094879, -0.32272568345069885, -0.20063048601150513, -0.24340853095054626, -0.3722503185272217, -0.1931561380624771, 0.3950960636138916, -0.2958455979824066, 0.2528689503669739, -0.11756224185228348, -0.2101004272699356, -0.1779503971338272, -0.47201043367385864, -0.10274569690227509, 0.2386797070503235, 0.26414191722869873, 0.1729045957326889, -0.04200572147965431, 0.36136168241500854, -0.6206261515617371, 0.29943105578422546, 0.18296577036380768, -0.11638835817575455, 0.114884153008461, 0.05378718674182892, 0.2133236676454544, 0.08678250014781952, -0.11692526936531067, 0.16836968064308167, -0.05905286595225334, -0.014313729479908943, 0.3365096151828766, -0.36891597509384155, 0.3253723680973053, -0.2826526165008545, 0.037240754812955856, 0.22013679146766663, -0.11274156719446182, -0.43968939781188965, -0.22016754746437073, 0.17403528094291687, 0.14908041059970856, -0.1108863353729248, -0.4020349979400635, 0.1299976110458374, 0.005802634637802839, -0.39609378576278687, -0.1459903120994568, 0.39685726165771484, -0.01603306271135807, -0.12497640401124954, -0.15539143979549408, -0.01966378651559353, 0.10079880058765411, 0.22278249263763428, -0.2305995225906372, 0.20117710530757904, -0.25087565183639526, 0.021771589294075966, -0.08185677975416183, 0.3278175890445709, 0.0673634484410286, -0.001510963193140924, 0.18697263300418854, -0.22211602330207825, -0.06431376934051514, -0.39945700764656067, 0.5800673961639404, 0.17286989092826843, 0.11208894103765488, 0.15619568526744843, 0.18525147438049316, -0.30442047119140625, -0.11992888152599335, -0.011087847873568535, 0.1350955218076706, -0.11089466512203217, -0.4100244343280792, -0.3892834782600403, 0.5749381184577942, 0.04820896312594414, -0.03414613753557205, 0.29530972242355347, 0.0565154142677784, -0.29029664397239685, 0.6706215143203735, 0.20020198822021484, 1.225578784942627, -0.17598895728588104, 0.2871076464653015, 0.054876189678907394, -0.10766847431659698, 0.5923517346382141, -0.42020517587661743, -0.21309834718704224, -0.5059723854064941, 0.3042323887348175, -0.11025373637676239, -0.14544348418712616, 0.1546524614095688, 0.33180344104766846, -0.06618154048919678, 0.10463544726371765, 0.275322288274765, -0.045740872621536255, 0.1352182924747467, 0.32488906383514404, -0.21979494392871857, -0.4144166111946106, -0.15021666884422302, -0.09407120198011398, -0.12854498624801636, 0.046928562223911285, 0.11305557936429977, -0.41741541028022766, -0.32015901803970337, 0.0748211070895195, -0.4243670701980591, 0.10035252571105957, -0.008133945055305958, -0.22221967577934265, -0.2661270499229431, -0.6077929735183716, 0.3085106611251831, 0.3429754972457886, 0.09421619772911072, -0.007780144456773996, -0.28515559434890747, 0.09172771871089935, -0.2546536922454834, -0.39086252450942993, 0.27973127365112305, -0.12988542020320892, 0.07501446455717087, -0.1504436433315277, -0.20562122762203217, 0.15109066665172577, 0.05339076742529869, -0.28714779019355774, -0.006581712979823351, -0.2250330001115799, 0.49902966618537903, -0.1214831992983818, -0.060345590114593506, 0.32706519961357117, -0.41001513600349426, 0.005482974927872419, 0.0734868273139, 0.3840419054031372, 0.10894736647605896, 0.052778210490942, 0.2629636228084564, -0.25314009189605713, -0.045207079499959946, 0.17500554025173187, 0.36659032106399536, -0.2257859706878662, 0.5001032948493958, 0.4196084439754486, -0.15374985337257385, -0.29532983899116516, 0.27615246176719666, 0.3888963758945465, -0.8429340720176697, -0.10552550107240677, -0.008340717293322086, 0.0986776351928711, 0.01622936688363552, 0.10067235678434372, -0.07750066369771957, -0.04655391350388527, 0.010311947204172611, -0.542080819606781, -0.26967620849609375, 0.1499103307723999, -0.09666280448436737, 0.07544462382793427, 0.39812320470809937, 0.09652518481016159, 0.041537296026945114, 0.13556693494319916, -0.20735123753547668, 0.08010676503181458, 0.5916287302970886, 0.1752157360315323, 0.10622408986091614, 0.09708879142999649, 0.019882533699274063, 0.02555638551712036, -0.024785324931144714, 0.11710461229085922, 0.11115162819623947, -0.13142049312591553, -0.015419227071106434, 0.20284496247768402, -0.06086160987615585, -0.2965806722640991, -0.30982351303100586, -0.45137426257133484, -0.2555304169654846, -0.19112586975097656, -0.14729641377925873, -0.15576721727848053, -0.1328599452972412, 0.13090796768665314, 0.07797173410654068, 0.014391588978469372, -0.4827415347099304, 0.2256816029548645, -0.21510623395442963, 0.251577764749527, -0.09725794941186905, 0.1495196521282196, -0.0002766309480648488, -0.050234105437994, 0.04456501081585884, 0.1290828138589859, 0.24697421491146088, 0.1895880252122879, 0.5016093254089355, -0.1193627417087555, -0.04702061414718628, -0.3728879392147064, 0.20447029173374176, 0.2573818862438202, 0.09038421511650085, 0.06380901485681534, 0.2950476109981537, 0.0918235182762146, -0.22265124320983887, 0.2888725996017456, -0.23050346970558167, -0.2013942450284958, -0.18630293011665344, 0.12676480412483215, -0.02107030153274536, -0.19755667448043823, -0.31091293692588806, -0.1603405624628067, 0.18026599287986755, -0.015163905918598175, -0.35536912083625793, 0.19000428915023804, 0.22559216618537903, -0.04182180017232895, -0.13116112351417542, 0.22774869203567505, 0.25549593567848206, -0.06322725117206573, -0.05395884066820145, 0.15032382309436798, 0.729856014251709, 0.26674455404281616, 0.14226235449314117, -0.16784678399562836, 0.041503921151161194, 0.43847525119781494, 0.24947483837604523, 0.014321703463792801, 0.23090074956417084, 0.5041621327400208, 0.6058552861213684, 0.10812108963727951, -0.04871401935815811, 0.2923034727573395, 0.1237492561340332, 0.03654143959283829, -0.459958553314209, -0.06597098708152771, -0.04848242923617363, -0.27365216612815857, -0.19247587025165558, -0.07600425183773041, -0.14734037220478058, 0.370712548494339, 0.15805166959762573, -0.0557744987308979, 0.16497446596622467, 0.07142505049705505, 0.3248579204082489, -0.24928683042526245, 0.08210790902376175, -0.12004118412733078, 0.015363607555627823, 0.037474095821380615, 0.43583187460899353, 0.33153751492500305, 0.1721891462802887, -0.13928714394569397, 0.14918875694274902, -0.00022725500457454473, -0.10717980563640594, 0.1763482391834259, 0.19603390991687775, 0.40060141682624817, 0.03843921050429344, 0.0731000006198883, 0.045242708176374435, 0.027987927198410034, -0.12604181468486786, 0.3155635595321655, -0.21342460811138153, 0.09988753497600555, -0.15060222148895264, 0.07479587942361832, -0.37400439381599426, -0.16975484788417816, 0.09984654933214188, -0.16131100058555603, 0.3633498549461365, 0.6061457991600037, -0.23281364142894745, -0.04642970487475395, -0.11821340769529343, 0.01120659802109003, -0.103349968791008, 0.4182170033454895, 0.3434804379940033, 0.127003014087677, -0.2906666100025177, -0.2813270390033722, -0.04703807458281517, 0.10857357829809189, -0.03955502435564995, -0.23724158108234406, -0.05329851061105728, 0.03458274155855179, 0.26135826110839844, 0.4125305712223053, 0.04884747788310051, 0.1342536062002182, -0.11785823851823807, 0.4718044400215149, -0.28801408410072327, -0.04197244346141815, 0.2731885313987732, 0.5164780616760254, -0.16281382739543915, -0.10243458300828934, 0.15228983759880066, 0.3586501479148865, -0.06448942422866821, -0.23673926293849945, -0.08922397345304489, 0.3275837004184723, -0.17799562215805054, 0.22377057373523712, -0.09077372401952744, 0.15293243527412415, -0.051674678921699524, -0.12333441525697708, 0.0046204156242311, -0.08744850754737854, 0.10571614652872086, 0.44399377703666687, 0.32237687706947327, -0.19108979403972626, -0.26522549986839294, -0.001539864344522357, 0.0471138060092926, 0.2749514579772949, -0.28760018944740295, -0.050847843289375305, -0.27267855405807495, -0.02337668091058731, -0.2617737650871277, 0.14743797481060028, -0.034875594079494476, 0.28765204548835754, -0.21054619550704956, 0.11583275347948074, -0.12690186500549316, -0.13028763234615326, 0.09160202741622925, -0.2509089410305023, 0.03996608033776283, 0.26852115988731384, 0.21695423126220703, 0.5681503415107727, 0.3674049973487854, -0.5660690665245056, 0.21065376698970795, 0.23423723876476288, 0.0702546015381813, 0.1036049947142601, 0.06293860822916031, 0.0025858976878225803, 0.10403262078762054, -0.01696898601949215, 0.39650610089302063, -0.058482538908720016, -0.03820488601922989, -0.4614564776420593, -0.49836644530296326 ]
https://github.com/huggingface/datasets/issues/192
[Question] Create Apache Arrow dataset from raw text file
Hello @mrm8488 and @lhoestq Is there a way to convert a dataset to Apache arrow format (locally/personal use) & use it before sending it to hugging face? Thanks :)
Hi guys, I have gathered and preprocessed about 2GB of COVID papers from CORD dataset @ Kggle. I have seen you have a text dataset as "Crime and punishment" in Apache arrow format. Do you have any script to do it from a raw txt file (preprocessed as for BERT like) or any guide? Is the worth of send it to you and add it to the NLP library? Thanks, Manu
29
[Question] Create Apache Arrow dataset from raw text file Hi guys, I have gathered and preprocessed about 2GB of COVID papers from CORD dataset @ Kggle. I have seen you have a text dataset as "Crime and punishment" in Apache arrow format. Do you have any script to do it from a raw txt file (preprocessed as for BERT like) or any guide? Is the worth of send it to you and add it to the NLP library? Thanks, Manu Hello @mrm8488 and @lhoestq Is there a way to convert a dataset to Apache arrow format (locally/personal use) & use it before sending it to hugging face? Thanks :)
[ -0.10410269349813461, -0.08714812994003296, -0.09965071082115173, 0.1544029861688614, -0.3756238520145416, 0.1595185101032257, 0.1173253133893013, 0.4278322160243988, -0.043595850467681885, -0.2937571108341217, 0.07305129617452621, 0.11510495096445084, -0.08337912708520889, 0.14750340580940247, 0.2515306770801544, -0.153853178024292, -0.08801751583814621, 0.21654418110847473, -0.29338303208351135, 0.015711648389697075, 0.27406200766563416, -0.02474883757531643, -0.0877850204706192, 0.012239886447787285, -0.17144334316253662, -0.23625536262989044, -0.12816721200942993, -0.003314283676445484, -0.14507755637168884, -0.33768293261528015, -0.025384997949004173, -0.029353545978665352, 0.5116749405860901, 0.20511527359485626, -0.00012170406262157485, -0.34656041860580444, 0.1274176388978958, -0.0041155628859996796, -0.24111872911453247, -0.20824940502643585, -0.15198726952075958, -0.009041264653205872, 0.19447478652000427, -0.3356015384197235, -0.060021352022886276, -0.3797876238822937, 0.030286962166428566, -0.46885746717453003, 0.839243471622467, 0.3928319811820984, 0.0757717564702034, -0.1602264940738678, 0.18871402740478516, 0.21793437004089355, 0.34135064482688904, 0.5511060357093811, -0.3402373194694519, -0.07742201536893845, 0.14995832741260529, 0.03337464854121208, -0.16508620977401733, 0.19850091636180878, 0.043988876044750214, -0.15324650704860687, 0.35406506061553955, -0.03162282332777977, -0.2806333005428314, -0.22104321420192719, 0.12315788120031357, 0.32636189460754395, 0.4954083561897278, -0.4750736355781555, -0.04503782466053963, -0.01197313517332077, 0.20504654943943024, -0.4249294400215149, -0.2257205843925476, 0.5759357213973999, -0.45890548825263977, 0.4854777455329895, 0.14960326254367828, -0.40089502930641174, -0.4561849534511566, -0.07455494999885559, 0.3925384283065796, -0.11834438145160675, -0.10440382361412048, -0.06381271034479141, -0.142126202583313, 0.16296027600765228, 0.22700193524360657, -0.3375427722930908, -0.1277935802936554, 0.26903942227363586, -0.042774733155965805, -0.21517300605773926, -0.5973441004753113, -0.19733555614948273, 0.18537579476833344, 0.07733018696308136, 0.38570892810821533, -0.11290831118822098, -0.04760950803756714, -0.2783755362033844, 0.07719910144805908, -0.07277897000312805, 0.4339829981327057, 0.13293208181858063, -0.007071338593959808, -0.2632567584514618, 0.17075762152671814, -0.11871568113565445, -0.09468048065900803, 0.017399169504642487, -0.1367945671081543, 0.20157092809677124, -0.010499740950763226, -0.2780987024307251, -0.033356498926877975, -0.019908951595425606, -0.3864953815937042, -0.07699979841709137, -0.21380358934402466, 0.026128295809030533, 0.20136502385139465, -0.1467558890581131, 0.22471104562282562, 0.3776320815086365, 0.052054181694984436, -0.28021395206451416, 0.17620933055877686, -0.018022168427705765, -0.2320820689201355, 0.28165552020072937, 0.08700871467590332, 0.16538067162036896, -0.1353076845407486, -0.06850408017635345, -0.05829614773392677, -0.1452392339706421, 0.25249114632606506, -0.036169957369565964, 0.21758878231048584, -0.06669128686189651, 0.20696470141410828, 0.016298165544867516, -0.12016603350639343, -0.007247621659189463, -0.07581128180027008, 0.3549622595310211, -0.18190038204193115, 0.03534134849905968, -0.044504985213279724, 0.015734555199742317, 0.07062359154224396, -0.08456835150718689, -0.2357393056154251, 0.16044282913208008, -0.0032511884346604347, 0.33366283774375916, 0.0702999159693718, 0.2868444621562958, 0.060523632913827896, -0.14232556521892548, -0.061350222676992416, 0.4451697766780853, -0.3385568857192993, 0.14817193150520325, -0.09775742143392563, 0.20304572582244873, -0.11506365984678268, 0.3096964359283447, -0.22147971391677856, 0.28552648425102234, -0.02173238806426525, 0.248059943318367, 0.760906457901001, -0.00319566298276186, 0.19536173343658447, 0.4988290071487427, 0.03261254355311394, -0.3793383240699768, 0.09234829246997833, -0.08416328579187393, -0.08584754914045334, -0.029711242765188217, -0.28686562180519104, 0.5281491875648499, 0.11185703426599503, -0.016950568184256554, -0.11940508335828781, -0.17205417156219482, -0.11659643054008484, -0.05393878370523453, -0.23231519758701324, -0.05482989177107811, 0.09218209981918335, -0.48238301277160645, 0.4895268976688385, -0.33566978573799133, 0.4532501697540283, 0.237852081656456, 0.02512098103761673, -0.039329126477241516, 0.1749616116285324, -0.031960178166627884, -0.09866148233413696, -0.3560512661933899, -0.3879682719707489, 0.21412938833236694, -0.326967716217041, -0.23821257054805756, -0.26968297362327576, -0.125290185213089, -0.06862170249223709, 0.08229734003543854, -0.01979629136621952, -0.23527994751930237, -0.15579476952552795, 0.014363083057105541, -0.21273447573184967, 0.31802859902381897, -0.11018742620944977, 0.1491357535123825, -0.2236449420452118, 0.11636775732040405, 0.1057080551981926, -0.00439384113997221, 0.237599179148674, 0.13671815395355225, -0.23089133203029633, 0.05242209509015083, -0.05458331108093262, 0.0032962250988930464, -0.07111271470785141, 0.17900675535202026, 0.5197834968566895, -0.29207944869995117, 0.07352687418460846, -0.6826449632644653, 0.31045371294021606, 0.0175312552601099, 0.21841339766979218, -0.10102833807468414, -0.46540260314941406, 0.48078009486198425, -0.0809444710612297, 0.19033698737621307, -0.1441081315279007, 0.0432979092001915, -0.13720974326133728, -0.11262873560190201, -0.16778384149074554, 0.14391833543777466, -0.06783543527126312, 0.3569740355014801, 0.21481426060199738, -0.09206991642713547, 0.09972012042999268, 0.031152866780757904, -0.1916440725326538, -0.19726866483688354, 0.23297186195850372, 0.21440370380878448, -0.23323801159858704, -0.07656899839639664, -0.06118118762969971, 0.04091804474592209, 0.11759529262781143, 0.1491021066904068, 0.021691640838980675, 0.0765998512506485, 0.2202986627817154, -0.13715334236621857, 0.33411863446235657, -0.1321941465139389, 0.3828594982624054, -0.1595306247472763, -0.16526341438293457, 0.021198788657784462, -0.0997905433177948, -0.1932457536458969, -0.10108207911252975, 0.14014816284179688, -0.08165836334228516, -0.03822110965847969, -0.1377265453338623, -0.40400806069374084, -0.3725431561470032, 0.26179975271224976, -0.03137892484664917, 0.253950834274292, -0.07426447421312332, -0.4246182143688202, 0.1473333239555359, -0.15592814981937408, -0.07536590844392776, 0.5126543045043945, 0.07902255654335022, 0.06046098470687866, -0.2836170196533203, -0.3982830345630646, -0.16304101049900055, 0.07741996645927429, 0.3069847524166107, 0.4057582914829254, 0.34915271401405334, -0.08209376782178879, 0.19486494362354279, -0.1915401667356491, 0.09424611181020737, 0.027147360146045685, -0.034980982542037964, -0.30891916155815125, -0.15635012090206146, 0.16644422709941864, -0.5549852848052979, 0.11955330520868301, -0.18406997621059418, 0.11688795685768127, -0.2389354258775711, -0.27920156717300415, -0.07527454942464828, -0.02139112539589405, 0.22103837132453918, -0.4112195670604706, -0.13717730343341827, -0.08414352685213089, 0.4223146140575409, 0.23396289348602295, 0.250032514333725, -0.04165755584836006, 0.13203328847885132, 0.17086564004421234, -0.23383009433746338, -0.07233113795518875, 0.19056548178195953, -0.15384617447853088, 0.2510855495929718, -0.36824527382850647, -0.5960057377815247, -0.06363940238952637, -0.015842650085687637, 0.021004540845751762, 0.31264904141426086, -0.42155441641807556, 0.5755351185798645, -0.03536474332213402, 0.05393073707818985, -0.1453755497932434, -0.005995717830955982, 0.20333744585514069, -0.03361830115318298, 0.044098157435655594, -0.03009047731757164, 0.08547737449407578, 0.31540393829345703, -0.16932444274425507, 0.3524860441684723, 0.4192032814025879, 0.13491706550121307, 0.08188965916633606, 0.6222636699676514, 0.24559447169303894, 0.10145347565412521, 0.2365688532590866, -0.024821169674396515, 0.0064175790175795555, 0.026079639792442322, 0.05190190300345421, -0.14045022428035736, 0.19511979818344116, 0.12384891510009766, 0.07361648976802826, 0.2893810570240021, -0.09587361663579941, -0.2961913049221039, -0.2651779353618622, -0.3516785502433777, -0.20116353034973145, 0.29700028896331787, -0.3599577248096466, 0.27954256534576416, -0.15597060322761536, -0.21718305349349976, -0.06307173520326614, -0.3113391697406769, 0.059967041015625, 0.12868718802928925, 0.31906723976135254, 0.1490602046251297, -0.27069488167762756, 0.34821581840515137, -0.6118110418319702, 0.3497487008571625, 0.051624100655317307, -0.10029686987400055, 0.05672137811779976, 0.03056354820728302, 0.14685751497745514, 0.08642497658729553, -0.17882418632507324, 0.11098466068506241, -0.02531636878848076, 0.009351846762001514, 0.2899271547794342, -0.4226803779602051, 0.39033493399620056, -0.228807270526886, 0.11045870929956436, 0.012455775402486324, 0.12471452355384827, -0.37960654497146606, -0.2781965136528015, 0.13440127670764923, 0.24216315150260925, -0.1416960507631302, -0.36481329798698425, 0.2568531632423401, 0.008762945421040058, -0.1879391372203827, -0.11056026816368103, 0.3730458915233612, 0.020872727036476135, -0.07553911209106445, -0.19689346849918365, -0.05982212722301483, 0.15742194652557373, 0.12727147340774536, -0.32880979776382446, 0.10629303008317947, -0.27296218276023865, 0.026045678183436394, 0.05754629522562027, 0.3884016275405884, 0.008848216384649277, 0.08767405152320862, 0.17699582874774933, -0.21610431373119354, -0.06899240612983704, -0.3054358661174774, 0.6498405337333679, 0.22233247756958008, 0.12618425488471985, 0.1377878487110138, 0.2107263058423996, -0.2504124343395233, -0.07103011012077332, -0.14615558087825775, 0.04115201532840729, -0.06889710575342178, -0.426135390996933, -0.34913864731788635, 0.5717192888259888, 0.053953636437654495, 0.002720082877203822, 0.19814132153987885, 0.08373111486434937, -0.27170035243034363, 0.5873882174491882, 0.2792503833770752, 1.1855225563049316, -0.2742815613746643, 0.27304917573928833, -0.13550738990306854, -0.31601810455322266, 0.4253913462162018, -0.27867963910102844, -0.25471818447113037, -0.5003958344459534, 0.4487970769405365, -0.07695724070072174, -0.153147891163826, 0.429431676864624, 0.41808274388313293, -0.04471508413553238, 0.023450076580047607, 0.21344681084156036, -0.1962738037109375, 0.11166691780090332, 0.4447539746761322, -0.33992743492126465, -0.3009551167488098, -0.09718891978263855, -0.05439264327287674, -0.1266823559999466, 0.14949734508991241, 0.1293238252401352, -0.44057825207710266, -0.37254348397254944, 0.08562374114990234, -0.2790435552597046, 0.0675969123840332, -0.16045615077018738, -0.23008498549461365, -0.30945253372192383, -0.6526294350624084, 0.38459140062332153, 0.303102970123291, 0.1611373871564865, 0.01803254336118698, -0.14568142592906952, 0.06527040898799896, -0.30593761801719666, -0.21463309228420258, 0.14000838994979858, -0.0970301553606987, 0.014986231923103333, -0.17341424524784088, -0.14730991423130035, 0.13271565735340118, 0.07229068875312805, -0.2237471044063568, 0.03694796562194824, -0.20663891732692719, 0.5542369484901428, -0.22024871408939362, -0.05737181380391121, 0.2082994282245636, -0.38716939091682434, 0.09555824846029282, 0.07478750497102737, 0.34375086426734924, 0.010053154081106186, -0.01984357088804245, 0.2923181354999542, -0.19841355085372925, 0.06222645193338394, 0.19615599513053894, 0.20381395518779755, -0.19679854810237885, 0.43999508023262024, 0.38451382517814636, -0.09716916084289551, -0.3226233422756195, 0.5589423775672913, 0.4894661605358124, -0.7103242874145508, -0.05827104672789574, -0.0873330682516098, 0.17481070756912231, 0.0075121112167835236, 0.07378464937210083, -0.04773588478565216, -0.024275539442896843, -0.04158999025821686, -0.4507727324962616, -0.30290886759757996, 0.021305326372385025, -0.05329263210296631, 0.09318296611309052, 0.3139227628707886, 0.018152549862861633, 0.0753164067864418, 0.04191593825817108, -0.20519845187664032, 0.1590941995382309, 0.6357872486114502, 0.17778629064559937, 0.027765924111008644, 0.042296573519706726, -0.056716181337833405, -0.00015137423179112375, -0.004827890079468489, 0.13855087757110596, 0.07041745632886887, -0.17438171803951263, -0.032969195395708084, 0.2171880155801773, -0.0901581346988678, -0.32971855998039246, -0.3392930328845978, -0.4449428617954254, -0.154939666390419, -0.03558647632598877, -0.13537022471427917, -0.08628867566585541, -0.06770666688680649, 0.16228139400482178, 0.1766919493675232, -0.0913199782371521, -0.5623156428337097, 0.1294584572315216, -0.10235680639743805, 0.21732500195503235, -0.13209497928619385, 0.16996720433235168, 0.09857216477394104, -0.05798961594700813, 0.047921571880578995, 0.009129863232374191, 0.3277357518672943, 0.1631401926279068, 0.35374194383621216, -0.17176233232021332, 0.04795100912451744, -0.20626850426197052, 0.3029596507549286, 0.1416812688112259, -0.008553911931812763, 0.07447542995214462, 0.29503971338272095, 0.07731839269399643, -0.20862682163715363, 0.23932339251041412, -0.2136077880859375, -0.2218504697084427, -0.26494908332824707, 0.05970841273665428, 0.004555628169327974, 0.026558035984635353, -0.41372552514076233, -0.2439277619123459, 0.040048837661743164, 0.009618189185857773, -0.27304908633232117, 0.34262531995773315, 0.14183786511421204, -0.036722030490636826, -0.012312526814639568, 0.22880791127681732, 0.16126643121242523, -0.05761465057730675, -0.10032609850168228, 0.14080803096294403, 0.7164048552513123, 0.2588323652744293, 0.22971655428409576, -0.1495770364999771, 0.10609381645917892, 0.34074926376342773, 0.17358440160751343, 0.06045139580965042, 0.30843067169189453, 0.501695990562439, 0.49053341150283813, -0.06533724814653397, 0.10396638512611389, 0.23329581320285797, 0.026200024411082268, 0.09245134145021439, -0.2975194454193115, -0.09762128442525864, -0.07974164932966232, -0.27052736282348633, -0.26703548431396484, -0.06595683097839355, -0.12805059552192688, 0.37830913066864014, 0.224216029047966, 0.006519431713968515, 0.09486543387174606, 0.07628700137138367, 0.30902570486068726, -0.1967896819114685, -0.08400562405586243, -0.047925520688295364, 0.06425434350967407, 0.014820144511759281, 0.4466158151626587, 0.14851869642734528, 0.11663597077131271, -0.18219169974327087, 0.21957455575466156, -0.0323844775557518, -0.13474969565868378, 0.25331515073776245, 0.24931177496910095, 0.43743449449539185, 0.02465965785086155, 0.16236284375190735, 0.03777848184108734, 0.0600285567343235, 0.010893844068050385, 0.19674983620643616, -0.16068632900714874, -0.10829733312129974, -0.04838661849498749, -0.0743742510676384, -0.455786794424057, -0.21377234160900116, 0.07236232608556747, -0.1384686380624771, 0.20172777771949768, 0.6342796683311462, -0.30233120918273926, -0.05422493442893028, -0.2243727743625641, 0.012459943071007729, -0.11517751216888428, 0.42381519079208374, 0.3466531038284302, 0.14342562854290009, -0.29574811458587646, -0.28634220361709595, -0.06824088096618652, 0.11662812530994415, -0.023076988756656647, -0.04253241792321205, -0.0962422788143158, -0.031165892258286476, 0.27951785922050476, 0.5127829909324646, 0.13443009555339813, 0.12040817737579346, -0.17573845386505127, 0.48935049772262573, -0.2379414290189743, -0.01999349147081375, 0.4501173198223114, 0.5878043174743652, -0.15019358694553375, -0.1882658302783966, 0.32243508100509644, 0.6150752902030945, -0.06042821705341339, -0.19979678094387054, 0.007830449379980564, 0.33421823382377625, -0.17057205736637115, 0.20960384607315063, -0.013460722751915455, 0.09708001464605331, 0.02499525435268879, -0.09997033327817917, 0.14133478701114655, 0.03287096694111824, 0.12586292624473572, 0.5581209063529968, 0.25094422698020935, -0.17178672552108765, -0.2090209424495697, -0.0683218315243721, 0.015021709725260735, 0.23496048152446747, -0.33540868759155273, -0.08596171438694, -0.3126036822795868, -0.09726390242576599, -0.2712614834308624, 0.041893623769283295, 0.016884218901395798, 0.30374711751937866, -0.17290425300598145, 0.012211902998387814, -0.20690639317035675, 0.023852337151765823, 0.11828358471393585, -0.2757699191570282, 0.03076789900660515, 0.3645254969596863, 0.1029348373413086, 0.6225427389144897, 0.15047995746135712, -0.48482584953308105, 0.14814776182174683, 0.31665951013565063, -0.022683389484882355, 0.2637162506580353, 0.09053988009691238, 0.05713618919253349, 0.13081413507461548, 0.010113729164004326, 0.5547729134559631, -0.05681570991873741, 0.004993907641619444, -0.5158871412277222, -0.40870675444602966 ]
https://github.com/huggingface/datasets/issues/192
[Question] Create Apache Arrow dataset from raw text file
> Is there a way to convert a dataset to Apache arrow format (locally/personal use) & use it before sending it to hugging face? Sure, to get a dataset in arrow format you can either: - [load from local files (txt, json, csv)](https://huggingface.co/nlp/loading_datasets.html?highlight=csv#from-local-files) - OR [load from python data (dict, pandas)](https://huggingface.co/nlp/loading_datasets.html?highlight=csv#from-in-memory-data) - OR [create your own dataset script](https://huggingface.co/nlp/loading_datasets.html?highlight=csv#using-a-custom-dataset-loading-script)
Hi guys, I have gathered and preprocessed about 2GB of COVID papers from CORD dataset @ Kggle. I have seen you have a text dataset as "Crime and punishment" in Apache arrow format. Do you have any script to do it from a raw txt file (preprocessed as for BERT like) or any guide? Is the worth of send it to you and add it to the NLP library? Thanks, Manu
58
[Question] Create Apache Arrow dataset from raw text file Hi guys, I have gathered and preprocessed about 2GB of COVID papers from CORD dataset @ Kggle. I have seen you have a text dataset as "Crime and punishment" in Apache arrow format. Do you have any script to do it from a raw txt file (preprocessed as for BERT like) or any guide? Is the worth of send it to you and add it to the NLP library? Thanks, Manu > Is there a way to convert a dataset to Apache arrow format (locally/personal use) & use it before sending it to hugging face? Sure, to get a dataset in arrow format you can either: - [load from local files (txt, json, csv)](https://huggingface.co/nlp/loading_datasets.html?highlight=csv#from-local-files) - OR [load from python data (dict, pandas)](https://huggingface.co/nlp/loading_datasets.html?highlight=csv#from-in-memory-data) - OR [create your own dataset script](https://huggingface.co/nlp/loading_datasets.html?highlight=csv#using-a-custom-dataset-loading-script)
[ -0.11623457819223404, -0.1841173768043518, -0.07161260396242142, 0.17511966824531555, -0.3280259072780609, 0.13923661410808563, 0.11795967817306519, 0.381742924451828, 0.002251590834930539, -0.2727222442626953, 0.05691486969590187, 0.09839979559183121, -0.07723776251077652, 0.21921460330486298, 0.31098002195358276, -0.05802769958972931, -0.10293187946081161, 0.18729236721992493, -0.28998038172721863, 0.03419673070311546, 0.24861127138137817, 0.023550081998109818, -0.06273787468671799, 0.014808127656579018, -0.18604959547519684, -0.21685869991779327, -0.12376035004854202, 0.10292910039424896, -0.11955978721380234, -0.40562522411346436, 0.045382339507341385, -0.02501893974840641, 0.5199137926101685, 0.17655058205127716, -0.00012255783076398075, -0.30129384994506836, 0.14918212592601776, -0.008922027423977852, -0.2559584677219391, -0.3521428406238556, -0.09115894883871078, -0.05141794681549072, 0.3022277355194092, -0.27631500363349915, -0.1366523951292038, -0.36661356687545776, 0.044754065573215485, -0.46327662467956543, 0.8944171071052551, 0.38236644864082336, 0.0668816938996315, -0.06973101943731308, 0.12017328292131424, 0.1746978461742401, 0.29358920454978943, 0.6498957872390747, -0.29852011799812317, 0.017333848401904106, 0.2269863337278366, -0.027604324743151665, -0.18706895411014557, 0.21038731932640076, 0.03350196033716202, -0.16015833616256714, 0.41826921701431274, 0.032942820340394974, -0.31972914934158325, -0.2667316794395447, 0.10459914803504944, 0.2918638288974762, 0.4435725808143616, -0.5081270933151245, -0.06863618642091751, -0.15882951021194458, 0.19593703746795654, -0.4047844707965851, -0.20237360894680023, 0.591252326965332, -0.45607444643974304, 0.4850003719329834, 0.11115609854459763, -0.3301268517971039, -0.45467573404312134, -0.02336692623794079, 0.4130195081233978, -0.11923098564147949, -0.1424248069524765, -0.034221142530441284, -0.08781936764717102, 0.11395931988954544, 0.14466233551502228, -0.34028294682502747, -0.08196868747472763, 0.2702946662902832, -0.030417220667004585, -0.148184671998024, -0.568871796131134, -0.11655516177415848, 0.29623928666114807, 0.1948118954896927, 0.35812410712242126, -0.14922933280467987, -0.08302738517522812, -0.2412634789943695, 0.11707867681980133, -0.03598489612340927, 0.4621577858924866, 0.12294930219650269, -0.0228782519698143, -0.1033877283334732, 0.21610605716705322, -0.11804761737585068, -0.07146725058555603, -0.007159936707466841, -0.19580847024917603, 0.12521176040172577, 0.06821320205926895, -0.35838761925697327, -0.06369709223508835, -0.06358227878808975, -0.31606054306030273, -0.06485728919506073, -0.16414400935173035, 0.121300607919693, 0.15003016591072083, -0.14854158461093903, 0.2026013284921646, 0.4209899306297302, 0.002455300185829401, -0.234660342335701, 0.1313377022743225, -0.01091756671667099, -0.28929054737091064, 0.34125465154647827, 0.15145541727542877, 0.041026193648576736, -0.03367765247821808, -0.0651327520608902, 0.04187648743391037, -0.18663957715034485, 0.21144548058509827, 0.003503225976601243, 0.2554565370082855, -0.027918757870793343, 0.225520521402359, 0.05125439539551735, -0.0815240815281868, -0.0155559116974473, -0.1415526419878006, 0.31942251324653625, -0.19716139137744904, -0.0013486410025507212, -0.019552281126379967, -0.01256589125841856, -0.0712689533829689, -0.11709246039390564, -0.29664355516433716, 0.16671741008758545, -0.05533001944422722, 0.3410833775997162, 0.05154578760266304, 0.30411309003829956, -0.00443054037168622, -0.15689967572689056, 0.06064519286155701, 0.575158417224884, -0.27983954548835754, 0.07748064398765564, -0.06367827206850052, 0.19212456047534943, -0.13204027712345123, 0.3210466206073761, -0.24032816290855408, 0.2045174539089203, -0.10772120207548141, 0.21508751809597015, 0.6738166809082031, -0.11156921833753586, 0.16387350857257843, 0.5188357830047607, 0.05644973739981651, -0.23707449436187744, 0.11921439319849014, -0.11641427129507065, -0.07434342056512833, -0.002806695643812418, -0.3140348494052887, 0.5571509003639221, 0.08260875195264816, 0.01825551502406597, -0.12995703518390656, -0.2152365744113922, -0.09700965136289597, 0.004683067090809345, -0.23405659198760986, -0.014258431270718575, 0.06036318838596344, -0.5500678420066833, 0.4852306544780731, -0.3260752260684967, 0.4003869295120239, 0.23203279078006744, 0.002529915887862444, 0.022987814620137215, 0.13997317850589752, -0.03116598352789879, -0.26830294728279114, -0.25660693645477295, -0.31449970602989197, 0.22397677600383759, -0.45106664299964905, -0.24881400167942047, -0.3250749707221985, -0.13213951885700226, -0.1072455644607544, -0.021121950820088387, -0.06251724809408188, -0.20799534022808075, -0.10984964668750763, 0.053020548075437546, -0.21073831617832184, 0.36422210931777954, -0.05452951788902283, 0.2473112940788269, -0.3691830337047577, 0.1250728815793991, 0.13142317533493042, -0.010291432030498981, 0.23380397260189056, 0.176345095038414, -0.18964023888111115, 0.03401162475347519, 0.007023621816188097, 0.08414443582296371, 0.010062050074338913, 0.20413382351398468, 0.4599420726299286, -0.27060630917549133, 0.12420657277107239, -0.7130151987075806, 0.2794923782348633, 0.007425840478390455, 0.18464599549770355, -0.09204741567373276, -0.48254671692848206, 0.5747975707054138, -0.14492934942245483, 0.23111045360565186, -0.14596496522426605, 0.02744235284626484, -0.13340258598327637, -0.077724888920784, -0.22561131417751312, 0.06281176209449768, -0.03213789314031601, 0.3259415626525879, 0.3570249080657959, -0.12372776120901108, 0.0172533318400383, 0.03389221429824829, -0.22877894341945648, -0.22015053033828735, 0.15034855902194977, 0.2633676826953888, -0.24480031430721283, -0.07236389815807343, -0.032712455838918686, -0.00866828765720129, 0.11415620893239975, 0.17046155035495758, -0.02507430873811245, 0.12046796828508377, 0.20249825716018677, -0.14061786234378815, 0.30452245473861694, -0.10037080943584442, 0.4107741415500641, -0.17753581702709198, -0.1377764195203781, -0.012871265411376953, -0.12090454995632172, -0.23075710237026215, -0.16143406927585602, 0.10699716955423355, -0.17705515027046204, -0.008562144823372364, -0.16584239900112152, -0.4955018162727356, -0.41690942645072937, 0.23199604451656342, -0.06793065369129181, 0.2515639364719391, -0.16291961073875427, -0.3335672616958618, 0.12842433154582977, -0.1341257244348526, -0.10173007100820541, 0.4734622538089752, -0.007049552630633116, 0.016295352950692177, -0.2771504819393158, -0.38154301047325134, -0.16488194465637207, 0.06597542762756348, 0.36865049600601196, 0.41833508014678955, 0.3976899981498718, -0.12005135416984558, 0.13771846890449524, -0.1911897510290146, 0.014920582994818687, 0.04858463257551193, -0.09082561731338501, -0.21377722918987274, -0.12403123825788498, 0.13249251246452332, -0.5243592262268066, 0.0738314613699913, -0.11730535328388214, 0.08368214964866638, -0.23695909976959229, -0.28719258308410645, -0.09819792956113815, 0.013112323358654976, 0.14931638538837433, -0.3908486068248749, -0.16516441106796265, -0.10185368359088898, 0.502617597579956, 0.25109145045280457, 0.19040580093860626, 0.04375829920172691, 0.10605818033218384, 0.1319572478532791, -0.2570900321006775, -0.10649918764829636, 0.18408609926700592, -0.20154999196529388, 0.29774656891822815, -0.39212650060653687, -0.5845621824264526, -0.11088468134403229, -0.006885068025439978, 0.03418790549039841, 0.28568175435066223, -0.46180614829063416, 0.41268250346183777, -0.07110340893268585, 0.07117278128862381, -0.11788593232631683, -0.010395599529147148, 0.20418325066566467, -0.02525159902870655, 0.08497018367052078, -0.020421605557203293, -0.017281759530305862, 0.29869189858436584, -0.20029963552951813, 0.28786906599998474, 0.4254688620567322, 0.1723226010799408, 0.05806053802371025, 0.5829063653945923, 0.30137309432029724, 0.10892781615257263, 0.28402915596961975, -0.08005806803703308, 0.05361035466194153, -0.026313263922929764, 0.013504818081855774, -0.20941531658172607, 0.12492375820875168, 0.14099793136119843, 0.10728851705789566, 0.3072510063648224, -0.036929357796907425, -0.33212587237358093, -0.28665295243263245, -0.4147660732269287, -0.21308183670043945, 0.30389243364334106, -0.3870818316936493, 0.23485973477363586, -0.1826094686985016, -0.18359370529651642, -0.0711640864610672, -0.3148888647556305, 0.050903234630823135, 0.223403662443161, 0.35175684094429016, 0.20501132309436798, -0.26657727360725403, 0.4281856417655945, -0.6509034037590027, 0.36899179220199585, 0.0584447979927063, -0.1497984528541565, 0.05009743571281433, -0.004402024205774069, 0.17924967408180237, 0.0758812353014946, -0.1050071269273758, 0.16946415603160858, -0.018786074593663216, -0.022987572476267815, 0.14634069800376892, -0.5188401341438293, 0.39623966813087463, -0.17134439945220947, 0.027793027460575104, 0.012689119204878807, 0.19453543424606323, -0.43167978525161743, -0.34926384687423706, 0.11258293688297272, 0.2232152372598648, -0.12154310941696167, -0.331159383058548, 0.2177026867866516, -0.06027273088693619, -0.24204523861408234, -0.06904469430446625, 0.37267550826072693, 0.033900439739227295, -0.047613710165023804, -0.13076543807983398, -0.04701978340744972, 0.12143260985612869, 0.11338749527931213, -0.2611371576786041, 0.13783729076385498, -0.28408902883529663, 0.09319937229156494, 0.14680559933185577, 0.4258970022201538, 0.07082593441009521, 0.2117328941822052, 0.13413195312023163, -0.34086161851882935, -0.12266544252634048, -0.22335761785507202, 0.621730387210846, 0.27249881625175476, 0.15294727683067322, 0.09791528433561325, 0.28170016407966614, -0.21737466752529144, -0.1763012856245041, -0.09220187366008759, 0.029144268482923508, -0.06998015940189362, -0.42138558626174927, -0.3758067190647125, 0.5728220343589783, 0.03647726774215698, -0.021539069712162018, 0.1576511710882187, 0.14217351377010345, -0.28869274258613586, 0.5798207521438599, 0.24262966215610504, 1.1859331130981445, -0.28146785497665405, 0.31370699405670166, -0.03596223518252373, -0.36061787605285645, 0.5188568830490112, -0.20172934234142303, -0.27785560488700867, -0.4951658546924591, 0.38379696011543274, -0.04490673542022705, -0.18553100526332855, 0.37482815980911255, 0.376426637172699, -0.030154578387737274, 0.12418890744447708, 0.166664257645607, -0.121605783700943, 0.05953569710254669, 0.47545430064201355, -0.3123103082180023, -0.34697622060775757, -0.1508464515209198, -0.07249150425195694, -0.14247477054595947, 0.232831671833992, 0.1278475970029831, -0.45531782507896423, -0.36892184615135193, 0.06390176713466644, -0.2599179446697235, 0.10600897669792175, -0.2166537195444107, -0.2464064210653305, -0.263786256313324, -0.638937771320343, 0.4127868711948395, 0.3639930486679077, 0.22060644626617432, -0.04784112051129341, -0.21090304851531982, 0.08878754824399948, -0.3631049692630768, -0.18496012687683105, 0.12916606664657593, -0.14812707901000977, 0.03022756241261959, -0.1457625776529312, -0.11701958626508713, 0.035430632531642914, 0.1157597079873085, -0.17766128480434418, -0.019819606095552444, -0.1916961967945099, 0.4527113735675812, -0.2664206624031067, -0.10803959518671036, 0.19887109100818634, -0.2783825397491455, 0.08723249286413193, 0.034055132418870926, 0.31136366724967957, 0.04984861984848976, -0.07525179535150528, 0.2784469425678253, -0.22621776163578033, 0.02781749702990055, 0.22859454154968262, 0.16524992883205414, -0.2648727595806122, 0.44301149249076843, 0.4258752465248108, -0.09707011282444, -0.25460851192474365, 0.46422380208969116, 0.5814706087112427, -0.8045473694801331, -0.02388317510485649, -0.0502198189496994, 0.1676587164402008, 0.024947986006736755, 0.11201496422290802, -0.01312509085983038, -0.09906873852014542, 0.03518993780016899, -0.46549445390701294, -0.32298707962036133, 0.08474759012460709, -0.057635433971881866, 0.10016036033630371, 0.3351861238479614, 0.10849282145500183, 0.05705320090055466, 0.06976903229951859, -0.19205914437770844, 0.17547237873077393, 0.6514863967895508, 0.16727960109710693, 0.12899930775165558, 0.044356491416692734, 0.05866916477680206, -0.010352068580687046, -0.02942405641078949, 0.15425308048725128, 0.03348403796553612, -0.1404445916414261, -0.05531333014369011, 0.22434088587760925, -0.11151158809661865, -0.3583466410636902, -0.30706554651260376, -0.35294052958488464, -0.05047788843512535, -0.14498591423034668, -0.0487511046230793, -0.06242743507027626, -0.048514436930418015, 0.11967755854129791, 0.09116733074188232, -0.0660633072257042, -0.5188177227973938, 0.16587625443935394, -0.12332584708929062, 0.21302905678749084, -0.027788620442152023, 0.1678663045167923, -0.014582050032913685, -0.0924045667052269, 0.06336047500371933, 0.019591573625802994, 0.34707528352737427, 0.20046566426753998, 0.33112573623657227, -0.11110337823629379, -0.01902942918241024, -0.20461244881153107, 0.3492092788219452, 0.14204499125480652, 0.04396377503871918, 0.04862741008400917, 0.3234451711177826, 0.06121789664030075, -0.1930478811264038, 0.24553310871124268, -0.17272059619426727, -0.18861496448516846, -0.25194215774536133, 0.0988532155752182, 0.03157610446214676, 0.10778149962425232, -0.4496343433856964, -0.23619265854358673, 0.17203684151172638, 0.06814946979284286, -0.2589748799800873, 0.27036571502685547, 0.14384528994560242, -0.039956722408533096, -0.007923359051346779, 0.2794294059276581, 0.2292509824037552, -0.08278074860572815, -0.14109349250793457, 0.17375065386295319, 0.6221063733100891, 0.35926955938339233, 0.2903221845626831, -0.2367408275604248, 0.06427197903394699, 0.3366841971874237, 0.18424731492996216, 0.0393265001475811, 0.20483262836933136, 0.6484077572822571, 0.5068117380142212, -0.0744929239153862, 0.11960915476083755, 0.2935684621334076, 0.049009453505277634, 0.03089696727693081, -0.25908687710762024, -0.08743663877248764, -0.12445323169231415, -0.19977812469005585, -0.2683987021446228, -0.15598931908607483, -0.07701190561056137, 0.32390695810317993, 0.194288432598114, -0.028334541246294975, 0.028101205825805664, 0.0917002484202385, 0.30897948145866394, -0.17272520065307617, 0.02200634777545929, 0.0015084290644153953, 0.010933653451502323, 0.040897615253925323, 0.4318581819534302, 0.24735650420188904, 0.0948859453201294, -0.12778699398040771, 0.2288217842578888, 0.026228640228509903, -0.09294784814119339, 0.21523043513298035, 0.19966153800487518, 0.396843820810318, 0.035862430930137634, 0.18723462522029877, 0.037547849118709564, 0.06753212958574295, -0.05572819709777832, 0.1819804608821869, -0.15538400411605835, -0.07866637408733368, -0.1503336876630783, -0.051529109477996826, -0.3914929926395416, -0.198607936501503, 0.06892614811658859, -0.17805640399456024, 0.20385095477104187, 0.5880893468856812, -0.24701930582523346, -0.049937114119529724, -0.16663596034049988, 0.00043457603896968067, -0.03191368654370308, 0.4568953514099121, 0.36623191833496094, 0.14163050055503845, -0.30244699120521545, -0.3249179422855377, -0.13995298743247986, 0.09046243876218796, -0.014410941861569881, -0.09643489122390747, -0.05399179086089134, -0.0555766299366951, 0.22305022180080414, 0.4540475606918335, 0.023300519213080406, 0.17148026823997498, -0.12571406364440918, 0.4537195861339569, -0.22176802158355713, -0.010113283060491085, 0.37255924940109253, 0.5218804478645325, -0.10128170251846313, -0.13284315168857574, 0.2542346715927124, 0.5032437443733215, -0.09317484498023987, -0.157328799366951, 0.056906770914793015, 0.2313569039106369, -0.25062295794487, 0.20016852021217346, 0.03537285700440407, 0.1330995112657547, -0.007061309646815062, -0.1503179967403412, 0.09966892004013062, -0.027995437383651733, 0.059136100113391876, 0.6108750700950623, 0.24947233498096466, -0.10200654715299606, -0.2170572280883789, -0.11341237276792526, -0.05578738451004028, 0.2859751880168915, -0.2583237588405609, -0.1287272870540619, -0.3091922998428345, -0.06209065765142441, -0.2419237196445465, 0.025370623916387558, 0.0015789661556482315, 0.4026339650154114, -0.15525579452514648, 0.06434924155473709, -0.16876113414764404, -0.04738330468535423, 0.2202807068824768, -0.3101285398006439, 0.08213483542203903, 0.32826748490333557, 0.1244426891207695, 0.6768018007278442, 0.10875978320837021, -0.5213273763656616, 0.19877082109451294, 0.2741398811340332, 0.00410881545394659, 0.21930517256259918, 0.0936860516667366, 0.05322783812880516, 0.07330220192670822, -0.03502911329269409, 0.45862334966659546, -0.02953355759382248, -0.029490958899259567, -0.4290831983089447, -0.43094953894615173 ]
https://github.com/huggingface/datasets/issues/189
[Question] BERT-style multiple choice formatting
Hi @sarahwie, can you details this a little more? I'm not sure I understand what you refer to and what you mean when you say "Previously, this was done by passing a list of InputFeatures to the dataloader instead of a list of InputFeature"
Hello, I am wondering what the equivalent formatting of a dataset should be to allow for multiple-choice answering prediction, BERT-style. Previously, this was done by passing a list of `InputFeatures` to the dataloader instead of a list of `InputFeature`, where `InputFeatures` contained lists of length equal to the number of answer choices in the MCQ instead of single items. I'm a bit confused on what the output of my feature conversion function should be when using `dataset.map()` to ensure similar behavior. Thanks!
44
[Question] BERT-style multiple choice formatting Hello, I am wondering what the equivalent formatting of a dataset should be to allow for multiple-choice answering prediction, BERT-style. Previously, this was done by passing a list of `InputFeatures` to the dataloader instead of a list of `InputFeature`, where `InputFeatures` contained lists of length equal to the number of answer choices in the MCQ instead of single items. I'm a bit confused on what the output of my feature conversion function should be when using `dataset.map()` to ensure similar behavior. Thanks! Hi @sarahwie, can you details this a little more? I'm not sure I understand what you refer to and what you mean when you say "Previously, this was done by passing a list of InputFeatures to the dataloader instead of a list of InputFeature"
[ 0.18983325362205505, -0.4081771969795227, -0.02667146734893322, -0.07893255352973938, 0.1755470484495163, -0.11440084129571915, 0.29332637786865234, 0.23464754223823547, -0.11313829571008682, -0.006701129954308271, -0.07507651299238205, 0.5407446622848511, -0.25638601183891296, 0.21988390386104584, 0.0012710931478068233, -0.4498932659626007, 0.16272614896297455, 0.13243594765663147, -0.1213601678609848, -0.11142545193433762, -0.22482644021511078, -0.18424533307552338, -0.25302571058273315, -0.10311741381883621, -0.1576479971408844, 0.06863229721784592, -0.3666488528251648, -0.2160356044769287, -0.3754041790962219, -0.3336400091648102, -0.19366523623466492, 0.28235217928886414, -0.21831142902374268, -0.03667671978473663, -0.00012514082482084632, -0.4015752375125885, -0.1773587018251419, -0.174785777926445, -0.11024003475904465, -0.11791319400072098, -0.8842170834541321, -0.0070363678969442844, 0.38440072536468506, 0.01858721300959587, 0.007791478652507067, 0.026045018807053566, 0.02693788707256317, -0.30778056383132935, 0.48101407289505005, -0.00877204816788435, 0.07200628519058228, -0.03722210228443146, 0.036857470870018005, 0.06419794261455536, 0.13089828193187714, 0.5086039900779724, 0.01867241971194744, -0.04318336024880409, 0.6300452351570129, 0.0035536789800971746, -0.20570042729377747, 0.3613415062427521, 0.09690272063016891, 0.10612980276346207, 0.34405022859573364, -0.026202095672488213, -0.0527590773999691, -0.0893978476524353, -0.23088891804218292, 0.3512217402458191, 0.7839899063110352, -0.09999584406614304, -0.26931771636009216, -0.03707228600978851, 0.021386830136179924, 0.06871086359024048, -0.18641172349452972, 0.014608637429773808, -0.14011618494987488, -0.06115395948290825, -0.4859077036380768, -0.32537612318992615, -0.3561825752258301, 0.06938924640417099, -0.27644792199134827, 0.42656078934669495, 0.008110071532428265, -0.03517956659197807, -0.052210889756679535, -0.2725643217563629, -0.021641544997692108, -0.43927112221717834, 0.3904174268245697, 0.1420876532793045, 0.1063593178987503, -0.379153311252594, -0.18745790421962738, 0.07733535766601562, 0.3710349500179291, -0.29466238617897034, -0.09317895770072937, -0.0026960226241499186, -0.5204654335975647, -0.012472649104893208, 0.5711631774902344, 0.002248713979497552, -0.0457250215113163, 0.07216697186231613, -0.34575319290161133, -0.2913959324359894, -0.2567959129810333, 0.04164733737707138, 0.12521569430828094, 0.22841861844062805, 0.0716412216424942, 0.12182886153459549, 0.05681386590003967, -0.32156211137771606, 0.06382255256175995, 0.07790389657020569, -0.3623914122581482, -0.2042137086391449, 0.03367432951927185, 0.15451663732528687, -0.08497172594070435, 0.5170227289199829, -0.19649600982666016, 0.20776665210723877, 0.08788161724805832, -0.356895387172699, 0.1389588713645935, -0.1019803062081337, -0.527290940284729, -0.2540569305419922, 0.1949182003736496, 0.2871018052101135, -0.010821481235325336, 0.3692454695701599, 0.3783961236476898, -0.23002678155899048, 0.07069478929042816, 0.024962684139609337, 0.564206600189209, -0.18388107419013977, -0.35279935598373413, 0.3267548382282257, 0.2906136214733124, 0.3297918140888214, -0.08868246525526047, 0.501160740852356, -0.2108786553144455, -0.014086611568927765, -0.15334878861904144, 0.017584431916475296, 0.21666336059570312, -0.21159720420837402, 0.041119106113910675, 0.35120153427124023, 0.04503424093127251, -0.33556750416755676, 0.13014279305934906, -0.5346511602401733, 0.049892738461494446, -0.29566824436187744, 0.00257532368414104, 0.4049507975578308, -0.46270596981048584, 0.09012237936258316, -0.19625265896320343, 0.23597653210163116, 0.22379319369792938, 0.26253268122673035, -0.07536940276622772, -0.05960937961935997, -0.011267870664596558, 0.3569810390472412, 0.3155226409435272, -0.1638091802597046, -0.31837067008018494, 0.229786217212677, 0.160812646150589, 0.11740345507860184, -0.03637678548693657, 0.23446953296661377, 0.15473641455173492, 0.26902490854263306, 0.05115644261240959, -0.1387484222650528, -0.30224478244781494, 0.36625200510025024, -0.018000775948166847, -0.04290363937616348, 0.12097184360027313, -0.04484029859304428, -0.07994305342435837, -0.14226490259170532, -0.08018451929092407, 0.1147543266415596, 0.2186724990606308, -0.205487459897995, -0.0574701651930809, -0.42732080817222595, 0.1527346968650818, 0.02900855429470539, -0.12003222107887268, -0.17264285683631897, -0.63584303855896, -0.07989735901355743, -0.0288462545722723, -0.18700334429740906, 0.22696945071220398, -0.2735061049461365, -0.2580793499946594, 0.05828803777694702, -0.3363605737686157, 0.10470392554998398, -0.013620543293654919, 0.02001439966261387, -0.24578633904457092, -0.2998508810997009, -0.2771872878074646, -0.06258799880743027, 0.5232807397842407, 0.018669918179512024, 0.05305343121290207, 0.11633868515491486, 0.0966973826289177, 0.002988348715007305, 0.13223688304424286, 0.3557765483856201, 0.19228705763816833, 0.2523918151855469, -0.03136144205927849, 0.2370244711637497, 0.45958828926086426, -0.04832536727190018, -0.038657598197460175, 0.11578815430402756, 0.4056984484195709, -0.23331955075263977, 0.3231179714202881, 0.17757347226142883, 0.09546288102865219, -0.0138601865619421, -0.8099329471588135, 0.1893579065799713, -0.07852783799171448, 0.11698033660650253, -0.14940395951271057, 0.2087835967540741, -0.19297188520431519, 0.2378271520137787, -0.060355789959430695, -0.4787614643573761, -0.12163136154413223, 0.26565805077552795, 0.15038207173347473, 0.07501335442066193, -0.4771553874015808, 0.27814239263534546, 0.3794211447238922, -0.01787433959543705, 0.22816415131092072, -0.06022470444440842, -0.017266787588596344, -0.09874331206083298, 0.4197058081626892, -0.2194468230009079, 0.23195821046829224, 0.15621227025985718, -0.06431907415390015, -0.1458723396062851, 0.008576455526053905, -0.12822562456130981, -0.0028875048737972975, -0.03820915147662163, -0.10356073826551437, 0.10288160294294357, 0.19065600633621216, -0.10545192658901215, 0.002276940969750285, 0.07328981906175613, -0.0567629300057888, -0.0949905589222908, -0.1219477728009224, 0.005622597876936197, -0.5228873491287231, -0.016359154134988785, -0.15502722561359406, -0.5842822790145874, -0.19491077959537506, -0.18189583718776703, 0.2008170783519745, -0.19690655171871185, -0.08773338794708252, 0.006360611412674189, 0.08728300780057907, 0.3509708642959595, 0.13916289806365967, 0.0905078649520874, 0.07922912389039993, -0.017442205920815468, 0.06981471180915833, 0.0482323095202446, 0.004030866548418999, 0.03277776017785072, 0.19473348557949066, 0.10559038817882538, 0.011595897376537323, -0.35372400283813477, 0.015374623239040375, 0.19428400695323944, -0.049460653215646744, 0.010338283143937588, 0.33580514788627625, 0.08736906200647354, -0.4954448640346527, 0.13991810381412506, 0.14058904349803925, 0.3530600965023041, -0.052326176315546036, 0.030835386365652084, 0.3473845422267914, 0.017780223861336708, 0.09395381808280945, -0.3488006293773651, -0.1854473054409027, 0.04607227444648743, 0.09167945384979248, -0.024083873257040977, 0.10844918340444565, -0.23966246843338013, 0.1782774180173874, -0.1272115260362625, -0.3487854301929474, -0.24792636930942535, -0.20571382343769073, 0.11752548068761826, 0.04890187084674835, -0.14117582142353058, -0.017388172447681427, -0.21726776659488678, 0.1409630924463272, -0.13892069458961487, 0.18003664910793304, 0.27049025893211365, -0.2778683602809906, -0.0454745814204216, 0.12244541198015213, 0.09823621064424515, 0.42160651087760925, 0.3028854727745056, 0.3918447494506836, 0.08133598417043686, -0.07759658992290497, 0.10318708419799805, 0.11008727550506592, 0.321149617433548, 0.009788180701434612, 0.14529140293598175, 0.04953543841838837, -0.13081257045269012, 0.78508061170578, 0.47630515694618225, 0.14479032158851624, 0.11252351105213165, 0.00275333016179502, -0.16256745159626007, -0.1379137933254242, -0.10453350841999054, 0.26983121037483215, -0.0767843946814537, 0.19053742289543152, 0.4577786326408386, 0.08809152245521545, -0.06342148035764694, 0.03595435619354248, 0.29231977462768555, -0.2363467663526535, -0.2863628566265106, 0.42527392506599426, -0.17180947959423065, 0.25057122111320496, 0.014178016223013401, -0.11715511232614517, -0.2502310276031494, -0.12013685703277588, -0.06141119450330734, -0.11164309084415436, 0.34356606006622314, -0.3986707627773285, -0.09759087860584259, -0.25619882345199585, -0.15704910457134247, 0.29891860485076904, 0.47733354568481445, 0.10383449494838715, -0.013509053736925125, -0.0769287645816803, 0.16502852737903595, 0.34249287843704224, 0.517250120639801, -0.039350129663944244, -0.11622565984725952, -0.09208127856254578, 0.0562993623316288, 0.17325037717819214, -0.03681081160902977, -0.6540711522102356, 0.32207444310188293, -0.2926158607006073, 0.13349077105522156, -0.2254885584115982, -0.35192805528640747, 0.17181231081485748, 0.14538060128688812, -0.10473261773586273, -0.2261105477809906, 0.1381576806306839, 0.5807726383209229, 0.0008979708072729409, 0.25213271379470825, 0.22878926992416382, -0.1592707633972168, 0.19195708632469177, -0.09325652569532394, 0.18506529927253723, -0.006664058193564415, 0.3489798307418823, 0.21433115005493164, 0.06443051248788834, 0.16485755145549774, -0.39185938239097595, 0.22942419350147247, 0.15006153285503387, 0.19573767483234406, 0.20301271975040436, -0.6772521734237671, -0.06086090952157974, 0.08496527373790741, -0.11696690320968628, 0.6384950876235962, 0.3110851049423218, 0.2727367579936981, -0.08074910938739777, 0.09046310186386108, -0.014860443770885468, -0.1867108941078186, -0.07817555218935013, 0.3036338984966278, 0.08140677213668823, -0.2078109085559845, -0.416169136762619, 0.38397490978240967, -0.11780273914337158, -0.2680467963218689, -0.1368650645017624, 0.5329195261001587, 0.0063750967383384705, 0.032525524497032166, 0.41089510917663574, 0.8881409764289856, -0.18266144394874573, 0.3090895116329193, 0.13335886597633362, 0.07515903562307358, 0.3711300492286682, -0.13379494845867157, 0.22359661757946014, -0.22098933160305023, 0.1054655984044075, -0.1843821108341217, -0.1960986703634262, -0.00030952002271078527, 0.5590389370918274, -0.6933904886245728, 0.2738441526889801, -0.15490444004535675, 0.15904997289180756, 0.1415613293647766, -0.062247276306152344, 0.31148844957351685, -0.1365654021501541, -0.13445068895816803, -0.012795763090252876, -0.20866385102272034, -0.13177894055843353, -0.15400853753089905, -0.04845930635929108, 0.03196560963988304, 0.005609406623989344, -0.31481584906578064, -0.06546603888273239, -0.5026325583457947, -0.22931405901908875, 0.07046796381473541, -0.13271157443523407, 0.4928193986415863, 0.47302302718162537, 0.11012987047433853, 0.12637826800346375, -0.08461318165063858, -0.18574944138526917, 0.13388431072235107, 0.2864260673522949, -0.23144851624965668, 0.10738258808851242, 0.5054354071617126, -0.02876119874417782, -0.3097275495529175, 0.23615434765815735, -0.12689250707626343, -0.0892120897769928, -0.4850243926048279, 0.1137210875749588, 0.7107089757919312, -0.5381379723548889, -0.4216561019420624, -0.03158736601471901, 0.12211253494024277, 0.028233230113983154, -0.08895724266767502, 0.26383456587791443, 0.11011478304862976, 0.5779934525489807, 0.23609013855457306, 0.157501682639122, 0.24454250931739807, -0.13318346440792084, 0.14268632233142853, -0.10161645710468292, 0.3163881301879883, 0.20700937509536743, -0.14301596581935883, -0.1376817673444748, 0.08183921873569489, 0.22919102013111115, -0.10338233411312103, 0.13760650157928467, -0.2527719736099243, -0.08193761855363846, -0.03888706490397453, 0.33960410952568054, 0.34378546476364136, 0.3037850260734558, -0.009304904378950596, 0.049746379256248474, -0.11958938837051392, 0.2682492136955261, -0.22673344612121582, 0.1715279370546341, 0.19719402492046356, -0.15251551568508148, 0.06678912788629532, 0.2607312798500061, -0.17686215043067932, -0.0941503643989563, 0.114686518907547, 0.3973579406738281, -0.002525784308090806, -0.20133496820926666, -0.22507244348526, -0.05165533348917961, -0.040725432336330414, -0.22952818870544434, -0.22013945877552032, -0.12144709378480911, -0.10964029282331467, 0.13896630704402924, 0.09552177786827087, -0.07645732909440994, -0.12439626455307007, -0.12791919708251953, -0.008751492947340012, -0.124936543405056, -0.0020421319641172886, -0.06993533670902252, -0.16730555891990662, 0.33069726824760437, 0.13859660923480988, -0.34176239371299744, -0.19411781430244446, 0.12415028363466263, 0.3273497223854065, -0.04631177335977554, 0.17618295550346375, 0.12686513364315033, 0.06882349401712418, -0.2199273556470871, -0.07351956516504288, 0.43982845544815063, 0.279468297958374, 0.03449907526373863, 0.028155971318483353, 0.09866365045309067, 0.163574680685997, 0.09105563163757324, 0.29285115003585815, 0.05899430811405182, 0.0481066070497036, 0.09166978299617767, 0.04486212134361267, 0.08739276975393295, -0.10868420451879501, -0.218077152967453, 0.5362926125526428, 0.10885220766067505, 0.16727381944656372, 0.35398221015930176, 0.058678314089775085, 0.01752314530313015, 0.37596365809440613, -0.02889471873641014, -0.35815805196762085, 0.0637383759021759, 0.014865649864077568, 0.6164770126342773, -0.09675861150026321, 0.060041092336177826, 0.15797939896583557, 0.2112785130739212, 0.06314340978860855, -0.003104474162682891, 0.07378432899713516, 0.21520595252513885, 0.11956258863210678, -0.1352907419204712, 0.49863892793655396, -0.016308991238474846, 0.21603018045425415, 0.22185875475406647, 0.10694896429777145, 0.44822296500205994, 0.3513822555541992, 0.13832300901412964, -0.2611650228500366, -0.5070871710777283, -0.00582129368558526, 0.08295252174139023, -0.14927959442138672, -0.18763814866542816, -0.023020895197987556, -0.220364511013031, -0.34096360206604004, -0.030820239335298538, -0.693973183631897, -0.1361895501613617, 0.46857306361198425, 0.32330021262168884, 0.5033946633338928, -0.12030564993619919, -0.04962632805109024, 0.105863556265831, 0.29202592372894287, -0.12352444976568222, 0.18553689122200012, 0.033744655549526215, 0.027988296002149582, 0.10734627395868301, 0.341978520154953, -0.0009038878488354385, 0.2166767567396164, -0.18185113370418549, -0.031961310654878616, -0.3361883759498596, -0.08698061853647232, 0.2088310569524765, -0.23594452440738678, 0.22233764827251434, -0.07854901999235153, 0.15231919288635254, 0.011358083225786686, 0.05159839242696762, 0.15602988004684448, 0.06995560973882675, 0.006509854458272457, -0.2356918454170227, -0.5836459398269653, 0.21649886667728424, -0.25015988945961, 0.08592205494642258, -0.03139951825141907, -0.6951826810836792, -0.02692946419119835, 0.05683095380663872, -0.10170379281044006, -0.08000952750444412, 0.14819759130477905, -0.012619337067008018, 0.26396530866622925, -0.029805989935994148, 0.36169713735580444, 0.014449600130319595, -0.32095739245414734, -0.17841556668281555, -0.2791799306869507, -0.11175796389579773, 0.24921168386936188, -0.12218495458364487, 0.07240942120552063, -0.02137734740972519, 0.05432417243719101, 0.3683215379714966, 0.05567153915762901, -0.5797141194343567, -0.25978371500968933, 0.4736153185367584, -0.1560892015695572, 0.16441193222999573, 0.2888821065425873, 0.25900229811668396, -0.02821209840476513, -0.06761839985847473, 0.27827540040016174, 0.31077009439468384, -0.1515820175409317, -0.2304983288049698, 0.010738945566117764, -0.24945513904094696, -0.059917762875556946, -0.22761529684066772, 0.36596646904945374, 0.27638158202171326, 0.10707338154315948, -0.20322756469249725, 0.2864719033241272, 0.4057309031486511, 0.11927662044763565, 0.09432721138000488, 0.11116784065961838, 0.6024981141090393, -0.14940732717514038, 0.2883361279964447, -0.05454830452799797, 0.02351175993680954, 0.09809593111276627, -0.16385097801685333, -0.6133046746253967, -0.022443752735853195, -0.023944243788719177, -0.32011792063713074, 0.13848373293876648, 0.41740304231643677, -0.12012837082147598, 0.18614080548286438, -0.24419806897640228, 0.2617795467376709, 0.13630367815494537, -0.22383356094360352, -0.34966328740119934, -0.30819305777549744, -0.045921895653009415, -0.24816477298736572, -0.014777674339711666, -0.7189199328422546, -0.24108226597309113, 0.23456324636936188, -0.11553258448839188, -0.1438896656036377, 0.04211645573377609, -0.0393071249127388, -0.16625702381134033, -0.36295104026794434, 0.6250200867652893, 0.036941900849342346, 0.10133004933595657, -0.31193286180496216, -0.4519559144973755 ]
https://github.com/huggingface/datasets/issues/189
[Question] BERT-style multiple choice formatting
I think I've resolved it. For others' reference: to convert from using the [`MultipleChoiceDataset` class](https://github.com/huggingface/transformers/blob/a34a9896ac2a4a33ff9cd805c76eed914c8d8965/examples/multiple-choice/utils_multiple_choice.py#L82)/[`run_multiple_choice.py`](https://github.com/huggingface/transformers/blob/a34a9896ac2a4a33ff9cd805c76eed914c8d8965/examples/multiple-choice/run_multiple_choice.py) script in Huggingface Transformers, I've done the following for hellaswag: 1. converted the `convert_examples_to_features()` function to only take one input and return a dictionary rather than a list: ``` def convert_examples_to_features(example, tokenizer, max_length): choices_inputs = defaultdict(list) for ending_idx, ending in enumerate(example['endings']['ending']): text_a = example['ctx'] text_b = ending inputs = tokenizer.encode_plus( text_a, text_b, add_special_tokens=True, max_length=max_length, pad_to_max_length=True, return_overflowing_tokens=True, ) if "num_truncated_tokens" in inputs and inputs["num_truncated_tokens"] > 0: logger.info( "Attention! you are cropping tokens (swag task is ok). " "If you are training ARC and RACE and you are poping question + options," "you need to try to use a bigger max seq length!" ) for key in inputs: choices_inputs[key].append(inputs[key]) choices_inputs['label'] = int(example['label']) return choices_inputs ``` 2. apply this directly (instance-wise) to dataset, convert dataset to torch tensors. Dataset is then ready to be passed to `Trainer` instance. ``` dataset['train'] = dataset['train'].map(lambda x: convert_examples_to_features(x, tokenizer, max_length), batched=False) columns = ['input_ids', 'token_type_ids', 'attention_mask', 'label'] dataset['train'].set_format(type='torch', columns=columns) ```
Hello, I am wondering what the equivalent formatting of a dataset should be to allow for multiple-choice answering prediction, BERT-style. Previously, this was done by passing a list of `InputFeatures` to the dataloader instead of a list of `InputFeature`, where `InputFeatures` contained lists of length equal to the number of answer choices in the MCQ instead of single items. I'm a bit confused on what the output of my feature conversion function should be when using `dataset.map()` to ensure similar behavior. Thanks!
168
[Question] BERT-style multiple choice formatting Hello, I am wondering what the equivalent formatting of a dataset should be to allow for multiple-choice answering prediction, BERT-style. Previously, this was done by passing a list of `InputFeatures` to the dataloader instead of a list of `InputFeature`, where `InputFeatures` contained lists of length equal to the number of answer choices in the MCQ instead of single items. I'm a bit confused on what the output of my feature conversion function should be when using `dataset.map()` to ensure similar behavior. Thanks! I think I've resolved it. For others' reference: to convert from using the [`MultipleChoiceDataset` class](https://github.com/huggingface/transformers/blob/a34a9896ac2a4a33ff9cd805c76eed914c8d8965/examples/multiple-choice/utils_multiple_choice.py#L82)/[`run_multiple_choice.py`](https://github.com/huggingface/transformers/blob/a34a9896ac2a4a33ff9cd805c76eed914c8d8965/examples/multiple-choice/run_multiple_choice.py) script in Huggingface Transformers, I've done the following for hellaswag: 1. converted the `convert_examples_to_features()` function to only take one input and return a dictionary rather than a list: ``` def convert_examples_to_features(example, tokenizer, max_length): choices_inputs = defaultdict(list) for ending_idx, ending in enumerate(example['endings']['ending']): text_a = example['ctx'] text_b = ending inputs = tokenizer.encode_plus( text_a, text_b, add_special_tokens=True, max_length=max_length, pad_to_max_length=True, return_overflowing_tokens=True, ) if "num_truncated_tokens" in inputs and inputs["num_truncated_tokens"] > 0: logger.info( "Attention! you are cropping tokens (swag task is ok). " "If you are training ARC and RACE and you are poping question + options," "you need to try to use a bigger max seq length!" ) for key in inputs: choices_inputs[key].append(inputs[key]) choices_inputs['label'] = int(example['label']) return choices_inputs ``` 2. apply this directly (instance-wise) to dataset, convert dataset to torch tensors. Dataset is then ready to be passed to `Trainer` instance. ``` dataset['train'] = dataset['train'].map(lambda x: convert_examples_to_features(x, tokenizer, max_length), batched=False) columns = ['input_ids', 'token_type_ids', 'attention_mask', 'label'] dataset['train'].set_format(type='torch', columns=columns) ```
[ 0.1992313712835312, -0.5757999420166016, 0.010819805786013603, -0.06993655860424042, 0.2859193980693817, -0.15366728603839874, 0.2446015179157257, 0.24167542159557343, -0.009871233254671097, 0.01801912486553192, -0.1549077332019806, 0.4558008909225464, -0.2713591456413269, 0.07535450905561447, 0.12558700144290924, -0.37995243072509766, 0.15262740850448608, 0.06899341940879822, -0.2016223967075348, -0.06751982122659683, -0.11300859600305557, 0.028114499524235725, -0.16656041145324707, -0.02064732275903225, -0.35132792592048645, 0.046457089483737946, -0.36015427112579346, -0.06562928110361099, -0.302508682012558, -0.39813727140426636, -0.13272543251514435, 0.19807372987270355, -0.25676771998405457, 0.11358483135700226, -0.00013058971671853215, -0.31119462847709656, -0.21208462119102478, -0.214484304189682, -0.0953419879078865, -0.03154289349913597, -0.720818281173706, -0.011711298488080502, 0.3938957750797272, 0.10779374837875366, -0.025060908868908882, -0.024407371878623962, 0.0035535588394850492, -0.13218991458415985, 0.6255208849906921, -0.02035323902964592, 0.009229087270796299, 0.04004593566060066, 0.08174990862607956, 0.0990011915564537, 0.16138499975204468, 0.5654865503311157, -0.07321808487176895, 0.08077124506235123, 0.5695481300354004, -0.0895330011844635, -0.23184435069561005, 0.5064967274665833, 0.03164716437458992, 0.07505466043949127, 0.29418256878852844, 0.1139432042837143, -0.015605716966092587, -0.08051599562168121, -0.22278228402137756, 0.4199119806289673, 0.5713578462600708, -0.30700621008872986, -0.4449508786201477, -0.28117117285728455, 0.07755286991596222, -0.0533093623816967, -0.18529567122459412, 0.04425524175167084, -0.24635545909404755, 0.0007776309503242373, -0.3870478570461273, -0.14347636699676514, -0.2480069100856781, 0.035836365073919296, -0.26923590898513794, 0.511878490447998, -0.05074466019868851, 0.021875252947211266, -0.006453304085880518, -0.23134659230709076, -0.15190981328487396, -0.37972119450569153, 0.30629515647888184, 0.20800676941871643, -0.04423924535512924, -0.39680689573287964, -0.1380617320537567, -0.07668580859899521, 0.4508565068244934, -0.24334178864955902, -0.008572511374950409, 0.02310919016599655, -0.4934573173522949, -0.0227420162409544, 0.6486945152282715, 0.041839443147182465, -0.011994633823633194, -0.002554632956162095, -0.23035380244255066, -0.08692970126867294, -0.1043047159910202, 0.06293808668851852, 0.1495053917169571, 0.14896339178085327, 0.12328223139047623, 0.10730747878551483, 0.024612823501229286, -0.26230356097221375, -0.0393705889582634, 0.01235000416636467, -0.32854604721069336, -0.13844531774520874, 0.13970713317394257, 0.24745799601078033, -0.12764616310596466, 0.5036818981170654, -0.12741424143314362, 0.30862611532211304, -0.009251223877072334, -0.09967109560966492, 0.09143628925085068, -0.1582871377468109, -0.4034571647644043, -0.10639693588018417, 0.21177957952022552, 0.330198734998703, 0.09043706953525543, 0.4027232527732849, 0.5290737748146057, -0.26109346747398376, 0.021613966673612595, 0.014538898132741451, 0.4841372072696686, 0.022282354533672333, -0.33286720514297485, 0.4092327058315277, 0.3574821650981903, 0.2704082429409027, -0.1167818233370781, 0.371844083070755, -0.17706017196178436, -0.03748033568263054, -0.016465844586491585, -0.059155866503715515, 0.12640254199504852, -0.16889476776123047, -0.1565994769334793, 0.3634796738624573, 0.12413904070854187, -0.1976224035024643, 0.08903394639492035, -0.505884051322937, -0.012677952647209167, -0.18312644958496094, 0.1596045196056366, 0.532459020614624, -0.2871997654438019, -0.14317400753498077, -0.023124273866415024, 0.14544884860515594, 0.07705996185541153, 0.3667641580104828, -0.0832999125123024, -0.0956675261259079, -0.028542257845401764, 0.40206581354141235, 0.201906219124794, -0.2790745496749878, -0.28862300515174866, 0.18329250812530518, 0.3203340172767639, 0.3041921555995941, -0.12321694940328598, 0.13023123145103455, 0.020968083292245865, 0.22887691855430603, 0.10841477662324905, -0.16214565932750702, -0.17627978324890137, 0.42187467217445374, -0.059204909950494766, -0.13189473748207092, 0.07925207912921906, -0.03650522604584694, -0.10871604830026627, -0.2211264967918396, -0.2119847685098648, 0.15718524158000946, 0.3730956017971039, -0.1357681006193161, -0.15732967853546143, -0.19157466292381287, 0.14541801810264587, -0.08411446958780289, -0.0321691632270813, -0.18415378034114838, -0.6730770468711853, 0.03187301382422447, -0.05881858617067337, -0.013976769521832466, -0.0235596876591444, -0.2371213585138321, -0.33111828565597534, 0.15139466524124146, -0.35942012071609497, -0.06123209372162819, -0.06678393483161926, -0.011039543896913528, -0.09729611873626709, -0.2155429720878601, -0.2320961207151413, 0.05262337252497673, 0.48802927136421204, 0.11189039796590805, -0.011932128109037876, 0.1936451941728592, 0.1937323361635208, -0.12392794340848923, 0.06306754052639008, 0.3638550043106079, 0.23855482041835785, 0.18986521661281586, 0.015624863095581532, 0.21068798005580902, 0.533371090888977, -0.07160791754722595, -0.12715260684490204, 0.23659737408161163, 0.45395633578300476, -0.1962519735097885, 0.15113325417041779, 0.20681196451187134, 0.19265982508659363, -0.047839634120464325, -0.7790425419807434, 0.25326138734817505, 0.013714666478335857, 0.2599428594112396, -0.0932527557015419, 0.192877858877182, -0.07523883134126663, 0.3315354883670807, -0.05952507257461548, -0.3936673104763031, -0.2014770209789276, 0.12461196631193161, 0.1480497419834137, -0.029560692608356476, -0.4988957345485687, 0.32104891538619995, 0.37384921312332153, -0.052154842764139175, 0.2573852837085724, -0.057632993906736374, -0.10829749703407288, -0.15855813026428223, 0.37332460284233093, -0.3535538911819458, 0.2837998569011688, 0.06916766613721848, -0.08367971330881119, -0.12158432602882385, -0.042423348873853683, -0.03475892171263695, 0.027157699689269066, 0.06743305921554565, -0.07191852480173111, 0.13305765390396118, 0.1848292350769043, -0.22043867409229279, -0.08031614869832993, -0.005193554330617189, -0.08777395635843277, -0.08954968303442001, -0.20640714466571808, 0.09672369807958603, -0.5124542117118835, 0.10381167382001877, -0.28017592430114746, -0.5547009706497192, -0.3657357096672058, -0.2290549874305725, 0.2649710774421692, -0.0899188369512558, -0.1655491292476654, 0.05860481411218643, 0.11672332882881165, 0.09931167215108871, 0.19559474289417267, -0.006007936783134937, -0.08945012092590332, 0.04601364582777023, 0.101840540766716, -0.051306743174791336, -0.047684043645858765, -0.04785295948386192, 0.0376473031938076, 0.12882840633392334, -0.11319982260465622, -0.45335331559181213, -0.008423865772783756, 0.3111930191516876, -0.11082735657691956, 0.1549980789422989, 0.37735605239868164, 0.10291551053524017, -0.4873548150062561, 0.04879304766654968, 0.212269589304924, 0.2985992729663849, -0.003608193015679717, 0.11281466484069824, 0.30032771825790405, -0.02294725365936756, -0.012344110757112503, -0.19157715141773224, -0.13206127285957336, -0.0038980254903435707, 0.10018575936555862, 0.02695542760193348, 0.15248613059520721, -0.16705460846424103, 0.2679470181465149, -0.035998404026031494, -0.3355773389339447, -0.24153491854667664, -0.14004886150360107, 0.08696035295724869, 0.14672940969467163, -0.08437187224626541, -0.05225026607513428, -0.2838633358478546, 0.048275336623191833, -0.01886799745261669, 0.1145024374127388, 0.1831580102443695, -0.3538355529308319, -0.08702810108661652, -0.06288919597864151, 0.17463314533233643, 0.4848901629447937, 0.28579121828079224, 0.26472368836402893, 0.11941147595643997, -0.056173454970121384, -0.006128234323114157, 0.16391129791736603, 0.348511666059494, 0.04665423557162285, 0.0049122110940515995, -0.09494286775588989, 0.001006687874905765, 1.0838123559951782, 0.4843832552433014, 0.2481924444437027, -0.003679476911202073, -0.11817070096731186, -0.12630312144756317, -0.15671615302562714, -0.09003333747386932, 0.11810865998268127, -0.07367558032274246, 0.302700400352478, 0.46514374017715454, 0.2442464828491211, 0.012060086242854595, 0.05346597731113434, 0.3858359158039093, -0.16218261420726776, -0.45231735706329346, 0.39658859372138977, -0.07453886419534683, 0.14584742486476898, 0.018084555864334106, 0.06413067877292633, -0.1832006871700287, -0.12235745787620544, -0.07621177285909653, 0.04760630428791046, 0.3101426064968109, -0.19234801828861237, -0.04028236120939255, -0.2107444554567337, -0.25974375009536743, 0.2814568877220154, 0.5223882794380188, 0.12692907452583313, -0.0026260772719979286, -0.14970150589942932, 0.20839805901050568, 0.23716409504413605, 0.6356257796287537, 0.17852969467639923, -0.2098158299922943, -0.009058625437319279, -0.1218133419752121, -0.010507178492844105, -0.12758614122867584, -0.6342934370040894, 0.1838952898979187, -0.27258023619651794, 0.22634340822696686, -0.3975050449371338, -0.4662688970565796, -0.07813085615634918, 0.117588110268116, -0.1353406012058258, -0.10769933462142944, -0.05650561302900314, 0.4158494174480438, -0.1580483317375183, 0.32520991563796997, 0.3609260022640228, -0.14264905452728271, 0.17127570509910583, -0.06450602412223816, 0.2282080203294754, -0.05428919568657875, 0.3435719609260559, 0.33272475004196167, 0.13303744792938232, 0.09885817021131516, -0.36606478691101074, 0.22141118347644806, 0.0841439738869667, -0.009226199239492416, 0.32175832986831665, -0.5721675157546997, -0.2904271185398102, 0.149000346660614, -0.04958123713731766, 0.4967503845691681, 0.4423184096813202, 0.2168344408273697, -0.050980113446712494, 0.2642027735710144, 0.07970554381608963, -0.13998617231845856, 0.11154863238334656, 0.2770269811153412, 0.08728725463151932, -0.17061857879161835, -0.39830249547958374, 0.4744042158126831, -0.06491956859827042, -0.28020209074020386, 0.013690364547073841, 0.5781548619270325, -0.0878419354557991, -0.07902681082487106, 0.29351478815078735, 0.9808765053749084, -0.2313358187675476, 0.39982378482818604, 0.40193766355514526, 0.02625681459903717, 0.31943854689598083, -0.28214573860168457, 0.26689982414245605, -0.22960367798805237, 0.11477584391832352, -0.12683069705963135, -0.18965299427509308, 0.09106993675231934, 0.4854971468448639, -0.6954207420349121, 0.4039885103702545, -0.06320980191230774, 0.0795762911438942, 0.06719359010457993, -0.020808108150959015, 0.1637314260005951, -0.25643831491470337, -0.26492926478385925, -0.06672491878271103, -0.25867417454719543, 0.0004014130390714854, -0.1023564338684082, -0.06116136908531189, -0.00667207408696413, -0.1209411770105362, -0.39329269528388977, -0.014796514064073563, -0.45530569553375244, -0.17547588050365448, 0.2911156415939331, -0.15258751809597015, 0.3953569829463959, 0.43452155590057373, 0.0012726468266919255, 0.14556823670864105, -0.1255720555782318, -0.02871054783463478, -0.08400893956422806, 0.4058331549167633, -0.26161566376686096, 0.09170577675104141, 0.5665057301521301, -0.0066858017817139626, -0.34218037128448486, 0.15743745863437653, -0.08092464506626129, -0.10101349651813507, -0.43462684750556946, 0.18062691390514374, 0.6312643885612488, -0.5566824078559875, -0.2709469497203827, 0.10570517927408218, 0.11234124004840851, -0.041594114154577255, -0.14434149861335754, 0.39041265845298767, -0.03776337206363678, 0.5050050616264343, 0.15269359946250916, 0.07537253946065903, 0.15192094445228577, -0.024606989696621895, 0.044685233384370804, -0.22016745805740356, 0.35747721791267395, 0.29706278443336487, -0.11576630920171738, -0.10506857931613922, -0.057065267115831375, 0.2073894441127777, -0.19489657878875732, 0.3421354293823242, -0.21411198377609253, -0.22223205864429474, -0.09522449225187302, 0.30014458298683167, 0.18653275072574615, 0.1520577371120453, 0.04173377528786659, -0.03233853727579117, -0.1166558712720871, 0.3286188542842865, -0.14338837563991547, 0.25442105531692505, 0.19086456298828125, -0.03195086494088173, 0.02839503064751625, 0.19577835500240326, -0.12895800173282623, -0.08749938756227493, 0.008276225067675114, 0.3492433726787567, -0.11323009431362152, -0.2655158042907715, -0.029820697382092476, 0.00357306026853621, -0.03897557780146599, -0.09551640599966049, -0.2530178427696228, -0.068989597260952, -0.2502743899822235, 0.1610957831144333, 0.047943390905857086, -0.12595920264720917, -0.026538213714957237, 0.031260862946510315, -0.023183219134807587, -0.23883187770843506, 0.1918046772480011, -0.08689844608306885, -0.14084844291210175, 0.3501245975494385, 0.007579805329442024, -0.21584546566009521, -0.13043969869613647, 0.027108220383524895, 0.18230821192264557, -0.03597751632332802, 0.33490604162216187, 0.15295879542827606, -0.035476185381412506, -0.23685018718242645, 0.03910716250538826, 0.4817050099372864, 0.15812689065933228, -0.013920990750193596, 0.07354775071144104, 0.08150321245193481, 0.07890559732913971, 0.017839733511209488, 0.28600430488586426, 0.00811080727726221, 0.07826036214828491, 0.04665002226829529, 0.1451670378446579, -0.006222646217793226, -0.10724986344575882, -0.29134538769721985, 0.45848003029823303, 0.10486231744289398, 0.20024646818637848, 0.40678226947784424, 0.01729259453713894, -0.003219203557819128, 0.3959062099456787, -0.053129296749830246, -0.27697813510894775, 0.06952805817127228, 0.011238755658268929, 0.6116737127304077, -0.16785180568695068, 0.12471392750740051, 0.07526886463165283, 0.21174606680870056, 0.17982585728168488, 0.03575972840189934, -0.18006381392478943, 0.2081255465745926, -0.13592562079429626, -0.15321308374404907, 0.5411537289619446, -0.13623937964439392, 0.01491106953471899, -0.006787335034459829, 0.021940764039754868, 0.45900511741638184, 0.23120072484016418, 0.2578004002571106, -0.3163489103317261, -0.35048866271972656, 0.0007420003530569375, 0.1056131049990654, -0.11325490474700928, -0.19692203402519226, 0.2010507583618164, -0.11095364391803741, -0.30959588289260864, -0.06364286690950394, -0.5783088207244873, -0.32991817593574524, 0.48669734597206116, 0.22484973073005676, 0.5121279358863831, -0.16760624945163727, -0.11800158768892288, 0.14495760202407837, 0.3679756820201874, -0.1525886058807373, 0.15408657491207123, 0.30220603942871094, -0.039084095507860184, 0.01862221769988537, 0.2026042640209198, 0.08953460305929184, 0.13583548367023468, -0.20486107468605042, -0.014349549077451229, -0.30121245980262756, -0.02742975763976574, 0.18805700540542603, -0.15484294295310974, 0.18413937091827393, 0.026714030653238297, 0.22319282591342926, -0.060947488993406296, 0.06802745908498764, -0.11772597581148148, 0.031708210706710815, 0.1584732085466385, -0.18284806609153748, -0.3853939175605774, -0.013145849108695984, -0.1711117923259735, 0.04769676923751831, -0.0006280641537159681, -0.7032930850982666, -0.06327446550130844, 0.07265041023492813, -0.06937984377145767, -0.0340653657913208, 0.0992378443479538, -0.04312045872211456, 0.20425190031528473, 0.1544470638036728, 0.24497507512569427, 0.14178621768951416, -0.27186909317970276, 0.12096352875232697, -0.27578336000442505, -0.1867605447769165, 0.2845391035079956, -0.2318635731935501, 0.03978758305311203, 0.02078201062977314, 0.10187926888465881, 0.341089129447937, 0.06822308152914047, -0.5816969871520996, -0.11385758966207504, 0.46572351455688477, -0.20126909017562866, 0.13621586561203003, 0.1844775676727295, 0.2668825387954712, 0.10629098117351532, -0.12843134999275208, 0.21247491240501404, 0.35382941365242004, -0.21196241676807404, -0.17079578340053558, -0.07130260765552521, -0.3326975703239441, -0.05884705111384392, -0.24034403264522552, 0.529384195804596, 0.2018613964319229, 0.12241364270448685, -0.268033891916275, 0.14065052568912506, 0.28605446219444275, 0.0567031092941761, -0.028977951034903526, 0.027318008244037628, 0.7622186541557312, -0.09328239411115646, 0.10506969690322876, -0.11119573563337326, 0.21598055958747864, 0.12799881398677826, -0.32997480034828186, -0.49485668540000916, 0.0894031822681427, -0.025288499891757965, -0.23564669489860535, 0.1079355850815773, 0.4834580719470978, -0.1225612610578537, 0.1367385983467102, -0.26708266139030457, 0.12856663763523102, 0.3199918270111084, -0.2631487250328064, -0.3871840238571167, -0.3999743163585663, -0.07116056233644485, -0.035582587122917175, -0.03292044624686241, -0.8910267949104309, -0.17213274538516998, 0.11249134689569473, -0.03815877065062523, -0.253605455160141, 0.08522715419530869, -0.022356748580932617, -0.27161073684692383, -0.37771672010421753, 0.6619979739189148, 0.056897856295108795, 0.03396184742450714, -0.22671622037887573, -0.5398954153060913 ]
https://github.com/huggingface/datasets/issues/188
When will the remaining math_dataset modules be added as dataset objects
Hi @tylerroost, we don't have a timeline for this at the moment. If you want to give it a look we would be happy to review a PR on it. Also, the library is one week old so everything is quite barebones, in particular the doc. You should expect some bumps on the road. To get you started, you can check the datasets scripts in the `./datasets` folder on the repo and find the one on math_datasets that will need to be modified. Then you should check the original repository on the math_dataset to see where the other files to download are located and what is the expected format for the various parts of the dataset. To get a general overview on how datasets scripts are written and used, you can read the nice tutorial on how to add a new dataset for TensorFlow Dataset [here](https://www.tensorflow.org/datasets/add_dataset), our API is not exactly identical but it can give you a high-level overview.
Currently only the algebra_linear_1d is supported. Is there a timeline for making the other modules supported. If no timeline is established, how can I help?
160
When will the remaining math_dataset modules be added as dataset objects Currently only the algebra_linear_1d is supported. Is there a timeline for making the other modules supported. If no timeline is established, how can I help? Hi @tylerroost, we don't have a timeline for this at the moment. If you want to give it a look we would be happy to review a PR on it. Also, the library is one week old so everything is quite barebones, in particular the doc. You should expect some bumps on the road. To get you started, you can check the datasets scripts in the `./datasets` folder on the repo and find the one on math_datasets that will need to be modified. Then you should check the original repository on the math_dataset to see where the other files to download are located and what is the expected format for the various parts of the dataset. To get a general overview on how datasets scripts are written and used, you can read the nice tutorial on how to add a new dataset for TensorFlow Dataset [here](https://www.tensorflow.org/datasets/add_dataset), our API is not exactly identical but it can give you a high-level overview.
[ -0.2745405435562134, -0.07471401244401932, -0.2828439772129059, 0.08163303881883621, 0.10131561011075974, 0.08511606603860855, 0.16227023303508759, 0.1815713793039322, 0.15496547520160675, 0.02899930812418461, 0.09069553017616272, 0.25426122546195984, -0.24301400780677795, 0.09133534878492355, 0.2556447982788086, -0.34201574325561523, 0.20134373009204865, -0.0719502717256546, -0.2481055110692978, -0.35922670364379883, -0.10869918763637543, 0.09392420202493668, -0.2666562795639038, -0.21100160479545593, -0.13362126052379608, -0.060057032853364944, -0.24860870838165283, -0.04428121820092201, -0.4253517687320709, -0.23080044984817505, 0.06616618484258652, 0.29805609583854675, 0.231758251786232, 0.35712137818336487, -0.00010253640357404947, -0.17914707958698273, 0.19525396823883057, 0.06595984846353531, -0.09675760567188263, -0.35875776410102844, -0.3629665970802307, -0.2668958604335785, 0.178465336561203, 0.1084667518734932, 0.0992727056145668, 0.09038355946540833, -0.039133451879024506, -0.40964093804359436, 0.02792554721236229, 0.35471293330192566, 0.23869970440864563, 0.35676470398902893, 0.41479402780532837, -0.1960519254207611, 0.23967798054218292, 0.06696385890245438, -0.18212394416332245, 0.08796816319227219, 0.4211680293083191, 0.11522797495126724, 0.3046659827232361, 0.14688746631145477, 0.1961059272289276, 0.10997598618268967, 0.3800205588340759, 0.018322696909308434, 0.013787508010864258, -0.5149787664413452, -0.3262500762939453, 0.2852602005004883, 0.7866033315658569, -0.2344970554113388, -0.23747482895851135, -0.1045815721154213, -0.03371284529566765, -0.2809472382068634, 0.036547791212797165, 0.03145712614059448, 0.21467287838459015, -0.16369548439979553, 0.06738559901714325, -0.403988778591156, -0.3353688716888428, 0.40255048871040344, 0.13977473974227905, 0.48645955324172974, 0.12364166975021362, 0.12610092759132385, 0.09881304949522018, -0.14917470514774323, 0.39519426226615906, -0.003663887968286872, 0.2163075953722, 0.2663557231426239, -0.2195390909910202, -0.2165071815252304, -0.09907487034797668, -0.0851171612739563, 0.018533574417233467, -0.18484316766262054, -0.14839690923690796, 0.3597193658351898, -0.3094291388988495, 0.25821036100387573, 0.2134397327899933, -0.15015865862369537, -0.1043488085269928, -0.004754893947392702, 0.42633056640625, -0.033695537596940994, 0.07132341712713242, 0.19717846810817719, -0.1961677223443985, 0.00761146005243063, -0.2757907509803772, 0.027275199070572853, 0.029524553567171097, -0.012821726500988007, 0.1225251704454422, -0.383207231760025, 0.09562388807535172, -0.3619908094406128, -0.04778074100613594, -0.08453338593244553, -0.3333868086338043, 0.17141391336917877, -0.009539090096950531, 0.027227310463786125, -0.1498737633228302, -0.02315676398575306, -0.07249989360570908, 0.15157656371593475, -0.2152208685874939, 0.01989307813346386, 0.2334243506193161, -0.10189913958311081, 0.08815018087625504, 0.012576744891703129, 0.20164798200130463, 0.15531906485557556, 0.18760822713375092, 0.10861674696207047, -0.13326986134052277, 0.2291128933429718, -0.169180229306221, 0.038062408566474915, -0.2693697214126587, 0.18624190986156464, -0.09615359455347061, -0.05022178962826729, -0.14967146515846252, -0.21687643229961395, -0.16860410571098328, 0.26260432600975037, -0.14511269330978394, -0.35408103466033936, -0.2239726334810257, 0.3776286840438843, -0.28075411915779114, 0.20390303432941437, -0.17724832892417908, -0.1275354027748108, -0.09541007876396179, -0.33919617533683777, 0.005437128711491823, 0.0747961550951004, -0.4191736578941345, 0.1555749475955963, -0.132332906126976, -0.05833638459444046, -0.23669791221618652, -0.054188910871744156, -0.1949496567249298, 0.01520797424018383, -0.09529175609350204, -0.12077126652002335, 0.4563644528388977, -0.5523672699928284, -0.06465864926576614, -0.16999676823616028, 0.22772066295146942, -0.22985900938510895, -0.1641165316104889, 0.1930217295885086, 0.039701007306575775, -0.14828282594680786, -0.20391355454921722, 0.3211022913455963, -0.1425628811120987, -0.2917490303516388, -0.2070937603712082, -0.11056303232908249, 0.05193554237484932, 0.11169528216123581, 0.36458057165145874, 0.017507178708910942, 0.16154691576957703, 0.21329595148563385, 0.014443299733102322, 0.10343026369810104, 0.03156650811433792, 0.2074110060930252, 0.41448670625686646, 0.21465077996253967, 0.4653908908367157, -0.42616915702819824, -0.4392792880535126, 0.13807770609855652, 0.1953706592321396, 0.18815839290618896, 0.10498832166194916, 0.08624697476625443, -0.004375908058136702, 0.10294652730226517, -0.12055452167987823, 0.31788066029548645, 0.1447257548570633, -0.0405566580593586, 0.21057097613811493, -0.14622442424297333, -0.5968188047409058, 0.09261253476142883, -0.1041228398680687, 0.4140123128890991, -0.4701017141342163, 0.30474191904067993, 0.015342125669121742, -0.1535005271434784, 0.16933943331241608, 0.2441733330488205, 0.0057321577332913876, -0.4595026969909668, 0.08033004403114319, 0.29646867513656616, 0.16137093305587769, 0.2357257753610611, -0.033806558698415756, 0.4449405372142792, 0.42598724365234375, -0.06091896444559097, 0.31197893619537354, -0.31688207387924194, -0.02650781348347664, -0.05693719536066055, 0.010205555707216263, 0.21473072469234467, -0.03721524402499199, -0.09473754465579987, 0.2684358060359955, 0.016563696786761284, 0.16712172329425812, 0.26641368865966797, -0.042163532227277756, -0.21579477190971375, 0.030043987557291985, -0.008428025059401989, -0.021350113674998283, -0.13367794454097748, -0.29696124792099, 0.22897666692733765, 0.6034263372421265, -0.12870298326015472, 0.035603880882263184, 0.061941564083099365, -0.2499329000711441, -0.008935240097343922, 0.0916372686624527, 0.27020013332366943, 0.1916397511959076, 0.1497255563735962, 0.21871399879455566, -0.01896939054131508, -0.29128867387771606, -0.12581434845924377, 0.03649391606450081, 0.19965335726737976, 0.14305661618709564, -0.060219671577215195, -0.013700210489332676, 0.005899767857044935, -0.09947994351387024, -0.019223885610699654, -0.045814093202352524, 0.47150611877441406, 0.07942918688058853, -0.18269282579421997, -0.4099804759025574, -0.18739186227321625, -0.07155577093362808, -0.1772991418838501, -0.1551743745803833, -0.306515634059906, 0.032820556312799454, -0.01460027601569891, 0.19896581768989563, 0.01642337068915367, 0.19826900959014893, 0.0992315486073494, -0.26712265610694885, 0.3675658106803894, -0.27087822556495667, 0.002239764202386141, -0.2175016850233078, 0.190299391746521, 0.2565094232559204, -0.08611202985048294, 0.4182341992855072, -0.034009017050266266, 0.23329444229602814, -0.13853515684604645, -0.6403660178184509, 0.1763879358768463, -0.1495903581380844, -0.02689199522137642, -0.02025003917515278, -0.0834309533238411, 0.1799280196428299, 0.11929915845394135, 0.07279103994369507, -0.29184484481811523, -0.05149528756737709, -0.044087156653404236, -0.19708730280399323, -0.1112024337053299, -0.13188116252422333, -0.5581411123275757, -0.4593663215637207, -0.2616685926914215, 0.17642129957675934, 0.3262958228588104, 0.0808149054646492, 0.16245858371257782, 0.08413638919591904, 0.1643994152545929, -0.06314384937286377, 0.03947453945875168, 0.1127869188785553, -0.25843721628189087, 0.4067577123641968, -0.3071102797985077, -0.3843965530395508, 0.4961082339286804, -0.10421451926231384, 0.3492160141468048, 0.21272261440753937, -0.3054114878177643, -0.1986849308013916, -0.025309422984719276, -0.01599125750362873, 0.16639313101768494, -0.01960626244544983, 0.23422183096408844, -0.009280478581786156, 0.052656129002571106, -0.2524115741252899, -0.14401841163635254, -0.08615756779909134, 0.37696877121925354, 0.029445532709360123, 0.1434265673160553, 0.3347582519054413, -0.03322376310825348, 0.430533230304718, 0.036448683589696884, -0.24522405862808228, 0.05040501803159714, 0.16883574426174164, 0.10867717117071152, 0.011382132768630981, -0.050217874348163605, 0.07712594419717789, 0.03132681921124458, 0.2990891933441162, 0.04891548678278923, 0.12088093906641006, -0.10746316611766815, -0.3140336871147156, 0.3371356427669525, -0.17844299972057343, -0.0018074819818139076, -0.0018802665872499347, -0.25169384479522705, 0.32957369089126587, 0.08878378570079803, -0.1960984766483307, -0.2880052626132965, -0.14876827597618103, 0.15746653079986572, 0.3545743525028229, 0.24198244512081146, -0.12902607023715973, -0.5378569960594177, -0.18528170883655548, -0.17318914830684662, 0.41747575998306274, -0.10197805613279343, 0.1792781502008438, -0.19229303300380707, 0.1106681376695633, 0.11774308234453201, 0.0859810933470726, 0.18783043324947357, -0.47661423683166504, -0.5271236896514893, 0.27916473150253296, -0.13627296686172485, -0.23418816924095154, -0.024256816133856773, -0.31063762307167053, 0.1136726438999176, -0.029084060341119766, 0.448890745639801, 0.015480444766581059, -0.3676786422729492, 0.24308720231056213, 0.13085727393627167, 0.085910364985466, -0.14056412875652313, 0.26299428939819336, -0.010136082768440247, -0.11732517927885056, -0.26204484701156616, -0.105214424431324, -0.11233877390623093, -0.11672724783420563, -0.5168625712394714, 0.07123570889234543, -0.33272668719291687, 0.18476171791553497, -0.047048419713974, -0.056590307503938675, 0.27492955327033997, 0.03683128580451012, 0.3168193995952606, -0.14537112414836884, 0.4666672348976135, -0.011513283476233482, -0.38040393590927124, -0.1696537286043167, 0.1740654557943344, -0.091621994972229, 0.1956474334001541, 0.1504792720079422, -0.052981939166784286, 0.18283647298812866, -0.06081851199269295, -0.031034929677844048, -0.1255170702934265, 0.16413484513759613, 0.4193398952484131, 0.22790870070457458, -0.2852485775947571, -0.6071878671646118, 0.11472297459840775, 0.016152100637555122, -0.04859548062086105, 0.4207157492637634, -0.13392069935798645, -0.17450836300849915, 0.11098339408636093, 0.041056402027606964, 0.5863260626792908, -0.2631886899471283, -0.2927810549736023, 0.03474948927760124, -0.0862998366355896, 0.4732598066329956, -0.6111394762992859, 0.17423975467681885, -0.0342671237885952, 0.10870564728975296, -0.08511234074831009, -0.2548709511756897, 0.21094462275505066, 0.07227673381567001, -0.20136019587516785, 0.09954261034727097, 0.17709042131900787, 0.13808012008666992, -0.1942106932401657, 0.6597743630409241, -0.06312791258096695, -0.14037339389324188, -0.09899760782718658, 0.2516934275627136, -0.028307924047112465, 0.052181556820869446, -0.06026068702340126, -0.19137834012508392, -0.031877510249614716, 0.181133434176445, -0.2849520444869995, -0.1687602698802948, -0.22131694853305817, 0.008319268934428692, 0.09131795167922974, -0.04329517483711243, 0.3468939960002899, 0.09623108059167862, 0.5915399193763733, -0.03202419355511665, -0.2376021146774292, 0.27973881363868713, -0.26822778582572937, -0.11788764595985413, -0.12288801372051239, 0.017207173630595207, 0.37216323614120483, -0.296741783618927, -0.012082227505743504, -0.07735522091388702, -0.0002787299163173884, -0.24259790778160095, -0.36513471603393555, -0.0714225247502327, 0.329862117767334, 0.09708566963672638, 0.1923026293516159, 0.303346186876297, -0.09278234094381332, 0.007560431491583586, 0.24786615371704102, 0.4511789083480835, 0.014724775217473507, 0.40766412019729614, 0.04798441007733345, 0.0005298939649946988, -0.15134301781654358, -0.15054936707019806, 0.2708602249622345, -0.2877453863620758, 0.07257480919361115, 0.43702995777130127, -0.19491249322891235, -0.20974552631378174, -0.19884450733661652, 0.18380053341388702, 0.05712946876883507, 0.0781836062669754, 0.15595877170562744, 0.09627263247966766, 0.19285620748996735, 0.0999172031879425, -0.023509008809924126, -0.41114315390586853, -0.4108961522579193, -0.293220192193985, -0.22436898946762085, 0.2566511631011963, 0.049461688846349716, 0.2794497609138489, -0.19637519121170044, -0.07223746180534363, 0.17411211133003235, 0.14937108755111694, -0.3791826367378235, 0.05701463297009468, 0.1513359248638153, 0.03618870675563812, 0.22612474858760834, -0.07250051200389862, -0.036121685057878494, -0.07174035161733627, 0.09386365860700607, -0.02794797345995903, -0.4679449796676636, -0.18472188711166382, 0.042026497423648834, 0.10021454840898514, -0.04263713210821152, -0.22871805727481842, 0.1266714483499527, -0.31415069103240967, -0.282792866230011, -0.024909403175115585, -0.03976761922240257, -0.3050768971443176, -0.14212068915367126, 0.33117467164993286, 0.18199747800827026, 0.1702483594417572, 0.5582084059715271, 0.09514033794403076, 0.21245074272155762, 0.20485855638980865, 0.40218791365623474, 0.18952016532421112, 0.2601065933704376, -0.08211435377597809, -0.1481981873512268, 0.07816378772258759, -0.1382056474685669, -0.17945781350135803, 0.39828890562057495, -0.16925737261772156, 0.15359129011631012, 0.03235561028122902, 0.10144001245498657, 0.2642422020435333, 0.021073395386338234, 0.036964595317840576, 0.14142902195453644, 0.21310999989509583, -0.1344299018383026, -0.05589975416660309, 0.4027065336704254, 0.4979761242866516, 0.10187961906194687, 0.07686421275138855, 0.5620501637458801, 0.1583322286605835, 0.2849802076816559, 0.17854437232017517, 0.23345720767974854, -0.14217449724674225, 0.4086757302284241, -0.10305356234312057, -0.05373323708772659, 0.0734400525689125, 0.08991985023021698, 0.3814929723739624, 0.0946706086397171, 0.3289203941822052, 0.18405811488628387, 0.4397294819355011, 0.1460438370704651, -0.08007954061031342, 0.5982283353805542, -0.3191632628440857, 0.06604879349470139, 0.16271136701107025, 0.06078672036528587, 0.09383008629083633, 0.07741829752922058, 0.46906551718711853, -0.00939022097736597, -0.13898488879203796, -0.054034870117902756, -0.1167001947760582, -0.10878189653158188, -0.005930845160037279, -0.04140695556998253, -0.17655418813228607, -0.2063528299331665, -0.0955452173948288, 0.042455628514289856, -0.24463951587677002, 0.10841132700443268, -0.00678416108712554, 0.02507498860359192, -0.18302905559539795, 0.34995150566101074, -0.08808033168315887, -0.06958433240652084, 0.20820191502571106, 0.44161081314086914, -0.2254703938961029, -0.20558901131153107, -0.09230701625347137, -0.07757951319217682, -0.06999122351408005, 0.02039334736764431, -0.35975363850593567, 0.30949512124061584, 0.0517013780772686, -0.1929580420255661, 0.2077178806066513, 0.4019235372543335, 0.07114530354738235, 0.3664083778858185, -0.11097901314496994, 0.15951333940029144, -0.053555529564619064, -0.02282078191637993, -0.28334107995033264, -0.02114742062985897, -0.4163028299808502, -0.08433801680803299, 0.11024491488933563, -0.15187110006809235, -0.04534958302974701, -0.04589655250310898, -0.47221624851226807, -0.5669900178909302, 0.5651461482048035, -0.18957000970840454, -0.15351344645023346, -0.25356101989746094, 0.10814565420150757, -0.09533153474330902, 0.25117409229278564, 0.1214740052819252, 0.45501285791397095, 0.05822421610355377, -0.3774457275867462, -0.2873682379722595, 0.00567632494494319, 0.149158775806427, 0.2815350592136383, -0.11302047222852707, 0.19985315203666687, 0.24011628329753876, -0.03342727571725845, 0.43038779497146606, -0.2973402440547943, 0.09230455011129379, -0.004098258446902037, -0.327586829662323, 0.009855114854872227, -0.20289508998394012, 0.041549645364284515, 0.1104595884680748, 0.25907883048057556, 0.07082254439592361, -0.26882967352867126, -0.02574295736849308, -0.41214823722839355, -0.02335910126566887, 0.33951467275619507, 0.08142495155334473, 0.4212687313556671, 0.1432819664478302, 0.29541251063346863, 0.18000566959381104, -0.04555322974920273, -0.41668370366096497, -0.16399523615837097, -0.01975839212536812, 0.23975980281829834, -0.12995566427707672, 0.08060023933649063, -0.19504323601722717, 0.01444647554308176, -0.16572168469429016, 0.14214079082012177, -0.16518983244895935, -0.3549889326095581, 0.03663397580385208, 0.06656534224748611, -0.08965376764535904, 0.032776348292827606, 0.40789395570755005, 0.13496799767017365, 0.23357634246349335, -0.24015283584594727, -0.21379755437374115, 0.1258513331413269, 0.05555209890007973, -0.0777759775519371, -0.24822509288787842, -0.19380585849285126, -0.18507109582424164, 0.26290977001190186, 0.11359754204750061, -0.8757448196411133, -0.21726404130458832, 0.2978820204734802, -0.09050814807415009, 0.06773237884044647, 0.26117226481437683, 0.460812509059906, -0.011697954498231411, -0.19920367002487183, 0.087603859603405, 0.004510651342570782, -0.05850180238485336, -0.13027030229568481, -0.15369893610477448 ]
https://github.com/huggingface/datasets/issues/187
[Question] How to load wikipedia ? Beam runner ?
I have seen that somebody is hard working on easierly loadable wikipedia. #129 Maybe I should wait a few days for that version ?
When `nlp.load_dataset('wikipedia')`, I got * `WARNING:nlp.builder:Trying to generate a dataset using Apache Beam, yet no Beam Runner or PipelineOptions() has been provided. Please pass a nlp.DownloadConfig(beam_runner=...) object to the builder.download_and_prepare(download_config=...) method. Default values will be used.` * `AttributeError: 'NoneType' object has no attribute 'size'` Could somebody tell me what should I do ? # Env On Colab, ``` git clone https://github.com/huggingface/nlp cd nlp pip install -q . ``` ``` %pip install -q apache_beam mwparserfromhell -> ERROR: pydrive 1.3.1 has requirement oauth2client>=4.0.0, but you'll have oauth2client 3.0.0 which is incompatible. ERROR: google-api-python-client 1.7.12 has requirement httplib2<1dev,>=0.17.0, but you'll have httplib2 0.12.0 which is incompatible. ERROR: chainer 6.5.0 has requirement typing-extensions<=3.6.6, but you'll have typing-extensions 3.7.4.2 which is incompatible. ``` ``` pip install -q apache-beam[interactive] ERROR: google-colab 1.0.0 has requirement ipython~=5.5.0, but you'll have ipython 5.10.0 which is incompatible. ``` # The whole message ``` WARNING:nlp.builder:Trying to generate a dataset using Apache Beam, yet no Beam Runner or PipelineOptions() has been provided. Please pass a nlp.DownloadConfig(beam_runner=...) object to the builder.download_and_prepare(download_config=...) method. Default values will be used. Downloading and preparing dataset wikipedia/20200501.aa (download: Unknown size, generated: Unknown size, total: Unknown size) to /root/.cache/huggingface/datasets/wikipedia/20200501.aa/1.0.0... --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.DoFnRunner.process() 44 frames /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.PerWindowInvoker.invoke_process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.PerWindowInvoker._invoke_process_per_window() /usr/local/lib/python3.6/dist-packages/apache_beam/io/iobase.py in process(self, element, init_result) 1081 writer.write(e) -> 1082 return [window.TimestampedValue(writer.close(), timestamp.MAX_TIMESTAMP)] 1083 /usr/local/lib/python3.6/dist-packages/apache_beam/io/filebasedsink.py in close(self) 422 def close(self): --> 423 self.sink.close(self.temp_handle) 424 return self.temp_shard_path /usr/local/lib/python3.6/dist-packages/apache_beam/io/parquetio.py in close(self, writer) 537 if len(self._buffer[0]) > 0: --> 538 self._flush_buffer() 539 if self._record_batches_byte_size > 0: /usr/local/lib/python3.6/dist-packages/apache_beam/io/parquetio.py in _flush_buffer(self) 569 for b in x.buffers(): --> 570 size = size + b.size 571 self._record_batches_byte_size = self._record_batches_byte_size + size AttributeError: 'NoneType' object has no attribute 'size' During handling of the above exception, another exception occurred: AttributeError Traceback (most recent call last) <ipython-input-9-340aabccefff> in <module>() ----> 1 dset = nlp.load_dataset('wikipedia') /usr/local/lib/python3.6/dist-packages/nlp/load.py in load_dataset(path, name, version, data_dir, data_files, split, cache_dir, download_config, download_mode, ignore_verifications, save_infos, **config_kwargs) 518 download_mode=download_mode, 519 ignore_verifications=ignore_verifications, --> 520 save_infos=save_infos, 521 ) 522 /usr/local/lib/python3.6/dist-packages/nlp/builder.py in download_and_prepare(self, download_config, download_mode, ignore_verifications, save_infos, dl_manager, **download_and_prepare_kwargs) 370 verify_infos = not save_infos and not ignore_verifications 371 self._download_and_prepare( --> 372 dl_manager=dl_manager, verify_infos=verify_infos, **download_and_prepare_kwargs 373 ) 374 # Sync info /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _download_and_prepare(self, dl_manager, verify_infos) 770 with beam.Pipeline(runner=beam_runner, options=beam_options,) as pipeline: 771 super(BeamBasedBuilder, self)._download_and_prepare( --> 772 dl_manager, pipeline=pipeline, verify_infos=False 773 ) # TODO{beam} verify infos 774 /usr/local/lib/python3.6/dist-packages/apache_beam/pipeline.py in __exit__(self, exc_type, exc_val, exc_tb) 501 def __exit__(self, exc_type, exc_val, exc_tb): 502 if not exc_type: --> 503 self.run().wait_until_finish() 504 505 def visit(self, visitor): /usr/local/lib/python3.6/dist-packages/apache_beam/pipeline.py in run(self, test_runner_api) 481 return Pipeline.from_runner_api( 482 self.to_runner_api(use_fake_coders=True), self.runner, --> 483 self._options).run(False) 484 485 if self._options.view_as(TypeOptions).runtime_type_check: /usr/local/lib/python3.6/dist-packages/apache_beam/pipeline.py in run(self, test_runner_api) 494 finally: 495 shutil.rmtree(tmpdir) --> 496 return self.runner.run_pipeline(self, self._options) 497 498 def __enter__(self): /usr/local/lib/python3.6/dist-packages/apache_beam/runners/direct/direct_runner.py in run_pipeline(self, pipeline, options) 128 runner = BundleBasedDirectRunner() 129 --> 130 return runner.run_pipeline(pipeline, options) 131 132 /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in run_pipeline(self, pipeline, options) 553 554 self._latest_run_result = self.run_via_runner_api( --> 555 pipeline.to_runner_api(default_environment=self._default_environment)) 556 return self._latest_run_result 557 /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in run_via_runner_api(self, pipeline_proto) 563 # TODO(pabloem, BEAM-7514): Create a watermark manager (that has access to 564 # the teststream (if any), and all the stages). --> 565 return self.run_stages(stage_context, stages) 566 567 @contextlib.contextmanager /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in run_stages(self, stage_context, stages) 704 stage, 705 pcoll_buffers, --> 706 stage_context.safe_coders) 707 metrics_by_stage[stage.name] = stage_results.process_bundle.metrics 708 monitoring_infos_by_stage[stage.name] = ( /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in _run_stage(self, worker_handler_factory, pipeline_components, stage, pcoll_buffers, safe_coders) 1071 cache_token_generator=cache_token_generator) 1072 -> 1073 result, splits = bundle_manager.process_bundle(data_input, data_output) 1074 1075 def input_for(transform_id, input_id): /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in process_bundle(self, inputs, expected_outputs) 2332 2333 with UnboundedThreadPoolExecutor() as executor: -> 2334 for result, split_result in executor.map(execute, part_inputs): 2335 2336 split_result_list += split_result /usr/lib/python3.6/concurrent/futures/_base.py in result_iterator() 584 # Careful not to keep a reference to the popped future 585 if timeout is None: --> 586 yield fs.pop().result() 587 else: 588 yield fs.pop().result(end_time - time.monotonic()) /usr/lib/python3.6/concurrent/futures/_base.py in result(self, timeout) 430 raise CancelledError() 431 elif self._state == FINISHED: --> 432 return self.__get_result() 433 else: 434 raise TimeoutError() /usr/lib/python3.6/concurrent/futures/_base.py in __get_result(self) 382 def __get_result(self): 383 if self._exception: --> 384 raise self._exception 385 else: 386 return self._result /usr/local/lib/python3.6/dist-packages/apache_beam/utils/thread_pool_executor.py in run(self) 42 # If the future wasn't cancelled, then attempt to execute it. 43 try: ---> 44 self._future.set_result(self._fn(*self._fn_args, **self._fn_kwargs)) 45 except BaseException as exc: 46 # Even though Python 2 futures library has #set_exection(), /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in execute(part_map) 2329 self._registered, 2330 cache_token_generator=self._cache_token_generator) -> 2331 return bundle_manager.process_bundle(part_map, expected_outputs) 2332 2333 with UnboundedThreadPoolExecutor() as executor: /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in process_bundle(self, inputs, expected_outputs) 2243 process_bundle_descriptor_id=self._bundle_descriptor.id, 2244 cache_tokens=[next(self._cache_token_generator)])) -> 2245 result_future = self._worker_handler.control_conn.push(process_bundle_req) 2246 2247 split_results = [] # type: List[beam_fn_api_pb2.ProcessBundleSplitResponse] /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in push(self, request) 1557 self._uid_counter += 1 1558 request.instruction_id = 'control_%s' % self._uid_counter -> 1559 response = self.worker.do_instruction(request) 1560 return ControlFuture(request.instruction_id, response) 1561 /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/sdk_worker.py in do_instruction(self, request) 413 # E.g. if register is set, this will call self.register(request.register)) 414 return getattr(self, request_type)( --> 415 getattr(request, request_type), request.instruction_id) 416 else: 417 raise NotImplementedError /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/sdk_worker.py in process_bundle(self, request, instruction_id) 448 with self.maybe_profile(instruction_id): 449 delayed_applications, requests_finalization = ( --> 450 bundle_processor.process_bundle(instruction_id)) 451 monitoring_infos = bundle_processor.monitoring_infos() 452 monitoring_infos.extend(self.state_cache_metrics_fn()) /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/bundle_processor.py in process_bundle(self, instruction_id) 837 for data in data_channel.input_elements(instruction_id, 838 expected_transforms): --> 839 input_op_by_transform_id[data.transform_id].process_encoded(data.data) 840 841 # Finish all operations. /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/bundle_processor.py in process_encoded(self, encoded_windowed_values) 214 decoded_value = self.windowed_coder_impl.decode_from_stream( 215 input_stream, True) --> 216 self.output(decoded_value) 217 218 def try_split(self, fraction_of_remainder, total_buffer_size): /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/operations.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.worker.operations.Operation.output() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/operations.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.worker.operations.Operation.output() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/operations.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.worker.operations.SingletonConsumerSet.receive() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/operations.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.worker.operations.DoOperation.process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/operations.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.worker.operations.DoOperation.process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.DoFnRunner.process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.DoFnRunner._reraise_augmented() /usr/local/lib/python3.6/dist-packages/future/utils/__init__.py in raise_with_traceback(exc, traceback) 417 if traceback == Ellipsis: 418 _, _, traceback = sys.exc_info() --> 419 raise exc.with_traceback(traceback) 420 421 else: /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.DoFnRunner.process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.PerWindowInvoker.invoke_process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.PerWindowInvoker._invoke_process_per_window() /usr/local/lib/python3.6/dist-packages/apache_beam/io/iobase.py in process(self, element, init_result) 1080 for e in bundle[1]: # values 1081 writer.write(e) -> 1082 return [window.TimestampedValue(writer.close(), timestamp.MAX_TIMESTAMP)] 1083 1084 /usr/local/lib/python3.6/dist-packages/apache_beam/io/filebasedsink.py in close(self) 421 422 def close(self): --> 423 self.sink.close(self.temp_handle) 424 return self.temp_shard_path /usr/local/lib/python3.6/dist-packages/apache_beam/io/parquetio.py in close(self, writer) 536 def close(self, writer): 537 if len(self._buffer[0]) > 0: --> 538 self._flush_buffer() 539 if self._record_batches_byte_size > 0: 540 self._write_batches(writer) /usr/local/lib/python3.6/dist-packages/apache_beam/io/parquetio.py in _flush_buffer(self) 568 for x in arrays: 569 for b in x.buffers(): --> 570 size = size + b.size 571 self._record_batches_byte_size = self._record_batches_byte_size + size AttributeError: 'NoneType' object has no attribute 'size' [while running 'train/Save to parquet/Write/WriteImpl/WriteBundles'] ```
24
[Question] How to load wikipedia ? Beam runner ? When `nlp.load_dataset('wikipedia')`, I got * `WARNING:nlp.builder:Trying to generate a dataset using Apache Beam, yet no Beam Runner or PipelineOptions() has been provided. Please pass a nlp.DownloadConfig(beam_runner=...) object to the builder.download_and_prepare(download_config=...) method. Default values will be used.` * `AttributeError: 'NoneType' object has no attribute 'size'` Could somebody tell me what should I do ? # Env On Colab, ``` git clone https://github.com/huggingface/nlp cd nlp pip install -q . ``` ``` %pip install -q apache_beam mwparserfromhell -> ERROR: pydrive 1.3.1 has requirement oauth2client>=4.0.0, but you'll have oauth2client 3.0.0 which is incompatible. ERROR: google-api-python-client 1.7.12 has requirement httplib2<1dev,>=0.17.0, but you'll have httplib2 0.12.0 which is incompatible. ERROR: chainer 6.5.0 has requirement typing-extensions<=3.6.6, but you'll have typing-extensions 3.7.4.2 which is incompatible. ``` ``` pip install -q apache-beam[interactive] ERROR: google-colab 1.0.0 has requirement ipython~=5.5.0, but you'll have ipython 5.10.0 which is incompatible. ``` # The whole message ``` WARNING:nlp.builder:Trying to generate a dataset using Apache Beam, yet no Beam Runner or PipelineOptions() has been provided. Please pass a nlp.DownloadConfig(beam_runner=...) object to the builder.download_and_prepare(download_config=...) method. Default values will be used. Downloading and preparing dataset wikipedia/20200501.aa (download: Unknown size, generated: Unknown size, total: Unknown size) to /root/.cache/huggingface/datasets/wikipedia/20200501.aa/1.0.0... --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.DoFnRunner.process() 44 frames /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.PerWindowInvoker.invoke_process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.PerWindowInvoker._invoke_process_per_window() /usr/local/lib/python3.6/dist-packages/apache_beam/io/iobase.py in process(self, element, init_result) 1081 writer.write(e) -> 1082 return [window.TimestampedValue(writer.close(), timestamp.MAX_TIMESTAMP)] 1083 /usr/local/lib/python3.6/dist-packages/apache_beam/io/filebasedsink.py in close(self) 422 def close(self): --> 423 self.sink.close(self.temp_handle) 424 return self.temp_shard_path /usr/local/lib/python3.6/dist-packages/apache_beam/io/parquetio.py in close(self, writer) 537 if len(self._buffer[0]) > 0: --> 538 self._flush_buffer() 539 if self._record_batches_byte_size > 0: /usr/local/lib/python3.6/dist-packages/apache_beam/io/parquetio.py in _flush_buffer(self) 569 for b in x.buffers(): --> 570 size = size + b.size 571 self._record_batches_byte_size = self._record_batches_byte_size + size AttributeError: 'NoneType' object has no attribute 'size' During handling of the above exception, another exception occurred: AttributeError Traceback (most recent call last) <ipython-input-9-340aabccefff> in <module>() ----> 1 dset = nlp.load_dataset('wikipedia') /usr/local/lib/python3.6/dist-packages/nlp/load.py in load_dataset(path, name, version, data_dir, data_files, split, cache_dir, download_config, download_mode, ignore_verifications, save_infos, **config_kwargs) 518 download_mode=download_mode, 519 ignore_verifications=ignore_verifications, --> 520 save_infos=save_infos, 521 ) 522 /usr/local/lib/python3.6/dist-packages/nlp/builder.py in download_and_prepare(self, download_config, download_mode, ignore_verifications, save_infos, dl_manager, **download_and_prepare_kwargs) 370 verify_infos = not save_infos and not ignore_verifications 371 self._download_and_prepare( --> 372 dl_manager=dl_manager, verify_infos=verify_infos, **download_and_prepare_kwargs 373 ) 374 # Sync info /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _download_and_prepare(self, dl_manager, verify_infos) 770 with beam.Pipeline(runner=beam_runner, options=beam_options,) as pipeline: 771 super(BeamBasedBuilder, self)._download_and_prepare( --> 772 dl_manager, pipeline=pipeline, verify_infos=False 773 ) # TODO{beam} verify infos 774 /usr/local/lib/python3.6/dist-packages/apache_beam/pipeline.py in __exit__(self, exc_type, exc_val, exc_tb) 501 def __exit__(self, exc_type, exc_val, exc_tb): 502 if not exc_type: --> 503 self.run().wait_until_finish() 504 505 def visit(self, visitor): /usr/local/lib/python3.6/dist-packages/apache_beam/pipeline.py in run(self, test_runner_api) 481 return Pipeline.from_runner_api( 482 self.to_runner_api(use_fake_coders=True), self.runner, --> 483 self._options).run(False) 484 485 if self._options.view_as(TypeOptions).runtime_type_check: /usr/local/lib/python3.6/dist-packages/apache_beam/pipeline.py in run(self, test_runner_api) 494 finally: 495 shutil.rmtree(tmpdir) --> 496 return self.runner.run_pipeline(self, self._options) 497 498 def __enter__(self): /usr/local/lib/python3.6/dist-packages/apache_beam/runners/direct/direct_runner.py in run_pipeline(self, pipeline, options) 128 runner = BundleBasedDirectRunner() 129 --> 130 return runner.run_pipeline(pipeline, options) 131 132 /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in run_pipeline(self, pipeline, options) 553 554 self._latest_run_result = self.run_via_runner_api( --> 555 pipeline.to_runner_api(default_environment=self._default_environment)) 556 return self._latest_run_result 557 /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in run_via_runner_api(self, pipeline_proto) 563 # TODO(pabloem, BEAM-7514): Create a watermark manager (that has access to 564 # the teststream (if any), and all the stages). --> 565 return self.run_stages(stage_context, stages) 566 567 @contextlib.contextmanager /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in run_stages(self, stage_context, stages) 704 stage, 705 pcoll_buffers, --> 706 stage_context.safe_coders) 707 metrics_by_stage[stage.name] = stage_results.process_bundle.metrics 708 monitoring_infos_by_stage[stage.name] = ( /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in _run_stage(self, worker_handler_factory, pipeline_components, stage, pcoll_buffers, safe_coders) 1071 cache_token_generator=cache_token_generator) 1072 -> 1073 result, splits = bundle_manager.process_bundle(data_input, data_output) 1074 1075 def input_for(transform_id, input_id): /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in process_bundle(self, inputs, expected_outputs) 2332 2333 with UnboundedThreadPoolExecutor() as executor: -> 2334 for result, split_result in executor.map(execute, part_inputs): 2335 2336 split_result_list += split_result /usr/lib/python3.6/concurrent/futures/_base.py in result_iterator() 584 # Careful not to keep a reference to the popped future 585 if timeout is None: --> 586 yield fs.pop().result() 587 else: 588 yield fs.pop().result(end_time - time.monotonic()) /usr/lib/python3.6/concurrent/futures/_base.py in result(self, timeout) 430 raise CancelledError() 431 elif self._state == FINISHED: --> 432 return self.__get_result() 433 else: 434 raise TimeoutError() /usr/lib/python3.6/concurrent/futures/_base.py in __get_result(self) 382 def __get_result(self): 383 if self._exception: --> 384 raise self._exception 385 else: 386 return self._result /usr/local/lib/python3.6/dist-packages/apache_beam/utils/thread_pool_executor.py in run(self) 42 # If the future wasn't cancelled, then attempt to execute it. 43 try: ---> 44 self._future.set_result(self._fn(*self._fn_args, **self._fn_kwargs)) 45 except BaseException as exc: 46 # Even though Python 2 futures library has #set_exection(), /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in execute(part_map) 2329 self._registered, 2330 cache_token_generator=self._cache_token_generator) -> 2331 return bundle_manager.process_bundle(part_map, expected_outputs) 2332 2333 with UnboundedThreadPoolExecutor() as executor: /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in process_bundle(self, inputs, expected_outputs) 2243 process_bundle_descriptor_id=self._bundle_descriptor.id, 2244 cache_tokens=[next(self._cache_token_generator)])) -> 2245 result_future = self._worker_handler.control_conn.push(process_bundle_req) 2246 2247 split_results = [] # type: List[beam_fn_api_pb2.ProcessBundleSplitResponse] /usr/local/lib/python3.6/dist-packages/apache_beam/runners/portability/fn_api_runner.py in push(self, request) 1557 self._uid_counter += 1 1558 request.instruction_id = 'control_%s' % self._uid_counter -> 1559 response = self.worker.do_instruction(request) 1560 return ControlFuture(request.instruction_id, response) 1561 /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/sdk_worker.py in do_instruction(self, request) 413 # E.g. if register is set, this will call self.register(request.register)) 414 return getattr(self, request_type)( --> 415 getattr(request, request_type), request.instruction_id) 416 else: 417 raise NotImplementedError /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/sdk_worker.py in process_bundle(self, request, instruction_id) 448 with self.maybe_profile(instruction_id): 449 delayed_applications, requests_finalization = ( --> 450 bundle_processor.process_bundle(instruction_id)) 451 monitoring_infos = bundle_processor.monitoring_infos() 452 monitoring_infos.extend(self.state_cache_metrics_fn()) /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/bundle_processor.py in process_bundle(self, instruction_id) 837 for data in data_channel.input_elements(instruction_id, 838 expected_transforms): --> 839 input_op_by_transform_id[data.transform_id].process_encoded(data.data) 840 841 # Finish all operations. /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/bundle_processor.py in process_encoded(self, encoded_windowed_values) 214 decoded_value = self.windowed_coder_impl.decode_from_stream( 215 input_stream, True) --> 216 self.output(decoded_value) 217 218 def try_split(self, fraction_of_remainder, total_buffer_size): /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/operations.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.worker.operations.Operation.output() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/operations.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.worker.operations.Operation.output() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/operations.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.worker.operations.SingletonConsumerSet.receive() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/operations.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.worker.operations.DoOperation.process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/worker/operations.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.worker.operations.DoOperation.process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.DoFnRunner.process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.DoFnRunner._reraise_augmented() /usr/local/lib/python3.6/dist-packages/future/utils/__init__.py in raise_with_traceback(exc, traceback) 417 if traceback == Ellipsis: 418 _, _, traceback = sys.exc_info() --> 419 raise exc.with_traceback(traceback) 420 421 else: /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.DoFnRunner.process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.PerWindowInvoker.invoke_process() /usr/local/lib/python3.6/dist-packages/apache_beam/runners/common.cpython-36m-x86_64-linux-gnu.so in apache_beam.runners.common.PerWindowInvoker._invoke_process_per_window() /usr/local/lib/python3.6/dist-packages/apache_beam/io/iobase.py in process(self, element, init_result) 1080 for e in bundle[1]: # values 1081 writer.write(e) -> 1082 return [window.TimestampedValue(writer.close(), timestamp.MAX_TIMESTAMP)] 1083 1084 /usr/local/lib/python3.6/dist-packages/apache_beam/io/filebasedsink.py in close(self) 421 422 def close(self): --> 423 self.sink.close(self.temp_handle) 424 return self.temp_shard_path /usr/local/lib/python3.6/dist-packages/apache_beam/io/parquetio.py in close(self, writer) 536 def close(self, writer): 537 if len(self._buffer[0]) > 0: --> 538 self._flush_buffer() 539 if self._record_batches_byte_size > 0: 540 self._write_batches(writer) /usr/local/lib/python3.6/dist-packages/apache_beam/io/parquetio.py in _flush_buffer(self) 568 for x in arrays: 569 for b in x.buffers(): --> 570 size = size + b.size 571 self._record_batches_byte_size = self._record_batches_byte_size + size AttributeError: 'NoneType' object has no attribute 'size' [while running 'train/Save to parquet/Write/WriteImpl/WriteBundles'] ``` I have seen that somebody is hard working on easierly loadable wikipedia. #129 Maybe I should wait a few days for that version ?
[ -0.151693657040596, 0.0031179734505712986, -0.01721995137631893, 0.3954123258590698, -0.06617990881204605, 0.05573397874832153, 0.04334845393896103, 0.33450591564178467, 0.34885478019714355, 0.11682835966348648, 0.3199314475059509, 0.17322218418121338, -0.1478527933359146, -0.08189928531646729, -0.01621277816593647, -0.4596615731716156, -0.05198840796947479, 0.24634312093257904, 0.036584511399269104, -0.05598796159029007, -0.2965039908885956, 0.26790305972099304, -0.2995697259902954, 0.10403278470039368, -0.062527135014534, -0.14766912162303925, 0.15097150206565857, 0.02395063266158104, -0.29176852107048035, -0.3756920397281647, 0.04113128408789635, -0.22728393971920013, 0.276458740234375, 0.149385005235672, -0.00011269282549619675, -0.03307882696390152, 0.6667638421058655, 0.013357669115066528, -0.44841110706329346, -0.3978945314884186, -0.13652655482292175, -0.40732458233833313, 0.45148614048957825, -0.27835771441459656, 0.010811317712068558, -0.18324361741542816, 0.31934037804603577, 0.17982664704322815, 0.41272208094596863, 0.29741474986076355, 0.21510018408298492, 0.013292597606778145, 0.31618088483810425, 0.02897258661687374, 0.43313151597976685, 0.023455947637557983, 0.03768175095319748, -0.0828801766037941, -0.13199615478515625, 0.03415637090802193, -0.12720993161201477, 0.15120497345924377, -0.2554621398448944, 0.08563674241304398, 0.3495388329029083, -0.19634704291820526, -0.07118009030818939, -0.5439115762710571, 0.13230374455451965, 0.14068518579006195, 0.6836255788803101, -0.0462401919066906, 0.16607220470905304, 0.09736326336860657, 0.11649153381586075, 0.0578923374414444, 0.23870915174484253, 0.32420703768730164, -0.42637380957603455, -0.307910293340683, -0.2179972529411316, -0.2496739625930786, -0.2927874028682709, 0.46477872133255005, 0.04294848442077637, 0.5524105429649353, -0.06267870962619781, 0.22764933109283447, -0.04183027893304825, -0.023902731016278267, -0.06886515021324158, -0.12929455935955048, 0.028790965676307678, 0.30135399103164673, 0.06049501523375511, 0.09357427060604095, 0.021501585841178894, 0.22951214015483856, 0.1968696415424347, -0.04341677203774452, -0.17343543469905853, -0.14541487395763397, 0.1463513821363449, 0.1465436816215515, 0.033378347754478455, 0.25672200322151184, 0.1367916613817215, -0.16623738408088684, 0.18984878063201904, 0.2154504805803299, -0.06594224274158478, 0.08795039355754852, -0.03179527074098587, -0.17879745364189148, -0.4661918878555298, 0.023530906066298485, 0.26094889640808105, -0.17632582783699036, -0.03626012057065964, 0.04170391336083412, -0.4259178042411804, -0.2416982799768448, -0.029363814741373062, 0.3228066861629486, -0.21127068996429443, 0.3525010347366333, 0.3740544617176056, 0.07673772424459457, -0.31700360774993896, -0.3271936774253845, 0.02015337347984314, 0.2933019995689392, -0.2867268919944763, 0.08495870232582092, 0.2305872142314911, 0.30473366379737854, 0.3511444330215454, -0.017778735607862473, -0.07731505483388901, 0.008209096267819405, 0.1638401299715042, -0.08172626048326492, -0.21706622838974, 0.23867292702198029, 0.3142967224121094, 0.27012449502944946, 0.05848691612482071, -0.332556813955307, -0.07982154190540314, 0.254853755235672, -0.33965858817100525, -0.07505931705236435, -0.0339420884847641, 0.16418908536434174, -0.1057850643992424, 0.01375927496701479, -0.3482522666454315, 0.24756571650505066, 0.030831562355160713, -0.28464385867118835, -0.17346782982349396, -0.16041061282157898, -0.08583085238933563, -0.35577037930488586, 0.2626804709434509, 0.335042804479599, 0.131585955619812, -0.1066659688949585, -0.0884825810790062, 0.16788513958454132, 0.2250353991985321, 0.056384842842817307, -0.12499355524778366, 0.47264978289604187, -0.04442116245627403, -0.07877182960510254, 0.7422001361846924, -0.13415741920471191, -0.17248478531837463, 0.23401010036468506, 0.13185381889343262, 0.019068745896220207, 0.06142447516322136, -0.06255600601434708, 0.12045765668153763, 0.08603455871343613, -0.051529474556446075, 0.5668156147003174, -0.011338646523654461, 0.03431916609406471, -0.40983110666275024, -0.25147753953933716, 0.17665830254554749, 0.06843876838684082, 0.38605496287345886, 0.1507289707660675, 0.010939043946564198, 0.5279049277305603, 0.0819583311676979, -0.03969411179423332, 0.026423240080475807, 0.31999671459198, -0.29706114530563354, -0.09037414193153381, -0.021785547956824303, 0.06996910274028778, -0.17513729631900787, -0.025476504117250443, -0.3948400616645813, 0.15085020661354065, 0.15849962830543518, 0.10617557168006897, -0.35405784845352173, -0.14297135174274445, -0.05245694890618324, -0.15489013493061066, 0.16197283565998077, 0.18550892174243927, 0.09486580640077591, 0.15069252252578735, 0.05413711816072464, 0.06682353466749191, -0.12752558290958405, -0.0409192331135273, -0.6120954751968384, 0.18550734221935272, -0.2736169099807739, -0.07680417597293854, 0.11377458274364471, 0.21021729707717896, 0.24915462732315063, 0.011265655048191547, -0.12344212830066681, 0.006875910330563784, -0.20115110278129578, -0.06263715028762817, -0.02723286859691143, -0.07050889730453491, 0.2022545486688614, -0.30023106932640076, 0.31451138854026794, 0.29330945014953613, 0.15260125696659088, -0.13237051665782928, 0.08773339539766312, -0.07689134776592255, 0.11841434240341187, 0.34047192335128784, 0.07403912395238876, 0.10327699035406113, 0.004244550596922636, -0.06271012127399445, -0.0017371755093336105, 0.22774799168109894, 0.3068399429321289, 0.4291991889476776, 0.09851831197738647, -0.10778782516717911, 0.0011931288754567504, -0.20937272906303406, 0.3699369430541992, 0.04473289102315903, 0.044053446501493454, 0.26717719435691833, -0.4355997145175934, -0.015432687476277351, 0.11218643933534622, -0.16287510097026825, 0.1669238954782486, 0.22491246461868286, -0.15550874173641205, -0.03987416252493858, -0.08139263838529587, -0.12739376723766327, 0.2003765106201172, 0.17848215997219086, 0.3696187436580658, 0.1193413957953453, 0.08972032368183136, 0.0779440701007843, -0.07833750545978546, -0.13671456277370453, -0.14669497311115265, 0.3170481026172638, -0.3040980100631714, 0.07823984324932098, -0.02340966835618019, -0.47671744227409363, -0.2430296540260315, 0.04412423074245453, -0.45233386754989624, -0.3893645703792572, -0.04385954514145851, 0.26631447672843933, 0.1046174094080925, 0.09521634876728058, 0.34473276138305664, 0.14382682740688324, -0.01657269522547722, -0.09474804997444153, -0.16188697516918182, -0.2388002574443817, -0.4626998007297516, 0.060326848179101944, 0.3643648326396942, 0.2716529369354248, 0.2625262439250946, -0.035812169313430786, 0.0011127227917313576, -0.016718665137887, -0.26169806718826294, 0.03566254675388336, -0.0869230329990387, 0.19745071232318878, -0.10585213452577591, 0.4281998574733734, -0.06742076575756073, -0.1935691088438034, 0.26940399408340454, -0.035429082810878754, -0.0946541279554367, 0.07878173142671585, -0.13136239349842072, 0.01989568956196308, -0.018720518797636032, -0.39383915066719055, -0.19605647027492523, -0.5722577571868896, 0.03628240153193474, 0.501702606678009, 0.15746314823627472, 0.1577187478542328, 0.22311848402023315, 0.07921379804611206, 0.3722139000892639, 0.3003678321838379, -0.07484975457191467, 0.08280817419290543, 0.3569963872432709, -0.24258437752723694, -0.45372599363327026, 0.006872965954244137, -0.253528356552124, 0.209784135222435, 0.1102369949221611, -0.5408006310462952, -0.06307720392942429, -0.036177199333906174, -0.09409040957689285, -0.07280664145946503, 0.09220873564481735, 0.402617484331131, -0.02519356645643711, 0.06426370143890381, 0.03355047479271889, -0.004459687974303961, -0.15063703060150146, -0.2979057729244232, 0.5397564172744751, 0.3464970290660858, 0.3857063353061676, -0.12387070804834366, 0.9763133525848389, 0.11937141418457031, -0.01083992701023817, 0.21248701214790344, 0.15789906680583954, 0.044725485146045685, -0.20025980472564697, -0.21275098621845245, 0.14212355017662048, -0.015238258056342602, -0.051309067755937576, 0.37091273069381714, -0.09053631126880646, -0.2564305365085602, -0.28977712988853455, -0.05816497281193733, -0.2621937394142151, -0.12282552570104599, 0.23912300169467926, 0.04297126457095146, 0.1968315690755844, -0.11754629015922546, 0.047273602336645126, -0.060971930623054504, -0.37045368552207947, 0.08413860201835632, 0.2635551989078522, -0.059675224125385284, 0.08931051194667816, 0.23530299961566925, -0.1747618019580841, -0.500149667263031, 0.4032238721847534, 0.029550744220614433, 0.09460243582725525, -0.23251847922801971, -0.02591860666871071, 0.17269589006900787, -0.043921180069446564, 0.22944073379039764, -0.17277593910694122, -0.1800784468650818, 0.3131050765514374, 0.2789316177368164, -0.49311912059783936, -0.007160308305174112, -0.22605569660663605, 0.2164933830499649, 0.3188696801662445, -0.06501603126525879, -0.44683459401130676, 0.01183990016579628, 0.35860759019851685, 0.26255926489830017, -0.16119292378425598, -0.06786514073610306, -0.4241851568222046, -0.10542388260364532, -0.3580247163772583, -0.16014021635055542, -0.1432640552520752, 0.37815916538238525, 0.014349212870001793, 0.08232944458723068, -0.051163963973522186, -0.075337253510952, 0.030097167938947678, 0.05655326694250107, 0.131614550948143, 0.15762104094028473, -0.3129267990589142, -0.17733922600746155, 0.25707873702049255, -0.3667420446872711, 0.20740579068660736, 0.3398808240890503, 0.02243984304368496, 0.009034660644829273, 0.16051527857780457, 0.10108721256256104, 0.12117061764001846, 0.038661688566207886, -0.09372314065694809, -0.0935019701719284, -0.020338034257292747, -0.10446936637163162, 0.048884227871894836, 0.20298327505588531, -0.18986760079860687, -0.19062060117721558, -0.18611255288124084, 0.9091110229492188, 0.11542623490095139, -0.0842638686299324, 0.39913392066955566, -0.10438898205757141, -0.3004433214664459, 0.4643958806991577, 0.292350172996521, 1.0337499380111694, -0.05213111266493797, 0.07410811632871628, 0.35504236817359924, 0.1453927904367447, 0.5975178480148315, -0.5634702444076538, 0.24681447446346283, -0.46771350502967834, 0.19664157927036285, -0.042075593024492264, -0.08887271583080292, 0.18185439705848694, -0.025461917743086815, -0.4229593575000763, 0.30799633264541626, 0.21473188698291779, 0.0531659834086895, -0.004552381578832865, 0.49724245071411133, -0.0075475783087313175, -0.2034151405096054, -0.20266906917095184, 0.11879844963550568, -0.4527357518672943, 0.4196886420249939, -0.31372296810150146, -0.07739096134901047, -0.03727024421095848, -0.2664511501789093, -0.32122600078582764, 0.23819111287593842, -0.2842641770839691, 0.2605055272579193, -0.07888054102659225, -0.40404394268989563, 0.13645227253437042, 0.20011760294437408, -0.07382778078317642, 0.22187964618206024, -0.36752083897590637, 0.30768585205078125, -0.3412059545516968, -0.3970908224582672, -0.16703733801841736, 0.30388787388801575, 0.3126794993877411, -0.20412328839302063, -0.09111820161342621, 0.26975324749946594, -0.15746261179447174, -0.1934327632188797, -0.11627263575792313, -0.04585758596658707, 0.40364253520965576, -0.08228084444999695, -0.28408196568489075, 0.0886298343539238, -0.11740061640739441, 0.10527415573596954, 0.12881609797477722, -0.15789073705673218, -0.02384245954453945, -0.11178715527057648, 0.4168771505355835, -0.12245702743530273, 0.1530793458223343, 0.21601298451423645, 0.35127541422843933, -0.0435100793838501, 0.6242594718933105, 0.24895085394382477, -0.12671321630477905, -0.19333410263061523, -0.08216233551502228, -0.3634004592895508, -0.22032591700553894, -0.13269227743148804, 0.061043016612529755, 0.026035446673631668, -0.16175736486911774, 0.3119335472583771, 0.04571117088198662, 0.14814680814743042, 0.08541948348283768, -0.47049397230148315, -0.08573392033576965, 0.36615797877311707, -0.21597418189048767, -0.026713917031884193, 0.12702395021915436, -0.11448001116514206, 0.46333301067352295, 0.13523519039154053, -0.25165560841560364, -0.19684088230133057, -0.2700384557247162, 0.20194405317306519, 0.33353719115257263, -0.03152146562933922, -0.09637937694787979, 0.18977418541908264, 0.03872344642877579, -0.18776541948318481, -0.19673025608062744, -0.1924375295639038, 0.07814862579107285, 0.16214889287948608, 0.20964877307415009, -0.058084242045879364, -0.10040754824876785, -0.2537476122379303, 0.04045746475458145, -0.17626875638961792, -0.1704917699098587, -0.28467410802841187, -0.16244164109230042, 0.07903806120157242, -0.06264567375183105, 0.28314584493637085, -0.39182183146476746, 0.21595941483974457, -0.15372933447360992, 0.4537641108036041, 0.11765002459287643, 0.04202544316649437, 0.5127456188201904, -0.05549119785428047, -0.742989182472229, 0.03110523149371147, 0.013880020938813686, 0.11196313798427582, 0.2513464391231537, -0.10180898755788803, 0.07591332495212555, -0.05008946731686592, -0.18661361932754517, 0.017427248880267143, -0.12645657360553741, 0.0703163743019104, 0.1133556067943573, 0.15643495321273804, -0.16165652871131897, 0.36316391825675964, 0.39460426568984985, 0.17278116941452026, -0.17091768980026245, 0.07866676151752472, 0.03425207361578941, -0.09170572459697723, -0.0037516900338232517, 0.22608576714992523, -0.005977886263281107, -0.05571694299578667, 0.02320878580212593, -0.12723250687122345, -0.025319501757621765, -0.013457526452839375, 0.01118869986385107, -0.020251518115401268, 0.1281440705060959, 0.11533665657043457, -0.14658227562904358, 0.18406178057193756, 0.05437787249684334, 0.006913813296705484, 0.015742264688014984, -0.18640635907649994, 0.23530153930187225, 0.15203320980072021, 0.37113532423973083, 0.04706316068768501, 0.049872931092977524, 0.04481019079685211, 0.11762654781341553, -0.14441314339637756, -0.40655189752578735, 0.18063049018383026, -0.12935365736484528, 0.12624388933181763, -0.4547828137874603, -0.07528630644083023, -0.28914448618888855, 0.14150774478912354, 0.10705230385065079, -0.27662789821624756, 0.05717267468571663, 0.2894270122051239, -0.3486204445362091, -0.40836644172668457, 0.43975427746772766, 0.14451606571674347, 0.0271508377045393, -0.1534154862165451, 0.36439886689186096, -0.15128156542778015, 0.01810949482023716, -0.08264714479446411, 0.3836381435394287, 0.3436252474784851, 0.5497422218322754, -0.4759595990180969, -0.05491875112056732, 0.055612847208976746, -0.15051960945129395, -0.07915592938661575, 0.11907359957695007, 0.41159501671791077, 0.23406964540481567, 0.18227386474609375, 0.13737614452838898, -0.23148714005947113, -0.03167314454913139, 0.18226882815361023, -0.10571838170289993, -0.08262036740779877, 0.16354778409004211, 0.0025952819269150496, -0.1768181473016739, -0.24999602138996124, -0.029853418469429016, -0.6789047122001648, 0.08470112085342407, -0.017861977219581604, 0.19374942779541016, -0.09923865646123886, -0.3172447681427002, 0.02822941169142723, -0.19291317462921143, 0.4531595706939697, 0.33714962005615234, 0.3244170546531677, -0.3690620958805084, -0.316669762134552, -0.3890509605407715, 0.15989811718463898, -0.3162235617637634, -0.32014793157577515, -0.0997476801276207, 0.23386456072330475, -0.15514890849590302, 0.3409184515476227, -0.23970557749271393, -0.0007632964407093823, 0.07648119330406189, 0.029247336089611053, -0.26590412855148315, -0.16083385050296783, 0.04977137967944145, -0.11939380317926407, -0.30618563294410706, -0.12973977625370026, -0.00499378889799118, -0.27574965357780457, 0.06248539313673973, -0.026223525404930115, -0.2599046528339386, 0.2788181006908417, 0.053318269550800323, 0.2739657163619995, 0.07670798152685165, 0.6266771554946899, -0.10283458977937698, -0.06840343773365021, -0.0048559242859482765, -0.06971412897109985, -0.06355009227991104, 0.369738906621933, 0.036819349974393845, 0.5085331797599792, -0.0811917632818222, 0.17832891643047333, -0.35964682698249817, 0.4930324852466583, -0.2634871006011963, 0.14966392517089844, -0.11382648348808289, -0.13014152646064758, -0.23576200008392334, 0.3155687749385834, -0.03986441716551781, 0.08727336674928665, -0.11603278666734695, 0.07417324930429459, -0.4722546935081482, -0.21190254390239716, 0.3164297938346863, -0.7915250062942505, -0.32634249329566956, 0.11869248002767563, 0.09361732751131058, -0.08433309197425842, 0.023439448326826096, -0.5125062465667725, -0.06366899609565735, 0.2648862600326538, -0.07161718606948853, -0.1045154258608818, 0.036423906683921814, -0.11945299804210663, -0.05821135640144348, -0.11786211282014847, -0.22919057309627533, -0.12916326522827148, -0.2262614518404007, -0.18604294955730438, -0.30095985531806946 ]
https://github.com/huggingface/datasets/issues/186
Weird-ish: Not creating unique caches for different phases
Looks like a duplicate of #120. This is already fixed on master. We'll do a new release on pypi soon
Sample code: ```python import nlp dataset = nlp.load_dataset('boolq') def func1(x): return x def func2(x): return None train_output = dataset["train"].map(func1) valid_output = dataset["validation"].map(func1) print() print(len(train_output), len(valid_output)) # Output: 9427 9427 ``` The map method in both cases seem to be pointing to the same cache, so the latter call based on the validation data will return the processed train data cache. What's weird is that the following doesn't seem to be an issue: ```python train_output = dataset["train"].map(func2) valid_output = dataset["validation"].map(func2) print() print(len(train_output), len(valid_output)) # 9427 3270 ```
20
Weird-ish: Not creating unique caches for different phases Sample code: ```python import nlp dataset = nlp.load_dataset('boolq') def func1(x): return x def func2(x): return None train_output = dataset["train"].map(func1) valid_output = dataset["validation"].map(func1) print() print(len(train_output), len(valid_output)) # Output: 9427 9427 ``` The map method in both cases seem to be pointing to the same cache, so the latter call based on the validation data will return the processed train data cache. What's weird is that the following doesn't seem to be an issue: ```python train_output = dataset["train"].map(func2) valid_output = dataset["validation"].map(func2) print() print(len(train_output), len(valid_output)) # 9427 3270 ``` Looks like a duplicate of #120. This is already fixed on master. We'll do a new release on pypi soon
[ 0.31512728333473206, -0.1353009045124054, -0.005951723549515009, 0.20902203023433685, 0.09835135191679001, -0.18703681230545044, 0.26089388132095337, 0.3269939720630646, -0.11203468590974808, -0.12795446813106537, -0.10739775747060776, 0.42777031660079956, 0.0819041058421135, -0.20389345288276672, 0.09390942007303238, 0.06840937584638596, 0.16564147174358368, 0.11475648730993271, -0.11452031135559082, -0.04316292703151703, -0.17623722553253174, 0.10245171934366226, -0.08972615003585815, -0.11245474219322205, -0.4616793990135193, 0.1460486650466919, -0.2929145097732544, -0.1581559032201767, 0.21726307272911072, -0.3920985460281372, 0.36049801111221313, 0.07840566337108612, -0.2556280791759491, 0.2644159495830536, -0.00011643479228951037, -0.039437536150217056, 0.18988557159900665, -0.060470204800367355, -0.09947331249713898, 0.15082624554634094, 0.020187849178910255, -0.14972253143787384, 0.16033527255058289, -0.3118847906589508, -0.016639070585370064, 0.11477650701999664, 0.05911470949649811, 0.1327991783618927, 0.5603551268577576, 0.06728798896074295, 0.19152870774269104, 0.2768820822238922, -0.30456361174583435, 0.19945140182971954, 0.1582948863506317, -0.09230409562587738, -0.19662749767303467, -0.07651348412036896, 0.01852801814675331, -0.27605998516082764, -0.03462275117635727, 0.3285544812679291, -0.09903604537248611, 0.09711853414773941, -0.01888144761323929, 0.09337884187698364, 0.16540858149528503, -0.19152803719043732, 0.08571892231702805, 0.06397523730993271, -0.12460657954216003, -0.32999730110168457, -0.2282964140176773, -0.3659476339817047, -0.22737930715084076, -0.3725854158401489, 0.13380685448646545, 0.12148881703615189, -0.1510031521320343, -0.11508475244045258, -0.2887340784072876, 0.25837573409080505, 0.18000300228595734, 0.06546581536531448, 0.1485069841146469, 0.34907928109169006, 0.17052233219146729, 0.003995365463197231, 0.20543481409549713, 0.16464833915233612, -0.022904587909579277, -0.1676708459854126, 0.0393318347632885, 0.38251474499702454, -0.32864564657211304, -0.13506928086280823, 0.2580425441265106, -0.037614598870277405, -0.020884132012724876, -0.035171277821063995, 0.29355984926223755, -0.05806310474872589, 0.03003242425620556, 0.03780967369675636, 0.14594827592372894, 0.4408576488494873, -0.05845896154642105, 0.26226186752319336, 0.03631458804011345, -0.11190104484558105, -0.4782034456729889, 0.07153692841529846, 0.3043484091758728, 0.0876687541604042, 0.3045758008956909, 0.05559048801660538, -0.07608956098556519, -0.09867508709430695, 0.042736250907182693, 0.0919157937169075, -0.36047303676605225, -0.1665520817041397, 0.04151659458875656, 0.12388278543949127, 0.010513640940189362, 0.1854729950428009, -0.2612519860267639, 0.04123657941818237, -0.5030214190483093, 0.1690860092639923, -0.3158501386642456, 0.011568352580070496, -0.4746894836425781, 0.3021249771118164, 0.24822424352169037, -0.20883476734161377, 0.2746181786060333, 0.3141807019710541, -0.2436193972826004, -0.2717684507369995, 0.29848557710647583, -0.2096518874168396, 0.3557654321193695, 0.09198487550020218, -0.11236432939767838, 0.20934966206550598, -0.052222371101379395, 0.17951080203056335, -0.20357491075992584, 0.10834082216024399, -0.10235651582479477, -0.18052832782268524, 0.34438708424568176, 0.16676528751850128, -0.2625376284122467, 0.15600068867206573, 0.3252079486846924, 0.21365152299404144, 0.7869164943695068, -0.38525721430778503, 0.1696179360151291, -0.07720289379358292, -0.26347315311431885, -0.3530147075653076, -0.03340945020318031, 0.2723200023174286, 0.1240132749080658, -0.3135716915130615, 0.16955463588237762, 0.23233555257320404, 0.0021237735636532307, 0.536086916923523, -0.16725347936153412, 0.15477752685546875, -0.3208006024360657, 0.03022231161594391, 0.42413291335105896, -0.29382672905921936, -0.46321913599967957, 0.12359775602817535, -0.010486197657883167, 0.1751442700624466, 0.11865287274122238, 0.1941884607076645, 0.11643951386213303, -0.2828816771507263, 0.20412351191043854, 0.04053018242120743, -0.10490675270557404, 0.09167338162660599, -0.37813296914100647, 0.014631930738687515, 0.36010071635246277, -0.24554404616355896, 0.00028567822300828993, 0.06188805028796196, -0.17130638659000397, -0.00008949417679104954, 0.26358166337013245, 0.0021526766940951347, 0.1317972093820572, 0.0214458666741848, -0.055076345801353455, -0.18047355115413666, 0.07052109390497208, -0.18271572887897491, 0.010328990407288074, 0.20688147842884064, -0.29167306423187256, -0.030908558517694473, 0.24033528566360474, -0.11940125375986099, 0.1090473011136055, -0.27650561928749084, -0.26970338821411133, -0.49192124605178833, 0.17864233255386353, 0.05129995942115784, 0.48496001958847046, -0.008755461312830448, -0.08701686561107635, 0.2997876703739166, 0.39124882221221924, -0.11880869418382645, -0.3205806016921997, 0.11166036874055862, -0.0021976525895297527, -0.16491961479187012, -0.3592391312122345, 0.5168808698654175, 0.2101716548204422, -0.008701741695404053, -0.12740252912044525, -0.04566918686032295, 0.005072710569947958, 0.08217803388834, -0.2594386637210846, 0.08567119389772415, 0.10755643248558044, 0.14894217252731323, 0.03267953544855118, 0.11418800055980682, -0.03314938023686409, -0.27372342348098755, 0.12674428522586823, 0.4829832911491394, 0.33497685194015503, 0.10381148755550385, -0.00017368566477671266, -0.08665511012077332, -0.06366544216871262, -0.1597079336643219, -0.22713622450828552, -0.34744197130203247, 0.25259506702423096, -0.03174927830696106, 0.42569899559020996, 0.27518168091773987, -0.04970972239971161, 0.27717718482017517, 0.5075845122337341, 0.1860629916191101, -0.022791942581534386, -0.09178189933300018, -0.015192250721156597, -0.26299044489860535, -0.09401454031467438, 0.21855512261390686, 0.5466564893722534, 0.09933163225650787, 0.14615938067436218, 0.18917647004127502, -0.32118386030197144, -0.20500116050243378, -0.029941141605377197, -0.12286622822284698, 0.06329682469367981, -0.042650241404771805, 0.372071236371994, -0.08435295522212982, -0.2066480666399002, 0.3466504216194153, 0.18771202862262726, -0.12854504585266113, 0.0377684161067009, 0.21131888031959534, -0.23826146125793457, -0.4322882294654846, -0.39855045080184937, -0.2013999968767166, -0.12059210240840912, -0.3052797317504883, 0.12759985029697418, 0.3070927858352661, 0.09445134550333023, 0.2894099950790405, -0.20783603191375732, 0.08653610199689865, -0.1703331172466278, -0.15077190101146698, 0.0021727054845541716, -0.3452710211277008, -0.31340834498405457, 0.03228013217449188, -0.016492845490574837, -0.27436086535453796, 0.2854226529598236, 0.018409553915262222, -0.17341504991054535, -0.14144840836524963, -0.39166557788848877, 0.20791587233543396, -0.008682504296302795, -0.20053057372570038, -0.1986219733953476, -0.3684796392917633, 0.09001265466213226, 0.012221753597259521, 0.3374743163585663, -0.36552634835243225, -0.2437644898891449, 0.055022940039634705, -0.08614606410264969, 0.10925415903329849, -0.27967268228530884, -0.10740357637405396, -0.14487729966640472, -0.04877825081348419, 0.20083190500736237, -0.2825668156147003, 0.2501550614833832, 0.3470873236656189, -0.25498172640800476, -0.09772118180990219, -0.15965349972248077, -0.16263151168823242, -0.46251705288887024, -0.03685716539621353, 0.21641822159290314, -0.31619638204574585, -0.18905670940876007, -0.38392335176467896, -0.34720781445503235, 0.26691120862960815, 0.17019592225551605, -0.4356415271759033, -0.43334972858428955, -0.06118972226977348, 0.07350410521030426, 0.17087793350219727, 0.049322787672281265, 0.38772571086883545, -0.17816771566867828, -0.018233617767691612, -0.15335536003112793, -0.3007083237171173, 0.2078815996646881, 0.3963834345340729, 0.35678863525390625, -0.1756996512413025, 0.22636643052101135, 0.13107091188430786, 0.5557454228401184, 0.6314631700515747, 0.06918271631002426, 0.37321674823760986, -0.013673664070665836, 0.2152680903673172, -0.26185542345046997, -0.19998620450496674, 0.1839577853679657, -0.15264496207237244, -0.2572272717952728, 0.1832878291606903, 0.1761987805366516, -0.5797125697135925, -0.0666341483592987, 0.16522343456745148, -0.2786339819431305, -0.3541928231716156, 0.09116538614034653, -0.31370463967323303, 0.22862091660499573, 0.3108310103416443, 0.23395323753356934, -0.4184882640838623, -0.1160450428724289, -0.12488027662038803, -0.30661094188690186, 0.145527184009552, 0.09425253421068192, -0.09625311940908432, -0.24743641912937164, -0.4561162292957306, 0.2061755210161209, 0.24228817224502563, 0.12783102691173553, 0.022665008902549744, -0.23083990812301636, 0.06999538838863373, 0.01940426044166088, 0.47368669509887695, -0.03395957127213478, -0.00395257817581296, 0.1012183353304863, -0.07110397517681122, -0.4034295678138733, -0.27929311990737915, 0.23190322518348694, 0.4614941477775574, 0.4238903820514679, 0.280887246131897, -0.2117939442396164, -0.107090525329113, -0.07216458022594452, 0.13284073770046234, -0.23876836895942688, -0.39052334427833557, -0.23795558512210846, -0.06291782855987549, 0.06605024635791779, 0.07129424065351486, -0.12635454535484314, -0.18954655528068542, -0.19810223579406738, -0.07180879265069962, 0.19534040987491608, -0.039148710668087006, 0.03154447302222252, 0.24644145369529724, 0.2672860622406006, -0.12621742486953735, 0.17920587956905365, 0.17825384438037872, -0.10764848440885544, -0.4886571764945984, 0.411167711019516, -0.18163643777370453, 0.3603590428829193, 0.18541871011257172, -0.05674765631556511, -0.02005329541862011, 0.4066057503223419, 0.19946962594985962, 0.14544092118740082, -0.07756051421165466, -0.07732496410608292, -0.026072409003973007, -0.2263960838317871, 0.3206343650817871, 0.0029677601996809244, -0.2382083386182785, -0.45535337924957275, 0.2363225370645523, 0.15792153775691986, -0.14968997240066528, 0.2283804714679718, -0.611373245716095, -0.11635436117649078, 0.39091721177101135, -0.19715014100074768, 0.7841211557388306, 0.02420222759246826, 0.47903725504875183, 0.22325576841831207, -0.05244666710495949, 0.3801923096179962, -0.04663291946053505, 0.363625705242157, -0.39555588364601135, 0.18653641641139984, 0.12171275168657303, -0.17433688044548035, -0.3330659866333008, 0.021303121000528336, 0.06636742502450943, 0.43155157566070557, 0.2073083370923996, 0.4841274321079254, 0.02028064616024494, -0.21137680113315582, 0.20552659034729004, 0.020139843225479126, -0.32142770290374756, 0.17260459065437317, 0.05149513855576515, 0.24724800884723663, -0.06316377967596054, -0.22587209939956665, -0.0719219222664833, -0.16542288661003113, -0.03188101202249527, 0.2956012189388275, -0.13948620855808258, 0.40801385045051575, 0.35625526309013367, -0.2011682391166687, -0.6638274192810059, 0.09609916806221008, 0.13754110038280487, -0.39595767855644226, -0.015468145720660686, -0.11183487623929977, 0.47419604659080505, 0.41549742221832275, -0.05152164399623871, -0.2530067563056946, 0.36641210317611694, -0.07084542512893677, -0.04830017685890198, -0.05500293895602226, -0.008599490858614445, -0.408769816160202, -0.2352360635995865, 0.3175221085548401, -0.3334675133228302, 0.050755083560943604, 0.1347249299287796, 0.1830376386642456, -0.07473701238632202, -0.11483446508646011, 0.11003774404525757, 0.07397695630788803, -0.12934580445289612, 0.3518597185611725, -0.35450470447540283, -0.10786081850528717, 0.03488926589488983, 0.28572145104408264, 0.2867709696292877, 0.08036472648382187, 0.36261263489723206, -0.2709967792034149, 0.06172782555222511, -0.036722809076309204, -0.33266276121139526, -0.31491509079933167, -0.32015693187713623, 0.10790388286113739, -0.25350451469421387, -0.18748392164707184, 0.10250499844551086, 0.007121643517166376, 0.12069142609834671, 0.37156033515930176, -0.08053586632013321, -0.18930572271347046, -0.5517708659172058, -0.0539889857172966, -0.10335443168878555, 0.16097642481327057, -0.017603009939193726, 0.6225496530532837, 0.07026710361242294, 0.3935421109199524, -0.28945863246917725, -0.020571837201714516, -0.015656696632504463, 0.22764624655246735, -0.13291803002357483, -0.4048028290271759, -0.05415891483426094, 0.14537984132766724, 0.04365763068199158, 0.33607983589172363, -0.48687249422073364, -0.2018185257911682, -0.16914953291416168, 0.1019672304391861, 0.14670836925506592, 0.1432407945394516, -0.2514541447162628, 0.13051357865333557, 0.3261987268924713, -0.34623983502388, 0.1161704882979393, 0.16987378895282745, 0.02248508483171463, 0.19154693186283112, 0.16412387788295746, 0.06718508899211884, 0.14418604969978333, 0.11707290261983871, -0.14140474796295166, -0.05994485318660736, 0.16280736029148102, 0.18861624598503113, -0.13985177874565125, 0.15012967586517334, -0.1367887705564499, -0.16258274018764496, -0.11337701231241226, -0.0793284997344017, -0.10075557231903076, -0.27143847942352295, -0.12925611436367035, 0.0730399563908577, 0.21172864735126495, 0.22056743502616882, -0.08022341877222061, -0.17486293613910675, -0.10348466038703918, 0.16781772673130035, 0.06923587620258331, 0.10698191821575165, 0.35733723640441895, 0.2053389847278595, 0.07064788043498993, 0.3598752021789551, 0.4193611443042755, -0.0002605749177746475, 0.468264102935791, 0.0346510075032711, -0.0801459327340126, -0.4715936481952667, 0.26750659942626953, 0.010570721700787544, 0.19505758583545685, 0.08407580852508545, 0.348564475774765, -0.28407397866249084, -0.07184179872274399, 0.06369445472955704, 0.2873610258102417, 0.160660520195961, -0.029453329741954803, -0.07150264084339142, 0.08202347904443741, 0.01956474967300892, -0.2752048075199127, 0.02920747548341751, -0.09377871453762054, 0.25051403045654297, 0.04363810271024704, 0.3715866506099701, -0.154639333486557, -0.18234418332576752, -0.42075127363204956, 0.3165487051010132, -0.28920936584472656, -0.2860289514064789, 0.2699081599712372, -0.300624817609787, 0.27491578459739685, 0.14489205181598663, -0.04462646692991257, 0.19261585175991058, 0.7117822766304016, 0.060089077800512314, -0.027329353615641594, -0.15319880843162537, -0.31560951471328735, 0.2999633848667145, 0.3962082862854004, -0.10624194145202637, 0.3474762737751007, 0.06683923304080963, -0.3460542559623718, -0.15965600311756134, 0.1851087510585785, 0.3952046036720276, 0.2746727764606476, -0.284920334815979, 0.16531194746494293, 0.3119156062602997, 0.0437200702726841, -0.14279311895370483, 0.28440603613853455, 0.464445024728775, -0.2753711938858032, 0.1001332625746727, 0.0773630440235138, -0.16198544204235077, 0.3374537527561188, 0.09224097430706024, 0.26263728737831116, -0.20471571385860443, 0.5761632919311523, -0.05505999177694321, -0.07937569916248322, 0.00896922592073679, 0.20302976667881012, -0.5066412687301636, -0.12212084233760834, 0.06195857748389244, 0.05762336403131485, 0.16316378116607666, -0.22196756303310394, 0.09016372263431549, 0.10871478915214539, 0.3901248872280121, 0.12524504959583282, 0.13043196499347687, -0.2856215238571167, -0.016745034605264664, -0.5698696970939636, 0.3329300284385681, -0.3890036344528198, 0.24580825865268707, -0.18367396295070648, 0.2748158276081085, -0.26683545112609863, 0.03942573070526123, 0.14697878062725067, -0.21078917384147644, 0.18485157191753387, 0.10188589990139008, -0.45321252942085266, 0.025243505835533142, 0.07228831946849823, -0.1955050826072693, 0.25490930676460266, -0.23168125748634338, 0.021740231662988663, -0.01908258907496929, 0.13105100393295288, 0.0037308516912162304, -0.0033552173990756273, 0.2672175467014313, 0.5061389207839966, 0.28892892599105835, 0.42512625455856323, 0.2001941055059433, 0.040501903742551804, -0.18851473927497864, -0.3905124068260193, 0.032601166516542435, -0.031582023948431015, -0.24561211466789246, -0.0693584680557251, 0.33418580889701843, -0.3125980794429779, -0.11305247992277145, -0.2842993140220642, 0.22645701467990875, 0.08093848079442978, -0.3933439254760742, 0.17809881269931793, -0.0364459753036499, -0.1294269561767578, 0.060526106506586075, 0.030793631449341774, 0.7254055738449097, 0.016145482659339905, 0.4419613480567932, -0.18536292016506195, -0.25712719559669495, 0.381931334733963, -0.6724092960357666, -0.26144281029701233, 0.0037971956189721823, 0.3030821979045868, 0.09342247247695923, -0.1773826628923416, -0.5309107899665833, -0.12437997758388519, 0.17774006724357605, -0.1675693690776825, -0.12121253460645676, 0.18059739470481873, -0.19370536506175995, -0.03503301367163658, -0.010602850466966629, 0.16422653198242188, 0.09443073719739914, -0.26849403977394104, 0.27279263734817505, -0.25110703706741333 ]
https://github.com/huggingface/datasets/issues/183
[Bug] labels of glue/ax are all -1
This is the test set given by the Glue benchmark. The labels are not provided, and therefore set to -1.
``` ax = nlp.load_dataset('glue', 'ax') for i in range(30): print(ax['test'][i]['label'], end=', ') ``` ``` -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, ```
20
[Bug] labels of glue/ax are all -1 ``` ax = nlp.load_dataset('glue', 'ax') for i in range(30): print(ax['test'][i]['label'], end=', ') ``` ``` -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, ``` This is the test set given by the Glue benchmark. The labels are not provided, and therefore set to -1.
[ 0.2502778470516205, -0.36407655477523804, -0.061697643250226974, 0.13931097090244293, -0.06844886392354965, -0.0784343034029007, 0.34460070729255676, 0.26304805278778076, 0.37012946605682373, 0.19107188284397125, -0.2715182602405548, 0.684173583984375, -0.06559257209300995, 0.11889583617448807, 0.18207645416259766, -0.11214101314544678, -0.10166774690151215, 0.3471296727657318, -0.15045088529586792, -0.1217602863907814, -0.26120510697364807, 0.3785853385925293, -0.2576518654823303, 0.14147385954856873, -0.17962245643138885, -0.10173988342285156, -0.15014229714870453, 0.07673332095146179, -0.03307449072599411, -0.40146902203559875, 0.10599248856306076, 0.14746470749378204, -0.10761843621730804, -0.12033428251743317, -0.00010571897291811183, -0.286956787109375, 0.2894553244113922, 0.07167448848485947, 0.032456304877996445, 0.12928707897663116, -0.14674852788448334, -0.2972312867641449, -0.03970690816640854, -0.3864680528640747, -0.057140301913022995, -0.0009060862939804792, 0.04592839628458023, -0.12836821377277374, 0.1665225327014923, 0.16479063034057617, 0.24084700644016266, 0.15829423069953918, 0.0880788043141365, 0.13617262244224548, 0.45657652616500854, -0.476816862821579, 0.04431513324379921, 0.2460300624370575, 0.09880738705396652, -0.3428689241409302, -0.016030726954340935, 0.45544907450675964, 0.1296474188566208, 0.1253586709499359, 0.03968087583780289, 0.14423783123493195, 0.03318944573402405, -0.35992252826690674, -0.08386991918087006, 0.4264651834964752, 0.13150835037231445, -0.3524520993232727, -0.3100077509880066, -0.20165713131427765, 0.21391217410564423, -0.3125109374523163, -0.06473931670188904, 0.14203506708145142, -0.01629771664738655, 0.009706767275929451, 0.037663187831640244, 0.10280227661132812, -0.12196788936853409, 0.1391846388578415, 0.13890403509140015, 0.5842685699462891, -0.020311454311013222, 0.06995396316051483, 0.369263619184494, -0.0291570033878088, -0.18502463400363922, 0.09218551963567734, -0.020079748705029488, 0.17599351704120636, -0.28078651428222656, -0.19722893834114075, 0.1095440685749054, 0.10077425837516785, 0.021075474098324776, 0.044597383588552475, 0.2250245064496994, 0.03807009756565094, 0.20733854174613953, 0.16420292854309082, 0.08044945448637009, 0.2856958508491516, 0.38481634855270386, 0.2837713956832886, 0.028997106477618217, 0.062447454780340195, -0.12825195491313934, 0.20890238881111145, -0.014154486358165741, -0.10458941757678986, 0.2733451724052429, -0.09539518505334854, -0.0748538076877594, -0.12386669218540192, -0.3194423019886017, -0.025794045999646187, -0.2851247191429138, 0.002068463247269392, -0.17965783178806305, 0.24445238709449768, 0.027863161638379097, 0.015347889624536037, -0.03464910015463829, 0.09597476571798325, -0.15768207609653473, -0.16153304278850555, -0.2969713509082794, 0.37302058935165405, -0.23291729390621185, -0.25091084837913513, 0.14201587438583374, 0.20792824029922485, 0.28334102034568787, 0.02848709747195244, -0.13069738447666168, 0.037280045449733734, 0.14972753822803497, -0.020442349836230278, 0.2775793671607971, 0.11669644713401794, -0.0686403214931488, 0.10494789481163025, -0.010465960018336773, -0.3798038959503174, -0.27661022543907166, -0.08015425503253937, 0.09214683622121811, 0.07429850101470947, 0.023174403235316277, 0.2386457473039627, -0.17580634355545044, 0.02703120745718479, 0.18054549396038055, 0.09859385341405869, -0.09635763615369797, 0.20633026957511902, 0.14499203860759735, -0.14029522240161896, -0.11530105024576187, -0.1897801011800766, 0.291601300239563, 0.2018367350101471, -0.30105817317962646, -0.11395082622766495, 0.08180438727140427, -0.03524620085954666, -0.1266513466835022, 0.0524621345102787, 0.22159940004348755, 0.005408497992902994, -0.26821544766426086, 0.22213934361934662, 0.29200905561447144, -0.5699794888496399, -0.46014702320098877, 0.02611774392426014, -0.0982159823179245, -0.11541435867547989, 0.08403782546520233, 0.07993191480636597, -0.02733754925429821, 0.2073451429605484, 0.3057434856891632, 0.27592071890830994, 0.14640164375305176, -0.014055615290999413, -0.3354366719722748, 0.14553871750831604, 0.38180625438690186, 0.1446693390607834, -0.1724829077720642, -0.16292215883731842, -0.2000306248664856, -0.1649155616760254, 0.2638298571109772, 0.11475209891796112, -0.08811508864164352, 0.26813000440597534, 0.012011575512588024, -0.35355862975120544, 0.1252511590719223, -0.146011620759964, -0.23448701202869415, 0.24857302010059357, -0.23344486951828003, 0.3903799057006836, 0.27938413619995117, -0.13543899357318878, -0.39104682207107544, -0.009539817459881306, -0.021590683609247208, -0.22261139750480652, 0.2764017581939697, 0.02215607464313507, 0.3353918790817261, 0.061245836317539215, -0.09688244760036469, 0.3609968423843384, 0.09802162647247314, 0.041608333587646484, -0.36970967054367065, 0.09687378257513046, -0.0590873658657074, -0.04003338888287544, 0.24726365506649017, 0.5104695558547974, 0.2441357523202896, -0.05340401455760002, 0.13028854131698608, 0.3582191467285156, -0.02366030029952526, 0.012895838357508183, 0.08359035849571228, -0.017381837591528893, -0.04296481981873512, -0.2957713007926941, 0.09626934677362442, 0.16723622381687164, 0.08046342432498932, -0.04020118713378906, -0.15581445395946503, 0.3673767149448395, 0.09383957833051682, 0.051863767206668854, 0.11298225075006485, 0.0980573445558548, -0.09670699387788773, -0.29650360345840454, -0.3145453631877899, -0.13945312798023224, 0.33484122157096863, -0.2254427820444107, 0.35534968972206116, 0.04052095115184784, -0.2502380311489105, 0.20897160470485687, 0.44670554995536804, 0.024678420275449753, 0.20379212498664856, -0.11887607723474503, -0.2853447496891022, -0.10063906013965607, 0.054819826036691666, 0.5218161940574646, 0.16740678250789642, 0.4208613336086273, 0.01765083149075508, -0.059704698622226715, -0.3939405381679535, -0.0734596997499466, 0.16773934662342072, -0.08496905863285065, 0.07969265431165695, -0.08924123644828796, 0.20133495330810547, -0.1482599377632141, -0.18963713943958282, 0.05562856048345566, -0.24803279340267181, 0.2343175858259201, -0.2628103196620941, -0.19544865190982819, -0.08484257012605667, -0.4971703886985779, -0.2795572876930237, -0.29548323154449463, -0.14680400490760803, -0.5272939205169678, 0.13580918312072754, -0.2413763552904129, -0.02068505249917507, 0.22529415786266327, 0.05436062812805176, 0.16112087666988373, -0.18304085731506348, 0.07843808829784393, -0.20593921840190887, -0.1355559527873993, -0.21072636544704437, 0.15992191433906555, -0.2740514278411865, 0.520328938961029, 0.39891332387924194, -0.09670893102884293, 0.07538869976997375, 0.319004625082016, -0.21436874568462372, -0.0005633768159896135, -0.17378319799900055, 0.2465425729751587, 0.4136919379234314, 0.0018736866768449545, -0.127711221575737, -0.012991977855563164, 0.24294660985469818, -0.43880903720855713, -0.0056831007823348045, -0.06721764802932739, 0.20725597441196442, -0.03913981840014458, -0.14822231233119965, -0.40482693910598755, -0.3648361563682556, -0.22120994329452515, 0.23954327404499054, 0.2358979731798172, 0.25962013006210327, 0.1151956245303154, -0.3311711549758911, 0.32700538635253906, -0.1680484116077423, -0.03360207378864288, -0.15168657898902893, -0.05473482236266136, 0.24902035295963287, -0.10521707683801651, -0.29829564690589905, -0.2496383786201477, -0.1970980167388916, 0.22305883467197418, -0.29266539216041565, -0.21160593628883362, -0.1770162731409073, -0.07414986938238144, -0.14019392430782318, 0.027845093980431557, 0.22281521558761597, 0.11768713593482971, 0.03263407200574875, -0.15178051590919495, -0.1452549695968628, -0.24856771528720856, 0.33657950162887573, 0.052055880427360535, -0.1408713459968567, 0.03853704780340195, 0.03523235023021698, 0.014001336880028248, 0.17996087670326233, 0.11784433573484421, -0.12312270700931549, 0.19788780808448792, -0.2788589298725128, 0.29525047540664673, 0.23055298626422882, -0.25402912497520447, 0.010372273623943329, 0.31672465801239014, 0.26697054505348206, 0.13920320570468903, 0.11717017740011215, -0.32901662588119507, -0.09989454597234726, 0.11832193285226822, -0.26283541321754456, -0.05908059701323509, 0.03934871405363083, -0.10333249717950821, -0.21532249450683594, 0.020680006593465805, -0.036073118448257446, -0.10807149112224579, -0.17250454425811768, -0.04275650531053543, -0.3998684883117676, -0.08388248085975647, -0.03833531215786934, -0.6121565103530884, -0.1214403510093689, -0.5194762349128723, 0.15431684255599976, 0.04839479550719261, 0.47348713874816895, -0.010718682780861855, -0.09485885500907898, -0.031376928091049194, 0.030421029776334763, 0.5715872645378113, -0.4146973490715027, 0.21319133043289185, 0.4603511691093445, 0.18213896453380585, -0.6949966549873352, 0.005641167052090168, -0.5489999651908875, -0.11691171675920486, 0.6284412741661072, 0.26226386427879333, -0.24899615347385406, -0.1502830982208252, 0.2737720310688019, 0.01633804850280285, -0.09192755818367004, -0.16651655733585358, -0.2935797870159149, -0.22302326560020447, -0.036795105785131454, 0.14788106083869934, -0.009105734527111053, 0.2636661231517792, 0.19725759327411652, -0.07814434170722961, 0.0583055354654789, -0.21258723735809326, 0.14469777047634125, 0.22399598360061646, 0.2999827563762665, 0.06814354658126831, -0.038512926548719406, 0.11713040620088577, 0.21222597360610962, -0.678682804107666, 0.12109210342168808, -0.5201428532600403, 0.35461097955703735, -0.2175755798816681, -0.2540552020072937, 0.465171217918396, 0.16192786395549774, 0.046617623418569565, -0.16148285567760468, -0.5788983702659607, 0.21016021072864532, -0.047842368483543396, 0.020869791507720947, 0.2084854245185852, 0.16397510468959808, -0.23401503264904022, -0.075669065117836, 0.5497071146965027, 0.0684775561094284, -0.22469250857830048, 0.33281832933425903, 0.19439494609832764, -0.3568423390388489, 0.41191157698631287, -0.08889126032590866, 0.9092607498168945, -0.09022440761327744, -0.007592036388814449, 0.25458523631095886, -0.36367499828338623, 0.6806232929229736, 0.14083026349544525, 0.12344533950090408, -0.5268470644950867, -0.39480623602867126, 0.06874214112758636, -0.08320699632167816, -0.12789766490459442, 0.27551546692848206, -0.22262027859687805, 0.41558218002319336, 0.07301419228315353, -0.19334641098976135, 0.022332942113280296, 0.38627687096595764, 0.0031904629431664944, -0.1048494353890419, -0.34154433012008667, 0.15480980277061462, -0.032970476895570755, 0.1674894094467163, -0.06146250292658806, -0.019678613170981407, -0.13237015902996063, -0.21145190298557281, -0.04834187030792236, 0.19770091772079468, -0.2182677835226059, -0.16570332646369934, 0.3732074499130249, -0.030513174831867218, -0.45243945717811584, 0.13093148171901703, 0.1828852891921997, 0.12453830242156982, 0.006406495813280344, 0.05262504145503044, 0.17659805715084076, 0.0018508562352508307, 0.5183131098747253, 0.25722214579582214, 0.29605770111083984, -0.1017708107829094, -0.05791672691702843, -0.10081510990858078, 0.08597913384437561, -0.10533226281404495, -0.06399103254079819, 0.05866871774196625, -0.05259699001908302, -0.14443068206310272, -0.3400135040283203, 0.0604078471660614, -0.013499947264790535, -0.19766484200954437, 0.19822049140930176, 0.09399646520614624, -0.4110831320285797, 0.27004340291023254, 0.09358714520931244, -0.3252715468406677, -0.03687022626399994, 0.3700679540634155, -0.059865403920412064, 0.43673062324523926, 0.17269733548164368, 0.029365450143814087, -0.016895228996872902, -0.2826912999153137, -0.06539253145456314, 0.3214578926563263, -0.22965192794799805, 0.0666494145989418, -0.12342473119497299, -0.10889716446399689, -0.02172076888382435, 0.035222701728343964, 0.19642871618270874, 0.1793636530637741, 0.0944259837269783, -0.4698393642902374, -0.127084419131279, 0.1616881936788559, 0.06652854382991791, 0.3171401023864746, 0.037787046283483505, -0.09414634108543396, 0.21790006756782532, 0.3500138521194458, -0.43584203720092773, -0.023670760914683342, -0.20669129490852356, -0.07812866568565369, 0.4145660996437073, -0.023621143773198128, 0.07423621416091919, 0.05426977947354317, 0.31571972370147705, 0.07997865229845047, -0.09340883791446686, -0.3097279667854309, -0.18688136339187622, 0.019749615341424942, -0.07820957154035568, -0.011729962192475796, 0.025764543563127518, -0.12001340836286545, -0.07672463357448578, -0.06595707684755325, -0.0914933830499649, -0.14658717811107635, 0.007815038785338402, 0.4473215937614441, 0.0004195283690933138, 0.0969926044344902, 0.0640295222401619, 0.1166827380657196, -0.021098095923662186, -0.2281685322523117, 0.03219021484255791, 0.16215156018733978, -0.05169977992773056, -0.09866417944431305, -0.15055575966835022, -0.047165077179670334, -0.38828244805336, -0.06876639276742935, 0.006775883492082357, -0.1605016142129898, -0.09582359343767166, 0.0773942694067955, 0.6340232491493225, 0.010784690268337727, -0.2965526282787323, -0.19216470420360565, 0.027980729937553406, 0.3456473648548126, -0.3949339687824249, -0.126902237534523, 0.4542474150657654, 0.2467411905527115, 0.2646099627017975, 0.35296595096588135, 0.13911183178424835, -0.08405221253633499, -0.1915542334318161, 0.00400015152990818, 0.29133257269859314, -0.4078510105609894, -0.16344156861305237, 0.17371119558811188, -0.14265140891075134, -0.1698189377784729, 0.32067784667015076, 0.05786522850394249, 0.168035089969635, 0.083892822265625, 0.39065858721733093, 0.576187014579773, 0.1416967511177063, 0.030426671728491783, -0.05772697553038597, 0.11515709012746811, -0.07071812450885773, 0.5745172500610352, -0.027649829164147377, 0.2495676875114441, 0.15677998960018158, 0.5897645950317383, -0.26135602593421936, -0.21765825152397156, -0.04092131182551384, 0.21259474754333496, -0.03424624726176262, -0.21949514746665955, 0.13715608417987823, -0.07145749777555466, -0.38834118843078613, -0.0043716044165194035, 0.014580149203538895, -0.23136131465435028, 0.38037240505218506, -0.04845985770225525, 0.12762011587619781, -0.2602135241031647, -0.1190982460975647, 0.02486773580312729, 0.40308573842048645, -0.39799290895462036, 0.06432221084833145, -0.2506301999092102, 0.09554299712181091, -0.017597749829292297, 0.35474029183387756, 0.1716807633638382, 0.01775095984339714, -0.25124460458755493, -0.08426308631896973, -0.19496102631092072, -0.11999712884426117, -0.2027815282344818, 0.2279926836490631, 0.24091891944408417, 0.018384570255875587, 0.4225253760814667, 0.18160413205623627, -0.22556307911872864, 0.0491376630961895, 0.1672343760728836, 0.09844788163900375, -0.007040185388177633, 0.49314796924591064, -0.10184979438781738, -0.2557190954685211, -0.26668086647987366, -0.05131073668599129, -0.40268993377685547, -0.06665568053722382, -0.11750071495771408, -0.030853325501084328, 0.17395567893981934, -0.3835694491863251, 0.14161594212055206, -0.04973562806844711, 0.4453084468841553, 0.012516305781900883, 0.29891666769981384, -0.21124085783958435, -0.21040160953998566, -0.6262405514717102, -0.022991569712758064, -0.09374181181192398, -0.5266149044036865, 0.2152058631181717, 0.02959825098514557, -0.06426091492176056, 0.020876655355095863, 0.04277653247117996, 0.5819960832595825, -0.16318999230861664, -0.1065303236246109, -0.42306721210479736, 0.11542148143053055, -0.18298152089118958, -0.05020393803715706, 0.14122670888900757, -0.29087457060813904, 0.011112341657280922, -0.23848788440227509, 0.1909264475107193, 0.07287725061178207, 0.025542670860886574, 0.06769722700119019, 0.27018487453460693, 0.13873106241226196, 0.043671831488609314, 0.2616764307022095, -0.14459004998207092, 0.08573217689990997, -0.10288732498884201, -0.01684618555009365, -0.29823511838912964, 0.039442457258701324, 0.1075923964381218, 0.33385181427001953, 0.04120611771941185, 0.02199254184961319, -0.07287892699241638, 0.09334114193916321, 0.16829843819141388, 0.2568712830543518, 0.03382585942745209, 0.19437935948371887, 0.14239339530467987, -0.04101568087935448, 0.0953967273235321, 0.3407493531703949, 0.005374308209866285, 0.05500326305627823, -0.012975076213479042, -0.3489190638065338, 0.4604705274105072, -0.32872432470321655, -0.1328890025615692, 0.09408235549926758, 0.3477376103401184, 0.26250365376472473, -0.24081891775131226, -0.5395863652229309, -0.13690753281116486, 0.3368123173713684, 0.042627964168787, -0.29551684856414795, 0.31263697147369385, -0.048021964728832245, 0.12107101827859879, 0.05420394986867905, -0.015983637422323227, 0.29871201515197754, -0.25793614983558655, -0.10434528440237045, -0.2538380026817322 ]
https://github.com/huggingface/datasets/issues/181
Cannot upload my own dataset
It's my misunderstanding. I cannot just upload a csv. I need to write a dataset loading script too.
I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ```
18
Cannot upload my own dataset I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ``` It's my misunderstanding. I cannot just upload a csv. I need to write a dataset loading script too.
[ 0.1318960040807724, -0.12678278982639313, 0.05668966844677925, -0.028536954894661903, 0.1169990599155426, -0.15414656698703766, 0.3969919979572296, -0.06210583820939064, -0.21482880413532257, 0.25181323289871216, -0.08670604974031448, 0.1588994711637497, -0.17180001735687256, 0.29421207308769226, 0.3824804425239563, 0.10448437929153442, 0.0907260850071907, 0.31321385502815247, 0.039624035358428955, -0.12798340618610382, -0.2627108693122864, 0.05864519625902176, -0.016203980892896652, -0.07316692173480988, -0.07528486847877502, -0.18676739931106567, -0.13502033054828644, 0.2678757607936859, -0.2258433997631073, -0.3100970387458801, 0.4256775379180908, 0.20961816608905792, 0.34658488631248474, 0.5284515023231506, -0.00012068287469446659, -0.06405141949653625, 0.17346465587615967, -0.24845094978809357, -0.5683867931365967, -0.35898587107658386, 0.08358648419380188, -0.1341313272714615, 0.08103367686271667, -0.2865523397922516, 0.22721834480762482, 0.05291759595274925, 0.050873491913080215, 0.13236580789089203, 0.5366078615188599, 0.4095074534416199, 0.12447773665189743, 0.06690984964370728, -0.2169531136751175, -0.10151197761297226, 0.008196079172194004, 0.16872473061084747, -0.03125663101673126, 0.2934637665748596, 0.3192078769207001, -0.4081140160560608, 0.2917585074901581, -0.24268227815628052, -0.03904305025935173, -0.007285440806299448, 0.5370673537254333, 0.054553668946027756, -0.053798429667949677, -0.42572593688964844, 0.004693928174674511, 0.1571439653635025, 0.2818419933319092, 0.033648986369371414, -0.32898274064064026, -0.11295104771852493, 0.1115846261382103, -0.5599851608276367, 0.2091520130634308, 0.3053469657897949, -0.425655722618103, -0.03852654993534088, -0.20674799382686615, 0.33666881918907166, -0.39196524024009705, 0.47593656182289124, 0.11568745225667953, 0.3350122272968292, -0.03723040968179703, 0.1301804780960083, 0.28618332743644714, -0.0934036448597908, -0.009626006707549095, 0.07325895875692368, 0.025191210210323334, 0.29447096586227417, -0.16644208133220673, -0.18143430352210999, -0.03056635893881321, -0.2799898386001587, -0.1150284931063652, 0.09093812853097916, 0.46870890259742737, -0.18926240503787994, -0.21843726933002472, 0.16301044821739197, 0.11493644118309021, 0.3294890522956848, -0.02343839779496193, 0.052270855754613876, 0.1015273928642273, -0.02500060386955738, 0.3100627362728119, -0.0896846130490303, -0.2538236379623413, -0.10262618958950043, -0.14759550988674164, -0.023878026753664017, 0.27968505024909973, 0.12876662611961365, 0.012593479827046394, -0.1362917423248291, 0.055091071873903275, 0.0633087083697319, -0.16161294281482697, 0.32300928235054016, -0.08835189789533615, -0.26840245723724365, 0.0020207262132316828, 0.27830973267555237, 0.1604539304971695, -0.2968744933605194, 0.10166545957326889, 0.33483898639678955, -0.10520272701978683, -0.15593522787094116, 0.38596177101135254, 0.026123657822608948, 0.1852371245622635, -0.05975554883480072, 0.22655616700649261, -0.14647097885608673, 0.022026164457201958, -0.03487221896648407, -0.1694040447473526, 0.014915183186531067, 0.23729680478572845, 0.3928620219230652, -0.03909885510802269, -0.5896612405776978, -0.11377377808094025, 0.09599344432353973, -0.4078330099582672, -0.3410406708717346, -0.4582658112049103, 0.056706830859184265, 0.04793150722980499, -0.18805932998657227, -0.35079944133758545, -0.13403895497322083, 0.10075774788856506, -0.27891290187835693, -0.0907556414604187, 0.010323682799935341, -0.0874781534075737, -0.13204549252986908, 0.061368897557258606, 0.04111165180802345, -0.5356765985488892, 0.26288166642189026, -0.08810202032327652, 0.552069902420044, 0.35892918705940247, 0.29697826504707336, -0.3132718503475189, 0.3023732900619507, -0.04386701062321663, 0.17262259125709534, 0.5356268286705017, -0.4726382791996002, -0.22461484372615814, 0.007902617566287518, -0.03681112825870514, -0.17610977590084076, 0.11352511495351791, 0.19710083305835724, 0.12219608575105667, -0.0631401389837265, 0.09179908037185669, 0.3859570622444153, 0.05037333443760872, 0.018724042922258377, -0.19749246537685394, -0.16739971935749054, 0.25019311904907227, -0.026912853121757507, -0.16875801980495453, 0.11825593560934067, 0.2636355459690094, 0.0289665125310421, 0.03021087311208248, -0.3133385479450226, 0.1195906549692154, 0.21816276013851166, 0.33692947030067444, -0.18201369047164917, -0.15409104526042938, 0.28230714797973633, -0.4028118848800659, 0.1659926474094391, 0.1726573258638382, 0.34748631715774536, -0.05526848882436752, -0.044146474450826645, -0.43384432792663574, -0.1111387237906456, -0.02036922052502632, -0.025351950898766518, 0.04577552527189255, 0.11559592932462692, 0.022934842854738235, -0.23246599733829498, -0.27188125252723694, 0.10688889771699905, -0.26044580340385437, 0.04938413202762604, -0.2114427238702774, -0.039563607424497604, -0.32546594738960266, -0.06348615884780884, 0.25600001215934753, 0.11972158402204514, 0.20380035042762756, -0.012564628385007381, -0.16738714277744293, 0.10784434527158737, -0.42461222410202026, 0.26699548959732056, -0.030855463817715645, 0.11985618621110916, 0.007746225222945213, -0.006524687632918358, -0.2597018778324127, -0.07866550981998444, 0.22222702205181122, -0.2051437795162201, -0.3840412199497223, 0.16649077832698822, -0.0096189696341753, 0.06530938297510147, 0.12813472747802734, 0.008446881547570229, 0.29795196652412415, -0.18777604401111603, 0.06347141414880753, 0.1433498114347458, 0.35993725061416626, 0.11930426210165024, 0.3806168735027313, -0.021553343161940575, -0.12040477991104126, -0.15498393774032593, 0.16857102513313293, 0.010845290496945381, 0.13222640752792358, 0.05839194729924202, 0.06194048374891281, -0.07862579077482224, 0.3246953785419464, 0.5969870686531067, 0.3846932649612427, 0.03554395213723183, -0.06499963253736496, -0.008058897219598293, -0.10091284662485123, -0.23450545966625214, -0.025496482849121094, -0.043070245534181595, 0.042327508330345154, 0.16191336512565613, 0.3972330093383789, 0.2295970916748047, -0.09001194685697556, 0.027445001527667046, 0.13210530579090118, 0.42635685205459595, -0.21294213831424713, 0.26301267743110657, -0.1049758717417717, -0.4120582342147827, 0.28124263882637024, -0.2462049275636673, -0.06726311147212982, -0.15471328794956207, -0.18189232051372528, 0.13285882771015167, 0.10234684497117996, -0.14529000222682953, 0.1561025232076645, 0.37681615352630615, -0.11184244602918625, -0.2587057650089264, -0.23002497851848602, -0.09443702548742294, -0.20480398833751678, 0.02289159782230854, 0.28228187561035156, -0.0796215757727623, 0.3960423469543457, 0.29778990149497986, 0.059125591069459915, 0.003640009555965662, -0.30800002813339233, -0.01848365180194378, -0.04487858712673187, 0.10120924562215805, 0.09834824502468109, 0.38181543350219727, 0.01183316856622696, -0.5280571579933167, 0.1464269906282425, 0.07743412256240845, -0.2765725553035736, 0.24847912788391113, -0.021304475143551826, -0.35785603523254395, -0.48331737518310547, -0.025970444083213806, -0.23736825585365295, -0.37235894799232483, 0.23394307494163513, 0.34234684705734253, 0.20104266703128815, 0.6263757944107056, 0.03308162838220596, 0.10552327334880829, 0.3673253655433655, 0.15832513570785522, 0.03778160735964775, -0.14254717528820038, 0.4890121817588806, -0.2197069227695465, -0.41547152400016785, 0.1447935402393341, -0.15794166922569275, 0.5447526574134827, -0.07129497826099396, -0.3519461452960968, -0.18232190608978271, 0.08118414133787155, 0.1464410424232483, -0.03778696805238724, -0.03920019790530205, 0.3170754909515381, -0.2450893223285675, 0.06349494308233261, -0.07407782971858978, -0.34562551975250244, 0.11119983345270157, 0.631762683391571, 0.2669854760169983, 0.0019000984029844403, 0.46685686707496643, 0.0014622043818235397, 0.7395946979522705, 0.07472337037324905, -0.29264503717422485, 0.41096189618110657, -0.10046782344579697, 0.2557591199874878, -0.1761990785598755, -0.3271771967411041, 0.02762295864522457, 0.1809423863887787, -0.12938350439071655, 0.12235864251852036, -0.05677149444818497, -0.07399090379476547, -0.4756852686405182, 0.3063320219516754, -0.29412996768951416, 0.049686651676893234, 0.010090081952512264, 0.10700743645429611, -0.009691514074802399, -0.07850060611963272, -0.22861391305923462, -0.3366197943687439, -0.48897427320480347, -0.18894614279270172, 0.2726038992404938, -0.21989959478378296, 0.17207488417625427, -0.6565855145454407, -0.13269838690757751, -0.645001232624054, 0.30821242928504944, -0.16790398955345154, 0.3023480772972107, -0.18335306644439697, 0.18022285401821136, 0.17020992934703827, -0.06942977011203766, 0.1479046791791916, -0.3889737129211426, -0.20179766416549683, -0.10400507599115372, 0.31818920373916626, -0.42693424224853516, -0.06536194682121277, -0.07280018925666809, 0.0820649042725563, 0.4818257987499237, 0.20355437695980072, -0.014485529623925686, 0.0970025435090065, -0.19211529195308685, -0.18257421255111694, -0.20688672363758087, -0.21783803403377533, -0.08103331178426743, -0.6945073008537292, -0.36785969138145447, -0.5016219615936279, 0.12273707240819931, 0.139009490609169, 0.14689341187477112, -0.030241912230849266, 0.21095657348632812, 0.08789093047380447, 0.4853441119194031, 0.004931288305670023, -0.059936441481113434, 0.6008955836296082, -0.16515682637691498, -0.47179046273231506, 0.496304988861084, -0.11771070957183838, 0.3781428337097168, 0.09147689491510391, -0.15693651139736176, -0.13819988071918488, -0.3253125250339508, 0.47440215945243835, 0.2339237481355667, 0.10316658020019531, 0.28175294399261475, -0.06286690384149551, 0.026009269058704376, -0.1506825089454651, 0.4732232391834259, 0.342796266078949, -0.13644374907016754, -0.3455304801464081, -0.14745089411735535, 0.48506879806518555, -0.20866644382476807, -0.06152811273932457, 0.22016842663288116, -0.20480911433696747, -0.23179934918880463, 0.21763408184051514, 0.16481974720954895, 1.2438881397247314, -0.05959576740860939, 0.4256589412689209, 0.07456768304109573, -0.03275956213474274, 0.7743014097213745, -0.16633498668670654, 0.06815854460000992, -0.23189252614974976, 0.1349937915802002, -0.14140096306800842, -0.061253536492586136, -0.11391089856624603, 0.1102679967880249, -0.22976018488407135, 0.10841342061758041, -0.11354122310876846, -0.04289872944355011, 0.0006199050112627447, 0.451251357793808, 0.10172685235738754, -0.34147417545318604, -0.17380928993225098, 0.06134425848722458, -0.25388625264167786, 0.31335458159446716, -0.17983408272266388, -0.21803885698318481, -0.3014194369316101, -0.2326200306415558, -0.18177863955497742, 0.24452164769172668, -0.2698897421360016, 0.07795136421918869, 0.0035695654805749655, -0.32298919558525085, 0.17310458421707153, 0.46305543184280396, 0.5105305910110474, -0.06357845664024353, -0.06871617585420609, -0.009666434489190578, -0.08557639271020889, -0.04236757010221481, -0.0929514542222023, 0.06018447130918503, 0.1363019198179245, -0.29558834433555603, 0.07367736846208572, 0.32115551829338074, -0.0851583406329155, -0.3732730448246002, -0.21020686626434326, 0.19486312568187714, 0.05941179394721985, -0.13404139876365662, -0.353651225566864, 0.06596378237009048, -0.21248780190944672, -0.015740210190415382, 0.004534305538982153, 0.03315125033259392, 0.08301077038049698, -0.09015682339668274, 0.08961178362369537, -0.19795213639736176, 0.09447414427995682, 0.06312530487775803, -0.0416487380862236, -0.04899268597364426, 0.7007595896720886, -0.12110782414674759, 0.08024582266807556, -0.04312048479914665, -0.2052937150001526, -0.33442604541778564, -0.14639467000961304, -0.07787693291902542, 0.3191206157207489, -0.030446503311395645, 0.039158403873443604, 0.6615248918533325, 0.39813438057899475, 0.1872178018093109, 0.0680561289191246, -0.11110655218362808, -0.323156476020813, -0.31334179639816284, -0.03441181033849716, 0.20288468897342682, -0.025104796513915062, 0.21315938234329224, 0.20983292162418365, 0.2966141700744629, -0.21475601196289062, -0.18158242106437683, -0.20305897295475006, 0.17764775454998016, 0.7618896961212158, -0.013124203309416771, -0.11110201478004456, -0.01039541233330965, -0.033159494400024414, 0.0095929941162467, -0.1480642706155777, -0.10335219651460648, -0.3793175220489502, 0.18479838967323303, 0.530653715133667, -0.11960484087467194, -0.014833870343863964, -0.23938076198101044, -0.11199570447206497, -0.12136786431074142, -0.32977423071861267, 0.041650619357824326, -0.014645803719758987, 0.20553633570671082, -0.3102741539478302, 0.39256688952445984, -0.1536855399608612, 0.17300890386104584, 0.02032848261296749, 0.2511647343635559, 0.06056530401110649, -0.052163396030664444, 0.15323275327682495, 0.07449077069759369, -0.39298877120018005, 0.006555031053721905, -0.1775217205286026, -0.09909837692975998, 0.025729240849614143, -0.41810235381126404, 0.15793192386627197, -0.2215922623872757, 0.02227184548974037, 0.04608582705259323, -0.5020130276679993, 0.21895766258239746, 0.44168978929519653, 0.08857282251119614, 0.040704675018787384, 0.16635127365589142, 0.07060414552688599, 0.2138996124267578, 0.024118464440107346, 0.2750566899776459, -0.0007004319340921938, -0.3031488358974457, -0.04899626970291138, -0.22703419625759125, -0.11426646262407303, -0.27213311195373535, 0.043391305953264236, -0.060878656804561615, -0.4242880046367645, 0.004608695860952139, 0.17042547464370728, 0.1590811014175415, -0.1035110279917717, 0.15583407878875732, -0.19883930683135986, 0.09525319188833237, 0.05639875307679176, -0.11588557064533234, 0.1838594675064087, -0.23433241248130798, 0.12833629548549652, -0.045181889086961746, -0.007980618625879288, 0.3380734324455261, 0.16224469244480133, 0.6163568496704102, -0.34986597299575806, -0.011332114227116108, -0.26698705554008484, 0.05803229659795761, 0.023866893723607063, 0.17970991134643555, -0.660415768623352, -0.005375952459871769, 0.06842368096113205, -0.08650720864534378, 0.03860209509730339, -0.32679834961891174, -0.05037999153137207, 0.26297029852867126, -0.16344895958900452, 0.1078675389289856, -0.13042381405830383, 0.19259321689605713, 0.1316489577293396, -0.11263591051101685, 0.3684442937374115, 0.003190580289810896, -0.03597094863653183, -0.06830333918333054, 0.05866580083966255, 0.5559297800064087, 0.6175882816314697, 0.1681138575077057, 0.2705540359020233, 0.23017145693302155, -0.006906915921717882, 0.07748176157474518, 0.16673292219638824, 0.46167081594467163, 0.3019828796386719, 0.1201508566737175, 0.030812909826636314, -0.12133550643920898, 0.14843043684959412, 0.05331902205944061, 0.0956461951136589, -0.3158130943775177, 0.22804559767246246, -0.02910594455897808, -0.3997560739517212, -0.1564083844423294, 0.4017167091369629, -0.45790740847587585, -0.014089539647102356, 0.19009722769260406, 0.18232698738574982, 0.36742445826530457, -0.23387102782726288, 0.022305460646748543, -0.18495407700538635, 0.3975648581981659, 0.4656152129173279, -0.03612707555294037, -0.1464557647705078, -0.21958963572978973, -0.48715418577194214, 0.17760160565376282, -0.22294281423091888, -0.20073817670345306, -0.3758898377418518, 0.17404401302337646, 0.062224626541137695, -0.018662473186850548, 0.07831179350614548, 0.08619589358568192, -0.2741784453392029, 0.02323915995657444, -0.3622923791408539, -0.020600583404302597, -0.12250672280788422, -0.0025123432278633118, -0.07682990282773972, -0.16656649112701416, 0.03956833854317665, 0.11009873449802399, 0.01287685614079237, 0.024147963151335716, 0.24240922927856445, 0.13704907894134521, 0.36731070280075073, 0.4786680042743683, -0.07187185436487198, 0.7685263752937317, 0.06704562157392502, 0.02057361602783203, 0.049299370497465134, -0.04225950688123703, -0.17712531983852386, -0.0018189174588769674, 0.1784062385559082, 0.3781343400478363, -0.05883779749274254, 0.34986382722854614, -0.19776208698749542, 0.06250420212745667, -0.022588172927498817, -0.5348160266876221, 0.055299222469329834, -0.006338216830044985, -0.0871974527835846, 0.23023846745491028, 0.4238302409648895, 0.08845046907663345, 0.07060226052999496, -0.14375503361225128, -0.4335862994194031, -0.2854062616825104, 0.4140370488166809, -0.5086390972137451, -0.300311416387558, 0.22361122071743011, 0.2395336925983429, -0.02934304066002369, 0.05272843316197395, -0.2230433225631714, -0.08505084365606308, 0.13561636209487915, -0.25086042284965515, -0.05188949406147003, 0.3241965174674988, -0.5186465978622437, 0.016272209584712982, -0.05528751015663147, 0.30498990416526794, 0.2024189978837967, -0.37965112924575806, -0.07080454379320145, -0.21507945656776428 ]
https://github.com/huggingface/datasets/issues/181
Cannot upload my own dataset
I now try with the sample `datasets/csv` folder. nlp-cli upload csv The error is still the same ``` 2020-05-21 17:20:56.394659: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/csv/csv.py to S3 under filename csv/csv.py and namespace korakot About to upload file /content/csv/dummy/0.0.0/dummy_data.zip to S3 under filename csv/dummy/0.0.0/dummy_data.zip and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ```
I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ```
116
Cannot upload my own dataset I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ``` I now try with the sample `datasets/csv` folder. nlp-cli upload csv The error is still the same ``` 2020-05-21 17:20:56.394659: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/csv/csv.py to S3 under filename csv/csv.py and namespace korakot About to upload file /content/csv/dummy/0.0.0/dummy_data.zip to S3 under filename csv/dummy/0.0.0/dummy_data.zip and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ```
[ 0.05260507017374039, -0.028109177947044373, 0.05533065274357796, -0.002076217671856284, 0.1455783098936081, -0.18475791811943054, 0.33769601583480835, -0.04772253334522247, -0.1926683783531189, 0.17972227931022644, -0.13032948970794678, 0.24369046092033386, -0.21156200766563416, 0.2791496515274048, 0.3693571388721466, 0.06986444443464279, 0.013265009969472885, 0.2872202694416046, 0.012621965259313583, -0.10832814127206802, -0.25432002544403076, 0.09193557500839233, -0.01930961012840271, -0.08095873892307281, -0.12342570722103119, -0.19380082190036774, -0.08119554072618484, 0.26023662090301514, -0.22382287681102753, -0.36161234974861145, 0.49111834168434143, 0.2347916215658188, 0.3647100329399109, 0.5793414115905762, -0.00011945607548113912, -0.048161860555410385, 0.16207441687583923, -0.22125191986560822, -0.6242924928665161, -0.4079533517360687, 0.02053179033100605, -0.15109434723854065, 0.07041117548942566, -0.25698813796043396, 0.18926163017749786, 0.02318582497537136, 0.021707406267523766, 0.0470513254404068, 0.4896421432495117, 0.40669864416122437, 0.13775812089443207, 0.12786884605884552, -0.22636938095092773, -0.13977250456809998, 0.0078122676350176334, 0.1732068806886673, -0.058755338191986084, 0.2595105469226837, 0.33482596278190613, -0.40046393871307373, 0.2734425961971283, -0.19609776139259338, 0.012751558795571327, -0.03790632635354996, 0.48874005675315857, -0.02175416611135006, -0.06072641909122467, -0.3897155225276947, -0.011177551001310349, 0.15420448780059814, 0.359345942735672, -0.017493410035967827, -0.39059364795684814, -0.1249680295586586, 0.13513141870498657, -0.48830851912498474, 0.24852284789085388, 0.2740955054759979, -0.4021526277065277, -0.027209019288420677, -0.26199376583099365, 0.3479146361351013, -0.4030779004096985, 0.44457802176475525, 0.10771244764328003, 0.37084653973579407, -0.037059519439935684, 0.09139953553676605, 0.2884840667247772, -0.15400110185146332, 0.0222611241042614, 0.038142263889312744, 0.02801339514553547, 0.26813408732414246, -0.21565788984298706, -0.15431590378284454, 0.011640980839729309, -0.23738588392734528, -0.08751226216554642, 0.11485116183757782, 0.42785215377807617, -0.12593218684196472, -0.2469717115163803, 0.14052730798721313, 0.10184363275766373, 0.3375079333782196, 0.023060964420437813, 0.12450456619262695, 0.13154950737953186, 0.003659794107079506, 0.24482949078083038, -0.07746660709381104, -0.2943651080131531, -0.08862198889255524, -0.11200959980487823, -0.05212016776204109, 0.2688553035259247, 0.07469891011714935, -0.007578806485980749, -0.10847937315702438, 0.07862377911806107, 0.042311180382966995, -0.08323124051094055, 0.3526003062725067, -0.059402596205472946, -0.2365913987159729, -0.02016693912446499, 0.2576303780078888, 0.14143484830856323, -0.26095741987228394, 0.06589074432849884, 0.2955012321472168, -0.12431138008832932, -0.1489529311656952, 0.4141167104244232, 0.023556025698781013, 0.1915086954832077, -0.06857948750257492, 0.21296533942222595, -0.11405139416456223, 0.05686986818909645, 0.029537387192249298, -0.1648234874010086, 0.017245996743440628, 0.22356106340885162, 0.3928877115249634, -0.05233939737081528, -0.5404901504516602, -0.1403442770242691, 0.023450613021850586, -0.3477194905281067, -0.3577897250652313, -0.4238155484199524, 0.07509467005729675, 0.020578205585479736, -0.2402537614107132, -0.33749985694885254, -0.09723229706287384, 0.09148174524307251, -0.2820450961589813, -0.09507915377616882, -0.045189060270786285, -0.1562713384628296, -0.11635982990264893, 0.11207778751850128, 0.06296442449092865, -0.4992433488368988, 0.21195167303085327, -0.07937360554933548, 0.5196458697319031, 0.2760166823863983, 0.30383020639419556, -0.3427070379257202, 0.23301370441913605, -0.06637613475322723, 0.184780552983284, 0.5762099623680115, -0.496898353099823, -0.3117567002773285, 0.04671144112944603, -0.03056415729224682, -0.11799032986164093, 0.07224146276712418, 0.17157937586307526, 0.11363746225833893, -0.08424404263496399, 0.10218207538127899, 0.4385216534137726, 0.06725231558084488, 0.04018228501081467, -0.22083260118961334, -0.1542978435754776, 0.2543802261352539, 0.036984529346227646, -0.1506790667772293, 0.08312279731035233, 0.26047369837760925, 0.06287048012018204, 0.010455841198563576, -0.3288802206516266, 0.14159730076789856, 0.2101689726114273, 0.35805752873420715, -0.1686781793832779, -0.12786732614040375, 0.21730726957321167, -0.4310542941093445, 0.21843940019607544, 0.18919144570827484, 0.34858348965644836, -0.13365644216537476, -0.03646441549062729, -0.4364340007305145, -0.10224460065364838, -0.10477036237716675, -0.04856071248650551, 0.04172839596867561, 0.12958480417728424, 0.0738745704293251, -0.18000994622707367, -0.22281496226787567, 0.15982887148857117, -0.2588527500629425, 0.06837044656276703, -0.21001912653446198, 0.035861872136592865, -0.27387937903404236, -0.09568676352500916, 0.26330816745758057, 0.16703729331493378, 0.2156517505645752, -0.015993326902389526, -0.20170988142490387, 0.0932929590344429, -0.4534428119659424, 0.28580012917518616, 0.017559455707669258, 0.10015660524368286, 0.036241479218006134, -0.05520585924386978, -0.21381926536560059, -0.10987970978021622, 0.17156188189983368, -0.15249386429786682, -0.367725670337677, 0.1745692938566208, -0.00448804022744298, 0.036084216088056564, 0.16663603484630585, 0.010148579254746437, 0.2959839701652527, -0.17794907093048096, 0.09211907535791397, 0.09837885946035385, 0.41748350858688354, 0.0931713730096817, 0.4033573567867279, -0.011018367484211922, -0.14266139268875122, -0.08572405576705933, 0.16825512051582336, -0.0006489527877420187, 0.15058505535125732, 0.07491134107112885, 0.02422843687236309, -0.05125855654478073, 0.30666786432266235, 0.6624007821083069, 0.3770544230937958, 0.08588395267724991, -0.02636568434536457, -0.05195987597107887, -0.12508544325828552, -0.21065744757652283, -0.0400322750210762, -0.07790257781744003, 0.0588909350335598, 0.18291515111923218, 0.44176456332206726, 0.23638851940631866, -0.13617070019245148, -0.04841203615069389, 0.1532732993364334, 0.45341357588768005, -0.19562582671642303, 0.25837647914886475, -0.1609424650669098, -0.4299411475658417, 0.24248960614204407, -0.2555159032344818, -0.06268616020679474, -0.1816956251859665, -0.2170829027891159, 0.14613687992095947, 0.10717090219259262, -0.08416301012039185, 0.10070127993822098, 0.3315700590610504, -0.10592741519212723, -0.28663918375968933, -0.2544068396091461, -0.11374948173761368, -0.1918024718761444, 0.027764219790697098, 0.325008362531662, -0.08840218186378479, 0.4124760031700134, 0.3248806595802307, 0.08864615112543106, -0.04613914713263512, -0.3268381357192993, -0.004597220569849014, -0.03740989416837692, 0.12089087814092636, 0.12138725817203522, 0.3563718795776367, 0.008665414527058601, -0.5445613265037537, 0.19488072395324707, 0.0784645527601242, -0.24362556636333466, 0.2516605257987976, -0.03253842145204544, -0.3485884666442871, -0.46765097975730896, -0.02082686685025692, -0.27371272444725037, -0.3687703609466553, 0.23967112600803375, 0.2940939664840698, 0.20991680026054382, 0.5754038691520691, 0.017712565138936043, 0.1252834051847458, 0.32062774896621704, 0.12391535937786102, 0.06834254413843155, -0.1860465258359909, 0.5255689024925232, -0.24559198319911957, -0.4252486824989319, 0.1025654748082161, -0.1634906679391861, 0.5601881742477417, -0.00971814151853323, -0.350139856338501, -0.1784343719482422, 0.059479452669620514, 0.2183055430650711, -0.023406868800520897, -0.06744986772537231, 0.2789645493030548, -0.20497187972068787, 0.061262767761945724, -0.08193539083003998, -0.3369683623313904, 0.13447405397891998, 0.6249411106109619, 0.2558959424495697, -0.04228245094418526, 0.4419952929019928, 0.03808320686221123, 0.7028173804283142, 0.10719956457614899, -0.3385477364063263, 0.44943198561668396, -0.03851423040032387, 0.18772763013839722, -0.13236820697784424, -0.3693922162055969, 0.0650610625743866, 0.22544586658477783, -0.06130951642990112, 0.13972711563110352, -0.022240621969103813, -0.10212096571922302, -0.4349321722984314, 0.2301696240901947, -0.25660955905914307, 0.04861358925700188, -0.005602943245321512, 0.10512040555477142, 0.055510494858026505, -0.06783225387334824, -0.21281220018863678, -0.3788442313671112, -0.4829713702201843, -0.16287356615066528, 0.27499496936798096, -0.23916321992874146, 0.17479705810546875, -0.6903293132781982, -0.09328904002904892, -0.6602334976196289, 0.3001752197742462, -0.13872285187244415, 0.331415057182312, -0.2023535519838333, 0.20676757395267487, 0.11335979402065277, -0.07665057480335236, 0.1280316859483719, -0.46501660346984863, -0.1679803729057312, -0.07704360038042068, 0.27476802468299866, -0.47235941886901855, -0.04191236570477486, -0.027605319395661354, 0.07491360604763031, 0.4365335702896118, 0.23509712517261505, 0.018999626860022545, 0.07955653220415115, -0.17663241922855377, -0.20305976271629333, -0.15370045602321625, -0.223358616232872, -0.03689245134592056, -0.6711584329605103, -0.3523871600627899, -0.447581946849823, 0.06721371412277222, 0.14565296471118927, 0.14058217406272888, -0.03764292970299721, 0.18601667881011963, 0.11092162877321243, 0.5076418519020081, 0.04284243285655975, -0.046259768307209015, 0.6295244097709656, -0.09305144846439362, -0.45713555812835693, 0.4725039005279541, -0.009942492470145226, 0.3894802927970886, 0.03559461981058121, -0.19427144527435303, -0.16963039338588715, -0.28503361344337463, 0.4578326642513275, 0.17230446636676788, 0.08548000454902649, 0.21727702021598816, -0.10695070773363113, 0.028125548735260963, -0.15494082868099213, 0.4699552059173584, 0.338965505361557, -0.09147369861602783, -0.4181351065635681, -0.1839522123336792, 0.4495735466480255, -0.17907272279262543, -0.03373914211988449, 0.24281229078769684, -0.1924961805343628, -0.2763458490371704, 0.20061789453029633, 0.11939840018749237, 1.091148853302002, -0.07678544521331787, 0.37884521484375, 0.12213286012411118, -0.048747748136520386, 0.7742520570755005, -0.14554975926876068, 0.045703303068876266, -0.23358765244483948, 0.10043346136808395, -0.1133074089884758, -0.05880850553512573, -0.06760561466217041, 0.1411718726158142, -0.23739932477474213, 0.08419696241617203, -0.04000595584511757, -0.11120809614658356, 0.007258252240717411, 0.5523269176483154, 0.07270980626344681, -0.3941994607448578, -0.22026029229164124, 0.08384991437196732, -0.311298668384552, 0.30937930941581726, -0.16995708644390106, -0.21965540945529938, -0.2857399582862854, -0.18248222768306732, -0.2072548121213913, 0.2760235667228699, -0.2514284551143646, 0.09939209371805191, 0.048872899264097214, -0.35540124773979187, 0.319200724363327, 0.4633529484272003, 0.5242566466331482, -0.07650511711835861, -0.0971658006310463, 0.04989144206047058, -0.1302102655172348, -0.014175839722156525, -0.038776934146881104, 0.03059144876897335, 0.1902915984392166, -0.2695980370044708, 0.042753905057907104, 0.2530287802219391, -0.02573995664715767, -0.31226563453674316, -0.1774221509695053, 0.14603418111801147, 0.04673401266336441, -0.16329127550125122, -0.35304486751556396, 0.08665362745523453, -0.1731005758047104, -0.04445960745215416, 0.02780868299305439, 0.02873481996357441, 0.07372667640447617, -0.0680365115404129, 0.0925288125872612, -0.2007799744606018, 0.08086918294429779, 0.10644624382257462, -0.1156516820192337, -0.05884043127298355, 0.7411607503890991, -0.060061264783144, 0.07929163426160812, -0.08564355224370956, -0.19852153956890106, -0.3181912302970886, -0.14075738191604614, -0.08977241814136505, 0.2795090973377228, -0.030942857265472412, 0.08318237960338593, 0.6403136849403381, 0.3384791314601898, 0.10922069847583771, 0.16181719303131104, -0.1482795625925064, -0.3228498697280884, -0.3060603737831116, -0.0438087098300457, 0.19313786923885345, -0.10730323195457458, 0.19159731268882751, 0.20194768905639648, 0.2981015145778656, -0.22825753688812256, -0.1637885868549347, -0.20056729018688202, 0.1670423448085785, 0.7540923953056335, -0.011614820919930935, -0.06655453145503998, -0.036773957312107086, -0.018900178372859955, 0.0023946736473590136, -0.1610790640115738, -0.11697687953710556, -0.38146358728408813, 0.17557691037654877, 0.48093992471694946, -0.18953624367713928, -0.04248155280947685, -0.24042090773582458, -0.09950759261846542, -0.14178162813186646, -0.2967747747898102, 0.06756307929754257, -0.03371114283800125, 0.2408895194530487, -0.26431041955947876, 0.3360063433647156, -0.0966443344950676, 0.21474190056324005, 0.00656738830730319, 0.17916543781757355, 0.08747607469558716, -0.022730806842446327, 0.19590188562870026, 0.07748442143201828, -0.3794806897640228, -0.03813604637980461, -0.20470403134822845, -0.0682208240032196, 0.028971701860427856, -0.4525890648365021, 0.13075216114521027, -0.11226857453584671, 0.04397634044289589, 0.035785119980573654, -0.4939005672931671, 0.1632532924413681, 0.3917880654335022, 0.10325788706541061, 0.01492580771446228, 0.11850079894065857, 0.16980482637882233, 0.2038433849811554, 0.02007768675684929, 0.24454757571220398, 0.034877970814704895, -0.29139766097068787, -0.05173511430621147, -0.20435799658298492, -0.05848478153347969, -0.26753944158554077, 0.12125523388385773, -0.09284653514623642, -0.37827280163764954, -0.04326402395963669, 0.1533942073583603, 0.1886671930551529, -0.1023826003074646, 0.19348900020122528, -0.2057679444551468, 0.12228456884622574, 0.032752249389886856, -0.07289709150791168, 0.21817117929458618, -0.2738109529018402, 0.11131202429533005, -0.06723932176828384, -0.023773400112986565, 0.3217809796333313, 0.14993290603160858, 0.6510178446769714, -0.3514946699142456, -0.005461555905640125, -0.2674216032028198, 0.11025845259428024, 0.009168733842670918, 0.1439596563577652, -0.629501223564148, -0.025503946468234062, 0.05435958132147789, -0.08338459581136703, 0.013343201950192451, -0.33831581473350525, -0.036366209387779236, 0.2841474413871765, -0.1411563754081726, 0.05001046881079674, -0.10629621893167496, 0.21016445755958557, 0.12682916224002838, -0.1563309282064438, 0.41823020577430725, 0.017136648297309875, -0.10469215363264084, -0.048103608191013336, 0.07154374569654465, 0.5066622495651245, 0.58162522315979, 0.13121743500232697, 0.2830178141593933, 0.1813770830631256, 0.033742789179086685, 0.060610584914684296, 0.19215276837348938, 0.4136764407157898, 0.2930334210395813, 0.1462852656841278, 0.029231643304228783, -0.11405064165592194, 0.07913730293512344, 0.049899984151124954, 0.05707652121782303, -0.3188946545124054, 0.2812894284725189, -0.031187105923891068, -0.33619245886802673, -0.16785822808742523, 0.39295390248298645, -0.4424993395805359, 0.011144219897687435, 0.2605811357498169, 0.22482042014598846, 0.3537735641002655, -0.2567594647407532, 0.024390410631895065, -0.1696019470691681, 0.41601407527923584, 0.48843154311180115, -0.07816869020462036, -0.14322051405906677, -0.2934463620185852, -0.5263960361480713, 0.16107000410556793, -0.2162851244211197, -0.23151785135269165, -0.356878399848938, 0.1811113804578781, 0.05726088583469391, -0.09626909345388412, 0.058155301958322525, 0.1449396014213562, -0.2326546013355255, -0.004405546002089977, -0.44343632459640503, 0.014719798229634762, -0.153797909617424, -0.014265604317188263, -0.03849327191710472, -0.16828298568725586, 0.029524296522140503, 0.09202343970537186, 0.018767889589071274, -0.006674217060208321, 0.22007009387016296, 0.13326187431812286, 0.34432676434516907, 0.4730622470378876, -0.0944240465760231, 0.7783322334289551, 0.09284882247447968, 0.004617420490831137, -0.006964905187487602, -0.06584292650222778, -0.16452087461948395, 0.0251687653362751, 0.136099174618721, 0.4669880270957947, -0.08531728386878967, 0.37317201495170593, -0.22988460958003998, 0.03581169620156288, -0.031407590955495834, -0.5221917033195496, 0.07673176378011703, 0.009969313628971577, -0.0682159811258316, 0.24942493438720703, 0.4241003096103668, 0.13799166679382324, 0.09841875731945038, -0.10654082149267197, -0.4279007911682129, -0.3111039102077484, 0.4142976701259613, -0.5499932169914246, -0.33335086703300476, 0.19432902336120605, 0.22382153570652008, -0.05248129740357399, 0.022692807018756866, -0.2831411063671112, -0.07024049013853073, 0.15611468255519867, -0.23596225678920746, -0.019298389554023743, 0.3056486248970032, -0.4318903386592865, 0.07046443969011307, -0.05291147157549858, 0.31289124488830566, 0.19471456110477448, -0.380187451839447, -0.0402463898062706, -0.1632498949766159 ]
https://github.com/huggingface/datasets/issues/181
Cannot upload my own dataset
We haven't tested the dataset upload feature yet cc @julien-c This is on our short/mid-term roadmap though
I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ```
17
Cannot upload my own dataset I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ``` We haven't tested the dataset upload feature yet cc @julien-c This is on our short/mid-term roadmap though
[ 0.010165617801249027, -0.014697292819619179, 0.06399780511856079, -0.047251515090465546, 0.1368081271648407, -0.1440623700618744, 0.3703482449054718, -0.035264384001493454, -0.17155195772647858, 0.153878852725029, -0.10310687124729156, 0.2589794099330902, -0.2308884561061859, 0.3168731927871704, 0.3755457103252411, 0.09489191323518753, 0.01937192678451538, 0.27177372574806213, -0.02011636272072792, -0.09850392490625381, -0.22678814828395844, 0.0701330304145813, -0.0022227352019399405, -0.09574955701828003, -0.14261679351329803, -0.21162675321102142, -0.09032746404409409, 0.2581227719783783, -0.2345554530620575, -0.34122931957244873, 0.4868590235710144, 0.23599037528038025, 0.34359171986579895, 0.5804949402809143, -0.00011795771570177749, -0.051996491849422455, 0.17183135449886322, -0.2157052457332611, -0.6393764615058899, -0.3554876446723938, -0.006130785681307316, -0.1342170536518097, 0.06067825108766556, -0.2517399489879608, 0.192875474691391, 0.027462592348456383, 0.03491760417819023, 0.03088012896478176, 0.4822885990142822, 0.4076225459575653, 0.14891530573368073, 0.14572227001190186, -0.17812171578407288, -0.15002445876598358, -0.002832615515217185, 0.1824352741241455, -0.0589720755815506, 0.2539564371109009, 0.358696848154068, -0.4081486463546753, 0.26444190740585327, -0.1995869129896164, 0.03678077459335327, -0.04424063861370087, 0.48748576641082764, -0.009249813854694366, -0.07491940259933472, -0.4059666395187378, -0.00484011135995388, 0.1771373748779297, 0.3769996762275696, -0.060027215629816055, -0.3945053815841675, -0.1456923931837082, 0.15505386888980865, -0.4790180027484894, 0.25372326374053955, 0.2788682281970978, -0.4247060716152191, -0.02253003790974617, -0.27115753293037415, 0.31374505162239075, -0.40935495495796204, 0.42216619849205017, 0.1056477501988411, 0.3526955842971802, -0.03370974585413933, 0.0666818767786026, 0.24374695122241974, -0.15369358658790588, 0.011768266558647156, 0.06547364592552185, 0.0214677844196558, 0.2782629132270813, -0.21666398644447327, -0.19502879679203033, 0.002794372383505106, -0.240585058927536, -0.058181293308734894, 0.12915611267089844, 0.40473249554634094, -0.12968337535858154, -0.24420137703418732, 0.1506330817937851, 0.13368745148181915, 0.3241250813007355, 0.0014827169943600893, 0.12236382067203522, 0.12233413755893707, -0.005846336483955383, 0.2769157886505127, -0.05174222216010094, -0.27427026629447937, -0.05080210790038109, -0.11301085352897644, -0.06884713470935822, 0.29202789068222046, 0.06772579252719879, -0.005541233345866203, -0.12934695184230804, 0.09735915809869766, 0.016586456447839737, -0.10256084054708481, 0.343923419713974, -0.0819454938173294, -0.2613027095794678, -0.053804900497198105, 0.23263388872146606, 0.13708680868148804, -0.2900107800960541, 0.05705784633755684, 0.2758283317089081, -0.11549311131238937, -0.15443263947963715, 0.4356735646724701, 0.012412016279995441, 0.19565072655677795, -0.05445248633623123, 0.20219503343105316, -0.06607900559902191, 0.07224190980195999, 0.01198643073439598, -0.16287995874881744, 0.008681237697601318, 0.2195928543806076, 0.3847881257534027, -0.06849830597639084, -0.5333331823348999, -0.14358186721801758, 0.026160849258303642, -0.35323697328567505, -0.34732839465141296, -0.4594143033027649, 0.08847449719905853, 0.0037647350691258907, -0.28174886107444763, -0.3244447410106659, -0.07150458544492722, 0.0896536260843277, -0.2517109811306, -0.09521620720624924, -0.046300485730171204, -0.1421903371810913, -0.12248452007770538, 0.12956172227859497, 0.08803895860910416, -0.5273436307907104, 0.18487119674682617, -0.08601419627666473, 0.4892788827419281, 0.27127528190612793, 0.2785511612892151, -0.30868059396743774, 0.2438846379518509, -0.08011314272880554, 0.18975171446800232, 0.5997761487960815, -0.4807259142398834, -0.33468496799468994, 0.04278616979718208, -0.037316642701625824, -0.09002692997455597, 0.03727806732058525, 0.19343505799770355, 0.10801960527896881, -0.09636015444993973, 0.09529583156108856, 0.44792360067367554, 0.04567834734916687, 0.01669199950993061, -0.21615584194660187, -0.16694800555706024, 0.2624110281467438, 0.040793050080537796, -0.15550212562084198, 0.07853072881698608, 0.2734289765357971, 0.08999720960855484, 0.02415444701910019, -0.3324550688266754, 0.15371623635292053, 0.19960275292396545, 0.3573826849460602, -0.17576833069324493, -0.1370927095413208, 0.2060299813747406, -0.4696052670478821, 0.21702052652835846, 0.1805570423603058, 0.349084734916687, -0.12102710455656052, -0.018303686752915382, -0.44371429085731506, -0.1005718857049942, -0.1240299791097641, -0.058668650686740875, 0.049087874591350555, 0.11639916896820068, 0.054980237036943436, -0.17006120085716248, -0.2535099983215332, 0.18731704354286194, -0.2769084870815277, 0.061253029853105545, -0.21974225342273712, 0.045059237629175186, -0.2696872055530548, -0.12417402118444443, 0.2587276101112366, 0.18245325982570648, 0.16576650738716125, 0.00609542103484273, -0.18364877998828888, 0.10668767243623734, -0.4784855544567108, 0.26239827275276184, 0.03369062393903732, 0.1265537589788437, 0.04126192256808281, -0.06761738657951355, -0.18328310549259186, -0.14366260170936584, 0.17080532014369965, -0.1465516984462738, -0.40138399600982666, 0.14321798086166382, -0.027937497943639755, 0.004521084018051624, 0.16433103382587433, 0.0004223847354296595, 0.279594749212265, -0.17933104932308197, 0.11594358831644058, 0.10681777447462082, 0.3852049708366394, 0.06651005893945694, 0.40208306908607483, -0.041343703866004944, -0.12613648176193237, -0.03237977623939514, 0.21093051135540009, 0.019254527986049652, 0.18925891816616058, 0.0736541897058487, 0.01379319652915001, -0.08014342188835144, 0.2915305197238922, 0.6511790752410889, 0.3804543614387512, 0.1110786572098732, -0.0320216603577137, -0.056274279952049255, -0.10829749703407288, -0.18812252581119537, -0.06301793456077576, -0.11955726891756058, 0.07156752794981003, 0.18021786212921143, 0.4236961901187897, 0.2545206546783447, -0.13882191479206085, -0.03602733835577965, 0.12983305752277374, 0.45807692408561707, -0.19086578488349915, 0.23381736874580383, -0.13966161012649536, -0.43804123997688293, 0.2469320297241211, -0.22983048856258392, -0.09044582396745682, -0.1691736876964569, -0.20334842801094055, 0.13439300656318665, 0.1450602114200592, -0.0874415934085846, 0.11892256140708923, 0.3378387987613678, -0.08473848551511765, -0.2890056371688843, -0.260663777589798, -0.09300468862056732, -0.20018258690834045, 0.053276196122169495, 0.32226434350013733, -0.10707732290029526, 0.4432593286037445, 0.33645904064178467, 0.08638137578964233, -0.054917193949222565, -0.3232796788215637, 0.00857230182737112, -0.06074177473783493, 0.12722721695899963, 0.10323645919561386, 0.3434315621852875, -0.0030984249897301197, -0.5138465166091919, 0.1799698770046234, 0.06420642137527466, -0.2702476382255554, 0.22155016660690308, -0.02242877148091793, -0.3159380555152893, -0.44382166862487793, -0.05534730851650238, -0.29095259308815, -0.3391730487346649, 0.25748729705810547, 0.32961562275886536, 0.1998351514339447, 0.5830571055412292, 0.02713247761130333, 0.13877184689044952, 0.32144874334335327, 0.1015031635761261, 0.04909173771739006, -0.20185203850269318, 0.5142776370048523, -0.25085246562957764, -0.46681058406829834, 0.1272696554660797, -0.13542209565639496, 0.5388170480728149, -0.02618454582989216, -0.35499605536460876, -0.20387764275074005, 0.06519963592290878, 0.2458018958568573, -0.03664572909474373, -0.05192951112985611, 0.2837885916233063, -0.19145295023918152, 0.044388484209775925, -0.07229132950305939, -0.33189165592193604, 0.14527878165245056, 0.6041275262832642, 0.2712104618549347, -0.05854354053735733, 0.4339742362499237, 0.04342501610517502, 0.7070648074150085, 0.13270670175552368, -0.3311067223548889, 0.44850072264671326, -0.0690917819738388, 0.20655743777751923, -0.14012017846107483, -0.3747256100177765, 0.13324427604675293, 0.23546594381332397, -0.03365195915102959, 0.11279822885990143, -0.030974699184298515, -0.0822751373052597, -0.4156801998615265, 0.22088822722434998, -0.2339075803756714, 0.04839329421520233, -0.016791323199868202, 0.1189957857131958, 0.04606042057275772, -0.0811108723282814, -0.22336693108081818, -0.3637567460536957, -0.5096309781074524, -0.12451834976673126, 0.24588757753372192, -0.24647212028503418, 0.2104227989912033, -0.6700916290283203, -0.05532244220376015, -0.6859098672866821, 0.28730276226997375, -0.149904265999794, 0.30860117077827454, -0.21619366109371185, 0.19936619699001312, 0.09411243349313736, -0.0823153629899025, 0.15053750574588776, -0.47544965147972107, -0.15888775885105133, -0.08539845794439316, 0.28484705090522766, -0.47482773661613464, -0.026424318552017212, -0.051321182399988174, 0.07265124469995499, 0.4124589264392853, 0.19209937751293182, 0.039509367197752, 0.07596687972545624, -0.14139555394649506, -0.216168612241745, -0.17694418132305145, -0.2280469387769699, -0.04576987400650978, -0.6439386010169983, -0.3358868956565857, -0.463667631149292, 0.056371066719293594, 0.1376699060201645, 0.1639884114265442, -0.06866511702537537, 0.1555037945508957, 0.09471635520458221, 0.505619466304779, 0.04792429879307747, -0.06624256074428558, 0.6384864449501038, -0.09328262507915497, -0.42609357833862305, 0.4850154221057892, 0.0034878000151365995, 0.38921934366226196, 0.02805510349571705, -0.14984716475009918, -0.15592491626739502, -0.2998330593109131, 0.4713129997253418, 0.13734710216522217, 0.08705965429544449, 0.1848033219575882, -0.09927714616060257, 0.02812325209379196, -0.1918717920780182, 0.5012039542198181, 0.3349217474460602, -0.08053188025951385, -0.42228931188583374, -0.18815988302230835, 0.43667906522750854, -0.16746826469898224, -0.0022929843980818987, 0.24263441562652588, -0.18027789890766144, -0.26839762926101685, 0.17485596239566803, 0.14205549657344818, 1.0450090169906616, -0.09819278866052628, 0.3577187657356262, 0.11459795385599136, -0.030823074281215668, 0.7806400656700134, -0.18001143634319305, 0.04540587589144707, -0.2553684115409851, 0.09903968870639801, -0.11995553970336914, -0.04338899999856949, -0.0456891693174839, 0.13680218160152435, -0.25287240743637085, 0.0857430100440979, -0.029485180974006653, -0.14704464375972748, 0.005726405885070562, 0.5664036273956299, 0.06178729981184006, -0.39426031708717346, -0.20463387668132782, 0.10850342363119125, -0.317049115896225, 0.29799193143844604, -0.16285033524036407, -0.2076191008090973, -0.28265753388404846, -0.15511329472064972, -0.2209421545267105, 0.30353108048439026, -0.2398693561553955, 0.07949474453926086, 0.05891745164990425, -0.3198932409286499, 0.33261775970458984, 0.4678554832935333, 0.5250082612037659, -0.08236995339393616, -0.10083205252885818, 0.037575650960206985, -0.16709356009960175, -0.02705296128988266, 0.009090117178857327, 0.050292328000068665, 0.22911953926086426, -0.29069823026657104, 0.030212432146072388, 0.2814529240131378, -0.0009340145043097436, -0.30204159021377563, -0.1756664514541626, 0.13425423204898834, 0.0712273120880127, -0.16748608648777008, -0.3050799071788788, 0.10473206639289856, -0.13321667909622192, -0.03923343867063522, 0.040146246552467346, 0.047616973519325256, 0.06888943165540695, -0.028613876551389694, 0.07775545865297318, -0.18398161232471466, 0.05718260258436203, 0.11787600070238113, -0.11957719922065735, -0.07269201427698135, 0.7520655989646912, -0.05399371311068535, 0.09325054287910461, -0.10994242876768112, -0.18213345110416412, -0.2736780345439911, -0.13815349340438843, -0.09949097782373428, 0.27688896656036377, -0.02094828337430954, 0.09042204916477203, 0.6383233070373535, 0.32130298018455505, 0.09763450920581818, 0.13834577798843384, -0.14989447593688965, -0.35234442353248596, -0.3358486294746399, -0.0205377284437418, 0.18586939573287964, -0.1201789453625679, 0.14587873220443726, 0.21818073093891144, 0.31043463945388794, -0.2514615058898926, -0.19500397145748138, -0.1998571902513504, 0.1583554893732071, 0.7369638681411743, -0.01766371913254261, -0.012900647707283497, -0.06114610657095909, 0.006987984292209148, -0.030109228566288948, -0.17383715510368347, -0.13647127151489258, -0.3693220019340515, 0.1645641177892685, 0.4616594910621643, -0.1914130300283432, -0.056229885667562485, -0.21856892108917236, -0.12847866117954254, -0.13379068672657013, -0.269339919090271, 0.06040286272764206, -0.057563118636608124, 0.24150052666664124, -0.22885163128376007, 0.3202771246433258, -0.08824973553419113, 0.17375928163528442, 0.03331238403916359, 0.16574044525623322, 0.07580552995204926, -0.019645702093839645, 0.19066917896270752, 0.07480891048908234, -0.33523082733154297, -0.026200080290436745, -0.17743359506130219, -0.07269701361656189, 0.017779523506760597, -0.4788331985473633, 0.10806454718112946, -0.13837306201457977, 0.05096149072051048, 0.022260962054133415, -0.49828463792800903, 0.13783903419971466, 0.3878340721130371, 0.11959458887577057, -0.002074212534353137, 0.11067019402980804, 0.20888115465641022, 0.20254860818386078, 0.013453075662255287, 0.23718994855880737, 0.021878035739064217, -0.24977128207683563, -0.07446973025798798, -0.19819703698158264, -0.0792117491364479, -0.28159669041633606, 0.0992148295044899, -0.1124681904911995, -0.3867097496986389, -0.04920406639575958, 0.16625823080539703, 0.19723796844482422, -0.09329351782798767, 0.21389225125312805, -0.2051725685596466, 0.11387532204389572, 0.032077446579933167, -0.061955299228429794, 0.22813265025615692, -0.2797224521636963, 0.13537012040615082, -0.09851952642202377, 0.005144830793142319, 0.3340873718261719, 0.14589481055736542, 0.6321764588356018, -0.3162313997745514, -0.01611190475523472, -0.24867331981658936, 0.07460804283618927, 0.017596669495105743, 0.167875736951828, -0.6061635613441467, -0.02913355641067028, 0.05296030268073082, -0.08525756001472473, 0.010776528157293797, -0.3678249418735504, -0.04822467267513275, 0.275940477848053, -0.15788526833057404, 0.05883338302373886, -0.0975588858127594, 0.1953687071800232, 0.13777555525302887, -0.12739373743534088, 0.4026271104812622, -0.008653298020362854, -0.09197316318750381, -0.05486815422773361, 0.11255784332752228, 0.46995019912719727, 0.5798242688179016, 0.14910899102687836, 0.31919750571250916, 0.20295313000679016, 0.033974163234233856, 0.06528519839048386, 0.18311437964439392, 0.3685241639614105, 0.2624928653240204, 0.13088937103748322, 0.0407571941614151, -0.10797873139381409, 0.05925644189119339, 0.027629371732473373, 0.07058043032884598, -0.3081683814525604, 0.28486961126327515, -0.019958244636654854, -0.32388433814048767, -0.1779017299413681, 0.397659033536911, -0.42802757024765015, 0.008005810901522636, 0.30593663454055786, 0.21202580630779266, 0.3255811333656311, -0.2719443142414093, 0.03361441567540169, -0.18920107185840607, 0.45011427998542786, 0.4820174276828766, -0.09710134565830231, -0.10160031169652939, -0.2517859637737274, -0.5138532519340515, 0.15989546477794647, -0.19470754265785217, -0.26225316524505615, -0.3417939841747284, 0.18984560668468475, 0.03611770644783974, -0.08899537473917007, 0.10701450705528259, 0.1390070617198944, -0.21020257472991943, -0.02003595046699047, -0.4624020457267761, 0.013603978790342808, -0.1590491384267807, 0.00987574178725481, -0.02020551636815071, -0.1756240874528885, 0.043980322778224945, 0.06408078223466873, 0.015800785273313522, -0.0002465331635903567, 0.1899520456790924, 0.1183924525976181, 0.3266213536262512, 0.4968600571155548, -0.10089084506034851, 0.758901834487915, 0.1091519296169281, -0.008117723278701305, -0.012254760600626469, -0.060708776116371155, -0.21167826652526855, 0.07008469849824905, 0.12792463600635529, 0.4384544789791107, -0.08241152763366699, 0.3603608012199402, -0.22094275057315826, 0.07509882003068924, -0.03312685340642929, -0.5146053433418274, 0.08451039344072342, 0.022847414016723633, -0.04442775994539261, 0.276464581489563, 0.41493019461631775, 0.09248579293489456, 0.09413088113069534, -0.09509778022766113, -0.4262842833995819, -0.27461379766464233, 0.3977464437484741, -0.5510284900665283, -0.29202544689178467, 0.1532052904367447, 0.2142147719860077, -0.032668303698301315, -0.00940574798732996, -0.3166787326335907, -0.06710311025381088, 0.17874476313591003, -0.22816728055477142, -0.011713657528162003, 0.2962079644203186, -0.4524933099746704, 0.09072312712669373, -0.046880193054676056, 0.33023375272750854, 0.17732542753219604, -0.37144556641578674, -0.016200171783566475, -0.14145898818969727 ]
https://github.com/huggingface/datasets/issues/181
Cannot upload my own dataset
Even if I fix the `TypeError: __init__() got an unexpected keyword argument 'cdn'` error, it looks like it still uploads to `https://s3.amazonaws.com/models.huggingface.co/bert/<namespace>/<dataset_name>` instead of `https://s3.amazonaws.com/datasets.huggingface.co/nlp/<namespace>/<dataset_name>`
I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ```
25
Cannot upload my own dataset I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ``` Even if I fix the `TypeError: __init__() got an unexpected keyword argument 'cdn'` error, it looks like it still uploads to `https://s3.amazonaws.com/models.huggingface.co/bert/<namespace>/<dataset_name>` instead of `https://s3.amazonaws.com/datasets.huggingface.co/nlp/<namespace>/<dataset_name>`
[ 0.07664904743432999, -0.05189044028520584, 0.05664351209998131, -0.013381031341850758, 0.1621726006269455, -0.18692173063755035, 0.315628319978714, -0.02782159112393856, -0.19460242986679077, 0.19591812789440155, -0.1271059662103653, 0.23585735261440277, -0.2207300215959549, 0.29263460636138916, 0.3533962070941925, 0.07266585528850555, 0.00557507062330842, 0.29510387778282166, 0.00900519173592329, -0.09385800361633301, -0.21986141800880432, 0.09480854868888855, 0.019349943846464157, -0.0798717588186264, -0.1348169595003128, -0.1881406456232071, -0.07919929176568985, 0.27564141154289246, -0.19723379611968994, -0.3347654342651367, 0.44861459732055664, 0.22274284064769745, 0.33252379298210144, 0.600294291973114, -0.00011811571312136948, -0.05863684415817261, 0.16570204496383667, -0.2280743420124054, -0.6185942888259888, -0.37377214431762695, 0.07301397621631622, -0.14274370670318604, 0.09190243482589722, -0.2383841872215271, 0.18933629989624023, 0.03741718456149101, 0.01646963506937027, 0.06538469344377518, 0.5281577110290527, 0.3778299391269684, 0.14735272526741028, 0.10313688218593597, -0.17057624459266663, -0.09499446302652359, 0.027343420311808586, 0.18931706249713898, -0.056042879819869995, 0.29447799921035767, 0.33864378929138184, -0.3992355167865753, 0.29277944564819336, -0.19877924025058746, -0.0034405335318297148, -0.0603887140750885, 0.4824242889881134, -0.02794663794338703, 0.007829021662473679, -0.4104377329349518, -0.029902294278144836, 0.17739343643188477, 0.3018234074115753, -0.030178504064679146, -0.359176367521286, -0.11388027667999268, 0.16878528892993927, -0.48414039611816406, 0.2560546398162842, 0.2725909352302551, -0.4102139472961426, -0.038525864481925964, -0.25709471106529236, 0.3279774487018585, -0.41808053851127625, 0.4426950514316559, 0.11960510164499283, 0.33937373757362366, -0.022680621594190598, 0.07667797803878784, 0.2550290822982788, -0.12778852880001068, 0.007881402038037777, 0.09094329178333282, 0.021747838705778122, 0.2719115614891052, -0.19632501900196075, -0.18465940654277802, -0.02068881131708622, -0.2897575795650482, -0.10948622971773148, 0.09983275085687637, 0.45074576139450073, -0.13097938895225525, -0.24737398326396942, 0.14947135746479034, 0.09313520044088364, 0.3291783928871155, -0.007071911357343197, 0.09553560614585876, 0.12827549874782562, 0.02128220908343792, 0.2753557562828064, -0.06610692292451859, -0.28716927766799927, -0.07212226837873459, -0.1261586844921112, -0.047326404601335526, 0.2747190594673157, 0.08483222872018814, 0.01176806353032589, -0.09488458186388016, 0.0725315511226654, 0.06044616550207138, -0.08017189055681229, 0.3419945538043976, -0.11174194514751434, -0.24170075356960297, 0.00597187876701355, 0.23593322932720184, 0.13921067118644714, -0.26889821887016296, 0.07524184137582779, 0.3023250997066498, -0.11196798086166382, -0.1445469856262207, 0.404756635427475, 0.008400018326938152, 0.1989603191614151, -0.046806685626506805, 0.18748801946640015, -0.13644443452358246, 0.06247757747769356, 0.006644637323915958, -0.17632445693016052, 0.03343520686030388, 0.22954972088336945, 0.38577982783317566, -0.06581928580999374, -0.5420820116996765, -0.13114000856876373, 0.03096676617860794, -0.37687358260154724, -0.3573209047317505, -0.4452533721923828, 0.0883500799536705, 0.04142153635621071, -0.23677773773670197, -0.3071364760398865, -0.11866731196641922, 0.09791082888841629, -0.2934204638004303, -0.09514958411455154, -0.046619437634944916, -0.1209680363535881, -0.13289502263069153, 0.11520441621541977, 0.04702841863036156, -0.47550636529922485, 0.2196236252784729, -0.08374650776386261, 0.5124330520629883, 0.32622963190078735, 0.2987669110298157, -0.3122525215148926, 0.27115723490715027, -0.06317272782325745, 0.17326481640338898, 0.5793394446372986, -0.4712454676628113, -0.2747650146484375, 0.02794957533478737, -0.03406190499663353, -0.10284217447042465, 0.05465628206729889, 0.17109665274620056, 0.08083738386631012, -0.0993407741189003, 0.07489421218633652, 0.4404076933860779, 0.07226981222629547, 0.046246208250522614, -0.23941272497177124, -0.1674555391073227, 0.2583915591239929, 0.033971235156059265, -0.16789831221103668, 0.07635156810283661, 0.24125425517559052, 0.0896945670247078, 0.023809174075722694, -0.33165574073791504, 0.13372574746608734, 0.2027381956577301, 0.35548877716064453, -0.15696808695793152, -0.1437087059020996, 0.23550467193126678, -0.40783795714378357, 0.18856647610664368, 0.12364498525857925, 0.352277934551239, -0.10161994397640228, -0.010192207992076874, -0.41855642199516296, -0.09694615006446838, -0.09851150214672089, -0.023751718923449516, 0.06040216237306595, 0.10851248353719711, 0.05050361901521683, -0.18064585328102112, -0.22650833427906036, 0.14043690264225006, -0.3063787519931793, 0.05729077383875847, -0.21424975991249084, 0.04307178407907486, -0.28264319896698, -0.11205241829156876, 0.2531871199607849, 0.17547647655010223, 0.19479499757289886, -0.015050319023430347, -0.19718362390995026, 0.09308925271034241, -0.45916521549224854, 0.3227212429046631, -0.0023456693161278963, 0.13259734213352203, 0.030503802001476288, -0.02273780293762684, -0.2367979884147644, -0.0917934849858284, 0.17833475768566132, -0.16936562955379486, -0.3360210657119751, 0.13825179636478424, -0.029992733150720596, 0.0167362280189991, 0.1587330549955368, 0.0029891717713326216, 0.2892494797706604, -0.18945999443531036, 0.11126956343650818, 0.1314769685268402, 0.40158364176750183, 0.0971415713429451, 0.38154563307762146, -0.045128896832466125, -0.13328278064727783, -0.08485881239175797, 0.18308334052562714, 0.023188957944512367, 0.16448433697223663, 0.07439377903938293, 0.020963724702596664, -0.09218311309814453, 0.30198606848716736, 0.5998584032058716, 0.360566645860672, 0.07620102167129517, -0.056455664336681366, -0.06078986078500748, -0.1149059534072876, -0.22583232820034027, -0.03621281310915947, -0.08589198440313339, 0.064612478017807, 0.2199162244796753, 0.4395747184753418, 0.2659415602684021, -0.13194307684898376, -0.011397895403206348, 0.10586413741111755, 0.46060752868652344, -0.19288837909698486, 0.24376621842384338, -0.1272563338279724, -0.4152645468711853, 0.24683143198490143, -0.27434229850769043, -0.08276362717151642, -0.17546424269676208, -0.19944074749946594, 0.1508495807647705, 0.09969239681959152, -0.11422175914049149, 0.17351239919662476, 0.3297785520553589, -0.10300759226083755, -0.2672962546348572, -0.28693389892578125, -0.08144994080066681, -0.20950110256671906, 0.04389668256044388, 0.33913367986679077, -0.10610675811767578, 0.42552515864372253, 0.3308124840259552, 0.05794118717312813, -0.03152487799525261, -0.31745123863220215, -0.012977869249880314, -0.03839245066046715, 0.09033801406621933, 0.08890731632709503, 0.37470728158950806, 0.011475899256765842, -0.5478413105010986, 0.19541127979755402, 0.054333969950675964, -0.25022077560424805, 0.24789635837078094, -0.031557098031044006, -0.31541702151298523, -0.47585976123809814, 0.0024254266172647476, -0.2676188349723816, -0.3631880283355713, 0.2492438405752182, 0.33015796542167664, 0.23138810694217682, 0.5911980271339417, 0.019704528152942657, 0.13168852031230927, 0.319717675447464, 0.15069156885147095, 0.045514702796936035, -0.1869828999042511, 0.4966631829738617, -0.23607194423675537, -0.42654773592948914, 0.1076657772064209, -0.1438557356595993, 0.5702791213989258, -0.04440787434577942, -0.3675759434700012, -0.18725794553756714, 0.08487430214881897, 0.17660588026046753, -0.07252097874879837, -0.07167650759220123, 0.29713737964630127, -0.1921672224998474, 0.0524088516831398, -0.05853414535522461, -0.2975662648677826, 0.1311016082763672, 0.6311604380607605, 0.2633402943611145, -0.05882125347852707, 0.4399242699146271, 0.017980385571718216, 0.6985472440719604, 0.10503231734037399, -0.3126606345176697, 0.44941067695617676, -0.03828560188412666, 0.1811317354440689, -0.14026778936386108, -0.34749236702919006, 0.08872748911380768, 0.2385050505399704, -0.042702194303274155, 0.14452439546585083, -0.043684475123882294, -0.07860308140516281, -0.44182538986206055, 0.2470041960477829, -0.2635873556137085, 0.05894435569643974, 0.00958127249032259, 0.10640180110931396, 0.03759400174021721, -0.10671697556972504, -0.24084407091140747, -0.3441416919231415, -0.5165199637413025, -0.12177100032567978, 0.2631739377975464, -0.24034810066223145, 0.18618792295455933, -0.6647047996520996, -0.122229665517807, -0.6593778133392334, 0.3021117150783539, -0.1535397171974182, 0.2957099378108978, -0.2063988596200943, 0.18639634549617767, 0.12765613198280334, -0.07750535756349564, 0.10924883931875229, -0.4198971688747406, -0.18771369755268097, -0.09577611833810806, 0.2693742513656616, -0.47796308994293213, -0.055133722722530365, -0.03794213756918907, 0.0814751535654068, 0.45575204491615295, 0.20855829119682312, 0.012650715187191963, 0.09506627917289734, -0.16743093729019165, -0.22884570062160492, -0.17120318114757538, -0.2424793392419815, -0.04781600460410118, -0.6381927728652954, -0.34167659282684326, -0.4831202030181885, 0.04853597655892372, 0.16847561299800873, 0.15164034068584442, -0.07459653168916702, 0.18773427605628967, 0.09917864203453064, 0.4985150694847107, 0.02120433747768402, -0.05744665488600731, 0.6246474385261536, -0.11789736151695251, -0.44503286480903625, 0.48908817768096924, -0.03532331809401512, 0.3584005832672119, 0.06729526817798615, -0.1445763260126114, -0.17467577755451202, -0.2812870442867279, 0.44760578870773315, 0.1808011829853058, 0.10797961056232452, 0.21889789402484894, -0.11514239758253098, 0.04811498150229454, -0.18779723346233368, 0.49670079350471497, 0.334112286567688, -0.09505525976419449, -0.3954769968986511, -0.156389519572258, 0.45640987157821655, -0.19788695871829987, -0.0036151763051748276, 0.25355783104896545, -0.22835582494735718, -0.28129085898399353, 0.2037985771894455, 0.16435497999191284, 1.1384713649749756, -0.09754567593336105, 0.37321898341178894, 0.12780317664146423, -0.037440236657857895, 0.7614157199859619, -0.18077360093593597, 0.052558910101652145, -0.25659966468811035, 0.14564958214759827, -0.10637703537940979, -0.05241253599524498, -0.08326452225446701, 0.08726075291633606, -0.23800896108150482, 0.07538549602031708, -0.06170401722192764, -0.13530084490776062, -0.0051054758951067924, 0.571159839630127, 0.10040008276700974, -0.3970247209072113, -0.2120526283979416, 0.09203223884105682, -0.3342668414115906, 0.3082543611526489, -0.16845138370990753, -0.23394031822681427, -0.30767351388931274, -0.14399343729019165, -0.1809542179107666, 0.28737354278564453, -0.25357165932655334, 0.07502065598964691, 0.06265701353549957, -0.330896258354187, 0.2630811035633087, 0.43935146927833557, 0.503719687461853, -0.0771641954779625, -0.09232044965028763, 0.02560354955494404, -0.13528966903686523, -0.049908675253391266, -0.026899190619587898, 0.07493110001087189, 0.19116373360157013, -0.3115493953227997, 0.07954223453998566, 0.2915779650211334, -0.025028912350535393, -0.30229324102401733, -0.18620507419109344, 0.15034233033657074, 0.09515649825334549, -0.14537641406059265, -0.3303045332431793, 0.09154950827360153, -0.18308034539222717, -0.031860340386629105, 0.04029693454504013, 0.059760868549346924, 0.04360359534621239, -0.0649104043841362, 0.0955715999007225, -0.18126723170280457, 0.07073958218097687, 0.09366316348314285, -0.0932907685637474, -0.05471505597233772, 0.7555968165397644, -0.07153259962797165, 0.09503095597028732, -0.08538433164358139, -0.21003100275993347, -0.32625430822372437, -0.1590140461921692, -0.09987189620733261, 0.27881181240081787, -0.05178218334913254, 0.08019924908876419, 0.6546945571899414, 0.3361331522464752, 0.12761598825454712, 0.1409846693277359, -0.13919930160045624, -0.34428784251213074, -0.3336314260959625, -0.025998974218964577, 0.17864979803562164, -0.10026039928197861, 0.20802229642868042, 0.1933751106262207, 0.33699774742126465, -0.24393419921398163, -0.18857641518115997, -0.1968168318271637, 0.1518179029226303, 0.7084037661552429, -0.03436828777194023, -0.06733902543783188, -0.02548195980489254, -0.00267820805311203, -0.025168485939502716, -0.170773446559906, -0.12747357785701752, -0.37970808148384094, 0.16888125240802765, 0.4832162857055664, -0.18876056373119354, -0.06686519086360931, -0.2162545919418335, -0.10766875743865967, -0.1371474713087082, -0.3264831304550171, 0.03570926934480667, -0.04568564146757126, 0.2057017832994461, -0.22629836201667786, 0.34786275029182434, -0.090091772377491, 0.20739376544952393, 0.0170756746083498, 0.1803193837404251, 0.0580897182226181, -0.040845081210136414, 0.17120887339115143, 0.06229710578918457, -0.3547424376010895, -0.03789805248379707, -0.18474212288856506, -0.1004062220454216, 0.02578580193221569, -0.4382697641849518, 0.10995222628116608, -0.15907247364521027, 0.04341196268796921, 0.03248496726155281, -0.5039015412330627, 0.1512499302625656, 0.3939330279827118, 0.10969102382659912, 0.04253867268562317, 0.12714268267154694, 0.16199155151844025, 0.2084401547908783, -0.017391663044691086, 0.2256091982126236, 0.039038967341184616, -0.265863835811615, -0.05207964405417442, -0.16882634162902832, -0.12271887809038162, -0.2723003923892975, 0.07296270877122879, -0.14059045910835266, -0.41853320598602295, -0.05919424816966057, 0.14822839200496674, 0.16802474856376648, -0.10553285479545593, 0.15312983095645905, -0.20561054348945618, 0.08979497104883194, 0.018969701603055, -0.08495764434337616, 0.23121830821037292, -0.2662845849990845, 0.07315904647111893, -0.08859801292419434, -0.032750606536865234, 0.3354671001434326, 0.1424107700586319, 0.6379612684249878, -0.3282475173473358, -0.02502335049211979, -0.24423611164093018, 0.10706325620412827, 0.00512257544323802, 0.18222734332084656, -0.6148513555526733, -0.019357090815901756, 0.042651135474443436, -0.10466207563877106, 0.030355872586369514, -0.3352692425251007, -0.0621042437851429, 0.30565348267555237, -0.17846336960792542, 0.09636345505714417, -0.14240045845508575, 0.21848930418491364, 0.16331422328948975, -0.16215969622135162, 0.4087912142276764, 0.03424404561519623, -0.10550687462091446, -0.08757492899894714, 0.08006307482719421, 0.4988904893398285, 0.5946882963180542, 0.11233969777822495, 0.3067477345466614, 0.21389812231063843, 0.037581585347652435, 0.078985795378685, 0.204816997051239, 0.43595850467681885, 0.28091001510620117, 0.11451367288827896, 0.04079459235072136, -0.12631744146347046, 0.07522374391555786, 0.031144630163908005, 0.10622459650039673, -0.2927732765674591, 0.3019168972969055, -0.010780733078718185, -0.3658277988433838, -0.17470870912075043, 0.3855644464492798, -0.46778810024261475, 0.023126307874917984, 0.2493371069431305, 0.21936698257923126, 0.33851975202560425, -0.2580968737602234, 0.03067835606634617, -0.17712587118148804, 0.4321081042289734, 0.4894350469112396, -0.08823558688163757, -0.12464883178472519, -0.28623396158218384, -0.4793255627155304, 0.18860940635204315, -0.1928083747625351, -0.21018420159816742, -0.35276103019714355, 0.18883036077022552, 0.04558388143777847, -0.10253942757844925, 0.07314874976873398, 0.09914860874414444, -0.22263766825199127, -0.029267704114317894, -0.454449862241745, 0.04934367164969444, -0.17575675249099731, -0.03394376114010811, -0.028623439371585846, -0.16214430332183838, 0.07061149924993515, 0.05983763560652733, 0.034420643001794815, 0.02024291828274727, 0.22016964852809906, 0.11978621780872345, 0.3783574402332306, 0.47854796051979065, -0.09864506125450134, 0.724032998085022, 0.08315073698759079, -0.00944974459707737, 0.026824085041880608, -0.05517999082803726, -0.17598691582679749, 0.039511922746896744, 0.1294776350259781, 0.47135043144226074, -0.07734785228967667, 0.3596561849117279, -0.24028132855892181, 0.07416427880525589, -0.0500345304608345, -0.5376129746437073, 0.08918165415525436, -0.004217804875224829, -0.04423810541629791, 0.26622524857521057, 0.4257822334766388, 0.11622243374586105, 0.1147407591342926, -0.12512753903865814, -0.4667600989341736, -0.283077210187912, 0.418811172246933, -0.5340776443481445, -0.3015986382961273, 0.19961701333522797, 0.22690314054489136, -0.06853359192609787, 0.05043429508805275, -0.3018624782562256, -0.06729265302419662, 0.1631373018026352, -0.22798317670822144, -0.015587744303047657, 0.31345611810684204, -0.4714374542236328, 0.07941395044326782, -0.03801816701889038, 0.2798067033290863, 0.20968572795391083, -0.35224273800849915, -0.04154472425580025, -0.154948428273201 ]
https://github.com/huggingface/datasets/issues/181
Cannot upload my own dataset
@lhoestq The endpoints in https://github.com/huggingface/nlp/blob/master/src/nlp/hf_api.py should be (depending on the type of file): ``` POST /api/datasets/presign GET /api/datasets/listObjs DELETE /api/datasets/deleteObj POST /api/metrics/presign GET /api/metrics/listObjs DELETE /api/metrics/deleteObj ``` In addition to this, @thomwolf cleaned up the objects with dataclasses but you should revert this and re-align to the hf_api that's in this branch of transformers: https://github.com/huggingface/transformers/pull/4632 (so that potential new JSON attributes in the API output don't break existing versions of any library)
I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ```
72
Cannot upload my own dataset I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ``` @lhoestq The endpoints in https://github.com/huggingface/nlp/blob/master/src/nlp/hf_api.py should be (depending on the type of file): ``` POST /api/datasets/presign GET /api/datasets/listObjs DELETE /api/datasets/deleteObj POST /api/metrics/presign GET /api/metrics/listObjs DELETE /api/metrics/deleteObj ``` In addition to this, @thomwolf cleaned up the objects with dataclasses but you should revert this and re-align to the hf_api that's in this branch of transformers: https://github.com/huggingface/transformers/pull/4632 (so that potential new JSON attributes in the API output don't break existing versions of any library)
[ 0.09444843232631683, -0.030215099453926086, 0.047693099826574326, -0.021573474630713463, 0.15865516662597656, -0.19878609478473663, 0.3093150854110718, -0.028614412993192673, -0.20045983791351318, 0.17955058813095093, -0.14158780872821808, 0.26487016677856445, -0.21903906762599945, 0.2876109778881073, 0.3653644025325775, 0.08437500149011612, -0.012491869740188122, 0.28119343519210815, -0.014132038690149784, -0.0984327495098114, -0.21194270253181458, 0.10448187589645386, 0.032021664083004, -0.04935764893889427, -0.11954166740179062, -0.17534534633159637, -0.09639151394367218, 0.2744736075401306, -0.20591913163661957, -0.3228282332420349, 0.4480946660041809, 0.22257733345031738, 0.30274707078933716, 0.577003538608551, -0.00011726109369192272, -0.04289494827389717, 0.1430101990699768, -0.22344964742660522, -0.6293329000473022, -0.38263648748397827, 0.07990779727697372, -0.1367632895708084, 0.07053184509277344, -0.2529914081096649, 0.19080835580825806, 0.041705381125211716, 0.0007069570128805935, 0.06395947188138962, 0.5112627148628235, 0.3838021457195282, 0.15527501702308655, 0.08335090428590775, -0.19444267451763153, -0.08382143825292587, 0.03142401948571205, 0.18943150341510773, -0.044448841363191605, 0.29738882184028625, 0.36469700932502747, -0.399980753660202, 0.2879972755908966, -0.19582869112491608, -0.004945149179548025, -0.09528350830078125, 0.47748538851737976, -0.024430278688669205, 0.0022098885383456945, -0.4087411165237427, -0.004733027424663305, 0.14790265262126923, 0.2761019170284271, -0.04426797106862068, -0.34744665026664734, -0.10564188659191132, 0.14628273248672485, -0.4893600642681122, 0.25155338644981384, 0.3014509677886963, -0.4247397184371948, -0.029061494395136833, -0.2817883789539337, 0.32458725571632385, -0.41372907161712646, 0.4369986355304718, 0.11532535403966904, 0.3545939028263092, -0.021374084055423737, 0.07461932301521301, 0.2307596057653427, -0.13691888749599457, -0.02224859781563282, 0.07378754019737244, 0.012210077606141567, 0.28070199489593506, -0.1889120489358902, -0.18760992586612701, -0.01863323710858822, -0.28084832429885864, -0.09787121415138245, 0.11185163259506226, 0.4561244249343872, -0.13783949613571167, -0.23980040848255157, 0.15594901144504547, 0.09088446199893951, 0.32344213128089905, -0.006549427285790443, 0.09779734164476395, 0.12052928656339645, 0.0003338170354254544, 0.28369060158729553, -0.06052282080054283, -0.27953943610191345, -0.07793816179037094, -0.14036314189434052, -0.03552086651325226, 0.2655665874481201, 0.08274928480386734, -0.0020344641525298357, -0.0889303907752037, 0.06566981971263885, 0.06152790039777756, -0.08327460289001465, 0.3429426848888397, -0.08083100616931915, -0.2568763792514801, -0.004063366446644068, 0.25254544615745544, 0.14727097749710083, -0.26811543107032776, 0.07059146463871002, 0.32219526171684265, -0.12113519012928009, -0.14960528910160065, 0.40277335047721863, 0.01671188324689865, 0.20685142278671265, -0.05475425347685814, 0.16550391912460327, -0.14268659055233002, 0.06256692856550217, 0.01568044163286686, -0.15282835066318512, 0.03313657268881798, 0.23737314343452454, 0.35915085673332214, -0.07919708639383316, -0.5343111753463745, -0.1254030019044876, 0.049367886036634445, -0.38119015097618103, -0.3526667058467865, -0.45391860604286194, 0.09591714292764664, 0.04416825622320175, -0.22488218545913696, -0.3203067183494568, -0.10148455202579498, 0.09989216923713684, -0.26112425327301025, -0.07177548110485077, -0.026748761534690857, -0.11899155378341675, -0.12630924582481384, 0.12081341445446014, 0.03390290588140488, -0.4636066257953644, 0.2092759609222412, -0.046157028526067734, 0.5230323672294617, 0.29660844802856445, 0.28856420516967773, -0.3173556625843048, 0.2671288549900055, -0.06915697455406189, 0.18908919394016266, 0.5893783569335938, -0.4831951856613159, -0.2531137466430664, 0.050312262028455734, -0.037071384489536285, -0.11455698311328888, 0.029296545311808586, 0.1607624739408493, 0.08552654832601547, -0.09906060993671417, 0.07839330285787582, 0.42903444170951843, 0.07837430387735367, 0.04569414258003235, -0.24622973799705505, -0.18085865676403046, 0.224577397108078, 0.030915699899196625, -0.17134493589401245, 0.0790286436676979, 0.26193147897720337, 0.05616142600774765, 0.02804369293153286, -0.34000593423843384, 0.14545132219791412, 0.1951552778482437, 0.33123818039894104, -0.16132688522338867, -0.14345738291740417, 0.24710357189178467, -0.4086432456970215, 0.18435846269130707, 0.11573022603988647, 0.3616282343864441, -0.14487585425376892, 0.002963240025565028, -0.42881831526756287, -0.09242136776447296, -0.08533667773008347, -0.03767787665128708, 0.06950760632753372, 0.09063832461833954, 0.06854327768087387, -0.16730612516403198, -0.22225156426429749, 0.11700570583343506, -0.3020484149456024, 0.0560997799038887, -0.18789005279541016, 0.06197570264339447, -0.2840302288532257, -0.11223091185092926, 0.22809508442878723, 0.18706244230270386, 0.20362743735313416, -0.015008680522441864, -0.1789752095937729, 0.09054966270923615, -0.4720057547092438, 0.3271062970161438, 0.04591049998998642, 0.1694822907447815, 0.023193662986159325, -0.026097606867551804, -0.21184560656547546, -0.11862582713365555, 0.18914294242858887, -0.15958721935749054, -0.32982540130615234, 0.1566563844680786, -0.02223804220557213, 0.008161724545061588, 0.16949927806854248, 0.012336867861449718, 0.28302493691444397, -0.19770628213882446, 0.11460234224796295, 0.1360907405614853, 0.4101683795452118, 0.09705732017755508, 0.3762931823730469, -0.037503041326999664, -0.13971656560897827, -0.0750536099076271, 0.15518172085285187, 0.00046218084753490984, 0.1710677146911621, 0.0722014307975769, 0.02527892030775547, -0.094702810049057, 0.288423091173172, 0.5951281189918518, 0.35346856713294983, 0.07945892214775085, -0.05248229205608368, -0.04475058987736702, -0.11151432245969772, -0.22756001353263855, -0.016284378245472908, -0.07748296111822128, 0.04854539409279823, 0.21171681582927704, 0.43869173526763916, 0.26055291295051575, -0.13261158764362335, -0.016882631927728653, 0.11479952931404114, 0.4540286958217621, -0.19068312644958496, 0.26633220911026, -0.09921867400407791, -0.4326513409614563, 0.2592723071575165, -0.2881086468696594, -0.08621837943792343, -0.182784765958786, -0.18338283896446228, 0.12537148594856262, 0.0859251543879509, -0.1114078164100647, 0.1523815542459488, 0.351053386926651, -0.09527862071990967, -0.27169281244277954, -0.2880833148956299, -0.1030789315700531, -0.20516343414783478, 0.04893951490521431, 0.3304884433746338, -0.11064460873603821, 0.41045957803726196, 0.34332275390625, 0.06626499444246292, -0.04624009504914284, -0.32678014039993286, -0.0032144642900675535, -0.04425834119319916, 0.09379108250141144, 0.0913999155163765, 0.33954891562461853, 0.004957741592079401, -0.5441508293151855, 0.18702338635921478, 0.05857492610812187, -0.24736614525318146, 0.259773850440979, -0.03783915191888809, -0.3145045340061188, -0.49806341528892517, 0.010082417167723179, -0.2758827209472656, -0.3460385501384735, 0.23943129181861877, 0.33021280169487, 0.23633921146392822, 0.593953549861908, 0.018189238384366035, 0.131915882229805, 0.31638675928115845, 0.13424095511436462, 0.03354407846927643, -0.20578527450561523, 0.48493099212646484, -0.2360689342021942, -0.43476200103759766, 0.1069926843047142, -0.1374835968017578, 0.5780153274536133, -0.04567057639360428, -0.39297613501548767, -0.19886080920696259, 0.09783364087343216, 0.2063959240913391, -0.06614187359809875, -0.08298248797655106, 0.2729761600494385, -0.1973799765110016, 0.039316192269325256, -0.06291264295578003, -0.2984916567802429, 0.15413492918014526, 0.6249305009841919, 0.2360728234052658, -0.07148406654596329, 0.4156576693058014, 0.050727952271699905, 0.6994932889938354, 0.08518002182245255, -0.3150878846645355, 0.4526676535606384, -0.03172583132982254, 0.17719611525535583, -0.13752295076847076, -0.3402385115623474, 0.09549253433942795, 0.23747211694717407, -0.04053667187690735, 0.1565457582473755, -0.028311360627412796, -0.0831950232386589, -0.4500623345375061, 0.22518055140972137, -0.25933757424354553, 0.04975167289376259, 0.01460089161992073, 0.0814526379108429, 0.03887375444173813, -0.09079187363386154, -0.2421218305826187, -0.35652685165405273, -0.5098559260368347, -0.12958145141601562, 0.2505665719509125, -0.23419040441513062, 0.18313686549663544, -0.6524077653884888, -0.1037067323923111, -0.6693301796913147, 0.2962004244327545, -0.14688903093338013, 0.27544599771499634, -0.21646258234977722, 0.18662232160568237, 0.13347889482975006, -0.08640004694461823, 0.0892622247338295, -0.4160825312137604, -0.17348743975162506, -0.09920826554298401, 0.29538002610206604, -0.46098917722702026, -0.04646872356534004, -0.03333204612135887, 0.08325408399105072, 0.484587162733078, 0.21635542809963226, 0.020408520475029945, 0.09795600175857544, -0.17281438410282135, -0.2555900514125824, -0.16451409459114075, -0.2275293916463852, -0.03326886147260666, -0.6557897329330444, -0.34370601177215576, -0.4793671667575836, 0.07115917652845383, 0.1532265841960907, 0.1519203633069992, -0.07051805406808853, 0.22515316307544708, 0.1137518361210823, 0.49375858902931213, 0.01778639294207096, -0.0832149013876915, 0.6168342232704163, -0.12393555045127869, -0.4665551781654358, 0.49015679955482483, -0.062034521251916885, 0.3874392807483673, 0.07472054660320282, -0.137711763381958, -0.1564883142709732, -0.3009519875049591, 0.44125694036483765, 0.1783338189125061, 0.10966116189956665, 0.21357350051403046, -0.13211655616760254, 0.05214957520365715, -0.19166453182697296, 0.47335830330848694, 0.33841729164123535, -0.09716443717479706, -0.37715110182762146, -0.13973112404346466, 0.44671180844306946, -0.2157983034849167, -0.007921336218714714, 0.23950178921222687, -0.21200741827487946, -0.2822958528995514, 0.21089613437652588, 0.1703662872314453, 1.1364245414733887, -0.07740283012390137, 0.3717598021030426, 0.13204924762248993, -0.05455544590950012, 0.756165623664856, -0.189636692404747, 0.024763785302639008, -0.24850215017795563, 0.17459559440612793, -0.10031937062740326, -0.044815290719270706, -0.08898164331912994, 0.11526598036289215, -0.2281901091337204, 0.057402901351451874, -0.04674174264073372, -0.16363777220249176, 0.007388527505099773, 0.5417446494102478, 0.09941370785236359, -0.37804168462753296, -0.2032104730606079, 0.10226572304964066, -0.3348422348499298, 0.300632506608963, -0.16814744472503662, -0.23469892144203186, -0.3243204653263092, -0.16216176748275757, -0.1864951103925705, 0.3050612211227417, -0.2374933362007141, 0.08102388679981232, 0.05311445891857147, -0.3184695541858673, 0.25533443689346313, 0.43584439158439636, 0.5033402442932129, -0.1063317060470581, -0.07909493893384933, 0.03086026944220066, -0.13777223229408264, -0.029955264180898666, 0.00021585235663224012, 0.07229527086019516, 0.17591899633407593, -0.3124386966228485, 0.06527704745531082, 0.296657919883728, -0.014302980154752731, -0.3096219599246979, -0.17401352524757385, 0.14716756343841553, 0.07625935971736908, -0.1434604674577713, -0.3124759793281555, 0.09920909255743027, -0.1777138113975525, -0.024742700159549713, 0.041210904717445374, 0.06617726385593414, 0.035941094160079956, -0.07256115972995758, 0.07392632216215134, -0.18735934793949127, 0.06512017548084259, 0.07385565340518951, -0.11527114361524582, -0.061694324016571045, 0.7477405071258545, -0.06718622148036957, 0.09660428017377853, -0.09921867400407791, -0.23556290566921234, -0.3396821618080139, -0.1887063831090927, -0.10450910031795502, 0.268494188785553, -0.04515475779771805, 0.06955849379301071, 0.6650156378746033, 0.3328908681869507, 0.14161746203899384, 0.16352076828479767, -0.1516106128692627, -0.3325672745704651, -0.3406677842140198, -0.03344905003905296, 0.16924843192100525, -0.08270024508237839, 0.19785837829113007, 0.20175254344940186, 0.3437690734863281, -0.2551347613334656, -0.1751261055469513, -0.213116854429245, 0.16516424715518951, 0.7187826633453369, -0.05416596308350563, -0.06109904497861862, -0.01915862038731575, 0.0032395923044532537, -0.044167179614305496, -0.1669817417860031, -0.13342894613742828, -0.3722269535064697, 0.16226893663406372, 0.4754086136817932, -0.20718461275100708, -0.07411085814237595, -0.20155292749404907, -0.10357882082462311, -0.12503787875175476, -0.3151991665363312, 0.04092114418745041, -0.05833408236503601, 0.21087117493152618, -0.2468242049217224, 0.35262587666511536, -0.0880851149559021, 0.19500480592250824, 0.019292667508125305, 0.18320366740226746, 0.07253312319517136, -0.024480164051055908, 0.164117693901062, 0.05056849122047424, -0.35458847880363464, -0.03148438036441803, -0.20351652801036835, -0.09130875021219254, 0.04711661487817764, -0.4444272518157959, 0.11381764709949493, -0.16836608946323395, 0.04783990979194641, 0.02164587564766407, -0.5189089775085449, 0.14512744545936584, 0.38480204343795776, 0.12011399120092392, 0.04832201823592186, 0.12300380319356918, 0.168443962931633, 0.21478180587291718, -0.010158488526940346, 0.2320229560136795, 0.062190908938646317, -0.2759912312030792, -0.05973144620656967, -0.1776445209980011, -0.10463615506887436, -0.26302629709243774, 0.07028733938932419, -0.12691187858581543, -0.4082081913948059, -0.05109411105513573, 0.13153181970119476, 0.1713169813156128, -0.1107318103313446, 0.1462879478931427, -0.2145097404718399, 0.09462927281856537, 0.028082218021154404, -0.08880063891410828, 0.2282719612121582, -0.2685786485671997, 0.07813689112663269, -0.10730578750371933, -0.04195205494761467, 0.33544427156448364, 0.14839206635951996, 0.6253678798675537, -0.30057933926582336, -0.02767796814441681, -0.24202090501785278, 0.14442461729049683, -0.0007231162744574249, 0.17657244205474854, -0.6177631616592407, -0.028876246884465218, 0.050703346729278564, -0.10465004295110703, 0.028325475752353668, -0.3496128022670746, -0.06082197651267052, 0.3098294138908386, -0.17741812765598297, 0.11428356915712357, -0.15285003185272217, 0.21429720520973206, 0.14840394258499146, -0.15133193135261536, 0.42992478609085083, 0.059626635164022446, -0.10391702502965927, -0.08217435330152512, 0.07708939164876938, 0.5147436857223511, 0.5829105973243713, 0.11593953520059586, 0.31674519181251526, 0.21503454446792603, 0.028208015486598015, 0.07165045291185379, 0.216790571808815, 0.43436065316200256, 0.2830870449542999, 0.1265617161989212, 0.04775439575314522, -0.12174860388040543, 0.07318612933158875, 0.05191708356142044, 0.1230233758687973, -0.2741186022758484, 0.30079132318496704, 0.0012679644860327244, -0.3716162443161011, -0.19212456047534943, 0.3825387954711914, -0.4631412923336029, 0.03657710924744606, 0.21874862909317017, 0.2108301967382431, 0.3433952331542969, -0.2772810757160187, 0.038886092603206635, -0.16345813870429993, 0.436238169670105, 0.4967527389526367, -0.06773627549409866, -0.12502819299697876, -0.28918665647506714, -0.4506285488605499, 0.18131107091903687, -0.19864536821842194, -0.229953333735466, -0.3568551540374756, 0.1961071491241455, 0.04849028214812279, -0.10085778683423996, 0.05383499711751938, 0.09990976005792618, -0.21706265211105347, -0.01697269268333912, -0.4612068831920624, 0.06106322631239891, -0.1587815135717392, -0.01481538824737072, -0.02872912399470806, -0.14931762218475342, 0.06392396986484528, 0.07336728274822235, 0.04292561486363411, 0.029615676030516624, 0.21092501282691956, 0.10708532482385635, 0.38921597599983215, 0.47182756662368774, -0.101585254073143, 0.7334184646606445, 0.0889015793800354, -0.002749854465946555, 0.012415150180459023, -0.051011715084314346, -0.17123179137706757, 0.02972528524696827, 0.1415664553642273, 0.4371808171272278, -0.08350149542093277, 0.35412919521331787, -0.2191765010356903, 0.053227584809064865, -0.051956869661808014, -0.5288753509521484, 0.10355649143457413, -0.005815076641738415, -0.03416036069393158, 0.2721147835254669, 0.40909549593925476, 0.11103535443544388, 0.11265738308429718, -0.10881705582141876, -0.4588083326816559, -0.2800886034965515, 0.3989618122577667, -0.5402225255966187, -0.3015979528427124, 0.19581826031208038, 0.23403999209403992, -0.060050830245018005, 0.04783572256565094, -0.3058384954929352, -0.059673972427845, 0.16777783632278442, -0.2149411141872406, -0.009056609123945236, 0.319765567779541, -0.4719265401363373, 0.0773255005478859, -0.03231985494494438, 0.2738976776599884, 0.20475371181964874, -0.35823583602905273, -0.03778081014752388, -0.16721460223197937 ]
https://github.com/huggingface/datasets/issues/181
Cannot upload my own dataset
New commands are ``` nlp-cli upload_dataset <path/to/dataset> nlp-cli upload_metric <path/to/metric> nlp-cli s3_datasets {rm, ls} nlp-cli s3_metrics {rm, ls} ``` Closing this issue.
I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ```
22
Cannot upload my own dataset I look into `nlp-cli` and `user.py` to learn how to upload my own data. It is supposed to work like this - Register to get username, password at huggingface.co - `nlp-cli login` and type username, passworld - I have a single file to upload at `./ttc/ttc_freq_extra.csv` - `nlp-cli upload ttc/ttc_freq_extra.csv` But I got this error. ``` 2020-05-21 16:33:52.722464: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1 About to upload file /content/ttc/ttc_freq_extra.csv to S3 under filename ttc/ttc_freq_extra.csv and namespace korakot Proceed? [Y/n] y Uploading... This might take a while if files are large Traceback (most recent call last): File "/usr/local/bin/nlp-cli", line 33, in <module> service.run() File "/usr/local/lib/python3.6/dist-packages/nlp/commands/user.py", line 234, in run token=token, filename=filename, filepath=filepath, organization=self.args.organization File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 141, in presign_and_upload urls = self.presign(token, filename=filename, organization=organization) File "/usr/local/lib/python3.6/dist-packages/nlp/hf_api.py", line 132, in presign return PresignedUrl(**d) TypeError: __init__() got an unexpected keyword argument 'cdn' ``` New commands are ``` nlp-cli upload_dataset <path/to/dataset> nlp-cli upload_metric <path/to/metric> nlp-cli s3_datasets {rm, ls} nlp-cli s3_metrics {rm, ls} ``` Closing this issue.
[ 0.10846686363220215, -0.04407903924584389, 0.05040140077471733, -0.02760578878223896, 0.1381332129240036, -0.18117083609104156, 0.326691597700119, -0.009185325354337692, -0.19502303004264832, 0.1800154745578766, -0.1299545168876648, 0.23764589428901672, -0.23067700862884521, 0.2919650375843048, 0.36800044775009155, 0.07485684007406235, 0.0062186154536902905, 0.2890031635761261, 0.011849310249090195, -0.10124019533395767, -0.21110163629055023, 0.10817064344882965, 0.035999517887830734, -0.06653718650341034, -0.1229436993598938, -0.17405153810977936, -0.10020694881677628, 0.27096477150917053, -0.2011144459247589, -0.3351971507072449, 0.44026118516921997, 0.22429464757442474, 0.32153770327568054, 0.5771898031234741, -0.00011768120020860806, -0.053284939378499985, 0.13914425671100616, -0.2255367487668991, -0.6261977553367615, -0.36731237173080444, 0.08496621996164322, -0.17352649569511414, 0.08267373591661453, -0.2572456896305084, 0.1832292079925537, 0.025091534480452538, 0.00859874952584505, 0.07702253758907318, 0.5234441161155701, 0.36876657605171204, 0.1549510806798935, 0.09582094103097916, -0.1830166131258011, -0.09871897101402283, 0.029122162610292435, 0.18727968633174896, -0.04247983917593956, 0.29659441113471985, 0.3687254786491394, -0.3961859941482544, 0.3029371201992035, -0.18616318702697754, -0.008992482908070087, -0.08873268216848373, 0.4844443202018738, -0.022030899301171303, 0.006850079633295536, -0.40278059244155884, -0.008083040826022625, 0.1580435335636139, 0.27216383814811707, -0.03827808052301407, -0.3466472625732422, -0.11687161773443222, 0.15635859966278076, -0.4871677756309509, 0.24212060868740082, 0.28327709436416626, -0.4218558669090271, -0.04651590809226036, -0.25824227929115295, 0.31141313910484314, -0.41079214215278625, 0.4234360456466675, 0.1253177374601364, 0.35030317306518555, -0.026140790432691574, 0.06101497262716293, 0.25930145382881165, -0.14530310034751892, -0.02124325931072235, 0.09257052093744278, 0.017638584598898888, 0.2819576561450958, -0.1919320672750473, -0.1905704140663147, -0.021572936326265335, -0.2899376451969147, -0.10397447645664215, 0.10910069942474365, 0.4545294940471649, -0.13084636628627777, -0.22314877808094025, 0.15165263414382935, 0.08718416839838028, 0.3312406837940216, 0.008624504320323467, 0.07960421591997147, 0.11092185229063034, 0.01273246482014656, 0.2880074083805084, -0.06930067390203476, -0.27576693892478943, -0.07849889248609543, -0.13718993961811066, -0.048890288919210434, 0.2654901146888733, 0.08655624836683273, 0.0027185112703591585, -0.08611790835857391, 0.07163924723863602, 0.03991139307618141, -0.07255198061466217, 0.33466601371765137, -0.09973735362291336, -0.2767595648765564, 0.008246120996773243, 0.24410729110240936, 0.12901826202869415, -0.2690322995185852, 0.07090898603200912, 0.30144989490509033, -0.12323897331953049, -0.1411055475473404, 0.4131867289543152, 0.002604223322123289, 0.19098684191703796, -0.04976858198642731, 0.1741701364517212, -0.11572030931711197, 0.07259535044431686, 0.026877403259277344, -0.15458202362060547, 0.026117052882909775, 0.23558521270751953, 0.3750024437904358, -0.07226500660181046, -0.5426656007766724, -0.12554562091827393, 0.04687367379665375, -0.37620678544044495, -0.348122775554657, -0.449422687292099, 0.09366770088672638, 0.0353323370218277, -0.23436014354228973, -0.2903648912906647, -0.08445347100496292, 0.09920245409011841, -0.27377137541770935, -0.08804371953010559, -0.017819838598370552, -0.10808909684419632, -0.14024370908737183, 0.11493581533432007, 0.05397984758019447, -0.4861149191856384, 0.2089095562696457, -0.06266996264457703, 0.513792097568512, 0.30137181282043457, 0.28101080656051636, -0.3043970465660095, 0.25999054312705994, -0.06236555054783821, 0.1773793250322342, 0.5754053592681885, -0.4743400514125824, -0.25073763728141785, 0.034278757870197296, -0.03656146302819252, -0.10301212966442108, 0.054686177521944046, 0.16820470988750458, 0.0883634015917778, -0.10266026854515076, 0.07414025813341141, 0.4095371961593628, 0.07595779001712799, 0.03001254051923752, -0.22451426088809967, -0.17478151619434357, 0.25426867604255676, 0.012387936003506184, -0.1504039168357849, 0.07483946532011032, 0.2523055076599121, 0.0847451239824295, 0.030770158395171165, -0.33520960807800293, 0.1500527411699295, 0.20149841904640198, 0.33596450090408325, -0.13525719940662384, -0.13722974061965942, 0.23809470236301422, -0.4048445522785187, 0.19656269252300262, 0.12716296315193176, 0.3596288859844208, -0.11084041744470596, 0.006236647721379995, -0.42009133100509644, -0.11388055235147476, -0.095737025141716, -0.03177253529429436, 0.06666672974824905, 0.09637987613677979, 0.07029160857200623, -0.18559114634990692, -0.23020467162132263, 0.1288207769393921, -0.3153892755508423, 0.061057768762111664, -0.19712720811367035, 0.05917000770568848, -0.2863599956035614, -0.11160070449113846, 0.21818305552005768, 0.17532144486904144, 0.20732025802135468, -0.01801823452115059, -0.17255271971225739, 0.089498370885849, -0.46536704897880554, 0.30873700976371765, 0.012426805682480335, 0.1415528953075409, 0.033568862825632095, -0.02158357761800289, -0.19961975514888763, -0.11788244545459747, 0.17727231979370117, -0.17059771716594696, -0.3431077003479004, 0.1499732881784439, -0.03314851224422455, 0.019479259848594666, 0.1668228954076767, -0.002481972798705101, 0.2705672085285187, -0.19244827330112457, 0.10810335725545883, 0.11901237815618515, 0.39312803745269775, 0.10056769847869873, 0.3735921382904053, -0.06234974414110184, -0.12617255747318268, -0.08031732589006424, 0.17376285791397095, 0.01459315326064825, 0.1585075706243515, 0.06848929077386856, 0.03325968608260155, -0.07810266315937042, 0.2860080897808075, 0.6209812164306641, 0.35156089067459106, 0.09832240641117096, -0.04840410128235817, -0.05061180889606476, -0.1180218979716301, -0.2226051390171051, -0.03233442083001137, -0.08653225749731064, 0.04504851996898651, 0.21368926763534546, 0.42927277088165283, 0.26288384199142456, -0.13860440254211426, 0.0028302406426519156, 0.11219168454408646, 0.441070020198822, -0.18067660927772522, 0.24720029532909393, -0.11076132953166962, -0.4309423863887787, 0.23817889392375946, -0.28240859508514404, -0.08930676430463791, -0.164419487118721, -0.19531308114528656, 0.11842501163482666, 0.11353164166212082, -0.12695936858654022, 0.16629907488822937, 0.3391794264316559, -0.11040627211332321, -0.25449445843696594, -0.2794031798839569, -0.08828532695770264, -0.2085285782814026, 0.05295977368950844, 0.3282945156097412, -0.11787407100200653, 0.4269637167453766, 0.3558066785335541, 0.06698260456323624, -0.048253122717142105, -0.34660568833351135, -0.012719860300421715, -0.06370187550783157, 0.07861333340406418, 0.0849282294511795, 0.34388962388038635, -0.0008328958647325635, -0.5517417192459106, 0.19058850407600403, 0.06303270906209946, -0.2525029480457306, 0.24601833522319794, -0.021941253915429115, -0.3063010573387146, -0.4941294491291046, -0.007458256557583809, -0.27835220098495483, -0.34387919306755066, 0.25040802359580994, 0.35123714804649353, 0.23518332839012146, 0.6135788559913635, 0.02083003893494606, 0.139483243227005, 0.31701117753982544, 0.1529635190963745, 0.033169668167829514, -0.18742351233959198, 0.4657982289791107, -0.253059446811676, -0.4329995810985565, 0.1104678362607956, -0.13290517032146454, 0.5748021602630615, -0.04160069674253464, -0.36591243743896484, -0.19975855946540833, 0.09144122898578644, 0.18816474080085754, -0.07725860923528671, -0.06613484770059586, 0.28565865755081177, -0.18216510117053986, 0.042107660323381424, -0.059119053184986115, -0.31170520186424255, 0.1501697599887848, 0.634050190448761, 0.2549777925014496, -0.07658562809228897, 0.42161354422569275, 0.04242322966456413, 0.7029531598091125, 0.0894530639052391, -0.3097902834415436, 0.4549662470817566, -0.028219074010849, 0.19479712843894958, -0.13296891748905182, -0.3579261600971222, 0.09596799314022064, 0.23466476798057556, -0.030786579474806786, 0.13775242865085602, -0.027219032868742943, -0.06776484847068787, -0.44664323329925537, 0.24750760197639465, -0.25212833285331726, 0.060791872441768646, -0.0016710951458662748, 0.1091473251581192, 0.04245854541659355, -0.10231301933526993, -0.23655220866203308, -0.3656546473503113, -0.5041989684104919, -0.12927523255348206, 0.24986431002616882, -0.2540321946144104, 0.19691483676433563, -0.6588916182518005, -0.12373936176300049, -0.6575391292572021, 0.28806063532829285, -0.14811094105243683, 0.2594555914402008, -0.21375109255313873, 0.17713625729084015, 0.13104209303855896, -0.0893455445766449, 0.11717246472835541, -0.41821831464767456, -0.1867838352918625, -0.09565205127000809, 0.2866272032260895, -0.4773975610733032, -0.05347100645303726, -0.04376180097460747, 0.08510912954807281, 0.4654023051261902, 0.22015418112277985, 0.03126194700598717, 0.0952756330370903, -0.1582920104265213, -0.24406804144382477, -0.1730843186378479, -0.2345629632472992, -0.04077920317649841, -0.6392327547073364, -0.3375431001186371, -0.4886775612831116, 0.04947761446237564, 0.15703986585140228, 0.1608157455921173, -0.07531002163887024, 0.21244412660598755, 0.11879786103963852, 0.4977768659591675, 0.022434018552303314, -0.06972049921751022, 0.6355298757553101, -0.1225438192486763, -0.44986531138420105, 0.48968493938446045, -0.0535588264465332, 0.37191247940063477, 0.050603389739990234, -0.12606146931648254, -0.18032985925674438, -0.29894495010375977, 0.4641976058483124, 0.18745794892311096, 0.11531272530555725, 0.21370334923267365, -0.11608636379241943, 0.049450840801000595, -0.1887468695640564, 0.47350144386291504, 0.34728601574897766, -0.09279077500104904, -0.37483152747154236, -0.16618962585926056, 0.44636350870132446, -0.18877387046813965, -0.0016304736491292715, 0.22985850274562836, -0.2005375772714615, -0.2884901463985443, 0.1962747871875763, 0.17903350293636322, 1.1350759267807007, -0.09616310149431229, 0.37867799401283264, 0.13602028787136078, -0.05242533981800079, 0.7858072519302368, -0.18570853769779205, 0.03413795679807663, -0.2420796900987625, 0.16550908982753754, -0.10266633331775665, -0.03763779252767563, -0.08714635670185089, 0.1050228402018547, -0.23912164568901062, 0.06721735745668411, -0.05727611109614372, -0.16555151343345642, -0.006033556070178747, 0.5472071170806885, 0.09571029245853424, -0.39484384655952454, -0.19912967085838318, 0.10654102265834808, -0.33136892318725586, 0.2973271310329437, -0.1884974241256714, -0.23479127883911133, -0.29894348978996277, -0.15512214601039886, -0.1712498515844345, 0.30104056000709534, -0.23888616263866425, 0.08907769620418549, 0.054780662059783936, -0.32584288716316223, 0.24529248476028442, 0.42029809951782227, 0.5057647228240967, -0.0947711244225502, -0.07163029164075851, 0.015904340893030167, -0.12948524951934814, -0.05805828794836998, -0.016992727294564247, 0.0785607174038887, 0.204513818025589, -0.3050248622894287, 0.06459339708089828, 0.28322574496269226, -0.004202621523290873, -0.2958988845348358, -0.1927383542060852, 0.17007623612880707, 0.10598528385162354, -0.14506235718727112, -0.32288095355033875, 0.10752295702695847, -0.1985757201910019, -0.023927772417664528, 0.043140243738889694, 0.05214391276240349, 0.02309442311525345, -0.06285283714532852, 0.07580862939357758, -0.1846112459897995, 0.06752022355794907, 0.09131108969449997, -0.08641243726015091, -0.08536357432603836, 0.7531553506851196, -0.0654556006193161, 0.10210202634334564, -0.09430187195539474, -0.23572085797786713, -0.344440758228302, -0.16373004019260406, -0.10431695729494095, 0.28870391845703125, -0.04541276767849922, 0.10916069895029068, 0.6310175657272339, 0.3479664623737335, 0.11863742023706436, 0.14224302768707275, -0.13004614412784576, -0.33897271752357483, -0.3504672348499298, -0.022793253883719444, 0.17645171284675598, -0.09358447790145874, 0.1920360028743744, 0.20667432248592377, 0.33862990140914917, -0.2541729509830475, -0.18684673309326172, -0.18249186873435974, 0.1629975289106369, 0.7249687910079956, -0.046619150787591934, -0.06739484518766403, -0.02037704922258854, 0.0019189182203263044, -0.040756627917289734, -0.1685422658920288, -0.13428914546966553, -0.38078394532203674, 0.16128729283809662, 0.47966650128364563, -0.202500581741333, -0.07540290802717209, -0.2050624042749405, -0.09480413794517517, -0.11984473466873169, -0.305656760931015, 0.043924156576395035, -0.0480843223631382, 0.22156158089637756, -0.232159823179245, 0.33715230226516724, -0.0559336394071579, 0.18995922803878784, 0.03067093901336193, 0.1726495772600174, 0.07241968810558319, -0.03402082622051239, 0.16806064546108246, 0.057489052414894104, -0.3541553318500519, -0.03275146707892418, -0.19122721254825592, -0.103845976293087, 0.029014505445957184, -0.44629940390586853, 0.1061885729432106, -0.17603440582752228, 0.04980066046118736, 0.022434690967202187, -0.4888654053211212, 0.13809145987033844, 0.3895101249217987, 0.11818040162324905, 0.03648433834314346, 0.11728806048631668, 0.16463932394981384, 0.218766987323761, -0.018171221017837524, 0.22868132591247559, 0.04256490245461464, -0.2698046565055847, -0.06452585756778717, -0.17235445976257324, -0.12308377027511597, -0.27268898487091064, 0.08677578717470169, -0.11697102338075638, -0.3999553322792053, -0.06926930695772171, 0.15473715960979462, 0.1715613603591919, -0.097404345870018, 0.14381533861160278, -0.20515550673007965, 0.10718589276075363, 0.03388464078307152, -0.08608327060937881, 0.22095447778701782, -0.2704433500766754, 0.08013695478439331, -0.08750218898057938, -0.027106905356049538, 0.3488442301750183, 0.14098139107227325, 0.6430829763412476, -0.30079030990600586, -0.038441840559244156, -0.23728369176387787, 0.12414265424013138, -0.0008916539372876287, 0.19300639629364014, -0.6118702292442322, -0.016116509214043617, 0.040175631642341614, -0.10860490798950195, 0.03986545652151108, -0.35414376854896545, -0.06759520620107651, 0.309472918510437, -0.17172783613204956, 0.11386804282665253, -0.1698199063539505, 0.20617708563804626, 0.14696478843688965, -0.14845134317874908, 0.39706501364707947, 0.054240211844444275, -0.10305584967136383, -0.07449574768543243, 0.08739268779754639, 0.5002291798591614, 0.6005570292472839, 0.11638987809419632, 0.32311055064201355, 0.20161046087741852, 0.038133248686790466, 0.07256437838077545, 0.2018141895532608, 0.4283689260482788, 0.27456265687942505, 0.1099807471036911, 0.04795939102768898, -0.12619198858737946, 0.06361647695302963, 0.04795260727405548, 0.09902076423168182, -0.29639777541160583, 0.29422879219055176, 0.0014727204106748104, -0.373447448015213, -0.1934806853532791, 0.3822582960128784, -0.48805636167526245, 0.028024353086948395, 0.24869614839553833, 0.215536966919899, 0.33565637469291687, -0.27084383368492126, 0.036630239337682724, -0.16707256436347961, 0.43749481439590454, 0.48262348771095276, -0.0939960852265358, -0.12573960423469543, -0.2834653854370117, -0.465683251619339, 0.17053495347499847, -0.19895115494728088, -0.22324588894844055, -0.3570999503135681, 0.20041172206401825, 0.04287050664424896, -0.12530548870563507, 0.0718621015548706, 0.0931006446480751, -0.22871510684490204, -0.058963388204574585, -0.472491592168808, 0.05648230016231537, -0.135480597615242, -0.02128111943602562, -0.02528051659464836, -0.1377783566713333, 0.05076063051819801, 0.055500324815511703, 0.03497475013136864, 0.025941763073205948, 0.2124396115541458, 0.10449168086051941, 0.39724162220954895, 0.48885905742645264, -0.11699507385492325, 0.7418338656425476, 0.09344246238470078, -0.010376805439591408, 0.011937758885324001, -0.037466153502464294, -0.1898796558380127, 0.05390346795320511, 0.12369749695062637, 0.4505586624145508, -0.09263240545988083, 0.3479197025299072, -0.22000618278980255, 0.07006711512804031, -0.040546055883169174, -0.5381633639335632, 0.11118791997432709, -0.0031197224743664265, -0.031219875440001488, 0.2671131193637848, 0.4321085214614868, 0.12411174178123474, 0.110828697681427, -0.10501426458358765, -0.4677984118461609, -0.27645811438560486, 0.40621402859687805, -0.5333340764045715, -0.2964300513267517, 0.18773140013217926, 0.23693422973155975, -0.08291369676589966, 0.036134686321020126, -0.2994183599948883, -0.07017748802900314, 0.16325390338897705, -0.22941242158412933, -0.012820384465157986, 0.3217461109161377, -0.4839881658554077, 0.06313849985599518, -0.03617709130048752, 0.24127358198165894, 0.22410330176353455, -0.3566042482852936, -0.033037252724170685, -0.15160402655601501 ]
https://github.com/huggingface/datasets/issues/179
[Feature request] separate split name and split instructions
If your dataset is a collection of sub-datasets, you should probably consider having one config per sub-dataset. For example for Glue, we have sst2, mnli etc. If you want to have multiple train sets (for example one per stage). The easiest solution would be to name them `nlp.Split("train_stage1")`, `nlp.Split("train_stage2")`, etc. or something like that.
Currently, the name of an nlp.NamedSplit is parsed in arrow_reader.py and used as the instruction. This makes it impossible to have several training sets, which can occur when: - A dataset corresponds to a collection of sub-datasets - A dataset was built in stages, adding new examples at each stage Would it be possible to have two separate fields in the Split class, a name /instruction and a unique ID that is used as the key in the builder's split_dict ?
54
[Feature request] separate split name and split instructions Currently, the name of an nlp.NamedSplit is parsed in arrow_reader.py and used as the instruction. This makes it impossible to have several training sets, which can occur when: - A dataset corresponds to a collection of sub-datasets - A dataset was built in stages, adding new examples at each stage Would it be possible to have two separate fields in the Split class, a name /instruction and a unique ID that is used as the key in the builder's split_dict ? If your dataset is a collection of sub-datasets, you should probably consider having one config per sub-dataset. For example for Glue, we have sst2, mnli etc. If you want to have multiple train sets (for example one per stage). The easiest solution would be to name them `nlp.Split("train_stage1")`, `nlp.Split("train_stage2")`, etc. or something like that.
[ 0.04338924586772919, -0.19811077415943146, 0.030185675248503685, 0.21069085597991943, 0.10683432966470718, 0.08471931517124176, 0.5492761135101318, 0.19509541988372803, -0.014409385621547699, 0.15259334444999695, 0.0023250970989465714, 0.45175155997276306, -0.2708510756492615, 0.32557156682014465, 0.019077444449067116, -0.40613803267478943, -0.06452751904726028, 0.061306584626436234, 0.23302127420902252, 0.10228580236434937, 0.09767555445432663, 0.2273228019475937, -0.03829320892691612, 0.2814406752586365, -0.23800848424434662, -0.22818610072135925, -0.2058771848678589, -0.1653982698917389, 0.018803033977746964, -0.40434202551841736, 0.17547115683555603, 0.17289216816425323, -0.24170561134815216, 0.03427776321768761, -0.00011900687968591228, -0.12621568143367767, 0.21351444721221924, -0.0090980539098382, -0.15672512352466583, -0.1734885722398758, -0.5583316683769226, -0.49402719736099243, 0.4417526423931122, -0.5981581807136536, -0.004088054411113262, -0.06665854156017303, -0.009363004006445408, 0.2533458471298218, 0.5818639993667603, -0.1610967367887497, 0.09534043073654175, -0.37281712889671326, -0.006457954179495573, 0.1218833327293396, 0.3031226396560669, 0.17062778770923615, -0.32061246037483215, -0.33287808299064636, -0.1921621561050415, 0.01641908660531044, -0.0143435625359416, 0.21867430210113525, 0.06726618111133575, -0.14753024280071259, 0.12505781650543213, 0.20889067649841309, -0.18132558465003967, -0.16720466315746307, -0.06251612305641174, 0.37087708711624146, -0.16584412753582, -0.18620947003364563, -0.21817921102046967, -0.7446478009223938, 0.19280828535556793, -0.39910298585891724, 0.009477135725319386, 0.2820793688297272, -0.023257751017808914, 0.0569058395922184, 0.22809386253356934, -0.30260971188545227, -0.1733475625514984, 0.07937303185462952, 0.15809054672718048, 0.6143718361854553, 0.08293595165014267, 0.20643454790115356, 0.19808441400527954, 0.01964546926319599, 0.19974546134471893, 0.13025808334350586, -0.04197680577635765, 0.1836852878332138, -0.2608931064605713, -0.3455762267112732, -0.39284199476242065, -0.08230011910200119, -0.21530209481716156, 0.14773288369178772, 0.2048059105873108, -0.037260860204696655, 0.08624082803726196, 0.13486889004707336, 0.24606934189796448, 0.07950537651777267, 0.43001291155815125, 0.6358776688575745, 0.11269086599349976, -0.019236857071518898, -0.3608032763004303, -0.024251269176602364, -0.06769775599241257, 0.06192421540617943, -0.17635320127010345, 0.10095897316932678, -0.09962914884090424, 0.23635649681091309, -0.05811112001538277, -0.29367464780807495, -0.11544916033744812, -0.2922120690345764, 0.27775993943214417, -0.007407102733850479, 0.13234078884124756, 0.055596426129341125, -0.32508039474487305, 0.13766512274742126, -0.190327450633049, -0.4007328748703003, -0.10605192184448242, 0.12324076890945435, -0.30392980575561523, 0.28528645634651184, -0.025047264993190765, 0.08957474678754807, 0.1397496908903122, -0.02392270602285862, -0.1711864173412323, -0.26183921098709106, 0.19998778402805328, -0.03077303245663643, 0.2510848939418793, 0.07993139326572418, 0.016892174258828163, 0.05753546953201294, -0.09818626195192337, -0.11058603972196579, -0.31324678659439087, -0.00434282049536705, -0.15221133828163147, -0.3459356725215912, 0.14371679723262787, 0.059594325721263885, 0.14884133636951447, 0.06756243109703064, 0.2856258451938629, 0.6215519309043884, 0.34704893827438354, -0.11526288837194443, 0.13643164932727814, 0.058719538152217865, -0.22138452529907227, -0.24143913388252258, -0.020820872858166695, -0.05500040203332901, 0.0404542051255703, -0.4513776898384094, -0.2863811254501343, 0.10820093005895615, -0.10519050806760788, 0.09117933362722397, -0.22937333583831787, 0.11082262545824051, -0.0266858097165823, 0.6001894474029541, 0.7154171466827393, -0.2713156044483185, 0.07817181199789047, 0.3304648697376251, 0.07260429114103317, -0.34571895003318787, 0.4756771922111511, 0.08520407974720001, 0.13212943077087402, -0.057622361928224564, -0.396837055683136, 0.1820664405822754, -0.38011738657951355, -0.17116215825080872, 0.20397421717643738, -0.5321475267410278, 0.16839459538459778, -0.09095210582017899, -0.055526427924633026, -0.37094515562057495, 0.020021768286824226, 0.2963106036186218, 0.4655568599700928, 0.03446798026561737, 0.23717808723449707, -0.17417529225349426, -0.07773016393184662, 0.2893904447555542, -0.1731615960597992, -0.3693116307258606, -0.2701151669025421, -0.010613588616251945, -0.17328216135501862, 0.056780703365802765, 0.013177584856748581, -0.40739691257476807, 0.04309285059571266, -0.2719109356403351, -0.16464754939079285, -0.28500720858573914, 0.046643614768981934, -0.2125973254442215, -0.058667752891778946, -0.34646669030189514, -0.3343038260936737, 0.17123690247535706, -0.06379195302724838, 0.2837934195995331, -0.31552425026893616, -0.1585674285888672, 0.19285030663013458, 0.00833495706319809, -0.23998557031154633, 0.5741868615150452, 0.2506142258644104, -0.02533869445323944, 0.06489785015583038, 0.35172727704048157, 0.38948336243629456, -0.07242319732904434, 0.023375356569886208, -0.12618185579776764, 0.253939688205719, -0.06732700020074844, -0.08393129706382751, -0.10601679980754852, -0.26386895775794983, -0.14377789199352264, -0.1017712727189064, 0.5019603967666626, 0.1218738853931427, 0.590939462184906, 0.2901533842086792, 0.05820629000663757, -0.06445087492465973, -0.2585403323173523, -0.1919446736574173, -0.46373245120048523, -0.2635038197040558, -0.3758358955383301, 0.08257977664470673, -0.043832167983055115, -0.28901010751724243, 0.24587661027908325, 0.4475729167461395, 0.14510782063007355, 0.016081850975751877, -0.2310430109500885, 0.11973941326141357, 0.17330650985240936, -0.20404227077960968, 0.7559657692909241, 0.19795002043247223, 0.16951534152030945, -0.0935155376791954, 0.05726432800292969, -0.042333148419857025, -0.18957479298114777, 0.27606838941574097, 0.2820304036140442, 0.079554483294487, -0.02713959850370884, -0.16982011497020721, -0.1431918889284134, 0.10303767025470734, -0.033338144421577454, 0.2293054312467575, -0.3285977840423584, -0.08615013211965561, -0.07723286747932434, -0.1250094622373581, -0.33749616146087646, -0.6012633442878723, -0.03178229555487633, -0.23478981852531433, -0.2417270541191101, 0.1579473465681076, -0.08859377354383469, -0.2108808159828186, 0.14174915850162506, -0.1397169679403305, 0.13324028253555298, -0.3738996684551239, 0.039971724152565, 0.1473178267478943, -0.3808603584766388, -0.0698835700750351, 0.025912106037139893, 0.05599702149629593, 0.034937307238578796, 0.2627149522304535, 0.32565441727638245, -0.28088316321372986, 0.1503288298845291, -0.33597153425216675, 0.17685151100158691, -0.3270227313041687, -0.2894212305545807, 0.22199642658233643, -0.31017279624938965, -0.004219421651214361, -0.468950092792511, 0.3010498881340027, -0.10614211112260818, -0.10014098137617111, -0.3176914155483246, 0.32188090682029724, 0.2116592824459076, -0.1271352767944336, -0.31906232237815857, -0.4289039075374603, -0.14181461930274963, 0.34559154510498047, -0.07541974633932114, 0.29651549458503723, 0.09804361313581467, -0.4208858907222748, -0.029327282682061195, -0.09281159192323685, -0.11690046638250351, 0.04345416650176048, 0.1445828527212143, -0.08969876915216446, -0.41021353006362915, 0.15662725269794464, 0.052333902567625046, -0.13682609796524048, 0.33396780490875244, -0.17611552774906158, -0.011461534537374973, 0.1888331174850464, -0.013530704192817211, -0.03493442386388779, 0.03572600707411766, 0.11508888006210327, 0.1559378206729889, 0.27365270256996155, 0.06651561707258224, -0.07385827600955963, -0.23240210115909576, 0.4348665773868561, 0.15245725214481354, 0.15153111517429352, -0.09006552398204803, 0.024157391861081123, -0.07993459701538086, 0.6646895408630371, 0.2682937681674957, 0.1996210813522339, -0.1777934581041336, -0.004738924093544483, -0.3417833149433136, 0.07829805463552475, -0.21145576238632202, 0.1659887135028839, -0.003021791810169816, -0.29930853843688965, 0.341092050075531, 0.030007073655724525, -0.06149280071258545, -0.2507540285587311, -0.06884215772151947, -0.0362500324845314, -0.5644192099571228, 0.06861479580402374, -0.18404830992221832, 0.21568116545677185, -0.19991718232631683, 0.0524551086127758, 0.0038426248356699944, 0.06245642155408859, -0.14635194838047028, -0.15383100509643555, -0.14873458445072174, -0.07147359848022461, -0.11421851068735123, -0.6203206181526184, -0.05381563678383827, -0.04224313423037529, 0.28871452808380127, -0.14281056821346283, -0.021647240966558456, -0.2228596806526184, 0.2955179810523987, 0.30176839232444763, 0.2860283851623535, -0.09012620151042938, -0.1666279286146164, 0.09333618730306625, 0.04214932769536972, 0.19225512444972992, -0.19830714166164398, 0.0910092294216156, 0.029229849576950073, 0.07483012229204178, 0.03935945779085159, -0.5721219182014465, -0.2948223054409027, 0.588744580745697, -0.24235455691814423, -0.33407726883888245, 0.03839506581425667, 0.03843766078352928, 0.13238564133644104, 0.07342962920665741, -0.040468115359544754, 0.10366427153348923, 0.12703469395637512, 0.1372135877609253, 0.08338845521211624, -0.03285440802574158, -0.2343682497739792, 0.08977660536766052, 0.1343218982219696, 0.021047260612249374, -0.05372067540884018, 0.08779150992631912, 0.29639932513237, 0.26572954654693604, -0.3124951124191284, 0.44037744402885437, -0.400066077709198, -0.0989203006029129, 0.009538134559988976, -0.2170742154121399, 0.538526713848114, 0.33155718445777893, 0.13701176643371582, 0.4471999406814575, 0.005818749777972698, 0.0808025375008583, -0.20520968735218048, -0.06847010552883148, 0.27846354246139526, -0.17185600101947784, -0.40903061628341675, -0.7786414623260498, 0.39444905519485474, 0.09553124755620956, -0.19228196144104004, 0.22138802707195282, 0.13551382720470428, -0.40848273038864136, 0.5755677223205566, 0.016211140900850296, 0.8848466277122498, -0.07890192419290543, 0.2104569673538208, -0.05240216851234436, -0.2898176610469818, 0.674556314945221, -0.09591908007860184, 0.13303448259830475, -0.36648884415626526, 0.29457226395606995, 0.057935524731874466, 0.10396654158830643, 0.23142297565937042, 0.4899131655693054, 0.006773074623197317, 0.2321578562259674, 0.08442994207143784, 0.26291415095329285, -0.3335981070995331, 0.3395187556743622, 0.14325371384620667, -0.11598385125398636, -0.08670655637979507, -0.048195578157901764, 0.0014831115258857608, -0.24703234434127808, 0.20888154208660126, -0.199101984500885, 0.01662462204694748, 0.17032180726528168, -0.25539764761924744, -0.00694996165111661, -0.12027224898338318, 0.21191957592964172, -0.12011650949716568, -0.4156666398048401, 0.39140889048576355, 0.17767105996608734, 0.2339615672826767, -0.1650223433971405, 0.019225813448429108, 0.4578537344932556, -0.04234394431114197, 0.09356594830751419, -0.17171630263328552, -0.012513748370110989, 0.3011680245399475, 0.05344686284661293, 0.027400951832532883, 0.031028084456920624, 0.01867331936955452, -0.08838093280792236, -0.2242787927389145, 0.2394067943096161, 0.44946640729904175, -0.5152657628059387, 0.20778262615203857, 0.031566403806209564, 0.0019562779925763607, -0.17893284559249878, 0.02122199907898903, -0.21939095854759216, -0.24939338862895966, 0.040107402950525284, -0.17997559905052185, -0.0684349536895752, 0.004961478989571333, -0.31928500533103943, 0.1781563013792038, 0.007905484177172184, 0.3882559835910797, -0.18149018287658691, 0.1604386419057846, -0.23996275663375854, 0.1463245153427124, 0.15137897431850433, -0.20615141093730927, 0.29052814841270447, 0.1530693769454956, -0.2651846408843994, 0.13528481125831604, -0.019382691010832787, 0.05809573829174042, 0.6687642931938171, -0.18001708388328552, 0.03185360133647919, -0.22082480788230896, -0.040196772664785385, -0.1035732701420784, 0.09732284396886826, 0.006829398684203625, 0.49609068036079407, -0.12097829580307007, 0.40659332275390625, -0.1664099544286728, -0.09653322398662567, 0.20132772624492645, 0.2539377510547638, -0.10842861980199814, 0.017031384631991386, 0.1527232527732849, 0.12926167249679565, 0.02038087695837021, -0.05497075989842415, 0.17742344737052917, -0.08827225118875504, -0.3262881636619568, 0.11676829308271408, 0.09450916200876236, 0.2027484029531479, -0.36953097581863403, -0.141166552901268, 0.10343468189239502, -0.15560424327850342, 0.20738081634044647, 0.06280748546123505, 0.260486900806427, 0.5012199878692627, 0.2707222104072571, 0.010650294832885265, -0.37880581617355347, 0.21698950231075287, -0.12480537593364716, 0.3370325565338135, 0.2525508403778076, 0.31090623140335083, -0.21063508093357086, 0.07289387285709381, 0.13318678736686707, -0.055774543434381485, 0.18610899150371552, -0.2795480191707611, -0.07307407259941101, 0.03606549650430679, -0.09714625775814056, -0.14056938886642456, 0.21551012992858887, 0.4737996459007263, 0.13721723854541779, -0.0349685437977314, 0.22364668548107147, 0.11047764122486115, -0.0722932443022728, 0.2131301313638687, 0.2800588309764862, -0.1797780692577362, 0.09948057681322098, 0.5531476140022278, 0.11321622878313065, 0.262257844209671, -0.4264288544654846, -0.06307661533355713, -0.10696768015623093, -0.030634470283985138, 0.5049610733985901, 0.828404426574707, 0.21459491550922394, -0.06918085366487503, 0.46372461318969727, -0.18284380435943604, 0.2680802345275879, 0.4525597095489502, -0.18141061067581177, 0.36137518286705017, 0.3433869481086731, -0.17254148423671722, 0.5454279184341431, 0.10762057453393936, -0.3695581257343292, 0.04664447903633118, -0.05818091705441475, -0.248765230178833, -0.12223390489816666, 0.7737444639205933, 0.043878160417079926, -0.33211201429367065, 0.17119194567203522, 0.31557562947273254, 0.1680949479341507, -0.2501698434352875, -0.13615261018276215, 0.11348205804824829, -0.3541956841945648, 0.0006025631446391344, -0.05430057644844055, -0.22777622938156128, 0.35432329773902893, 0.08226676285266876, 0.3009105920791626, -0.09902714192867279, -0.06769774854183197, 0.2107917070388794, 0.47529464960098267, -0.3517264723777771, -0.1351252943277359, -0.017433542758226395, -0.29152190685272217, 0.15518058836460114, -0.0719444528222084, 0.008162344805896282, -0.13468031585216522, -0.2549782693386078, -0.2885669469833374, 0.246069073677063, -0.10548307746648788, 0.02328447997570038, -0.06977805495262146, 0.12184911221265793, 0.2911975085735321, 0.2118600606918335, -0.023083319887518883, -0.03540928289294243, -0.2072150856256485, 0.2606799304485321, -0.018993614241480827, -0.05991294980049133, 0.5498870611190796, 0.23981600999832153, -0.18786679208278656, -0.025995628908276558, 0.22880105674266815, -0.2921214699745178, -0.017804067581892014, 0.35891613364219666, -0.06609898060560226, 0.08894086629152298, -0.1679060459136963, -0.022722044959664345, -0.0211720559746027, 0.485547810792923, -0.22243903577327728, -0.20367248356342316, -0.2754066586494446, -0.3702971339225769, -0.022414501756429672, -0.1403137594461441, -0.12521357834339142, -0.27419185638427734, -0.13138259947299957, 0.16895204782485962, -0.1570611596107483, -0.1679462045431137, 0.16192841529846191, 0.04043981432914734, 0.4329409897327423, 0.3385252058506012, -0.23166534304618835, 0.005767221096903086, -0.02158740907907486, 0.14577606320381165, 0.17384907603263855, 0.13576985895633698, -0.07442179322242737, 0.36563432216644287, -0.2971503436565399, 0.1566159576177597, -0.02288590371608734, 0.584936261177063, 0.13453155755996704, 0.0016775067197158933, -0.020735984668135643, 0.22752948105335236, 0.029862824827432632, 0.03711685910820961, 0.04069647565484047, 0.012698493897914886, -0.5554836392402649, 0.41142797470092773, 0.06802868098020554, 0.1626550257205963, -0.27481985092163086, 0.4120602309703827, -0.14148400723934174, 0.003723665839061141, -0.23200534284114838, 0.10678362846374512, -0.013923656195402145, 0.19690798223018646, -0.023011675104498863, 0.2964519262313843, 0.3759611248970032, 0.2861541509628296, -0.11745959520339966, -0.06586791574954987, -0.13472864031791687, -0.37831565737724304, 0.16204316914081573, -0.03662749379873276, -0.09991863369941711, 0.08974155783653259, 0.3389492928981781, 0.19934126734733582, 0.11615023016929626, -0.765256941318512, -0.2747170925140381, 0.1887984275817871, -0.24032434821128845, -0.024293990805745125, 0.14281323552131653, 0.2829596698284149, -0.1662369668483734, 0.01470612920820713, -0.13825169205665588, 0.279705286026001, -0.06312909722328186, 0.030536295846104622, -0.14354076981544495 ]
https://github.com/huggingface/datasets/issues/179
[Feature request] separate split name and split instructions
Thanks for the tip! I ended up setting up three different versions of the dataset with their own configs. for the named splits, I was trying with `nlp.Split("train-stage1")`, which fails. Changing to `nlp.Split("train_stage1")` works :) I looked for examples of what works in the code comments, it may be worth adding some examples of valid/invalid names in there?
Currently, the name of an nlp.NamedSplit is parsed in arrow_reader.py and used as the instruction. This makes it impossible to have several training sets, which can occur when: - A dataset corresponds to a collection of sub-datasets - A dataset was built in stages, adding new examples at each stage Would it be possible to have two separate fields in the Split class, a name /instruction and a unique ID that is used as the key in the builder's split_dict ?
58
[Feature request] separate split name and split instructions Currently, the name of an nlp.NamedSplit is parsed in arrow_reader.py and used as the instruction. This makes it impossible to have several training sets, which can occur when: - A dataset corresponds to a collection of sub-datasets - A dataset was built in stages, adding new examples at each stage Would it be possible to have two separate fields in the Split class, a name /instruction and a unique ID that is used as the key in the builder's split_dict ? Thanks for the tip! I ended up setting up three different versions of the dataset with their own configs. for the named splits, I was trying with `nlp.Split("train-stage1")`, which fails. Changing to `nlp.Split("train_stage1")` works :) I looked for examples of what works in the code comments, it may be worth adding some examples of valid/invalid names in there?
[ 0.10458503663539886, -0.09824613481760025, 0.027599284425377846, 0.2224782407283783, 0.11074420809745789, 0.11211378127336502, 0.5969956517219543, 0.23253028094768524, -0.03582202270627022, 0.11198779940605164, 0.01879107765853405, 0.4395620822906494, -0.21577922999858856, 0.2508060336112976, -0.04807354137301445, -0.37081459164619446, -0.03147456422448158, 0.09958340227603912, 0.3009320795536041, 0.11061368882656097, 0.07535476237535477, 0.2413926124572754, -0.04599401354789734, 0.27223193645477295, -0.2111959457397461, -0.18311923742294312, -0.23876751959323883, -0.15462426841259003, 0.028415407985448837, -0.4011644124984741, 0.19092652201652527, 0.1710231900215149, -0.29777076840400696, 0.11503423750400543, -0.00012200133642181754, -0.14937646687030792, 0.26977992057800293, -0.00883525051176548, -0.1772000789642334, -0.23157063126564026, -0.5108582973480225, -0.47503551840782166, 0.4412984549999237, -0.6071304082870483, 0.006469752173870802, -0.16628843545913696, 0.00986290629953146, 0.23457162082195282, 0.5989471673965454, -0.14206406474113464, 0.07940192520618439, -0.39332571625709534, 0.004251088015735149, 0.0590033233165741, 0.29850417375564575, 0.25531521439552307, -0.3581681549549103, -0.3896653652191162, -0.13251154124736786, 0.019385214895009995, 0.013584301806986332, 0.1901758909225464, 0.04804079234600067, -0.15089280903339386, 0.05642269179224968, 0.18820179998874664, -0.24059109389781952, -0.10517442226409912, -0.0314754843711853, 0.3205355405807495, -0.14700555801391602, -0.1331748515367508, -0.22489918768405914, -0.7789210081100464, 0.12296422570943832, -0.44449758529663086, 0.023179342970252037, 0.3083987832069397, -0.08555497974157333, 0.0007133097387850285, 0.15985225141048431, -0.2577967643737793, -0.15868929028511047, 0.07997331768274307, 0.187296062707901, 0.5755341053009033, 0.04268646985292435, 0.21303322911262512, 0.13065476715564728, 0.0514659658074379, 0.24592845141887665, 0.09748350828886032, 0.03707437962293625, 0.14819641411304474, -0.240221306681633, -0.3462635576725006, -0.3629278838634491, -0.06122719496488571, -0.24780479073524475, 0.07276277244091034, 0.1947961151599884, -0.09304298460483551, 0.03791997209191322, 0.10942403972148895, 0.3325088620185852, 0.030752351507544518, 0.36637771129608154, 0.5928642749786377, 0.09704628586769104, 0.07390140742063522, -0.3302590548992157, -0.018299609422683716, -0.0558229498565197, 0.10930285602807999, -0.23530174791812897, 0.08426590263843536, -0.10003924369812012, 0.16355445981025696, -0.04588336497545242, -0.25977033376693726, -0.09913969784975052, -0.2428596019744873, 0.2244429886341095, 0.03780536353588104, 0.11807239800691605, 0.039067432284355164, -0.3129239082336426, 0.20334015786647797, -0.12280908972024918, -0.4388543963432312, -0.08893830329179764, 0.09969503432512283, -0.30024784803390503, 0.22964981198310852, 0.10692691057920456, 0.14246748387813568, 0.18240630626678467, 0.008428450673818588, -0.14765790104866028, -0.24537293612957, 0.15634974837303162, 0.006644109264016151, 0.2317558079957962, 0.04239441081881523, 0.0324648879468441, 0.05623086169362068, -0.037742093205451965, -0.07236965000629425, -0.3021255135536194, 0.017727509140968323, -0.07855930924415588, -0.3099415898323059, 0.18636268377304077, 0.04043269157409668, 0.14403407275676727, 0.061936117708683014, 0.3620258569717407, 0.6019593477249146, 0.4055766761302948, -0.16478529572486877, 0.10734276473522186, 0.037772390991449356, -0.25963088870048523, -0.19345623254776, -0.04369606077671051, -0.013956135138869286, -0.010729662142693996, -0.44869235157966614, -0.38156476616859436, 0.16157037019729614, -0.12807022035121918, 0.07277079671621323, -0.14766012132167816, 0.058448079973459244, -0.002875157166272402, 0.6331421732902527, 0.5671976804733276, -0.2285178154706955, 0.0892338752746582, 0.33646753430366516, 0.11590448766946793, -0.3352683484554291, 0.48383790254592896, 0.05015707388520241, 0.10745836049318314, -0.10666823387145996, -0.42390763759613037, 0.18046174943447113, -0.5068804025650024, -0.2254238724708557, 0.17374908924102783, -0.5565367341041565, 0.17674870789051056, -0.07062932103872299, -0.03899882733821869, -0.3304854929447174, -0.0028007179498672485, 0.2975159287452698, 0.48496320843696594, 0.034014150500297546, 0.2780067026615143, -0.19330035150051117, -0.06715782731771469, 0.2642986476421356, -0.1216098740696907, -0.35723677277565, -0.3117901682853699, -0.04289237782359123, -0.12128956615924835, 0.03108038380742073, -0.037856172770261765, -0.38954946398735046, -0.012586669996380806, -0.2790812849998474, -0.1604304313659668, -0.2774960696697235, 0.031024886295199394, -0.2263088822364807, -0.061904218047857285, -0.33485183119773865, -0.3435288071632385, 0.1144685447216034, -0.08758199214935303, 0.26570925116539, -0.3278537690639496, -0.16811059415340424, 0.22280524671077728, -0.02559535577893257, -0.1821068525314331, 0.6047658920288086, 0.28619974851608276, -0.008092588745057583, 0.06207441911101341, 0.327442467212677, 0.391190767288208, -0.030648430809378624, 0.005734299309551716, -0.12919305264949799, 0.18323718011379242, -0.12301718443632126, -0.052457865327596664, -0.04394245892763138, -0.27993229031562805, -0.20842532813549042, -0.20216161012649536, 0.5459466576576233, 0.02758193016052246, 0.6292760372161865, 0.29110652208328247, 0.049969371408224106, -0.012831488624215126, -0.3354654610157013, -0.1722605973482132, -0.5019677877426147, -0.21168507635593414, -0.41578900814056396, 0.11526751518249512, -0.13372935354709625, -0.277699738740921, 0.2355119287967682, 0.4305519759654999, 0.17799261212348938, 0.03364662081003189, -0.23415634036064148, 0.13413700461387634, 0.1435285359621048, -0.19840024411678314, 0.6893719434738159, 0.1891097128391266, 0.15224939584732056, -0.04077427089214325, 0.08033344894647598, 0.033018387854099274, -0.18132717907428741, 0.2855433523654938, 0.25729379057884216, 0.08319481462240219, -0.0017402502708137035, -0.16712534427642822, -0.13735830783843994, 0.05301697179675102, 0.06068041920661926, 0.22533860802650452, -0.33681824803352356, -0.05381312966346741, -0.07482294738292694, -0.1295626312494278, -0.31246498227119446, -0.5548499226570129, -0.03342391178011894, -0.1863943487405777, -0.22623001039028168, 0.20903994143009186, 0.014330345205962658, -0.21730944514274597, 0.15765009820461273, -0.15921010076999664, 0.16644985973834991, -0.41059428453445435, 0.08151478320360184, 0.050145041197538376, -0.41217389702796936, -0.036571111530065536, 0.02297835610806942, 0.06295411288738251, 0.01340650487691164, 0.28876858949661255, 0.36025896668434143, -0.2948746681213379, 0.18425598740577698, -0.37269893288612366, 0.18451955914497375, -0.3997817933559418, -0.2675517201423645, 0.22958476841449738, -0.30351734161376953, 0.014626645483076572, -0.42857199907302856, 0.3069126009941101, -0.16745899617671967, -0.1307963877916336, -0.3343505263328552, 0.3282942473888397, 0.19106096029281616, -0.17773844301700592, -0.31340286135673523, -0.40244075655937195, -0.14577093720436096, 0.3905414938926697, -0.06979198753833771, 0.275059312582016, 0.07240861654281616, -0.47120168805122375, -0.028921015560626984, -0.0028673801571130753, -0.12118414044380188, 0.0688781589269638, 0.22075915336608887, -0.07444760948419571, -0.4078604280948639, 0.1298803985118866, 0.08459174633026123, -0.12469162791967392, 0.36802560091018677, -0.24733638763427734, -0.06267568469047546, 0.2727818489074707, 0.051286887377500534, -0.06166435033082962, 0.1613778918981552, 0.0468614436686039, 0.16062426567077637, 0.29216834902763367, 0.06726684421300888, -0.17208799719810486, -0.24137750267982483, 0.3609781563282013, 0.12154132127761841, 0.1340133100748062, -0.07184949517250061, 0.09102027863264084, -0.05060441419482231, 0.6642805337905884, 0.29654571413993835, 0.2599334120750427, -0.15238559246063232, 0.06831364333629608, -0.3416397273540497, 0.0813865140080452, -0.22828562557697296, 0.1414753943681717, -0.05142045021057129, -0.3744942247867584, 0.3664928674697876, 0.03204843029379845, -0.1365581750869751, -0.26294854283332825, -0.012270881794393063, -0.0825326070189476, -0.5328510403633118, 0.008546442724764347, -0.1830296516418457, 0.2694476246833801, -0.1307053118944168, 0.06471725553274155, 0.032200947403907776, -0.03419889137148857, -0.14848773181438446, -0.19269715249538422, -0.09611956030130386, -0.1442139446735382, -0.1351078301668167, -0.5458650588989258, -0.05232827365398407, -0.013712148182094097, 0.2454078644514084, -0.1310095489025116, -0.0434127151966095, -0.19963517785072327, 0.2979270815849304, 0.2764438986778259, 0.2861582934856415, -0.0740295946598053, -0.1080959364771843, 0.14678597450256348, 0.046864598989486694, 0.19412049651145935, -0.1502498984336853, 0.07547231763601303, 0.025733284652233124, 0.04000810161232948, 0.021056868135929108, -0.5988478064537048, -0.27710771560668945, 0.5790035128593445, -0.21796877682209015, -0.3166865110397339, 0.028383495286107063, 0.05465560033917427, 0.1528550535440445, 0.08562375605106354, -0.02849876508116722, 0.12638455629348755, 0.10292917490005493, 0.0625978633761406, 0.131519615650177, -0.03761422261595726, -0.1843014806509018, 0.12430398911237717, 0.1284984052181244, 0.00018779364472720772, -0.08899038285017014, 0.1555132269859314, 0.23965628445148468, 0.2841663360595703, -0.32764962315559387, 0.49771419167518616, -0.363668829202652, -0.10032572597265244, 0.021688221022486687, -0.21436485648155212, 0.5283656120300293, 0.4158989489078522, 0.14213882386684418, 0.38511332869529724, 0.04732183739542961, 0.047360680997371674, -0.19367416203022003, -0.11665058135986328, 0.32908570766448975, -0.2601209282875061, -0.4622156620025635, -0.7009155750274658, 0.4129932224750519, 0.0718451514840126, -0.22511838376522064, 0.2465713620185852, 0.15709926187992096, -0.37633752822875977, 0.5711109042167664, -0.01972142979502678, 0.8757337927818298, 0.0014641567831858993, 0.19628098607063293, 0.01991477608680725, -0.2845609486103058, 0.6315008997917175, -0.01509584579616785, 0.10312221944332123, -0.3777364492416382, 0.3019935190677643, 0.06101781129837036, 0.08488953113555908, 0.2285563200712204, 0.5355206727981567, 0.013650733977556229, 0.18368692696094513, 0.06374097615480423, 0.27864834666252136, -0.2928549647331238, 0.33803319931030273, 0.20744073390960693, -0.08355415612459183, -0.14133737981319427, -0.04221641644835472, 0.02688756212592125, -0.25858765840530396, 0.1981574445962906, -0.18899035453796387, -0.02284952625632286, 0.14109393954277039, -0.2769813537597656, 0.06596162170171738, -0.004031647462397814, 0.11964194476604462, -0.12575943768024445, -0.44830775260925293, 0.43208828568458557, 0.1650523990392685, 0.32499927282333374, -0.18011316657066345, -0.03807511553168297, 0.471480131149292, -0.058188989758491516, 0.09149448573589325, -0.16750414669513702, 0.02509329654276371, 0.2214224934577942, 0.029511284083127975, 0.10628629475831985, 0.026310840621590614, -0.02527465857565403, -0.06378037482500076, -0.2644970118999481, 0.21506069600582123, 0.4529539942741394, -0.5559496283531189, 0.17467345297336578, -0.01298394612967968, 0.052493371069431305, -0.2350744903087616, 0.01745951920747757, -0.23933294415473938, -0.21969915926456451, 0.016379496082663536, -0.18049436807632446, -0.04825228452682495, -0.02668444626033306, -0.2955884039402008, 0.23804031312465668, -0.015308116562664509, 0.4281604290008545, -0.11671429872512817, 0.16510877013206482, -0.23216113448143005, 0.07606164366006851, 0.17264467477798462, -0.2626328766345978, 0.28448686003685, 0.11356735229492188, -0.27455300092697144, 0.1506139039993286, -0.015604402869939804, 0.05226319655776024, 0.6398311257362366, -0.15392349660396576, 0.015465147793293, -0.20867978036403656, 0.004173132590949535, -0.06663259863853455, 0.06332922726869583, -0.045188065618276596, 0.4669845998287201, -0.19155891239643097, 0.4138610064983368, -0.13755100965499878, -0.01859297975897789, 0.2807844579219818, 0.26239317655563354, -0.15228065848350525, 0.009632370434701443, 0.19519418478012085, 0.07528966665267944, 0.005358508322387934, -0.006513004656881094, 0.20427751541137695, -0.06048782169818878, -0.31543099880218506, 0.12557701766490936, 0.13687163591384888, 0.19400665163993835, -0.3965693712234497, -0.16864915192127228, 0.07181933522224426, -0.1536898910999298, 0.17872698605060577, 0.06687228381633759, 0.28282567858695984, 0.4819811284542084, 0.3057691752910614, 0.11159995198249817, -0.36444994807243347, 0.18410398066043854, -0.12273823469877243, 0.3808996379375458, 0.2149306684732437, 0.27629393339157104, -0.16513992846012115, 0.09467706829309464, 0.10485759377479553, -0.04907878488302231, 0.19031277298927307, -0.3088620901107788, 0.013765617273747921, 0.03284564986824989, -0.030746977776288986, -0.14969120919704437, 0.25887104868888855, 0.5672736167907715, 0.1264316290616989, -0.0025875535793602467, 0.22976264357566833, 0.08614969253540039, -0.09101282805204391, 0.25636476278305054, 0.28159084916114807, -0.1594066172838211, 0.08424215763807297, 0.5541465878486633, 0.12478668987751007, 0.2655099928379059, -0.4000854194164276, -0.03358297422528267, -0.05661659315228462, -0.044363029301166534, 0.501385509967804, 0.8824549317359924, 0.20604345202445984, -0.019955584779381752, 0.4297122359275818, -0.1907728910446167, 0.2422768771648407, 0.4915965497493744, -0.1730971485376358, 0.3304673433303833, 0.3147026300430298, -0.12262701243162155, 0.5907779335975647, 0.19488142430782318, -0.40433797240257263, -0.003066928358748555, -0.016770387068390846, -0.23411890864372253, -0.1521468162536621, 0.7617435455322266, -0.011582126840949059, -0.32540491223335266, 0.09411877393722534, 0.26180583238601685, 0.1801459640264511, -0.20263704657554626, -0.16221727430820465, 0.055055685341358185, -0.35550782084465027, 0.008618003688752651, -0.08892423659563065, -0.206252321600914, 0.35118433833122253, 0.0373205691576004, 0.26485714316368103, -0.15261177718639374, -0.032471898943185806, 0.17793378233909607, 0.5029487609863281, -0.31820863485336304, -0.10205788910388947, -0.0024601786863058805, -0.2936455011367798, 0.19719001650810242, -0.09273358434438705, -0.005650539416819811, -0.14058050513267517, -0.32904762029647827, -0.35682880878448486, 0.20284217596054077, -0.1465296894311905, -0.00009966111974790692, 0.03662054240703583, 0.12525337934494019, 0.3037355840206146, 0.18481440842151642, -0.04838895797729492, -0.026214923709630966, -0.15178020298480988, 0.2818923890590668, -0.07489319145679474, -0.04711452126502991, 0.6482115387916565, 0.13582082092761993, -0.18381637334823608, 0.02307545207440853, 0.17055052518844604, -0.23164071142673492, -0.06651804596185684, 0.3410789668560028, -0.04825194925069809, 0.11722412705421448, -0.1129508689045906, -0.03360508009791374, -0.034838706254959106, 0.40514278411865234, -0.2068142294883728, -0.2867162823677063, -0.26146507263183594, -0.4266860783100128, -0.08076170086860657, -0.07397931814193726, -0.1493707150220871, -0.24868972599506378, -0.0993378534913063, 0.13811615109443665, -0.08726971596479416, -0.16104353964328766, 0.11898617446422577, -0.07627561688423157, 0.39194750785827637, 0.36234721541404724, -0.22487066686153412, 0.006812672130763531, -0.0320691280066967, 0.1607150137424469, 0.12720412015914917, 0.11123872548341751, -0.1088414117693901, 0.3447793424129486, -0.32984739542007446, 0.17180922627449036, -0.07115096598863602, 0.5625971555709839, 0.20206759870052338, -0.059335220605134964, 0.012389618903398514, 0.23775669932365417, 0.029855914413928986, -0.06352163851261139, 0.035342250019311905, -0.002857745159417391, -0.5393228530883789, 0.5002479553222656, 0.01725945435464382, 0.17635037004947662, -0.23865655064582825, 0.44471845030784607, -0.14618812501430511, -0.0446426086127758, -0.2594882845878601, 0.1652461588382721, -0.06921348720788956, 0.22428317368030548, -0.060415033251047134, 0.3944740891456604, 0.4205292761325836, 0.26409462094306946, -0.12344460934400558, -0.025244353339076042, -0.21176902949810028, -0.3554520905017853, 0.23516441881656647, -0.07453923672437668, -0.13716068863868713, 0.10333868861198425, 0.3187499940395355, 0.1716342568397522, 0.016238758340477943, -0.7228702306747437, -0.26910263299942017, 0.16507181525230408, -0.2762206196784973, 0.020327763631939888, 0.03865715488791466, 0.23195217549800873, -0.23970811069011688, 0.013042955659329891, -0.04580787196755409, 0.29867565631866455, -0.07945644110441208, 0.06468213349580765, -0.11545087397098541 ]
https://github.com/huggingface/datasets/issues/168
Loading 'wikitext' dataset fails
Hi, make sure you have a recent version of pyarrow. Are you using it in Google Colab? In this case, this error is probably the same as #128
Loading the 'wikitext' dataset fails with Attribute error: Code to reproduce (From example notebook): import nlp wikitext_dataset = nlp.load_dataset('wikitext') Error: --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) <ipython-input-17-d5d9df94b13c> in <module>() 11 12 # Load a dataset and print the first examples in the training set ---> 13 wikitext_dataset = nlp.load_dataset('wikitext') 14 print(wikitext_dataset['train'][0]) 6 frames /usr/local/lib/python3.6/dist-packages/nlp/load.py in load_dataset(path, name, version, data_dir, data_files, split, cache_dir, download_config, download_mode, ignore_verifications, save_infos, **config_kwargs) 518 download_mode=download_mode, 519 ignore_verifications=ignore_verifications, --> 520 save_infos=save_infos, 521 ) 522 /usr/local/lib/python3.6/dist-packages/nlp/builder.py in download_and_prepare(self, download_config, download_mode, ignore_verifications, save_infos, dl_manager, **download_and_prepare_kwargs) 363 verify_infos = not save_infos and not ignore_verifications 364 self._download_and_prepare( --> 365 dl_manager=dl_manager, verify_infos=verify_infos, **download_and_prepare_kwargs 366 ) 367 # Sync info /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _download_and_prepare(self, dl_manager, verify_infos, **prepare_split_kwargs) 416 try: 417 # Prepare split will record examples associated to the split --> 418 self._prepare_split(split_generator, **prepare_split_kwargs) 419 except OSError: 420 raise OSError("Cannot find data file. " + (self.MANUAL_DOWNLOAD_INSTRUCTIONS or "")) /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _prepare_split(self, split_generator) 594 example = self.info.features.encode_example(record) 595 writer.write(example) --> 596 num_examples, num_bytes = writer.finalize() 597 598 assert num_examples == num_examples, f"Expected to write {split_info.num_examples} but wrote {num_examples}" /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in finalize(self, close_stream) 173 def finalize(self, close_stream=True): 174 if self.pa_writer is not None: --> 175 self.write_on_file() 176 self.pa_writer.close() 177 if close_stream: /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in write_on_file(self) 124 else: 125 # All good --> 126 self._write_array_on_file(pa_array) 127 self.current_rows = [] 128 /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in _write_array_on_file(self, pa_array) 93 def _write_array_on_file(self, pa_array): 94 """Write a PyArrow Array""" ---> 95 pa_batch = pa.RecordBatch.from_struct_array(pa_array) 96 self._num_bytes += pa_array.nbytes 97 self.pa_writer.write_batch(pa_batch) AttributeError: type object 'pyarrow.lib.RecordBatch' has no attribute 'from_struct_array'
28
Loading 'wikitext' dataset fails Loading the 'wikitext' dataset fails with Attribute error: Code to reproduce (From example notebook): import nlp wikitext_dataset = nlp.load_dataset('wikitext') Error: --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) <ipython-input-17-d5d9df94b13c> in <module>() 11 12 # Load a dataset and print the first examples in the training set ---> 13 wikitext_dataset = nlp.load_dataset('wikitext') 14 print(wikitext_dataset['train'][0]) 6 frames /usr/local/lib/python3.6/dist-packages/nlp/load.py in load_dataset(path, name, version, data_dir, data_files, split, cache_dir, download_config, download_mode, ignore_verifications, save_infos, **config_kwargs) 518 download_mode=download_mode, 519 ignore_verifications=ignore_verifications, --> 520 save_infos=save_infos, 521 ) 522 /usr/local/lib/python3.6/dist-packages/nlp/builder.py in download_and_prepare(self, download_config, download_mode, ignore_verifications, save_infos, dl_manager, **download_and_prepare_kwargs) 363 verify_infos = not save_infos and not ignore_verifications 364 self._download_and_prepare( --> 365 dl_manager=dl_manager, verify_infos=verify_infos, **download_and_prepare_kwargs 366 ) 367 # Sync info /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _download_and_prepare(self, dl_manager, verify_infos, **prepare_split_kwargs) 416 try: 417 # Prepare split will record examples associated to the split --> 418 self._prepare_split(split_generator, **prepare_split_kwargs) 419 except OSError: 420 raise OSError("Cannot find data file. " + (self.MANUAL_DOWNLOAD_INSTRUCTIONS or "")) /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _prepare_split(self, split_generator) 594 example = self.info.features.encode_example(record) 595 writer.write(example) --> 596 num_examples, num_bytes = writer.finalize() 597 598 assert num_examples == num_examples, f"Expected to write {split_info.num_examples} but wrote {num_examples}" /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in finalize(self, close_stream) 173 def finalize(self, close_stream=True): 174 if self.pa_writer is not None: --> 175 self.write_on_file() 176 self.pa_writer.close() 177 if close_stream: /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in write_on_file(self) 124 else: 125 # All good --> 126 self._write_array_on_file(pa_array) 127 self.current_rows = [] 128 /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in _write_array_on_file(self, pa_array) 93 def _write_array_on_file(self, pa_array): 94 """Write a PyArrow Array""" ---> 95 pa_batch = pa.RecordBatch.from_struct_array(pa_array) 96 self._num_bytes += pa_array.nbytes 97 self.pa_writer.write_batch(pa_batch) AttributeError: type object 'pyarrow.lib.RecordBatch' has no attribute 'from_struct_array' Hi, make sure you have a recent version of pyarrow. Are you using it in Google Colab? In this case, this error is probably the same as #128
[ -0.11933788657188416, 0.0886276513338089, -0.008977650664746761, 0.21019726991653442, 0.3692685067653656, 0.026060238480567932, 0.44878119230270386, 0.41310203075408936, 0.43308186531066895, 0.012566424906253815, 0.12384431064128876, 0.39542996883392334, -0.1505494862794876, 0.021962128579616547, 0.12090560793876648, -0.4289896488189697, -0.033739328384399414, 0.36367329955101013, -0.04146643355488777, 0.06297355890274048, -0.1876838058233261, 0.10851654410362244, -0.2651917040348053, 0.255005806684494, -0.30572274327278137, -0.05702603608369827, 0.21592095494270325, 0.08206319063901901, -0.07891858369112015, -0.55900639295578, 0.2770833671092987, -0.21704953908920288, 0.0644511878490448, 0.11474639922380447, -0.00010995574848493561, 0.20530641078948975, 0.33269748091697693, 0.03021298721432686, -0.5799461603164673, -0.18957747519016266, -0.4773319959640503, -0.11893901973962784, 0.29743248224258423, -0.24970698356628418, -0.028513457626104355, 0.05389736220240593, 0.03591109812259674, -0.07441242039203644, 0.2967468500137329, 0.5315061807632446, 0.23788878321647644, 0.5236918926239014, 0.15454885363578796, 0.034493979066610336, -0.07856866717338562, -0.17057105898857117, -0.08211525529623032, 0.22396785020828247, 0.07387740910053253, -0.39154455065727234, -0.013374139554798603, 0.12692393362522125, 0.15117257833480835, 0.26370176672935486, 0.5083714723587036, 0.014393548481166363, 0.2405228316783905, -0.15062619745731354, 0.014946400187909603, 0.07640086859464645, 0.3353831470012665, -0.16493351757526398, -0.27160072326660156, -0.2243598848581314, 0.2476663589477539, -0.33475664258003235, 0.23653079569339752, -0.036676887422800064, -0.16220447421073914, -0.008197000250220299, -0.2692282497882843, -0.028147514909505844, -0.24182115495204926, 0.35268786549568176, -0.05821506306529045, 0.4996490478515625, 0.04358331486582756, 0.14274749159812927, -0.058065660297870636, -0.07545804977416992, 0.14037638902664185, -0.08709083497524261, -0.12930019199848175, 0.2578132152557373, -0.38315585255622864, -0.03646787628531456, -0.034946393221616745, -0.1169760525226593, -0.031624484807252884, 0.2635209858417511, 0.2199900895357132, -0.2248735874891281, 0.3238379955291748, 0.3154149651527405, 0.22298957407474518, 0.42046910524368286, 0.07276459038257599, 0.06359180808067322, -0.08280493319034576, 0.13602527976036072, -0.23372802138328552, -0.15548524260520935, -0.15553590655326843, -0.17655469477176666, 0.09289750456809998, -0.15460263192653656, 0.22602041065692902, -0.07111654430627823, -0.49847400188446045, 0.11472059786319733, -0.1434047520160675, 0.04629036784172058, 0.20511473715305328, 0.17948518693447113, -0.12763333320617676, 0.23223552107810974, 0.14412418007850647, 0.2825672924518585, -0.2519769072532654, -0.04313638061285019, -0.16628925502300262, 0.24921317398548126, -0.23438338935375214, -0.18224935233592987, 0.24275411665439606, 0.19340045750141144, 0.3994779884815216, -0.10116773843765259, 0.11126504093408585, -0.20175516605377197, 0.22737497091293335, -0.04551902040839195, -0.12887892127037048, 0.25665295124053955, -0.21606145799160004, 0.15489806234836578, 0.20159101486206055, -0.2824234366416931, -0.08482398837804794, -0.10174469649791718, -0.27263587713241577, -0.3080922067165375, -0.12625131011009216, 0.281415194272995, 0.14738169312477112, -0.24296516180038452, -0.15226291120052338, -0.06118733063340187, 0.14749260246753693, -0.3687240481376648, -0.11810583621263504, -0.46125921607017517, -0.22063837945461273, -0.023005429655313492, 0.3976428508758545, 0.20490917563438416, 0.131386861205101, -0.10897225141525269, -0.11542121320962906, 0.07785490155220032, -0.03179694712162018, 0.12271356582641602, -0.43835729360580444, 0.5421618223190308, -0.06105572730302811, 0.11886267364025116, 0.6413458585739136, -0.34278684854507446, -0.4578539729118347, 0.12356847524642944, -0.014279641211032867, 0.2093287706375122, -0.22536881268024445, -0.016720760613679886, -0.013226832263171673, -0.1845516413450241, 0.2123139649629593, 0.6192137002944946, -0.015043232589960098, 0.0824793204665184, -0.12689664959907532, -0.01554853655397892, 0.5002202987670898, 0.10985670238733292, 0.09252531081438065, -0.007928650826215744, -0.04170486703515053, 0.4966488778591156, -0.0013133209431543946, -0.12597094476222992, -0.014860233291983604, -0.013630262576043606, -0.06468662619590759, -0.104604572057724, -0.21782620251178741, -0.073886938393116, -0.27766478061676025, 0.3043038845062256, -0.11754702031612396, -0.0585317425429821, -0.04715042933821678, 0.1061868816614151, -0.34911301732063293, 0.16848920285701752, -0.423268586397171, -0.37731868028640747, 0.17209294438362122, 0.1551334410905838, -0.01082319114357233, 0.2384408712387085, -0.20474199950695038, 0.21943683922290802, -0.27402806282043457, 0.05550078675150871, -0.21892870962619781, 0.1612894982099533, -0.055882859975099564, -0.0424220934510231, -0.009479359723627567, 0.30879732966423035, 0.03833145648241043, -0.017996225506067276, -0.3062887489795685, 0.286030650138855, -0.1254843771457672, 0.08035652339458466, -0.03226462006568909, -0.14337000250816345, 0.06880593299865723, -0.07154339551925659, -0.17518380284309387, 0.20819160342216492, 0.3319510519504547, -0.18012623488903046, -0.04314308613538742, 0.11287619918584824, 0.12721261382102966, 0.08142238110303879, 0.09886393696069717, 0.16397275030612946, 0.1439925581216812, -0.06582631915807724, 0.10127968341112137, 0.07743124663829803, 0.2337101697921753, -0.09498753398656845, 0.20471026003360748, -0.15645122528076172, -0.16714543104171753, -0.11880633234977722, 0.31311914324760437, 0.17005811631679535, 0.1674172282218933, 0.08148688822984695, -0.27525240182876587, -0.09487471729516983, 0.07855787128210068, 0.1460483968257904, 0.4009645879268646, -0.051832713186740875, -0.10834138095378876, -0.05257111415266991, -0.4165973961353302, 0.07112560421228409, 0.1707080900669098, 0.03662936016917229, 0.20912867784500122, 0.3018558621406555, 0.04131574556231499, 0.08731992542743683, -0.29526659846305847, -0.35701724886894226, 0.2824332118034363, 0.44053006172180176, -0.22848272323608398, 0.2078799307346344, -0.0305313840508461, -0.0025109166745096445, -0.06784793734550476, -0.33363422751426697, -0.18862943351268768, -0.34299421310424805, -0.12133592367172241, 0.4337747097015381, 0.024118667468428612, 0.1810036450624466, -0.3977627158164978, -0.022050168365240097, 0.19475045800209045, -0.1147860512137413, -0.03137960657477379, -0.31579849123954773, -0.2693284749984741, 0.045942142605781555, 0.3818604350090027, 0.09107596427202225, 0.3510979115962982, -0.06677016615867615, -0.09253522008657455, -0.007033964619040489, -0.05841490253806114, -0.07556762546300888, -0.18372154235839844, 0.24404771625995636, 0.009521574713289738, 0.2855835556983948, -0.17702993750572205, -0.5220329761505127, 0.5017470121383667, -0.25570252537727356, -0.05778523162007332, 0.31510716676712036, 0.27111291885375977, -0.24919019639492035, -0.0794772058725357, -0.374688059091568, -0.1264500468969345, -0.28060564398765564, -0.05946861207485199, 0.07482945919036865, 0.06202014163136482, 0.3974682688713074, 0.07085619866847992, 0.21601079404354095, -0.002573415171355009, 0.11343801766633987, -0.04826263710856438, -0.0900639072060585, 0.1859847605228424, -0.31003302335739136, -0.31731799244880676, -0.12134335190057755, 0.00993572361767292, 0.3347616195678711, 0.11073707789182663, -0.3489590883255005, -0.21413186192512512, -0.10495806485414505, 0.08531273156404495, -0.2213255912065506, -0.040067099034786224, 0.18285974860191345, 0.07230282574892044, -0.06572254747152328, -0.08664463460445404, 0.03598904609680176, 0.07123992592096329, 0.15234632790088654, 0.13496197760105133, -0.006304681766778231, 0.08922669291496277, -0.08488419651985168, 0.6008992195129395, 0.21869337558746338, -0.2143774777650833, 0.30028587579727173, -0.10327647626399994, 0.005145025905221701, -0.20868363976478577, -0.43980398774147034, 0.2641644775867462, -0.11523281037807465, 0.1829284131526947, 0.17231453955173492, -0.07949270308017731, -0.22478418052196503, -0.19255878031253815, 0.06885252147912979, -0.17844729125499725, -0.27277931571006775, 0.14616899192333221, -0.21739473938941956, 0.278711199760437, -0.0688643828034401, 0.22253714501857758, -0.20522361993789673, -0.16798114776611328, 0.06985098868608475, 0.16693316400051117, -0.09054534137248993, 0.07967454195022583, -0.24400153756141663, -0.2761092185974121, -0.47995808720588684, -0.027839848771691322, 0.2410319298505783, 0.643750786781311, -0.08638402819633484, 0.037189047783613205, 0.0807223692536354, 0.08174136281013489, 0.13737380504608154, -0.23013576865196228, 0.1822637915611267, 0.2891305088996887, 0.031824760138988495, -0.3197470009326935, -0.06433095782995224, -0.1401122510433197, 0.35801708698272705, 0.07665061950683594, -0.05443219840526581, -0.03723124787211418, -0.16765417158603668, 0.13983193039894104, 0.25993993878364563, -0.18729501962661743, -0.04411264508962631, -0.09173404425382614, -0.0898098275065422, -0.30776384472846985, -0.006155079230666161, -0.009947403334081173, 0.3356599807739258, 0.12296963483095169, 0.26572561264038086, -0.5462993383407593, 0.11665675789117813, 0.09016667306423187, 0.17458726465702057, 0.09595341980457306, 0.057944491505622864, 0.18724635243415833, -0.16663426160812378, 0.183209627866745, -0.14590534567832947, 0.5633814334869385, 0.14737607538700104, -0.33079853653907776, -0.31953001022338867, 0.021663723513484, 0.16743426024913788, 0.18398089706897736, -0.018585436046123505, -0.1411214917898178, -0.276946485042572, -0.06146418675780296, -0.008709541521966457, 0.30457016825675964, 0.5610408186912537, 0.0026278814766556025, -0.3271811902523041, -0.2585411071777344, 0.48284009099006653, -0.2850036919116974, 0.11732959747314453, 0.32155776023864746, 0.05337808281183243, -0.3131917119026184, 0.08413904905319214, 0.07783257216215134, 0.6767759323120117, 0.1695389300584793, 0.3111467659473419, 0.3826707899570465, 0.18392953276634216, 0.4142702519893646, -0.043089646846055984, 0.28471478819847107, -0.3927352726459503, -0.13342542946338654, -0.08068593591451645, -0.06546245515346527, 0.10708130896091461, -0.00887344777584076, -0.025437481701374054, 0.1506209671497345, 0.15688923001289368, -0.0041417344473302364, 0.05714833736419678, 0.4492526948451996, -0.03878338634967804, -0.19681920111179352, -0.34702637791633606, 0.15601642429828644, -0.218000128865242, 0.29750892519950867, -0.09073349088430405, -0.06648434698581696, 0.05520431324839592, -0.06735192984342575, -0.2555583119392395, 0.2574388384819031, -0.29947420954704285, 0.4594575762748718, 0.057822249829769135, -0.47568145394325256, 0.1984158307313919, 0.20709721744060516, 0.17566342651844025, 0.005698441993445158, -0.2995951473712921, 0.2387789934873581, -0.2900906801223755, -0.2218656986951828, -0.06728668510913849, 0.11194741725921631, 0.4200740456581116, -0.1725509911775589, -0.14637738466262817, 0.1479404866695404, -0.12065477669239044, -0.2415027767419815, 0.10395903885364532, 0.036866601556539536, 0.35248854756355286, -0.39257586002349854, -0.5152192711830139, -0.0024203567299991846, -0.18207542598247528, -0.17490997910499573, 0.17449994385242462, 0.023450996726751328, -0.18903093039989471, 0.15189528465270996, 0.14982764422893524, -0.1989799439907074, 0.1164064109325409, 0.24804162979125977, -0.2495405673980713, 0.1462835967540741, 0.763757050037384, 0.033945709466934204, 0.012805157341063023, -0.31383034586906433, -0.05428868532180786, 0.12508492171764374, -0.4820001423358917, -0.11712706089019775, 0.1255035698413849, 0.16912545263767242, -0.14424709975719452, 0.4983648657798767, 0.06302112340927124, -0.018845530226826668, 0.2471855729818344, -0.5424144268035889, -0.10621847957372665, -0.04302152618765831, 0.017753472551703453, 0.23777538537979126, -0.0026242828462272882, 0.11488918960094452, 0.13550932705402374, 0.10338813066482544, -0.27263733744621277, 0.014253862202167511, -0.15762493014335632, 0.1379975527524948, 0.21783483028411865, -0.15653111040592194, 0.22328241169452667, -0.07694793492555618, 0.15638624131679535, -0.09584680199623108, -0.12303421646356583, -0.24829301238059998, -0.25483760237693787, 0.10011181235313416, 0.24556748569011688, -0.16452161967754364, -0.03067888505756855, -0.10233359783887863, 0.02367708459496498, -0.23110559582710266, -0.008462597616016865, -0.11772827059030533, -0.29334336519241333, 0.38943231105804443, 0.043897844851017, 0.2783689498901367, -0.1116899773478508, 0.12635605037212372, -0.3007572293281555, 0.14912119507789612, -0.018496258184313774, -0.07522967457771301, 0.27581310272216797, 0.09365364164113998, -0.46588754653930664, 0.005421184469014406, -0.40260615944862366, 0.037148136645555496, 0.36982107162475586, -0.45511066913604736, 0.15995857119560242, -0.1316279023885727, 0.09546560794115067, 0.4656839668750763, -0.2714088261127472, -0.0036662661004811525, 0.34966516494750977, 0.2378423660993576, -0.2974173426628113, -0.18925972282886505, 0.23999083042144775, 0.06134593486785889, 0.032459061592817307, 0.027239473536610603, 0.04666171222925186, -0.3037489652633667, -0.32123425602912903, 0.08927872031927109, -0.029339971020817757, -0.265601247549057, -0.011407657526433468, 0.3393266201019287, -0.002364314394071698, -0.05951515957713127, -0.05004987120628357, 0.21102137863636017, 0.035392459481954575, 0.38269418478012085, -0.29260048270225525, 0.20699812471866608, 0.05926388129591942, 0.04391023516654968, -0.11114337295293808, -0.21921710669994354, 0.05030057579278946, -0.157841295003891, 0.03383495658636093, 0.29528310894966125, 0.09981518238782883, 0.1530808061361313, -0.285504013299942, -0.12453760206699371, -0.20802302658557892, 0.22792354226112366, -0.17904223501682281, -0.04890627786517143, -0.24299491941928864, 0.021017028018832207, -0.025876162573695183, -0.09469947963953018, -0.14167042076587677, -0.03817928954958916, 0.3095564842224121, 0.2532219886779785, -0.1722697913646698, -0.45553529262542725, 0.10826699435710907, 0.3432939052581787, 0.18004795908927917, -0.24588418006896973, 0.33685368299484253, 0.19381506741046906, -0.06877820193767548, -0.01412354689091444, 0.39711013436317444, 0.4357686936855316, 0.33839377760887146, -0.3642207980155945, 0.08116745948791504, -0.2033807784318924, -0.0568217858672142, -0.02326009050011635, 0.26951277256011963, 0.10607278347015381, 0.11737696081399918, 0.4624232351779938, 0.12148629873991013, -0.12472685426473618, 0.04363814741373062, -0.08805640786886215, -0.2605327367782593, -0.26294752955436707, 0.5191484093666077, 0.185794860124588, -0.11438807845115662, -0.3206513226032257, 0.1899377405643463, -0.2749866843223572, -0.0809149518609047, 0.06285153329372406, 0.24058951437473297, 0.113153375685215, -0.21382735669612885, 0.07050874084234238, 0.00031574213062413037, 0.7061741352081299, 0.48772212862968445, -0.01771952025592327, -0.3658044636249542, -0.11652323603630066, -0.5989870429039001, 0.2178366482257843, -0.3639686405658722, -0.26901865005493164, 0.2833371162414551, -0.07233183085918427, -0.28808191418647766, 0.24852752685546875, 0.12761986255645752, 0.15819624066352844, -0.05727671831846237, 0.0983472391963005, -0.49076488614082336, -0.1486053764820099, -0.03065069019794464, -0.0076751429587602615, -0.0831807479262352, -0.25705254077911377, 0.1453229933977127, -0.4349938929080963, 0.1161026805639267, 0.011985591612756252, 0.06934838742017746, -0.0002399603690719232, 0.09664099663496017, 0.3544251024723053, 0.16532465815544128, 0.5012484788894653, -0.035560812801122665, 0.0036571817472577095, -0.19115544855594635, -0.15662235021591187, -0.22899216413497925, 0.13182324171066284, 0.08762864023447037, 0.5133113861083984, -0.026107320562005043, -0.06854572147130966, -0.2640267312526703, 0.40316781401634216, -0.08100741356611252, 0.08831649273633957, -0.13508076965808868, 0.07731454819440842, -0.07035040855407715, 0.22758440673351288, 0.26436471939086914, 0.4245404899120331, -0.17832162976264954, 0.20220930874347687, -0.27229106426239014, -0.4602454900741577, 0.32893526554107666, -0.6614815592765808, -0.5177910923957825, 0.07512339949607849, 0.11619257926940918, -0.1423347145318985, 0.05819396302103996, -0.6006348133087158, 0.0705668106675148, 0.24261273443698883, -0.11140530556440353, -0.3171199560165405, 0.21794475615024567, 0.016542308032512665, 0.08000080287456512, 0.043573711067438126, 0.44835665822029114, -0.14847348630428314, -0.15976785123348236, -0.004933878779411316, -0.1632518470287323 ]
https://github.com/huggingface/datasets/issues/168
Loading 'wikitext' dataset fails
Hi, The squad bug seems to be fixed, but the loading of the 'wikitext' still suffers from this problem (on Colab with pyarrow=0.17.1).
Loading the 'wikitext' dataset fails with Attribute error: Code to reproduce (From example notebook): import nlp wikitext_dataset = nlp.load_dataset('wikitext') Error: --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) <ipython-input-17-d5d9df94b13c> in <module>() 11 12 # Load a dataset and print the first examples in the training set ---> 13 wikitext_dataset = nlp.load_dataset('wikitext') 14 print(wikitext_dataset['train'][0]) 6 frames /usr/local/lib/python3.6/dist-packages/nlp/load.py in load_dataset(path, name, version, data_dir, data_files, split, cache_dir, download_config, download_mode, ignore_verifications, save_infos, **config_kwargs) 518 download_mode=download_mode, 519 ignore_verifications=ignore_verifications, --> 520 save_infos=save_infos, 521 ) 522 /usr/local/lib/python3.6/dist-packages/nlp/builder.py in download_and_prepare(self, download_config, download_mode, ignore_verifications, save_infos, dl_manager, **download_and_prepare_kwargs) 363 verify_infos = not save_infos and not ignore_verifications 364 self._download_and_prepare( --> 365 dl_manager=dl_manager, verify_infos=verify_infos, **download_and_prepare_kwargs 366 ) 367 # Sync info /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _download_and_prepare(self, dl_manager, verify_infos, **prepare_split_kwargs) 416 try: 417 # Prepare split will record examples associated to the split --> 418 self._prepare_split(split_generator, **prepare_split_kwargs) 419 except OSError: 420 raise OSError("Cannot find data file. " + (self.MANUAL_DOWNLOAD_INSTRUCTIONS or "")) /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _prepare_split(self, split_generator) 594 example = self.info.features.encode_example(record) 595 writer.write(example) --> 596 num_examples, num_bytes = writer.finalize() 597 598 assert num_examples == num_examples, f"Expected to write {split_info.num_examples} but wrote {num_examples}" /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in finalize(self, close_stream) 173 def finalize(self, close_stream=True): 174 if self.pa_writer is not None: --> 175 self.write_on_file() 176 self.pa_writer.close() 177 if close_stream: /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in write_on_file(self) 124 else: 125 # All good --> 126 self._write_array_on_file(pa_array) 127 self.current_rows = [] 128 /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in _write_array_on_file(self, pa_array) 93 def _write_array_on_file(self, pa_array): 94 """Write a PyArrow Array""" ---> 95 pa_batch = pa.RecordBatch.from_struct_array(pa_array) 96 self._num_bytes += pa_array.nbytes 97 self.pa_writer.write_batch(pa_batch) AttributeError: type object 'pyarrow.lib.RecordBatch' has no attribute 'from_struct_array'
23
Loading 'wikitext' dataset fails Loading the 'wikitext' dataset fails with Attribute error: Code to reproduce (From example notebook): import nlp wikitext_dataset = nlp.load_dataset('wikitext') Error: --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) <ipython-input-17-d5d9df94b13c> in <module>() 11 12 # Load a dataset and print the first examples in the training set ---> 13 wikitext_dataset = nlp.load_dataset('wikitext') 14 print(wikitext_dataset['train'][0]) 6 frames /usr/local/lib/python3.6/dist-packages/nlp/load.py in load_dataset(path, name, version, data_dir, data_files, split, cache_dir, download_config, download_mode, ignore_verifications, save_infos, **config_kwargs) 518 download_mode=download_mode, 519 ignore_verifications=ignore_verifications, --> 520 save_infos=save_infos, 521 ) 522 /usr/local/lib/python3.6/dist-packages/nlp/builder.py in download_and_prepare(self, download_config, download_mode, ignore_verifications, save_infos, dl_manager, **download_and_prepare_kwargs) 363 verify_infos = not save_infos and not ignore_verifications 364 self._download_and_prepare( --> 365 dl_manager=dl_manager, verify_infos=verify_infos, **download_and_prepare_kwargs 366 ) 367 # Sync info /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _download_and_prepare(self, dl_manager, verify_infos, **prepare_split_kwargs) 416 try: 417 # Prepare split will record examples associated to the split --> 418 self._prepare_split(split_generator, **prepare_split_kwargs) 419 except OSError: 420 raise OSError("Cannot find data file. " + (self.MANUAL_DOWNLOAD_INSTRUCTIONS or "")) /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _prepare_split(self, split_generator) 594 example = self.info.features.encode_example(record) 595 writer.write(example) --> 596 num_examples, num_bytes = writer.finalize() 597 598 assert num_examples == num_examples, f"Expected to write {split_info.num_examples} but wrote {num_examples}" /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in finalize(self, close_stream) 173 def finalize(self, close_stream=True): 174 if self.pa_writer is not None: --> 175 self.write_on_file() 176 self.pa_writer.close() 177 if close_stream: /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in write_on_file(self) 124 else: 125 # All good --> 126 self._write_array_on_file(pa_array) 127 self.current_rows = [] 128 /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in _write_array_on_file(self, pa_array) 93 def _write_array_on_file(self, pa_array): 94 """Write a PyArrow Array""" ---> 95 pa_batch = pa.RecordBatch.from_struct_array(pa_array) 96 self._num_bytes += pa_array.nbytes 97 self.pa_writer.write_batch(pa_batch) AttributeError: type object 'pyarrow.lib.RecordBatch' has no attribute 'from_struct_array' Hi, The squad bug seems to be fixed, but the loading of the 'wikitext' still suffers from this problem (on Colab with pyarrow=0.17.1).
[ -0.11933788657188416, 0.0886276513338089, -0.008977650664746761, 0.21019726991653442, 0.3692685067653656, 0.026060238480567932, 0.44878119230270386, 0.41310203075408936, 0.43308186531066895, 0.012566424906253815, 0.12384431064128876, 0.39542996883392334, -0.1505494862794876, 0.021962128579616547, 0.12090560793876648, -0.4289896488189697, -0.033739328384399414, 0.36367329955101013, -0.04146643355488777, 0.06297355890274048, -0.1876838058233261, 0.10851654410362244, -0.2651917040348053, 0.255005806684494, -0.30572274327278137, -0.05702603608369827, 0.21592095494270325, 0.08206319063901901, -0.07891858369112015, -0.55900639295578, 0.2770833671092987, -0.21704953908920288, 0.0644511878490448, 0.11474639922380447, -0.00010995574848493561, 0.20530641078948975, 0.33269748091697693, 0.03021298721432686, -0.5799461603164673, -0.18957747519016266, -0.4773319959640503, -0.11893901973962784, 0.29743248224258423, -0.24970698356628418, -0.028513457626104355, 0.05389736220240593, 0.03591109812259674, -0.07441242039203644, 0.2967468500137329, 0.5315061807632446, 0.23788878321647644, 0.5236918926239014, 0.15454885363578796, 0.034493979066610336, -0.07856866717338562, -0.17057105898857117, -0.08211525529623032, 0.22396785020828247, 0.07387740910053253, -0.39154455065727234, -0.013374139554798603, 0.12692393362522125, 0.15117257833480835, 0.26370176672935486, 0.5083714723587036, 0.014393548481166363, 0.2405228316783905, -0.15062619745731354, 0.014946400187909603, 0.07640086859464645, 0.3353831470012665, -0.16493351757526398, -0.27160072326660156, -0.2243598848581314, 0.2476663589477539, -0.33475664258003235, 0.23653079569339752, -0.036676887422800064, -0.16220447421073914, -0.008197000250220299, -0.2692282497882843, -0.028147514909505844, -0.24182115495204926, 0.35268786549568176, -0.05821506306529045, 0.4996490478515625, 0.04358331486582756, 0.14274749159812927, -0.058065660297870636, -0.07545804977416992, 0.14037638902664185, -0.08709083497524261, -0.12930019199848175, 0.2578132152557373, -0.38315585255622864, -0.03646787628531456, -0.034946393221616745, -0.1169760525226593, -0.031624484807252884, 0.2635209858417511, 0.2199900895357132, -0.2248735874891281, 0.3238379955291748, 0.3154149651527405, 0.22298957407474518, 0.42046910524368286, 0.07276459038257599, 0.06359180808067322, -0.08280493319034576, 0.13602527976036072, -0.23372802138328552, -0.15548524260520935, -0.15553590655326843, -0.17655469477176666, 0.09289750456809998, -0.15460263192653656, 0.22602041065692902, -0.07111654430627823, -0.49847400188446045, 0.11472059786319733, -0.1434047520160675, 0.04629036784172058, 0.20511473715305328, 0.17948518693447113, -0.12763333320617676, 0.23223552107810974, 0.14412418007850647, 0.2825672924518585, -0.2519769072532654, -0.04313638061285019, -0.16628925502300262, 0.24921317398548126, -0.23438338935375214, -0.18224935233592987, 0.24275411665439606, 0.19340045750141144, 0.3994779884815216, -0.10116773843765259, 0.11126504093408585, -0.20175516605377197, 0.22737497091293335, -0.04551902040839195, -0.12887892127037048, 0.25665295124053955, -0.21606145799160004, 0.15489806234836578, 0.20159101486206055, -0.2824234366416931, -0.08482398837804794, -0.10174469649791718, -0.27263587713241577, -0.3080922067165375, -0.12625131011009216, 0.281415194272995, 0.14738169312477112, -0.24296516180038452, -0.15226291120052338, -0.06118733063340187, 0.14749260246753693, -0.3687240481376648, -0.11810583621263504, -0.46125921607017517, -0.22063837945461273, -0.023005429655313492, 0.3976428508758545, 0.20490917563438416, 0.131386861205101, -0.10897225141525269, -0.11542121320962906, 0.07785490155220032, -0.03179694712162018, 0.12271356582641602, -0.43835729360580444, 0.5421618223190308, -0.06105572730302811, 0.11886267364025116, 0.6413458585739136, -0.34278684854507446, -0.4578539729118347, 0.12356847524642944, -0.014279641211032867, 0.2093287706375122, -0.22536881268024445, -0.016720760613679886, -0.013226832263171673, -0.1845516413450241, 0.2123139649629593, 0.6192137002944946, -0.015043232589960098, 0.0824793204665184, -0.12689664959907532, -0.01554853655397892, 0.5002202987670898, 0.10985670238733292, 0.09252531081438065, -0.007928650826215744, -0.04170486703515053, 0.4966488778591156, -0.0013133209431543946, -0.12597094476222992, -0.014860233291983604, -0.013630262576043606, -0.06468662619590759, -0.104604572057724, -0.21782620251178741, -0.073886938393116, -0.27766478061676025, 0.3043038845062256, -0.11754702031612396, -0.0585317425429821, -0.04715042933821678, 0.1061868816614151, -0.34911301732063293, 0.16848920285701752, -0.423268586397171, -0.37731868028640747, 0.17209294438362122, 0.1551334410905838, -0.01082319114357233, 0.2384408712387085, -0.20474199950695038, 0.21943683922290802, -0.27402806282043457, 0.05550078675150871, -0.21892870962619781, 0.1612894982099533, -0.055882859975099564, -0.0424220934510231, -0.009479359723627567, 0.30879732966423035, 0.03833145648241043, -0.017996225506067276, -0.3062887489795685, 0.286030650138855, -0.1254843771457672, 0.08035652339458466, -0.03226462006568909, -0.14337000250816345, 0.06880593299865723, -0.07154339551925659, -0.17518380284309387, 0.20819160342216492, 0.3319510519504547, -0.18012623488903046, -0.04314308613538742, 0.11287619918584824, 0.12721261382102966, 0.08142238110303879, 0.09886393696069717, 0.16397275030612946, 0.1439925581216812, -0.06582631915807724, 0.10127968341112137, 0.07743124663829803, 0.2337101697921753, -0.09498753398656845, 0.20471026003360748, -0.15645122528076172, -0.16714543104171753, -0.11880633234977722, 0.31311914324760437, 0.17005811631679535, 0.1674172282218933, 0.08148688822984695, -0.27525240182876587, -0.09487471729516983, 0.07855787128210068, 0.1460483968257904, 0.4009645879268646, -0.051832713186740875, -0.10834138095378876, -0.05257111415266991, -0.4165973961353302, 0.07112560421228409, 0.1707080900669098, 0.03662936016917229, 0.20912867784500122, 0.3018558621406555, 0.04131574556231499, 0.08731992542743683, -0.29526659846305847, -0.35701724886894226, 0.2824332118034363, 0.44053006172180176, -0.22848272323608398, 0.2078799307346344, -0.0305313840508461, -0.0025109166745096445, -0.06784793734550476, -0.33363422751426697, -0.18862943351268768, -0.34299421310424805, -0.12133592367172241, 0.4337747097015381, 0.024118667468428612, 0.1810036450624466, -0.3977627158164978, -0.022050168365240097, 0.19475045800209045, -0.1147860512137413, -0.03137960657477379, -0.31579849123954773, -0.2693284749984741, 0.045942142605781555, 0.3818604350090027, 0.09107596427202225, 0.3510979115962982, -0.06677016615867615, -0.09253522008657455, -0.007033964619040489, -0.05841490253806114, -0.07556762546300888, -0.18372154235839844, 0.24404771625995636, 0.009521574713289738, 0.2855835556983948, -0.17702993750572205, -0.5220329761505127, 0.5017470121383667, -0.25570252537727356, -0.05778523162007332, 0.31510716676712036, 0.27111291885375977, -0.24919019639492035, -0.0794772058725357, -0.374688059091568, -0.1264500468969345, -0.28060564398765564, -0.05946861207485199, 0.07482945919036865, 0.06202014163136482, 0.3974682688713074, 0.07085619866847992, 0.21601079404354095, -0.002573415171355009, 0.11343801766633987, -0.04826263710856438, -0.0900639072060585, 0.1859847605228424, -0.31003302335739136, -0.31731799244880676, -0.12134335190057755, 0.00993572361767292, 0.3347616195678711, 0.11073707789182663, -0.3489590883255005, -0.21413186192512512, -0.10495806485414505, 0.08531273156404495, -0.2213255912065506, -0.040067099034786224, 0.18285974860191345, 0.07230282574892044, -0.06572254747152328, -0.08664463460445404, 0.03598904609680176, 0.07123992592096329, 0.15234632790088654, 0.13496197760105133, -0.006304681766778231, 0.08922669291496277, -0.08488419651985168, 0.6008992195129395, 0.21869337558746338, -0.2143774777650833, 0.30028587579727173, -0.10327647626399994, 0.005145025905221701, -0.20868363976478577, -0.43980398774147034, 0.2641644775867462, -0.11523281037807465, 0.1829284131526947, 0.17231453955173492, -0.07949270308017731, -0.22478418052196503, -0.19255878031253815, 0.06885252147912979, -0.17844729125499725, -0.27277931571006775, 0.14616899192333221, -0.21739473938941956, 0.278711199760437, -0.0688643828034401, 0.22253714501857758, -0.20522361993789673, -0.16798114776611328, 0.06985098868608475, 0.16693316400051117, -0.09054534137248993, 0.07967454195022583, -0.24400153756141663, -0.2761092185974121, -0.47995808720588684, -0.027839848771691322, 0.2410319298505783, 0.643750786781311, -0.08638402819633484, 0.037189047783613205, 0.0807223692536354, 0.08174136281013489, 0.13737380504608154, -0.23013576865196228, 0.1822637915611267, 0.2891305088996887, 0.031824760138988495, -0.3197470009326935, -0.06433095782995224, -0.1401122510433197, 0.35801708698272705, 0.07665061950683594, -0.05443219840526581, -0.03723124787211418, -0.16765417158603668, 0.13983193039894104, 0.25993993878364563, -0.18729501962661743, -0.04411264508962631, -0.09173404425382614, -0.0898098275065422, -0.30776384472846985, -0.006155079230666161, -0.009947403334081173, 0.3356599807739258, 0.12296963483095169, 0.26572561264038086, -0.5462993383407593, 0.11665675789117813, 0.09016667306423187, 0.17458726465702057, 0.09595341980457306, 0.057944491505622864, 0.18724635243415833, -0.16663426160812378, 0.183209627866745, -0.14590534567832947, 0.5633814334869385, 0.14737607538700104, -0.33079853653907776, -0.31953001022338867, 0.021663723513484, 0.16743426024913788, 0.18398089706897736, -0.018585436046123505, -0.1411214917898178, -0.276946485042572, -0.06146418675780296, -0.008709541521966457, 0.30457016825675964, 0.5610408186912537, 0.0026278814766556025, -0.3271811902523041, -0.2585411071777344, 0.48284009099006653, -0.2850036919116974, 0.11732959747314453, 0.32155776023864746, 0.05337808281183243, -0.3131917119026184, 0.08413904905319214, 0.07783257216215134, 0.6767759323120117, 0.1695389300584793, 0.3111467659473419, 0.3826707899570465, 0.18392953276634216, 0.4142702519893646, -0.043089646846055984, 0.28471478819847107, -0.3927352726459503, -0.13342542946338654, -0.08068593591451645, -0.06546245515346527, 0.10708130896091461, -0.00887344777584076, -0.025437481701374054, 0.1506209671497345, 0.15688923001289368, -0.0041417344473302364, 0.05714833736419678, 0.4492526948451996, -0.03878338634967804, -0.19681920111179352, -0.34702637791633606, 0.15601642429828644, -0.218000128865242, 0.29750892519950867, -0.09073349088430405, -0.06648434698581696, 0.05520431324839592, -0.06735192984342575, -0.2555583119392395, 0.2574388384819031, -0.29947420954704285, 0.4594575762748718, 0.057822249829769135, -0.47568145394325256, 0.1984158307313919, 0.20709721744060516, 0.17566342651844025, 0.005698441993445158, -0.2995951473712921, 0.2387789934873581, -0.2900906801223755, -0.2218656986951828, -0.06728668510913849, 0.11194741725921631, 0.4200740456581116, -0.1725509911775589, -0.14637738466262817, 0.1479404866695404, -0.12065477669239044, -0.2415027767419815, 0.10395903885364532, 0.036866601556539536, 0.35248854756355286, -0.39257586002349854, -0.5152192711830139, -0.0024203567299991846, -0.18207542598247528, -0.17490997910499573, 0.17449994385242462, 0.023450996726751328, -0.18903093039989471, 0.15189528465270996, 0.14982764422893524, -0.1989799439907074, 0.1164064109325409, 0.24804162979125977, -0.2495405673980713, 0.1462835967540741, 0.763757050037384, 0.033945709466934204, 0.012805157341063023, -0.31383034586906433, -0.05428868532180786, 0.12508492171764374, -0.4820001423358917, -0.11712706089019775, 0.1255035698413849, 0.16912545263767242, -0.14424709975719452, 0.4983648657798767, 0.06302112340927124, -0.018845530226826668, 0.2471855729818344, -0.5424144268035889, -0.10621847957372665, -0.04302152618765831, 0.017753472551703453, 0.23777538537979126, -0.0026242828462272882, 0.11488918960094452, 0.13550932705402374, 0.10338813066482544, -0.27263733744621277, 0.014253862202167511, -0.15762493014335632, 0.1379975527524948, 0.21783483028411865, -0.15653111040592194, 0.22328241169452667, -0.07694793492555618, 0.15638624131679535, -0.09584680199623108, -0.12303421646356583, -0.24829301238059998, -0.25483760237693787, 0.10011181235313416, 0.24556748569011688, -0.16452161967754364, -0.03067888505756855, -0.10233359783887863, 0.02367708459496498, -0.23110559582710266, -0.008462597616016865, -0.11772827059030533, -0.29334336519241333, 0.38943231105804443, 0.043897844851017, 0.2783689498901367, -0.1116899773478508, 0.12635605037212372, -0.3007572293281555, 0.14912119507789612, -0.018496258184313774, -0.07522967457771301, 0.27581310272216797, 0.09365364164113998, -0.46588754653930664, 0.005421184469014406, -0.40260615944862366, 0.037148136645555496, 0.36982107162475586, -0.45511066913604736, 0.15995857119560242, -0.1316279023885727, 0.09546560794115067, 0.4656839668750763, -0.2714088261127472, -0.0036662661004811525, 0.34966516494750977, 0.2378423660993576, -0.2974173426628113, -0.18925972282886505, 0.23999083042144775, 0.06134593486785889, 0.032459061592817307, 0.027239473536610603, 0.04666171222925186, -0.3037489652633667, -0.32123425602912903, 0.08927872031927109, -0.029339971020817757, -0.265601247549057, -0.011407657526433468, 0.3393266201019287, -0.002364314394071698, -0.05951515957713127, -0.05004987120628357, 0.21102137863636017, 0.035392459481954575, 0.38269418478012085, -0.29260048270225525, 0.20699812471866608, 0.05926388129591942, 0.04391023516654968, -0.11114337295293808, -0.21921710669994354, 0.05030057579278946, -0.157841295003891, 0.03383495658636093, 0.29528310894966125, 0.09981518238782883, 0.1530808061361313, -0.285504013299942, -0.12453760206699371, -0.20802302658557892, 0.22792354226112366, -0.17904223501682281, -0.04890627786517143, -0.24299491941928864, 0.021017028018832207, -0.025876162573695183, -0.09469947963953018, -0.14167042076587677, -0.03817928954958916, 0.3095564842224121, 0.2532219886779785, -0.1722697913646698, -0.45553529262542725, 0.10826699435710907, 0.3432939052581787, 0.18004795908927917, -0.24588418006896973, 0.33685368299484253, 0.19381506741046906, -0.06877820193767548, -0.01412354689091444, 0.39711013436317444, 0.4357686936855316, 0.33839377760887146, -0.3642207980155945, 0.08116745948791504, -0.2033807784318924, -0.0568217858672142, -0.02326009050011635, 0.26951277256011963, 0.10607278347015381, 0.11737696081399918, 0.4624232351779938, 0.12148629873991013, -0.12472685426473618, 0.04363814741373062, -0.08805640786886215, -0.2605327367782593, -0.26294752955436707, 0.5191484093666077, 0.185794860124588, -0.11438807845115662, -0.3206513226032257, 0.1899377405643463, -0.2749866843223572, -0.0809149518609047, 0.06285153329372406, 0.24058951437473297, 0.113153375685215, -0.21382735669612885, 0.07050874084234238, 0.00031574213062413037, 0.7061741352081299, 0.48772212862968445, -0.01771952025592327, -0.3658044636249542, -0.11652323603630066, -0.5989870429039001, 0.2178366482257843, -0.3639686405658722, -0.26901865005493164, 0.2833371162414551, -0.07233183085918427, -0.28808191418647766, 0.24852752685546875, 0.12761986255645752, 0.15819624066352844, -0.05727671831846237, 0.0983472391963005, -0.49076488614082336, -0.1486053764820099, -0.03065069019794464, -0.0076751429587602615, -0.0831807479262352, -0.25705254077911377, 0.1453229933977127, -0.4349938929080963, 0.1161026805639267, 0.011985591612756252, 0.06934838742017746, -0.0002399603690719232, 0.09664099663496017, 0.3544251024723053, 0.16532465815544128, 0.5012484788894653, -0.035560812801122665, 0.0036571817472577095, -0.19115544855594635, -0.15662235021591187, -0.22899216413497925, 0.13182324171066284, 0.08762864023447037, 0.5133113861083984, -0.026107320562005043, -0.06854572147130966, -0.2640267312526703, 0.40316781401634216, -0.08100741356611252, 0.08831649273633957, -0.13508076965808868, 0.07731454819440842, -0.07035040855407715, 0.22758440673351288, 0.26436471939086914, 0.4245404899120331, -0.17832162976264954, 0.20220930874347687, -0.27229106426239014, -0.4602454900741577, 0.32893526554107666, -0.6614815592765808, -0.5177910923957825, 0.07512339949607849, 0.11619257926940918, -0.1423347145318985, 0.05819396302103996, -0.6006348133087158, 0.0705668106675148, 0.24261273443698883, -0.11140530556440353, -0.3171199560165405, 0.21794475615024567, 0.016542308032512665, 0.08000080287456512, 0.043573711067438126, 0.44835665822029114, -0.14847348630428314, -0.15976785123348236, -0.004933878779411316, -0.1632518470287323 ]
https://github.com/huggingface/datasets/issues/168
Loading 'wikitext' dataset fails
When you install `nlp` for the first time on a Colab runtime, it updates the `pyarrow` library that was already on colab. This update shows this message on colab: ``` WARNING: The following packages were previously imported in this runtime: [pyarrow] You must restart the runtime in order to use newly installed versions. ``` You just have to restart the runtime and it should be fine.
Loading the 'wikitext' dataset fails with Attribute error: Code to reproduce (From example notebook): import nlp wikitext_dataset = nlp.load_dataset('wikitext') Error: --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) <ipython-input-17-d5d9df94b13c> in <module>() 11 12 # Load a dataset and print the first examples in the training set ---> 13 wikitext_dataset = nlp.load_dataset('wikitext') 14 print(wikitext_dataset['train'][0]) 6 frames /usr/local/lib/python3.6/dist-packages/nlp/load.py in load_dataset(path, name, version, data_dir, data_files, split, cache_dir, download_config, download_mode, ignore_verifications, save_infos, **config_kwargs) 518 download_mode=download_mode, 519 ignore_verifications=ignore_verifications, --> 520 save_infos=save_infos, 521 ) 522 /usr/local/lib/python3.6/dist-packages/nlp/builder.py in download_and_prepare(self, download_config, download_mode, ignore_verifications, save_infos, dl_manager, **download_and_prepare_kwargs) 363 verify_infos = not save_infos and not ignore_verifications 364 self._download_and_prepare( --> 365 dl_manager=dl_manager, verify_infos=verify_infos, **download_and_prepare_kwargs 366 ) 367 # Sync info /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _download_and_prepare(self, dl_manager, verify_infos, **prepare_split_kwargs) 416 try: 417 # Prepare split will record examples associated to the split --> 418 self._prepare_split(split_generator, **prepare_split_kwargs) 419 except OSError: 420 raise OSError("Cannot find data file. " + (self.MANUAL_DOWNLOAD_INSTRUCTIONS or "")) /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _prepare_split(self, split_generator) 594 example = self.info.features.encode_example(record) 595 writer.write(example) --> 596 num_examples, num_bytes = writer.finalize() 597 598 assert num_examples == num_examples, f"Expected to write {split_info.num_examples} but wrote {num_examples}" /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in finalize(self, close_stream) 173 def finalize(self, close_stream=True): 174 if self.pa_writer is not None: --> 175 self.write_on_file() 176 self.pa_writer.close() 177 if close_stream: /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in write_on_file(self) 124 else: 125 # All good --> 126 self._write_array_on_file(pa_array) 127 self.current_rows = [] 128 /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in _write_array_on_file(self, pa_array) 93 def _write_array_on_file(self, pa_array): 94 """Write a PyArrow Array""" ---> 95 pa_batch = pa.RecordBatch.from_struct_array(pa_array) 96 self._num_bytes += pa_array.nbytes 97 self.pa_writer.write_batch(pa_batch) AttributeError: type object 'pyarrow.lib.RecordBatch' has no attribute 'from_struct_array'
66
Loading 'wikitext' dataset fails Loading the 'wikitext' dataset fails with Attribute error: Code to reproduce (From example notebook): import nlp wikitext_dataset = nlp.load_dataset('wikitext') Error: --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) <ipython-input-17-d5d9df94b13c> in <module>() 11 12 # Load a dataset and print the first examples in the training set ---> 13 wikitext_dataset = nlp.load_dataset('wikitext') 14 print(wikitext_dataset['train'][0]) 6 frames /usr/local/lib/python3.6/dist-packages/nlp/load.py in load_dataset(path, name, version, data_dir, data_files, split, cache_dir, download_config, download_mode, ignore_verifications, save_infos, **config_kwargs) 518 download_mode=download_mode, 519 ignore_verifications=ignore_verifications, --> 520 save_infos=save_infos, 521 ) 522 /usr/local/lib/python3.6/dist-packages/nlp/builder.py in download_and_prepare(self, download_config, download_mode, ignore_verifications, save_infos, dl_manager, **download_and_prepare_kwargs) 363 verify_infos = not save_infos and not ignore_verifications 364 self._download_and_prepare( --> 365 dl_manager=dl_manager, verify_infos=verify_infos, **download_and_prepare_kwargs 366 ) 367 # Sync info /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _download_and_prepare(self, dl_manager, verify_infos, **prepare_split_kwargs) 416 try: 417 # Prepare split will record examples associated to the split --> 418 self._prepare_split(split_generator, **prepare_split_kwargs) 419 except OSError: 420 raise OSError("Cannot find data file. " + (self.MANUAL_DOWNLOAD_INSTRUCTIONS or "")) /usr/local/lib/python3.6/dist-packages/nlp/builder.py in _prepare_split(self, split_generator) 594 example = self.info.features.encode_example(record) 595 writer.write(example) --> 596 num_examples, num_bytes = writer.finalize() 597 598 assert num_examples == num_examples, f"Expected to write {split_info.num_examples} but wrote {num_examples}" /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in finalize(self, close_stream) 173 def finalize(self, close_stream=True): 174 if self.pa_writer is not None: --> 175 self.write_on_file() 176 self.pa_writer.close() 177 if close_stream: /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in write_on_file(self) 124 else: 125 # All good --> 126 self._write_array_on_file(pa_array) 127 self.current_rows = [] 128 /usr/local/lib/python3.6/dist-packages/nlp/arrow_writer.py in _write_array_on_file(self, pa_array) 93 def _write_array_on_file(self, pa_array): 94 """Write a PyArrow Array""" ---> 95 pa_batch = pa.RecordBatch.from_struct_array(pa_array) 96 self._num_bytes += pa_array.nbytes 97 self.pa_writer.write_batch(pa_batch) AttributeError: type object 'pyarrow.lib.RecordBatch' has no attribute 'from_struct_array' When you install `nlp` for the first time on a Colab runtime, it updates the `pyarrow` library that was already on colab. This update shows this message on colab: ``` WARNING: The following packages were previously imported in this runtime: [pyarrow] You must restart the runtime in order to use newly installed versions. ``` You just have to restart the runtime and it should be fine.
[ -0.11933788657188416, 0.0886276513338089, -0.008977650664746761, 0.21019726991653442, 0.3692685067653656, 0.026060238480567932, 0.44878119230270386, 0.41310203075408936, 0.43308186531066895, 0.012566424906253815, 0.12384431064128876, 0.39542996883392334, -0.1505494862794876, 0.021962128579616547, 0.12090560793876648, -0.4289896488189697, -0.033739328384399414, 0.36367329955101013, -0.04146643355488777, 0.06297355890274048, -0.1876838058233261, 0.10851654410362244, -0.2651917040348053, 0.255005806684494, -0.30572274327278137, -0.05702603608369827, 0.21592095494270325, 0.08206319063901901, -0.07891858369112015, -0.55900639295578, 0.2770833671092987, -0.21704953908920288, 0.0644511878490448, 0.11474639922380447, -0.00010995574848493561, 0.20530641078948975, 0.33269748091697693, 0.03021298721432686, -0.5799461603164673, -0.18957747519016266, -0.4773319959640503, -0.11893901973962784, 0.29743248224258423, -0.24970698356628418, -0.028513457626104355, 0.05389736220240593, 0.03591109812259674, -0.07441242039203644, 0.2967468500137329, 0.5315061807632446, 0.23788878321647644, 0.5236918926239014, 0.15454885363578796, 0.034493979066610336, -0.07856866717338562, -0.17057105898857117, -0.08211525529623032, 0.22396785020828247, 0.07387740910053253, -0.39154455065727234, -0.013374139554798603, 0.12692393362522125, 0.15117257833480835, 0.26370176672935486, 0.5083714723587036, 0.014393548481166363, 0.2405228316783905, -0.15062619745731354, 0.014946400187909603, 0.07640086859464645, 0.3353831470012665, -0.16493351757526398, -0.27160072326660156, -0.2243598848581314, 0.2476663589477539, -0.33475664258003235, 0.23653079569339752, -0.036676887422800064, -0.16220447421073914, -0.008197000250220299, -0.2692282497882843, -0.028147514909505844, -0.24182115495204926, 0.35268786549568176, -0.05821506306529045, 0.4996490478515625, 0.04358331486582756, 0.14274749159812927, -0.058065660297870636, -0.07545804977416992, 0.14037638902664185, -0.08709083497524261, -0.12930019199848175, 0.2578132152557373, -0.38315585255622864, -0.03646787628531456, -0.034946393221616745, -0.1169760525226593, -0.031624484807252884, 0.2635209858417511, 0.2199900895357132, -0.2248735874891281, 0.3238379955291748, 0.3154149651527405, 0.22298957407474518, 0.42046910524368286, 0.07276459038257599, 0.06359180808067322, -0.08280493319034576, 0.13602527976036072, -0.23372802138328552, -0.15548524260520935, -0.15553590655326843, -0.17655469477176666, 0.09289750456809998, -0.15460263192653656, 0.22602041065692902, -0.07111654430627823, -0.49847400188446045, 0.11472059786319733, -0.1434047520160675, 0.04629036784172058, 0.20511473715305328, 0.17948518693447113, -0.12763333320617676, 0.23223552107810974, 0.14412418007850647, 0.2825672924518585, -0.2519769072532654, -0.04313638061285019, -0.16628925502300262, 0.24921317398548126, -0.23438338935375214, -0.18224935233592987, 0.24275411665439606, 0.19340045750141144, 0.3994779884815216, -0.10116773843765259, 0.11126504093408585, -0.20175516605377197, 0.22737497091293335, -0.04551902040839195, -0.12887892127037048, 0.25665295124053955, -0.21606145799160004, 0.15489806234836578, 0.20159101486206055, -0.2824234366416931, -0.08482398837804794, -0.10174469649791718, -0.27263587713241577, -0.3080922067165375, -0.12625131011009216, 0.281415194272995, 0.14738169312477112, -0.24296516180038452, -0.15226291120052338, -0.06118733063340187, 0.14749260246753693, -0.3687240481376648, -0.11810583621263504, -0.46125921607017517, -0.22063837945461273, -0.023005429655313492, 0.3976428508758545, 0.20490917563438416, 0.131386861205101, -0.10897225141525269, -0.11542121320962906, 0.07785490155220032, -0.03179694712162018, 0.12271356582641602, -0.43835729360580444, 0.5421618223190308, -0.06105572730302811, 0.11886267364025116, 0.6413458585739136, -0.34278684854507446, -0.4578539729118347, 0.12356847524642944, -0.014279641211032867, 0.2093287706375122, -0.22536881268024445, -0.016720760613679886, -0.013226832263171673, -0.1845516413450241, 0.2123139649629593, 0.6192137002944946, -0.015043232589960098, 0.0824793204665184, -0.12689664959907532, -0.01554853655397892, 0.5002202987670898, 0.10985670238733292, 0.09252531081438065, -0.007928650826215744, -0.04170486703515053, 0.4966488778591156, -0.0013133209431543946, -0.12597094476222992, -0.014860233291983604, -0.013630262576043606, -0.06468662619590759, -0.104604572057724, -0.21782620251178741, -0.073886938393116, -0.27766478061676025, 0.3043038845062256, -0.11754702031612396, -0.0585317425429821, -0.04715042933821678, 0.1061868816614151, -0.34911301732063293, 0.16848920285701752, -0.423268586397171, -0.37731868028640747, 0.17209294438362122, 0.1551334410905838, -0.01082319114357233, 0.2384408712387085, -0.20474199950695038, 0.21943683922290802, -0.27402806282043457, 0.05550078675150871, -0.21892870962619781, 0.1612894982099533, -0.055882859975099564, -0.0424220934510231, -0.009479359723627567, 0.30879732966423035, 0.03833145648241043, -0.017996225506067276, -0.3062887489795685, 0.286030650138855, -0.1254843771457672, 0.08035652339458466, -0.03226462006568909, -0.14337000250816345, 0.06880593299865723, -0.07154339551925659, -0.17518380284309387, 0.20819160342216492, 0.3319510519504547, -0.18012623488903046, -0.04314308613538742, 0.11287619918584824, 0.12721261382102966, 0.08142238110303879, 0.09886393696069717, 0.16397275030612946, 0.1439925581216812, -0.06582631915807724, 0.10127968341112137, 0.07743124663829803, 0.2337101697921753, -0.09498753398656845, 0.20471026003360748, -0.15645122528076172, -0.16714543104171753, -0.11880633234977722, 0.31311914324760437, 0.17005811631679535, 0.1674172282218933, 0.08148688822984695, -0.27525240182876587, -0.09487471729516983, 0.07855787128210068, 0.1460483968257904, 0.4009645879268646, -0.051832713186740875, -0.10834138095378876, -0.05257111415266991, -0.4165973961353302, 0.07112560421228409, 0.1707080900669098, 0.03662936016917229, 0.20912867784500122, 0.3018558621406555, 0.04131574556231499, 0.08731992542743683, -0.29526659846305847, -0.35701724886894226, 0.2824332118034363, 0.44053006172180176, -0.22848272323608398, 0.2078799307346344, -0.0305313840508461, -0.0025109166745096445, -0.06784793734550476, -0.33363422751426697, -0.18862943351268768, -0.34299421310424805, -0.12133592367172241, 0.4337747097015381, 0.024118667468428612, 0.1810036450624466, -0.3977627158164978, -0.022050168365240097, 0.19475045800209045, -0.1147860512137413, -0.03137960657477379, -0.31579849123954773, -0.2693284749984741, 0.045942142605781555, 0.3818604350090027, 0.09107596427202225, 0.3510979115962982, -0.06677016615867615, -0.09253522008657455, -0.007033964619040489, -0.05841490253806114, -0.07556762546300888, -0.18372154235839844, 0.24404771625995636, 0.009521574713289738, 0.2855835556983948, -0.17702993750572205, -0.5220329761505127, 0.5017470121383667, -0.25570252537727356, -0.05778523162007332, 0.31510716676712036, 0.27111291885375977, -0.24919019639492035, -0.0794772058725357, -0.374688059091568, -0.1264500468969345, -0.28060564398765564, -0.05946861207485199, 0.07482945919036865, 0.06202014163136482, 0.3974682688713074, 0.07085619866847992, 0.21601079404354095, -0.002573415171355009, 0.11343801766633987, -0.04826263710856438, -0.0900639072060585, 0.1859847605228424, -0.31003302335739136, -0.31731799244880676, -0.12134335190057755, 0.00993572361767292, 0.3347616195678711, 0.11073707789182663, -0.3489590883255005, -0.21413186192512512, -0.10495806485414505, 0.08531273156404495, -0.2213255912065506, -0.040067099034786224, 0.18285974860191345, 0.07230282574892044, -0.06572254747152328, -0.08664463460445404, 0.03598904609680176, 0.07123992592096329, 0.15234632790088654, 0.13496197760105133, -0.006304681766778231, 0.08922669291496277, -0.08488419651985168, 0.6008992195129395, 0.21869337558746338, -0.2143774777650833, 0.30028587579727173, -0.10327647626399994, 0.005145025905221701, -0.20868363976478577, -0.43980398774147034, 0.2641644775867462, -0.11523281037807465, 0.1829284131526947, 0.17231453955173492, -0.07949270308017731, -0.22478418052196503, -0.19255878031253815, 0.06885252147912979, -0.17844729125499725, -0.27277931571006775, 0.14616899192333221, -0.21739473938941956, 0.278711199760437, -0.0688643828034401, 0.22253714501857758, -0.20522361993789673, -0.16798114776611328, 0.06985098868608475, 0.16693316400051117, -0.09054534137248993, 0.07967454195022583, -0.24400153756141663, -0.2761092185974121, -0.47995808720588684, -0.027839848771691322, 0.2410319298505783, 0.643750786781311, -0.08638402819633484, 0.037189047783613205, 0.0807223692536354, 0.08174136281013489, 0.13737380504608154, -0.23013576865196228, 0.1822637915611267, 0.2891305088996887, 0.031824760138988495, -0.3197470009326935, -0.06433095782995224, -0.1401122510433197, 0.35801708698272705, 0.07665061950683594, -0.05443219840526581, -0.03723124787211418, -0.16765417158603668, 0.13983193039894104, 0.25993993878364563, -0.18729501962661743, -0.04411264508962631, -0.09173404425382614, -0.0898098275065422, -0.30776384472846985, -0.006155079230666161, -0.009947403334081173, 0.3356599807739258, 0.12296963483095169, 0.26572561264038086, -0.5462993383407593, 0.11665675789117813, 0.09016667306423187, 0.17458726465702057, 0.09595341980457306, 0.057944491505622864, 0.18724635243415833, -0.16663426160812378, 0.183209627866745, -0.14590534567832947, 0.5633814334869385, 0.14737607538700104, -0.33079853653907776, -0.31953001022338867, 0.021663723513484, 0.16743426024913788, 0.18398089706897736, -0.018585436046123505, -0.1411214917898178, -0.276946485042572, -0.06146418675780296, -0.008709541521966457, 0.30457016825675964, 0.5610408186912537, 0.0026278814766556025, -0.3271811902523041, -0.2585411071777344, 0.48284009099006653, -0.2850036919116974, 0.11732959747314453, 0.32155776023864746, 0.05337808281183243, -0.3131917119026184, 0.08413904905319214, 0.07783257216215134, 0.6767759323120117, 0.1695389300584793, 0.3111467659473419, 0.3826707899570465, 0.18392953276634216, 0.4142702519893646, -0.043089646846055984, 0.28471478819847107, -0.3927352726459503, -0.13342542946338654, -0.08068593591451645, -0.06546245515346527, 0.10708130896091461, -0.00887344777584076, -0.025437481701374054, 0.1506209671497345, 0.15688923001289368, -0.0041417344473302364, 0.05714833736419678, 0.4492526948451996, -0.03878338634967804, -0.19681920111179352, -0.34702637791633606, 0.15601642429828644, -0.218000128865242, 0.29750892519950867, -0.09073349088430405, -0.06648434698581696, 0.05520431324839592, -0.06735192984342575, -0.2555583119392395, 0.2574388384819031, -0.29947420954704285, 0.4594575762748718, 0.057822249829769135, -0.47568145394325256, 0.1984158307313919, 0.20709721744060516, 0.17566342651844025, 0.005698441993445158, -0.2995951473712921, 0.2387789934873581, -0.2900906801223755, -0.2218656986951828, -0.06728668510913849, 0.11194741725921631, 0.4200740456581116, -0.1725509911775589, -0.14637738466262817, 0.1479404866695404, -0.12065477669239044, -0.2415027767419815, 0.10395903885364532, 0.036866601556539536, 0.35248854756355286, -0.39257586002349854, -0.5152192711830139, -0.0024203567299991846, -0.18207542598247528, -0.17490997910499573, 0.17449994385242462, 0.023450996726751328, -0.18903093039989471, 0.15189528465270996, 0.14982764422893524, -0.1989799439907074, 0.1164064109325409, 0.24804162979125977, -0.2495405673980713, 0.1462835967540741, 0.763757050037384, 0.033945709466934204, 0.012805157341063023, -0.31383034586906433, -0.05428868532180786, 0.12508492171764374, -0.4820001423358917, -0.11712706089019775, 0.1255035698413849, 0.16912545263767242, -0.14424709975719452, 0.4983648657798767, 0.06302112340927124, -0.018845530226826668, 0.2471855729818344, -0.5424144268035889, -0.10621847957372665, -0.04302152618765831, 0.017753472551703453, 0.23777538537979126, -0.0026242828462272882, 0.11488918960094452, 0.13550932705402374, 0.10338813066482544, -0.27263733744621277, 0.014253862202167511, -0.15762493014335632, 0.1379975527524948, 0.21783483028411865, -0.15653111040592194, 0.22328241169452667, -0.07694793492555618, 0.15638624131679535, -0.09584680199623108, -0.12303421646356583, -0.24829301238059998, -0.25483760237693787, 0.10011181235313416, 0.24556748569011688, -0.16452161967754364, -0.03067888505756855, -0.10233359783887863, 0.02367708459496498, -0.23110559582710266, -0.008462597616016865, -0.11772827059030533, -0.29334336519241333, 0.38943231105804443, 0.043897844851017, 0.2783689498901367, -0.1116899773478508, 0.12635605037212372, -0.3007572293281555, 0.14912119507789612, -0.018496258184313774, -0.07522967457771301, 0.27581310272216797, 0.09365364164113998, -0.46588754653930664, 0.005421184469014406, -0.40260615944862366, 0.037148136645555496, 0.36982107162475586, -0.45511066913604736, 0.15995857119560242, -0.1316279023885727, 0.09546560794115067, 0.4656839668750763, -0.2714088261127472, -0.0036662661004811525, 0.34966516494750977, 0.2378423660993576, -0.2974173426628113, -0.18925972282886505, 0.23999083042144775, 0.06134593486785889, 0.032459061592817307, 0.027239473536610603, 0.04666171222925186, -0.3037489652633667, -0.32123425602912903, 0.08927872031927109, -0.029339971020817757, -0.265601247549057, -0.011407657526433468, 0.3393266201019287, -0.002364314394071698, -0.05951515957713127, -0.05004987120628357, 0.21102137863636017, 0.035392459481954575, 0.38269418478012085, -0.29260048270225525, 0.20699812471866608, 0.05926388129591942, 0.04391023516654968, -0.11114337295293808, -0.21921710669994354, 0.05030057579278946, -0.157841295003891, 0.03383495658636093, 0.29528310894966125, 0.09981518238782883, 0.1530808061361313, -0.285504013299942, -0.12453760206699371, -0.20802302658557892, 0.22792354226112366, -0.17904223501682281, -0.04890627786517143, -0.24299491941928864, 0.021017028018832207, -0.025876162573695183, -0.09469947963953018, -0.14167042076587677, -0.03817928954958916, 0.3095564842224121, 0.2532219886779785, -0.1722697913646698, -0.45553529262542725, 0.10826699435710907, 0.3432939052581787, 0.18004795908927917, -0.24588418006896973, 0.33685368299484253, 0.19381506741046906, -0.06877820193767548, -0.01412354689091444, 0.39711013436317444, 0.4357686936855316, 0.33839377760887146, -0.3642207980155945, 0.08116745948791504, -0.2033807784318924, -0.0568217858672142, -0.02326009050011635, 0.26951277256011963, 0.10607278347015381, 0.11737696081399918, 0.4624232351779938, 0.12148629873991013, -0.12472685426473618, 0.04363814741373062, -0.08805640786886215, -0.2605327367782593, -0.26294752955436707, 0.5191484093666077, 0.185794860124588, -0.11438807845115662, -0.3206513226032257, 0.1899377405643463, -0.2749866843223572, -0.0809149518609047, 0.06285153329372406, 0.24058951437473297, 0.113153375685215, -0.21382735669612885, 0.07050874084234238, 0.00031574213062413037, 0.7061741352081299, 0.48772212862968445, -0.01771952025592327, -0.3658044636249542, -0.11652323603630066, -0.5989870429039001, 0.2178366482257843, -0.3639686405658722, -0.26901865005493164, 0.2833371162414551, -0.07233183085918427, -0.28808191418647766, 0.24852752685546875, 0.12761986255645752, 0.15819624066352844, -0.05727671831846237, 0.0983472391963005, -0.49076488614082336, -0.1486053764820099, -0.03065069019794464, -0.0076751429587602615, -0.0831807479262352, -0.25705254077911377, 0.1453229933977127, -0.4349938929080963, 0.1161026805639267, 0.011985591612756252, 0.06934838742017746, -0.0002399603690719232, 0.09664099663496017, 0.3544251024723053, 0.16532465815544128, 0.5012484788894653, -0.035560812801122665, 0.0036571817472577095, -0.19115544855594635, -0.15662235021591187, -0.22899216413497925, 0.13182324171066284, 0.08762864023447037, 0.5133113861083984, -0.026107320562005043, -0.06854572147130966, -0.2640267312526703, 0.40316781401634216, -0.08100741356611252, 0.08831649273633957, -0.13508076965808868, 0.07731454819440842, -0.07035040855407715, 0.22758440673351288, 0.26436471939086914, 0.4245404899120331, -0.17832162976264954, 0.20220930874347687, -0.27229106426239014, -0.4602454900741577, 0.32893526554107666, -0.6614815592765808, -0.5177910923957825, 0.07512339949607849, 0.11619257926940918, -0.1423347145318985, 0.05819396302103996, -0.6006348133087158, 0.0705668106675148, 0.24261273443698883, -0.11140530556440353, -0.3171199560165405, 0.21794475615024567, 0.016542308032512665, 0.08000080287456512, 0.043573711067438126, 0.44835665822029114, -0.14847348630428314, -0.15976785123348236, -0.004933878779411316, -0.1632518470287323 ]
https://github.com/huggingface/datasets/issues/166
Add a method to shuffle a dataset
+1 for the naming convention About the `shuffle` method, from my understanding it should be done in `Dataloader` (better separation between dataset processing - usage)
Could maybe be a `dataset.shuffle(generator=None, seed=None)` signature method. Also, we could maybe have a clear indication of which method modify in-place and which methods return/cache a modified dataset. I kinda like torch conversion of having an underscore suffix for all the methods which modify a dataset in-place. What do you think?
25
Add a method to shuffle a dataset Could maybe be a `dataset.shuffle(generator=None, seed=None)` signature method. Also, we could maybe have a clear indication of which method modify in-place and which methods return/cache a modified dataset. I kinda like torch conversion of having an underscore suffix for all the methods which modify a dataset in-place. What do you think? +1 for the naming convention About the `shuffle` method, from my understanding it should be done in `Dataloader` (better separation between dataset processing - usage)
[ 0.15341980755329132, -0.037309981882572174, -0.06269705295562744, -0.12907074391841888, 0.17716194689273834, 0.08376326411962509, 0.25355952978134155, 0.23048630356788635, -0.21282389760017395, 0.3302725851535797, -0.07075712084770203, 0.7361389994621277, -0.22251449525356293, -0.20181241631507874, 0.023780381307005882, -0.08247515559196472, 0.2442183494567871, -0.2117859125137329, -0.16500301659107208, 0.027073252946138382, -0.11413025110960007, -0.38973167538642883, 0.023244095966219902, -0.1421937644481659, -0.19263304769992828, -0.04983321949839592, -0.13309867680072784, 0.04740719124674797, -0.24616187810897827, -0.2544476389884949, -0.46736687421798706, 0.7014350295066833, -0.1527414321899414, 0.4499368667602539, -0.00011275450378889218, -0.29292982816696167, 0.35525086522102356, 0.032352566719055176, -0.31195342540740967, 0.07475164532661438, -0.17222630977630615, -0.07203900814056396, 0.06607125699520111, -0.20177508890628815, 0.21404647827148438, 0.057774413377046585, 0.056992389261722565, -0.129842609167099, 0.16466796398162842, -0.08943448215723038, 0.17472471296787262, -0.09979461878538132, -0.3013956844806671, -0.01625067926943302, 0.5358918309211731, 0.5136160254478455, -0.15510833263397217, 0.11450301855802536, 0.5382509827613831, 0.3613964915275574, -0.18806764483451843, 0.15618135035037994, -0.09367480129003525, -0.0767192468047142, 0.4886153042316437, -0.2939813435077667, -0.5896613001823425, 0.04415925592184067, 0.12085699290037155, 0.17901340126991272, 0.6000992655754089, -0.20126236975193024, -0.4169185757637024, -0.24042931199073792, 0.2678801119327545, -0.13337333500385284, 0.05386345833539963, -0.19137103855609894, -0.012278229929506779, -0.10944899916648865, -0.13056576251983643, -0.15746819972991943, -0.005667958874255419, -0.058629631996154785, 0.4977303147315979, 0.5297003388404846, 0.1231476292014122, -0.01947876624763012, 0.1149403378367424, 0.0041558812372386456, 0.5591574907302856, -0.10236559063196182, 0.014496291056275368, 0.26156651973724365, -0.11023624241352081, -0.10021746903657913, -0.007673264946788549, 0.25632375478744507, 0.24947910010814667, 0.11320142447948456, 0.30541399121284485, 0.27013713121414185, -0.06867620348930359, 0.07461683452129364, -0.06943379342556, -0.07609081268310547, 0.03268691152334213, -0.04483657702803612, 0.28886014223098755, -0.31532323360443115, 0.17996270954608917, -0.11023584008216858, -0.151326984167099, -0.05018084868788719, 0.03848577290773392, 0.12834715843200684, -0.30146992206573486, -0.10199952125549316, 0.1980665624141693, -0.5291101336479187, 0.06318297237157822, -0.427242636680603, 0.07826317846775055, 0.11914778500795364, 0.29036256670951843, 0.11539556831121445, -0.15124498307704926, 0.04944971576333046, 0.29246819019317627, -0.12642426788806915, -0.06676341593265533, 0.03252582997083664, -0.4743705093860626, 0.23198041319847107, 0.1612491011619568, -0.21715739369392395, 0.1098427027463913, 0.08951130509376526, 0.15006819367408752, 0.22949780523777008, 0.16365423798561096, 0.21711936593055725, 0.32886335253715515, -0.17753833532333374, -0.448760062456131, -0.06934994459152222, -0.09622693806886673, 0.2949220538139343, -0.29651060700416565, 0.26207271218299866, -0.2736850678920746, -0.2621243894100189, -0.06397705525159836, 0.12658855319023132, -0.025156455114483833, -0.10126559436321259, -0.23621191084384918, 0.20045378804206848, 0.11881707608699799, -0.3033382296562195, 0.38085198402404785, 0.13568197190761566, 0.18717354536056519, -0.10355456918478012, -0.03098028525710106, 0.25986555218696594, -0.09520675241947174, -0.02041863091289997, -0.33892300724983215, -0.10294383764266968, 0.18109670281410217, 0.11230591684579849, -0.16153885424137115, 0.023971274495124817, -0.12279962003231049, -0.13101059198379517, 0.48198840022087097, 0.21702782809734344, -0.17160676419734955, -0.0348060205578804, -0.15349377691745758, -0.1682637482881546, 0.5057479739189148, 0.3218179941177368, -0.07960895448923111, -0.25737977027893066, 0.28341588377952576, -0.09260088205337524, -0.413171648979187, 0.037118930369615555, -0.2332092672586441, -0.13467749953269958, 0.3806229829788208, 0.23806552588939667, -0.09189771860837936, 0.001127741765230894, 0.04510430991649628, -0.4092567265033722, 0.1858004778623581, -0.13240307569503784, -0.12164749950170517, -0.2527748644351959, -0.029327258467674255, 0.3391714096069336, -0.08386564254760742, -0.0944199189543724, -0.03489343449473381, -0.15264250338077545, 0.28662416338920593, -0.10843127965927124, 0.02840794064104557, -0.29122623801231384, 0.03806915506720543, -0.39275699853897095, -0.004562702961266041, -0.1235525831580162, 0.03388357162475586, 0.24871081113815308, -0.2404136210680008, -0.23405387997627258, -0.28614458441734314, -0.19298118352890015, -0.16904833912849426, -0.12849821150302887, -0.22818215191364288, -0.05662832409143448, -0.04576092213392258, -0.04605216905474663, -0.2763004004955292, 0.20862270891666412, -0.03019898012280464, -0.08184048533439636, -0.17129084467887878, 0.26135414838790894, 0.08896934986114502, -0.13293997943401337, 0.4365851879119873, 0.34575068950653076, -0.06262771040201187, 0.09248374402523041, 0.1802338808774948, -0.05637483298778534, -0.01670404151082039, -0.09874457865953445, -0.29944732785224915, 0.2603793442249298, -0.26657652854919434, 0.0909133106470108, -0.2621859908103943, -0.3188469707965851, -0.047044787555933, 0.37200841307640076, -0.2649489939212799, 0.169795423746109, 0.11686860769987106, 0.20927342772483826, 0.3513680398464203, 0.16266904771327972, -0.32343101501464844, -0.13951683044433594, 0.26496994495391846, -0.15998435020446777, -0.0220340583473444, 0.3419930934906006, 0.2032807618379593, -0.24940215051174164, 0.14528746902942657, 0.06851331144571304, 0.3731105327606201, 0.3408050239086151, 0.012890034355223179, 0.07192329317331314, 0.27976229786872864, -0.15011709928512573, -0.048269931226968765, 0.16651837527751923, -0.28262200951576233, -0.011177447624504566, 0.17486043274402618, 0.12027407437562943, -0.15232893824577332, -0.38304734230041504, 0.0760490819811821, 0.03220731392502785, -0.036178816109895706, -0.07769910246133804, -0.0185251347720623, -0.07378356903791428, -0.200893834233284, -0.19713079929351807, -0.3390849828720093, -0.26242831349372864, 0.2674481272697449, 0.058490827679634094, -0.2952110171318054, 0.3880605697631836, 0.18326784670352936, 0.5487819314002991, -0.21524830162525177, -0.10795997828245163, -0.3561215400695801, -0.4411758780479431, 0.12635716795921326, 0.04095596447587013, 0.2258598655462265, -0.13597309589385986, 0.7025212049484253, -0.05100417509675026, -0.07998792827129364, -0.3822793960571289, -0.46982985734939575, -0.08981538563966751, -0.007440835237503052, -0.2571098208427429, 0.3828088045120239, -0.32706019282341003, 0.08849554508924484, -0.47281867265701294, -0.06897009909152985, -0.2552245259284973, -0.01026111375540495, 0.00735268322750926, -0.010248886421322823, 0.009277488104999065, -0.18420124053955078, -0.29757654666900635, -0.13124607503414154, -0.40497979521751404, 0.07406418025493622, -0.6318069100379944, -0.050030019134283066, 0.01203717291355133, -0.286834180355072, -0.13241486251354218, -0.22056840360164642, 0.16136981546878815, -0.13669824600219727, -0.7596734762191772, 0.19291138648986816, -0.20223389565944672, 0.057574741542339325, -0.14220532774925232, -0.06971224397420883, 0.028892502188682556, 0.6446996331214905, 0.2754819989204407, -0.1669301837682724, 0.14853544533252716, 0.2016545683145523, -0.0016153764445334673, 0.15614508092403412, 0.34704461693763733, 0.2847755551338196, -0.12292229384183884, 0.062341753393411636, -0.2354121208190918, -0.11293711513280869, -0.10726990550756454, 0.31545358896255493, 0.10719183832406998, 0.3211008906364441, 0.016647305339574814, 0.6867966651916504, -0.6602863073348999, -0.09058065712451935, 0.019348733127117157, 0.2002178281545639, 0.13724686205387115, -0.06796826422214508, -0.09096956253051758, 0.5424608588218689, 0.24561846256256104, -0.3071582019329071, 0.1921602338552475, -0.04839325323700905, -0.16012294590473175, 0.03776339441537857, 0.13110794126987457, -0.14749839901924133, -0.1425306349992752, 0.3427988290786743, -0.3273088335990906, 0.08625153452157974, -0.04006374254822731, 0.17369869351387024, -0.295199453830719, 0.00426009064540267, 0.14684410393238068, 0.20264197885990143, 0.32880911231040955, -0.27043139934539795, -0.43609878420829773, -0.02117975614964962, -0.12659446895122528, 0.2958836257457733, 0.24746254086494446, 0.12883813679218292, -0.013604347594082355, -0.1818046122789383, 0.122365303337574, 0.25762298703193665, 0.4825326204299927, -0.24698956310749054, -0.28336101770401, -0.005569244269281626, -0.016896691173315048, -0.11249516159296036, 0.015213599428534508, -0.07555719465017319, 0.14184704422950745, 0.07289339601993561, 0.0938347727060318, -0.22726136445999146, -0.17682978510856628, 0.1136343851685524, 0.19612251222133636, 0.16109341382980347, -0.33927229046821594, -0.0829639583826065, -0.16535867750644684, -0.15990160405635834, 0.19259148836135864, -0.023795219138264656, -0.3008369505405426, -0.05560804530978203, 0.058830611407756805, -0.03348199650645256, 0.3097735643386841, 0.02668420597910881, 0.12922686338424683, 0.03208744525909424, 0.12182671576738358, 0.044979773461818695, 0.01223845686763525, 0.37701883912086487, 0.4276655912399292, -0.20563429594039917, -0.48697629570961, -0.1709441989660263, 0.014515101909637451, -0.5441600680351257, 0.24406953155994415, 0.23828573524951935, 0.251432329416275, 0.29675278067588806, -0.1404077261686325, 0.17606545984745026, -0.21850675344467163, 0.07353833317756653, 0.05243738368153572, 0.00740347895771265, -0.5293939709663391, -0.9117101430892944, 0.5416221022605896, 0.37718671560287476, -0.32241395115852356, 0.5554153919219971, 0.0005211685784161091, -0.2684890329837799, 0.4150931239128113, 0.29519858956336975, 1.205547571182251, -0.08146879076957703, 0.18745920062065125, -0.029216691851615906, -0.37673425674438477, 0.2766074538230896, -0.180907741189003, 0.14940035343170166, -0.12933886051177979, -0.39364683628082275, 0.05398144572973251, -0.08897192031145096, 0.07474023103713989, 0.3279806673526764, -0.12489799410104752, 0.34531211853027344, -0.3932631015777588, 0.1730789691209793, 0.1445835828781128, 0.12142428010702133, -0.24394623935222626, -0.2914193272590637, -0.08227930217981339, 0.07282450050115585, -0.09159094840288162, -0.21026641130447388, -0.033486444503068924, -0.005069646518677473, -0.0670088455080986, 0.14088669419288635, -0.1443738341331482, -0.30578505992889404, 0.15224625170230865, -0.08004232496023178, -0.2814943492412567, -0.1759495884180069, 0.44333285093307495, 0.10858900099992752, 0.2535838484764099, 0.07984964549541473, 0.16678926348686218, 0.2920086681842804, 0.0875646248459816, 0.40555620193481445, -0.09556896984577179, -0.08928154408931732, 0.10746969282627106, 0.08212101459503174, -0.04872202128171921, 0.030049020424485207, 0.11088273674249649, -0.4286024272441864, -0.529884397983551, -0.08485477417707443, 0.5617538094520569, -0.1269349604845047, 0.09261056035757065, 0.01875244453549385, -0.10598676651716232, -0.017499925568699837, 0.13704335689544678, 0.2787863612174988, -0.25865602493286133, 0.3623964190483093, -0.21821698546409607, 0.019990630447864532, -0.12006457149982452, 0.10490286350250244, -0.2883319556713104, -0.12774567306041718, 0.21792304515838623, -0.21949732303619385, -0.1973455548286438, -0.1657760590314865, 0.11748882383108139, 0.23314915597438812, -0.004313383251428604, 0.03345128893852234, -0.13740426301956177, -0.2781904935836792, -0.15487925708293915, -0.1567090004682541, 0.21951714158058167, 0.34901174902915955, -0.03535425290465355, 0.11037494987249374, -0.46019017696380615, 0.07673456519842148, 0.17053338885307312, 0.3398685157299042, -0.1675647348165512, -0.046649154275655746, -0.1151118054986, -0.01923518255352974, -0.3286210000514984, -0.14945915341377258, 0.2203100472688675, 0.2718961536884308, 0.009397551417350769, 0.22101406753063202, -0.20140770077705383, -0.24054917693138123, 0.06552603840827942, 0.06110499054193497, -0.018643315881490707, -0.14807042479515076, -0.13654357194900513, 0.11171066015958786, -0.07000500708818436, 0.22942933440208435, 0.20375727117061615, -0.0818171352148056, -0.0028205697890371084, -0.25039294362068176, 0.026482373476028442, 0.2947494089603424, 0.005087785888463259, 0.4968263804912567, 0.6184056997299194, 0.08185280114412308, -0.21739862859249115, 0.2066221982240677, -0.02550266683101654, 0.03144066780805588, 0.12063372135162354, 0.14772583544254303, -0.02656259946525097, -0.020382247865200043, 0.11907122284173965, 0.3216325640678406, 0.18479935824871063, 0.06783641874790192, -0.09052778780460358, 0.2885018587112427, 0.19308529794216156, 0.4536784887313843, 0.099522665143013, 0.4589994549751282, 0.001219184952788055, -0.3575811982154846, 0.4314950704574585, 0.20655803382396698, -0.0190287996083498, -0.07362867891788483, -0.2535403072834015, -0.005758689250797033, 0.08054933696985245, -0.005396639462560415, -0.2296503484249115, 0.015039621852338314, -0.3449499309062958, -0.07759948074817657, 0.14512372016906738, -0.4788013696670532, 0.39619380235671997, 0.3324783742427826, -0.12170477211475372, -0.027955887839198112, 0.3708789646625519, 0.009444385766983032, -0.26959073543548584, 0.1324603259563446, 0.4628450870513916, 0.9553405046463013, 0.326776385307312, 0.16063319146633148, 0.13742738962173462, 0.14922742545604706, 0.1605956107378006, 0.6927059888839722, 0.3411797285079956, 0.08903813362121582, 0.385088711977005, 0.22731800377368927, 0.14771723747253418, -0.3402741849422455, 0.27715712785720825, 0.07764463126659393, 0.2560737431049347, 0.030500371009111404, -0.0663548931479454, -0.19421172142028809, -0.25191745162010193, -0.08347728103399277, 0.01261717639863491, 0.051534753292798996, 0.5799619555473328, 0.2349967658519745, 0.2591428756713867, -0.35283297300338745, 0.12664280831813812, 0.27182677388191223, 0.44117575883865356, -0.15010924637317657, -0.4055158793926239, 0.142024427652359, -0.17151236534118652, 0.7451763153076172, -0.019770467653870583, -0.0012754487106576562, 0.24583126604557037, 0.029737232252955437, 0.11833087354898453, -0.3746960461139679, -0.03710116073489189, -0.015620614401996136, 0.12326820939779282, 0.14358218014240265, -0.17992940545082092, -0.012089768424630165, 0.17657983303070068, -0.14344380795955658, 0.07518966495990753, -0.07518370449542999, -0.046274349093437195, -0.18357498943805695, 0.3442081809043884, 0.49961674213409424, -0.30887570977211, 0.10623538494110107, -0.0849187821149826, -0.19567659497261047, -0.1527647078037262, 0.040559399873018265, 0.16102319955825806, -0.2053283303976059, 0.09134728461503983, 0.05906746909022331, 0.29307594895362854, 0.2837912142276764, 0.2766536772251129, -0.0072680008597671986, -0.14981047809123993, -0.3939853012561798, -0.38947126269340515, -0.13016000390052795, 0.4760107100009918, 0.003611741354689002, -0.0354774184525013, -0.18319879472255707, 0.03684157878160477, -0.06707989424467087, 0.14567480981349945, -0.12240808457136154, 0.05333174020051956, 0.1979421228170395, -0.3570575416088104, -0.12122110277414322, -0.1313953697681427, -0.022572530433535576, 0.18365547060966492, -0.06100409850478172, -0.306797593832016, 0.4100828766822815, 0.044131744652986526, -0.3097112476825714, 0.14482349157333374, 0.06988386064767838, 0.11397387087345123, 0.21174189448356628, -0.019559452310204506, 0.388077974319458, -0.002460102317854762, -0.18213070929050446, 0.02638501115143299, -0.14211727678775787, -0.5724024772644043, 0.12119142711162567, 0.0075441873632371426, -0.011270591989159584, -0.0383661650121212, -0.20877611637115479, -0.44213390350341797, 0.5485702157020569, -0.35532131791114807, 0.08550769835710526, -0.7277726531028748, 0.15805144608020782, 0.24726291000843048, 0.08188704401254654, -0.04858022928237915, 0.11439915746450424, 0.0510067418217659, 0.27076035737991333, -0.22954663634300232, 0.19431641697883606, 0.5219246745109558, 0.16950589418411255, -0.24273888766765594, -0.3214069604873657, 0.21311461925506592, -0.08961247652769089, 0.02953283116221428, -0.3195727467536926, -0.08377420902252197, 0.23225107789039612, 0.003436184022575617, 0.07984823733568192, 0.383951872587204, 0.12844973802566528, -0.04665743559598923, -0.22277213633060455, -0.0014488390879705548, -0.12010563164949417, -0.030344294384121895, -0.11485431343317032, -0.2884332537651062 ]
https://github.com/huggingface/datasets/issues/166
Add a method to shuffle a dataset
+1 for shuffle in `Dataloader`. Some `Dataloader` just store idxs of dataset and just shuffle those idxs, which might(?) be faster than do shuffle in dataset, especially when doing shuffle every epoch. Also +1 for the naming convention.
Could maybe be a `dataset.shuffle(generator=None, seed=None)` signature method. Also, we could maybe have a clear indication of which method modify in-place and which methods return/cache a modified dataset. I kinda like torch conversion of having an underscore suffix for all the methods which modify a dataset in-place. What do you think?
38
Add a method to shuffle a dataset Could maybe be a `dataset.shuffle(generator=None, seed=None)` signature method. Also, we could maybe have a clear indication of which method modify in-place and which methods return/cache a modified dataset. I kinda like torch conversion of having an underscore suffix for all the methods which modify a dataset in-place. What do you think? +1 for shuffle in `Dataloader`. Some `Dataloader` just store idxs of dataset and just shuffle those idxs, which might(?) be faster than do shuffle in dataset, especially when doing shuffle every epoch. Also +1 for the naming convention.
[ 0.07046911120414734, 0.016060497611761093, -0.08107688277959824, -0.13309356570243835, 0.1867508590221405, -0.018764972686767578, 0.36125412583351135, 0.24943767488002777, -0.13167954981327057, 0.39812397956848145, -0.11517035216093063, 0.7459425926208496, -0.32974597811698914, -0.2149703949689865, -0.04452412948012352, -0.07981546223163605, 0.2653796076774597, -0.10314817726612091, -0.20524269342422485, 0.05622423440217972, -0.12470738589763641, -0.3481490910053253, 0.02367376908659935, -0.14052952826023102, -0.08968815952539444, 0.01218604575842619, -0.11303704977035522, 0.07146795094013214, -0.273320734500885, -0.21861381828784943, -0.445068895816803, 0.7157345414161682, -0.1049824059009552, 0.5668812394142151, -0.00010911504068644717, -0.25887197256088257, 0.34147241711616516, 0.032926879823207855, -0.3560408055782318, 0.14131318032741547, -0.13315828144550323, -0.0783197209239006, 0.03868082910776138, -0.2729431986808777, 0.22207041084766388, -0.034398388117551804, 0.11589914560317993, -0.11278633028268814, 0.11098942905664444, -0.14737366139888763, 0.18193326890468597, -0.09936544299125671, -0.3846821188926697, 0.0020909288432449102, 0.5785194635391235, 0.4610337018966675, -0.20912036299705505, 0.18245267868041992, 0.5745747685432434, 0.3853919804096222, -0.33382847905158997, 0.19154955446720123, -0.13094298541545868, -0.10194747895002365, 0.5779693722724915, -0.30015528202056885, -0.539605438709259, 0.06476461887359619, -0.03939216956496239, 0.20368947088718414, 0.6476196050643921, -0.1902359127998352, -0.4480341672897339, -0.15105906128883362, 0.18167026340961456, -0.04967651888728142, 0.030789978802204132, -0.2421192079782486, -0.019576311111450195, -0.0853261724114418, -0.037378471344709396, -0.11297043412923813, -0.00256237736903131, -0.13636109232902527, 0.4217974543571472, 0.5597809553146362, 0.15498225390911102, -0.02861437201499939, 0.23609445989131927, -0.02223958633840084, 0.5372044444084167, -0.06967498362064362, 0.07333335280418396, 0.2372979074716568, -0.21732088923454285, -0.18081369996070862, -0.01553327776491642, 0.21171049773693085, 0.18797318637371063, 0.10574372857809067, 0.3026648759841919, 0.2721395790576935, 0.013139449991285801, 0.1095648929476738, -0.11448836326599121, -0.03855234757065773, 0.033464789390563965, -0.10943514108657837, 0.3762948215007782, -0.22859545052051544, 0.1526610553264618, -0.0816257894039154, -0.2623734772205353, -0.017839159816503525, 0.10215211659669876, 0.1414058655500412, -0.41330796480178833, 0.05687219277024269, 0.17915013432502747, -0.49136921763420105, 0.15620632469654083, -0.4261990189552307, 0.10317118465900421, 0.18962278962135315, 0.18701602518558502, 0.13204005360603333, -0.1550372987985611, -0.012790009379386902, 0.21155139803886414, -0.1847407966852188, -0.10223614424467087, 0.04918687418103218, -0.507145881652832, 0.21948522329330444, 0.10138154774904251, -0.20319077372550964, 0.11597611010074615, 0.03873945400118828, 0.13651782274246216, 0.3439455032348633, 0.1533028483390808, 0.20069316029548645, 0.1924789696931839, -0.25562238693237305, -0.43917229771614075, -0.1295638084411621, -0.14438511431217194, 0.3152361512184143, -0.2803954780101776, 0.24230872094631195, -0.29306846857070923, -0.23662246763706207, -0.11380854994058609, 0.17173616588115692, -0.0018634925363585353, -0.061999328434467316, -0.22932656109333038, 0.24639564752578735, 0.07400444895029068, -0.30563902854919434, 0.4155261516571045, 0.07851913571357727, 0.17123475670814514, -0.1233358383178711, -0.07254987210035324, 0.2618195116519928, -0.11127045750617981, -0.018436439335346222, -0.2302527278661728, -0.16228745877742767, 0.2056487798690796, 0.1142255887389183, -0.1920030117034912, 0.007894791662693024, -0.141953244805336, -0.14744576811790466, 0.4907088577747345, 0.1812632828950882, -0.15881480276584625, -0.012354603968560696, -0.11443611979484558, -0.10257530212402344, 0.6056383848190308, 0.3467119634151459, 0.056494370102882385, -0.3341640532016754, 0.28792890906333923, -0.06120159849524498, -0.3102134168148041, 0.09815284609794617, -0.22475828230381012, -0.11841282248497009, 0.38146352767944336, 0.26114967465400696, -0.12281882017850876, -0.009176995605230331, 0.005660836584866047, -0.3373621702194214, 0.19471324980258942, -0.1716936081647873, -0.07312598079442978, -0.2363463193178177, -0.08199675381183624, 0.30896949768066406, -0.015054577961564064, -0.06498894095420837, -0.0631176233291626, -0.11575783789157867, 0.2917594611644745, 0.01764991320669651, 0.1858094483613968, -0.28128284215927124, 0.02902127616107464, -0.298960417509079, -0.015439005568623543, -0.17469626665115356, 0.06162843480706215, 0.2616628110408783, -0.2221287339925766, -0.24753659963607788, -0.2646149694919586, -0.17950385808944702, -0.27028486132621765, -0.16231724619865417, -0.25778427720069885, -0.10035726428031921, -0.04210984706878662, 0.027301941066980362, -0.21384727954864502, 0.22047048807144165, -0.130250483751297, -0.07318146526813507, -0.15486404299736023, 0.24804937839508057, 0.11062821000814438, 0.0012288902653381228, 0.30390554666519165, 0.3766377866268158, -0.023014141246676445, 0.0787244364619255, 0.3076276183128357, -0.08393970876932144, 0.012866821140050888, -0.11360371112823486, -0.2761874496936798, 0.25423145294189453, -0.23281945288181305, 0.08788284659385681, -0.2713022530078888, -0.3496970236301422, -0.008261473849415779, 0.359683096408844, -0.33686259388923645, 0.2372020184993744, 0.12871108949184418, 0.2788377106189728, 0.4074001610279083, 0.2508472204208374, -0.3290271461009979, -0.1328149288892746, 0.3740353286266327, -0.12329505383968353, -0.07980760931968689, 0.3697067201137543, 0.12849660217761993, -0.2949620187282562, 0.1104046106338501, 0.08891547471284866, 0.3699173033237457, 0.34708455204963684, 0.0725460797548294, -0.03381386026740074, 0.33118048310279846, -0.13318000733852386, -0.010054422542452812, 0.09692011773586273, -0.13696539402008057, -0.020879527553915977, 0.10166261345148087, 0.058117788285017014, -0.19984035193920135, -0.5052341818809509, 0.03356492146849632, 0.09178820252418518, 0.03976175934076309, -0.07727978378534317, 0.05568936839699745, -0.05265926569700241, -0.14146274328231812, -0.1758357286453247, -0.28012600541114807, -0.2714926302433014, 0.25312918424606323, 0.10748493671417236, -0.33829304575920105, 0.33368685841560364, 0.18314313888549805, 0.531672477722168, -0.1584073305130005, -0.058418627828359604, -0.3735503554344177, -0.4471939504146576, 0.15386396646499634, 0.07673260569572449, 0.22851309180259705, -0.18271352350711823, 0.6215425133705139, -0.03552662953734398, -0.063880056142807, -0.3513755798339844, -0.3830469250679016, -0.06246757507324219, -0.0042406488209962845, -0.2293923795223236, 0.3913438022136688, -0.2218208909034729, 0.11684223264455795, -0.4907086491584778, -0.005511086899787188, -0.31851720809936523, -0.10767810791730881, 0.02714819461107254, 0.025268299505114555, 0.013321924954652786, -0.16682860255241394, -0.24212731420993805, -0.1718025505542755, -0.4843711256980896, 0.03587116673588753, -0.5416010022163391, -0.0025587088894098997, 0.013053643517196178, -0.2118820995092392, -0.10356830805540085, -0.11099376529455185, 0.15742072463035583, -0.154494047164917, -0.8162312507629395, 0.22808250784873962, -0.1825128197669983, 0.07005339115858078, -0.22892823815345764, -0.11542683094739914, 0.08359403163194656, 0.6077491641044617, 0.2217944711446762, -0.06396849453449249, 0.13453032076358795, 0.13789765536785126, 0.039365001022815704, 0.16842596232891083, 0.2622101604938507, 0.30263346433639526, -0.18170034885406494, 0.02077954076230526, -0.22535987198352814, -0.08990482985973358, -0.12235324084758759, 0.2259427011013031, 0.17040754854679108, 0.36077171564102173, -0.015084953047335148, 0.739267110824585, -0.6775938272476196, -0.053754765540361404, -0.02820468693971634, 0.1690758764743805, 0.1355459839105606, -0.031218798831105232, -0.14977462589740753, 0.5597658157348633, 0.21935677528381348, -0.30563148856163025, 0.2308705598115921, -0.08712230622768402, -0.24886932969093323, -0.011342350393533707, 0.0699889287352562, -0.041593998670578, -0.0578061081469059, 0.3630124628543854, -0.30870166420936584, 0.013248709961771965, -0.06105046346783638, 0.12251829355955124, -0.326917827129364, 0.009556652046740055, 0.15242880582809448, 0.22981391847133636, 0.26550573110580444, -0.24354426562786102, -0.4098934829235077, 0.04423016309738159, -0.21469588577747345, 0.1624688357114792, 0.2266131341457367, 0.07951004803180695, 0.09746851772069931, -0.27765214443206787, 0.0666256695985794, 0.20896762609481812, 0.5011020302772522, -0.12202287465333939, -0.2938186228275299, -0.018135465681552887, -0.035046596080064774, -0.08999684453010559, -0.02498531900346279, -0.11665723472833633, 0.12557902932167053, 0.020660780370235443, 0.097683846950531, -0.26590216159820557, -0.16656813025474548, 0.0744640976190567, 0.13256339728832245, 0.07664009183645248, -0.33591681718826294, -0.14411619305610657, -0.0007811808027327061, -0.15488092601299286, 0.17788535356521606, -0.09627881646156311, -0.2856132984161377, -0.061437275260686874, -0.021943526342511177, 0.03640851005911827, 0.2939797043800354, 0.07353529334068298, 0.21434831619262695, 0.05764323100447655, 0.07084579020738602, 0.048228438943624496, -0.06505824625492096, 0.2817291021347046, 0.43656736612319946, -0.14282384514808655, -0.4651133418083191, -0.10632561892271042, 0.13580022752285004, -0.4808529317378998, 0.2497146874666214, 0.11656679213047028, 0.21496322751045227, 0.1901760697364807, -0.14731983840465546, 0.21787600219249725, -0.26834022998809814, -0.010570462793111801, 0.06111469119787216, 0.0282115638256073, -0.4887731075286865, -0.9323244690895081, 0.5213589668273926, 0.36476418375968933, -0.2582475244998932, 0.5743241310119629, 0.004262933041900396, -0.20411065220832825, 0.4123575985431671, 0.3018024265766144, 1.151923656463623, -0.09614524990320206, 0.2185094952583313, 0.10118727385997772, -0.2109861522912979, 0.2955073416233063, -0.22981436550617218, 0.15573978424072266, -0.15916015207767487, -0.3745519518852234, 0.06611080467700958, -0.13276307284832, 0.049897048622369766, 0.26341700553894043, -0.2560465931892395, 0.2536979615688324, -0.30841678380966187, 0.23445896804332733, 0.020436536520719528, 0.19144269824028015, -0.28746286034584045, -0.32861530780792236, -0.050824135541915894, 0.09983649849891663, -0.05817970260977745, -0.300087571144104, -0.04737616330385208, -0.0546853132545948, -0.09370715171098709, 0.20287580788135529, -0.10154663026332855, -0.25969353318214417, 0.13058173656463623, -0.05969173461198807, -0.26577726006507874, -0.09930046647787094, 0.366740882396698, 0.03436373174190521, 0.12549418210983276, -0.014878549613058567, 0.15132851898670197, 0.24826094508171082, 0.03901851549744606, 0.4666064381599426, -0.04503755643963814, -0.13180342316627502, 0.2440778762102127, 0.06336638331413269, 0.005416215397417545, 0.04518687725067139, 0.11685121804475784, -0.4874970316886902, -0.42852020263671875, -0.04317472130060196, 0.5997399687767029, -0.1540207862854004, 0.11646611988544464, 0.049157384783029556, -0.19502560794353485, -0.04816142097115517, 0.15840928256511688, 0.2610594928264618, -0.23258595168590546, 0.2867862582206726, -0.12047930806875229, 0.04268612340092659, -0.12510357797145844, 0.08967906981706619, -0.16128090023994446, -0.11361168324947357, 0.19274280965328217, -0.23611854016780853, -0.280558705329895, -0.19183996319770813, 0.15163587033748627, 0.21312901377677917, -0.029194459319114685, 0.1103459894657135, -0.23565159738063812, -0.3742777705192566, -0.0794694721698761, -0.19264288246631622, 0.2327069342136383, 0.3861812651157379, -0.07074157148599625, 0.1231982409954071, -0.493431955575943, 0.08805893361568451, 0.09681469202041626, 0.35483744740486145, -0.2407088279724121, -0.09278304874897003, -0.10960792005062103, -0.01845543645322323, -0.3788086771965027, -0.25162872672080994, 0.15548302233219147, 0.19931134581565857, 0.008189604617655277, 0.11616213619709015, -0.22874008119106293, -0.10670087486505508, 0.09685711562633514, 0.08086853474378586, -0.0796147882938385, -0.1717635840177536, -0.17849396169185638, 0.09644435346126556, -0.09953492879867554, 0.21651533246040344, 0.21649381518363953, -0.06108001992106438, -0.09319717437028885, -0.24933065474033356, 0.0512288436293602, 0.26158323884010315, -0.059605903923511505, 0.4868052005767822, 0.6710013151168823, 0.09724459797143936, -0.2060214877128601, 0.1600845605134964, 0.07366198301315308, 0.0598275288939476, 0.0875624343752861, 0.12772886455059052, -0.036418233066797256, -0.09296619147062302, 0.10473310947418213, 0.43972092866897583, 0.15501375496387482, 0.13724665343761444, -0.002952976617962122, 0.2808755934238434, 0.19764983654022217, 0.3586435914039612, 0.18620434403419495, 0.4341937303543091, 0.041120536625385284, -0.3968985974788666, 0.4125511944293976, 0.2342456579208374, -0.03565257787704468, -0.13068020343780518, -0.20733655989170074, -0.11624646931886673, 0.13849018514156342, 0.01038201991468668, -0.14836159348487854, 0.006765217985957861, -0.21708597242832184, -0.13100185990333557, 0.0985453873872757, -0.4322226643562317, 0.3699081540107727, 0.24644213914871216, -0.10153907537460327, -0.07392758131027222, 0.44252723455429077, 0.01766732707619667, -0.18382646143436432, 0.16107520461082458, 0.3885464668273926, 0.8879129886627197, 0.3873404860496521, 0.14933794736862183, 0.09435481578111649, 0.17374666035175323, 0.14424842596054077, 0.7236197590827942, 0.23297740519046783, 0.11240517348051071, 0.3068797290325165, 0.2533046007156372, 0.10200030356645584, -0.22233836352825165, 0.27811336517333984, 0.14876244962215424, 0.1322818249464035, 0.01139193493872881, -0.03571079671382904, -0.22587217390537262, -0.3367291986942291, -0.09032930433750153, -0.013228057883679867, 0.08498972654342651, 0.5246777534484863, 0.1999397873878479, 0.2989272177219391, -0.2885635495185852, 0.1400090605020523, 0.20761993527412415, 0.434596985578537, -0.21771687269210815, -0.42913103103637695, 0.1927117109298706, -0.11657316982746124, 0.6738451719284058, 0.05358901247382164, -0.03115135058760643, 0.2896180748939514, 0.05447361245751381, 0.074894979596138, -0.3837047219276428, -0.03711368143558502, -0.06795184314250946, 0.10829351842403412, 0.10009725391864777, -0.23522496223449707, 0.041720930486917496, 0.17708905041217804, -0.19993330538272858, 0.01671491377055645, 0.03517809882760048, -0.14711080491542816, -0.12542842328548431, 0.23321986198425293, 0.5211764574050903, -0.23741331696510315, 0.10508289188146591, -0.020696256309747696, -0.3141877055168152, -0.1590210646390915, -0.0296090766787529, 0.0701117143034935, -0.16472966969013214, 0.047691710293293, 0.08573605120182037, 0.24409615993499756, 0.3486327528953552, 0.27942270040512085, 0.009575264528393745, -0.11959245800971985, -0.41816869378089905, -0.40806201100349426, -0.1301291435956955, 0.44738802313804626, 0.02127646468579769, -0.023290976881980896, -0.03652353212237358, 0.03883378207683563, -0.039017993956804276, 0.16553053259849548, -0.13707773387432098, 0.01628235913813114, 0.11755911260843277, -0.36923354864120483, -0.10993676632642746, -0.09451630711555481, -0.0524839386343956, 0.18723364174365997, -0.12909874320030212, -0.2749149203300476, 0.31997251510620117, 0.10629604756832123, -0.24907800555229187, 0.09010932594537735, 0.15677808225154877, 0.18383319675922394, 0.2313622534275055, -0.10574964433908463, 0.22225689888000488, -0.013971598818898201, -0.18763765692710876, 0.06642043590545654, -0.08607617020606995, -0.4891238510608673, 0.009831330738961697, 0.0942465141415596, 0.08463631570339203, -0.02032407931983471, -0.22575277090072632, -0.4579549729824066, 0.5146232843399048, -0.3050031065940857, 0.08379428833723068, -0.7182447910308838, 0.2188802808523178, 0.23858249187469482, 0.05179176479578018, -0.15571339428424835, 0.12745331227779388, 0.1018626019358635, 0.22148765623569489, -0.3064771294593811, 0.16515332460403442, 0.44851773977279663, 0.10766973346471786, -0.23511825501918793, -0.28615665435791016, 0.1609083116054535, -0.043487146496772766, 0.17592766880989075, -0.30663496255874634, 0.002416707342490554, 0.21606115996837616, 0.011468611657619476, 0.10717825591564178, 0.4145711660385132, 0.05682259052991867, 0.005425390787422657, -0.20953324437141418, 0.02015521004796028, -0.14458610117435455, -0.030570605769753456, -0.11404632031917572, -0.28600773215293884 ]
https://github.com/huggingface/datasets/issues/166
Add a method to shuffle a dataset
As you might already know the issue of dataset shuffling came up in the nlp code [walkthrough](https://youtu.be/G3pOvrKkFuk?t=3204) by Yannic Kilcher
Could maybe be a `dataset.shuffle(generator=None, seed=None)` signature method. Also, we could maybe have a clear indication of which method modify in-place and which methods return/cache a modified dataset. I kinda like torch conversion of having an underscore suffix for all the methods which modify a dataset in-place. What do you think?
20
Add a method to shuffle a dataset Could maybe be a `dataset.shuffle(generator=None, seed=None)` signature method. Also, we could maybe have a clear indication of which method modify in-place and which methods return/cache a modified dataset. I kinda like torch conversion of having an underscore suffix for all the methods which modify a dataset in-place. What do you think? As you might already know the issue of dataset shuffling came up in the nlp code [walkthrough](https://youtu.be/G3pOvrKkFuk?t=3204) by Yannic Kilcher
[ 0.11682182550430298, 0.047994934022426605, -0.008072469383478165, -0.2174368053674698, 0.12888213992118835, 0.02868986316025257, 0.13635922968387604, 0.36064305901527405, -0.2647390365600586, 0.2979814112186432, 0.011636115610599518, 0.8511587381362915, -0.35360419750213623, -0.23209847509860992, 0.2013930380344391, -0.07049967348575592, 0.15841278433799744, -0.10356619209051132, -0.24149124324321747, 0.05236378312110901, -0.13913819193840027, -0.3828112781047821, -0.07822736352682114, -0.20003661513328552, -0.13133969902992249, -0.09400425106287003, -0.08385585993528366, 0.13230076432228088, -0.1840188354253769, -0.29359665513038635, -0.4079645872116089, 0.6355045437812805, -0.10218527168035507, 0.14192359149456024, -0.00010813648987095803, -0.27224454283714294, 0.32626375555992126, 0.1310778707265854, -0.4321398437023163, 0.04468626528978348, -0.08185317367315292, -0.09962701797485352, 0.00863680150359869, -0.1526377946138382, 0.06552902609109879, 0.13389582931995392, 0.16543442010879517, -0.14641205966472626, 0.24235843122005463, 0.050310149788856506, 0.20914483070373535, -0.03290902450680733, -0.19010041654109955, 0.15847380459308624, 0.5000157952308655, 0.4068683385848999, -0.10122709721326828, 0.028628354892134666, 0.4395243227481842, 0.22933202981948853, -0.26579925417900085, 0.2744676470756531, -0.06643842905759811, -0.18149198591709137, 0.3078679144382477, -0.28791841864585876, -0.5689671039581299, 0.07428515702486038, 0.0025417557917535305, 0.17878617346286774, 0.5037117004394531, -0.285334974527359, -0.39036983251571655, -0.35262981057167053, 0.33025169372558594, -0.18849986791610718, 0.019990336149930954, -0.09641754627227783, -0.05681654065847397, -0.065743088722229, -0.24524752795696259, -0.15989483892917633, 0.03680567443370819, -0.02175346575677395, 0.5161421298980713, 0.6055989861488342, 0.20056715607643127, -0.0033857650123536587, 0.2449982762336731, 0.08339690417051315, 0.5854989886283875, -0.009341344237327576, 0.03303399682044983, 0.29964423179626465, -0.14720872044563293, -0.20837311446666718, 0.055004023015499115, 0.29483696818351746, 0.23643532395362854, 0.14461302757263184, 0.32707303762435913, 0.28509965538978577, -0.0413028784096241, 0.03735368326306343, 0.005551974754780531, 0.03812026232481003, -0.05310165882110596, 0.054955754429101944, 0.28106892108917236, -0.3015076518058777, 0.23124872148036957, 0.00397071847692132, -0.049036744982004166, -0.08059734851121902, -0.010976872406899929, 0.13244368135929108, -0.1945452243089676, -0.06849855929613113, 0.03612923249602318, -0.581583559513092, -0.04963230341672897, -0.3245076835155487, 0.14889627695083618, 0.10779424756765366, 0.27703598141670227, 0.14072154462337494, -0.10792281478643417, -0.04817373305559158, 0.1835097372531891, -0.2393871545791626, -0.10505368560552597, 0.025838855654001236, -0.49392688274383545, 0.17078785598278046, 0.17374014854431152, -0.07522283494472504, 0.13901478052139282, 0.11294428259134293, 0.06702842563390732, 0.23749524354934692, 0.07884696871042252, 0.2658466696739197, 0.40570104122161865, -0.2036510705947876, -0.438070684671402, -0.13343015313148499, -0.10840387642383575, 0.2805593013763428, -0.37445250153541565, 0.2528228461742401, -0.2730662226676941, -0.23034346103668213, -0.09223329275846481, 0.15652374923229218, -0.09696920216083527, -0.2698228359222412, -0.1759152114391327, 0.3530564308166504, 0.06264561414718628, -0.4234466850757599, 0.3546224534511566, 0.07788459211587906, 0.062280479818582535, -0.004047938622534275, 0.013162248767912388, 0.31766319274902344, 0.046954430639743805, 0.020499443635344505, -0.24038155376911163, 0.001836559153161943, 0.2378980815410614, 0.23913849890232086, -0.1998988687992096, -0.021302664652466774, -0.1616418957710266, -0.06126968190073967, 0.5229774713516235, 0.1896459311246872, -0.20807014405727386, 0.013186221942305565, -0.20950424671173096, -0.09727875143289566, 0.4091677963733673, 0.3883785009384155, -0.12241657078266144, -0.3362310230731964, 0.2075742483139038, 0.11739445477724075, -0.4104662239551544, 0.14842136204242706, -0.35707616806030273, -0.07999227941036224, 0.4687561094760895, 0.3144294023513794, -0.24346399307250977, 0.0008652854594402015, 0.07365517318248749, -0.15948039293289185, 0.294075608253479, -0.1337457299232483, -0.13703735172748566, -0.23566249012947083, -0.0692659467458725, 0.12855036556720734, -0.17178983986377716, -0.11686261743307114, -0.10402937233448029, -0.23739519715309143, 0.22362865507602692, -0.05178000032901764, 0.11644184589385986, -0.285942405462265, 0.08745825290679932, -0.33856984972953796, 0.07401040941476822, -0.1682894229888916, 0.09142850339412689, 0.11526419222354889, -0.10391492396593094, -0.16962872445583344, -0.22858038544654846, -0.06513673812150955, -0.1469428390264511, -0.18812045454978943, -0.3230235278606415, 0.02998736873269081, -0.11075791716575623, -0.05123124271631241, -0.07592608034610748, 0.46679195761680603, -0.08376175910234451, -0.0623999647796154, -0.09606405347585678, 0.2955297529697418, -0.0001284318568650633, -0.26771280169487, 0.37398579716682434, 0.2960720956325531, 0.03366652503609657, 0.03514605388045311, 0.09655198454856873, 0.03362180292606354, -0.05559789389371872, -0.0028625335544347763, -0.3500954806804657, 0.3163101375102997, -0.16839580237865448, 0.028979981318116188, -0.18219539523124695, -0.25976261496543884, 0.06748044490814209, 0.22245848178863525, -0.33422860503196716, 0.16626094281673431, 0.2199171632528305, 0.21182896196842194, 0.281310498714447, 0.22115714848041534, -0.4753613770008087, 0.06712479144334793, 0.4927184581756592, -0.054044995456933975, 0.01846528984606266, 0.34422391653060913, 0.2549564838409424, -0.3329373896121979, 0.12259705364704132, 0.12657085061073303, 0.2766241431236267, 0.4469538927078247, 0.0526612251996994, 0.14309236407279968, 0.09512909501791, -0.16515378654003143, -0.04433983936905861, -0.06922058761119843, -0.17121687531471252, -0.028135454282164574, 0.23661468923091888, 0.14595475792884827, -0.28768253326416016, -0.32734742760658264, -0.04693524166941643, 0.18919755518436432, -0.11021775007247925, -0.058832231909036636, -0.13308188319206238, -0.3062000572681427, -0.3187926113605499, -0.09166983515024185, -0.339528352022171, -0.3069959878921509, 0.3917933702468872, 0.010386759415268898, -0.320739209651947, 0.40423503518104553, 0.1464310586452484, 0.5348385572433472, -0.2747848629951477, -0.011294062249362469, -0.2897605895996094, -0.4303379952907562, -0.08097214251756668, 0.07382337749004364, 0.2886367440223694, -0.04778332635760307, 0.7124965786933899, 0.07893449068069458, -0.13275279104709625, -0.3044416904449463, -0.5318077206611633, -0.14086131751537323, 0.03139031305909157, -0.21374176442623138, 0.34212639927864075, -0.39242348074913025, 0.09011007845401764, -0.3882690370082855, -0.18767446279525757, -0.34289202094078064, -0.08024153113365173, 0.12296098470687866, 0.0008936544763855636, 0.02240022085607052, -0.29992949962615967, -0.34487825632095337, -0.07238531857728958, -0.4497670531272888, 0.0770968422293663, -0.6393796801567078, 0.07205867022275925, 0.05661339312791824, -0.23646968603134155, -0.16641613841056824, -0.37278982996940613, 0.007764334790408611, -0.127971351146698, -0.5969660878181458, 0.16613715887069702, -0.1103721633553505, -0.06128214672207832, -0.2556629478931427, -0.05604007840156555, 0.01381267886608839, 0.6228569746017456, 0.19484549760818481, -0.13971590995788574, 0.10687966644763947, 0.07648130506277084, -0.0811649039387703, 0.1502828747034073, 0.4343815743923187, 0.16205017268657684, -0.12345772236585617, 0.11876938492059708, -0.1826428323984146, 0.042115386575460434, -0.015398562885820866, 0.3286912143230438, 0.034825362265110016, 0.2597437798976898, 0.1325589418411255, 0.5310884118080139, -0.5287258625030518, -0.08033175766468048, 0.05685665085911751, 0.17602039873600006, 0.25830429792404175, -0.09530063718557358, 0.0022561966907233, 0.5387940406799316, 0.2906547784805298, -0.2538532018661499, 0.2548823058605194, -0.026976650580763817, -0.3092075288295746, 0.12093125283718109, 0.1135091632604599, -0.09752099215984344, -0.15819938480854034, 0.467640221118927, -0.3359432518482208, 0.028550995513796806, 0.054026708006858826, 0.12251276522874832, -0.26753970980644226, -0.16778606176376343, 0.00724573852494359, 0.04410345107316971, 0.3468678891658783, -0.1331421434879303, -0.42491427063941956, -0.045601196587085724, -0.29521653056144714, 0.19519048929214478, 0.26279449462890625, 0.27446967363357544, -0.015354511328041553, -0.17939533293247223, 0.13794802129268646, 0.12701371312141418, 0.3991086184978485, -0.18520726263523102, -0.15685276687145233, -0.08336551487445831, 0.12594930827617645, -0.18887260556221008, 0.053663741797208786, -0.11359232664108276, 0.10668975859880447, 0.16241750121116638, 0.05929332226514816, -0.2636672854423523, -0.1953766942024231, 0.137511745095253, 0.08628306537866592, 0.16331222653388977, -0.3234395384788513, -0.07841911911964417, -0.19659315049648285, -0.13586924970149994, 0.2577885687351227, -0.006181434728205204, -0.2898067831993103, -0.03264253959059715, 0.08358705043792725, -0.05086013302206993, 0.22913555800914764, -0.008726988919079304, 0.17341488599777222, 0.0596778579056263, 0.21080242097377777, -0.02444530464708805, -0.008577482774853706, 0.3802027702331543, 0.2015167623758316, -0.17109210789203644, -0.35532519221305847, 0.0014371216529980302, -0.08021998405456543, -0.3615974187850952, 0.25072285532951355, 0.2108691930770874, 0.21007823944091797, 0.26145607233047485, -0.14829246699810028, 0.15075673162937164, -0.23374539613723755, -0.030189139768481255, -0.06791061162948608, 0.007533286232501268, -0.5339677333831787, -0.8507201075553894, 0.5450389385223389, 0.32437360286712646, -0.3054795563220978, 0.44373372197151184, -0.024500772356987, -0.3562778830528259, 0.3782590329647064, 0.329973965883255, 1.0034146308898926, -0.07321766763925552, 0.2827601432800293, 0.1653262972831726, -0.3735775053501129, 0.42987215518951416, -0.1853654980659485, 0.29490500688552856, -0.19988380372524261, -0.26987752318382263, 0.10410719364881516, -0.0036484466399997473, -0.02802206017076969, 0.28348803520202637, -0.23576699197292328, 0.31240272521972656, -0.16876879334449768, 0.029425228014588356, 0.12953586876392365, 0.21414139866828918, -0.26988545060157776, -0.227239727973938, -0.24185191094875336, 0.13649655878543854, -0.22734884917736053, -0.21245190501213074, 0.021456871181726456, -0.1538650095462799, 0.0076255276799201965, 0.06421853601932526, -0.1773070991039276, -0.11582379788160324, 0.11999876797199249, -0.13377797603607178, -0.19474442303180695, -0.06841447204351425, 0.5051288604736328, 0.015424305573105812, 0.25004348158836365, 0.07930770516395569, 0.14134986698627472, 0.23059940338134766, 0.07139043509960175, 0.3418619930744171, -0.046026069670915604, -0.036250751465559006, 0.1730344146490097, 0.025741666555404663, 0.0010232338681817055, -0.05954526737332344, 0.13033832609653473, -0.6264166235923767, -0.47416746616363525, -0.008947601541876793, 0.320463091135025, -0.025615176185965538, 0.13314706087112427, 0.23690831661224365, -0.07707636058330536, -0.08411937206983566, 0.15529386699199677, 0.23977650701999664, -0.24776574969291687, 0.3586953580379486, -0.11311710625886917, -0.03689296543598175, -0.07945580035448074, 0.15876717865467072, -0.23023171722888947, 0.015225759707391262, 0.3005436360836029, -0.23293372988700867, -0.13038411736488342, -0.2351306527853012, -0.019168036058545113, 0.23810681700706482, -0.06477142870426178, 0.15195606648921967, -0.13655047118663788, -0.22427047789096832, -0.06621544808149338, -0.1361084133386612, 0.05216231569647789, 0.28071829676628113, -0.05464275926351547, 0.01373661495745182, -0.39536479115486145, 0.02080049179494381, 0.0989731103181839, 0.3021582365036011, -0.11496877670288086, 0.14674462378025055, -0.1508699357509613, 0.005486074835062027, -0.378298282623291, -0.24021001160144806, 0.11466891318559647, 0.31972330808639526, 0.07316653430461884, 0.1575394868850708, -0.2105623185634613, -0.21844375133514404, 0.06116204708814621, 0.04334103688597679, -0.04984719306230545, -0.23751045763492584, -0.1392945945262909, 0.06823787093162537, -0.0031457513105124235, 0.20376434922218323, 0.15675513446331024, -0.12760603427886963, -0.03966999799013138, -0.2093828022480011, -0.11454752087593079, 0.176520437002182, -0.01926150731742382, 0.5665611624717712, 0.5050375461578369, 0.10235331207513809, 0.009776938706636429, 0.1858241707086563, -0.14052699506282806, -0.11232603341341019, 0.019576726481318474, 0.2767688035964966, -0.013695599511265755, 0.05107124522328377, 0.02432316727936268, 0.2786237895488739, 0.14185374975204468, 0.08936906605958939, -0.0814656987786293, 0.19317273795604706, 0.10954701900482178, 0.4665202796459198, 0.061830274760723114, 0.3582407534122467, 0.08235417306423187, -0.48556745052337646, 0.4337492287158966, 0.21816928684711456, -0.1612999439239502, -0.0022095455788075924, -0.17344088852405548, 0.07349107414484024, 0.1654215306043625, -0.003153015859425068, -0.10503248870372772, -0.06873256713151932, -0.2876117527484894, -0.07359498739242554, 0.14936858415603638, -0.5486272573471069, 0.2823893129825592, 0.2594435214996338, -0.12507177889347076, -0.16082508862018585, 0.21707968413829803, -0.09828022867441177, -0.16002945601940155, 0.1616702526807785, 0.36011210083961487, 1.0041593313217163, 0.3350100815296173, 0.2119104564189911, 0.024365492165088654, 0.14911016821861267, 0.09338070452213287, 0.6433224678039551, 0.3787454068660736, 0.14951598644256592, 0.4294312596321106, 0.23128452897071838, 0.1897774040699005, -0.3021346628665924, 0.26007208228111267, 0.166106179356575, 0.1658439338207245, 0.06580915302038193, -0.11546698957681656, -0.1193489357829094, -0.26598110795021057, -0.0077394456602633, -0.0538184680044651, 0.10528252273797989, 0.5309209823608398, 0.1861804723739624, 0.1603645235300064, -0.41128915548324585, 0.20345577597618103, 0.2584473192691803, 0.4700572192668915, -0.18684883415699005, -0.2418614625930786, 0.025069210678339005, -0.2161089926958084, 0.7542921900749207, 0.041959524154663086, 0.024506790563464165, 0.28092649579048157, 0.04929599165916443, 0.032787661999464035, -0.41480621695518494, 0.009417344816029072, -0.15027864277362823, 0.1793755739927292, 0.2596071660518646, -0.12145692110061646, 0.0218115895986557, 0.13871453702449799, -0.11151232570409775, 0.20139579474925995, -0.046703189611434937, -0.07099413126707077, -0.08586221188306808, 0.3180973529815674, 0.5659241676330566, -0.2598888874053955, 0.04079611226916313, 0.029230797663331032, -0.1360316425561905, -0.06514092534780502, 0.006684079300612211, 0.14587543904781342, -0.17006884515285492, -0.05686282739043236, 0.0926966741681099, 0.3524608314037323, 0.403199166059494, 0.26844581961631775, -0.10742642730474472, -0.11384610086679459, -0.36881527304649353, -0.5009532570838928, -0.1890980750322342, 0.3571245074272156, -0.05158931016921997, -0.028194965794682503, -0.02227247692644596, -0.018850645050406456, -0.018899673596024513, -0.044772565364837646, 0.03546401485800743, 0.013845539651811123, 0.07077297568321228, -0.3659392297267914, -0.013891463167965412, -0.20218093693256378, -0.07325293868780136, 0.14310510456562042, -0.1260942965745926, -0.27879366278648376, 0.314703106880188, 0.044638242572546005, -0.22550247609615326, 0.2509477436542511, -0.08069916069507599, 0.14602090418338776, 0.033939238637685776, 0.1352875828742981, 0.1892540007829666, 0.04944802448153496, -0.11370181292295456, -0.1537809818983078, -0.10955619812011719, -0.6963853240013123, 0.005131698679178953, 0.058589234948158264, 0.054866813123226166, -0.06912785023450851, -0.10537717491388321, -0.41169610619544983, 0.5068255066871643, -0.2618783414363861, 0.08110793679952621, -0.5822643637657166, 0.17405766248703003, 0.19211320579051971, 0.19045811891555786, -0.14750568568706512, 0.2194092571735382, 0.09773163497447968, 0.18745550513267517, -0.23057420551776886, 0.10078958421945572, 0.4633548855781555, 0.021655425429344177, -0.17914241552352905, -0.216499462723732, 0.24227537214756012, 0.01254197396337986, -0.14014942944049835, -0.3644137680530548, -0.17112572491168976, 0.17598293721675873, -0.041666626930236816, 0.2560969889163971, 0.4652126133441925, 0.046045467257499695, -0.04631755128502846, -0.18348334729671478, 0.0035771045368164778, -0.06137636676430702, -0.045356303453445435, 0.00507642887532711, -0.4476947486400604 ]
https://github.com/huggingface/datasets/issues/163
[Feature request] Add cos-e v1.0
Sounds good, @mariamabarham do you want to give a look? I think we should have two configurations so we can allow either version of the dataset to be loaded with the `1.0` version being the default maybe. Cc some authors of the great cos-e: @nazneenrajani @bmccann
I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf).
46
[Feature request] Add cos-e v1.0 I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf). Sounds good, @mariamabarham do you want to give a look? I think we should have two configurations so we can allow either version of the dataset to be loaded with the `1.0` version being the default maybe. Cc some authors of the great cos-e: @nazneenrajani @bmccann
[ -0.25224724411964417, -0.2662770450115204, -0.12816211581230164, -0.3826158940792084, -0.41568827629089355, -0.10087808221578598, 0.019268780946731567, 0.2192731350660324, -0.08714974671602249, 0.3809593915939331, 0.05702600255608559, 0.3763355314731598, -0.015546808950603008, 0.3852824866771698, -0.0764765664935112, 0.14052660763263702, 0.08472937345504761, 0.3658883571624756, -0.08661047369241714, -0.13832923769950867, -0.2019471526145935, 0.11041351407766342, -0.06942540407180786, -0.05444715544581413, 0.14053836464881897, 0.17084471881389618, -0.19724233448505402, -0.07244247943162918, -0.22426988184452057, -0.20771391689777374, 0.35505375266075134, 0.45551031827926636, 0.22513900697231293, 0.06983251124620438, -0.00010584220581222326, -0.23818445205688477, 0.4025997817516327, -0.07296527922153473, -0.3339165151119232, -0.2384468913078308, -0.41790133714675903, -0.3771292567253113, 0.06569387018680573, -0.09597756713628769, 0.06359320133924484, 0.09290825575590134, 0.16502238810062408, -0.19661813974380493, 0.06413621455430984, 0.011999391950666904, 0.30262696743011475, 0.17412418127059937, -0.08758492767810822, -0.34514889121055603, 0.3349485397338867, 0.449106901884079, -0.35373684763908386, -0.17945313453674316, 0.31677186489105225, 0.013062856160104275, -0.06734270602464676, 0.41108182072639465, 0.2277282476425171, -0.1817954033613205, -0.025220628827810287, -0.06428226083517075, 0.1812652349472046, 0.007419890724122524, 0.14173486828804016, 0.1168774664402008, 0.5061965584754944, 0.12980081140995026, -0.41553935408592224, 0.28998300433158875, 0.004278996028006077, -0.49353086948394775, 0.26659244298934937, -0.1062392145395279, -0.029614154249429703, 0.06423833221197128, 0.18722118437290192, -0.3388349413871765, -0.4957718551158905, 0.20651458203792572, -0.07795700430870056, 0.0905710831284523, -0.15790334343910217, -0.1536710560321808, 0.12449917197227478, 0.10767868906259537, -0.043482277542352676, 0.04136797413229942, -0.17636604607105255, 0.20438635349273682, -0.24313992261886597, -0.25906258821487427, 0.2833954393863678, 0.001078931731171906, 0.45887547731399536, 0.08281146734952927, 0.154005229473114, 0.04369635134935379, -0.0718052089214325, 0.10415893793106079, 0.1537679135799408, 0.20156952738761902, 0.46074262261390686, -0.05985747650265694, 0.39925891160964966, 0.2871788740158081, 0.476828008890152, 0.0420064814388752, 0.1551857441663742, 0.03422985225915909, -0.06027299165725708, -0.0319339893758297, 0.07255350798368454, -0.5815814733505249, -0.09604071825742722, 0.05002417787909508, -0.016389036551117897, 0.0166013166308403, -0.03814131021499634, 0.09351347386837006, 0.2691534459590912, 0.22098198533058167, 0.2916280925273895, -0.0643903836607933, -0.03568077087402344, -0.4872020184993744, -0.29533958435058594, -0.07494291663169861, -0.2445651739835739, 0.2993689775466919, 0.327759325504303, -0.4222894310951233, 0.13820572197437286, -0.3130995035171509, 0.2833765149116516, 0.04933526739478111, -0.18007919192314148, 0.3390869200229645, -0.22517696022987366, 0.12475133687257767, 0.057131633162498474, 0.08530967682600021, -0.2785825729370117, 0.5125001072883606, -0.1971520334482193, 0.12465782463550568, -0.3382830023765564, -0.44855234026908875, -0.3868960738182068, 0.28708532452583313, -0.0740516185760498, -0.15551422536373138, -0.1518208384513855, 0.36664536595344543, -0.2398143708705902, -0.1644044816493988, -0.0805048793554306, 0.3010803163051605, -0.27182281017303467, -0.12088683247566223, 0.02714027464389801, 0.14325301349163055, -0.1281464695930481, -0.04009632393717766, -0.5082139372825623, -0.09364712238311768, -0.1425672322511673, -0.029935738071799278, -0.12085220962762833, -0.2735488712787628, -0.1699308156967163, -0.17351989448070526, 0.3994155526161194, -0.33321884274482727, -0.14639067649841309, 0.23286594450473785, -0.028698723763227463, -0.24974243342876434, 0.14834187924861908, -0.1565067172050476, 0.0467190146446228, -0.27039507031440735, -0.228197380900383, 0.04900587350130081, -0.14814819395542145, -0.1040414348244667, -0.14730118215084076, -0.4412725567817688, -0.18274036049842834, 0.2603660225868225, 0.24740350246429443, 0.03346296772360802, 0.22792431712150574, -0.39961373805999756, 0.09001251310110092, -0.16994144022464752, 0.17561861872673035, -0.019764386117458344, 0.7601085901260376, 0.032792869955301285, 0.15846048295497894, 0.06392256915569305, -0.3966706097126007, 0.07201731950044632, 0.090555340051651, 0.1692187637090683, 0.22833293676376343, -0.3766869008541107, -0.03511645644903183, -0.04920860752463341, 0.0924888476729393, -0.09495951235294342, 0.2041812539100647, 0.1204761490225792, -0.13210278749465942, -0.015266706235706806, -0.028222637251019478, 0.25926652550697327, -0.24022382497787476, -0.15056875348091125, -0.07579410076141357, 0.16699030995368958, -0.18648074567317963, 0.03914570435881615, -0.18251173198223114, 0.12863817811012268, -0.07770595699548721, -0.0659889280796051, -0.12712711095809937, 0.3903431296348572, -0.09755901992321014, 0.23098590970039368, 0.055191799998283386, 0.3234146535396576, 0.14523349702358246, -0.17544732987880707, 0.07596839964389801, -0.1623835563659668, -0.10134494304656982, 0.18791493773460388, -0.22592833638191223, 0.38410717248916626, 0.06922294199466705, -0.15123452246189117, 0.009792857803404331, -0.008774581365287304, 0.062455467879772186, -0.09672992676496506, -0.36420097947120667, -0.11718843877315521, 0.18756090104579926, 0.11140123009681702, -0.44378700852394104, -0.28570112586021423, -0.30886122584342957, -0.08802053332328796, 0.03273776173591614, -0.06471413373947144, 0.11028403788805008, 0.20297089219093323, -0.0012313614133745432, -0.23741556704044342, 0.30815067887306213, -0.10598635673522949, 0.2734644114971161, 0.3854377269744873, -0.16695116460323334, -0.021352330222725868, -0.20481480658054352, 0.11400572210550308, 0.1383959800004959, 0.2815913259983063, 0.04978850111365318, 0.10224246978759766, 0.3348846733570099, 0.09427259117364883, -0.3502333462238312, -0.308944970369339, -0.045419320464134216, 0.057697687298059464, -0.15502671897411346, -0.19572260975837708, -0.03369687497615814, 0.0711766853928566, 0.0014666501665487885, 0.1478462517261505, 0.16372084617614746, -0.19144974648952484, 0.4073890745639801, 0.26924824714660645, -0.22280806303024292, 0.32431355118751526, 0.036760956048965454, 0.8047566413879395, -0.12738488614559174, 0.18538615107536316, -0.2448902279138565, 0.05357374623417854, -0.31329861283302307, 0.2546367049217224, -0.22676561772823334, -0.22785437107086182, 0.46085333824157715, -0.10080960392951965, 0.24583852291107178, -0.11814724653959274, -0.7101550102233887, 0.09934397041797638, -0.10333838313817978, -0.0934278815984726, 0.030384935438632965, -0.21468698978424072, -0.0738753005862236, 0.019448844715952873, 0.05901803448796272, -0.44859644770622253, -0.4825385510921478, -0.10999597609043121, -0.24806146323680878, 0.13051088154315948, -0.3649931848049164, -0.48683127760887146, 0.1457367092370987, -0.2606458067893982, -0.2076708972454071, -0.052967369556427, 0.16154085099697113, 0.26981109380722046, 0.12356352806091309, -0.0581195168197155, 0.06886522471904755, 0.05131303891539574, -0.38603153824806213, 0.04582156240940094, 0.34243741631507874, -0.3188912868499756, -0.4158433973789215, 0.1217731386423111, 0.011760883033275604, 0.15939804911613464, 0.1757790446281433, -0.3779384195804596, -0.3949556350708008, 0.275278240442276, 0.12953287363052368, -0.14654043316841125, 0.06901132315397263, 0.5344857573509216, 0.06983834505081177, -0.21253998577594757, -0.07770739495754242, -0.4181550145149231, 0.18073123693466187, 0.4679787755012512, 0.143210768699646, -0.29487988352775574, 0.2655486762523651, 0.19312961399555206, 0.32020488381385803, 0.16470728814601898, -0.07552259415388107, 0.12141604721546173, 0.13372135162353516, 0.21813784539699554, -0.20591577887535095, 0.01071433536708355, 0.01658337004482746, 0.05266035348176956, 0.17323482036590576, 0.5349008440971375, 0.19639094173908234, -0.40721455216407776, -0.37295639514923096, -0.22823022305965424, 0.09779666364192963, 0.0007366268546320498, -0.1938166469335556, 0.16490212082862854, 0.4404818117618561, -0.06908103823661804, -0.15247684717178345, -0.32904037833213806, -0.10789810866117477, 0.09796715527772903, 0.34180477261543274, 0.15771318972110748, -0.08698428422212601, 0.09859805554151535, 0.1773262470960617, -0.28397229313850403, 0.35610535740852356, 0.14722830057144165, -0.004486479330807924, 0.005454095546156168, -0.10187996178865433, 0.15186353027820587, 0.019838888198137283, 0.1859121173620224, -0.2136530727148056, -0.16913346946239471, -0.1400272250175476, -0.08662836253643036, 0.08970759809017181, -0.06182192638516426, -0.111711286008358, -0.2805306017398834, -0.0905054435133934, 0.05194638669490814, 0.027261465787887573, -0.14532747864723206, 0.6831695437431335, -0.17081350088119507, 0.010884147137403488, -0.2852765917778015, 0.050516121089458466, -0.23356208205223083, -0.0009023428428918123, -0.017939191311597824, -0.1901806890964508, 0.0005647466750815511, -0.10173072665929794, -0.15057475864887238, 0.23931929469108582, 0.013858205638825893, -0.006709089037030935, 0.14193889498710632, -0.14012408256530762, 0.011414074338972569, 0.11479271203279495, 0.08129554986953735, 0.047842640429735184, 0.37294796109199524, 0.12629081308841705, 0.22759829461574554, -0.2665453255176544, 0.17605388164520264, -0.1937774419784546, 0.06537343561649323, 0.2491942197084427, 0.030140884220600128, -0.06442731618881226, -0.29717180132865906, -0.22939139604568481, -0.12345936894416809, -0.20089145004749298, 0.2619432806968689, 0.041544556617736816, -0.37265822291374207, 0.028384987264871597, 0.32451045513153076, 0.021752752363681793, -0.27290016412734985, 0.2598053216934204, 0.21533307433128357, -0.24798963963985443, 0.33155807852745056, 0.0006588782998733222, 0.7347700595855713, 0.24774301052093506, -0.06344792246818542, 0.2377440184354782, -0.04784703254699707, 0.4994877874851227, -0.15337014198303223, 0.013707942329347134, 0.07889310270547867, -0.13275177776813507, -0.09385058283805847, 0.2984381914138794, 0.37705013155937195, -0.09206005930900574, -0.3057999908924103, 0.1782730519771576, 0.36761510372161865, 0.2488742470741272, 0.08150095492601395, 0.18558430671691895, 0.33593547344207764, -0.3099149465560913, 0.1831713318824768, 0.18838095664978027, 0.00758065702393651, 0.16981589794158936, 0.0008189335349015892, -0.036344531923532486, 0.01405798178166151, -0.3261224329471588, -0.22565610706806183, -0.14075037837028503, 0.344286173582077, -0.00929958838969469, -0.035762373358011246, -0.07223028689622879, 0.43538060784339905, -0.3726445138454437, 0.207982137799263, 0.19100742042064667, -0.28965577483177185, 0.1257748156785965, 0.16576583683490753, 0.053725019097328186, -0.06958981603384018, -0.09448564797639847, 0.077021025121212, -0.16798190772533417, -0.09140422195196152, 0.18939553201198578, 0.21163731813430786, -0.1577853560447693, -0.0802488699555397, -0.14002549648284912, -0.032127443701028824, 0.04893331229686737, 0.3648076057434082, 0.5734385848045349, 0.19827178120613098, -0.28032052516937256, 0.24688148498535156, 0.22918599843978882, -0.1116868108510971, -0.07789948582649231, 0.15205664932727814, -0.10007655620574951, -0.20936448872089386, 0.4151041507720947, 0.41105833649635315, -0.40592581033706665, -0.04722759127616882, -0.08289723843336105, -0.08956631273031235, -0.2743074297904968, -0.04644768312573433, 0.12019768357276917, -0.5605069994926453, -0.19973421096801758, 0.06941709667444229, -0.20039421319961548, 0.18112961947917938, 0.7872421741485596, 0.14098186790943146, 0.17524492740631104, -0.05200910568237305, -0.018956856802105904, -0.18221916258335114, 0.06492505222558975, -0.4313204884529114, 0.3021855652332306, -0.026818621903657913, -0.15485018491744995, -0.11379797011613846, 0.03980981558561325, -0.4568732976913452, -0.192204087972641, 0.26230764389038086, -0.13239455223083496, 0.372192919254303, 0.31053921580314636, 0.26454269886016846, 0.0035349037498235703, 0.21191680431365967, -0.14141185581684113, -0.33388087153434753, -0.2028176188468933, 0.06465192139148712, 0.09212338924407959, 0.06677127629518509, 0.17337298393249512, -0.03824377804994583, -0.11493047326803207, -0.2642992436885834, 0.032551076263189316, 0.26225918531417847, -0.029773715883493423, -0.1581730991601944, 0.18510586023330688, -0.24483998119831085, -0.07198487967252731, 0.15404702723026276, 0.38236004114151, 0.18938863277435303, 0.08267606794834137, -0.06526104360818863, 0.42594343423843384, -0.007294509559869766, -0.1126653328537941, 0.04660404101014137, 0.378113329410553, -0.27740931510925293, 0.30024346709251404, 0.018675293773412704, -0.02458520047366619, -0.08572182059288025, -0.03785381838679314, 0.13398224115371704, -0.41137969493865967, -0.07079403102397919, 0.028907058760523796, 0.03366865590214729, 0.3719577193260193, -0.2801135778427124, 0.05250128358602524, -0.0581754669547081, -0.05517246946692467, 0.21768657863140106, 0.19678077101707458, 0.3475117087364197, 0.015416697598993778, 0.7643713355064392, 0.03258528560400009, 0.012449795380234718, -0.053610850125551224, 0.45254966616630554, 0.10950743407011032, 0.045196544378995895, 0.09279991686344147, 0.08463882654905319, 0.06753072142601013, 0.17613504827022552, -0.0687374398112297, 0.11972717195749283, -0.08243265002965927, 0.4220128059387207, 0.5082917213439941, 0.009178321808576584, -0.33825480937957764, 0.07828831672668457, 0.2829192578792572, 0.0896713137626648, 0.03318409249186516, 0.30095821619033813, -0.1659034639596939, 0.011287045665085316, 0.07916238158941269, -0.7445870041847229, -0.15216273069381714, 0.020008113235235214, -0.1553734391927719, -0.15849334001541138, -0.13477909564971924, -0.406162828207016, -0.06609311699867249, 0.16439181566238403, 0.007545987609773874, 0.15479020774364471, -0.19950591027736664, 0.015690436586737633, -0.10653693228960037, -0.00798280443996191, 0.11110534518957138, 0.22404608130455017, 0.026627108454704285, 0.3685714900493622, -0.49482378363609314, -0.2772008776664734, 0.3008657395839691, 0.3290163576602936, -0.01646159216761589, -0.07080629467964172, 0.3832436203956604, 0.1276649832725525, 0.08333251625299454, -0.1014527902007103, 0.04092063382267952, 0.08495252579450607, 0.09764701873064041, 0.146663635969162, 0.2090858668088913, 0.2040061205625534, -0.13698969781398773, 0.24193686246871948, -0.07487759739160538, -0.13943202793598175, -0.2820708751678467, 0.11750295758247375, 0.16903412342071533, 0.0901317447423935, 0.03028159774839878, 0.16643081605434418, 0.041157204657793045, -0.37090566754341125, -0.07294442504644394, -0.05760603025555611, -0.12956935167312622, -0.1690276712179184, 0.1355428695678711, -0.1693425327539444, -0.04852965474128723, 0.22995752096176147, 0.007202232722193003, -0.013600819744169712, -0.4024968147277832, 0.05314406380057335, 0.35809582471847534, 0.1434372067451477, -0.0768827497959137, 0.0670180469751358, 0.22650213539600372, 0.259703665971756, -0.12988053262233734, 0.1573919951915741, -0.25790947675704956, 0.36999547481536865, 0.09451108425855637, -0.30971941351890564, 0.1618374139070511, 0.45948630571365356, -0.028369272127747536, -0.11244360357522964, -0.15681952238082886, -0.10699551552534103, 0.047833651304244995, 0.0735950767993927, -0.24181421101093292, 0.07018902897834778, 0.13543926179409027, -0.0018166357185691595, 0.535402238368988, 0.025703078135848045, 0.17436566948890686, -0.21326428651809692, -0.04778820276260376, -0.4498699903488159, -0.20575033128261566, -0.08338765054941177, 0.028017450124025345, 0.08583541214466095, 0.08767566084861755, -0.0047048283740878105, 0.05368756502866745, -0.03177979588508606, 0.27985140681266785, -0.014120352454483509, 0.00022554157476406544, -0.24943330883979797, 0.47038963437080383, -0.37131333351135254, -0.2708432972431183, -0.04918542504310608, -0.15531067550182343, 0.11307983845472336, -0.17197906970977783, -0.10465744882822037, -0.17666929960250854, 0.2870737612247467, -0.07113718241453171, 0.051782917231321335, -0.18545478582382202, 0.3238445222377777, 0.3122849464416504, -0.00006310753815341741, -0.33270421624183655, 0.06873585283756256, 0.19568853080272675, 0.17945131659507751, -0.13641560077667236, 0.2679246664047241, -0.07479602843523026, 0.023499902337789536, -0.1861780732870102, -0.011183187365531921, -0.19011494517326355, -0.06747494637966156, 0.032859884202480316, -0.3748263716697693 ]
https://github.com/huggingface/datasets/issues/163
[Feature request] Add cos-e v1.0
cos_e v1.0 is related to CQA v1.0 but only CQA v1.11 dataset is available on their website. Indeed their is lots of ids in cos_e v1, which are not in CQA v1.11 or the other way around. @sarahwie, @thomwolf, @nazneenrajani, @bmccann do you know where I can find CQA v1.0
I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf).
50
[Feature request] Add cos-e v1.0 I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf). cos_e v1.0 is related to CQA v1.0 but only CQA v1.11 dataset is available on their website. Indeed their is lots of ids in cos_e v1, which are not in CQA v1.11 or the other way around. @sarahwie, @thomwolf, @nazneenrajani, @bmccann do you know where I can find CQA v1.0
[ -0.02323293499648571, -0.18676243722438812, -0.17846323549747467, -0.3313111364841461, -0.5435763001441956, 0.06571663916110992, -0.17831730842590332, 0.12594039738178253, -0.3440611958503723, 0.46897250413894653, 0.23541802167892456, 0.12405304610729218, 0.07490003108978271, 0.42129889130592346, 0.01564243622124195, 0.08941134065389633, 0.1170303225517273, 0.40649187564849854, 0.156411811709404, -0.07623076438903809, -0.046966202557086945, 0.3286639451980591, -0.268859326839447, -0.012009824626147747, 0.16894973814487457, 0.11247790604829788, -0.32557418942451477, -0.26800331473350525, -0.14709492027759552, -0.16988083720207214, 0.4301478862762451, 0.2873972952365875, 0.3542037308216095, -0.00829679612070322, -0.0001180448816739954, -0.24944999814033508, 0.2678096890449524, -0.06388585269451141, -0.2248019427061081, -0.28847646713256836, -0.39872676134109497, -0.37567535042762756, -0.12163138389587402, -0.01597958616912365, 0.13137966394424438, 0.0580621175467968, 0.20080436766147614, 0.015675347298383713, -0.1259213536977768, 0.1071854829788208, 0.18949463963508606, 0.18075960874557495, 0.1411362886428833, -0.37958770990371704, 0.17057491838932037, 0.3191809058189392, -0.27347421646118164, -0.2285313755273819, 0.2699383497238159, 0.025855710729956627, 0.30776649713516235, 0.48162439465522766, 0.2874336540699005, -0.3409464955329895, -0.17208893597126007, -0.013270877301692963, 0.11673440784215927, -0.12508030235767365, 0.2783507704734802, -0.045265987515449524, 0.5306006073951721, 0.17983447015285492, -0.3857193887233734, 0.4131040573120117, 0.04533262923359871, -0.47312355041503906, 0.2673744559288025, -0.013816768303513527, 0.06820397078990936, 0.11895697563886642, 0.2374476045370102, -0.3053060472011566, -0.43842682242393494, 0.16605913639068604, -0.16326332092285156, 0.0052228220738470554, -0.11229739338159561, -0.18681909143924713, 0.06563159078359604, 0.045105595141649246, -0.048754893243312836, 0.18406380712985992, -0.25097814202308655, 0.26039373874664307, -0.22966554760932922, -0.3830535113811493, 0.2817988395690918, 0.021808341145515442, 0.3297160565853119, -0.034621741622686386, 0.15862123668193817, -0.21944288909435272, -0.24751417338848114, 0.0007928924169391394, 0.12062620371580124, 0.3228638768196106, 0.3589962124824524, 0.052947454154491425, 0.14591944217681885, 0.23537777364253998, 0.44060102105140686, -0.040402285754680634, 0.0712786540389061, 0.01292085275053978, -0.16717663407325745, 0.05689515545964241, 0.06021237000823021, -0.6049149036407471, -0.1965155303478241, -0.004578236490488052, -0.23373329639434814, 0.06281226128339767, -0.14380186796188354, 0.03192555904388428, 0.4180451035499573, 0.21806325018405914, 0.22002330422401428, -0.19355079531669617, -0.009807290509343147, -0.6159117817878723, -0.27655524015426636, -0.014677955769002438, -0.0672636404633522, 0.3063664734363556, 0.21016515791416168, -0.361815482378006, 0.22055011987686157, -0.3328658640384674, 0.22424130141735077, 0.05530916154384613, -0.15858890116214752, 0.436467707157135, -0.22660820186138153, 0.18580681085586548, -0.022986527532339096, 0.06077193468809128, -0.2817750573158264, 0.49024346470832825, -0.21706923842430115, 0.1842230260372162, -0.3824465274810791, -0.43082091212272644, -0.5014216899871826, 0.14038513600826263, 0.050710294395685196, -0.19810707867145538, 0.03673640638589859, 0.24654188752174377, -0.2765118479728699, -0.17123249173164368, 0.02272367663681507, 0.29645469784736633, -0.17472507059574127, -0.13263075053691864, -0.005532918497920036, 0.1790468841791153, -0.26312336325645447, -0.06219679117202759, -0.3929760754108429, 0.0770600438117981, -0.1144099086523056, 0.05869674310088158, -0.06112257391214371, -0.31643980741500854, -0.1432291716337204, -0.2597504258155823, 0.33863118290901184, -0.6128677129745483, -0.5238029360771179, 0.14918266236782074, -0.03131825104355812, -0.2901611030101776, 0.21898722648620605, 0.005280482582747936, 0.007204203866422176, -0.16219724714756012, -0.20951691269874573, -0.03626927360892296, -0.2873411774635315, -0.333136647939682, 0.021973755210638046, -0.4311452805995941, -0.07969656586647034, 0.10743052512407303, 0.09247304499149323, 0.09353815019130707, 0.23239195346832275, -0.6215126514434814, 0.14444561302661896, -0.01956114172935486, 0.010354569181799889, 0.06936360150575638, 0.6898933053016663, 0.07786308974027634, 0.26113927364349365, 0.05161323398351669, -0.3183966279029846, 0.05620035156607628, -0.21524697542190552, 0.06954575330018997, 0.2679485082626343, -0.3772903382778168, -0.1154729425907135, 0.0034698948729783297, 0.16920095682144165, -0.28243616223335266, 0.04788865149021149, 0.04885225370526314, 0.10802208632230759, 0.0012773934286087751, 0.013500618748366833, 0.33671069145202637, -0.15173602104187012, -0.17996639013290405, -0.10303966701030731, 0.17719748616218567, -0.195079505443573, 0.008516152389347553, 0.016285602003335953, 0.2220599204301834, 0.10666114091873169, 0.019013283774256706, -0.08943834900856018, 0.19991281628608704, -0.16442212462425232, 0.25171536207199097, 0.2599520981311798, 0.4301054775714874, 0.22773417830467224, -0.22602073848247528, 0.06586575508117676, -0.09690889716148376, -0.1978672444820404, 0.35074302554130554, -0.010788246057927608, 0.25207433104515076, 0.1498488485813141, -0.07520967721939087, -0.007432149723172188, 0.2028426229953766, 0.031593792140483856, 0.04165678098797798, -0.20685191452503204, -0.1189875453710556, 0.21474160254001617, 0.10000555217266083, -0.34825125336647034, -0.3811200261116028, -0.20631039142608643, 0.16081732511520386, 0.1386513113975525, -0.11609704047441483, -0.0007918394985608757, 0.11394800990819931, -0.02068331278860569, -0.08092055469751358, 0.346027135848999, -0.10363273322582245, 0.09568565338850021, 0.2869536280632019, -0.2408585101366043, -0.1860809177160263, -0.00962152797728777, 0.1602334976196289, 0.15206782519817352, 0.2799004316329956, -0.2412099689245224, -0.1741049885749817, 0.3792100250720978, 0.1788063943386078, -0.12696224451065063, -0.23240984976291656, -0.014872080646455288, 0.05603449046611786, 0.028574971482157707, -0.09625102579593658, -0.13883927464485168, 0.005258068908005953, -0.06992211937904358, -0.0594109445810318, 0.18317686021327972, -0.19776979088783264, 0.27721282839775085, 0.17876549065113068, 0.07255703210830688, 0.27476873993873596, -0.00020035906345583498, 0.776387095451355, -0.22260735929012299, 0.30331772565841675, -0.07585015147924423, 0.13949935138225555, -0.42412352561950684, 0.13569289445877075, -0.2917245328426361, -0.17309953272342682, 0.4106995165348053, -0.09878092259168625, 0.333431214094162, -0.08852579444646835, -0.8045451045036316, -0.0008278439636342227, -0.10774533450603485, -0.10973905026912689, 0.11922898143529892, -0.313337504863739, -0.09177692979574203, 0.024535540491342545, -0.025782927870750427, -0.2734341323375702, -0.48534730076789856, -0.164653480052948, -0.3983460068702698, 0.0878530964255333, -0.17175938189029694, -0.38223880529403687, -0.04387272521853447, -0.07672493904829025, -0.12428012490272522, -0.12437772005796432, 0.26520416140556335, 0.12538957595825195, -0.00853987317532301, -0.07314718514680862, 0.17700503766536713, -0.04666570574045181, -0.19722265005111694, 0.20906880497932434, 0.33739587664604187, -0.3322138786315918, -0.32960522174835205, 0.3415122628211975, -0.021702595055103302, 0.2813383936882019, 0.011478787288069725, -0.42799150943756104, -0.1520685851573944, 0.06688933819532394, 0.13095764815807343, 0.051632028073072433, -0.036550503224134445, 0.7915481328964233, -0.11600464582443237, -0.07611801475286484, -0.027554836124181747, -0.5734850168228149, 0.1022515594959259, 0.5704919099807739, 0.24227374792099, -0.11164876818656921, 0.3174266517162323, 0.1585416942834854, 0.5127479434013367, 0.3592984974384308, 0.010425216518342495, 0.022733936086297035, 0.1787562519311905, 0.04132813587784767, -0.12431827187538147, -0.23321403563022614, 0.030882395803928375, 0.08355014771223068, 0.08108741044998169, 0.35798510909080505, 0.17127497494220734, -0.25755178928375244, -0.23424024879932404, -0.43134576082229614, -0.04863084480166435, 0.13530905544757843, -0.34676018357276917, 0.08997038751840591, 0.6666926741600037, -0.05727138742804527, -0.1794421672821045, -0.48417335748672485, -0.05204249173402786, 0.1998935043811798, 0.4482497274875641, 0.3122659921646118, -0.002332007046788931, 0.1590178906917572, -0.0632472038269043, -0.29903727769851685, 0.38223668932914734, 0.09644635021686554, 0.07628419250249863, -0.023621540516614914, 0.057052452117204666, 0.16065460443496704, 0.003988228738307953, 0.20746296644210815, -0.11883389204740524, -0.20663711428642273, -0.08955521881580353, -0.04198891296982765, 0.32143884897232056, -0.03235861659049988, 0.03982214629650116, -0.24847358465194702, -0.10593654215335846, -0.1875954270362854, 0.19501066207885742, -0.0317000150680542, 0.661746621131897, -0.2799369990825653, -0.17864874005317688, -0.11406060308218002, -0.056554052978754044, -0.07454974204301834, 0.04568130150437355, -0.058856017887592316, -0.26100653409957886, -0.01762564294040203, -0.03928040713071823, -0.17231251299381256, 0.2192324697971344, -0.2052268534898758, 0.0071362778544425964, 0.04859590157866478, -0.06891928613185883, -0.018689116463065147, 0.08036047220230103, 0.10861840099096298, -0.028514139354228973, 0.23418088257312775, 0.0030013469513505697, 0.39899662137031555, -0.3178146481513977, 0.1732623130083084, -0.044184811413288116, 0.22356382012367249, 0.1530674397945404, -0.031643737107515335, 0.07378581166267395, -0.1780216544866562, -0.3110802173614502, -0.2239464670419693, -0.41653239727020264, 0.35408511757850647, 0.09462448954582214, -0.20846572518348694, 0.0869981050491333, 0.24481096863746643, 0.09113103151321411, -0.2714417576789856, 0.33740630745887756, 0.39424213767051697, -0.013790697790682316, 0.17862364649772644, -0.040982283651828766, 0.782871425151825, 0.38780489563941956, -0.11633782833814621, 0.08909591287374496, -0.1653096228837967, 0.30338573455810547, 0.02732103317975998, 0.10152129828929901, 0.165683314204216, -0.12508714199066162, -0.13380086421966553, 0.19046875834465027, 0.3938685953617096, -0.19231145083904266, -0.2761596739292145, 0.3090684711933136, 0.2963235676288605, 0.3791378438472748, 0.18517270684242249, 0.046606555581092834, 0.3087802529335022, -0.012175876647233963, 0.22883780300617218, 0.08569778501987457, 0.17001846432685852, 0.44463011622428894, -0.015017685480415821, -0.149693563580513, -0.07364744693040848, -0.24480433762073517, -0.31907719373703003, -0.24239026010036469, 0.1053241714835167, -0.12534378468990326, 0.06371266394853592, -0.13501380383968353, 0.479557603597641, -0.18676088750362396, 0.4332153797149658, 0.019766712561249733, -0.318354994058609, 0.11428608745336533, 0.24963220953941345, -0.2459229975938797, 0.05072203278541565, -0.032061316072940826, -0.2735676169395447, -0.19744296371936798, -0.12815222144126892, 0.4675576984882355, 0.2643013298511505, -0.18208633363246918, -0.29033833742141724, -0.17927522957324982, -0.009771977551281452, 0.16735287010669708, 0.37810957431793213, 0.4931950271129608, 0.2174016535282135, -0.23251928389072418, 0.13885164260864258, 0.14542190730571747, -0.12461980432271957, -0.33353549242019653, -0.1084514632821083, 0.12606148421764374, -0.15928490459918976, 0.20721007883548737, 0.42981231212615967, -0.32244277000427246, 0.07744569331407547, -0.12384378165006638, 0.03547756373882294, -0.13220739364624023, -0.2717389464378357, 0.030616676434874535, -0.43420037627220154, -0.15630154311656952, 0.03174436464905739, -0.17103658616542816, 0.1918586790561676, 0.9666600227355957, 0.06149475648999214, 0.20077769458293915, -0.18781888484954834, -0.015118599869310856, -0.20959633588790894, 0.17482054233551025, -0.5250990986824036, 0.21809245645999908, -0.1759723722934723, -0.07979467511177063, -0.16315677762031555, 0.2668922543525696, -0.2849743664264679, -0.22890253365039825, 0.17995333671569824, -0.15777553617954254, 0.3327464461326599, 0.2898227870464325, 0.1722332239151001, 0.11206047981977463, 0.04333662614226341, -0.13135670125484467, -0.14088349044322968, -0.04713470861315727, 0.10009501874446869, 0.12241394072771072, 0.0846710205078125, 0.3393620252609253, -0.037495002150535583, -0.18758778274059296, -0.19110754132270813, 0.19971632957458496, 0.24985694885253906, -0.0930987298488617, 0.08678851276636124, 0.24760368466377258, -0.29602277278900146, -0.1283227503299713, -0.08523260802030563, 0.5458859801292419, 0.2416352778673172, 0.13735003769397736, 0.1575447916984558, 0.3756001591682434, -0.09079961478710175, -0.038903094828128815, -0.07435794174671173, 0.4465693235397339, -0.3893511891365051, 0.2567906975746155, 0.036853671073913574, 0.09243869036436081, -0.13186980783939362, 0.03822946548461914, -0.2625071406364441, -0.5318458676338196, -0.012972931377589703, 0.1900569200515747, -0.05724000185728073, 0.20679323375225067, -0.36436280608177185, 0.022032978013157845, -0.26102885603904724, -0.09930068999528885, -0.02965007722377777, 0.2112356275320053, 0.37755802273750305, 0.07074391841888428, 0.8590605854988098, 0.11811655014753342, -0.11029195785522461, -0.0721646249294281, 0.4866429567337036, 0.22771117091178894, 0.39008045196533203, 0.13891077041625977, 0.09025684744119644, 0.0017616688273847103, 0.10719459503889084, -0.16587112843990326, -0.07671118527650833, -0.13972194492816925, 0.4415699541568756, 0.454982727766037, 0.32754531502723694, -0.20021578669548035, -0.20803426206111908, 0.24367699027061462, 0.18482738733291626, 0.014334077946841717, 0.23309212923049927, -0.08115053921937943, 0.12781958281993866, 0.09431014209985733, -0.7621111869812012, -0.09892009198665619, 0.07419566810131073, -0.16120032966136932, -0.15801450610160828, -0.12280502915382385, -0.2658146917819977, -0.11310730874538422, 0.3679632842540741, 0.0979592576622963, 0.226737380027771, -0.2759630084037781, -0.14949993789196014, -0.05910404771566391, -0.2040785551071167, 0.2876170873641968, 0.25948917865753174, 0.0613488107919693, 0.48437637090682983, -0.5644538402557373, -0.30071717500686646, 0.42977404594421387, 0.21083788573741913, 0.019138116389513016, -0.2560524046421051, 0.2916175127029419, 0.0555403009057045, -0.02176857925951481, -0.26670530438423157, 0.23909790813922882, 0.18271829187870026, -0.14026692509651184, 0.1108374148607254, -0.014656143262982368, 0.062215372920036316, 0.043425094336271286, -0.058801330626010895, 0.012008662335574627, -0.08062709122896194, -0.4757099151611328, 0.13423475623130798, -0.20503151416778564, 0.16521906852722168, -0.12193158268928528, 0.4493805468082428, 0.08370654284954071, -0.3802337348461151, -0.2866358160972595, -0.04445837065577507, -0.22272135317325592, 0.01435281615704298, 0.05502232536673546, -0.4033515453338623, -0.16033098101615906, 0.3848736882209778, 0.1999773532152176, -0.12101370841264725, -0.6013123393058777, 0.06163457781076431, 0.257350891828537, 0.1335512101650238, -0.13102930784225464, 0.11507020145654678, 0.26026254892349243, 0.14775680005550385, -0.2571090757846832, 0.23596125841140747, -0.014066101051867008, 0.29102134704589844, 0.09884369373321533, -0.16290096938610077, 0.08296132832765579, 0.3361522853374481, 0.003927521873265505, -0.24153627455234528, -0.013702915981411934, -0.12546353042125702, 0.28384748101234436, -0.12509246170520782, -0.14821293950080872, -0.029125146567821503, 0.09563394635915756, 0.12166129052639008, 0.4608387053012848, 0.09020771831274033, 0.021487874910235405, -0.06572093069553375, -0.14753466844558716, -0.40121620893478394, -0.06006242334842682, 0.03288824111223221, -0.1326180398464203, 0.014206317253410816, 0.05187305435538292, -0.13590921461582184, 0.3132939636707306, -0.009249288588762283, 0.43694525957107544, 0.12656542658805847, -0.024377290159463882, -0.14826251566410065, 0.4578137695789337, -0.4662647247314453, -0.1687982976436615, 0.28169554471969604, -0.11046410351991653, 0.0842430368065834, -0.11323792487382889, -0.10509437322616577, -0.01785527728497982, 0.17952178418636322, -0.15866203606128693, 0.22237525880336761, -0.2001095563173294, 0.4053359031677246, 0.2907795011997223, -0.03121282532811165, -0.35276666283607483, -0.0743962973356247, 0.18313200771808624, 0.27748629450798035, -0.017893696203827858, 0.38301995396614075, 0.17758134007453918, 0.059883713722229004, -0.1240595281124115, 0.266304075717926, -0.18348103761672974, -0.0043845875188708305, 0.04803823307156563, -0.299056738615036 ]
https://github.com/huggingface/datasets/issues/163
[Feature request] Add cos-e v1.0
@mariamabarham I'm also not sure where to find CQA 1.0. Perhaps it's not possible to include this version of the dataset. I'll close the issue if that's the case.
I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf).
29
[Feature request] Add cos-e v1.0 I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf). @mariamabarham I'm also not sure where to find CQA 1.0. Perhaps it's not possible to include this version of the dataset. I'll close the issue if that's the case.
[ -0.2492516189813614, -0.10364528000354767, -0.13912147283554077, -0.42108508944511414, -0.5307979583740234, -0.06874842196702957, -0.07874950766563416, 0.07906591892242432, -0.2197636514902115, 0.4390557110309601, 0.1513826698064804, 0.29369908571243286, 0.06534161418676376, 0.4375331699848175, 0.00677889259532094, 0.2519572675228119, 0.1544279307126999, 0.3981415331363678, -0.03590266406536102, -0.081110879778862, -0.09899769723415375, 0.3010998070240021, -0.22698070108890533, -0.059495504945516586, 0.06923222541809082, -0.020368196070194244, -0.272861510515213, -0.1656869351863861, -0.15396244823932648, -0.15395841002464294, 0.4113926291465759, 0.3626807630062103, 0.20779551565647125, 0.00431181350722909, -0.00011051368346670642, -0.21097931265830994, 0.2824908196926117, -0.12771229445934296, -0.23003928363323212, -0.12380530685186386, -0.33321258425712585, -0.3251549303531647, 0.008640189655125141, 0.0028277591336518526, 0.08679977804422379, 0.1362750232219696, 0.12899689376354218, -0.10482513159513474, 0.0015904215397313237, 0.1707078069448471, 0.26446297764778137, 0.26653358340263367, -0.005566049832850695, -0.39218857884407043, 0.16724230349063873, 0.27682921290397644, -0.31645846366882324, -0.14915384352207184, 0.3068578541278839, -0.017131920903921127, 0.08296902477741241, 0.5250520706176758, 0.2620766758918762, -0.17759619653224945, -0.12935487926006317, -0.05383484438061714, 0.12355459481477737, -0.11475835740566254, 0.3260292708873749, 0.055493999272584915, 0.5153294801712036, 0.14862225949764252, -0.4204438328742981, 0.383330374956131, -0.001999185187742114, -0.3895682990550995, 0.2576100528240204, -0.05857387185096741, 0.0061521572060883045, 0.05968845635652542, 0.14559179544448853, -0.15127761662006378, -0.3991321623325348, 0.1424819976091385, -0.17090915143489838, 0.031246913596987724, -0.1322280764579773, -0.2196558117866516, 0.08026821166276932, 0.09614326804876328, -0.10330715775489807, 0.20984381437301636, -0.2582561671733856, 0.18841126561164856, -0.277146577835083, -0.2824537456035614, 0.3214116394519806, 0.06597057729959488, 0.4131600856781006, -0.06235172599554062, 0.15197372436523438, -0.09950979053974152, -0.11116881668567657, 0.14640194177627563, 0.05179976299405098, 0.25040239095687866, 0.23542837798595428, -0.05337193235754967, 0.32314926385879517, 0.22540916502475739, 0.4800471067428589, -0.03999696671962738, 0.16129687428474426, 0.04168181121349335, -0.21857859194278717, 0.0124273132532835, -0.026502322405576706, -0.6751198172569275, -0.2794661223888397, 0.0029018851928412914, -0.07580011337995529, 0.059032049030065536, -0.0788477435708046, 0.14038997888565063, 0.3872992992401123, 0.11087072640657425, 0.23839251697063446, -0.03799457475543022, -0.06861971318721771, -0.5039106011390686, -0.3114446997642517, 0.04990089684724808, -0.0885661393404007, 0.2114316076040268, 0.2986612319946289, -0.22972550988197327, 0.306023508310318, -0.3308390974998474, 0.3673436939716339, 0.11446870863437653, -0.20475664734840393, 0.366653710603714, -0.1926843225955963, 0.28552544116973877, 0.06034709885716438, 0.04286782070994377, -0.22038331627845764, 0.35715654492378235, -0.2228282392024994, 0.09571941196918488, -0.26771533489227295, -0.4235020875930786, -0.47084832191467285, 0.19991588592529297, 0.05588282644748688, -0.22937530279159546, 0.04921968653798103, 0.3272184729576111, -0.20584699511528015, -0.17717047035694122, -0.02569678984582424, 0.3373240530490875, -0.10757403075695038, -0.12294215708971024, 0.11131095141172409, 0.18788529932498932, -0.2905886173248291, -0.009798011742532253, -0.36860162019729614, 0.04430435597896576, -0.044148221611976624, 0.04356656223535538, -0.10054685920476913, -0.22311905026435852, -0.11662525683641434, -0.23496823012828827, 0.30609071254730225, -0.37219464778900146, -0.41131791472435, 0.13090968132019043, 0.05861261859536171, -0.34151050448417664, 0.12815269827842712, -0.1533481627702713, 0.01939266361296177, -0.1947709619998932, -0.17797856032848358, 0.1028161495923996, -0.18041373789310455, -0.16533124446868896, -0.02364656887948513, -0.4142698645591736, -0.12499187141656876, 0.2005961686372757, 0.12091860920190811, 0.06461574882268906, 0.27017340064048767, -0.611605703830719, 0.055760789662599564, -0.13917455077171326, 0.11108922213315964, 0.0837540253996849, 0.7464853525161743, -0.046157434582710266, 0.18957947194576263, 0.1458413451910019, -0.31344035267829895, 0.023550964891910553, -0.2664182782173157, 0.1402667611837387, 0.11664498597383499, -0.3408266305923462, -0.09279312938451767, 0.0023917427752166986, 0.06440760940313339, -0.33980387449264526, 0.13861176371574402, 0.09465918689966202, 0.050836775451898575, 0.09796710312366486, 0.12586471438407898, 0.309498131275177, -0.19311794638633728, -0.22526556253433228, 0.00183396867942065, 0.1426393836736679, -0.2699933350086212, -0.10479789227247238, 0.14433225989341736, 0.28665801882743835, 0.025029083713889122, -0.04019142687320709, -0.16687466204166412, 0.2657508850097656, -0.16327407956123352, 0.2737017869949341, 0.0959019660949707, 0.3830801844596863, 0.1436949223279953, -0.14276307821273804, 0.07344518601894379, -0.11707842350006104, -0.12005702406167984, 0.3190229535102844, -0.11595245450735092, 0.3486652076244354, 0.08622628450393677, -0.08114521205425262, -0.03529343381524086, 0.18887202441692352, 0.1800479143857956, -0.03710193186998367, -0.21818657219409943, -0.10062822699546814, 0.194736048579216, 0.0921783521771431, -0.3050861060619354, -0.3667621910572052, -0.24421155452728271, 0.05284825339913368, 0.18153144419193268, -0.008826361037790775, 0.09464158117771149, 0.1603843718767166, -0.038517557084560394, -0.16388599574565887, 0.44971251487731934, -0.09222906827926636, 0.10848819464445114, 0.3049636483192444, -0.20059606432914734, -0.15271058678627014, -0.14874981343746185, 0.04772194102406502, 0.129911407828331, 0.26417914032936096, -0.08511517941951752, -0.049024514853954315, 0.3393491208553314, 0.14000314474105835, -0.3533048927783966, -0.2279835045337677, -0.05839714780449867, 0.028281498700380325, 0.0008938511600717902, -0.09523627907037735, -0.13379181921482086, 0.006519919261336327, 0.03023514710366726, 0.0011714436113834381, 0.13668015599250793, -0.23665054142475128, 0.36741745471954346, 0.1476508527994156, -0.11774155497550964, 0.38227707147598267, -0.007209445349872112, 0.7501001358032227, -0.1117478534579277, 0.30824923515319824, -0.18576043844223022, 0.12192919105291367, -0.42350342869758606, 0.2297057807445526, -0.28927668929100037, -0.12859049439430237, 0.4955849051475525, -0.09794554859399796, 0.4712076187133789, -0.10985393822193146, -0.7647410035133362, 0.06731855869293213, -0.12290865927934647, -0.015936627984046936, 0.014705231413245201, -0.2569582760334015, -0.15914709866046906, -0.01935483328998089, -0.049182143062353134, -0.33155882358551025, -0.4355299472808838, -0.05172494053840637, -0.4801783263683319, -0.005570733454078436, -0.3519708216190338, -0.45951053500175476, -0.021146131679415703, -0.21627868711948395, -0.19869275391101837, -0.21795013546943665, 0.1730872392654419, 0.2513790428638458, 0.025441640987992287, 0.050335269421339035, -0.015436293557286263, 0.007242842111736536, -0.2543783485889435, 0.12977904081344604, 0.49869686365127563, -0.3865445852279663, -0.5030512809753418, 0.2911548614501953, -0.0433313250541687, 0.3502505421638489, -0.006292901001870632, -0.35988497734069824, -0.2641656994819641, 0.000894879107363522, 0.14793294668197632, -0.009468075819313526, 0.03826894611120224, 0.7621442079544067, -0.1158134788274765, -0.19106900691986084, -0.11626727133989334, -0.5392227172851562, 0.1955433040857315, 0.49827268719673157, 0.24280840158462524, -0.27890607714653015, 0.38106828927993774, 0.20991535484790802, 0.3193809688091278, 0.35116782784461975, -0.056519344449043274, 0.121635802090168, 0.16339349746704102, 0.14771686494350433, -0.19633907079696655, -0.0166929941624403, 0.14289222657680511, 0.07544197142124176, 0.14860881865024567, 0.5334338545799255, 0.12728312611579895, -0.37974217534065247, -0.2531403601169586, -0.2787436842918396, -0.05657949671149254, 0.052457790821790695, -0.2920553982257843, 0.01573851704597473, 0.5925151109695435, -0.027571244165301323, -0.16437633335590363, -0.29470551013946533, -0.024257773533463478, 0.18267162144184113, 0.4032699763774872, 0.22562946379184723, -0.011568904854357243, 0.0843055248260498, -0.07599929720163345, -0.3604990839958191, 0.38195058703422546, 0.05675142630934715, 0.10147171467542648, -0.1281963735818863, 0.061538659036159515, 0.16979165375232697, 0.01039549708366394, 0.23315414786338806, -0.19572308659553528, -0.270768940448761, -0.09753907471895218, -0.03323459252715111, 0.20889633893966675, -0.07565899193286896, -0.060406189411878586, -0.17605648934841156, -0.058455243706703186, -0.17866380512714386, 0.03458090126514435, -0.09470604360103607, 0.5656786561012268, -0.2811230421066284, -0.045757029205560684, -0.09696969389915466, -0.017560169100761414, -0.2522434890270233, -0.05329679697751999, 0.018999850377440453, -0.14617544412612915, -0.029697058722376823, -0.0632442981004715, -0.20307795703411102, 0.3560209274291992, -0.1344764530658722, 0.005193919874727726, 0.20223873853683472, 0.01302633062005043, 0.05348793789744377, 0.06713123619556427, 0.0382126085460186, 0.031946904957294464, 0.20771455764770508, 0.022880936041474342, 0.3638216555118561, -0.14250902831554413, 0.2619233727455139, -0.25988489389419556, 0.2579028606414795, 0.17956480383872986, -0.053154438734054565, 0.0472467802464962, -0.20904166996479034, -0.23196910321712494, -0.11197536438703537, -0.26540473103523254, 0.2837344706058502, 0.03603328391909599, -0.14099210500717163, 0.12018943578004837, 0.24428658187389374, 0.02649674564599991, -0.24268624186515808, 0.23383928835391998, 0.3414635956287384, -0.1066095381975174, 0.31874993443489075, 0.01441032811999321, 0.7144328355789185, 0.28476572036743164, -0.15132971107959747, 0.1777026653289795, -0.09784656763076782, 0.28467857837677, -0.08520686626434326, 0.03084232099354267, 0.14362390339374542, -0.029414216056466103, -0.1080329418182373, 0.2240849882364273, 0.47330933809280396, -0.18071278929710388, -0.4074779152870178, 0.23934665322303772, 0.37372520565986633, 0.25265422463417053, 0.1791730523109436, 0.02582426369190216, 0.27924439311027527, -0.19411848485469818, 0.2228650599718094, 0.16126234829425812, 0.11865916848182678, 0.24784338474273682, -0.007930280640721321, -0.1217644065618515, -0.10707350075244904, -0.33475080132484436, -0.2222694456577301, -0.1555018424987793, 0.17588740587234497, -0.137101411819458, 0.12958714365959167, -0.1396809220314026, 0.5002398490905762, -0.21630369126796722, 0.3512283265590668, 0.07694292813539505, -0.39390143752098083, 0.1984987109899521, 0.12374652922153473, -0.0937175303697586, -0.04847005009651184, 0.011539527215063572, -0.09306968748569489, -0.3150055408477783, -0.040458619594573975, 0.2745906412601471, 0.13504855334758759, -0.22948993742465973, -0.1504879891872406, -0.15534131228923798, -0.051747456192970276, 0.19557838141918182, 0.3932061791419983, 0.4967651069164276, 0.20558060705661774, -0.2776391804218292, 0.21186205744743347, 0.1932184398174286, -0.09961529821157455, -0.09029276669025421, 0.0016793938120827079, 0.05238150805234909, -0.2097775936126709, 0.20240309834480286, 0.3447614908218384, -0.20526725053787231, 0.11226953566074371, -0.0696331337094307, -0.07564307749271393, -0.23330862820148468, -0.3203873038291931, -0.01394443865865469, -0.5064805150032043, -0.2149946093559265, 0.04626059904694557, -0.2646351456642151, 0.16258898377418518, 0.9665587544441223, 0.019335635006427765, 0.08566265553236008, -0.15413005650043488, -0.12778082489967346, -0.27205607295036316, 0.09145724773406982, -0.6348618268966675, 0.3025995194911957, -0.09700216352939606, -0.010685326531529427, -0.058876629918813705, 0.2275671660900116, -0.400564581155777, -0.23105911910533905, 0.08580867946147919, -0.07548965513706207, 0.17271439731121063, 0.18505443632602692, 0.3002167344093323, 0.20005688071250916, 0.10802128911018372, -0.03642868623137474, -0.2978534400463104, -0.14092525839805603, 0.06706120818853378, 0.09022767096757889, 0.08517438173294067, 0.26510581374168396, 0.015573768876492977, -0.09612809121608734, -0.36367708444595337, 0.14022478461265564, 0.3158975839614868, -0.2554153501987457, -0.08340796083211899, 0.2138916701078415, -0.2956834137439728, -0.1531856507062912, 0.07409316301345825, 0.268174946308136, 0.20325881242752075, 0.0001839410251704976, 0.02265312150120735, 0.4017670154571533, -0.07402633875608444, -0.12795953452587128, 0.0669444352388382, 0.3990156352519989, -0.32437655329704285, 0.23266425728797913, 0.057870764285326004, 0.021776793524622917, -0.07467729598283768, -0.02741868793964386, -0.11537857353687286, -0.5985838770866394, -0.052579473704099655, 0.08984797447919846, 0.027268588542938232, 0.27411359548568726, -0.40830352902412415, 0.10977049916982651, -0.23368966579437256, -0.06540721654891968, 0.17333342134952545, 0.12680703401565552, 0.5387253761291504, -0.043540939688682556, 0.9180060029029846, 0.10789868235588074, -0.07989372313022614, -0.07315359264612198, 0.3522467315196991, 0.09844809770584106, 0.19314135611057281, 0.20261429250240326, 0.16971656680107117, 0.017392121255397797, 0.08312444388866425, -0.06407205015420914, 0.010315518826246262, -0.09518510848283768, 0.3211268186569214, 0.4013405740261078, 0.15263168513774872, -0.3137335777282715, 0.03715796023607254, 0.16140033304691315, 0.19644980132579803, 0.016423862427473068, 0.21400709450244904, -0.09558103233575821, 0.1256728321313858, 0.11404736340045929, -0.85201096534729, -0.14799995720386505, -0.00028107588877901435, -0.15192845463752747, -0.1547405868768692, -0.1215369701385498, -0.2612948417663574, -0.10409160703420639, 0.29159030318260193, 0.10824879258871078, 0.29016774892807007, -0.29568856954574585, -0.08855556696653366, -0.07152863591909409, -0.0806170254945755, 0.14213530719280243, 0.14287453889846802, 0.08948025852441788, 0.5019296407699585, -0.49050262570381165, -0.25601011514663696, 0.30503714084625244, 0.304168164730072, 0.16102759540081024, -0.2079572081565857, 0.28158438205718994, 0.18219377100467682, 0.019643358886241913, -0.2664215564727783, 0.16778196394443512, 0.2630542814731598, -0.0011913884663954377, 0.12151792645454407, 0.10152985155582428, 0.11109867691993713, -0.06895548105239868, 0.019005026668310165, -0.11042580008506775, -0.00463815126568079, -0.2969081997871399, 0.1431904137134552, -0.07266102731227875, 0.12516352534294128, -0.08303342759609222, 0.24224282801151276, 0.06060127913951874, -0.33828580379486084, -0.23554345965385437, -0.14563855528831482, -0.21853677928447723, -0.14484553039073944, 0.11242623627185822, -0.2741735577583313, 0.020280664786696434, 0.2404281049966812, 0.25404441356658936, -0.026827899739146233, -0.5908099412918091, 0.002973642898723483, 0.2741870582103729, 0.13339398801326752, 0.03358670696616173, 0.0834173858165741, 0.25750473141670227, 0.20823340117931366, -0.26136985421180725, 0.29738911986351013, -0.14023014903068542, 0.3001265823841095, -0.008812443353235722, -0.3560074269771576, 0.06686946749687195, 0.26842185854911804, -0.008376084268093109, -0.18063199520111084, -0.04371901601552963, -0.11608517915010452, 0.2443632036447525, 0.006212277337908745, -0.218147411942482, 0.014315287582576275, 0.08671717345714569, 0.04670054465532303, 0.47451063990592957, 0.20910735428333282, 0.12969060242176056, -0.14030103385448456, -0.13858993351459503, -0.41770780086517334, -0.09693184494972229, 0.04019993171095848, -0.16448217630386353, 0.07427629828453064, 0.2251124083995819, -0.042595114558935165, 0.1396811455488205, -0.003688121447339654, 0.4698244631290436, 0.07916274666786194, -0.08481202274560928, -0.12159710377454758, 0.5496423244476318, -0.4163469076156616, -0.1756494641304016, 0.12344039976596832, -0.13445429503917694, 0.16413356363773346, -0.18473394215106964, -0.14443282783031464, -0.1051093265414238, 0.27850866317749023, -0.22749821841716766, 0.0910431295633316, -0.08435239642858505, 0.31331154704093933, 0.11957185715436935, -0.010894750244915485, -0.338371604681015, 0.041730888187885284, 0.11309714615345001, 0.2314244955778122, -0.08588790893554688, 0.44417452812194824, 0.10862798988819122, 0.01142162550240755, -0.14661464095115662, 0.36385518312454224, -0.15389685332775116, -0.04098968580365181, 0.043749526143074036, -0.31981801986694336 ]
https://github.com/huggingface/datasets/issues/163
[Feature request] Add cos-e v1.0
You can now do ```python from nlp import load_dataset cos_e = load_dataset("cos_e", "v1.0") ``` Thanks @mariamabarham !
I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf).
17
[Feature request] Add cos-e v1.0 I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf). You can now do ```python from nlp import load_dataset cos_e = load_dataset("cos_e", "v1.0") ``` Thanks @mariamabarham !
[ -0.25247398018836975, -0.26478496193885803, -0.11422256380319595, -0.4864543080329895, -0.3794749677181244, -0.20802795886993408, 0.08028761297464371, 0.18926985561847687, 0.1177971139550209, 0.3005789816379547, -0.15577569603919983, 0.5226975083351135, -0.14769305288791656, 0.4125772714614868, 0.14620853960514069, 0.10172688215970993, 0.03274611011147499, 0.36265653371810913, -0.14609380066394806, -0.2017548680305481, -0.24998566508293152, 0.22434002161026, -0.17047907412052155, -0.054653652012348175, 0.21928025782108307, 0.0769510492682457, -0.29594308137893677, 0.07935701310634613, -0.06169765815138817, -0.19533933699131012, 0.3291562795639038, 0.32923299074172974, 0.18306899070739746, 0.06467704474925995, -0.00010421490151202306, -0.23515966534614563, 0.25748446583747864, -0.1284157782793045, -0.2869001626968384, -0.18150916695594788, -0.18946315348148346, -0.4737761914730072, 0.22126206755638123, -0.21765995025634766, 0.10919737070798874, 0.12459331750869751, 0.20315469801425934, -0.12437333911657333, 0.12705610692501068, 0.2281513214111328, 0.32328739762306213, 0.40828630328178406, -0.034266382455825806, -0.18497438728809357, 0.2077932059764862, 0.27063989639282227, -0.2086317539215088, -0.012949217110872269, 0.15916523337364197, -0.2754570543766022, -0.19682717323303223, 0.4057950973510742, 0.1401132345199585, -0.09584784507751465, -0.1321590691804886, -0.06573233753442764, 0.1608225554227829, 0.021586479619145393, 0.010287469252943993, 0.03543566167354584, 0.2350434958934784, -0.049714911729097366, -0.3959870934486389, 0.14017361402511597, 0.028912648558616638, -0.566857099533081, 0.0933171734213829, -0.08714453876018524, -0.12451322376728058, -0.14305652678012848, 0.2074774205684662, -0.1623462736606598, -0.4726058840751648, 0.2572449743747711, 0.004829579032957554, 0.29475754499435425, -0.06364818662405014, -0.20170080661773682, 0.22664408385753632, 0.04250159114599228, -0.07998202741146088, 0.15150269865989685, -0.03282922878861427, 0.3442158102989197, -0.5325772762298584, -0.2802599370479584, 0.3405757248401642, 0.09036599844694138, 0.2678714990615845, 0.04260275885462761, 0.29281696677207947, 0.1416623890399933, -0.003866570768877864, 0.19317752122879028, 0.030217958614230156, 0.24012543261051178, 0.29266491532325745, -0.04129928722977638, 0.42817920446395874, 0.2653527557849884, 0.3137468099594116, 0.0563645176589489, 0.07044044137001038, -0.10898221284151077, -0.14500035345554352, -0.06849303096532822, 0.008153818547725677, -0.46858859062194824, -0.26381605863571167, -0.05357560142874718, -0.055846378207206726, 0.1455134153366089, -0.013917938806116581, 0.27329736948013306, 0.17012718319892883, 0.24480260908603668, 0.43004289269447327, 0.009215160273015499, -0.2248598337173462, -0.39519909024238586, -0.29168084263801575, 0.09738122671842575, -0.3190461993217468, 0.08400393277406693, 0.3731900751590729, -0.20060047507286072, 0.27423202991485596, -0.4334008991718292, 0.25433459877967834, 0.07114213705062866, -0.19919119775295258, 0.2366589456796646, -0.11447160691022873, 0.061436738818883896, 0.22511300444602966, -0.042978931218385696, -0.23836006224155426, 0.20623290538787842, -0.2083875685930252, -0.08082154393196106, -0.31519249081611633, -0.3829510509967804, -0.4068826138973236, 0.2756212055683136, -0.05516820028424263, -0.31489723920822144, -0.13434965908527374, 0.32343706488609314, -0.1739307940006256, -0.18654800951480865, -0.036193445324897766, 0.13921807706356049, -0.254175066947937, -0.12859705090522766, 0.0910562127828598, 0.1125255823135376, 0.054605964571237564, -0.13686439394950867, -0.5214043855667114, 0.001388277392834425, -0.010002242401242256, 0.0492771752178669, 0.021899739280343056, -0.12459694594144821, -0.130900040268898, -0.13090990483760834, 0.579158365726471, -0.36789393424987793, -0.06391055136919022, 0.12525269389152527, 0.053750764578580856, -0.2881128489971161, 0.09406495094299316, -0.1714257001876831, -0.14280198514461517, -0.10669828951358795, -0.10121865570545197, 0.341888964176178, -0.07049576938152313, 0.09957355260848999, -0.11652079224586487, -0.3035581409931183, -0.012261444702744484, 0.30136382579803467, 0.2912398874759674, -0.08109161257743835, 0.23300935328006744, -0.15536558628082275, 0.20717039704322815, -0.27829286456108093, 0.034141380339860916, 0.054512809962034225, 0.5859819054603577, 0.046290457248687744, 0.09907370060682297, -0.05174309015274048, -0.31716135144233704, 0.016678931191563606, -0.11427254229784012, 0.29519733786582947, 0.11283326894044876, -0.2925978899002075, -0.03222913667559624, 0.05407179519534111, 0.11803635209798813, -0.1114928349852562, 0.26141393184661865, 0.11125155538320541, 0.023204412311315536, 0.07920379191637039, 0.07408259063959122, 0.3704972267150879, -0.1423625648021698, -0.1394183337688446, -0.25196871161460876, 0.19473204016685486, -0.21140143275260925, -0.09095337241888046, 0.0903235375881195, 0.3076283633708954, -0.014147050678730011, -0.01649150252342224, -0.09004524350166321, 0.34207871556282043, -0.19772452116012573, 0.3473967909812927, 0.06613938510417938, 0.11991211026906967, 0.03165224939584732, -0.18702678382396698, 0.10462124645709991, 0.020478153601288795, 0.10501959919929504, 0.19834351539611816, -0.18742139637470245, 0.47520485520362854, 0.0901666134595871, -0.05945897474884987, 0.09390728175640106, 0.032953985035419464, 0.15092258155345917, -0.2320464551448822, -0.30256155133247375, 0.012772483751177788, 0.38707539439201355, 0.11589349806308746, -0.3680187463760376, -0.3732738792896271, -0.37219536304473877, -0.06634058803319931, 0.15661004185676575, 0.005696177948266268, 0.21773922443389893, 0.1688700169324875, -0.022963572293519974, -0.24440841376781464, 0.23325835168361664, -0.09595482796430588, 0.3129545748233795, 0.3339042067527771, -0.08845458924770355, 0.03661453351378441, -0.34952232241630554, 0.03818830847740173, 0.184769406914711, 0.21045885980129242, -0.0037195715121924877, 0.18175502121448517, 0.4397560954093933, -0.033547479659318924, -0.48965147137641907, -0.41225677728652954, -0.22404542565345764, 0.04827684164047241, -0.24939976632595062, -0.012750890105962753, -0.06169255077838898, -0.14514486491680145, -0.05505021661520004, 0.09441985934972763, 0.09586106240749359, -0.3651609718799591, 0.2836422026157379, 0.33103182911872864, -0.2519615590572357, 0.19968418776988983, -0.01126991119235754, 0.6269267201423645, -0.04075019434094429, 0.1283239722251892, -0.23835818469524384, -0.0503174364566803, -0.24117912352085114, 0.2618352472782135, -0.24112769961357117, 0.031092504039406776, 0.48307761549949646, -0.18815037608146667, 0.2633422017097473, 0.05125856027007103, -0.6950429081916809, 0.019227860495448112, -0.12177242338657379, -0.06724274903535843, 0.10748764872550964, -0.3113570511341095, -0.10114439576864243, -0.06949933618307114, 0.043600648641586304, -0.6364915370941162, -0.34482088685035706, -0.013514435850083828, -0.23367244005203247, 0.1660439521074295, -0.40941816568374634, -0.43483325839042664, -0.10463457554578781, -0.3718425929546356, -0.12796851992607117, 0.09901633113622665, 0.18222947418689728, 0.39012715220451355, -0.00585034629330039, 0.1749870777130127, 0.06103974208235741, 0.05882646143436432, -0.3134269416332245, 0.17933981120586395, 0.42067235708236694, -0.40305981040000916, -0.5246385335922241, 0.008747668005526066, -0.09246662259101868, 0.32258492708206177, 0.09169090539216995, -0.11638464778661728, -0.3931528329849243, 0.1710185706615448, 0.03368848189711571, -0.1324879676103592, 0.27087709307670593, 0.4044109880924225, -0.032161418348550797, -0.21041612327098846, -0.1765943318605423, -0.24583227932453156, 0.150851771235466, 0.45187434554100037, 0.13134096562862396, -0.24895334243774414, 0.2793436050415039, 0.08622337877750397, 0.11332707852125168, 0.10454542934894562, -0.16755050420761108, 0.2823273837566376, -0.027697859331965446, 0.20288503170013428, -0.001617696601897478, -0.0004303008900023997, 0.07328096032142639, 0.0729726180434227, 0.22535473108291626, 0.5441485643386841, 0.14013390243053436, -0.4167070686817169, -0.3592554032802582, -0.30449479818344116, -0.03829605132341385, -0.0881943330168724, -0.05756612494587898, 0.2055097222328186, 0.35951921343803406, -0.028068702667951584, -0.1992827206850052, -0.4128100872039795, -0.20246325433254242, -0.012766615487635136, 0.23737002909183502, 0.004086571745574474, -0.06508563458919525, 0.09634526818990707, 0.0386570505797863, -0.49792617559432983, 0.38196200132369995, 0.11513090133666992, 0.07316866517066956, 0.038020066916942596, -0.08453972637653351, 0.06384678930044174, -0.10432208329439163, 0.187059685587883, -0.1114230751991272, 0.00545691279694438, -0.056445103138685226, -0.10295866429805756, -0.17706327140331268, 0.056419216096401215, -0.37846603989601135, -0.19693288207054138, 0.07580847293138504, -0.009619825519621372, -0.000914979085791856, -0.17845605313777924, 0.6600252985954285, -0.08007165789604187, -0.05716071277856827, -0.1971150040626526, -0.012537197209894657, -0.4126780927181244, -0.07509056478738785, -0.0036728112027049065, -0.15633101761341095, 0.21791595220565796, -0.120223268866539, -0.08209189772605896, 0.17336593568325043, -0.01764693483710289, 0.11866375803947449, 0.28284594416618347, -0.07113201171159744, 0.12168481945991516, -0.015217212028801441, -0.025802413001656532, 0.07832588255405426, 0.11322911083698273, 0.11903868615627289, 0.1873854249715805, -0.06267848610877991, 0.050075940787792206, -0.05787036567926407, 0.29155245423316956, 0.18741922080516815, 0.09294907748699188, 0.022943198680877686, -0.3587075471878052, -0.09099864214658737, -0.08881648629903793, -0.006467881612479687, 0.28515705466270447, 0.23027606308460236, -0.2992841899394989, 0.01836075261235237, 0.40848487615585327, 0.11278720200061798, -0.22551417350769043, 0.13825827836990356, 0.035436227917671204, -0.43261751532554626, 0.5216683745384216, -0.047607678920030594, 0.6814635396003723, 0.2664853632450104, 0.06667958199977875, 0.4089909791946411, 0.019803693518042564, 0.619162380695343, -0.12821902334690094, 0.09045027941465378, 0.008613633923232555, -0.07493513077497482, -0.028318963944911957, 0.25809162855148315, 0.2802240252494812, -0.1453399807214737, -0.302685409784317, 0.35803529620170593, 0.4741484224796295, -0.02822926826775074, 0.06607978045940399, 0.19065847992897034, 0.26750051975250244, -0.3572767376899719, -0.1759166419506073, 0.19538584351539612, -0.05740028992295265, 0.19830577075481415, -0.009514258243143559, -0.07632477581501007, -0.03352498635649681, -0.34461966156959534, -0.09483549743890762, -0.004619645420461893, 0.40876227617263794, 0.06272813677787781, -0.013135116547346115, 0.00018215621821582317, 0.35457536578178406, -0.251066654920578, 0.2068144977092743, 0.12703579664230347, -0.23320062458515167, 0.13416174054145813, 0.03041515126824379, 0.033476002514362335, -0.057564567774534225, 0.06580711156129837, 0.17728327214717865, -0.21683046221733093, -0.09247511625289917, 0.13309289515018463, 0.2801993787288666, -0.22764916718006134, 0.07511981576681137, -0.059924524277448654, -0.09679807722568512, 0.09913405030965805, 0.27097466588020325, 0.5199604630470276, 0.04232712835073471, -0.2932431995868683, 0.24941058456897736, 0.21730127930641174, -0.24735358357429504, 0.1259591430425644, 0.14907799661159515, -0.1499183624982834, -0.12614040076732635, 0.4295511841773987, 0.32879021763801575, -0.17596828937530518, 0.16236048936843872, 0.028476493433117867, -0.0659819096326828, -0.265457421541214, -0.2450835257768631, -0.0956837460398674, -0.5750576257705688, -0.06528180837631226, 0.11142838001251221, -0.18903690576553345, 0.20219182968139648, 0.7269641757011414, -0.03920890390872955, 0.20746299624443054, -0.10957187414169312, -0.25682294368743896, -0.27607446908950806, 0.05067749693989754, -0.45215514302253723, 0.22447267174720764, 0.1467542052268982, 0.0574888214468956, -0.057873159646987915, 0.08658470958471298, -0.4668601453304291, -0.2237444818019867, 0.07354486733675003, -0.011188284493982792, 0.3666559159755707, 0.08408203721046448, 0.3317530155181885, 0.07002545893192291, 0.25914931297302246, 0.017092570662498474, -0.18132854998111725, -0.271454393863678, -0.12523701786994934, 0.07885979115962982, 0.0218804944306612, 0.17113493382930756, 0.09806948155164719, -0.05675014480948448, -0.27304205298423767, -0.05970102921128273, 0.18825013935565948, -0.08539971709251404, -0.186428040266037, 0.3759683072566986, -0.20979605615139008, -0.13131263852119446, 0.37401965260505676, 0.2997761368751526, 0.20191803574562073, 0.013176705688238144, -0.10522761195898056, 0.4919644594192505, -0.06271237879991531, -0.10722273588180542, 0.04128710925579071, 0.1994611918926239, -0.2113649994134903, 0.1068020910024643, 0.08347363770008087, 0.0005003410624340177, -0.2596774399280548, -0.20306313037872314, 0.21257956326007843, -0.2958906888961792, 0.01585974171757698, 0.0698932483792305, 0.07521279156208038, 0.3328961133956909, -0.4336082637310028, 0.15516607463359833, -0.0699423998594284, -0.15005697309970856, 0.38073331117630005, 0.1659955531358719, 0.25933748483657837, -0.14491328597068787, 0.6386485695838928, -0.011185668408870697, -0.019398875534534454, 0.06352314352989197, 0.2700660526752472, 0.007332291454076767, 0.08318625390529633, 0.007491759955883026, 0.06729592382907867, 0.002586598973721266, 0.2614637017250061, -0.02446972019970417, 0.043506935238838196, -0.026396824046969414, 0.19555123150348663, 0.3980354964733124, -0.004165648482739925, -0.2185801863670349, -0.040635257959365845, 0.26924821734428406, 0.13353000581264496, 0.22033332288265228, 0.12551991641521454, 0.013996127061545849, -0.026474693790078163, 0.07339467853307724, -0.6518133282661438, 0.10224466770887375, 0.04633386433124542, -0.2674895226955414, -0.19240356981754303, -0.0601472444832325, -0.528367817401886, -0.03477101773023605, 0.16022349894046783, -0.09570533782243729, 0.11657224595546722, -0.05166484788060188, 0.07509500533342361, -0.1783222109079361, -0.0644860491156578, 0.15219172835350037, 0.2097458839416504, -0.07242745906114578, 0.26005929708480835, -0.31430912017822266, -0.13356681168079376, 0.11316905170679092, 0.3317929804325104, 0.25361526012420654, -0.07983961701393127, 0.25138047337532043, 0.1893158107995987, 0.07819001376628876, -0.06341101974248886, 0.015541641972959042, 0.12749674916267395, 0.17518645524978638, 0.15679284930229187, 0.34565579891204834, 0.19967445731163025, -0.2353813499212265, 0.10903971642255783, 0.018087217584252357, -0.18517619371414185, -0.18787050247192383, 0.29476118087768555, 0.2506920397281647, 0.12793835997581482, -0.1549277901649475, 0.0566546805202961, -0.22269326448440552, -0.3151549994945526, -0.05655500292778015, -0.04162653163075447, -0.06579119712114334, -0.2692042589187622, 0.14210399985313416, -0.1014510914683342, 0.24289904534816742, 0.2020774781703949, 0.30529677867889404, 0.07524054497480392, -0.4313351809978485, -0.16718439757823944, 0.21204787492752075, 0.1855374425649643, -0.2335086315870285, -0.04269825667142868, 0.1669415384531021, 0.14643697440624237, -0.24126192927360535, 0.10201074182987213, -0.13331803679466248, 0.42490431666374207, -0.0015242679510265589, -0.3902808725833893, 0.15905426442623138, 0.14354091882705688, -0.09171497821807861, -0.03733183443546295, -0.25922462344169617, -0.3049333691596985, -0.1383424550294876, 0.13726076483726501, -0.2919713854789734, 0.09674257040023804, 0.00660020811483264, 0.1337219476699829, 0.4191931188106537, 0.029712989926338196, 0.277936726808548, -0.33394208550453186, -0.006887573748826981, -0.3844675123691559, -0.25012120604515076, -0.06227720156311989, 0.046573709696531296, 0.09622674435377121, 0.17797420918941498, 0.020027566701173782, 0.18667908012866974, 0.049494922161102295, 0.39270541071891785, 0.03392958268523216, -0.2314748913049698, -0.042331065982580185, 0.5437834858894348, -0.3518790602684021, -0.2071860283613205, -0.07825611531734467, -0.0048724813386797905, 0.03916778787970543, -0.13390249013900757, -0.05519828200340271, -0.30653682351112366, 0.32572460174560547, -0.33518847823143005, -0.05862820893526077, -0.07195737957954407, 0.34767183661460876, 0.15555158257484436, -0.006517232861369848, -0.33956560492515564, 0.16316482424736023, 0.14248700439929962, 0.16235244274139404, -0.03403221815824509, 0.39861759543418884, -0.14811287820339203, -0.012055295519530773, -0.21766582131385803, -0.168108731508255, -0.06432762742042542, -0.08481419086456299, 0.13156428933143616, -0.4343090057373047 ]
https://github.com/huggingface/datasets/issues/163
[Feature request] Add cos-e v1.0
@mariamabarham Just wanted to note that default behavior `cos_e = load_dataset("cos_e")` now loads `v1.0`. Not sure if this is intentional (but the flag specification does work as intended).
I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf).
28
[Feature request] Add cos-e v1.0 I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf). @mariamabarham Just wanted to note that default behavior `cos_e = load_dataset("cos_e")` now loads `v1.0`. Not sure if this is intentional (but the flag specification does work as intended).
[ -0.2263442426919937, -0.2859196066856384, -0.11841704696416855, -0.49191954731941223, -0.45303866267204285, -0.25085997581481934, 0.2163064181804657, 0.22147080302238464, -0.03563763201236725, 0.3164638578891754, 0.07013238221406937, 0.4571443498134613, 0.03245658800005913, 0.35331857204437256, -0.16244779527187347, 0.4214378297328949, 0.11718586832284927, 0.2206687182188034, -0.03011218272149563, -0.2556172013282776, -0.2271457016468048, 0.1309097409248352, -0.16859953105449677, -0.12527354061603546, 0.16298657655715942, 0.28877493739128113, -0.25442370772361755, -0.09574659913778305, -0.04146132990717888, -0.18337279558181763, 0.402875691652298, 0.34786275029182434, 0.17024455964565277, 0.064628466963768, -0.00011004394036717713, -0.13638851046562195, 0.45679572224617004, -0.06395550072193146, -0.27537062764167786, -0.15361222624778748, -0.3556925356388092, -0.3650524318218231, 0.07075220346450806, -0.07159361988306046, 0.21727699041366577, 0.1627269983291626, 0.14945967495441437, -0.031101441010832787, 0.04020896553993225, 0.16627268493175507, 0.2691469192504883, 0.2860795557498932, -0.0316774845123291, -0.3328508138656616, 0.21256165206432343, 0.47362658381462097, -0.3600449562072754, -0.0748889297246933, 0.043695494532585144, -0.097093366086483, -0.1418289691209793, 0.3820017874240875, 0.2254558503627777, -0.17204174399375916, -0.26014792919158936, -0.030674099922180176, 0.2146509438753128, 0.13055475056171417, 0.26056668162345886, 0.056903060525655746, 0.3529522120952606, 0.13864944875240326, -0.26492413878440857, 0.25002145767211914, -0.05883856490254402, -0.42920050024986267, 0.22503916919231415, -0.16231626272201538, 0.021105149760842323, 0.004745615646243095, 0.354120135307312, -0.18666183948516846, -0.5334498286247253, 0.1561819165945053, -0.2166714072227478, 0.17142988741397858, -0.21249546110630035, -0.14087404310703278, 0.09422025084495544, 0.10761158168315887, -0.0929834172129631, 0.11195201426744461, -0.20708131790161133, 0.1804274022579193, -0.3474440574645996, -0.10264940559864044, 0.3085196614265442, -0.06392929702997208, 0.42677024006843567, 0.15892153978347778, 0.1950724571943283, 0.1905687004327774, -0.015698445960879326, 0.15070229768753052, 0.011564020998775959, 0.368662565946579, 0.5664445757865906, 0.06401573121547699, 0.40305113792419434, 0.14254805445671082, 0.40475764870643616, 0.016548268496990204, 0.0781296044588089, 0.021182309836149216, -0.002248266479000449, -0.1615668088197708, 0.11037484556436539, -0.5184798836708069, -0.18017295002937317, -0.03196050599217415, -0.10153331607580185, 0.09271857142448425, 0.0045884898863732815, 0.2791188657283783, 0.2888183891773224, 0.1953786015510559, 0.32147732377052307, -0.022119669243693352, 0.0167545136064291, -0.28218910098075867, -0.2884853184223175, -0.08126652985811234, -0.2485475093126297, 0.20286834239959717, 0.199064239859581, -0.3987301290035248, 0.20908896625041962, -0.3644558787345886, 0.4094012379646301, -0.06413277983665466, -0.22893200814723969, 0.3046184480190277, -0.035715263336896896, 0.2634488642215729, 0.03881171718239784, -0.017300302162766457, -0.09945616126060486, 0.3986673951148987, -0.16598151624202728, 0.11357633024454117, -0.2175620049238205, -0.48887625336647034, -0.2699544429779053, 0.26509180665016174, -0.18048830330371857, -0.252689927816391, -0.24294327199459076, 0.4173630475997925, -0.2470453530550003, -0.007607806008309126, 0.013684500008821487, 0.14105455577373505, -0.3214110732078552, -0.20329055190086365, -0.08240289986133575, 0.19734472036361694, -0.04360007867217064, -0.08860594034194946, -0.6579606533050537, -0.09929870069026947, 0.015475055202841759, -0.1841048151254654, -0.018392963334918022, -0.3088859021663666, -0.24293731153011322, -0.0975847989320755, 0.4091990888118744, -0.5409092307090759, -0.08889524638652802, 0.3007032573223114, -0.05289744958281517, -0.21085987985134125, 0.19971732795238495, -0.3001004755496979, -0.15993109345436096, -0.2848050594329834, -0.31128448247909546, 0.05204416438937187, -0.03445454686880112, -0.051054853945970535, -0.07111449539661407, -0.34644144773483276, -0.3489781320095062, 0.25857147574424744, 0.3637622892856598, 0.03293349966406822, 0.35678935050964355, -0.41883406043052673, 0.15049850940704346, -0.05560559034347534, 0.13320381939411163, -0.02552761882543564, 0.6704113483428955, 0.10371121764183044, 0.19556772708892822, 0.1381073147058487, -0.3721388280391693, 0.09697646647691727, 0.18270868062973022, 0.15224707126617432, 0.20965860784053802, -0.36246711015701294, -0.15044757723808289, 0.14429223537445068, -0.006205368787050247, -0.19019679725170135, 0.18903659284114838, 0.16745969653129578, -0.06673926115036011, -0.03944779559969902, -0.04032135382294655, 0.4120464026927948, -0.33331143856048584, -0.08935010433197021, 0.16916735470294952, 0.06708600372076035, -0.05908951535820961, -0.007216553203761578, -0.052230916917324066, 0.09110058844089508, -0.0032766906078904867, 0.008714437484741211, -0.1518043428659439, 0.3978514075279236, -0.01469998899847269, 0.13733519613742828, 0.07569342106580734, 0.20276600122451782, 0.12991948425769806, -0.11296900361776352, 0.015641065314412117, 0.031187446787953377, -0.060542721301317215, 0.1088653951883316, -0.21376988291740417, 0.36783504486083984, -0.04400163143873215, 0.0786852017045021, -0.0030530672520399094, -0.08954416215419769, 0.026837438344955444, -0.07286764681339264, -0.5226483941078186, -0.2180047631263733, 0.24215637147426605, 0.10695768147706985, -0.4233148992061615, -0.23976370692253113, -0.22046203911304474, -0.08050587028265, 0.1762453019618988, -0.10011542588472366, 0.15004470944404602, 0.3162190020084381, 0.10816868394613266, -0.19115209579467773, 0.27675700187683105, -0.16653181612491608, 0.40634238719940186, 0.31514108180999756, -0.10612604022026062, -0.04802083969116211, -0.3054264187812805, -0.05322883278131485, 0.13648025691509247, 0.21421436965465546, 0.12367841601371765, 0.12573565542697906, 0.4169747829437256, -0.1495237797498703, -0.5176547169685364, -0.41359439492225647, -0.06272361427545547, -0.16659320890903473, -0.19666247069835663, -0.1665612906217575, -0.06399542093276978, 0.09422177076339722, -0.10740800201892853, 0.03863605484366417, 0.08334194123744965, -0.23620034754276276, 0.2636570930480957, 0.22466303408145905, -0.06135547161102295, 0.23576168715953827, 0.02540084719657898, 0.6541305184364319, -0.02252563089132309, 0.13648205995559692, -0.15872788429260254, 0.06659913808107376, -0.4203680157661438, 0.19410860538482666, -0.5060153007507324, -0.22600552439689636, 0.4339205324649811, -0.007591599132865667, 0.30970650911331177, -0.06009937450289726, -0.6933542490005493, 0.07365821301937103, -0.19269125163555145, -0.13785633444786072, 0.06659673154354095, -0.28835171461105347, -0.11153620481491089, 0.0407223105430603, 0.08749260008335114, -0.39442071318626404, -0.3563586473464966, -0.08152005821466446, -0.2302687168121338, 0.27909955382347107, -0.34588760137557983, -0.4851733148097992, 0.23486745357513428, -0.33899667859077454, -0.11091086268424988, -0.0839218869805336, 0.1287860870361328, 0.5446903109550476, 0.13814029097557068, 0.08742589503526688, 0.12513798475265503, -0.014643293805420399, -0.4106576144695282, 0.23893611133098602, 0.3401024341583252, -0.3349623680114746, -0.4000644087791443, 0.10895130783319473, 0.09359079599380493, 0.15444882214069366, 0.2097408026456833, -0.18663319945335388, -0.5588257312774658, 0.18018390238285065, 0.15571722388267517, -0.20453016459941864, 0.1585506945848465, 0.3840799033641815, -0.019267883151769638, -0.1956893652677536, -0.18685121834278107, -0.3143158555030823, 0.22574381530284882, 0.30816155672073364, 0.12987008690834045, -0.16425515711307526, 0.26787930727005005, 0.1611645221710205, 0.14897847175598145, 0.0820319801568985, 0.01911984756588936, 0.050538770854473114, 0.11084679514169693, 0.3510951101779938, -0.025693004950881004, 0.05527964606881142, 0.13333190977573395, 0.08192222565412521, 0.20691217482089996, 0.46736425161361694, 0.1389230191707611, -0.23098979890346527, -0.2874266505241394, -0.25231650471687317, 0.16300874948501587, 0.0316605269908905, -0.16662411391735077, 0.18828554451465607, 0.43356022238731384, -0.129059299826622, -0.09658744931221008, -0.3269830644130707, 0.032860733568668365, 0.06107472628355026, 0.4213995635509491, 0.09682717174291611, -0.17435413599014282, -0.0047747306525707245, 0.14035294950008392, -0.3281846344470978, 0.3598349988460541, 0.1082928329706192, 0.05134522542357445, -0.0035431666765362024, -0.20926575362682343, 0.2115895003080368, -0.029143327847123146, 0.43104201555252075, -0.18089962005615234, 0.01370184775441885, -0.11118075251579285, -0.07789857685565948, -0.09302577376365662, -0.16937553882598877, -0.11786756664514542, -0.3060424029827118, -0.03530928120017052, -0.035561300814151764, -0.05281471461057663, -0.029260344803333282, 0.5762627124786377, -0.2663777470588684, 0.07449111342430115, -0.2874296009540558, -0.20819991827011108, -0.42952269315719604, -0.10088349878787994, 0.004545544274151325, -0.059622831642627716, -0.09537112712860107, -0.22082702815532684, -0.1356155127286911, 0.26978790760040283, -0.2274513989686966, -0.1419636309146881, 0.1268462985754013, -0.25425612926483154, -0.0007775581325404346, -0.033055003732442856, 0.10001509636640549, -0.11615092307329178, 0.3154969811439514, 0.016632255166769028, 0.2869824171066284, -0.37325596809387207, 0.10811769962310791, -0.1803288608789444, 0.2171071171760559, 0.271982878446579, 0.001341125345788896, 0.145433247089386, -0.4411625564098358, -0.23803608119487762, -0.013947062194347382, -0.08445091545581818, 0.21031050384044647, -0.004942504223436117, -0.20481297373771667, -0.04624088108539581, 0.39969974756240845, 0.05420319736003876, -0.2666935324668884, 0.27243226766586304, 0.14732725918293, -0.2883753776550293, 0.4483867883682251, 0.011259361170232296, 0.6613928079605103, 0.2898103594779968, -0.21789176762104034, 0.29653051495552063, -0.21648292243480682, 0.567108690738678, -0.18582585453987122, 0.13965533673763275, 0.09834523499011993, -0.1416414976119995, 0.0241186935454607, 0.26291337609291077, 0.3644726574420929, -0.04198592156171799, -0.32919666171073914, 0.2483590692281723, 0.21675215661525726, 0.3008890748023987, 0.2127337008714676, 0.1163179874420166, 0.08536642044782639, -0.30704644322395325, 0.001339540584012866, 0.18750150501728058, 0.08409218490123749, 0.2702901065349579, -0.04117576405405998, 0.09877122193574905, 0.02850968949496746, -0.37169915437698364, -0.22798755764961243, 0.031631652265787125, 0.4609616696834564, 0.11891832947731018, 0.012707752175629139, 0.0654437392950058, 0.23905444145202637, -0.3236115276813507, 0.2711406648159027, 0.10313165187835693, -0.08830945938825607, 0.29943227767944336, 0.31932541728019714, 0.010496728122234344, -0.17676013708114624, -0.15102766454219818, 0.024321960285305977, -0.15881282091140747, -0.20199954509735107, 0.2272275686264038, 0.17419680953025818, -0.23866282403469086, -0.20172835886478424, 0.033908989280462265, -0.05447155609726906, 0.1423088163137436, 0.5334233641624451, 0.43454304337501526, 0.10620912164449692, -0.17012625932693481, 0.2046644687652588, 0.12692950665950775, -0.11475098878145218, 0.0031190156005322933, 0.1529107391834259, -0.19069130718708038, -0.18325175344944, 0.5322158336639404, 0.279170960187912, -0.13204467296600342, -0.0629272535443306, -0.1157122403383255, 0.016295485198497772, -0.2150593250989914, -0.18904046714305878, 0.14092497527599335, -0.39371028542518616, -0.12774847447872162, 0.16813963651657104, -0.28275325894355774, 0.24938450753688812, 0.8095267415046692, 0.07201117277145386, 0.20189467072486877, -0.22826364636421204, -0.1281670480966568, -0.0254081841558218, -0.07955536991357803, -0.4771961569786072, 0.18120896816253662, -0.02050521969795227, -0.05071505159139633, -0.08789864927530289, -0.012812324799597263, -0.39563411474227905, -0.24205568432807922, 0.19528833031654358, -0.09695474058389664, 0.40214791893959045, 0.21221564710140228, 0.4604627788066864, 0.18588727712631226, 0.21235574781894684, 0.015105808153748512, -0.2088432013988495, -0.1787458062171936, -0.01365662645548582, 0.10726433247327805, 0.08703876286745071, 0.2987448573112488, -0.19903330504894257, -0.18296313285827637, -0.2923521399497986, 0.02228664420545101, 0.33509352803230286, 0.10991968959569931, -0.19674529135227203, 0.08827067166566849, -0.45101508498191833, 0.010249475948512554, 0.34226444363594055, 0.1872560828924179, 0.17638732492923737, 0.09170470386743546, -0.06608226150274277, 0.3762289583683014, -0.1261286735534668, -0.026595599949359894, 0.12284691631793976, 0.34496548771858215, -0.2599087059497833, 0.1292828619480133, 0.008576865307986736, 0.08216625452041626, -0.19888195395469666, -0.1735459417104721, 0.016401223838329315, -0.30185824632644653, 0.006896988023072481, -0.056638166308403015, -0.1672634482383728, 0.303483784198761, -0.23948150873184204, 0.06966451555490494, -0.2762705981731415, -0.16598942875862122, 0.2079402506351471, 0.2772286534309387, 0.25617900490760803, -0.09345296025276184, 0.7590206861495972, 0.1068866029381752, 0.141437366604805, -0.0747293010354042, 0.22079692780971527, 0.1693819910287857, 0.04263382405042648, 0.16868571937084198, 0.27581408619880676, -0.0878944844007492, 0.36049672961235046, -0.07754961401224136, 0.3008498251438141, 0.03680356964468956, 0.2739909291267395, 0.5524520874023438, 0.05506005138158798, -0.31329652667045593, 0.10778699070215225, 0.4623662233352661, 0.05116426572203636, 0.058433838188648224, 0.18636412918567657, -0.22566922008991241, -0.020918238908052444, 0.1821376234292984, -0.7633122205734253, -0.17677178978919983, 0.08755525946617126, -0.27711573243141174, -0.1974257230758667, -0.10568856447935104, -0.47328004240989685, -0.006775857415050268, 0.2994881570339203, 0.050781723111867905, -0.016120750457048416, -0.20815548300743103, 0.008193980902433395, -0.060377661138772964, -0.277809739112854, 0.07377641648054123, 0.2569640576839447, 0.006414429750293493, 0.20321451127529144, -0.4178890585899353, -0.13610689342021942, 0.3770045340061188, 0.13934075832366943, 0.13725177943706512, -0.08849287033081055, 0.5053772926330566, 0.11966467648744583, 0.102590411901474, -0.09160086512565613, -0.009294171817600727, 0.18863524496555328, 0.12328115850687027, 0.1045762225985527, 0.18037132918834686, 0.19206154346466064, -0.18656638264656067, 0.15649458765983582, -0.035235386341810226, -0.24298490583896637, -0.24431394040584564, 0.18434250354766846, 0.10635688155889511, 0.15036258101463318, -0.08765789866447449, 0.22046828269958496, 0.032328732311725616, -0.3462279736995697, -0.062112320214509964, 0.020574338734149933, -0.12764078378677368, -0.14604635536670685, 0.11986365169286728, -0.0651048943400383, -0.12488248944282532, 0.16994702816009521, -0.11498861014842987, 0.0625869557261467, -0.37307971715927124, -0.004312611650675535, 0.3659427762031555, 0.1877307891845703, 0.07066179066896439, 0.05478473752737045, 0.14213654398918152, 0.14274713397026062, -0.25380030274391174, 0.12222804874181747, -0.4326278269290924, 0.5907481908798218, -0.08773498237133026, -0.18280228972434998, 0.14077673852443695, 0.4150138199329376, -0.11227644234895706, -0.1277846395969391, -0.2960698902606964, -0.21165800094604492, 0.06730302423238754, 0.10888557881116867, -0.33292466402053833, -0.12729646265506744, 0.1839389204978943, 0.03788955509662628, 0.44412922859191895, 0.04809069260954857, 0.21306054294109344, -0.12853895127773285, -0.07115848362445831, -0.5014454126358032, -0.13466595113277435, -0.02391618676483631, 0.09259212017059326, -0.13384434580802917, -0.029839111492037773, 0.043161265552043915, 0.11951347440481186, 0.02203463762998581, 0.34258833527565, 0.0461307093501091, -0.09075728058815002, -0.09505613148212433, 0.4244762659072876, -0.5147940516471863, -0.3080167770385742, -0.10602758079767227, -0.1027715876698494, 0.06682370603084564, -0.12948089838027954, -0.04089593514800072, -0.24251917004585266, 0.34958985447883606, -0.22609254717826843, -0.08503848314285278, -0.08611113578081131, 0.4343441128730774, 0.20387002825737, 0.08677276968955994, -0.22049041092395782, 0.09851120412349701, 0.22752772271633148, 0.19771848618984222, -0.27606460452079773, 0.1508464813232422, -0.0448225662112236, 0.01777232438325882, -0.2207440733909607, 0.023583484813570976, -0.17097090184688568, 0.015042563900351524, 0.15334360301494598, -0.3238043189048767 ]
https://github.com/huggingface/datasets/issues/163
[Feature request] Add cos-e v1.0
> @mariamabarham Just wanted to note that default behavior `cos_e = load_dataset("cos_e")` now loads `v1.0`. Not sure if this is intentional (but the flag specification does work as intended). In the new version of `nlp`, if you try `cos_e = load_dataset("cos_e")` it throws this error: ``` ValueError: Config name is missing. Please pick one among the available configs: ['v1.0', 'v1.11'] Example of usage: `load_dataset('cos_e', 'v1.0')` ``` For datasets with at least two configurations, we now force the user to pick one (no default)
I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf).
83
[Feature request] Add cos-e v1.0 I noticed the second release of cos-e (v1.11) is included in this repo. I wanted to request inclusion of v1.0, since this is the version on which results are reported on in [the paper](https://www.aclweb.org/anthology/P19-1487/), and v1.11 has noted [annotation](https://github.com/salesforce/cos-e/issues/2) [issues](https://arxiv.org/pdf/2004.14546.pdf). > @mariamabarham Just wanted to note that default behavior `cos_e = load_dataset("cos_e")` now loads `v1.0`. Not sure if this is intentional (but the flag specification does work as intended). In the new version of `nlp`, if you try `cos_e = load_dataset("cos_e")` it throws this error: ``` ValueError: Config name is missing. Please pick one among the available configs: ['v1.0', 'v1.11'] Example of usage: `load_dataset('cos_e', 'v1.0')` ``` For datasets with at least two configurations, we now force the user to pick one (no default)
[ -0.1603744775056839, -0.19055931270122528, -0.04092991352081299, -0.5251528024673462, -0.5301491618156433, -0.20913541316986084, 0.07917019724845886, 0.3027110993862152, -0.08486749231815338, 0.2993150055408478, 0.0913366749882698, 0.5883418917655945, -0.1273890733718872, 0.3007555603981018, 0.1448276937007904, 0.19229498505592346, 0.049381282180547714, 0.3573561906814575, 0.016091199591755867, -0.1463334560394287, -0.28475192189216614, 0.26689523458480835, -0.15900380909442902, 0.022066157311201096, 0.17621365189552307, 0.18231673538684845, -0.3217964470386505, -0.02632436342537403, -0.06103832647204399, -0.2656818926334381, 0.4338547885417938, 0.2968844175338745, 0.24196332693099976, -0.15443074703216553, -0.00011306983651593328, -0.2487541288137436, 0.5088636875152588, -0.06220341846346855, -0.4297464191913605, -0.30758795142173767, -0.3339642286300659, -0.5184953212738037, 0.16902555525302887, -0.09877857565879822, 0.18584366142749786, 0.15064726769924164, 0.36876365542411804, -0.03254261985421181, -0.0012717362260445952, 0.10816007852554321, 0.22739481925964355, 0.2823913097381592, -0.08409370481967926, -0.2982403039932251, 0.11302653700113297, 0.3750303387641907, -0.30863261222839355, -0.1634404957294464, -0.04228060692548752, -0.2713530957698822, -0.05244768410921097, 0.3673298954963684, 0.2016817331314087, -0.22526875138282776, -0.19585596024990082, -0.04784739390015602, 0.14524711668491364, 0.04428587108850479, 0.078597292304039, 0.09422317892313004, 0.3499737083911896, 0.04661160707473755, -0.2632308602333069, 0.04236535355448723, 0.07684063166379929, -0.5963776111602783, 0.28150567412376404, -0.06955378502607346, -0.17123281955718994, -0.0754082128405571, 0.2659721374511719, -0.2741566598415375, -0.42526867985725403, 0.3804301917552948, -0.060804009437561035, 0.2607443928718567, -0.066587433218956, -0.06787726283073425, 0.23503431677818298, 0.08089528232812881, -0.10468441247940063, 0.02762945368885994, -0.14414776861667633, 0.2545185983181, -0.41611698269844055, -0.14212243258953094, 0.4012798070907593, 0.04142022877931595, 0.2769635021686554, 0.14983630180358887, 0.2790862023830414, 0.012851260602474213, -0.05144159123301506, 0.1853979527950287, 0.08551666140556335, 0.31466537714004517, 0.6022409796714783, 0.028706977143883705, 0.2610609531402588, 0.09194550663232803, 0.5128617882728577, 0.1002449169754982, 0.07537537068128586, 0.0205945186316967, -0.16389109194278717, -0.12788325548171997, 0.11135794967412949, -0.3812931180000305, -0.12265650182962418, -0.09454710781574249, -0.09384182095527649, 0.06300147622823715, 0.02480914629995823, 0.19082999229431152, 0.2591685950756073, 0.2565685510635376, 0.34428519010543823, -0.07276105135679245, -0.13242383301258087, -0.38469743728637695, -0.2491019070148468, -0.05333331599831581, -0.39312615990638733, 0.2702557146549225, 0.3587176501750946, -0.23135268688201904, 0.2805840075016022, -0.40262845158576965, 0.1932351291179657, 0.01676781289279461, -0.2188313752412796, 0.31003543734550476, -0.16564689576625824, 0.07988977432250977, 0.0468822680413723, 0.01372320856899023, -0.30024266242980957, 0.2614302635192871, -0.255144864320755, 0.09527262300252914, -0.2514297068119049, -0.5217928290367126, -0.2898743152618408, 0.21310198307037354, -0.2102159708738327, -0.3397212028503418, -0.16480180621147156, 0.4888763427734375, -0.2243463695049286, -0.2618389427661896, -0.019775765016674995, 0.07405783236026764, -0.37112289667129517, -0.1326874941587448, -0.08441752940416336, 0.1536419838666916, 0.032274965196847916, -0.23951125144958496, -0.658933162689209, 0.023701617494225502, 0.0504046268761158, -0.10603703558444977, -0.10561566799879074, -0.26578083634376526, -0.14965063333511353, -0.059897083789110184, 0.7229757905006409, -0.4373762607574463, -0.11549936234951019, 0.2689365744590759, -0.06760742515325546, -0.20526228845119476, 0.10484437644481659, -0.13712194561958313, -0.23001356422901154, -0.15859846770763397, -0.17310063540935516, 0.2563081979751587, -0.05131072923541069, -0.10487085580825806, -0.08494441956281662, -0.3723750412464142, -0.08402244746685028, 0.2475237250328064, 0.2598857283592224, 0.01023722905665636, 0.265401691198349, -0.17162573337554932, 0.20318888127803802, -0.06941133737564087, 0.14847150444984436, 0.038213811814785004, 0.44097602367401123, 0.042967651039361954, 0.09988891333341599, 0.10250699520111084, -0.5050921440124512, 0.07856296747922897, -0.04806382209062576, 0.2747199833393097, 0.323030561208725, -0.2149769812822342, -0.04981633275747299, -0.05138079449534416, 0.06278447806835175, -0.2429581582546234, 0.12956197559833527, 0.29166939854621887, -0.0453818254172802, -0.02740437723696232, -0.0881027951836586, 0.47716766595840454, -0.21469978988170624, -0.1701604723930359, -0.22500786185264587, 0.07947064191102982, -0.13179494440555573, -0.03842928260564804, 0.00116040650755167, 0.30664336681365967, -0.005750616546720266, 0.03146074712276459, -0.08589591830968857, 0.34104013442993164, -0.17072239518165588, 0.2327122688293457, 0.002011618809774518, 0.051782362163066864, 0.1553289294242859, -0.1710367053747177, 0.07149919122457504, -0.03874818980693817, -0.05914714187383652, 0.17416724562644958, -0.2358752191066742, 0.4087848663330078, 0.05555146560072899, -0.004942735191434622, 0.015664739534258842, -0.02518545836210251, 0.024581462144851685, -0.21304510533809662, -0.5066871643066406, -0.16084495186805725, 0.3907119035720825, 0.09037655591964722, -0.41927701234817505, -0.21827338635921478, -0.4219493567943573, -0.053911201655864716, 0.2659190893173218, 0.010921623557806015, 0.06860537827014923, 0.346120685338974, 0.19898231327533722, -0.2583043575286865, 0.2876759171485901, -0.033756546676158905, 0.5213518738746643, 0.35590964555740356, -0.14374160766601562, 0.05537904053926468, -0.49707189202308655, -0.03433484956622124, 0.19374403357505798, 0.2227170765399933, 0.1649829000234604, 0.03451317176222801, 0.4799373745918274, 0.0247183870524168, -0.4151759445667267, -0.4975750148296356, 0.030082112178206444, -0.0667785033583641, -0.17099332809448242, -0.0767858475446701, -0.23566094040870667, -0.24855449795722961, -0.13437774777412415, 0.15273745357990265, -0.006642815191298723, -0.28227078914642334, 0.2530551254749298, 0.22281822562217712, -0.09684938192367554, 0.2704640328884125, -0.08819800615310669, 0.716515839099884, -0.1732388138771057, 0.1325022578239441, -0.2111666351556778, -0.019733861088752747, -0.4386223554611206, 0.14359359443187714, -0.3673209249973297, -0.12818725407123566, 0.43063703179359436, 0.03370577096939087, 0.24169102311134338, 0.10172466188669205, -0.656621515750885, 0.07370705902576447, -0.16741971671581268, -0.08497527241706848, 0.07531516999006271, -0.38130757212638855, -0.1796533614397049, -0.09716538339853287, 0.09173255413770676, -0.34687861800193787, -0.3567782938480377, -0.06084062159061432, -0.17782935500144958, 0.19364352524280548, -0.38614892959594727, -0.5043296813964844, -0.0028923486825078726, -0.3854028880596161, -0.060332465916872025, -0.0328722707927227, 0.26456305384635925, 0.7287464737892151, -0.029272636398673058, 0.060911331325769424, -0.018814528360962868, 0.05397898703813553, -0.3222590982913971, 0.41111958026885986, 0.24176225066184998, -0.36633065342903137, -0.3256417214870453, -0.09408179670572281, -0.00901124719530344, 0.2759550213813782, 0.13772433996200562, -0.20071959495544434, -0.37605220079421997, 0.1728869378566742, 0.09563282877206802, -0.18955008685588837, 0.1695757359266281, 0.4947325587272644, 0.035685159265995026, -0.14348995685577393, -0.11991380155086517, -0.39781200885772705, 0.22548779845237732, 0.5513760447502136, 0.3134692907333374, -0.09973803907632828, 0.23797841370105743, 0.12890534102916718, 0.05465374514460564, 0.18605907261371613, -0.03154803812503815, 0.213314026594162, 0.0751069188117981, 0.302755743265152, -0.06819268316030502, -0.04883895814418793, 0.1862047165632248, 0.11914422363042831, 0.3289925456047058, 0.4975138008594513, 0.10180965811014175, -0.4926542043685913, -0.3609602153301239, -0.2420576959848404, 0.12460731714963913, 0.0034180390648543835, -0.06890585273504257, 0.060732219368219376, 0.4087851941585541, 0.026040149852633476, -0.2615889310836792, -0.413523405790329, -0.17632927000522614, -0.10904993116855621, 0.4381890296936035, 0.03299254551529884, 0.0128650376573205, -0.08266282081604004, 0.026466401293873787, -0.4243900179862976, 0.2321767359972, 0.06501653790473938, 0.2848561406135559, -0.054201483726501465, -0.13267475366592407, 0.11208350211381912, 0.08530955016613007, 0.26006919145584106, -0.23703275620937347, -0.12139759957790375, -0.011055175215005875, 0.12417043000459671, -0.19625386595726013, -0.06323307752609253, -0.16072867810726166, -0.17356668412685394, 0.07299986481666565, 0.01652752421796322, 0.012415716424584389, -0.18408197164535522, 0.5894098877906799, -0.19198822975158691, 0.027728402987122536, -0.33776766061782837, -0.15445470809936523, -0.2920677959918976, -0.09544720500707626, -0.08875241875648499, -0.18691982328891754, 0.14225344359874725, -0.20191358029842377, -0.0053324648179113865, 0.22868692874908447, -0.09486288577318192, -0.005436966195702553, 0.22884564101696014, -0.24832862615585327, 0.13451334834098816, -0.09508802741765976, 0.017488475888967514, 0.09789527207612991, 0.03977128490805626, -0.04592747613787651, 0.23973777890205383, -0.0904294028878212, 0.18363934755325317, 0.041714467108249664, 0.34280717372894287, 0.2911255657672882, 0.0761275365948677, 0.15024402737617493, -0.38553085923194885, -0.25055670738220215, -0.04244711995124817, -0.08079559355974197, 0.24254734814167023, -0.0665716826915741, -0.3306765854358673, -0.09095665067434311, 0.49124675989151, 0.06609102338552475, -0.33063361048698425, 0.3240831792354584, 0.0861424133181572, -0.3378443121910095, 0.4092397689819336, -0.1289200335741043, 0.7491977214813232, 0.3950461447238922, 0.037740807980298996, 0.32539916038513184, -0.0398075170814991, 0.8740349411964417, -0.03195828199386597, 0.2768111824989319, 0.0714413970708847, 0.052608881145715714, 0.03373265266418457, 0.24986253678798676, 0.37604930996894836, -0.015048308297991753, -0.3114893436431885, 0.3749997019767761, 0.5791009068489075, 0.0683802142739296, 0.1932913362979889, 0.3513813018798828, 0.14927445352077484, -0.4552728235721588, -0.16807572543621063, 0.12077060341835022, -0.10806309431791306, 0.38529008626937866, -0.014131552539765835, -0.006578164175152779, 0.016940301284193993, -0.3202202320098877, -0.2105741649866104, 0.04029235243797302, 0.3929307162761688, 0.05890282616019249, 0.07368302345275879, 0.02825682796537876, 0.4237445890903473, -0.23016357421875, 0.1826665997505188, 0.16094551980495453, -0.20402799546718597, 0.16447234153747559, 0.1395418792963028, -0.018578585237264633, -0.07313785701990128, -0.14397500455379486, 0.08395621180534363, -0.17628887295722961, -0.23792897164821625, 0.10621460527181625, 0.1955350786447525, -0.2047397345304489, -0.11868369579315186, -0.007554013282060623, -0.03843315690755844, 0.2452455461025238, 0.2921878695487976, 0.5004429221153259, 0.10781104117631912, -0.2376679927110672, 0.17232942581176758, 0.09804871678352356, -0.10802887380123138, 0.023980578407645226, 0.20026299357414246, -0.15120668709278107, -0.1128026619553566, 0.5301060676574707, 0.21330399811267853, -0.18939733505249023, 0.10306354612112045, -0.14518451690673828, 0.07328219711780548, -0.19599482417106628, -0.15487858653068542, -0.032985713332891464, -0.5144611597061157, -0.1993071436882019, 0.18461062014102936, -0.23774148523807526, 0.3092065751552582, 0.721386194229126, 0.016554540023207664, 0.2494659274816513, -0.10885801911354065, -0.18847353756427765, -0.053461797535419464, 0.02773416042327881, -0.5827197432518005, 0.16098639369010925, 0.13868652284145355, 0.05973074957728386, -0.03966493904590607, 0.07866760343313217, -0.36422857642173767, -0.2929230332374573, 0.12223733961582184, 0.047586988657712936, 0.5287963151931763, 0.1722332239151001, 0.2284761518239975, 0.08470815420150757, 0.13575172424316406, -0.10030117630958557, -0.18807871639728546, -0.15801309049129486, -0.136435404419899, 0.1190502718091011, 0.1717752367258072, 0.17103160917758942, -0.17962630093097687, -0.19826237857341766, -0.20716483891010284, -0.024119030684232712, 0.1777469515800476, -0.07091853767633438, -0.1936512142419815, 0.33471935987472534, -0.46812644600868225, -0.045132558792829514, 0.38918665051460266, 0.3572051525115967, 0.16981032490730286, 0.04913078248500824, -0.028829241171479225, 0.5353574156761169, -0.03685251250863075, -0.051141660660505295, -0.08615097403526306, 0.24315856397151947, -0.32708391547203064, 0.24757257103919983, -0.09058984369039536, -0.007488209754228592, -0.221811443567276, -0.22819671034812927, -0.03033953160047531, -0.30786868929862976, 0.03573698177933693, -0.04483167827129364, 0.06395334750413895, 0.23791800439357758, -0.3148876130580902, 0.05522339418530464, -0.03594920039176941, -0.029131878167390823, 0.18411999940872192, 0.21877309679985046, 0.32171350717544556, -0.11545950174331665, 0.6415473222732544, -0.04418754577636719, 0.014910978265106678, -0.1816294938325882, 0.2836932837963104, 0.19050993025302887, 0.16893890500068665, 0.08880208432674408, 0.1863638162612915, 0.008465138264000416, 0.3330290913581848, -0.06245894357562065, 0.22023755311965942, -0.04858461022377014, 0.3439468443393707, 0.5547144412994385, 0.07117666304111481, -0.3583298921585083, -0.05430210009217262, 0.3893848955631256, 0.14579084515571594, 0.12391278892755508, 0.1341467946767807, 0.007062557619065046, -0.07100199162960052, 0.18138766288757324, -0.6420035362243652, 0.05338019132614136, 0.03196069598197937, -0.3519986867904663, -0.38255956768989563, -0.09261003136634827, -0.5104483962059021, -0.06215190887451172, 0.14173391461372375, -0.0074940952472388744, 0.032942626625299454, -0.07868805527687073, 0.023907452821731567, -0.20594535768032074, -0.05458845943212509, 0.22706551849842072, 0.16826972365379333, -0.02340839058160782, 0.3689230978488922, -0.5310978293418884, -0.2112865298986435, 0.19069553911685944, 0.24770773947238922, 0.20929516851902008, -0.03900201618671417, 0.405857115983963, 0.11202484369277954, 0.1455620527267456, -0.0707632526755333, -0.05981991067528725, 0.1610332876443863, 0.35448962450027466, 0.04199765995144844, 0.261322557926178, 0.09808734804391861, -0.09979250282049179, 0.14007237553596497, 0.043431371450424194, -0.29908323287963867, -0.21187923848628998, 0.3249712884426117, 0.17535436153411865, 0.18041366338729858, -0.02330038696527481, 0.37575432658195496, -0.13017255067825317, -0.2383153885602951, 0.08112385123968124, 0.10263857990503311, -0.10894843190908432, -0.24176271259784698, 0.09944423288106918, -0.07116492837667465, 0.007067068945616484, 0.18419744074344635, -0.030556920915842056, 0.02113489806652069, -0.41406717896461487, -0.12865127623081207, 0.2505950331687927, 0.11373472213745117, -0.12676385045051575, 0.09774120897054672, 0.15977394580841064, 0.04668646678328514, -0.2836061418056488, -0.12935735285282135, -0.05016298219561577, 0.5626105666160583, -0.019448792561888695, -0.2112596482038498, 0.1510002166032791, 0.35848715901374817, -0.02639128267765045, -0.12816943228244781, -0.1886707842350006, -0.2875412702560425, 0.04389853775501251, -0.00815906748175621, -0.3241661489009857, 0.01152632012963295, 0.04369772970676422, 0.23319746553897858, 0.3404272198677063, 0.10100140422582626, 0.2775696814060211, -0.15534451603889465, -0.08501600474119186, -0.5489777326583862, -0.18481042981147766, -0.0563795380294323, 0.0664726197719574, -0.0956382229924202, 0.052344366908073425, -0.034796543419361115, 0.34384262561798096, -0.010121267288923264, 0.3170662820339203, -0.020813705399632454, -0.09819822013378143, -0.03269500285387039, 0.40111249685287476, -0.47487980127334595, -0.15526598691940308, -0.021512264385819435, -0.01749444007873535, 0.03464525192975998, -0.06058569252490997, -0.07859817147254944, -0.23899789154529572, 0.3525125980377197, -0.16081756353378296, -0.04476721212267876, -0.020303623750805855, 0.4243360757827759, 0.2557459771633148, 0.07693609595298767, -0.25946545600891113, -0.01638028770685196, 0.21026988327503204, 0.1324421614408493, -0.1359894722700119, 0.2562757432460785, -0.15202423930168152, -0.025490229949355125, -0.20269489288330078, -0.10973001271486282, -0.1159677729010582, -0.15834107995033264, 0.11542078852653503, -0.349846214056015 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
usually you can replace `download` in your dataset script with `download_and_prepare()` - could you share the code for your dataset here? :-)
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
22
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. usually you can replace `download` in your dataset script with `download_and_prepare()` - could you share the code for your dataset here? :-)
[ -0.29844045639038086, 0.2715887427330017, -0.025401674211025238, -0.17361287772655487, 0.11133858561515808, -0.03249125927686691, 0.3496930003166199, 0.30469509959220886, 0.09639108926057816, 0.1327766478061676, 0.1502358764410019, 0.18139110505580902, -0.29633527994155884, 0.05446058139204979, 0.16967663168907166, 0.0301237590610981, -0.0018449394265189767, 0.23788054287433624, -0.10852712392807007, 0.2644939422607422, -0.15700730681419373, 0.025940915569663048, -0.08346935361623764, 0.0347202830016613, -0.007923226803541183, 0.14469146728515625, -0.15397809445858002, 0.015508492477238178, -0.373642235994339, -0.9120109677314758, 0.4226640462875366, 0.08653821051120758, 0.18526966869831085, 0.4016505777835846, -0.00012305592827033252, -0.07805042713880539, 0.5848660469055176, -0.0812644362449646, -0.6614248752593994, -0.24752673506736755, 0.17996811866760254, -0.4401648938655853, 0.33093446493148804, -0.030891045928001404, 0.056430570781230927, -0.14283616840839386, 0.22567318379878998, 0.036393582820892334, 0.06255222856998444, 0.3939165472984314, 0.14157570898532867, 0.27843010425567627, 0.10396631062030792, -0.22894836962223053, -0.1139817088842392, 0.06613495200872421, -0.13955572247505188, 0.5098291635513306, 0.33754780888557434, 0.08086355030536652, -0.026145553216338158, -0.11726588755846024, 0.0758390873670578, 0.4826371967792511, 0.17982535064220428, -0.020105795934796333, 0.45073366165161133, -0.026417598128318787, -0.19099822640419006, 0.0009865248575806618, 0.8557761907577515, -0.557236909866333, -0.28589102625846863, 0.020644869655370712, 0.09727909415960312, -0.1694418489933014, 0.21763037145137787, -0.20579788088798523, -0.32464438676834106, 0.01700473204255104, -0.1948697566986084, -0.337063729763031, -0.09025173634290695, 0.3106403946876526, -0.21952900290489197, 0.1833120882511139, 0.06803254783153534, 0.21227127313613892, 0.055997684597969055, -0.09097342193126678, 0.23839010298252106, -0.022679699584841728, -0.006122444290667772, 0.2574823796749115, -0.08425547182559967, -0.24596400558948517, -0.3158247768878937, -0.20292505621910095, 0.15998384356498718, 0.05829789116978645, -0.48271316289901733, -0.18354415893554688, 0.04289029911160469, 0.2146054357290268, 0.11196036636829376, 0.317733496427536, 0.3439334034919739, 0.28459277749061584, 0.19506855309009552, 0.1744561791419983, 0.2424226999282837, 0.15372450649738312, -0.07111586630344391, -0.39698177576065063, 0.10665496438741684, 0.08801122009754181, 0.24009370803833008, -0.2806534469127655, -0.1707746982574463, -0.2619645595550537, -0.3381374478340149, 0.009782247245311737, -0.07732278108596802, 0.2922811508178711, -0.09014855325222015, -0.23386314511299133, 0.04907820746302605, 0.44676926732063293, -0.19828669726848602, -0.15629234910011292, -0.02728927507996559, 0.10530050098896027, -0.21463638544082642, 0.049913376569747925, 0.3252113163471222, 0.014613310806453228, 0.4042742848396301, -0.2409370094537735, -0.15935105085372925, -0.04674298316240311, 0.10865764319896698, 0.059232473373413086, -0.2635217010974884, 0.30100664496421814, -0.07218058407306671, 0.08340983092784882, 0.048662688583135605, -0.27629339694976807, -0.2149084508419037, 0.23098868131637573, 0.06791497021913528, -0.4953612983226776, -0.023342089727520943, 0.1133398488163948, -0.5093958377838135, -0.059858739376068115, -0.06913214176893234, -0.16433396935462952, 0.10568103939294815, -0.3929360508918762, -0.030814146623015404, -0.4348370134830475, -0.37226366996765137, -0.11169154942035675, -0.03351253271102905, 0.2969774603843689, -0.33695119619369507, -0.2700844705104828, -0.44567787647247314, -0.3823715150356293, 0.2131667584180832, -0.25797051191329956, -0.07269981503486633, 0.2081032544374466, -0.41918906569480896, -0.029213760048151016, 0.5602767467498779, -0.5683712363243103, -0.36130160093307495, 0.44704416394233704, -0.4189874231815338, -0.2590119540691376, 0.18650442361831665, -0.0008603418827988207, 0.06650573760271072, -0.3295423686504364, -0.14305251836776733, 0.06461130827665329, 0.0423731729388237, -0.13622736930847168, -0.11024294793605804, -0.2676571309566498, 0.19914446771144867, 0.26322752237319946, -0.09678233414888382, 0.13125525414943695, 0.14763695001602173, 0.10275781154632568, 0.3266957402229309, 0.14259015023708344, 0.17181050777435303, -0.08835884183645248, 0.23017220199108124, -0.22566978633403778, -0.14068517088890076, -0.10932862758636475, -0.6099810004234314, 0.2383185476064682, -0.06623521447181702, 0.20779499411582947, -0.14170926809310913, 0.023358389735221863, -0.6258833408355713, 0.01586984470486641, -0.059171222150325775, -0.2286381870508194, 0.006851642858237028, 0.1909550577402115, 0.11202634871006012, 0.018214566633105278, 0.03710727021098137, 0.0626557469367981, -0.21972577273845673, 0.17596451938152313, 0.010265305638313293, -0.0012483586324378848, -0.10338080674409866, 0.03828245773911476, 0.23613692820072174, 0.22019892930984497, 0.08650259673595428, -0.31085482239723206, -0.04248622804880142, 0.2862294614315033, 0.27485087513923645, 0.20396490395069122, 0.1929636001586914, -0.07467367500066757, 0.13896974921226501, 0.15383584797382355, 0.1740688532590866, 0.21679866313934326, 0.08699431270360947, -0.2102239429950714, 0.08055426180362701, 0.11069183796644211, 0.19249321520328522, -0.02169324830174446, -0.17491620779037476, 0.026054412126541138, 0.4049426317214966, 0.12913775444030762, -0.026237325742840767, -0.38228341937065125, 0.10046067088842392, 0.29537415504455566, 0.6366955637931824, -0.047104936093091965, 0.366362601518631, -0.10841751843690872, -0.19691742956638336, -0.34861817955970764, 0.008201361633837223, 0.16020745038986206, -0.15804894268512726, -0.09225594252347946, -0.005740282125771046, 0.2688913345336914, 0.8805614113807678, -0.026490002870559692, 0.1716458648443222, 0.14539185166358948, -0.11143899708986282, -0.20710675418376923, 0.18835251033306122, 0.07700680196285248, 0.04825577884912491, 0.14204147458076477, -0.01417926698923111, -0.3831111192703247, -0.08877193927764893, -0.053082533180713654, 0.45894336700439453, 0.484392374753952, -0.3677523136138916, -0.19629882276058197, -0.24734264612197876, -0.36711904406547546, -0.41309091448783875, -0.3259521722793579, -0.152617409825325, -0.3070310950279236, -0.1207633689045906, 0.27662524580955505, 0.13867036998271942, 0.2270921915769577, -0.2515861690044403, 0.18580736219882965, -0.2386372834444046, -0.3114602863788605, 0.09593988955020905, -0.1713583767414093, -0.3402532637119293, 0.030635185539722443, 0.2242271602153778, 0.03131996467709541, 0.4151515066623688, -0.14686715602874756, -0.21823152899742126, -0.2512979805469513, -0.2445472776889801, 0.12566052377223969, 0.08897516876459122, 0.5211690664291382, 0.459918349981308, -0.027610044926404953, 0.4571678936481476, 0.11719503998756409, 0.034604236483573914, -0.26613301038742065, -0.4669434726238251, -0.15766939520835876, -0.019803231582045555, 0.0831543505191803, -0.03829893842339516, -0.47527745366096497, -0.5595125555992126, -0.0779832974076271, 0.01707637868821621, 0.10753872990608215, 0.02640489861369133, 0.044132035225629807, -0.09966850280761719, 0.20519216358661652, 0.025716695934534073, -0.11663968861103058, -0.14416977763175964, 0.0730152353644371, 0.020061222836375237, -0.17576852440834045, -0.17963048815727234, 0.1131969690322876, -0.06287679076194763, 0.08196181803941727, -0.15483304858207703, -0.41781681776046753, -0.34866905212402344, 0.07465802133083344, 0.18385641276836395, 0.28053751587867737, -0.2723369002342224, 0.28959473967552185, 0.24983246624469757, 0.14815491437911987, -0.12449974566698074, -0.3925286829471588, 0.15622250735759735, 0.019437246024608612, 0.25050458312034607, 0.1468769609928131, 0.482563853263855, -0.19213831424713135, 0.7506566047668457, 0.11803735047578812, -0.16310954093933105, 0.020595410838723183, -0.2556125819683075, 0.041471920907497406, -0.05395025759935379, -0.12413839250802994, 0.3178110718727112, -0.13975505530834198, -0.24665527045726776, 0.19574576616287231, -0.22311292588710785, -0.05246172845363617, -0.09552178531885147, 0.2213321030139923, -0.06543803215026855, -0.02543662302196026, -0.05347732827067375, -0.2093602567911148, 0.05761226639151573, 0.03788955882191658, -0.07096198201179504, -0.19440779089927673, -0.2140195667743683, 0.0048631965182721615, 0.31353676319122314, 0.004929577000439167, 0.16027186810970306, -0.30051475763320923, 0.01857459731400013, -0.44620153307914734, 0.3039751946926117, 0.2712741792201996, 0.5984054207801819, 0.08295039087533951, 0.07675036042928696, 0.12854933738708496, 0.10277190059423447, 0.214049831032753, -0.028820855543017387, 0.12243358045816422, 0.25168922543525696, -0.13396185636520386, -0.08915599435567856, -0.24068473279476166, -0.0658818855881691, 0.21251852810382843, -0.029731804504990578, 0.09589340537786484, -0.350087970495224, -0.08926238119602203, 0.2652932107448578, 0.1409556269645691, -0.08590056747198105, -0.35511037707328796, -0.12330726534128189, -0.07265676558017731, -0.15032172203063965, -0.21273328363895416, -0.17580446600914001, 0.10044043511152267, -0.12927833199501038, 0.2820556163787842, 0.01495383121073246, -0.09018884599208832, -0.08661564439535141, 0.11567679792642593, 0.14000989496707916, -0.1844521462917328, 0.3376259505748749, 0.1773081123828888, 0.04688049480319023, 0.4023391604423523, 0.31333476305007935, -0.10161557048559189, -0.2746429741382599, 0.02688930369913578, -0.0069023058749735355, 0.3866168260574341, 0.12779398262500763, 0.11361671984195709, 0.04185861721634865, -0.26975369453430176, -0.14895153045654297, -0.29846101999282837, 0.1984383761882782, 0.3825054168701172, -0.09436485916376114, -0.05138183757662773, -0.3670406639575958, 0.6012393236160278, 0.2057899683713913, -0.15702876448631287, 0.47727301716804504, -0.25194334983825684, -0.2803325355052948, -0.0551794171333313, 0.07219060510396957, 0.8358832001686096, 0.1205170750617981, 0.16542430222034454, 0.37753617763519287, -0.26247909665107727, 0.3626868426799774, 0.09037943929433823, 0.16148610413074493, -0.5544981956481934, -0.08480004966259003, 0.024787534028291702, -0.23059777915477753, 0.24379819631576538, -0.10523279756307602, -0.23620693385601044, 0.37085577845573425, 0.04794004186987877, 0.2831428050994873, 0.25457656383514404, 0.3559739291667938, -0.11178558319807053, -0.04452484846115112, -0.10924924910068512, 0.10165832191705704, -0.16587036848068237, 0.40507540106773376, -0.23478902876377106, 0.08953950554132462, -0.08646884560585022, -0.012371436692774296, -0.2658645212650299, -0.1177593469619751, -0.14358888566493988, 0.0613325871527195, 0.07406354695558548, -0.1460016369819641, 0.10794640332460403, 0.1995038092136383, 0.13345986604690552, -0.16214776039123535, -0.16512355208396912, 0.15120474994182587, -0.1544622927904129, 0.10339926183223724, 0.3309921622276306, 0.057362258434295654, 0.23977400362491608, -0.3374665379524231, -0.17271889746189117, 0.1334928572177887, -0.23813052475452423, -0.13328047096729279, -0.09552449733018875, -0.1865774393081665, 0.10564973950386047, -0.3508002460002899, -0.3200393319129944, -0.14020806550979614, 0.07292886078357697, -0.14415010809898376, 0.04000476375222206, 0.2728863060474396, -0.3790871798992157, -0.006066022906452417, -0.12708009779453278, -0.3006230592727661, 0.01657925546169281, 0.47354406118392944, -0.061931345611810684, -0.16933684051036835, 0.7641562819480896, -0.17778785526752472, 0.06867405027151108, -0.1367509514093399, 0.06282448768615723, 0.011607315391302109, -0.13863568007946014, 0.025742609053850174, -0.04889152571558952, -0.18072549998760223, 0.10184179246425629, 0.5352264642715454, 0.3329124450683594, 0.004821314476430416, 0.0025992386508733034, -0.4617112874984741, -0.4422576129436493, -0.09019365906715393, -0.15435552597045898, 0.18460498750209808, -0.07652498781681061, 0.13444040715694427, 0.21724900603294373, 0.10431430488824844, -0.20098243653774261, -0.17669570446014404, 0.007593640126287937, 0.24287132918834686, -0.12657630443572998, 0.043921347707509995, -0.026722947135567665, 0.08731262385845184, 0.031429849565029144, -0.02703716605901718, -0.3348521292209625, -0.07218381762504578, -0.06499527394771576, 0.20209333300590515, 0.09678618609905243, -0.028630798682570457, 0.07220404595136642, -0.15105348825454712, -0.02545272558927536, 0.17819692194461823, -0.026853246614336967, 0.08564912527799606, -0.09723051637411118, 0.15172472596168518, 0.323880136013031, -0.07167673856019974, -0.0500326007604599, 0.21732375025749207, 0.056248653680086136, 0.16201654076576233, -0.06300607323646545, 0.2864285707473755, -0.13696040213108063, 0.05460214987397194, 0.20251184701919556, 0.10919065028429031, 0.00826989859342575, 0.25383952260017395, 0.11799744516611099, -0.1682083159685135, 0.14139583706855774, -0.049902476370334625, 0.3557969927787781, 0.7263293862342834, -0.2195245921611786, -0.12481648474931717, 0.4329027533531189, 0.056365642696619034, -0.25470423698425293, -0.04399574175477028, 0.3098616302013397, 0.13776229321956635, 0.04623576998710632, 0.21163645386695862, 0.0015199186746031046, -0.03864835575222969, 0.1667882204055786, 0.136115163564682, 0.12866854667663574, -0.22629186511039734, -0.05903846025466919, 0.4483225345611572, -0.22031785547733307, 0.16248919069766998, 0.39000236988067627, 0.06479001045227051, 0.21330390870571136, 0.2186935544013977, 0.0370323583483696, 0.6142650842666626, 0.05644610524177551, 0.017884615808725357, 0.08402588963508606, -0.049764249473810196, -0.0006024215254001319, 0.27663731575012207, -0.03003878891468048, -0.08608400076627731, 0.32754427194595337, 0.5549076199531555, -0.3502872884273529, -0.15907354652881622, -0.26408109068870544, -0.08660256117582321, 0.03199830278754234, -0.1442382037639618, -0.285094290971756, 0.08155746012926102, -0.2161397784948349, -0.13049784302711487, -0.1188771203160286, -0.3093909025192261, 0.1627066731452942, 0.017042912542819977, -0.39833638072013855, -0.2444036602973938, -0.29097846150398254, 0.19365999102592468, -0.10940323770046234, -0.2699061930179596, 0.22087353467941284, -0.013040437363088131, 0.03285647928714752, 0.29115352034568787, 0.32487085461616516, 0.4699459671974182, 0.1625402569770813, -0.31814834475517273, -0.23251259326934814, -0.1214054599404335, 0.0972113236784935, 0.035079870373010635, 0.38069844245910645, -0.049887340515851974, -0.10703082382678986, 0.4631485939025879, 0.07265851646661758, 0.03713390976190567, 0.2186434268951416, 0.43211260437965393, -0.48619788885116577, -0.17150621116161346, 0.7934108376502991, -0.07552880048751831, -0.14116635918617249, -0.12957756221294403, 0.25525858998298645, -0.2604233920574188, 0.13434624671936035, 0.09955931454896927, -0.0734473392367363, 0.13018867373466492, 0.02547362633049488, 0.03947751596570015, -0.10640526562929153, 0.3546755015850067, 0.31101474165916443, -0.04381674900650978, -0.36947259306907654, -0.2513987123966217, -0.7986477017402649, 0.22746266424655914, -0.11356884986162186, -0.3236985206604004, -0.06485434621572495, -0.2603072226047516, 0.18936985731124878, 0.048111192882061005, -0.14149132370948792, 0.19518372416496277, 0.0548698790371418, 0.1351766735315323, -0.18994446098804474, -0.3179086744785309, -0.28872302174568176, -0.023948905989527702, -0.10388270020484924, -0.38692688941955566, 0.11374405026435852, -0.3304968774318695, -0.07958602905273438, 0.11841258406639099, -0.00978133175522089, 0.47797977924346924, -0.005791679956018925, 0.3654899001121521, -0.11248235404491425, 0.7752764225006104, 0.18795110285282135, 0.06328219920396805, -0.16494369506835938, 0.12217647582292557, -0.12064740806818008, 0.13483795523643494, -0.07033583521842957, 0.10441012680530548, -0.11896292120218277, 0.5212562084197998, -0.44372087717056274, 0.4605462849140167, -0.028102917596697807, 0.07408181577920914, -0.3021467924118042, 0.045647867023944855, -0.25546136498451233, 0.12183515727519989, 0.34445950388908386, 0.1411459892988205, -0.09161032736301422, 0.35506948828697205, -0.21494273841381073, -0.19039171934127808, 0.3962551951408386, -0.10348974168300629, -0.09398656338453293, 0.02151637151837349, 0.2646278142929077, -0.05172327905893326, -0.140091672539711, -0.4393872916698456, 0.19632719457149506, 0.29076045751571655, -0.09371230006217957, 0.11838958412408829, 0.26023393869400024, 0.28375154733657837, 0.13402162492275238, 0.018327342346310616, 0.9827129244804382, -0.05974318087100983, 0.1760433316230774, -0.10538399964570999, -0.22786012291908264 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
I have an initial version here: https://github.com/EntilZha/nlp/tree/master/datasets/qanta Thats pretty close to what I'll do as a PR, but still want to do some more sanity checks/tests (just got tests passing). I figured out how to get all tests passing by adding a download command and some finagling with the data zip https://github.com/EntilZha/nlp/blob/master/tests/utils.py#L127
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
52
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. I have an initial version here: https://github.com/EntilZha/nlp/tree/master/datasets/qanta Thats pretty close to what I'll do as a PR, but still want to do some more sanity checks/tests (just got tests passing). I figured out how to get all tests passing by adding a download command and some finagling with the data zip https://github.com/EntilZha/nlp/blob/master/tests/utils.py#L127
[ -0.1962164044380188, 0.40003663301467896, -0.03146868571639061, -0.1644131988286972, 0.044283874332904816, -0.1332930475473404, 0.2794852554798126, 0.37617382407188416, 0.009489011019468307, 0.08292659372091293, 0.20643217861652374, 0.15806853771209717, -0.30056875944137573, 0.06804094463586807, 0.1369246542453766, 0.0012007486075162888, -0.056462906301021576, 0.16989615559577942, -0.09795354306697845, 0.2125062346458435, -0.12722063064575195, 0.07482463121414185, 0.020457282662391663, 0.10806168615818024, -0.0279990341514349, 0.08875596523284912, -0.2158946692943573, 0.03439050167798996, -0.38094669580459595, -0.7858073711395264, 0.3818259537220001, 0.07057212293148041, 0.1851997673511505, 0.3587057590484619, -0.0001226895401487127, -0.08625637739896774, 0.5880787372589111, -0.05607656016945839, -0.7038211226463318, -0.23162031173706055, 0.24540702998638153, -0.4804559648036957, 0.30185362696647644, -0.021507857367396355, 0.18851712346076965, -0.1316383183002472, 0.24420835077762604, 0.10435124486684799, 0.16175146400928497, 0.32244816422462463, 0.11873658746480942, 0.27763259410858154, 0.051467739045619965, -0.17972403764724731, -0.10094481706619263, 0.08600436896085739, -0.0892680287361145, 0.5028811097145081, 0.3648625314235687, 0.03111984208226204, -0.1123633086681366, -0.13111865520477295, 0.0922636017203331, 0.49313539266586304, 0.125021830201149, -0.02229432389140129, 0.501173734664917, 0.01911216415464878, -0.21289396286010742, 0.04104846715927124, 0.7716667652130127, -0.5064083933830261, -0.37548547983169556, -0.07861018925905228, 0.05649541690945625, -0.13751418888568878, 0.260755717754364, -0.1600877344608307, -0.3946453332901001, 0.019819963723421097, -0.16637592017650604, -0.28037208318710327, -0.06712199002504349, 0.2935136556625366, -0.28309366106987, 0.26829126477241516, 0.07774811238050461, 0.17557835578918457, 0.08950234949588776, -0.10085011273622513, 0.257343590259552, -0.06183681637048721, 0.024405231699347496, 0.2348814755678177, -0.03021564893424511, -0.22387483716011047, -0.26101967692375183, -0.2221420854330063, 0.192213773727417, 0.08718465268611908, -0.46262574195861816, -0.19067612290382385, -0.031207416206598282, 0.14959871768951416, 0.1416144222021103, 0.4171293079853058, 0.33842331171035767, 0.23890084028244019, 0.24328815937042236, 0.1627921313047409, 0.24343957006931305, 0.2634247839450836, -0.05981258675456047, -0.3704582154750824, -0.03249732032418251, 0.09223353117704391, 0.14825262129306793, -0.3287044167518616, -0.15334022045135498, -0.22707168757915497, -0.38531777262687683, -0.046314191073179245, -0.054925546050071716, 0.23764850199222565, -0.10888967663049698, -0.14659355580806732, 0.04901169613003731, 0.41204389929771423, -0.2063007950782776, -0.1511351317167282, -0.029488692060112953, 0.044526323676109314, -0.1915409415960312, -0.012614103965461254, 0.30864498019218445, 0.05575856938958168, 0.3780354857444763, -0.36712145805358887, -0.11766038835048676, -0.0343116819858551, 0.11314684897661209, 0.18523584306240082, -0.2689652442932129, 0.2899585962295532, -0.1555456668138504, 0.052032142877578735, 0.07850415259599686, -0.29270139336586, -0.22106733918190002, 0.16356506943702698, 0.12039454281330109, -0.4569105803966522, 0.037894830107688904, 0.09306175261735916, -0.6060150265693665, -0.06466449052095413, 0.012727136723697186, -0.11572936922311783, 0.09138728678226471, -0.3326219916343689, 0.026576446369290352, -0.39136913418769836, -0.20526567101478577, -0.0672704204916954, -0.09972656518220901, 0.239104762673378, -0.2667430639266968, -0.33723214268684387, -0.4481429159641266, -0.32592299580574036, 0.1292017102241516, -0.22480514645576477, -0.07712365686893463, 0.2149806022644043, -0.37946194410324097, -0.008516068570315838, 0.549782931804657, -0.5216635465621948, -0.35860756039619446, 0.44549551606178284, -0.42921745777130127, -0.2467300295829773, 0.23419708013534546, -0.04898207634687424, -0.0027358396910130978, -0.3621472716331482, -0.08949577808380127, 0.08227457851171494, 0.04962048679590225, -0.17818880081176758, -0.1884840875864029, -0.34391698241233826, 0.11360983550548553, 0.220098078250885, -0.03988165780901909, 0.08876338601112366, 0.09676407277584076, 0.054097097367048264, 0.35139426589012146, 0.11439753323793411, 0.15504473447799683, -0.14401863515377045, 0.17716705799102783, -0.23787640035152435, -0.17119525372982025, -0.05177222564816475, -0.5992658138275146, 0.2523057460784912, -0.11843055486679077, 0.21302375197410583, -0.04645262658596039, -0.011181975714862347, -0.5071431398391724, 0.003305689198896289, 0.01984049752354622, -0.2721043527126312, 0.0020704620983451605, 0.2044224590063095, 0.12961775064468384, -0.047958437353372574, -0.010605939663946629, 0.03222338482737541, -0.21072953939437866, 0.1925332099199295, -0.009809226728975773, -0.020286353304982185, -0.04618968814611435, 0.06240842118859291, 0.2495947629213333, 0.25558868050575256, 0.05471820384263992, -0.34098246693611145, 0.007953260093927383, 0.28200313448905945, 0.20832444727420807, 0.18946510553359985, 0.18919208645820618, -0.07011060416698456, 0.1703489124774933, 0.26392483711242676, 0.16458021104335785, 0.26295530796051025, -0.01518195029348135, -0.2123047411441803, -0.08616410940885544, 0.24468617141246796, 0.1680414080619812, 0.04419174790382385, -0.12836235761642456, 0.06571459770202637, 0.323811411857605, 0.11381228268146515, -0.13485534489154816, -0.2280585914850235, 0.1960444301366806, 0.28523680567741394, 0.6019870042800903, -0.05045628920197487, 0.3259405195713043, -0.04792163893580437, -0.27811822295188904, -0.3706563413143158, 0.017936602234840393, 0.11389987170696259, -0.09621313959360123, -0.056275587528944016, 0.09414354711771011, 0.36090877652168274, 0.8490371108055115, -0.03925830498337746, 0.2709169387817383, 0.0866793766617775, -0.18442854285240173, -0.23165827989578247, 0.1509414166212082, 0.11443711072206497, -0.02646586485207081, 0.1556096225976944, 0.0009209238342009485, -0.4011898934841156, -0.0890222117304802, -0.00013104175741318613, 0.43312111496925354, 0.39993149042129517, -0.34717369079589844, -0.10980546474456787, -0.30210116505622864, -0.4445500075817108, -0.3985893428325653, -0.3630511164665222, -0.2080824226140976, -0.38136446475982666, -0.06145554408431053, 0.22124408185482025, -0.0000810364363132976, 0.23552817106246948, -0.19494999945163727, 0.28329363465309143, -0.3094022274017334, -0.3343958854675293, 0.11647369712591171, -0.20119445025920868, -0.4004580080509186, 0.038076844066381454, 0.28132161498069763, 0.07048669457435608, 0.44269314408302307, -0.10782049596309662, -0.297360897064209, -0.2132592499256134, -0.3188501000404358, 0.1328553408384323, 0.06095792353153229, 0.5699260234832764, 0.476179838180542, -0.01488469634205103, 0.42041054368019104, 0.07882976531982422, 0.042043741792440414, -0.23354840278625488, -0.4335298538208008, -0.23417878150939941, 0.09182939678430557, 0.09691432863473892, -0.1396644413471222, -0.39729273319244385, -0.52669358253479, -0.08559026569128036, 0.17034383118152618, 0.127034991979599, 0.0732078030705452, 0.09398307651281357, -0.1114279106259346, 0.2511831820011139, 0.08683197945356369, -0.18744944036006927, -0.11685296148061752, 0.07108961790800095, -0.009084474295377731, -0.14291846752166748, -0.18716339766979218, 0.0666738748550415, -0.1160896047949791, 0.0868186354637146, -0.1572064310312271, -0.48865631222724915, -0.44232678413391113, 0.15424875915050507, 0.23099425435066223, 0.3925914168357849, -0.20678284764289856, 0.2817329168319702, 0.20180130004882812, 0.12314736843109131, -0.10951633006334305, -0.3401254713535309, 0.16022495925426483, 0.09377182275056839, 0.3110191524028778, 0.13703356683254242, 0.34237030148506165, -0.11723019182682037, 0.7626636028289795, 0.11879157274961472, -0.09411487728357315, -0.03337553143501282, -0.13440778851509094, 0.015389122068881989, -0.01784033142030239, -0.04059560224413872, 0.3350594937801361, -0.2001541554927826, -0.31042638421058655, 0.20721690356731415, -0.11684413254261017, -0.03705428168177605, -0.13537539541721344, 0.289180189371109, -0.01741296797990799, -0.06730537861585617, 0.0038469242863357067, -0.21916760504245758, 0.008076525293290615, 0.09175167977809906, -0.028248609974980354, -0.19920840859413147, -0.15857325494289398, 0.03104414790868759, 0.30560776591300964, 0.009298949502408504, 0.17336349189281464, -0.2502236068248749, 0.006672599818557501, -0.38045933842658997, 0.3209521174430847, 0.30832692980766296, 0.4791105091571808, 0.08994140475988388, 0.04050209000706673, 0.1281771957874298, 0.023768724873661995, 0.2255101203918457, 0.07952552288770676, 0.16413183510303497, 0.25303003191947937, -0.13103421032428741, -0.04519340395927429, -0.24799904227256775, -0.061088692396879196, 0.06379716098308563, 0.004301052074879408, 0.09554502367973328, -0.40971267223358154, -0.09873350709676743, 0.18825578689575195, 0.08911528438329697, -0.07824177294969559, -0.23516075313091278, -0.07499676942825317, -0.0863107219338417, -0.11131929606199265, -0.21148739755153656, -0.22183185815811157, 0.017018204554915428, -0.21502092480659485, 0.40636956691741943, -0.0038533478509634733, -0.08811715990304947, -0.06015439331531525, 0.09694578498601913, 0.08987754583358765, -0.2820565104484558, 0.2924054265022278, 0.26512542366981506, 0.07484505325555801, 0.3037792444229126, 0.31530171632766724, -0.09081118553876877, -0.21888962388038635, 0.02536877617239952, 0.05520324781537056, 0.42117199301719666, 0.35065004229545593, 0.10434029251337051, -0.05036744475364685, -0.23994220793247223, -0.2073725014925003, -0.2343253642320633, 0.19697505235671997, 0.28577369451522827, -0.03091180883347988, 0.07850922644138336, -0.31849759817123413, 0.7194918394088745, 0.22033101320266724, -0.2501451373100281, 0.4313563406467438, -0.21806496381759644, -0.26280879974365234, -0.06578270345926285, 0.12758244574069977, 0.9033374786376953, 0.14794974029064178, 0.23772503435611725, 0.4199581742286682, -0.26912349462509155, 0.46432656049728394, 0.07282888889312744, 0.1551104635000229, -0.4947300851345062, -0.00016493158182129264, 0.019680967554450035, -0.18237504363059998, 0.20951975882053375, -0.09575729072093964, -0.22470085322856903, 0.37494614720344543, 0.146523118019104, 0.2810683846473694, 0.3701074421405792, 0.30879998207092285, -0.2517711818218231, -0.03366045281291008, -0.18580879271030426, 0.09027866274118423, -0.14980097115039825, 0.40192821621894836, -0.2816433608531952, -0.0003147864481434226, -0.1442372351884842, -0.04424754157662392, -0.25294482707977295, -0.11223667114973068, -0.1734396070241928, 0.06214199960231781, 0.15585318207740784, -0.13206034898757935, -0.06616007536649704, 0.18333126604557037, 0.08269034326076508, -0.21432577073574066, -0.1497819870710373, 0.125441774725914, -0.1433178037405014, 0.16847257316112518, 0.3818269371986389, 0.04532889649271965, 0.32668229937553406, -0.31696557998657227, -0.05816670134663582, 0.06710881739854813, -0.2803245782852173, -0.12472692131996155, -0.13294465839862823, -0.17246976494789124, 0.14165620505809784, -0.2999444901943207, -0.14256227016448975, -0.10251213610172272, -0.009897447191178799, -0.11590281128883362, 0.022395405918359756, 0.28106245398521423, -0.35529664158821106, -0.06101107969880104, -0.11896102130413055, -0.229909285902977, 0.009613190777599812, 0.4747704565525055, -0.06723813712596893, -0.13764449954032898, 0.6765257120132446, -0.21037662029266357, 0.06588239222764969, -0.1299181431531906, 0.06201256439089775, -0.1010732352733612, -0.11184319853782654, 0.07678360491991043, -0.16967491805553436, -0.3008669316768646, 0.04371461272239685, 0.5850263833999634, 0.44189485907554626, 0.01240659411996603, -0.07040780782699585, -0.46289220452308655, -0.41166549921035767, -0.0003565764636732638, -0.14149700105190277, 0.20699220895767212, -0.21216511726379395, 0.2207392007112503, 0.16328176856040955, 0.04603132978081703, -0.18724991381168365, -0.201705664396286, 0.049465011805295944, 0.2059495449066162, -0.13847650587558746, 0.04342780262231827, -0.015561915002763271, 0.14493170380592346, -0.03392514958977699, 0.06349050253629684, -0.3615332543849945, -0.07694458216428757, -0.10113611817359924, 0.205429345369339, 0.11184282600879669, -0.0389489121735096, 0.15560312569141388, -0.11734632402658463, 0.039722029119729996, 0.21439765393733978, -0.06621638685464859, -0.04835908114910126, -0.09268877655267715, 0.11267165839672089, 0.34582746028900146, 0.09795007854700089, -0.006950722075998783, 0.30149081349372864, 0.10003815591335297, 0.1763516664505005, -0.11613292992115021, 0.26699644327163696, -0.23393245041370392, 0.08855492621660233, 0.20248933136463165, 0.1292089819908142, -0.019432740285992622, 0.2600131034851074, 0.13960111141204834, -0.11417266726493835, 0.08041876554489136, -0.1625293344259262, 0.3586401343345642, 0.8415080308914185, -0.17826047539710999, -0.13109172880649567, 0.47036680579185486, 0.032728150486946106, -0.19339051842689514, 0.027116669341921806, 0.2276790738105774, 0.20254090428352356, 0.039652951061725616, 0.2661378085613251, 0.010690231807529926, -0.13747438788414001, 0.19611214101314545, 0.14074355363845825, 0.09474228322505951, -0.1786988526582718, -0.07266798615455627, 0.5383139848709106, -0.2553040087223053, 0.1620742231607437, 0.3502655029296875, 0.0004751363885588944, 0.21568617224693298, 0.08506053686141968, 0.025451358407735825, 0.5338088870048523, -0.012156912125647068, 0.05823533609509468, 0.11353061348199844, 0.01912609115242958, -0.04424288496375084, 0.259392112493515, 0.051443591713905334, -0.01177574135363102, 0.3216552734375, 0.6301923394203186, -0.4306804835796356, -0.17472806572914124, -0.2778608798980713, -0.0494091622531414, 0.08228189498186111, -0.149983748793602, -0.39644888043403625, 0.04359782114624977, -0.2213105857372284, -0.07336188107728958, -0.14795038104057312, -0.3185242712497711, 0.2299109399318695, 0.003296698909252882, -0.3555910885334015, -0.3461932837963104, -0.3967624306678772, 0.14084655046463013, -0.03580840304493904, -0.27528664469718933, 0.25086456537246704, 0.02940678969025612, 0.018532320857048035, 0.28296446800231934, 0.28391772508621216, 0.44415634870529175, 0.08628559857606888, -0.4387357831001282, -0.25408488512039185, -0.1913307011127472, 0.13583241403102875, -0.07762172073125839, 0.39957278966903687, -0.05981661379337311, -0.16802260279655457, 0.4992051124572754, 0.05886784940958023, 0.04353220760822296, 0.26643142104148865, 0.3724867105484009, -0.5206156969070435, -0.1689816564321518, 0.8208152651786804, -0.06794864684343338, -0.16675154864788055, -0.027103710919618607, 0.17767925560474396, -0.2319854497909546, 0.08259518444538116, -0.035704031586647034, -0.04750058799982071, 0.18776071071624756, 0.11940551549196243, 0.04107683151960373, -0.17806358635425568, 0.3992721140384674, 0.3388812839984894, -0.07150029391050339, -0.3425285816192627, -0.21393299102783203, -0.8131747841835022, 0.08850466459989548, -0.10057635605335236, -0.37371405959129333, -0.06355058401823044, -0.23741276562213898, 0.25356847047805786, 0.1047837883234024, -0.2574350833892822, 0.1771562099456787, -0.0015824732836335897, 0.12594372034072876, -0.2627957761287689, -0.365987628698349, -0.21408496797084808, -0.025764409452676773, -0.029740480706095695, -0.40254855155944824, 0.07842409610748291, -0.31337136030197144, -0.09926488250494003, 0.2237178534269333, -0.039395153522491455, 0.4940950870513916, 0.07308284193277359, 0.25377795100212097, -0.012197772972285748, 0.6868556141853333, 0.24270272254943848, -0.0313035324215889, -0.2473425716161728, 0.17876677215099335, -0.13330844044685364, 0.12535804510116577, -0.10701501369476318, 0.1350855976343155, -0.17025771737098694, 0.6143175959587097, -0.41418296098709106, 0.4506012499332428, 0.07495135068893433, 0.07818985730409622, -0.28697723150253296, 0.04797068238258362, -0.34138748049736023, 0.14387986063957214, 0.32806554436683655, 0.30497342348098755, -0.11956615746021271, 0.34439727663993835, -0.2657005488872528, -0.1519375592470169, 0.41174817085266113, -0.2489464282989502, -0.05838595703244209, 0.05723579227924347, 0.21902203559875488, 0.0009552594856359065, -0.1586715430021286, -0.44350025057792664, 0.1210436299443245, 0.26714086532592773, -0.13255079090595245, 0.08481642603874207, 0.14585541188716888, 0.2846493422985077, 0.02644173987209797, -0.03574095666408539, 0.9360098242759705, -0.07236679643392563, 0.1544691026210785, -0.018570255488157272, -0.2431814968585968 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
I'm quite positive that you can just replace the `dl_manager.download()` statements here: https://github.com/EntilZha/nlp/blob/4d46443b65f1f756921db8275594e6af008a1de7/datasets/qanta/qanta.py#L194 with `dl_manager.download_and_extract()` even though you don't extract anything. I would prefer to avoid adding more functions to the MockDataLoadManager and keep it as simple as possible (It's already to complex now IMO). Could you check if you can replace the `download()` function?
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
55
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. I'm quite positive that you can just replace the `dl_manager.download()` statements here: https://github.com/EntilZha/nlp/blob/4d46443b65f1f756921db8275594e6af008a1de7/datasets/qanta/qanta.py#L194 with `dl_manager.download_and_extract()` even though you don't extract anything. I would prefer to avoid adding more functions to the MockDataLoadManager and keep it as simple as possible (It's already to complex now IMO). Could you check if you can replace the `download()` function?
[ -0.10087016224861145, 0.5024331212043762, -0.03979744389653206, -0.08978427946567535, 0.09680349379777908, -0.12240692973136902, 0.2872874438762665, 0.3135526776313782, -0.05438557267189026, 0.11574174463748932, 0.046182114630937576, 0.1649334579706192, -0.27445515990257263, -0.005963859613984823, 0.16595666110515594, -0.03701652213931084, -0.07802166044712067, 0.25511303544044495, -0.09114520251750946, 0.21499760448932648, -0.23069177567958832, 0.12446098029613495, -0.07093708962202072, 0.07457935810089111, 0.13653087615966797, 0.19969803094863892, -0.16580606997013092, 0.05431869998574257, -0.3160353899002075, -0.876468300819397, 0.34792110323905945, 0.18099525570869446, 0.2423819601535797, 0.23878878355026245, -0.00012480559234973043, -0.1538027971982956, 0.5954886078834534, -0.12650635838508606, -0.6724079847335815, -0.17283177375793457, 0.24162250757217407, -0.44368821382522583, 0.26434189081192017, -0.07161904126405716, 0.2790195047855377, -0.1683756560087204, 0.18214724957942963, 0.11028362065553665, 0.07405313849449158, 0.3863357901573181, 0.11946222186088562, 0.24075782299041748, 0.08585676550865173, -0.12019384652376175, -0.058449819684028625, 0.024083854630589485, -0.0008230159291997552, 0.6254271864891052, 0.4066478908061981, 0.021435411646962166, 0.021082354709506035, -0.010083990171551704, -0.061721447855234146, 0.45013120770454407, 0.24951261281967163, -0.04782029986381531, 0.49971386790275574, -0.0700686052441597, -0.16319656372070312, -0.013614766299724579, 0.6971597075462341, -0.5866326093673706, -0.3620380461215973, -0.054751206189394, 0.15433408319950104, -0.20426040887832642, 0.13494253158569336, -0.28096359968185425, -0.4394976496696472, 0.058052536100149155, -0.10193029791116714, -0.5488116145133972, 0.017848778516054153, 0.25423696637153625, -0.30119872093200684, 0.2867894172668457, 0.11398261040449142, 0.1454484909772873, 0.24242813885211945, -0.05278779938817024, 0.21367624402046204, -0.022138562053442, -0.08245529234409332, 0.16982325911521912, 0.0019522076472640038, -0.3044985234737396, -0.3149797022342682, -0.2155284583568573, 0.09691578894853592, 0.13492238521575928, -0.37371397018432617, -0.14380402863025665, -0.05634018033742905, 0.13417312502861023, 0.09189242124557495, 0.4165341854095459, 0.3381824195384979, 0.22063656151294708, 0.10820451378822327, 0.22823067009449005, 0.23546768724918365, 0.12068718671798706, 0.03867407143115997, -0.3611074984073639, 0.019834082573652267, 0.19145004451274872, 0.13478772342205048, -0.19446928799152374, -0.08906871825456619, -0.24023915827274323, -0.4562419652938843, -0.05777711048722267, -0.1779346466064453, 0.2095634490251541, -0.08753695338964462, -0.08982694894075394, 0.10732883214950562, 0.40889468789100647, -0.12385623157024384, -0.27147436141967773, 0.0032229716889560223, 0.19675232470035553, -0.09802553802728653, 0.11117035895586014, 0.21871031820774078, 0.024942509829998016, 0.3383959233760834, -0.37400108575820923, -0.034395981580019, -0.03799280524253845, 0.1008778065443039, -0.013926450163125992, -0.1752636432647705, 0.29795607924461365, -0.13329365849494934, 0.03545466437935829, 0.02031053602695465, -0.3518681824207306, -0.20858587324619293, 0.24709321558475494, 0.22873270511627197, -0.4571872651576996, -0.09473820775747299, 0.059910472482442856, -0.5005056262016296, -0.02239154651761055, -0.05310003086924553, -0.076795794069767, -0.07516738772392273, -0.5091549754142761, 0.15669932961463928, -0.3630225956439972, -0.2893626391887665, -0.05024648457765579, -0.059420350939035416, 0.37651327252388, -0.2985683083534241, -0.29981234669685364, -0.5405735373497009, -0.30797621607780457, 0.149720698595047, -0.244998961687088, -0.16400165855884552, 0.26623648405075073, -0.3465721309185028, 0.04316471144556999, 0.6672197580337524, -0.6826147437095642, -0.3817330598831177, 0.3879486918449402, -0.2968752086162567, -0.20163170993328094, 0.13723111152648926, -0.08174300193786621, 0.07141759246587753, -0.31405961513519287, -0.2028537094593048, -0.005927268415689468, 0.09159789979457855, -0.2129979431629181, -0.13809283077716827, -0.47395285964012146, 0.13925619423389435, 0.22118911147117615, 0.0007851556874811649, 0.24563640356063843, 0.03358960524201393, 0.14707604050636292, 0.36340445280075073, 0.1003376767039299, 0.13167637586593628, -0.15881609916687012, 0.24994680285453796, -0.18818053603172302, -0.18373259902000427, -0.08811551332473755, -0.5891332030296326, 0.21657982468605042, -0.16063576936721802, 0.22971317172050476, -0.07063758373260498, -0.0018131551332771778, -0.4160645008087158, -0.039894189685583115, 0.09383790194988251, -0.2353091537952423, -0.025887051597237587, 0.19697047770023346, 0.19552244246006012, 0.004465823993086815, -0.01573062315583229, 0.06437589228153229, -0.1620643585920334, 0.23343253135681152, 0.07400815933942795, -0.05165088176727295, -0.022028660401701927, 0.14289598166942596, 0.20402996242046356, 0.24361903965473175, 0.03090188466012478, -0.25026261806488037, -0.036677151918411255, 0.3134898245334625, 0.14573054015636444, 0.27442291378974915, 0.10608703643083572, -0.04028899595141411, 0.18484210968017578, 0.28031137585639954, 0.16047853231430054, 0.2589097321033478, 0.11254424601793289, -0.12400826811790466, 0.0815054401755333, 0.11401244252920151, 0.30637696385383606, 0.03690582141280174, -0.10140994936227798, 0.04883849248290062, 0.2526004910469055, 0.06619599461555481, -0.12181832641363144, -0.3277987241744995, -0.024629639461636543, 0.2636626362800598, 0.6258202195167542, -0.09830503910779953, 0.3067035377025604, -0.10810234397649765, -0.27656814455986023, -0.3729352355003357, 0.02712607942521572, 0.13545626401901245, -0.2431388646364212, -0.05934811756014824, 0.005909881554543972, 0.4128134846687317, 0.8891432285308838, -0.08765023201704025, 0.2127113789319992, 0.25769320130348206, 0.09946401417255402, -0.1476353108882904, 0.31021782755851746, 0.273175448179245, 0.03670435771346092, 0.1560714691877365, 0.017295531928539276, -0.24424812197685242, -0.0324268639087677, -0.09209571778774261, 0.330915629863739, 0.3591979742050171, -0.30033472180366516, -0.10634119808673859, -0.30636605620384216, -0.5244476199150085, -0.3691817820072174, -0.3438989222049713, -0.20262251794338226, -0.3367355465888977, -0.16702468693256378, 0.25610971450805664, -0.016801126301288605, 0.2510782778263092, -0.17695993185043335, 0.23437583446502686, -0.3581813871860504, -0.26189184188842773, 0.014088819734752178, -0.06981273740530014, -0.28657805919647217, 0.005503244232386351, 0.2328590452671051, 0.1618766337633133, 0.24962659180164337, -0.2545630931854248, -0.21234263479709625, -0.34648963809013367, -0.18361665308475494, 0.19877421855926514, 0.2477278709411621, 0.4137883484363556, 0.5279138684272766, -0.0702146589756012, 0.3567330837249756, 0.05945051088929176, -0.06608288735151291, -0.1920313835144043, -0.3716992139816284, -0.30342382192611694, -0.0005269242683425546, 0.27847978472709656, -0.08362536877393723, -0.3137311339378357, -0.6029019951820374, -0.14122115075588226, 0.034232206642627716, 0.07534819096326828, 0.1252206712961197, 0.030590686947107315, -0.24102658033370972, 0.22648802399635315, -0.06759880483150482, -0.23341916501522064, -0.20004798471927643, 0.1599310040473938, -0.009144952520728111, -0.16310198605060577, -0.04546184092760086, 0.07400716096162796, -0.10867732018232346, 0.05650421231985092, -0.23117774724960327, -0.4390263855457306, -0.49937358498573303, 0.20321010053157806, -0.011472797952592373, 0.14936859905719757, -0.18084704875946045, 0.2266942858695984, 0.24462617933750153, 0.14415551722049713, 0.007774581201374531, -0.4050814211368561, 0.12961599230766296, 0.1733422428369522, 0.25960150361061096, 0.2025206834077835, 0.4707225263118744, -0.18957793712615967, 0.6809969544410706, 0.08018959313631058, -0.06540750712156296, -0.006915408652275801, -0.19380633533000946, 0.12983419001102448, 0.09109845012426376, -0.07360187917947769, 0.2188321352005005, -0.19008886814117432, -0.2723953425884247, 0.26072928309440613, -0.3641332983970642, -0.09073998779058456, -0.1529102772474289, 0.21480999886989594, -0.18263716995716095, -0.1637749820947647, 0.07783829420804977, -0.2045094072818756, 0.0020857532508671284, 0.1484217345714569, 0.00008367608825210482, -0.21906661987304688, -0.058467570692300797, 0.10795843601226807, 0.2922639846801758, 0.09092121571302414, 0.11736774444580078, -0.28290224075317383, -0.007355594541877508, -0.4521167576313019, 0.3575472831726074, 0.3472757339477539, 0.4865114092826843, 0.2158142477273941, 0.039203789085149765, 0.0082472562789917, 0.040773604065179825, 0.21377934515476227, -0.029947955161333084, 0.013004861772060394, 0.2556818127632141, -0.13648857176303864, 0.061476826667785645, -0.22550521790981293, -0.09265473484992981, 0.0719015821814537, 0.08172386884689331, -0.07073070108890533, -0.31187525391578674, -0.15123626589775085, 0.3328525722026825, 0.1557932198047638, -0.1117551326751709, -0.2423667162656784, 0.00851057842373848, -0.15314200520515442, -0.1970021277666092, -0.19007055461406708, -0.24555879831314087, 0.1381567269563675, -0.12386205047369003, 0.44703012704849243, 0.013527016155421734, -0.09884051978588104, 0.08706270903348923, 0.024768998846411705, 0.187104731798172, -0.1544833481311798, 0.13126227259635925, 0.17856884002685547, 0.0028184344992041588, 0.3041520118713379, 0.2868061363697052, -0.02494579553604126, -0.26218181848526, 0.024909837171435356, 0.11324064433574677, 0.3784661293029785, 0.3422512412071228, 0.19741135835647583, 0.0638270378112793, -0.28435468673706055, -0.18205124139785767, -0.14466559886932373, 0.16747474670410156, 0.350755900144577, -0.1033928319811821, 0.05127176269888878, -0.23811398446559906, 0.6501335501670837, 0.15208756923675537, -0.2175854593515396, 0.29246801137924194, -0.19579973816871643, -0.25129613280296326, -0.12040631473064423, 0.1437392234802246, 0.9772183299064636, 0.17870500683784485, 0.19215767085552216, 0.37996527552604675, -0.2701803743839264, 0.5282023549079895, 0.12167239189147949, 0.11431528627872467, -0.4974672198295593, -0.036809105426073074, 0.01277151983231306, -0.21958334743976593, 0.2260391116142273, -0.15513776242733002, -0.23718251287937164, 0.35429805517196655, -0.010540962219238281, 0.28239917755126953, 0.3365882337093353, 0.32032984495162964, -0.22276438772678375, -0.026248833164572716, -0.161359965801239, 0.042556922882795334, -0.049717362970113754, 0.41778913140296936, -0.2992789149284363, 0.009875660762190819, -0.08630752563476562, -0.04552652686834335, -0.325615257024765, -0.20060066878795624, -0.29272374510765076, 0.017192082479596138, 0.08279824256896973, -0.13996846973896027, -0.010615960694849491, 0.17796418070793152, 0.08981534838676453, -0.21600760519504547, -0.099816232919693, 0.09243497252464294, -0.14039382338523865, 0.2634120583534241, 0.38699373602867126, 0.09648340195417404, 0.26869824528694153, -0.34610772132873535, -0.09964385628700256, 0.30464792251586914, -0.23491986095905304, -0.2469949573278427, -0.041801322251558304, -0.2066764086484909, 0.13844771683216095, -0.19164152443408966, -0.32677629590034485, -0.025652989745140076, 0.0002871699398383498, -0.13440831005573273, -0.00313161825761199, 0.16473489999771118, -0.3638894855976105, -0.008035514503717422, -0.2035222053527832, -0.25346142053604126, 0.04944055527448654, 0.4804394543170929, -0.02904648892581463, -0.22069257497787476, 0.6655129790306091, -0.2630751132965088, 0.1176690012216568, -0.12205177545547485, -0.007447515148669481, -0.09300586581230164, -0.10929890722036362, 0.07604288309812546, -0.17612077295780182, -0.17633578181266785, -0.010683165863156319, 0.582041323184967, 0.3484703302383423, 0.0852460265159607, -0.1574583649635315, -0.4427752196788788, -0.40297049283981323, 0.01759088598191738, -0.06714580208063126, 0.13793665170669556, -0.18048006296157837, 0.18105721473693848, 0.12914839386940002, 0.08300752937793732, -0.17618070542812347, -0.10881144553422928, 0.027718398720026016, 0.10157683491706848, -0.13341467082500458, 0.17750395834445953, 0.0053311679512262344, 0.14194698631763458, -0.06230149045586586, 0.17821498215198517, -0.2920299768447876, -0.052992161363363266, -0.11403021216392517, 0.22683246433734894, 0.02388901077210903, 0.06089262291789055, 0.1892111599445343, -0.04796633869409561, 0.015474945306777954, 0.24556481838226318, -0.10414738953113556, -0.03355506435036659, -0.05473897233605385, 0.21973326802253723, 0.3419252038002014, 0.042103543877601624, -0.1363520622253418, 0.23928897082805634, 0.2469780445098877, 0.21957504749298096, -0.13978244364261627, 0.37925589084625244, -0.11718590557575226, 0.06506086885929108, 0.28737080097198486, 0.04528328403830528, -0.07998504489660263, 0.2675742208957672, 0.09384781867265701, 0.030712014064192772, 0.11529319733381271, -0.07018156349658966, 0.2299199104309082, 0.8423095345497131, -0.24265408515930176, -0.11675264686346054, 0.40150582790374756, 0.009781691245734692, -0.23323564231395721, -0.08031421899795532, 0.2138211578130722, 0.06446327269077301, 0.009682416915893555, 0.2943810522556305, 0.049018606543540955, -0.01730167865753174, 0.31864240765571594, 0.09393873065710068, 0.08764534443616867, -0.14573350548744202, -0.07335031777620316, 0.5390340089797974, -0.26027238368988037, 0.2122848480939865, 0.392304927110672, 0.1399432271718979, 0.3845630884170532, 0.15995614230632782, -0.06687791645526886, 0.5641660690307617, -0.06525702029466629, 0.007685644552111626, 0.2044232338666916, -0.08264811336994171, -0.08975955843925476, 0.3919524550437927, 0.028776004910469055, 0.02241714671254158, 0.30100855231285095, 0.39953985810279846, -0.3488951325416565, -0.16715654730796814, -0.2033414989709854, 0.030660714954137802, 0.1244572252035141, -0.11075616627931595, -0.23745077848434448, 0.005826856475323439, -0.26031723618507385, -0.058110691606998444, -0.18110908567905426, -0.36725741624832153, 0.2537267208099365, -0.051365986466407776, -0.3961370289325714, -0.1772398054599762, -0.5898524522781372, 0.02973158098757267, -0.106915682554245, -0.26929736137390137, 0.051676880568265915, 0.03170190006494522, 0.07564829289913177, 0.22591058909893036, 0.4408133625984192, 0.3618394732475281, 0.10118624567985535, -0.32470789551734924, -0.3104083836078644, -0.20571963489055634, 0.168928325176239, -0.0008587106713093817, 0.26479336619377136, 0.06462644785642624, -0.18335413932800293, 0.4019778072834015, 0.034111566841602325, 0.051967207342386246, 0.18164241313934326, 0.4109756648540497, -0.47856712341308594, -0.2713511884212494, 0.9588332772254944, 0.018587833270430565, -0.07537417113780975, -0.1051437109708786, 0.2092435657978058, -0.24594160914421082, 0.15358887612819672, -0.05406968295574188, -0.02507912740111351, 0.1652688831090927, 0.20228461921215057, 0.011583652347326279, -0.048438575118780136, 0.2834642827510834, 0.2929684817790985, -0.06524450331926346, -0.299843430519104, -0.32373619079589844, -0.8680152893066406, 0.16102389991283417, 0.09295067191123962, -0.27846962213516235, -0.019421739503741264, -0.28995683789253235, 0.17587114870548248, 0.1876770257949829, -0.24990306794643402, 0.2563643455505371, -0.007288736756891012, -0.035454124212265015, -0.28572213649749756, -0.28918272256851196, -0.24124270677566528, -0.023864706978201866, -0.022986534982919693, -0.3951798975467682, 0.19543075561523438, -0.43772903084754944, -0.13601012527942657, 0.16049307584762573, -0.05110668018460274, 0.4187009036540985, 0.07762899994850159, 0.4448586404323578, -0.03417659178376198, 0.505246102809906, 0.15348978340625763, -0.050154708325862885, -0.04782101511955261, 0.2508711814880371, -0.1411208063364029, 0.11360668390989304, -0.07329348474740982, 0.09428660571575165, -0.17538884282112122, 0.6784105896949768, -0.4178822338581085, 0.4013878405094147, 0.011650784872472286, 0.12361098825931549, -0.49177613854408264, -0.0024986991193145514, -0.17377354204654694, 0.14623768627643585, 0.37973955273628235, 0.25357943773269653, -0.1776377409696579, 0.3676301836967468, -0.42262136936187744, 0.026184655725955963, 0.5046990513801575, -0.16471509635448456, -0.07719241827726364, -0.04225656017661095, 0.3168366551399231, -0.04624221473932266, -0.20655196905136108, -0.5365793704986572, 0.06525680422782898, 0.2432873696088791, -0.06352637708187103, 0.14099591970443726, 0.20620766282081604, 0.17869743704795837, 0.07907149195671082, -0.038183920085430145, 0.822300136089325, -0.0023479594383388758, 0.1193787232041359, -0.1942015439271927, -0.261188805103302 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
I might be doing something wrong, but swapping those two gives this error: ``` > with open(path) as f: E IsADirectoryError: [Errno 21] Is a directory: 'datasets/qanta/dummy/mode=first,char_skip=25/2018.4.18/dummy_data-zip-extracted/dummy_data' src/nlp/datasets/qanta/3d965403133687b819905ead4b69af7bcee365865279b2f797c79f809b4490c3/qanta.py:280: IsADirectoryError During handling of the above exception, another exception occurred: ``` So it seems like the directory name is getting passed. Is this not functioning as expected, or is there some caching happening maybe? I deleted the dummy files and re-ran the import script with no changes. I'm digging a bit in with a debugger, but no clear reason yet
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
88
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. I might be doing something wrong, but swapping those two gives this error: ``` > with open(path) as f: E IsADirectoryError: [Errno 21] Is a directory: 'datasets/qanta/dummy/mode=first,char_skip=25/2018.4.18/dummy_data-zip-extracted/dummy_data' src/nlp/datasets/qanta/3d965403133687b819905ead4b69af7bcee365865279b2f797c79f809b4490c3/qanta.py:280: IsADirectoryError During handling of the above exception, another exception occurred: ``` So it seems like the directory name is getting passed. Is this not functioning as expected, or is there some caching happening maybe? I deleted the dummy files and re-ran the import script with no changes. I'm digging a bit in with a debugger, but no clear reason yet
[ 0.05404652655124664, 0.07422345876693726, 0.009659906849265099, 0.17164680361747742, 0.01845763809978962, -0.1827673614025116, 0.4075562357902527, 0.18073609471321106, 0.03177332878112793, 0.24988539516925812, 0.038060907274484634, 0.06830506771802902, -0.06527341157197952, -0.2527972161769867, 0.016633503139019012, 0.15957880020141602, 0.1347908079624176, 0.04466508701443672, -0.2199355810880661, -0.0055109127424657345, -0.3515836298465729, 0.20710352063179016, -0.18735042214393616, 0.14597298204898834, -0.09561678767204285, 0.2706398069858551, -0.22023619711399078, 0.2536095678806305, -0.19941888749599457, -0.5675514936447144, 0.14673180878162384, 0.029544586315751076, 0.1657157838344574, 0.5746882557868958, -0.0001326538185821846, -0.07714255154132843, 0.5401037931442261, -0.1227792501449585, -0.668753981590271, -0.23048843443393707, 0.28792908787727356, -0.15296024084091187, 0.20211370289325714, -0.08125074952840805, 0.2706165611743927, -0.1741899847984314, 0.20436641573905945, -0.06535480171442032, 0.1248997151851654, 0.3529888391494751, 0.019501879811286926, 0.1126963421702385, -0.010982266627252102, 0.07280804961919785, 0.09343849867582321, 0.2516901195049286, -0.09757174551486969, 0.4664696455001831, 0.21144574880599976, 0.0007739223074167967, -0.0459158830344677, -0.015806833282113075, 0.01910911127924919, 0.46090877056121826, 0.3064596652984619, 0.18696454167366028, 0.46506157517433167, 0.10373485833406448, -0.012513875029981136, -0.03786511719226837, 0.6569065451622009, -0.4838959574699402, -0.4344104528427124, -0.0811828002333641, -0.12728415429592133, -0.20948459208011627, 0.38311174511909485, -0.1031869426369667, -0.3200293481349945, 0.09504419565200806, 0.01449093222618103, -0.2616746425628662, -0.035386670380830765, 0.2144245058298111, -0.37360259890556335, 0.3995920419692993, 0.034709542989730835, 0.19331222772598267, 0.07282900810241699, -0.04959544539451599, 0.3031978905200958, -0.34249162673950195, 0.04692405089735985, 0.16503369808197021, 0.09963355213403702, -0.004732599016278982, -0.16642382740974426, -0.06243536248803139, -0.14533266425132751, 0.034462735056877136, -0.4331529140472412, -0.13929222524166107, -0.18535898625850677, 0.21319697797298431, 0.18501053750514984, 0.4487714171409607, 0.2529617249965668, 0.49305298924446106, 0.17895826697349548, 0.1321789026260376, 0.13476191461086273, 0.18852677941322327, -0.007199963554739952, -0.44622087478637695, -0.0033653918653726578, 0.1070360466837883, 0.2182110846042633, -0.2600004971027374, -0.2631749212741852, -0.2860175669193268, -0.4661548435688019, -0.036149077117443085, -0.05770571529865265, 0.2854333519935608, -0.21941138803958893, -0.1112205982208252, 0.06850823014974594, 0.3858867883682251, -0.19950050115585327, 0.05821758881211281, 0.0005530173657462001, 0.03904857113957405, -0.25790175795555115, 0.03631291911005974, 0.09758501499891281, 0.11273707449436188, 0.2136283814907074, -0.32738807797431946, 0.13679832220077515, -0.1455705761909485, 0.07292021065950394, -0.0004475788737181574, 0.1181761622428894, 0.3578861355781555, -0.14977291226387024, 0.19033102691173553, 0.1012113094329834, -0.45273345708847046, -0.25621798634529114, 0.2993796169757843, -0.10430748015642166, -0.4040016829967499, 0.09910107403993607, -0.005106083117425442, -0.4261927008628845, 0.06087459996342659, 0.07882998138666153, -0.39645642042160034, 0.19351816177368164, -0.5260312557220459, 0.06719560921192169, -0.2752555012702942, -0.2804777920246124, -0.12022747844457626, -0.09894001483917236, 0.4806630611419678, -0.34747326374053955, -0.28770050406455994, -0.38061732053756714, -0.04328218847513199, 0.12351769208908081, -0.07172459363937378, -0.12091968208551407, 0.04395796358585358, -0.5625425577163696, 0.018757035955786705, 0.4358152449131012, -0.5859127640724182, -0.32398390769958496, 0.3623734414577484, -0.5026482939720154, -0.19662664830684662, 0.4976196885108948, -0.11811009794473648, -0.01931067928671837, -0.43628931045532227, 0.00029263939359225333, 0.04905888810753822, 0.06989649683237076, -0.041125889867544174, -0.17711284756660461, -0.20519663393497467, -0.025002844631671906, 0.08768326044082642, -0.17620688676834106, 0.10934759676456451, 0.045501891523599625, -0.10576467216014862, 0.4786713719367981, 0.07372569292783737, 0.09647486358880997, -0.0445866696536541, 0.20204515755176544, 0.05049104243516922, -0.041183315217494965, 0.25568610429763794, -0.5695759057998657, 0.1495256870985031, -0.1553897261619568, 0.12621258199214935, -0.24375763535499573, -0.0856456309556961, -0.36877018213272095, 0.06484130024909973, -0.14587122201919556, -0.15639863908290863, -0.1060829758644104, 0.5268378853797913, -0.01650206372141838, -0.06277445703744888, 0.04309321194887161, 0.08777216821908951, -0.14425581693649292, 0.24223819375038147, -0.11042865365743637, 0.005360633134841919, -0.05163919925689697, 0.08585384488105774, -0.0734892338514328, 0.030898265540599823, 0.22429078817367554, -0.40818122029304504, 0.03320874646306038, 0.3498994708061218, 0.2747834324836731, 0.46276605129241943, -0.022555237635970116, -0.13486912846565247, 0.1559964418411255, 0.24798308312892914, 0.3091239333152771, 0.344550758600235, 0.08198753744363785, -0.2899186313152313, -0.019193634390830994, 0.2700665295124054, 0.16893097758293152, 0.3189723491668701, -0.03178762272000313, -0.049406006932258606, 0.04129423946142197, 0.1339268684387207, -0.06262322515249252, -0.2984604239463806, 0.18931908905506134, 0.16936028003692627, 0.5996584892272949, 0.07128743082284927, 0.32798314094543457, 0.025675922632217407, -0.172720804810524, -0.27904796600341797, -0.08578597754240036, -0.000038109763409011066, -0.18876491487026215, -0.08370154350996017, 0.09133226424455643, 0.4056794047355652, 0.9681351184844971, -0.09375602006912231, 0.3328850269317627, 0.033049795776605606, -0.10807796567678452, -0.2747819125652313, 0.28881070017814636, 0.22763454914093018, -0.07494259625673294, 0.17337003350257874, -0.03139890357851982, -0.31840306520462036, -0.04010477289557457, -0.05030520260334015, 0.5000056624412537, 0.14260053634643555, -0.5107773542404175, 0.2243058681488037, -0.2545761168003082, -0.3097941279411316, -0.6035177111625671, -0.3119942247867584, -0.060538940131664276, -0.5021187663078308, -0.09551563113927841, 0.3961227536201477, -0.030483268201351166, 0.20940634608268738, -0.07476487755775452, 0.08459162712097168, -0.4432923197746277, -0.44742220640182495, -0.014093596488237381, 0.00826389528810978, -0.48457765579223633, -0.14611247181892395, 0.33262211084365845, -0.0875559151172638, 0.22779212892055511, -0.23387661576271057, -0.20161931216716766, -0.23572388291358948, -0.2972358167171478, 0.12632721662521362, 0.15794751048088074, 0.4473169147968292, 0.23238010704517365, 0.000206656739464961, 0.46073299646377563, 0.0031135049648582935, 0.1273820847272873, -0.20606102049350739, -0.3891749978065491, -0.15753836929798126, 0.1272813230752945, 0.1342286318540573, -0.1593412607908249, -0.39147770404815674, -0.4792369604110718, -0.03614691272377968, -0.0952894538640976, 0.05918224900960922, 0.12565308809280396, 0.248422309756279, -0.1625204086303711, 0.14527694880962372, 0.06382398307323456, -0.09572579711675644, -0.03304046392440796, 0.08952083438634872, 0.03843162581324577, -0.23889963328838348, -0.0015530497767031193, 0.04427346959710121, 0.0077286348678171635, 0.056634508073329926, -0.0904531255364418, -0.4622609317302704, -0.4703609049320221, 0.19328849017620087, 0.05559973046183586, 0.18048244714736938, -0.10260471701622009, 0.4552403390407562, 0.2974391281604767, 0.13758745789527893, -0.09040072560310364, -0.3450208604335785, 0.024569319561123848, 0.08051814138889313, 0.3078867793083191, 0.15233756601810455, 0.3881361782550812, -0.19256655871868134, 0.8443016409873962, 0.12629827857017517, 0.035267993807792664, 0.08385474234819412, -0.15377230942249298, 0.18944457173347473, -0.08218611776828766, -0.03154900670051575, 0.15041479468345642, -0.08360820263624191, -0.552213191986084, 0.3674585223197937, -0.05712158605456352, -0.23132149875164032, -0.06647635996341705, 0.22735415399074554, -0.14406530559062958, -0.15885131061077118, 0.07420085370540619, -0.19066934287548065, -0.012434421107172966, 0.16560159623622894, 0.0025183851830661297, -0.26383131742477417, -0.21585780382156372, 0.012425274588167667, 0.302284300327301, 0.02733442932367325, 0.08392205834388733, -0.33054378628730774, -0.12841497361660004, -0.14755143225193024, 0.3084339201450348, 0.33007675409317017, 0.41019463539123535, -0.01594269834458828, -0.1134256049990654, 0.2813282907009125, 0.007648846134543419, 0.3506707549095154, 0.06734572350978851, 0.212434321641922, 0.19533084332942963, -0.0024007055908441544, -0.11793316900730133, -0.18610504269599915, -0.07736077904701233, 0.1471853256225586, 0.10939319431781769, 0.18207956850528717, -0.23847314715385437, -0.26433053612709045, 0.20997920632362366, 0.1963241994380951, -0.1552479863166809, -0.21320797502994537, -0.20942910015583038, -0.12904751300811768, -0.1241963654756546, -0.3455806374549866, -0.13917340338230133, -0.06293459236621857, -0.09446270763874054, 0.38782799243927, 0.018096420913934708, -0.18357962369918823, 0.05521760880947113, 0.004763313103467226, 0.30929040908813477, -0.18026426434516907, 0.15318915247917175, 0.36620134115219116, 0.10744740068912506, 0.406026154756546, 0.38970834016799927, -0.18061485886573792, -0.1275581568479538, 0.025755705311894417, 0.09579324722290039, 0.3917337656021118, 0.6246488690376282, 0.04617830738425255, -0.18961088359355927, -0.2906067967414856, -0.21935437619686127, -0.2398291975259781, 0.09601426124572754, 0.40891221165657043, 0.009106487967073917, 0.10071638971567154, -0.28545671701431274, 0.5554717779159546, 0.20811231434345245, -0.3564368486404419, 0.5225518345832825, -0.0035547157749533653, -0.2918550372123718, 0.11549641191959381, -0.051186658442020416, 1.116386890411377, 0.29423409700393677, 0.4239662289619446, 0.36085981130599976, -0.2088661640882492, 0.09465114772319794, 0.11679045855998993, 0.17522218823432922, -0.46067366003990173, -0.02153204195201397, -0.01903378963470459, -0.33305877447128296, 0.23627984523773193, -0.03546988591551781, -0.09988705068826675, 0.40644270181655884, -0.04007881507277489, 0.35672709345817566, 0.1936497688293457, 0.1573721468448639, -0.2964204251766205, 0.02826472744345665, 0.02614477276802063, -0.04065537452697754, 0.2072831243276596, 0.45894956588745117, -0.16293075680732727, 0.03178030252456665, -0.2632966637611389, -0.0773332342505455, -0.03073139302432537, -0.07309716194868088, -0.13402125239372253, 0.07461783289909363, 0.022244365885853767, -0.2519102394580841, -0.2052021622657776, 0.3793676197528839, 0.19289584457874298, -0.22925761342048645, 0.022600343450903893, 0.1731870472431183, -0.1718834787607193, 0.3654879927635193, 0.35210564732551575, -0.04134286195039749, 0.30463114380836487, -0.13484010100364685, -0.13243763148784637, 0.1016063392162323, -0.2835322916507721, -0.1992294192314148, -0.24217508733272552, -0.09257854521274567, 0.22591644525527954, -0.32463282346725464, -0.148344025015831, -0.14957794547080994, -0.08646150678396225, -0.1007794514298439, -0.0521918348968029, 0.1412244439125061, -0.2876558303833008, 0.010073249228298664, -0.26084986329078674, -0.428896963596344, -0.0010189249878749251, 0.5187016129493713, -0.12068388611078262, -0.05585877224802971, 0.892748236656189, -0.2553897798061371, 0.07004021853208542, 0.03783545643091202, 0.16071435809135437, -0.22212928533554077, -0.265175998210907, 0.06714978814125061, -0.18132200837135315, -0.24391566216945648, 0.021687224507331848, 0.5689194202423096, 0.32607439160346985, 0.08179080486297607, -0.3220895528793335, -0.4693025052547455, -0.24841569364070892, 0.006860856898128986, -0.1358155906200409, 0.18076550960540771, -0.20048187673091888, 0.028380583971738815, 0.1821313500404358, -0.11743483692407608, -0.1076882854104042, -0.019981030374765396, 0.11945271492004395, 0.22964152693748474, -0.15153925120830536, 0.051610566675662994, 0.0627187192440033, 0.03995991498231888, -0.08371757715940475, 0.10385262966156006, -0.2190840244293213, 0.05673457682132721, -0.08646614849567413, 0.26456212997436523, 0.02870079316198826, 0.004550865851342678, 0.006185547914355993, -0.32587912678718567, 0.13018324971199036, 0.06811761111021042, -0.02315836027264595, 0.02889181300997734, -0.05095072090625763, 0.04266951233148575, 0.3008274734020233, 0.06145782396197319, 0.029404964298009872, 0.3885272145271301, 0.05780895799398422, 0.2583296597003937, -0.18535061180591583, 0.21913109719753265, -0.2305953949689865, 0.02076648361980915, 0.0404154472053051, 0.10006552189588547, -0.018764376640319824, 0.20925186574459076, 0.13397341966629028, -0.18022216856479645, 0.1182854175567627, -0.10023893415927887, 0.3639490008354187, 0.9179773926734924, -0.1468939185142517, -0.23009653389453888, 0.4129926860332489, -0.09078697860240936, -0.05701108276844025, -0.12697483599185944, 0.07886938750743866, 0.13585904240608215, -0.05844656005501747, 0.3512895405292511, -0.05280698835849762, -0.11743787676095963, 0.347430944442749, 0.21134482324123383, 0.20100295543670654, -0.23741601407527924, -0.10164026916027069, 0.40461286902427673, -0.3055979907512665, 0.4028892517089844, 0.47488725185394287, 0.06230880692601204, 0.20035035908222198, 0.3251497745513916, 0.08612222969532013, 0.5218536853790283, -0.14537419378757477, -0.02890690788626671, 0.1209086924791336, -0.10828809440135956, -0.029848983511328697, 0.2552216351032257, -0.005975879728794098, 0.044362086802721024, 0.3560176491737366, 0.5558944940567017, -0.4765012264251709, -0.10528762638568878, -0.2333439737558365, -0.13780564069747925, 0.016764096915721893, -0.2931417226791382, -0.21884974837303162, -0.03580266237258911, -0.1372567117214203, -0.03259814903140068, -0.198746457695961, -0.19986848533153534, 0.2898297905921936, 0.05615820363163948, -0.2658575773239136, -0.2286197394132614, -0.36382439732551575, 0.019897669553756714, 0.03998113423585892, -0.34758326411247253, 0.2545582056045532, -0.0002545866882428527, 0.1358792930841446, 0.297561377286911, 0.13145600259304047, 0.4875794053077698, 0.1934376358985901, -0.26962271332740784, -0.2617157995700836, -0.28797075152397156, -0.0018293434986844659, -0.07789882272481918, 0.3554931581020355, -0.029869375750422478, -0.09149397909641266, 0.3583313524723053, -0.01570928283035755, -0.005666935816407204, 0.22060039639472961, 0.4076077938079834, -0.2764085829257965, -0.14980651438236237, 0.809942901134491, -0.24034632742404938, -0.21010591089725494, -0.07573114335536957, 0.03368818014860153, -0.45921432971954346, 0.12718051671981812, 0.011043453589081764, -0.14476582407951355, 0.18752117455005646, 0.11137890815734863, -0.03916383907198906, -0.09055349975824356, 0.34214919805526733, 0.40220606327056885, -0.1118350699543953, -0.3300431966781616, -0.19953051209449768, -0.8534030318260193, 0.26502713561058044, -0.05009095370769501, -0.14365620911121368, -0.20968376100063324, -0.07952242344617844, 0.2140342891216278, -0.017104899510741234, -0.11020050942897797, 0.11080669611692429, 0.05849318951368332, 0.18325722217559814, -0.3073976933956146, -0.3775627315044403, -0.1809948980808258, 0.0396498367190361, -0.00847720168530941, -0.5915565490722656, 0.1994027942419052, -0.4175170063972473, -0.15232014656066895, 0.1619682013988495, -0.07171478867530823, 0.3689963221549988, 0.20901578664779663, 0.3645205795764923, -0.1086045429110527, 0.6161584258079529, 0.1779768317937851, -0.14680612087249756, -0.2628197968006134, 0.057859212160110474, -0.09725794941186905, 0.01836315542459488, -0.08927494287490845, 0.17368924617767334, -0.32428979873657227, 0.4118088185787201, -0.37827664613723755, 0.22013387084007263, 0.08509625494480133, -0.0034848612267524004, -0.22224700450897217, 0.18447351455688477, -0.35889050364494324, 0.17099454998970032, 0.3244397044181824, 0.32722944021224976, 0.04074619710445404, 0.32959210872650146, -0.30356165766716003, -0.2424767166376114, 0.6405372619628906, -0.09905644506216049, -0.1643476039171219, 0.1367870718240738, 0.29454505443573, -0.011158742941915989, -0.27710673213005066, -0.5036941766738892, 0.17269277572631836, 0.2413635104894638, 0.06933792680501938, 0.024125991389155388, 0.051933690905570984, 0.2529212534427643, 0.017192360013723373, -0.06566262990236282, 0.9180793762207031, 0.031080404296517372, 0.07061309367418289, 0.09902886301279068, -0.20347800850868225 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
From what I can tell here: https://github.com/huggingface/nlp/blob/master/tests/utils.py#L115 1. `data_url` is the correct http link 2. `path_to_dummy_data` is a directory, which is causing the issue That path comes from `download_dummy_data`, which I think assumes that the data comes from the zip file, but isn't aware of individual files. So it seems like it data manager needs to be aware if the url its getting is for a file or a zip/directory, and pass this information along. This might happen in `download_dummy_data`, but probably better to happen in `download_and_extract`? Maybe a simple check to see if `os.path.basename` returns the dummy data zip filename, if not then join paths with the basename of the url?
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
112
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. From what I can tell here: https://github.com/huggingface/nlp/blob/master/tests/utils.py#L115 1. `data_url` is the correct http link 2. `path_to_dummy_data` is a directory, which is causing the issue That path comes from `download_dummy_data`, which I think assumes that the data comes from the zip file, but isn't aware of individual files. So it seems like it data manager needs to be aware if the url its getting is for a file or a zip/directory, and pass this information along. This might happen in `download_dummy_data`, but probably better to happen in `download_and_extract`? Maybe a simple check to see if `os.path.basename` returns the dummy data zip filename, if not then join paths with the basename of the url?
[ -0.0789022222161293, 0.2855135500431061, 0.0002483105636201799, 0.052754513919353485, 0.05063500627875328, -0.26075419783592224, 0.2143820822238922, 0.27597543597221375, -0.054034050554037094, 0.08017387241125107, 0.2072281539440155, 0.22357366979122162, -0.22146843373775482, -0.03458164632320404, 0.12442849576473236, 0.054476719349622726, -0.10658448934555054, 0.08478714525699615, -0.22504183650016785, 0.1550334095954895, -0.24932251870632172, 0.17533212900161743, 0.0034216304775327444, 0.16723036766052246, 0.015414005145430565, 0.14610615372657776, -0.2237706482410431, 0.20182101428508759, -0.41943100094795227, -0.7930225133895874, 0.3121372163295746, 0.030213430523872375, 0.23527787625789642, 0.3840053677558899, -0.00012361556582618505, -0.09185260534286499, 0.5897997617721558, -0.1559889018535614, -0.733406126499176, -0.42152395844459534, 0.24337336421012878, -0.5072749257087708, 0.27730104327201843, -0.15510261058807373, 0.185713529586792, -0.06349404901266098, 0.29818055033683777, 0.2096101939678192, 0.18837369978427887, 0.4003964364528656, 0.07241325825452805, 0.30033037066459656, 0.0519985631108284, -0.11268042027950287, 0.0085213677957654, 0.17452716827392578, -0.05216129869222641, 0.5873820781707764, 0.20252344012260437, 0.07431633025407791, -0.15627051889896393, -0.07906906306743622, 0.10531989485025406, 0.5026988983154297, 0.15227755904197693, 0.06298036873340607, 0.39901256561279297, 0.046528689563274384, -0.26758721470832825, 0.08663015067577362, 0.6675561666488647, -0.5247458219528198, -0.3587069809436798, -0.07098479568958282, -0.07901224493980408, -0.057430826127529144, 0.38079017400741577, -0.1414194107055664, -0.41265881061553955, 0.09209899604320526, -0.10645019263029099, -0.1712895631790161, -0.031994882971048355, 0.3967050015926361, -0.29836466908454895, 0.4051719903945923, -0.018465658649802208, 0.16227851808071136, 0.1562567800283432, 0.05911758914589882, 0.25665122270584106, -0.17506061494350433, 0.17345483601093292, 0.18063035607337952, 0.10605955868959427, -0.20942452549934387, -0.2895835340023041, -0.2742269039154053, 0.14529183506965637, 0.11145263910293579, -0.30826324224472046, -0.19240978360176086, -0.06261337548494339, 0.22528623044490814, 0.23071354627609253, 0.32411637902259827, 0.17963629961013794, 0.26923829317092896, 0.2589324116706848, 0.30538997054100037, 0.27003729343414307, 0.16195186972618103, 0.007208426017314196, -0.33093082904815674, -0.12101900577545166, 0.1566588133573532, 0.20481820404529572, -0.21207603812217712, -0.19729726016521454, -0.24859660863876343, -0.4048672020435333, -0.12655924260616302, -0.0687747597694397, 0.3116096556186676, -0.19009633362293243, -0.10306048393249512, 0.09242883324623108, 0.3832481801509857, -0.1280384510755539, -0.07070978730916977, 0.028685620054602623, 0.04314412176609039, -0.24823282659053802, 0.028636638075113297, 0.29634830355644226, 0.2202128767967224, 0.4854152500629425, -0.29047390818595886, -0.16111810505390167, -0.05845450237393379, 0.12964653968811035, 0.0434606596827507, -0.04721199348568916, 0.21914850175380707, -0.0758894607424736, 0.18278968334197998, 0.09156704694032669, -0.3037325143814087, -0.2998290956020355, 0.17252498865127563, -0.07228994369506836, -0.40235960483551025, 0.10723482817411423, 0.06545607000589371, -0.5473222732543945, 0.021734042093157768, 0.09589525312185287, -0.16957278549671173, -0.05753936246037483, -0.35714277625083923, 0.047964971512556076, -0.23688282072544098, -0.25407907366752625, -0.15955165028572083, -0.013157341629266739, 0.33312276005744934, -0.1687350571155548, -0.3116174638271332, -0.4004591405391693, -0.2978602647781372, 0.21164411306381226, -0.19554123282432556, -0.11783377826213837, 0.27287039160728455, -0.4967011511325836, 0.06526059657335281, 0.616349458694458, -0.4722347557544708, -0.29971927404403687, 0.5806676745414734, -0.4525260329246521, -0.06820087879896164, 0.26381388306617737, -0.036954596638679504, -0.22122466564178467, -0.29773959517478943, -0.08449303358793259, 0.13490669429302216, 0.026060326024889946, -0.15002338588237762, -0.19787006080150604, -0.3536165654659271, 0.14594636857509613, 0.2388366460800171, -0.08798076212406158, 0.09268982708454132, 0.053085122257471085, -0.07039486616849899, 0.4528980255126953, 0.10668645054101944, 0.06376843899488449, -0.061137739568948746, 0.12078879773616791, -0.19682006537914276, -0.10716619342565536, 0.02618851140141487, -0.584591269493103, 0.1251315325498581, -0.31481266021728516, 0.19763067364692688, -0.18850234150886536, -0.07421603053808212, -0.5055438280105591, -0.018278859555721283, 0.02284945361316204, -0.2564983665943146, 0.009734151884913445, 0.3550093173980713, 0.09080639481544495, -0.049176380038261414, 0.032250527292490005, 0.015585353597998619, -0.2026490569114685, 0.2942970395088196, -0.010931855998933315, 0.06413038820028305, -0.0877615436911583, 0.11537972092628479, 0.2468138188123703, 0.25044649839401245, 0.16801969707012177, -0.4039676785469055, 0.03068997897207737, 0.31766217947006226, 0.19105173647403717, 0.1839984655380249, 0.20126739144325256, -0.06688758730888367, 0.16839489340782166, 0.19110065698623657, 0.08350783586502075, 0.469428151845932, 0.04755910485982895, -0.1295865774154663, 0.045802041888237, 0.20797240734100342, 0.15497945249080658, -0.0021297389175742865, -0.07428152114152908, 0.07928746193647385, 0.3448399305343628, 0.022174596786499023, -0.10134491324424744, -0.25451725721359253, 0.10405268520116806, 0.334933876991272, 0.6334263682365417, -0.006708168890327215, 0.2735133171081543, -0.10973963141441345, -0.14043645560741425, -0.31091761589050293, 0.04697315767407417, 0.14044471085071564, -0.18512596189975739, -0.1090434119105339, 0.0009037916315719485, 0.37554657459259033, 0.8084438443183899, -0.008676969446241856, 0.33784976601600647, 0.09863545745611191, -0.16344335675239563, -0.2317848652601242, 0.24697187542915344, 0.17262277007102966, -0.09168841689825058, 0.1258706897497177, -0.15991592407226562, -0.3018525540828705, -0.14503882825374603, 0.01993461884558201, 0.42337384819984436, 0.31813207268714905, -0.4547550678253174, -0.05683939903974533, -0.29740631580352783, -0.5187105536460876, -0.505165159702301, -0.31656894087791443, -0.2091602087020874, -0.3562687039375305, -0.05608263984322548, 0.2834823727607727, -0.13092882931232452, 0.16682440042495728, -0.11503655463457108, 0.2502075433731079, -0.32945579290390015, -0.3059852421283722, 0.05414053052663803, -0.09253250062465668, -0.43069520592689514, 0.0013973088935017586, 0.4053304195404053, 0.09085558354854584, 0.2507198750972748, -0.2317570000886917, -0.20343612134456635, -0.2620779871940613, -0.27703630924224854, 0.13477826118469238, 0.1804632693529129, 0.556559145450592, 0.44621336460113525, -0.0005815006443299353, 0.4154551327228546, 0.023447010666131973, 0.02131885476410389, -0.17712309956550598, -0.35947850346565247, -0.16657471656799316, 0.11461679637432098, 0.11637794226408005, -0.17482411861419678, -0.43526527285575867, -0.5285279154777527, -0.05853009968996048, 0.029240572825074196, 0.08637480437755585, 0.15409362316131592, 0.09188555181026459, -0.17155207693576813, 0.18889738619327545, 0.04282142594456673, -0.1449870467185974, -0.15512122213840485, -0.07492940127849579, 0.026332752779126167, -0.14079636335372925, -0.17391948401927948, 0.008674937300384045, -0.20956028997898102, 0.0579383485019207, -0.16223223507404327, -0.40622448921203613, -0.5565921068191528, 0.14096739888191223, 0.18127037584781647, 0.2534227967262268, -0.15074796974658966, 0.37731117010116577, 0.18244048953056335, 0.09180473536252975, -0.13267122209072113, -0.3363354206085205, 0.24598202109336853, 0.10700111091136932, 0.4016033113002777, 0.20600558817386627, 0.29570287466049194, -0.026322178542613983, 0.7256444096565247, 0.14918753504753113, -0.07229521125555038, 0.09984976798295975, -0.1860620379447937, 0.10383355617523193, 0.15628500282764435, 0.005900978576391935, 0.18452078104019165, -0.15817278623580933, -0.26709139347076416, 0.3186193108558655, -0.19768159091472626, -0.04666381701827049, -0.11971629410982132, 0.2846118211746216, -0.2596028745174408, -0.12736397981643677, 0.12735626101493835, -0.058602143079042435, -0.04352840781211853, 0.0980662927031517, -0.04892680421471596, -0.16124743223190308, -0.061358433216810226, 0.12111463397741318, 0.35151761770248413, 0.03328762203454971, 0.14570894837379456, -0.31239357590675354, -0.04856174811720848, -0.3304588496685028, 0.28733181953430176, 0.3822648525238037, 0.429167777299881, 0.058980636298656464, 0.013043485581874847, 0.09527906775474548, 0.0012859517009928823, 0.2868996262550354, 0.019052233546972275, 0.15850690007209778, 0.13101881742477417, -0.1548554003238678, 0.024731826037168503, -0.1552332192659378, -0.09564229100942612, 0.09846913814544678, 0.007518771570175886, 0.1507926732301712, -0.4683268070220947, -0.13879454135894775, 0.19019144773483276, 0.11471991240978241, -0.05802200362086296, -0.22207719087600708, -0.01593109220266342, -0.12120797485113144, -0.04730689153075218, -0.22213192284107208, -0.14633320271968842, -0.009873436763882637, -0.10142938792705536, 0.29026535153388977, 0.07743453979492188, -0.1481546014547348, -0.014837699942290783, 0.08857018500566483, 0.20450171828269958, -0.3280981779098511, 0.11342311650514603, 0.2718532979488373, 0.19918011128902435, 0.2238493263721466, 0.40062472224235535, -0.10000602900981903, -0.21022146940231323, 0.033992569893598557, 0.05974142998456955, 0.32148364186286926, 0.4006623327732086, 0.1043480858206749, -0.03905968368053436, -0.25136008858680725, -0.1566224992275238, -0.3356702923774719, 0.22357909381389618, 0.4120945930480957, -0.10309459269046783, 0.14459405839443207, -0.22911995649337769, 0.6646540760993958, 0.13159412145614624, -0.2383851855993271, 0.5089074373245239, -0.2838892340660095, -0.28065967559814453, -0.07736483216285706, 0.04677686095237732, 1.0281774997711182, 0.2466937005519867, 0.2519371807575226, 0.4927612543106079, -0.26039817929267883, 0.2928633987903595, 0.06944647431373596, 0.09858231246471405, -0.43218106031417847, 0.01908154785633087, -0.06405003368854523, -0.20192989706993103, 0.2541379928588867, 0.032369282096624374, -0.23340952396392822, 0.3385883867740631, 0.13275602459907532, 0.38943061232566833, 0.2429599016904831, 0.40020880103111267, -0.24319526553153992, -0.021532300859689713, -0.23148714005947113, 0.04449529945850372, -0.029153671115636826, 0.35789671540260315, -0.2018134593963623, -0.08174549788236618, -0.20169273018836975, -0.1322373002767563, -0.0011775937164202332, -0.08184593915939331, -0.39228355884552, 0.028829623013734818, 0.06671217083930969, -0.10534287989139557, -0.2333947718143463, 0.2026786059141159, 0.1094043031334877, -0.2765067219734192, -0.13125458359718323, 0.14214494824409485, -0.3156772553920746, 0.258056104183197, 0.3230246305465698, -0.006119485013186932, 0.2491149604320526, -0.3315944969654083, -0.026277165859937668, 0.07291943579912186, -0.3236355185508728, -0.19239948689937592, 0.07747034728527069, -0.21757853031158447, 0.07555178552865982, -0.3103995621204376, -0.2757737338542938, -0.1316888928413391, -0.08827850222587585, -0.11784443259239197, -0.004206841345876455, 0.2224298119544983, -0.35297083854675293, -0.07808355987071991, -0.06195734068751335, -0.2608759105205536, -0.005462296307086945, 0.415506511926651, -0.2048802226781845, -0.0484173521399498, 0.6307445168495178, -0.4179270267486572, 0.09337016195058823, -0.07985346764326096, 0.13358965516090393, -0.15759475529193878, -0.11624132096767426, 0.05989929661154747, -0.08218081295490265, -0.31458526849746704, -0.05027579888701439, 0.5699491500854492, 0.4229191839694977, 0.1599038541316986, -0.18029342591762543, -0.49785053730010986, -0.30219748616218567, 0.040707625448703766, -0.19161413609981537, 0.1929023116827011, -0.08632427453994751, 0.21488751471042633, 0.23817569017410278, 0.000838355568703264, -0.17753930389881134, -0.11061987280845642, -0.04482254013419151, 0.18705309927463531, -0.1764596700668335, 0.028116866946220398, 0.08777710050344467, 0.11093784868717194, -0.05144145339727402, 0.07127979397773743, -0.2663798928260803, -0.03337715193629265, -0.10417533665895462, 0.22602683305740356, 0.15439268946647644, -0.07422839850187302, 0.1739836484193802, -0.14891104400157928, 0.028481382876634598, 0.15573978424072266, -0.1324017345905304, -0.05118018761277199, -0.030449405312538147, 0.11647015064954758, 0.23405690491199493, 0.04433029145002365, -0.20257124304771423, 0.28245559334754944, -0.00937894731760025, 0.15508610010147095, -0.09460033476352692, 0.20901910960674286, -0.3223201632499695, 0.031602002680301666, 0.17144295573234558, 0.09846854209899902, -0.06700091809034348, 0.2769375741481781, 0.13658596575260162, 0.0017012528842315078, 0.002458317670971155, -0.04743044078350067, 0.3074745237827301, 0.7473391890525818, -0.20278292894363403, -0.1288806051015854, 0.4639030396938324, -0.0035370169207453728, -0.21185457706451416, 0.06120771914720535, 0.19137926399707794, 0.004371054470539093, 0.07066086679697037, 0.25566205382347107, 0.04539069905877113, 0.0798012912273407, 0.2536444664001465, 0.056786663830280304, 0.15905019640922546, -0.06254497915506363, -0.055671196430921555, 0.5163024663925171, -0.24758462607860565, 0.3452170789241791, 0.34482479095458984, -0.026771580800414085, 0.3172127902507782, 0.04373566061258316, -0.022251535207033157, 0.5483084321022034, -0.005632856860756874, 0.007813177071511745, 0.12488177418708801, 0.007366970181465149, -0.10260342061519623, 0.2061542123556137, 0.0983365848660469, 0.050279516726732254, 0.25428202748298645, 0.6443095803260803, -0.5301638245582581, -0.19425928592681885, -0.26956501603126526, -0.09844537824392319, 0.08751735836267471, -0.11784806847572327, -0.45354050397872925, -0.008799362927675247, -0.3257638216018677, -0.14106613397598267, -0.13782061636447906, -0.344389945268631, 0.2559642791748047, 0.003879806725308299, -0.34121793508529663, -0.2516854405403137, -0.43340250849723816, 0.13744859397411346, -0.04456023499369621, -0.3266587555408478, 0.2011280357837677, 0.03343760594725609, 0.03124566376209259, 0.2748129665851593, 0.299069344997406, 0.4600577652454376, 0.07712940126657486, -0.28695231676101685, -0.3478586673736572, -0.19405217468738556, 0.07553885132074356, -0.07693608105182648, 0.29652073979377747, 0.006351418327540159, -0.16871340572834015, 0.5234033465385437, 0.04897015169262886, 0.015593276359140873, 0.21652823686599731, 0.42398372292518616, -0.4900086522102356, -0.14891774952411652, 0.7677967548370361, -0.07951367646455765, -0.08629471808671951, -0.05311258137226105, -0.01245995331555605, -0.2715529203414917, 0.06506989151239395, -0.10145223140716553, -0.13405363261699677, 0.22263292968273163, 0.12485460937023163, 0.026110278442502022, -0.05931299179792404, 0.29370978474617004, 0.36874935030937195, -0.1291070431470871, -0.40220576524734497, -0.2841157615184784, -0.973096489906311, 0.07774326205253601, 0.019219042733311653, -0.3939654529094696, -0.0716053918004036, -0.2766568064689636, 0.21714988350868225, 0.11051171272993088, -0.21781666576862335, 0.20729738473892212, 0.24071747064590454, 0.045983269810676575, -0.2624927759170532, -0.31379956007003784, -0.24278892576694489, 0.03638782352209091, -0.04189537093043327, -0.4255925416946411, 0.11773139238357544, -0.22743502259254456, -0.14005208015441895, 0.28931474685668945, -0.040745142847299576, 0.3404310941696167, 0.07629331946372986, 0.28897297382354736, -0.07201764732599258, 0.6170154809951782, 0.10390680283308029, 0.010897784493863583, -0.11753727495670319, 0.21040000021457672, -0.1169169694185257, 0.06729249656200409, -0.07970187813043594, 0.20908226072788239, -0.18039101362228394, 0.6635027527809143, -0.5068441033363342, 0.4374552369117737, 0.04022597149014473, 0.08135387301445007, -0.2613564431667328, 0.10217373818159103, -0.3644089102745056, 0.13158933818340302, 0.35818013548851013, 0.28291061520576477, -0.12177609652280807, 0.30615150928497314, -0.1378706693649292, -0.19781377911567688, 0.39826470613479614, -0.12612152099609375, -0.04153181239962578, 0.04978022351861, 0.2727917730808258, -0.09282983839511871, -0.061109475791454315, -0.4452346861362457, 0.05422221124172211, 0.27372869849205017, 0.029210742563009262, 0.11292879283428192, 0.09593425691127777, 0.15679368376731873, 0.11953683942556381, -0.0622406080365181, 0.874388575553894, 0.04461171105504036, 0.01759820617735386, 0.0011316887103021145, -0.24424117803573608 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
I think the dataset script works correctly. Just the dummy data structure seems to be wrong. I will soon add more commands that should make the create of the dummy data easier. I'd recommend that you won't concentrate too much on the dummy data. If you manage to load the dataset correctly via: ```python # use local path to qanta nlp.load_dataset("./datasets/qanta") ``` then feel free to open a PR and we will look into the dummy data problem together :-) Also please make sure that the Version is in the format 1.0.0 (three numbers separated by two points) - not a date.
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
102
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. I think the dataset script works correctly. Just the dummy data structure seems to be wrong. I will soon add more commands that should make the create of the dummy data easier. I'd recommend that you won't concentrate too much on the dummy data. If you manage to load the dataset correctly via: ```python # use local path to qanta nlp.load_dataset("./datasets/qanta") ``` then feel free to open a PR and we will look into the dummy data problem together :-) Also please make sure that the Version is in the format 1.0.0 (three numbers separated by two points) - not a date.
[ -0.1612340807914734, 0.2551889419555664, 0.016922492533922195, -0.06889186054468155, -0.018732747063040733, -0.0799422487616539, 0.40359100699424744, 0.33871883153915405, 0.07540542632341385, 0.048679303377866745, 0.19280578196048737, 0.11071784049272537, -0.2993966341018677, 0.1028154045343399, 0.10059250146150589, -0.11661358177661896, 0.022554408758878708, 0.22912083566188812, -0.17938725650310516, 0.22091667354106903, -0.2534818649291992, 0.10033542662858963, -0.11082897335290909, 0.17837142944335938, -0.01125146634876728, 0.15829160809516907, -0.28702589869499207, 0.05930349975824356, -0.33930858969688416, -0.7205921411514282, 0.35619986057281494, 0.07208004593849182, 0.2591172456741333, 0.49416384100914, -0.00012244142999406904, -0.07432957738637924, 0.519700288772583, -0.14471903443336487, -0.675597608089447, -0.23255538940429688, 0.24015609920024872, -0.3556255102157593, 0.3359575867652893, 0.001461594714783132, 0.05675801634788513, -0.22579772770404816, 0.247934952378273, 0.11201948672533035, 0.22972865402698517, 0.3637806475162506, 0.13691145181655884, 0.29282674193382263, 0.0548730231821537, -0.12325616180896759, -0.14173874258995056, -0.01133990753442049, -0.11851325631141663, 0.5441452264785767, 0.34197744727134705, 0.033158667385578156, 0.007737208157777786, -0.05453261733055115, 0.08724608272314072, 0.3912864327430725, 0.19234435260295868, 0.00016575551126152277, 0.5235515832901001, -0.04455576464533806, -0.11648388206958771, 0.09077619016170502, 0.7365924715995789, -0.5787573456764221, -0.3857167065143585, -0.00001732414966681972, 0.051615096628665924, -0.11142732203006744, 0.17219942808151245, -0.10055965185165405, -0.2993268668651581, 0.05937664210796356, -0.10987219214439392, -0.2239535003900528, -0.10024017095565796, 0.2735379934310913, -0.21386946737766266, 0.2330450564622879, 0.060237735509872437, 0.22848187386989594, 0.0010027011157944798, -0.10714861750602722, 0.19523288309574127, -0.1300066113471985, -0.05614889785647392, 0.26843056082725525, -0.09157150983810425, -0.18151718378067017, -0.34438666701316833, -0.16297675669193268, 0.13957522809505463, -0.10119520872831345, -0.31825608015060425, -0.1162262111902237, 0.04142532870173454, 0.1830255538225174, 0.1564447432756424, 0.36657196283340454, 0.27113908529281616, 0.20313167572021484, 0.12128400057554245, 0.14138714969158173, 0.27186721563339233, 0.20314207673072815, -0.0669591948390007, -0.353377103805542, -0.012509787455201149, 0.09310347586870193, 0.20516307651996613, -0.2799246609210968, -0.16182009875774384, -0.21086379885673523, -0.3621453642845154, 0.011534121818840504, -0.12015778571367264, 0.32975876331329346, -0.19222189486026764, -0.18938478827476501, 0.07900389283895493, 0.5592824816703796, -0.20093823969364166, -0.2160234898328781, -0.014422201551496983, 0.02494044043123722, -0.21503886580467224, -0.06536278873682022, 0.23748543858528137, 0.009568185545504093, 0.4028274118900299, -0.25579291582107544, -0.2903880178928375, -0.0060512893833220005, 0.23149839043617249, 0.020161759108304977, -0.19140666723251343, 0.25587910413742065, -0.05009007453918457, -0.024772901087999344, 0.10018858313560486, -0.33269739151000977, -0.22785723209381104, 0.2212532013654709, 0.10084784030914307, -0.46670815348625183, -0.2118014097213745, 0.09065297991037369, -0.450385719537735, 0.021674595773220062, 0.017006613314151764, -0.04640205204486847, 0.09855174273252487, -0.3617475926876068, -0.08257236331701279, -0.4229152798652649, -0.29159048199653625, -0.11434746533632278, -0.057408612221479416, 0.3280137777328491, -0.43154069781303406, -0.2729131281375885, -0.4097929298877716, -0.2997346818447113, 0.10333587974309921, -0.26148882508277893, -0.07541269809007645, 0.2706588804721832, -0.39036861062049866, -0.10745123028755188, 0.5073692798614502, -0.5006061792373657, -0.25220558047294617, 0.4450427293777466, -0.39289534091949463, -0.19235138595104218, 0.1796450912952423, 0.03588695824146271, 0.0022721029818058014, -0.29832923412323, -0.12099097669124603, 0.07886017858982086, 0.08520542085170746, -0.13364578783512115, -0.16009731590747833, -0.3697935938835144, 0.13916604220867157, 0.270159512758255, -0.08963329344987869, 0.12558989226818085, 0.021427547559142113, -0.01105976477265358, 0.3364551067352295, 0.08014389127492905, 0.1863800436258316, -0.08931367844343185, 0.11018522083759308, -0.1883333921432495, -0.0934508889913559, -0.011489068157970905, -0.6761333346366882, 0.21052990853786469, -0.2046797275543213, 0.30025070905685425, -0.014471063390374184, 0.03403763845562935, -0.6020734310150146, 0.018951890990138054, -0.08776872605085373, -0.24328796565532684, 0.04227801039814949, 0.19854852557182312, 0.13480216264724731, 0.05541975051164627, 0.005590681917965412, 0.13680900633335114, -0.1502520591020584, 0.2549869120121002, -0.0785129964351654, -0.0289891567081213, -0.11474765092134476, 0.09138873219490051, 0.3021050691604614, 0.2819165885448456, 0.0884689912199974, -0.3099202513694763, 0.052437443286180496, 0.2828422784805298, 0.2346714287996292, 0.14354713261127472, 0.20161665976047516, -0.11179055273532867, 0.08561433106660843, 0.22560012340545654, 0.26149871945381165, 0.3159291744232178, 0.1025577187538147, -0.21182097494602203, -0.0016614483902230859, 0.24940180778503418, 0.2455371469259262, -0.004479769617319107, -0.2242424041032791, 0.05857342481613159, 0.29595938324928284, 0.1056823581457138, -0.10605505108833313, -0.31116771697998047, 0.0934353917837143, 0.3330495059490204, 0.5936198234558105, -0.09914630651473999, 0.19831408560276031, -0.035244639962911606, -0.26504382491111755, -0.3206389844417572, 0.11221975088119507, 0.1166355088353157, -0.28330516815185547, -0.10186287015676498, 0.032512109726667404, 0.27163201570510864, 0.8015950322151184, -0.055095870047807693, 0.0596001073718071, 0.113706573843956, -0.1535860002040863, -0.19464103877544403, 0.19420111179351807, 0.10261213779449463, -0.07959935069084167, 0.1113402247428894, -0.04581394046545029, -0.34773868322372437, -0.03733552619814873, -0.009879067540168762, 0.41767385601997375, 0.4067232310771942, -0.40734076499938965, -0.05698395147919655, -0.15354447066783905, -0.3752649426460266, -0.3857831656932831, -0.3960627019405365, -0.20940040051937103, -0.37914949655532837, -0.061007946729660034, 0.3446575701236725, 0.018215902149677277, 0.2503041923046112, -0.09401390701532364, 0.2637230455875397, -0.20558682084083557, -0.3096754848957062, 0.1600995808839798, -0.1387047916650772, -0.45876526832580566, 0.030165517702698708, 0.24290674924850464, 0.1665126234292984, 0.365800142288208, -0.2689116299152374, -0.21224702894687653, -0.2011978179216385, -0.3196375370025635, 0.14715693891048431, -0.010631361976265907, 0.6500735878944397, 0.4507490396499634, 0.05853479728102684, 0.29792651534080505, 0.11121915280818939, 0.00396703602746129, -0.13633912801742554, -0.3817054331302643, -0.08096275478601456, -0.02113894373178482, 0.06345333904027939, -0.11353909969329834, -0.43938854336738586, -0.5979087948799133, -0.028065716847777367, 0.07417415827512741, 0.11462879925966263, 0.12124387919902802, 0.1300855576992035, -0.015365725383162498, 0.18102000653743744, 0.031908612698316574, -0.1962921917438507, -0.19380247592926025, 0.053101543337106705, 0.08232861012220383, -0.17546862363815308, -0.2817414700984955, 0.1657278686761856, -0.21765875816345215, 0.08913864195346832, -0.26760897040367126, -0.46371662616729736, -0.46481844782829285, 0.24480992555618286, 0.16363143920898438, 0.3045330345630646, -0.20452989637851715, 0.29216697812080383, 0.2518637478351593, 0.17327241599559784, -0.12148860841989517, -0.3563106060028076, 0.20366673171520233, 0.015651805326342583, 0.3223896026611328, 0.19924722611904144, 0.3970176875591278, -0.2766503691673279, 0.7669876217842102, 0.1820708066225052, -0.08122323453426361, -0.03907862305641174, -0.21187689900398254, 0.09913432598114014, -0.0870857685804367, -0.05764676630496979, 0.23883303999900818, -0.13823354244232178, -0.20207838714122772, 0.3131054639816284, -0.17024505138397217, -0.00924388226121664, -0.12770049273967743, 0.23863403499126434, -0.11140123754739761, -0.0631123036146164, 0.05957996845245361, -0.16654394567012787, 0.03780478984117508, 0.09868120402097702, -0.014955833554267883, -0.2617679834365845, -0.19446423649787903, -0.05254652351140976, 0.3353537321090698, 0.00215329322963953, 0.11820770055055618, -0.3188008666038513, 0.011236878111958504, -0.4265466034412384, 0.2786065638065338, 0.2302430123090744, 0.5973353385925293, 0.12704825401306152, 0.08186463266611099, 0.10293969511985779, 0.10305596888065338, 0.2114655077457428, -0.007445040624588728, 0.09969300031661987, 0.21199876070022583, -0.0997578576207161, -0.21547380089759827, -0.24166694283485413, -0.08823610097169876, 0.1525653749704361, -0.04120707884430885, 0.12435618788003922, -0.42245280742645264, -0.06276505440473557, 0.23153835535049438, 0.19613608717918396, -0.18541748821735382, -0.301469087600708, -0.19508272409439087, -0.013291374780237675, -0.10633982717990875, -0.22760526835918427, -0.23696103692054749, 0.07546506077051163, -0.19839756190776825, 0.43460142612457275, 0.026548009365797043, -0.09503401070833206, -0.009466934017837048, 0.22040188312530518, 0.21856385469436646, -0.22556531429290771, 0.3471059203147888, 0.21265459060668945, 0.04422289878129959, 0.19378294050693512, 0.27334198355674744, 0.04669957980513573, -0.26991793513298035, 0.07171909511089325, -0.05596551299095154, 0.38829565048217773, 0.23094025254249573, 0.07861863821744919, -0.000054226482461672276, -0.22865766286849976, -0.15552133321762085, -0.26685723662376404, 0.1135370135307312, 0.3650880753993988, -0.012089106254279613, -0.057567473500967026, -0.32377952337265015, 0.5772800445556641, 0.12792934477329254, -0.22235459089279175, 0.3620983362197876, -0.22821210324764252, -0.23417429625988007, 0.1770772933959961, 0.0581519789993763, 0.892177939414978, 0.16323056817054749, 0.2606757879257202, 0.49016788601875305, -0.22215138375759125, 0.35116279125213623, 0.21474812924861908, 0.11312031000852585, -0.6522515416145325, -0.002710633911192417, 0.03967898339033127, -0.1463421881198883, 0.21416518092155457, -0.14346659183502197, -0.3366466462612152, 0.3871503174304962, 0.004389860201627016, 0.2776869833469391, 0.2860882580280304, 0.310657799243927, -0.17871209979057312, -0.027079835534095764, -0.1167604997754097, 0.08298254013061523, -0.08146379142999649, 0.37291330099105835, -0.23302648961544037, -0.0751851350069046, -0.2692853808403015, -0.06246271729469299, -0.19180411100387573, -0.13464787602424622, -0.16934482753276825, -0.052010249346494675, 0.11992678046226501, -0.11077151447534561, -0.041967377066612244, 0.2312595248222351, 0.24265508353710175, -0.14473354816436768, -0.22735688090324402, 0.05623307824134827, -0.26526400446891785, 0.075464628636837, 0.3290461003780365, 0.06114033982157707, 0.250312864780426, -0.2712115943431854, -0.17359286546707153, 0.10917025804519653, -0.24852506816387177, -0.0686345025897026, -0.14945879578590393, -0.20029316842556, 0.1652701199054718, -0.30027487874031067, -0.3457494378089905, -0.1171412244439125, -0.04909711331129074, -0.09219115227460861, 0.028591465204954147, 0.22841808199882507, -0.3406800925731659, -0.005800157785415649, -0.11406505852937698, -0.26731932163238525, -0.02843252755701542, 0.4541560411453247, -0.029473118484020233, -0.013765047304332256, 0.8067975044250488, -0.09696414321660995, 0.15098273754119873, -0.1248982697725296, 0.064055435359478, -0.0021404135040938854, -0.20873437821865082, 0.007498104125261307, -0.07276108115911484, -0.30545729398727417, 0.06333990395069122, 0.6002620458602905, 0.3502776622772217, -0.006405347026884556, -0.01581365242600441, -0.5503167510032654, -0.4705413281917572, -0.01743685081601143, -0.13372249901294708, 0.10882771760225296, -0.09000879526138306, 0.19828666746616364, 0.2599766254425049, 0.030202098190784454, -0.20928721129894257, -0.13878823816776276, -0.02385501191020012, 0.19493116438388824, -0.15603125095367432, 0.015625176951289177, 0.04091312363743782, 0.10925984382629395, -0.015660665929317474, 0.016997870057821274, -0.39186638593673706, -0.07138147950172424, -0.10062667727470398, 0.20227815210819244, 0.1108470931649208, -0.11428556591272354, 0.0633029192686081, -0.1228523701429367, -0.11143874377012253, 0.19947609305381775, -0.09420663863420486, 0.017175618559122086, -0.10651464760303497, 0.18531425297260284, 0.3287239670753479, 0.05133923143148422, -0.0904780924320221, 0.2512056231498718, 0.10262160003185272, 0.18970070779323578, -0.11579400300979614, 0.21868319809436798, -0.25777673721313477, 0.04440510272979736, 0.12394825369119644, 0.13644693791866302, -0.06430789828300476, 0.31939226388931274, 0.1831648051738739, -0.14177611470222473, 0.14575496315956116, -0.12332583218812943, 0.3411400616168976, 0.7627298831939697, -0.2596038281917572, -0.16030241549015045, 0.4350282847881317, 0.04208822548389435, -0.23856425285339355, 0.005389446392655373, 0.20815496146678925, 0.10267157852649689, 0.053421713411808014, 0.2619725167751312, 0.05471424385905266, -0.08130981773138046, 0.17981931567192078, 0.14861777424812317, 0.11086525768041611, -0.19002754986286163, -0.08867484331130981, 0.504673421382904, -0.29137900471687317, 0.23321588337421417, 0.37283995747566223, 0.042408812791109085, 0.16424599289894104, 0.16576358675956726, 0.05226315185427666, 0.4733753204345703, 0.055274296551942825, 0.05170294642448425, 0.11520903557538986, -0.026024725288152695, -0.004513718653470278, 0.3991515040397644, -0.009431418031454086, 0.008123713545501232, 0.31100350618362427, 0.594642698764801, -0.4088001549243927, -0.15509121119976044, -0.32456767559051514, -0.10484921932220459, 0.03521817922592163, -0.13292890787124634, -0.3802722990512848, 0.0711788460612297, -0.25378596782684326, -0.15523597598075867, -0.15357741713523865, -0.32617029547691345, 0.11912376433610916, 0.0025214990600943565, -0.3347841203212738, -0.23041388392448425, -0.46209484338760376, 0.2961376905441284, -0.1024322584271431, -0.21056650578975677, 0.2456555962562561, 0.06732194125652313, 0.13905905187129974, 0.3560238182544708, 0.3868747055530548, 0.4162226617336273, 0.15023018419742584, -0.4203752875328064, -0.1778593808412552, -0.15420076251029968, 0.10862697660923004, 0.07844631373882294, 0.3482345640659332, -0.01889878511428833, -0.18846814334392548, 0.5514341592788696, 0.06358040124177933, 0.04156536981463432, 0.2617349624633789, 0.43246617913246155, -0.41368380188941956, -0.1912737935781479, 0.7280195355415344, -0.1477658897638321, -0.12769806385040283, -0.11568319797515869, 0.17403192818164825, -0.3391329050064087, 0.18818505108356476, 0.0550403818488121, -0.011934862472116947, 0.2017199993133545, 0.004270872101187706, 0.03773851320147514, -0.17049330472946167, 0.4298184812068939, 0.39635393023490906, -0.030322715640068054, -0.40154916048049927, -0.241623654961586, -0.8262500762939453, 0.19436651468276978, -0.09572435915470123, -0.3178330957889557, -0.037830058485269547, -0.2311190515756607, 0.19117747247219086, 0.1463651955127716, -0.2504749000072479, 0.23906347155570984, -0.015761040151119232, 0.1646001636981964, -0.3513510227203369, -0.3869229853153229, -0.27075454592704773, 0.11317004263401031, -0.08452475070953369, -0.4501943588256836, 0.174981489777565, -0.24662654101848602, -0.09763600677251816, 0.16377083957195282, -0.019515471532940865, 0.45245975255966187, 0.10025571286678314, 0.31207072734832764, -0.04562393203377724, 0.5762502551078796, 0.1779058873653412, -0.04075108841061592, -0.23221510648727417, 0.1957598626613617, -0.14506034553050995, 0.0977168083190918, 0.049104396253824234, 0.1591259241104126, -0.12989266216754913, 0.5070659518241882, -0.36369580030441284, 0.4363592267036438, 0.08669731020927429, 0.12016750872135162, -0.4500046968460083, 0.17375865578651428, -0.22239628434181213, 0.13848957419395447, 0.24804364144802094, 0.26967716217041016, -0.10546994209289551, 0.4141567647457123, -0.1896025389432907, -0.05892498046159744, 0.42898741364479065, -0.20631760358810425, -0.10361842066049576, 0.06802386045455933, 0.19691073894500732, -0.11535141617059708, -0.1730976402759552, -0.41207343339920044, 0.2724250555038452, 0.245002880692482, -0.062201522290706635, 0.14686477184295654, 0.16532813012599945, 0.2617567181587219, 0.09281004965305328, -0.012294868007302284, 0.8619529604911804, -0.04305972903966904, 0.12368853390216827, -0.07380092889070511, -0.2325073778629303 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
The script loading seems to work fine so I'll work on getting a PR open after a few sanity checks on the data. On version, we currently have it versioned with YYYY.MM.DD scheme so it would be nice to not change that, but will it cause issues?
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
47
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. The script loading seems to work fine so I'll work on getting a PR open after a few sanity checks on the data. On version, we currently have it versioned with YYYY.MM.DD scheme so it would be nice to not change that, but will it cause issues?
[ -0.17673854529857635, 0.2111750990152359, -0.009480219334363937, -0.13179311156272888, 0.0734434649348259, -0.14275360107421875, 0.4176614284515381, 0.2573598623275757, 0.049973078072071075, 0.11905651539564133, 0.22333301603794098, 0.06963493674993515, -0.2887464463710785, 0.08405496925115585, 0.0976942703127861, -0.007681668736040592, 0.06393378973007202, 0.1413363218307495, -0.10444250702857971, 0.2076694518327713, -0.2553172707557678, -0.005777653772383928, -0.05395064130425453, 0.10496213287115097, -0.021723946556448936, 0.12302416563034058, -0.2095181792974472, 0.10340165346860886, -0.43610113859176636, -0.8527144193649292, 0.3802422285079956, 0.10694988071918488, 0.288835734128952, 0.394959032535553, -0.00012843437434639782, -0.05747820809483528, 0.5988144874572754, -0.10965307056903839, -0.625224769115448, -0.1496022492647171, 0.27855560183525085, -0.39122554659843445, 0.35908570885658264, -0.035221166908741, 0.15007977187633514, -0.18746663630008698, 0.21873193979263306, 0.04183859005570412, 0.14435122907161713, 0.28827789425849915, 0.07647917419672012, 0.2813303768634796, 0.06819605082273483, -0.21769149601459503, -0.11556960642337799, 0.0595376193523407, -0.08263542503118515, 0.568021833896637, 0.4420579671859741, 0.16450127959251404, -0.0479075089097023, -0.1347484588623047, 0.0992564707994461, 0.45589518547058105, 0.18909728527069092, -0.04519267752766609, 0.6421712040901184, 0.01721687987446785, -0.14175662398338318, -0.0005505550652742386, 0.9479650855064392, -0.5009497404098511, -0.3974159359931946, 0.05029507726430893, -0.011666958220303059, -0.12136584520339966, 0.2301284372806549, -0.19277647137641907, -0.2900266647338867, 0.04725116863846779, -0.1721954196691513, -0.306246280670166, -0.12384630739688873, 0.30030086636543274, -0.30258920788764954, 0.30529940128326416, 0.10334347933530807, 0.24936030805110931, 0.01884397491812706, -0.11828548461198807, 0.2843584418296814, -0.03635099157691002, -0.014641515910625458, 0.18634480237960815, 0.07852444052696228, -0.2858283519744873, -0.34592482447624207, -0.20418153703212738, 0.11312200874090195, -0.04123414680361748, -0.42331191897392273, -0.18879041075706482, -0.0349556989967823, 0.17621389031410217, 0.22279943525791168, 0.2793656587600708, 0.322689026594162, 0.1964389979839325, 0.24403131008148193, 0.07674892246723175, 0.3281920850276947, 0.26700055599212646, -0.07985171675682068, -0.40303072333335876, 0.016207333654165268, 0.05976143479347229, 0.21140815317630768, -0.3521576523780823, -0.08716639131307602, -0.26098182797431946, -0.3343003988265991, -0.04781261831521988, -0.13767439126968384, 0.254379540681839, -0.11698345839977264, -0.09509827941656113, 0.03040611557662487, 0.42387014627456665, -0.15298603475093842, -0.15913033485412598, -0.010798301547765732, 0.014280112460255623, -0.16778992116451263, -0.01822947897017002, 0.2101246416568756, 0.0889371857047081, 0.31579434871673584, -0.27740415930747986, -0.20177309215068817, 0.00356960273347795, -0.015546661801636219, 0.09739144146442413, -0.19562062621116638, 0.2981411814689636, -0.2475428730249405, -0.021107951179146767, 0.043806567788124084, -0.2993887662887573, -0.20705406367778778, 0.27791765332221985, 0.0965084657073021, -0.45383507013320923, -0.017877884209156036, 0.048682745546102524, -0.5866081118583679, -0.006349837873131037, -0.018573900684714317, -0.11635640263557434, 0.11945313960313797, -0.3823109269142151, 0.01369706355035305, -0.4451650083065033, -0.37953656911849976, -0.10745555907487869, -0.15624383091926575, 0.2226879745721817, -0.3308180570602417, -0.27616390585899353, -0.31168127059936523, -0.34300047159194946, 0.15970581769943237, -0.24245984852313995, -0.09673337638378143, 0.09309771656990051, -0.3562757670879364, -0.21318891644477844, 0.5165299773216248, -0.6141233444213867, -0.3168637156486511, 0.47944459319114685, -0.3933306634426117, -0.3192726671695709, 0.24876056611537933, -0.0207884032279253, 0.09857206791639328, -0.43709495663642883, -0.12158013135194778, -0.015840401872992516, -0.018848072737455368, -0.16953305900096893, -0.2131861448287964, -0.3432087004184723, 0.15771761536598206, 0.2769618332386017, -0.06715557724237442, 0.1459980458021164, 0.16151906549930573, 0.13526900112628937, 0.3999651372432709, 0.1997319757938385, 0.1501396745443344, -0.17016765475273132, 0.2423335611820221, -0.2819081246852875, -0.034802887588739395, -0.06494912505149841, -0.6823078393936157, 0.2038474977016449, -0.08313919603824615, 0.1682872176170349, 0.01756259799003601, -0.010357142426073551, -0.5097715854644775, -0.004371670540422201, 0.0005462606786750257, -0.18655119836330414, -0.07908085733652115, 0.11807793378829956, 0.17096984386444092, -0.06011904776096344, -0.029049653559923172, 0.008818023838102818, -0.21767611801624298, 0.1949048787355423, 0.06572531908750534, 0.021832238882780075, -0.046724554151296616, 0.05512044578790665, 0.20559047162532806, 0.2732929289340973, 0.05787929520010948, -0.3460526168346405, -0.0076025244779884815, 0.27380239963531494, 0.37539586424827576, 0.24678638577461243, 0.14574918150901794, 0.03405623883008957, 0.2007620632648468, 0.2431778460741043, 0.2701939642429352, 0.24104620516300201, 0.013896883465349674, -0.18850558996200562, 0.043078579008579254, 0.1935190111398697, 0.1659841239452362, -0.039776165038347244, -0.16236130893230438, -0.028106141835451126, 0.2738693058490753, 0.16490109264850616, -0.16081160306930542, -0.3247186839580536, 0.01635800302028656, 0.32111093401908875, 0.626117467880249, 0.021359724923968315, 0.22093932330608368, -0.10340210795402527, -0.24081885814666748, -0.36838632822036743, -0.035853974521160126, 0.13243229687213898, -0.1466740220785141, -0.13123944401741028, 0.034572500735521317, 0.2909609079360962, 0.8139573931694031, -0.05824098363518715, 0.14619846642017365, 0.09321124106645584, -0.08680161088705063, -0.2109648883342743, 0.13557201623916626, 0.13352128863334656, 0.0800890177488327, 0.13521067798137665, 0.024450497701764107, -0.33166101574897766, -0.012322158552706242, 0.050647031515836716, 0.4254480302333832, 0.37603074312210083, -0.3850041329860687, -0.14506089687347412, -0.18982532620429993, -0.4163716733455658, -0.46484285593032837, -0.2600250244140625, -0.24017728865146637, -0.32249850034713745, -0.11344274878501892, 0.3259066045284271, 0.01779020205140114, 0.20503245294094086, -0.16928575932979584, 0.2984502911567688, -0.2949589788913727, -0.27069273591041565, 0.12766176462173462, -0.20652702450752258, -0.37653347849845886, -0.005978043656796217, 0.22037045657634735, -0.0590689592063427, 0.4229103624820709, -0.15566520392894745, -0.3003934919834137, -0.15991602838039398, -0.3217743933200836, 0.14323265850543976, 0.1649392545223236, 0.5009920597076416, 0.4258592426776886, -0.019075673073530197, 0.46502485871315, 0.18073034286499023, 0.029476292431354523, -0.19941331446170807, -0.45201048254966736, -0.20943689346313477, -0.013930434361100197, 0.09201017767190933, -0.09755679965019226, -0.3627032935619354, -0.6268526315689087, -0.04943516477942467, 0.09246379882097244, 0.04649152234196663, 0.08684554696083069, 0.07008592784404755, -0.17430447041988373, 0.12629681825637817, 0.06489038467407227, -0.14631518721580505, -0.1392495036125183, 0.12885458767414093, -0.06379919499158859, -0.11443979293107986, -0.2529369294643402, 0.08709780126810074, -0.09722992032766342, -0.022104382514953613, -0.06927567720413208, -0.41635191440582275, -0.4980792701244354, 0.18607769906520844, 0.12020441889762878, 0.3934653103351593, -0.3054564595222473, 0.3518291711807251, 0.2290000319480896, 0.19614912569522858, -0.07248472422361374, -0.34929370880126953, 0.2188253104686737, 0.11768539249897003, 0.21085363626480103, 0.3140124976634979, 0.3968406617641449, -0.17006129026412964, 0.6948562860488892, 0.13153111934661865, -0.17439638078212738, -0.0818142369389534, -0.1846345216035843, 0.0265184473246336, -0.07165735214948654, 0.025269655510783195, 0.2892470955848694, -0.12185027450323105, -0.3842005431652069, 0.2331879734992981, -0.1759001463651657, -0.04469290003180504, -0.18192678689956665, 0.30313539505004883, 0.04550091549754143, -0.02356850914657116, -0.01765950582921505, -0.18742959201335907, 0.06210881844162941, 0.06367649137973785, -0.03822535276412964, -0.16917094588279724, -0.19289416074752808, 0.006161416415125132, 0.34943413734436035, 0.002803509822115302, 0.2017984390258789, -0.2370382845401764, 0.06796196103096008, -0.324955552816391, 0.32468441128730774, 0.3511456251144409, 0.6161115169525146, 0.12353872507810593, 0.00640374468639493, 0.1294785588979721, 0.04525180906057358, 0.1685182899236679, 0.015866152942180634, 0.20254449546337128, 0.1941896677017212, -0.15339437127113342, -0.10331513732671738, -0.27947351336479187, -0.0556245893239975, 0.1073632538318634, -0.014489945024251938, 0.20647211372852325, -0.37307003140449524, -0.1632528305053711, 0.2161456048488617, 0.13574638962745667, -0.10049237310886383, -0.3330315053462982, -0.0852661281824112, -0.017189422622323036, -0.03294232860207558, -0.2879054844379425, -0.2201475203037262, -0.0035187876783311367, -0.17048780620098114, 0.36092492938041687, -0.009871206246316433, -0.1548067033290863, -0.04507386311888695, 0.12206535786390305, 0.14888431131839752, -0.28252142667770386, 0.3064955770969391, 0.2818533778190613, 0.10912103205919266, 0.4292037785053253, 0.24100300669670105, -0.07917208969593048, -0.20654132962226868, 0.062448032200336456, -0.008029955439269543, 0.38697099685668945, 0.36695870757102966, 0.10100652277469635, -0.04046150669455528, -0.1914743036031723, -0.18607081472873688, -0.3599005341529846, 0.1049325168132782, 0.3145603835582733, -0.07839646935462952, -0.08095933496952057, -0.37498557567596436, 0.6471112966537476, 0.18893463909626007, -0.2820153832435608, 0.4349674880504608, -0.29321232438087463, -0.24159406125545502, 0.016749316826462746, 0.11472030729055405, 0.9402292966842651, 0.10115311294794083, 0.17937594652175903, 0.40740200877189636, -0.2569124400615692, 0.27772751450538635, 0.20907986164093018, 0.09308402985334396, -0.5795199871063232, 0.036909569054841995, 0.0066171991638839245, -0.18128736317157745, 0.22857515513896942, -0.1084260493516922, -0.14125226438045502, 0.3231202960014343, 0.06759408116340637, 0.34467124938964844, 0.2996441423892975, 0.3267157971858978, -0.1354597508907318, 0.03873352333903313, -0.1763438731431961, 0.014816279523074627, -0.16323506832122803, 0.3315986692905426, -0.24513894319534302, 0.020707150921225548, -0.20245060324668884, -0.0634012445807457, -0.2631097435951233, -0.20407146215438843, -0.1457936316728592, 0.029359713196754456, 0.05805611237883568, -0.17235910892486572, -0.00357824913226068, 0.26856595277786255, 0.18531247973442078, -0.2835442125797272, -0.11798469722270966, 0.06572867929935455, -0.14783523976802826, 0.1241728812456131, 0.35413286089897156, 0.09129980951547623, 0.3243592381477356, -0.34615588188171387, -0.041703708469867706, 0.07845740020275116, -0.3147275149822235, -0.11034616827964783, -0.16823716461658478, -0.16970397531986237, 0.17790314555168152, -0.35208630561828613, -0.2943039536476135, -0.14130592346191406, 0.02850237302482128, -0.05287932977080345, -0.02120472863316536, 0.2687150239944458, -0.24451075494289398, -0.011231345124542713, -0.13838419318199158, -0.2438735067844391, 0.014476073905825615, 0.44725751876831055, 0.03498436510562897, -0.10985369980335236, 0.6268343329429626, -0.2000741958618164, 0.07830142229795456, -0.08096454292535782, 0.04233285412192345, -0.030404355376958847, -0.13849714398384094, 0.07945191860198975, -0.06528616696596146, -0.30925360321998596, 0.03099931962788105, 0.6317124962806702, 0.45381924510002136, -0.022201132029294968, -0.018441034480929375, -0.41377219557762146, -0.4684457778930664, 0.060229506343603134, -0.13376905024051666, 0.15528957545757294, -0.1445983350276947, 0.18821322917938232, 0.10624844580888748, -0.04616064578294754, -0.1483946442604065, -0.1558258980512619, 0.17581871151924133, 0.22977714240550995, -0.18760889768600464, 0.04707852751016617, -0.02007254585623741, 0.04233887419104576, -0.06394973397254944, -0.004503554664552212, -0.37347763776779175, -0.00042900018161162734, -0.06305183470249176, 0.24354545772075653, 0.10599255561828613, -0.1003904715180397, 0.10395467281341553, -0.19135817885398865, -0.013094630092382431, 0.2529706358909607, -0.0426039919257164, 0.09640387445688248, -0.08035586029291153, 0.10699272900819778, 0.3446796238422394, 0.15482012927532196, -0.008971558883786201, 0.35565701127052307, 0.08667539805173874, 0.3026209771633148, -0.07837709784507751, 0.3181912899017334, -0.25813016295433044, 0.07906094938516617, 0.18990539014339447, 0.15776826441287994, -0.017042608931660652, 0.20191441476345062, 0.1141490787267685, -0.17145690321922302, 0.14508675038814545, -0.04361802339553833, 0.40425533056259155, 0.8169980645179749, -0.23315949738025665, -0.20942115783691406, 0.3954935371875763, -0.006465008947998285, -0.20162184536457062, 0.046727877110242844, 0.2742976248264313, 0.16639940440654755, -0.024561550468206406, 0.31011372804641724, 0.09837272018194199, -0.016525505110621452, 0.13899224996566772, 0.20030361413955688, 0.17520634829998016, -0.25456053018569946, -0.014315756969153881, 0.4361822307109833, -0.19119314849376678, 0.23482510447502136, 0.39259836077690125, 0.10050832480192184, 0.19480393826961517, 0.09433531761169434, 0.08113159984350204, 0.5245571136474609, 0.12521125376224518, 0.10351584106683731, 0.1761607825756073, -0.013916155323386192, 0.026047537103295326, 0.4011744558811188, -0.0035628683399409056, -0.04608440771698952, 0.40405791997909546, 0.6079491376876831, -0.3851248323917389, -0.2014324963092804, -0.29688191413879395, -0.19066813588142395, 0.05135670676827431, -0.13128723204135895, -0.25092726945877075, -0.021365763619542122, -0.1806858330965042, -0.0604485347867012, -0.24239714443683624, -0.28716710209846497, 0.2314751297235489, -0.004475128371268511, -0.27352237701416016, -0.2810951769351959, -0.36880800127983093, 0.1331091672182083, -0.029771286994218826, -0.24174071848392487, 0.28902751207351685, -0.010914476588368416, 0.04701489210128784, 0.39378565549850464, 0.25173360109329224, 0.40285080671310425, 0.13960044085979462, -0.357499361038208, -0.28657132387161255, -0.2061857134103775, 0.0576339028775692, 0.045156870037317276, 0.34859374165534973, -0.003062604693695903, -0.13039417564868927, 0.47411996126174927, 0.029031692072749138, 0.10622157901525497, 0.2274843007326126, 0.35508567094802856, -0.4659959673881531, -0.10022179782390594, 0.8348637223243713, -0.08329629898071289, -0.20123817026615143, 0.04727892950177193, 0.17994146049022675, -0.20123788714408875, 0.049357667565345764, -0.06658048927783966, 0.008686555549502373, 0.1635265350341797, 0.14323142170906067, -0.0007180774118751287, -0.09857775270938873, 0.2886532247066498, 0.3661208748817444, 0.025453556329011917, -0.2843554615974426, -0.21561042964458466, -0.746605396270752, 0.19427801668643951, -0.05663437768816948, -0.3099113404750824, -0.021590501070022583, -0.25138431787490845, 0.23229815065860748, 0.061980441212654114, -0.17731736600399017, 0.18057386577129364, 0.0964602380990982, 0.10784252732992172, -0.1980055272579193, -0.27597418427467346, -0.21555252373218536, 0.02278275415301323, -0.10384821146726608, -0.424919456243515, 0.11889957636594772, -0.357105016708374, -0.1592426598072052, 0.08887874335050583, 0.01766159012913704, 0.46478062868118286, 0.005589491222053766, 0.2521139979362488, -0.08387380093336105, 0.7441820502281189, 0.2783358097076416, -0.06739994883537292, -0.1677694171667099, 0.12654688954353333, -0.14369550347328186, 0.17403094470500946, -0.12162140756845474, 0.08444086462259293, -0.18291990458965302, 0.5333904027938843, -0.41235408186912537, 0.49777358770370483, -0.12797079980373383, 0.06901855021715164, -0.34033215045928955, 0.05988244339823723, -0.2637896239757538, 0.03139052912592888, 0.3083660900592804, 0.19540078938007355, -0.037802476435899734, 0.3024754524230957, -0.22645708918571472, -0.12371459603309631, 0.457865834236145, -0.17120207846164703, -0.03230823948979378, 0.08505822718143463, 0.2551976144313812, 0.05695425719022751, -0.14033232629299164, -0.4825673997402191, 0.18956714868545532, 0.24742940068244934, -0.09605918079614639, 0.12113327533006668, 0.1639314740896225, 0.25603291392326355, 0.014003340154886246, -0.023637430742383003, 1.0229690074920654, -0.11149543523788452, 0.0829375833272934, -0.14923928678035736, -0.2332775890827179 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
> The script loading seems to work fine so I'll work on getting a PR open after a few sanity checks on the data. > > On version, we currently have it versioned with YYYY.MM.DD scheme so it would be nice to not change that, but will it cause issues? It would cause issues for sure for the tests....not sure if it would also cause issues otherwise. I would prefer to keep the same version style as we have for other models. You could for example simply add version 1.0.0 and add a comment with the date you currently use for the versioning. What is your opinion regarding the version here @lhoestq @mariamabarham @thomwolf ?
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
115
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. > The script loading seems to work fine so I'll work on getting a PR open after a few sanity checks on the data. > > On version, we currently have it versioned with YYYY.MM.DD scheme so it would be nice to not change that, but will it cause issues? It would cause issues for sure for the tests....not sure if it would also cause issues otherwise. I would prefer to keep the same version style as we have for other models. You could for example simply add version 1.0.0 and add a comment with the date you currently use for the versioning. What is your opinion regarding the version here @lhoestq @mariamabarham @thomwolf ?
[ -0.07840639352798462, 0.21769832074642181, -0.007994765415787697, -0.2187173217535019, -0.016749747097492218, -0.15013808012008667, 0.45485737919807434, 0.3005516827106476, 0.01723027601838112, 0.0683666542172432, 0.24650821089744568, 0.07947704941034317, -0.34436482191085815, 0.09023833274841309, 0.06199243664741516, 0.004057256039232016, 0.1139812171459198, 0.08718735724687576, -0.06386037915945053, 0.15320907533168793, -0.23857004940509796, -0.04668717831373215, 0.05854009464383125, 0.18519844114780426, 0.010335407219827175, 0.11389530450105667, -0.24757252633571625, 0.0872463583946228, -0.41373854875564575, -0.8389002680778503, 0.3315025866031647, 0.1653556078672409, 0.38253918290138245, 0.42129871249198914, -0.00012696099292952567, -0.09459579735994339, 0.6248884201049805, -0.020553918555378914, -0.6024017930030823, -0.16927875578403473, 0.3652234673500061, -0.3112028241157532, 0.4110463857650757, 0.024780360981822014, 0.16198453307151794, -0.20271022617816925, 0.17064090073108673, 0.16206945478916168, 0.1313154548406601, 0.1334400177001953, 0.06785374879837036, 0.22151775658130646, 0.052436839789152145, -0.1449638158082962, -0.19909559190273285, 0.05197082832455635, -0.0658050999045372, 0.6167840361595154, 0.538267970085144, 0.1772146373987198, -0.021481351926922798, -0.0021334534976631403, 0.1414719820022583, 0.328044056892395, 0.20711073279380798, -0.0733766257762909, 0.7348248362541199, 0.061361029744148254, -0.13098594546318054, 0.008156141266226768, 0.9527409076690674, -0.4113695025444031, -0.5182250142097473, 0.05566876754164696, 0.04883899167180061, -0.1310151219367981, 0.18289019167423248, -0.16959132254123688, -0.1973838061094284, 0.1067793071269989, -0.18128985166549683, -0.32025784254074097, -0.22140276432037354, 0.2812292277812958, -0.2812555134296417, 0.2253076583147049, 0.08337686210870743, 0.24231453239917755, -0.10925696790218353, -0.1724252849817276, 0.260231614112854, -0.09966786950826645, -0.018706679344177246, 0.19757479429244995, 0.1821490377187729, -0.3591313660144806, -0.389056921005249, -0.11087005585432053, 0.14589837193489075, -0.10912835597991943, -0.3968164324760437, -0.27656814455986023, -0.0512404665350914, 0.09083858877420425, 0.25132593512535095, 0.22492338716983795, 0.4186558425426483, 0.11730152368545532, 0.23796430230140686, 0.028204094618558884, 0.3856794238090515, 0.27554261684417725, -0.03827471286058426, -0.31103453040122986, 0.03038773313164711, 0.07220078259706497, 0.18404798209667206, -0.4060976803302765, 0.04705066978931427, -0.23230348527431488, -0.3417428135871887, -0.13881896436214447, -0.11908876895904541, 0.1863187551498413, -0.11592963337898254, 0.000331586372340098, -0.0072271437384188175, 0.3783922791481018, -0.12641966342926025, -0.18677861988544464, -0.029238922521471977, -0.05743896961212158, -0.1304679960012436, -0.020339462906122208, 0.19088758528232574, 0.03255564346909523, 0.289774626493454, -0.2891443967819214, -0.18417130410671234, 0.020452288910746574, -0.08764714002609253, 0.12490967661142349, -0.2652382254600525, 0.2683076858520508, -0.40122702717781067, -0.015372781082987785, -0.07192354649305344, -0.3612964153289795, -0.15017567574977875, 0.2882687449455261, 0.07951976358890533, -0.43441006541252136, -0.015945926308631897, 0.052783094346523285, -0.48552608489990234, -0.012248165905475616, -0.029270147904753685, -0.05822455510497093, 0.14872434735298157, -0.3259884715080261, 0.043540358543395996, -0.4276157021522522, -0.3264010548591614, -0.1250607669353485, -0.20648141205310822, 0.23732160031795502, -0.37796443700790405, -0.24657829105854034, -0.21628819406032562, -0.32622402906417847, 0.05176319554448128, -0.20617209374904633, -0.09009325504302979, 0.034393470734357834, -0.1435721218585968, -0.24087361991405487, 0.5255957841873169, -0.6463603377342224, -0.28754159808158875, 0.501909077167511, -0.2739059329032898, -0.275549978017807, 0.2525528073310852, -0.024602163583040237, 0.1583234965801239, -0.5408316850662231, -0.15352246165275574, -0.10466641932725906, -0.03884086385369301, -0.17798317968845367, -0.33155348896980286, -0.4214683771133423, 0.20058327913284302, 0.334543913602829, -0.14667843282222748, 0.10944321751594543, 0.09240783005952835, 0.07713969796895981, 0.5037010312080383, 0.18341240286827087, 0.051096897572278976, -0.21777094900608063, 0.36563289165496826, -0.2329491376876831, -0.07816815376281738, -0.01453306432813406, -0.720613420009613, 0.18981395661830902, -0.10228998214006424, 0.09577107429504395, 0.13708673417568207, 0.005196824669837952, -0.4359965920448303, -0.023731883615255356, -0.013887777924537659, -0.2627042829990387, -0.08157558739185333, -0.061651427298784256, 0.05206858739256859, -0.029485782608389854, -0.010523254983127117, 0.022948499768972397, -0.20240561664104462, 0.22217707335948944, 0.08579199761152267, -0.0024018369149416685, -0.09218437969684601, 0.047818172723054886, 0.15576086938381195, 0.293927937746048, 0.07502295076847076, -0.302520215511322, -0.020341936498880386, 0.32047200202941895, 0.3806014955043793, 0.2452133297920227, 0.11209164559841156, 0.10057663917541504, 0.29977262020111084, 0.3446674942970276, 0.32449132204055786, 0.14844419062137604, -0.04383357614278793, -0.09661363810300827, -0.002045735949650407, 0.19905777275562286, 0.1993005871772766, -0.1508340835571289, -0.24920277297496796, -0.08458001166582108, 0.18606235086917877, 0.20408257842063904, -0.2294480949640274, -0.29849493503570557, -0.02239740826189518, 0.33672502636909485, 0.5842096209526062, 0.010334302671253681, 0.18908660113811493, -0.035989753901958466, -0.22549119591712952, -0.3813839852809906, -0.11914398521184921, 0.15917089581489563, -0.15317369997501373, -0.17685852944850922, 0.0373910516500473, 0.2746085226535797, 0.6558875441551208, -0.050809234380722046, 0.06169327720999718, 0.07968847453594208, -0.09127505868673325, -0.15820498764514923, 0.13297709822654724, 0.1951868236064911, 0.0797489583492279, 0.0744830071926117, 0.02701413817703724, -0.2232912927865982, 0.05465991422533989, 0.07418319582939148, 0.2964886426925659, 0.30123189091682434, -0.3033636212348938, -0.10878120362758636, -0.14369456470012665, -0.35501861572265625, -0.4484521448612213, -0.2716020941734314, -0.36121997237205505, -0.28391093015670776, -0.04440874233841896, 0.30553245544433594, -0.00804943311959505, 0.16161315143108368, -0.08846748620271683, 0.35138967633247375, -0.30625680088996887, -0.08021820336580276, 0.13091690838336945, -0.16871538758277893, -0.40213847160339355, -0.007545545697212219, 0.16932828724384308, -0.14140769839286804, 0.3990311026573181, -0.07358185946941376, -0.3104704022407532, -0.07311951369047165, -0.4179687798023224, 0.07629513740539551, 0.13683411478996277, 0.5680725574493408, 0.4437895119190216, -0.015858739614486694, 0.4559709429740906, 0.2141166627407074, 0.026539582759141922, -0.18789532780647278, -0.37475988268852234, -0.24569803476333618, 0.002452183049172163, 0.06221345067024231, -0.06288854777812958, -0.34115535020828247, -0.674794614315033, 0.01595262438058853, 0.14174216985702515, 0.041540272533893585, 0.09176651388406754, 0.09137094765901566, -0.20751534402370453, 0.009894615039229393, -0.03349146246910095, -0.12606987357139587, -0.09700209647417068, 0.14260676503181458, -0.11859004944562912, -0.11486148834228516, -0.21240922808647156, 0.09414664655923843, 0.05003815516829491, -0.049713797867298126, -0.15218909084796906, -0.38123002648353577, -0.5387282967567444, 0.27524468302726746, 0.09593893587589264, 0.44321680068969727, -0.3750832974910736, 0.31686413288116455, 0.25746023654937744, 0.21060150861740112, -0.030558103695511818, -0.3124101161956787, 0.2227209508419037, 0.20246177911758423, 0.2387276440858841, 0.34162265062332153, 0.42786845564842224, -0.17156003415584564, 0.6987610459327698, 0.26078903675079346, -0.21202370524406433, -0.23117226362228394, -0.005480967927724123, -0.12855055928230286, -0.09365084022283554, 0.03923335298895836, 0.2602944076061249, -0.11454768478870392, -0.36592742800712585, 0.28301042318344116, -0.07185554504394531, 0.0009104206110350788, -0.17153936624526978, 0.2326364666223526, 0.04826631769537926, -0.025848064571619034, -0.016991743817925453, -0.10419643670320511, 0.09018248319625854, 0.036169566214084625, -0.061962634325027466, -0.163551926612854, -0.24948860704898834, 0.07056630402803421, 0.45116737484931946, -0.007166197523474693, 0.23893795907497406, -0.1418851912021637, 0.14389045536518097, -0.3231489062309265, 0.29190781712532043, 0.36468976736068726, 0.5911158323287964, 0.20548345148563385, -0.06498810648918152, 0.16462929546833038, 0.03949254751205444, 0.16430646181106567, 0.068117655813694, 0.2253868579864502, 0.08400386571884155, -0.23833321034908295, -0.08051186800003052, -0.3215266466140747, 0.0013931782450526953, 0.04301387071609497, 0.06709764152765274, 0.2078157663345337, -0.37192270159721375, -0.22546431422233582, 0.19167442619800568, 0.11681389808654785, -0.14998020231723785, -0.32392776012420654, -0.04554982855916023, 0.13843339681625366, 0.07787808775901794, -0.3222939670085907, -0.3112127482891083, -0.09886188060045242, -0.23056074976921082, 0.35554516315460205, -0.0032918581273406744, -0.15369625389575958, -0.06030772998929024, 0.17602184414863586, 0.12314026057720184, -0.35025739669799805, 0.32314831018447876, 0.3036523163318634, 0.1975908726453781, 0.419914186000824, 0.11987718194723129, -0.12025206536054611, -0.10920149832963943, -0.038991253823041916, -0.09179157018661499, 0.3305351734161377, 0.446627676486969, 0.029408376663923264, -0.07550673186779022, -0.15196508169174194, -0.21442173421382904, -0.3695791959762573, -0.030505428090691566, 0.2852318286895752, -0.049796212464571, -0.06767588108778, -0.3444429636001587, 0.6544520258903503, 0.13295790553092957, -0.35399505496025085, 0.38949471712112427, -0.2603274881839752, -0.20492376387119293, 0.01446897815912962, 0.12303533405065536, 0.9271785616874695, 0.13575279712677002, 0.2261447161436081, 0.3561149537563324, -0.26288023591041565, 0.19106148183345795, 0.2073083519935608, -0.021173398941755295, -0.5951218605041504, 0.07956142723560333, -0.010454891249537468, -0.11937633901834488, 0.2597505748271942, -0.0681867003440857, -0.06725665181875229, 0.3064363896846771, 0.11517385393381119, 0.3037271201610565, 0.27472925186157227, 0.3166316747665405, -0.07457143813371658, 0.12467685341835022, -0.14576420187950134, -0.002733726054430008, -0.1987127810716629, 0.2215563952922821, -0.21738159656524658, -0.059121374040842056, -0.13456182181835175, -0.1353384107351303, -0.1971939206123352, -0.27885547280311584, -0.14172951877117157, 0.002586817368865013, 0.09939520806074142, -0.242243692278862, 0.09887224435806274, 0.207837775349617, 0.21243104338645935, -0.276932030916214, -0.10583090037107468, -0.006890570279210806, -0.09415480494499207, 0.19183121621608734, 0.2326047569513321, 0.11455299705266953, 0.4135674834251404, -0.3656519055366516, -0.029833296313881874, -0.0021501800511032343, -0.30152416229248047, -0.0788201317191124, -0.244759663939476, -0.2292693853378296, 0.20691914856433868, -0.3499625623226166, -0.24334916472434998, -0.07171518355607986, 0.060952845960855484, -0.04963534697890282, -0.016589585691690445, 0.29876700043678284, -0.16764646768569946, 0.002461120719090104, -0.0652824193239212, -0.14836598932743073, 0.003522098297253251, 0.4236813187599182, 0.035341206938028336, -0.12862040102481842, 0.5450727939605713, -0.17769360542297363, -0.012219224125146866, -0.06524535268545151, 0.0422196239233017, -0.056933995336294174, -0.26405856013298035, 0.02484985440969467, -0.0826936736702919, -0.254199355840683, 0.026647139340639114, 0.6445192098617554, 0.5140350461006165, -0.08816120028495789, 0.008510657586157322, -0.31155499815940857, -0.4337584972381592, 0.1188797876238823, -0.14966148138046265, 0.12794028222560883, -0.2016352266073227, 0.19356219470500946, 0.08962821960449219, -0.08079403638839722, -0.16443055868148804, -0.16892485320568085, 0.28729039430618286, 0.16615460813045502, -0.20746676623821259, 0.055749546736478806, -0.05666719749569893, 0.0003574356669560075, -0.07731940597295761, -0.09144643694162369, -0.43206557631492615, -0.009923156350851059, -0.06818331778049469, 0.2238517850637436, 0.09701970219612122, -0.09937217831611633, 0.0750126838684082, -0.18205660581588745, 0.032962292432785034, 0.3337462246417999, -0.029269937425851822, 0.05774608626961708, 0.040878910571336746, 0.2147262990474701, 0.31241950392723083, 0.24340109527111053, -0.02675935998558998, 0.41367068886756897, 0.05288263410329819, 0.2778564989566803, -0.03120342083275318, 0.30528873205184937, -0.23807887732982635, 0.03763607144355774, 0.07074980437755585, 0.19237954914569855, -0.18816961348056793, 0.32301488518714905, 0.04429030418395996, -0.19937720894813538, 0.17659597098827362, 0.008233627304434776, 0.38792884349823, 0.7847015857696533, -0.19644689559936523, -0.3009435832500458, 0.4536358416080475, 0.02780831605195999, -0.12452926486730576, 0.08317089825868607, 0.2623593509197235, 0.19259525835514069, -0.04848894849419594, 0.33698877692222595, 0.14231327176094055, 0.037863150238990784, 0.10244553536176682, 0.23270535469055176, 0.11558347195386887, -0.30714133381843567, -0.014102863147854805, 0.4850962162017822, -0.19111405313014984, 0.25854918360710144, 0.42612940073013306, 0.21961547434329987, 0.042595140635967255, 0.01736043579876423, 0.08998798578977585, 0.4031566083431244, 0.20004688203334808, 0.04172322899103165, 0.21260829269886017, 0.07285215705633163, -0.052006810903549194, 0.35862401127815247, 0.010092625394463539, -0.07322865724563599, 0.4328910708427429, 0.6263272762298584, -0.24576975405216217, -0.27351221442222595, -0.26523229479789734, -0.2573296129703522, 0.11760263890028, -0.15073885023593903, -0.3253522515296936, -0.03138638660311699, -0.17232517898082733, -0.06642258167266846, -0.3928258419036865, -0.23233869671821594, 0.2464117705821991, 0.06092766672372818, -0.24920044839382172, -0.352198988199234, -0.4377790689468384, 0.04229274392127991, 0.06692018359899521, -0.148677259683609, 0.24843694269657135, 0.00396727817133069, 0.0939527377486229, 0.4052651524543762, 0.24692608416080475, 0.24045221507549286, 0.09698589146137238, -0.3979862332344055, -0.33230021595954895, -0.24257369339466095, 0.0760236606001854, 0.16924108564853668, 0.2883457839488983, 0.02029680274426937, -0.1690835952758789, 0.4392831325531006, 0.004951359238475561, 0.17242638766765594, 0.36718448996543884, 0.30420127511024475, -0.41587042808532715, -0.11310689896345139, 0.8669894337654114, -0.010363774374127388, -0.2329440712928772, 0.14805182814598083, 0.19929912686347961, -0.20762725174427032, 0.12630192935466766, -0.041104044765233994, 0.05794123187661171, 0.12840145826339722, 0.15669399499893188, 0.01264475379139185, -0.031988464295864105, 0.18832430243492126, 0.3991011679172516, -0.01953926496207714, -0.2369241565465927, -0.12392780184745789, -0.5654705166816711, 0.07696279138326645, 0.051341038197278976, -0.2140645831823349, -0.018772656098008156, -0.24497385323047638, 0.2046445608139038, 0.06571236252784729, -0.11546369642019272, 0.022159822285175323, -0.050805602222681046, 0.14864738285541534, -0.17918527126312256, -0.19621089100837708, -0.1313892900943756, 0.11998558789491653, -0.11988729238510132, -0.3040076792240143, 0.051378000527620316, -0.31778350472450256, -0.19125236570835114, 0.018191730603575706, 0.0473661795258522, 0.46464958786964417, 0.03857524320483208, 0.13066141307353973, -0.06797123700380325, 0.5171780586242676, 0.3847467303276062, -0.17759005725383759, -0.1369653046131134, 0.10956744104623795, -0.12803304195404053, 0.1817883551120758, -0.13699176907539368, 0.1513928771018982, -0.12285523116588593, 0.42535245418548584, -0.3331279158592224, 0.41835811734199524, -0.22756338119506836, 0.00016316142864525318, -0.41708213090896606, 0.10477394610643387, -0.1155538558959961, -0.11772557348012924, 0.29186394810676575, 0.2757241129875183, -0.07171876728534698, 0.26153165102005005, -0.28110143542289734, -0.07436525076627731, 0.43028655648231506, -0.22998568415641785, 0.04756646975874901, 0.09626569598913193, 0.29810577630996704, 0.1159944236278534, -0.053857237100601196, -0.5310583114624023, 0.12829114496707916, 0.23650725185871124, -0.14698515832424164, 0.022672440856695175, 0.22228990495204926, 0.31165826320648193, -0.06905617564916611, -0.03913317993283272, 1.0794566869735718, -0.17351435124874115, 0.1126837432384491, -0.15644274652004242, -0.1975000500679016 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
Maybe use the YYYY.MM.DD as the config name ? That's what we are doing for wikipedia
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
16
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. Maybe use the YYYY.MM.DD as the config name ? That's what we are doing for wikipedia
[ -0.141180157661438, 0.2805643677711487, -0.010228937491774559, -0.17343877255916595, 0.09962110221385956, -0.056226328015327454, 0.3951389491558075, 0.3567788600921631, 0.13251608610153198, 0.16331030428409576, 0.25443243980407715, 0.1168983206152916, -0.26364395022392273, -0.032296158373355865, 0.08826752007007599, -0.02258087322115898, 0.024319864809513092, 0.15399642288684845, -0.06924323737621307, 0.2538091242313385, -0.15772661566734314, 0.027644295245409012, -0.07621083408594131, 0.003468480659648776, -0.03206659108400345, 0.1495646834373474, -0.12435517460107803, -0.029232658445835114, -0.3976539075374603, -0.8835688233375549, 0.3851564824581146, 0.08030350506305695, 0.15719856321811676, 0.4587191641330719, -0.0001239301636815071, -0.09218396246433258, 0.660364031791687, -0.09545914828777313, -0.7126164436340332, -0.11028579622507095, 0.21409565210342407, -0.4431757926940918, 0.31489893794059753, -0.12270388007164001, 0.07477767765522003, -0.12571509182453156, 0.24232231080532074, 0.002951042726635933, 0.042344335466623306, 0.27795830368995667, 0.11370956897735596, 0.20985428988933563, 0.1447017341852188, -0.20508593320846558, -0.06326285004615784, 0.07268252968788147, -0.12551940977573395, 0.5561836957931519, 0.29881590604782104, 0.08979528397321701, -0.011444400995969772, -0.08872734010219574, 0.08542582392692566, 0.474565714597702, 0.20247617363929749, -0.025697732344269753, 0.5517069697380066, 0.010823414660990238, -0.16354914009571075, -0.0021504529286175966, 0.9606100916862488, -0.5900263786315918, -0.28566688299179077, 0.05598229914903641, 0.061815399676561356, -0.04415784776210785, 0.22932542860507965, -0.1933583766222, -0.3710888922214508, 0.005593670066446066, -0.1138651892542839, -0.3460589647293091, -0.1390753835439682, 0.38633835315704346, -0.28163307905197144, 0.3920338749885559, 0.12043479084968567, 0.20180313289165497, 0.029666505753993988, -0.122710682451725, 0.17662954330444336, -0.03735435754060745, 0.052941229194402695, 0.2122679352760315, 0.05316222831606865, -0.29137110710144043, -0.361217737197876, -0.23290030658245087, 0.1176927462220192, -0.07169977575540543, -0.5114039182662964, -0.21405284106731415, 0.06013592705130577, 0.2136068344116211, 0.11943753063678741, 0.2886674404144287, 0.31643879413604736, 0.2632662057876587, 0.2690868079662323, 0.15910065174102783, 0.2307039201259613, 0.2268163114786148, -0.001236294279806316, -0.39059674739837646, 0.007877267897129059, 0.043516915291547775, 0.23724433779716492, -0.2792249619960785, -0.0018140148604288697, -0.2233661562204361, -0.3548259437084198, -0.03768961504101753, -0.15496304631233215, 0.2712012529373169, -0.21224357187747955, -0.2190592736005783, 0.035491421818733215, 0.3669433295726776, -0.19033758342266083, -0.08297501504421234, -0.030062152072787285, 0.10269130021333694, -0.20946714282035828, 0.04133981838822365, 0.2736009359359741, 0.11987373232841492, 0.381564199924469, -0.22116528451442719, -0.17527347803115845, -0.06266386061906815, 0.04889069125056267, 0.16503962874412537, -0.2584545314311981, 0.31381118297576904, -0.17547930777072906, 0.09148220717906952, 0.07570147514343262, -0.398679256439209, -0.26067492365837097, 0.1905805766582489, 0.08797729760408401, -0.3984543979167938, -0.0753718838095665, 0.09867946058511734, -0.5322722792625427, -0.04728288948535919, 0.03678005933761597, -0.10876550525426865, 0.16949868202209473, -0.43949946761131287, 0.0399811826646328, -0.32962048053741455, -0.32755348086357117, -0.12069281190633774, -0.044775158166885376, 0.22451521456241608, -0.3344631791114807, -0.34485405683517456, -0.39960163831710815, -0.2620830535888672, 0.19108612835407257, -0.27225616574287415, -0.051615409553050995, 0.17989279329776764, -0.3446759283542633, -0.09361156821250916, 0.5451647043228149, -0.5609608292579651, -0.3666340410709381, 0.4103250205516815, -0.39714255928993225, -0.3093583881855011, 0.21600043773651123, -0.0019483502255752683, 0.07214003056287766, -0.29543301463127136, -0.13503055274486542, 0.026698661968111992, 0.054789528250694275, -0.12957771122455597, -0.17708136141300201, -0.31291669607162476, 0.15713384747505188, 0.24047572910785675, -0.05776774138212204, 0.12762784957885742, 0.17676858603954315, 0.2212318629026413, 0.33956876397132874, 0.12165796756744385, 0.19460879266262054, -0.084800586104393, 0.20399156212806702, -0.17558185756206512, -0.07350839674472809, -0.09873486310243607, -0.6286865472793579, 0.21739527583122253, -0.04047866538167, 0.2884501814842224, -0.07102862000465393, -0.05127639323472977, -0.6030595898628235, -0.028186168521642685, 0.00714518129825592, -0.2543843388557434, 0.0009686918929219246, 0.21090392768383026, 0.12575198709964752, 0.07182266563177109, 0.08742720633745193, -0.05625365674495697, -0.20948576927185059, 0.18995793163776398, -0.039935994893312454, -0.016077112406492233, -0.11674582213163376, 0.11367644369602203, 0.12232333421707153, 0.20108278095722198, 0.032386187463998795, -0.2665897309780121, -0.011399648152291775, 0.18859533965587616, 0.3017374575138092, 0.266964852809906, 0.20126543939113617, -0.02864086627960205, 0.18482054769992828, 0.14330708980560303, 0.2356155514717102, 0.31909435987472534, 0.07314877957105637, -0.27954864501953125, 0.0007628898601979017, 0.13875500857830048, 0.2109912633895874, 0.003671384882181883, -0.22517845034599304, -0.008010493591427803, 0.3341831862926483, 0.12936246395111084, -0.1294441670179367, -0.3318292200565338, 0.09353794157505035, 0.3465369939804077, 0.5930154919624329, -0.03065650723874569, 0.3187485933303833, -0.17704841494560242, -0.19611361622810364, -0.3319162130355835, 0.010572312399744987, 0.1472771018743515, -0.1484431028366089, -0.13274501264095306, 0.004782126285135746, 0.26078298687934875, 0.7473021745681763, -0.017133647575974464, 0.21024437248706818, 0.08281274884939194, -0.1943521797657013, -0.18114586174488068, 0.1961626410484314, 0.14606839418411255, 0.058496203273534775, 0.11732576787471771, 0.008353723213076591, -0.35019412636756897, -0.04217161610722542, -0.010637227445840836, 0.46164819598197937, 0.42120227217674255, -0.32619917392730713, -0.12541928887367249, -0.20866435766220093, -0.4111088812351227, -0.45044612884521484, -0.4019864499568939, -0.31500232219696045, -0.3542330265045166, -0.10665392875671387, 0.2702238857746124, 0.08679535984992981, 0.2829831540584564, -0.22471629083156586, 0.3037927448749542, -0.2569120228290558, -0.35868939757347107, 0.0704028531908989, -0.2922024726867676, -0.33515968918800354, 0.011352027766406536, 0.2335166186094284, 0.0032294425182044506, 0.4592943489551544, -0.12466821819543839, -0.3249700963497162, -0.2268054336309433, -0.3663179576396942, 0.1080348864197731, 0.06434562802314758, 0.5244489908218384, 0.4165666699409485, 0.04030127078294754, 0.47716858983039856, 0.04909354820847511, 0.03497636690735817, -0.18719065189361572, -0.4240948557853699, -0.2361035943031311, -0.012633761391043663, 0.17698754370212555, -0.09272029250860214, -0.3785635828971863, -0.5373612642288208, -0.10313188284635544, 0.08804374933242798, 0.06720157712697983, 0.006694186013191938, 0.15287858247756958, -0.1076129749417305, 0.15140707790851593, 0.10712608695030212, -0.055706579238176346, -0.13396614789962769, 0.1255902200937271, 0.020721521228551865, -0.16886906325817108, -0.10974088311195374, 0.10617498308420181, -0.11159364879131317, 0.03736875578761101, -0.1364319622516632, -0.4577692449092865, -0.404999703168869, 0.10826626420021057, 0.16511787474155426, 0.40959516167640686, -0.2527126669883728, 0.29434141516685486, 0.2326202094554901, 0.14786021411418915, -0.09710413962602615, -0.38558489084243774, 0.18258604407310486, 0.03584718331694603, 0.25486353039741516, 0.2382112741470337, 0.4276253283023834, -0.1880529522895813, 0.8638381361961365, 0.12577718496322632, -0.09333564341068268, -0.06046348065137863, -0.22055964171886444, 0.0037173659075051546, -0.059209730476140976, -0.0444650761783123, 0.32291179895401, -0.11662055552005768, -0.3214450776576996, 0.2759306728839874, -0.2046300768852234, -0.1540333777666092, -0.13479170203208923, 0.28695493936538696, -0.0061087110079824924, -0.06751181185245514, -0.044665709137916565, -0.14884798228740692, 0.04530670493841171, 0.05406409874558449, -0.049502212554216385, -0.2289305180311203, -0.25206297636032104, 0.04696153104305267, 0.3166661560535431, -0.15776319801807404, 0.18858551979064941, -0.24663034081459045, 0.031643226742744446, -0.4474697411060333, 0.2941858768463135, 0.3331862688064575, 0.5381494760513306, 0.10278918594121933, -0.015301230363547802, 0.1133810430765152, 0.10491450130939484, 0.20852747559547424, -0.017759518697857857, 0.1344527155160904, 0.24128955602645874, -0.09606413543224335, -0.1246165782213211, -0.25974854826927185, -0.008719860576093197, 0.1897028684616089, -0.004295377992093563, 0.1342666745185852, -0.4184402823448181, -0.07045499980449677, 0.2634959816932678, 0.19597133994102478, -0.09762386232614517, -0.32441040873527527, -0.08676221966743469, -0.028297683224081993, -0.08012966811656952, -0.23945501446723938, -0.16850043833255768, 0.10118187218904495, -0.07583699375391006, 0.36228272318840027, 0.0368487648665905, -0.06702441722154617, -0.08236604183912277, 0.15544487535953522, 0.1467123180627823, -0.2071235626935959, 0.26483219861984253, 0.22875168919563293, 0.02747339941561222, 0.3402649164199829, 0.24137137830257416, -0.11826832592487335, -0.2586962878704071, -0.04757852852344513, 0.007216784171760082, 0.3554086685180664, 0.23109465837478638, 0.18838533759117126, -0.01939680427312851, -0.16011221706867218, -0.14804236590862274, -0.3491378128528595, 0.19839318096637726, 0.32603728771209717, -0.09454573690891266, -0.08157742023468018, -0.37446698546409607, 0.6699296832084656, 0.21066692471504211, -0.22166122496128082, 0.5757384300231934, -0.21588164567947388, -0.36131006479263306, -0.02096301130950451, 0.09729336202144623, 0.9390996098518372, 0.12304692715406418, 0.22184567153453827, 0.3271257281303406, -0.20360638201236725, 0.4068250358104706, 0.11397723853588104, 0.16315177083015442, -0.5522783398628235, 0.03558999300003052, 0.041692398488521576, -0.18077506124973297, 0.18681347370147705, -0.12864026427268982, -0.2766341269016266, 0.34582534432411194, 0.09094031900167465, 0.31623345613479614, 0.2804929316043854, 0.37360402941703796, -0.06282558292150497, -0.04738743230700493, -0.16777780652046204, 0.06752443313598633, -0.17312783002853394, 0.393238365650177, -0.250163733959198, 0.09912167489528656, -0.1918770670890808, -0.038452666252851486, -0.3120807111263275, -0.1937754601240158, -0.12246160954236984, 0.12016146630048752, 0.054419875144958496, -0.16867998242378235, -0.025051789358258247, 0.22702477872371674, 0.12529321014881134, -0.16949136555194855, -0.17709797620773315, 0.14197933673858643, -0.18136003613471985, 0.06414905935525894, 0.30314692854881287, 0.05139261856675148, 0.30210572481155396, -0.3474113941192627, -0.09088580310344696, 0.12196414172649384, -0.28413015604019165, -0.20523521304130554, -0.12637746334075928, -0.17083795368671417, 0.08427350968122482, -0.3267585337162018, -0.28286013007164, -0.1330205351114273, 0.046896763145923615, -0.1448686122894287, 0.02616160735487938, 0.2255212515592575, -0.31772395968437195, 0.0012352591147646308, -0.03404896706342697, -0.34487298130989075, 0.01830502413213253, 0.4467865526676178, -0.08538582921028137, -0.07896202802658081, 0.7237135767936707, -0.16906121373176575, 0.0381770022213459, -0.11864405870437622, 0.02304716408252716, -0.1872938871383667, -0.15366345643997192, 0.04960165545344353, -0.051594141870737076, -0.2729282081127167, 0.010612376034259796, 0.5785097479820251, 0.4408789277076721, 0.04504614323377609, -0.02072584070265293, -0.45889076590538025, -0.41634073853492737, 0.0077952793799340725, -0.18109863996505737, 0.20086121559143066, -0.08203019946813583, 0.18932244181632996, 0.15709957480430603, 0.015954120084643364, -0.1874680519104004, -0.16336217522621155, 0.04029538109898567, 0.2436019629240036, -0.13371004164218903, 0.022527379915118217, -0.0700734332203865, 0.08745704591274261, 0.0018108729273080826, -0.08830492943525314, -0.2857430875301361, -0.05251769348978996, -0.1365341991186142, 0.22036170959472656, 0.06873495876789093, -0.0029470936860889196, 0.08095471560955048, -0.19381675124168396, -0.005588045343756676, 0.16855372488498688, -0.05514528602361679, 0.10774005204439163, -0.09697463363409042, 0.19977757334709167, 0.33400028944015503, 0.04802260547876358, -0.019416028633713722, 0.2950715720653534, 0.11993785947561264, 0.2520240843296051, -0.047951798886060715, 0.33785951137542725, -0.19110265374183655, 0.06257543712854385, 0.17886348068714142, 0.20285284519195557, -0.08617793768644333, 0.17287035286426544, 0.08450338989496231, -0.10913441330194473, 0.10387425869703293, 0.04550354555249214, 0.3643861711025238, 0.7686907649040222, -0.2697174847126007, -0.14736369252204895, 0.43366602063179016, 0.043138351291418076, -0.207034632563591, -0.06151563301682472, 0.3559982478618622, 0.09675069153308868, 0.05956510826945305, 0.2429228574037552, -0.003486514324322343, -0.019756920635700226, 0.1555713415145874, 0.16882428526878357, 0.08332844078540802, -0.22388096153736115, -0.014182274229824543, 0.38112521171569824, -0.20433002710342407, 0.16130255162715912, 0.36529281735420227, 0.12268375605344772, 0.2558838725090027, 0.1170235276222229, 0.12785831093788147, 0.4928237497806549, 0.06652885675430298, 0.034910984337329865, 0.071229949593544, 0.07629580795764923, 0.0614205040037632, 0.2696569561958313, -0.09182707220315933, -0.021174617111682892, 0.37101131677627563, 0.5512710213661194, -0.3462454676628113, -0.13162802159786224, -0.27952107787132263, -0.1425870656967163, 0.04645572975277901, -0.18091298639774323, -0.24025419354438782, 0.013755124062299728, -0.13747720420360565, -0.11026789993047714, -0.15492691099643707, -0.3587169945240021, 0.24380111694335938, 0.008351651020348072, -0.36951687932014465, -0.30901628732681274, -0.21209539473056793, 0.17734505236148834, -0.00655336445197463, -0.31453678011894226, 0.18164925277233124, -0.06015787273645401, 0.009301981888711452, 0.2860340476036072, 0.2964823246002197, 0.45625099539756775, 0.08334717154502869, -0.4143472909927368, -0.25419533252716064, -0.18059153854846954, 0.06383571028709412, -0.02755093201994896, 0.34688565135002136, 0.037008918821811676, -0.1990603655576706, 0.49579232931137085, 0.06062581017613411, 0.035857222974300385, 0.13405029475688934, 0.44267964363098145, -0.5077695846557617, -0.12595975399017334, 0.6570250988006592, -0.06691086292266846, -0.17114847898483276, -0.01440021675080061, 0.268285870552063, -0.31312286853790283, 0.09172555804252625, -0.0007790211820974946, -0.04264144226908684, 0.1423482894897461, 0.07496274262666702, 0.025010205805301666, -0.17087195813655853, 0.4039270877838135, 0.31939437985420227, 0.019597064703702927, -0.2624469995498657, -0.2591545879840851, -0.7384795546531677, 0.1910724639892578, -0.11629606783390045, -0.34262776374816895, -0.07228435575962067, -0.19995787739753723, 0.24219270050525665, 0.04681188985705376, -0.1436450332403183, 0.1019250750541687, 0.029758501797914505, 0.14397871494293213, -0.1850312054157257, -0.31219449639320374, -0.20069940388202667, -0.022074073553085327, -0.06889425963163376, -0.40647295117378235, 0.11685853451490402, -0.36691638827323914, -0.09841734170913696, 0.07443844527006149, 0.0026450196746736765, 0.5323454141616821, 0.0727049708366394, 0.2942674458026886, -0.10212331265211105, 0.8140290975570679, 0.1740230917930603, -0.04890303313732147, -0.17330366373062134, 0.11475560814142227, -0.10631595551967621, 0.16286912560462952, -0.12736451625823975, 0.11199052631855011, -0.20888228714466095, 0.48685672879219055, -0.3890763223171234, 0.6018153429031372, -0.017861103639006615, 0.08193943649530411, -0.28256478905677795, 0.011468818411231041, -0.2549227178096771, 0.08918774873018265, 0.2956138253211975, 0.19525976479053497, -0.07665391266345978, 0.35480573773384094, -0.15306326746940613, -0.25798699259757996, 0.41756871342658997, -0.16063283383846283, -0.11916568130254745, 0.03131844103336334, 0.22146037220954895, -0.012548786588013172, -0.09062953293323517, -0.4313504993915558, 0.19755926728248596, 0.30649811029434204, -0.06527047604322433, 0.03529047966003418, 0.19647490978240967, 0.22549015283584595, 0.09795273840427399, -0.005873700138181448, 0.9912323951721191, -0.10930512845516205, 0.11685183644294739, -0.1647680103778839, -0.256541907787323 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
> Maybe use the YYYY.MM.DD as the config name ? That's what we are doing for wikipedia I'm not sure if this will work because the name should be unique and it seems that he has multiple config name in his data with the same version. As @patrickvonplaten suggested, I think you can add a comment about the version in the data description.
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
63
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. > Maybe use the YYYY.MM.DD as the config name ? That's what we are doing for wikipedia I'm not sure if this will work because the name should be unique and it seems that he has multiple config name in his data with the same version. As @patrickvonplaten suggested, I think you can add a comment about the version in the data description.
[ -0.08671984076499939, 0.273745596408844, 0.011248845607042313, -0.14076071977615356, 0.0627119392156601, -0.04816705361008644, 0.39387497305870056, 0.3804906904697418, 0.08933466672897339, 0.20729294419288635, 0.25848400592803955, 0.14728517830371857, -0.23994077742099762, -0.015442755073308945, 0.04751227796077728, -0.006765051744878292, -0.0585150420665741, 0.21216976642608643, -0.029032966122031212, 0.2738310694694519, -0.084898941218853, -0.01112412754446268, 0.07523125410079956, 0.0901787132024765, -0.04472048208117485, 0.11217653006315231, -0.12271810322999954, -0.06276867538690567, -0.2637859284877777, -0.924858033657074, 0.34500542283058167, 0.12437280267477036, 0.1424257457256317, 0.5053983926773071, -0.00012234803580213338, -0.05318630114197731, 0.6260414719581604, -0.10737137496471405, -0.699290931224823, -0.150202214717865, 0.2308252900838852, -0.4619854986667633, 0.28589946031570435, -0.18482576310634613, 0.034543514251708984, -0.13203270733356476, 0.23104502260684967, 0.050338588654994965, 0.09132016450166702, 0.14613552391529083, 0.1093360185623169, 0.14259831607341766, 0.11537960171699524, -0.16234879195690155, -0.10149726271629333, 0.0278070829808712, -0.09675868600606918, 0.5679475665092468, 0.2544052004814148, 0.13260518014431, 0.01634714938700199, -0.01072561927139759, 0.06912686675786972, 0.454142302274704, 0.13088621199131012, -0.05509152635931969, 0.526400089263916, 0.054601989686489105, -0.12745775282382965, 0.13458876311779022, 0.9771100282669067, -0.5941922664642334, -0.340108186006546, 0.009945991449058056, 0.1404201239347458, -0.003393951104953885, 0.24677704274654388, -0.17631345987319946, -0.36520880460739136, 0.008559276349842548, -0.09733223915100098, -0.2902485132217407, -0.1943589299917221, 0.34757253527641296, -0.2534489333629608, 0.4243464469909668, 0.13284392654895782, 0.21705570816993713, -0.05047548934817314, -0.23413099348545074, 0.14442528784275055, -0.04758596420288086, 0.06462268531322479, 0.12047538161277771, 0.06505198776721954, -0.31929996609687805, -0.47052884101867676, -0.3281736671924591, 0.11238016188144684, -0.17740800976753235, -0.40035566687583923, -0.24110478162765503, 0.06570754200220108, 0.20268851518630981, 0.12156537175178528, 0.19925355911254883, 0.38518908619880676, 0.2224496752023697, 0.2539424002170563, 0.1447163224220276, 0.17460262775421143, 0.20782192051410675, 0.023900508880615234, -0.4101082384586334, 0.0025964260566979647, 0.03073345310986042, 0.2849094867706299, -0.2973402440547943, 0.017244623973965645, -0.1251344382762909, -0.36313822865486145, -0.026332613080739975, -0.21659742295742035, 0.35842540860176086, -0.2512478232383728, -0.22439903020858765, 0.08188780397176743, 0.3014509975910187, -0.10217802226543427, -0.07199326157569885, -0.02329467050731182, 0.024575229734182358, -0.12462446838617325, 0.11025360971689224, 0.24170301854610443, 0.08670426160097122, 0.4142281413078308, -0.229518324136734, -0.15621726214885712, -0.125001922249794, -0.003739507868885994, 0.21193449199199677, -0.23602941632270813, 0.3118695616722107, -0.17207123339176178, 0.0903826430439949, 0.025926275178790092, -0.43858593702316284, -0.3004496395587921, 0.1951528787612915, 0.08729209005832672, -0.34400734305381775, -0.11515448242425919, 0.0968005359172821, -0.55954909324646, -0.08119646459817886, 0.1679890900850296, -0.12055958807468414, 0.13395187258720398, -0.46944504976272583, 0.0365447960793972, -0.2592834532260895, -0.30312782526016235, -0.15861903131008148, -0.0653037577867508, 0.270070344209671, -0.31470558047294617, -0.3589436411857605, -0.3312329351902008, -0.20738713443279266, 0.09957811236381531, -0.3180849254131317, -0.03698398917913437, 0.13217461109161377, -0.2359393835067749, -0.07427416741847992, 0.572056770324707, -0.5332984924316406, -0.3166053295135498, 0.41642817854881287, -0.34503814578056335, -0.24903054535388947, 0.2598501443862915, 0.024156855419278145, 0.16989918053150177, -0.26485475897789, -0.12308786809444427, -0.05039875954389572, 0.12207343429327011, -0.10380498319864273, -0.19102099537849426, -0.41373634338378906, 0.09981698542833328, 0.27179497480392456, -0.10852324217557907, 0.1087997779250145, 0.12809617817401886, 0.19002781808376312, 0.4033348560333252, 0.0872235968708992, 0.12643897533416748, -0.08443984389305115, 0.19085736572742462, -0.09954743832349777, -0.15517735481262207, -0.03792794793844223, -0.6501653790473938, 0.28278541564941406, -0.02253596857190132, 0.22074227035045624, 0.029344309121370316, -0.11142086982727051, -0.5947405099868774, -0.07118973881006241, -0.014744791202247143, -0.22981150448322296, 0.038247235119342804, 0.24519817531108856, 0.040380965918302536, 0.028741950169205666, 0.13230161368846893, -0.06919849663972855, -0.17278480529785156, 0.15835656225681305, -0.05750015005469322, -0.03594789281487465, -0.10883518308401108, 0.10054513812065125, 0.14131931960582733, 0.19206170737743378, 0.08820591121912003, -0.2097431719303131, -0.00825817883014679, 0.1989326924085617, 0.26645469665527344, 0.29947489500045776, 0.17117862403392792, -0.039752498269081116, 0.11774808168411255, 0.16865314543247223, 0.2221967726945877, 0.2807430624961853, 0.0006244280957616866, -0.26477786898612976, 0.012451709248125553, 0.1801752746105194, 0.28095564246177673, -0.021284230053424835, -0.31741616129875183, -0.0282784104347229, 0.31934472918510437, 0.06371920555830002, -0.16110476851463318, -0.2894611060619354, 0.04568502679467201, 0.3787924647331238, 0.558190107345581, -0.052863460034132004, 0.1500062495470047, -0.24769987165927887, -0.24770930409431458, -0.3083849847316742, 0.01466447301208973, 0.09064150601625443, -0.17407093942165375, -0.025309555232524872, 0.05884447693824768, 0.2664114832878113, 0.6372908353805542, 0.0469011552631855, 0.14139960706233978, 0.08129139989614487, -0.23515844345092773, -0.123539499938488, 0.17031624913215637, 0.18290230631828308, 0.126407191157341, 0.14134620130062103, -0.00915686134248972, -0.28563192486763, 0.02741549164056778, -0.03308280184864998, 0.4702056050300598, 0.32481294870376587, -0.3375099003314972, -0.10117241740226746, -0.23019495606422424, -0.3934481143951416, -0.5377240777015686, -0.43573370575904846, -0.4151231646537781, -0.301993191242218, -0.038001175969839096, 0.29891496896743774, 0.02140120044350624, 0.3244086802005768, -0.21723338961601257, 0.2908102869987488, -0.2720581293106079, -0.4005085825920105, 0.05991295725107193, -0.28607839345932007, -0.31616517901420593, 0.008332767523825169, 0.2427157759666443, -0.040145646780729294, 0.47745394706726074, -0.09524847567081451, -0.3994390070438385, -0.2494906485080719, -0.32829493284225464, 0.08962191641330719, 0.0833933874964714, 0.5332787036895752, 0.4108940660953522, 0.11481565237045288, 0.5304835438728333, -0.03753027319908142, 0.017749100923538208, -0.07441163063049316, -0.35614147782325745, -0.23214009404182434, -0.014168757945299149, 0.21951976418495178, -0.05585182458162308, -0.3374412953853607, -0.5183433890342712, -0.1113724634051323, 0.08031614869832993, 0.05806170031428337, 0.07833331823348999, 0.21258750557899475, -0.09202153235673904, 0.10448309034109116, 0.06396188586950302, 0.012659265659749508, -0.13171154260635376, 0.12492069602012634, 0.04393602907657623, -0.1699933409690857, -0.03391916677355766, 0.12768371403217316, -0.06895323842763901, 0.04899919033050537, -0.1415828913450241, -0.43566036224365234, -0.40003150701522827, 0.16557349264621735, 0.1715940535068512, 0.3695300221443176, -0.29879215359687805, 0.25390881299972534, 0.21395662426948547, 0.12991198897361755, -0.10910607129335403, -0.42786070704460144, 0.2279863953590393, 0.10129110515117645, 0.32825008034706116, 0.2778738737106323, 0.3687107563018799, -0.19837813079357147, 0.8586011528968811, 0.20692852139472961, -0.21062274277210236, -0.14095476269721985, -0.24552209675312042, -0.0657978504896164, 0.03593484312295914, 0.013525753282010555, 0.21882039308547974, -0.027355918660759926, -0.3129956126213074, 0.3082881569862366, -0.23085860908031464, -0.1324727088212967, -0.12156078219413757, 0.2083231657743454, -0.02576589025557041, -0.11315485835075378, 0.0600837878882885, -0.14215582609176636, 0.06087641417980194, 0.008019189350306988, -0.06661414355039597, -0.2422121912240982, -0.25388792157173157, 0.02873443067073822, 0.33859914541244507, -0.18706993758678436, 0.1998986303806305, -0.15251682698726654, 0.01847412995994091, -0.4077704846858978, 0.33840230107307434, 0.3403620421886444, 0.547407865524292, 0.12012167274951935, -0.035384401679039, 0.13025228679180145, 0.14203676581382751, 0.20741406083106995, -0.09370391815900803, 0.1443241834640503, 0.2877644896507263, -0.07581449300050735, -0.13675837218761444, -0.3179684579372406, 0.03660057857632637, 0.18235020339488983, 0.0333266519010067, 0.12281104177236557, -0.4709642827510834, -0.09846235066652298, 0.28636637330055237, 0.19446662068367004, -0.16313207149505615, -0.33340024948120117, -0.10241449624300003, 0.048791732639074326, -0.07874704897403717, -0.2163066267967224, -0.24037276208400726, 0.044299300760030746, -0.03885578736662865, 0.30862000584602356, 0.008862723596394062, -0.03533443808555603, -0.0612187497317791, 0.26634860038757324, 0.20314045250415802, -0.20016597211360931, 0.28932344913482666, 0.2616364657878876, 0.08416636288166046, 0.31533554196357727, 0.20172272622585297, -0.08268697559833527, -0.2403508573770523, -0.08250275999307632, 0.017410922795534134, 0.3056637644767761, 0.30834776163101196, 0.2403547316789627, -0.00855755340307951, -0.1876193732023239, -0.1448666900396347, -0.289258748292923, 0.16906285285949707, 0.29359254240989685, -0.1301058828830719, -0.1003357321023941, -0.3282800018787384, 0.7530408501625061, 0.17228107154369354, -0.2961042821407318, 0.5668509006500244, -0.23936691880226135, -0.41096031665802, -0.013560524210333824, 0.10263261944055557, 0.9836562871932983, 0.07682172954082489, 0.21855442225933075, 0.22350861132144928, -0.1969560980796814, 0.39025235176086426, 0.16065636277198792, 0.06136526167392731, -0.5651342272758484, 0.0880422592163086, 0.029159367084503174, -0.09895409643650055, 0.24448086321353912, -0.06411704421043396, -0.35170358419418335, 0.3407963812351227, 0.08218154311180115, 0.2972547709941864, 0.25180670619010925, 0.3703327476978302, -0.04620549827814102, -0.039256803691387177, -0.24055273830890656, 0.05103302374482155, -0.22437357902526855, 0.37193697690963745, -0.21973858773708344, 0.034164924174547195, -0.18881767988204956, -0.07785941660404205, -0.3279077112674713, -0.23957979679107666, -0.1501387059688568, 0.0694773718714714, 0.019197266548871994, -0.18754331767559052, -0.027795149013400078, 0.239035964012146, 0.06307165324687958, -0.12049339711666107, -0.1902083158493042, 0.11415395885705948, -0.20278127491474152, 0.08895660191774368, 0.23756326735019684, -0.017683355137705803, 0.3286876380443573, -0.34493783116340637, -0.11815705895423889, 0.15308457612991333, -0.18728309869766235, -0.23272433876991272, -0.0833960622549057, -0.1945824772119522, 0.04582151770591736, -0.34027716517448425, -0.24017763137817383, -0.07632226496934891, 0.027845434844493866, -0.10413629561662674, 0.03824272379279137, 0.16137807071208954, -0.32575690746307373, 0.017180269584059715, 0.054205406457185745, -0.33293408155441284, 0.046963758766651154, 0.3687140643596649, -0.08637402206659317, -0.0008776349714025855, 0.6735665202140808, -0.24754735827445984, 0.08735307306051254, -0.10800743103027344, 0.00033712037838995457, -0.3327264189720154, -0.13322988152503967, -0.04113619029521942, -0.04393399879336357, -0.2875959277153015, -0.06571567058563232, 0.5596664547920227, 0.5117820501327515, 0.10100788623094559, 0.06241808459162712, -0.38301143050193787, -0.34222716093063354, 0.075409434735775, -0.129617378115654, 0.26565811038017273, -0.0739363506436348, 0.2888072729110718, 0.09227912873029709, 0.04482703655958176, -0.20386141538619995, -0.14279797673225403, 0.01740557700395584, 0.23781272768974304, -0.16256770491600037, 0.12307775020599365, -0.1156911849975586, 0.028433891013264656, -0.014392982237040997, -0.10343105345964432, -0.26048579812049866, -0.04680173099040985, -0.16584137082099915, 0.22512228786945343, 0.03080495446920395, 0.08343648165464401, 0.012772759422659874, -0.18992915749549866, 0.09936905652284622, 0.17254605889320374, 0.004947949666529894, 0.11314799636602402, -0.06754762679338455, 0.27282893657684326, 0.32850587368011475, 0.11845210194587708, -0.12467195838689804, 0.275012731552124, 0.10709898918867111, 0.21082906424999237, -0.006237783934921026, 0.3608352243900299, -0.2627812922000885, 0.036621373146772385, 0.1833154261112213, 0.18572746217250824, -0.20793026685714722, 0.13544900715351105, 0.0163811594247818, -0.04313158988952637, 0.03326166421175003, 0.1499873548746109, 0.3344956040382385, 0.7781229019165039, -0.31009310483932495, -0.1703513115644455, 0.45299527049064636, 0.06668009608983994, -0.19272927939891815, -0.051059335470199585, 0.33120468258857727, 0.04032472148537636, 0.07683945447206497, 0.2602328062057495, 0.011644672602415085, -0.06465549767017365, 0.1997528225183487, 0.1704651117324829, -0.010798294097185135, -0.2592611312866211, 0.08984925597906113, 0.3426213562488556, -0.18214958906173706, 0.16570408642292023, 0.3916594684123993, 0.14826297760009766, 0.2784763276576996, 0.007382862735539675, 0.13599705696105957, 0.4200582802295685, 0.041429661214351654, -0.07995530962944031, 0.09235914796590805, 0.11966153979301453, 0.10311797261238098, 0.27148836851119995, -0.11963281780481339, 0.038716547191143036, 0.40465444326400757, 0.5331208109855652, -0.28116029500961304, -0.14102189242839813, -0.3413611352443695, -0.1561206728219986, 0.06328210979700089, -0.20392915606498718, -0.31007611751556396, 0.02806445024907589, -0.12455962598323822, -0.12496007233858109, -0.17175225913524628, -0.37852680683135986, 0.26906657218933105, 0.06951320916414261, -0.37244588136672974, -0.333720862865448, -0.21656247973442078, 0.24834291636943817, 0.09708008170127869, -0.35574087500572205, 0.1587713658809662, -0.023697778582572937, -0.03222324699163437, 0.29715946316719055, 0.3373473286628723, 0.4456098973751068, -0.01678207702934742, -0.3285938501358032, -0.2467666119337082, -0.16939695179462433, 0.09490437805652618, 0.03207847476005554, 0.20311395823955536, 0.010795685462653637, -0.30086565017700195, 0.550868570804596, 0.07143070548772812, 0.05135597661137581, 0.1238614022731781, 0.3914083242416382, -0.4377101957798004, -0.17570704221725464, 0.5832920670509338, -0.017454519867897034, -0.18201778829097748, 0.005244133062660694, 0.24484995007514954, -0.3289758265018463, 0.18183214962482452, 0.021824520081281662, 0.04144603759050369, 0.14532457292079926, 0.08675654232501984, 0.021520040929317474, -0.11364626884460449, 0.4138195812702179, 0.28322020173072815, -0.067356176674366, -0.26395875215530396, -0.22315619885921478, -0.7593225240707397, 0.180171936750412, -0.1377023160457611, -0.370753675699234, 0.0003883876488544047, -0.21278594434261322, 0.2463700920343399, 0.03339163586497307, -0.11697360128164291, -0.017687400802969933, -0.004546359647065401, 0.18568411469459534, -0.23927155137062073, -0.35890623927116394, -0.13322201371192932, 0.014740511775016785, 0.013599262572824955, -0.3658693730831146, 0.11531128734350204, -0.2808385193347931, -0.11550098657608032, 0.06102651357650757, 0.012041691690683365, 0.5453344583511353, 0.053664542734622955, 0.24777285754680634, -0.16881847381591797, 0.7727631330490112, 0.15961353480815887, -0.06455005705356598, -0.15310174226760864, 0.09544995427131653, -0.13104121387004852, 0.14249837398529053, -0.08216863125562668, 0.129333958029747, -0.1667127162218094, 0.49388009309768677, -0.3666071891784668, 0.5406870245933533, -0.09922441840171814, 0.08082917332649231, -0.221819207072258, -0.03709638491272926, -0.15141688287258148, 0.1198631227016449, 0.22769054770469666, 0.27160727977752686, -0.10734843462705612, 0.3570460081100464, -0.23412708938121796, -0.3536113202571869, 0.4814949333667755, -0.10037027299404144, -0.1550838053226471, 0.021687589585781097, 0.32063207030296326, -0.09507036209106445, 0.07444137334823608, -0.4297248125076294, 0.23768393695354462, 0.32410743832588196, -0.040602680295705795, 0.008705069310963154, 0.2114335149526596, 0.22685372829437256, 0.08315151929855347, 0.0017670484958216548, 1.0600346326828003, -0.10923629999160767, 0.023447591811418533, -0.2541798949241638, -0.311448872089386 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
Actually maybe our versioning format (inherited from tfds) is too strong for what we use it for? We could allow any string maybe? I see it more and more like an identifier for the user that we will back with a serious hashing/versioning system.- so we could let the user quite free on it.
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
54
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. Actually maybe our versioning format (inherited from tfds) is too strong for what we use it for? We could allow any string maybe? I see it more and more like an identifier for the user that we will back with a serious hashing/versioning system.- so we could let the user quite free on it.
[ -0.15275059640407562, 0.21056875586509705, 0.005355939269065857, -0.15212278068065643, 0.1695316880941391, -0.12297339737415314, 0.3444485664367676, 0.3772253096103668, 0.013979267328977585, 0.22968502342700958, 0.1677795946598053, 0.01522478275001049, -0.3342418074607849, 0.031307339668273926, 0.02281864546239376, -0.1002444177865982, -0.059414781630039215, 0.19700652360916138, -0.07303546369075775, 0.34329548478126526, -0.06011231988668442, -0.07316159456968307, 0.001971287652850151, 0.19343025982379913, 0.010818032547831535, 0.14580407738685608, -0.11535069346427917, -0.018754221498966217, -0.3885173201560974, -0.8583573698997498, 0.1343807727098465, 0.09185009449720383, 0.23209898173809052, 0.43688514828681946, -0.00012612351565621793, -0.07396578043699265, 0.5866605043411255, -0.024041924625635147, -0.7674416303634644, -0.16982939839363098, 0.34464311599731445, -0.3756413757801056, 0.253552109003067, -0.06577913463115692, 0.11372017115354538, -0.18384258449077606, 0.18135395646095276, 0.07539646327495575, 0.1361466348171234, 0.24555979669094086, 0.09919435530900955, 0.3106924295425415, 0.02813064679503441, -0.11110108345746994, -0.059569746255874634, 0.12354490906000137, -0.14380046725273132, 0.5142219662666321, 0.4895510971546173, 0.24067041277885437, -0.07860230654478073, -0.07082390040159225, 0.06261537224054337, 0.43598079681396484, 0.1371396780014038, -0.03264617174863815, 0.538478434085846, 0.04423658922314644, -0.19004222750663757, 0.1604461371898651, 0.894517183303833, -0.5086085796356201, -0.46555274724960327, -0.06761395186185837, 0.017939995974302292, -0.10367345064878464, 0.1076577827334404, -0.19388914108276367, -0.37744978070259094, 0.20891009271144867, -0.16695886850357056, -0.332184761762619, -0.09903969615697861, 0.28893008828163147, -0.17620806396007538, 0.23123683035373688, 0.12704646587371826, 0.25155335664749146, -0.02958393655717373, -0.06738045811653137, 0.15492773056030273, -0.05209431052207947, 0.015269262716174126, 0.13405314087867737, 0.08804331719875336, -0.33190926909446716, -0.4149751663208008, -0.29810646176338196, 0.1995217502117157, 0.025635965168476105, -0.5338367223739624, -0.1899031400680542, -0.06855525821447372, 0.1762411743402481, 0.08776427060365677, 0.2744113504886627, 0.36686980724334717, 0.22054262459278107, 0.31181737780570984, -0.010775068774819374, 0.2858647406101227, 0.21933391690254211, -0.005531912669539452, -0.27972158789634705, -0.0023882880341261625, 0.07539422065019608, 0.1061919778585434, -0.2118775099515915, -0.021384794265031815, -0.25974979996681213, -0.42282989621162415, -0.0277204941958189, -0.09297042340040207, 0.23292526602745056, -0.17843341827392578, -0.14660847187042236, -0.019973186776041985, 0.3341646194458008, -0.10098520666360855, -0.19496291875839233, -0.00010885216033784673, -0.020320644602179527, -0.24842937290668488, -0.060860831290483475, 0.22468259930610657, -0.13178715109825134, 0.3598700761795044, -0.35544338822364807, -0.1441565901041031, -0.038140442222356796, -0.022520659491419792, 0.030894549563527107, -0.17065273225307465, 0.2882431745529175, -0.23571926355361938, 0.01707245036959648, 0.0558258593082428, -0.30029040575027466, -0.21881386637687683, 0.21825015544891357, -0.003230085363611579, -0.43608349561691284, -0.13643601536750793, 0.06006936356425285, -0.5637038350105286, 0.021724311634898186, 0.047286294400691986, -0.16956304013729095, 0.18883542716503143, -0.2967648506164551, 0.14172504842281342, -0.2503831386566162, -0.3607434332370758, -0.1515684276819229, -0.17838239669799805, 0.22806371748447418, -0.3107922077178955, -0.2898622453212738, -0.2297290712594986, -0.3846811354160309, 0.07798049598932266, -0.13051722943782806, -0.07232149690389633, 0.20862610638141632, -0.3361049294471741, -0.035918623208999634, 0.6512773633003235, -0.5617482662200928, -0.28832465410232544, 0.6501526236534119, -0.399601012468338, -0.25953438878059387, 0.2737455666065216, -0.07270032912492752, 0.1499107927083969, -0.4444637596607208, -0.13253633677959442, 0.004221390932798386, -0.044506873935461044, -0.11314117163419724, -0.08715642243623734, -0.424634724855423, 0.20988085865974426, 0.20373652875423431, -0.06479071080684662, 0.22341375052928925, 0.12184417992830276, 0.1067122370004654, 0.4471714496612549, 0.035212986171245575, 0.16927528381347656, -0.21021677553653717, 0.2047894299030304, -0.1228681355714798, -0.10746610909700394, -0.13562557101249695, -0.585386335849762, 0.2420368492603302, -0.04733578860759735, 0.1534973382949829, -0.006374089978635311, -0.06405385583639145, -0.44171759486198425, 0.02850278653204441, 0.10575579851865768, -0.21586696803569794, -0.05539095029234886, 0.14584685862064362, 0.02506588026881218, -0.03469611704349518, -0.026319997385144234, -0.00754122156649828, -0.2913225293159485, 0.1625145822763443, 0.20891259610652924, -0.14848405122756958, -0.08332175016403198, 0.0995795950293541, 0.15043047070503235, 0.2522902488708496, -0.0697937160730362, -0.30531641840934753, 0.020699020475149155, 0.3233891725540161, 0.2586059868335724, 0.1494382619857788, 0.07985381036996841, 0.05053427815437317, 0.18621492385864258, 0.2668223977088928, 0.14637485146522522, 0.20671044290065765, 0.024877401068806648, -0.2785700857639313, 0.023999091237783432, 0.1552950143814087, 0.1547406017780304, -0.02467266656458378, -0.21239568293094635, -0.042883988469839096, 0.29666978120803833, 0.08589237928390503, -0.15950730443000793, -0.28861600160598755, 0.10071886330842972, 0.2132440060377121, 0.627163290977478, 0.013069737702608109, 0.2344064563512802, -0.16070246696472168, -0.1911907196044922, -0.41979601979255676, 0.047629304230213165, 0.2504354417324066, -0.08639710396528244, -0.06156637892127037, -0.038702208548784256, 0.2455977350473404, 0.8113173842430115, 0.007162834983319044, 0.2236735075712204, 0.05936557799577713, -0.06221948564052582, -0.20312698185443878, 0.21539635956287384, 0.18741698563098907, -0.019955186173319817, 0.14958515763282776, -0.009534542448818684, -0.2130720466375351, 0.01115147303789854, -0.029949406161904335, 0.3745068609714508, 0.38911446928977966, -0.45985060930252075, -0.07126305252313614, -0.24342291057109833, -0.463784784078598, -0.3513341546058655, -0.32140496373176575, -0.2830216884613037, -0.2886621952056885, -0.07606194168329239, 0.32052522897720337, 0.042950984090566635, 0.2218054085969925, -0.27010688185691833, 0.255303293466568, -0.26923424005508423, -0.2900586724281311, 0.03959781676530838, -0.13836514949798584, -0.37830349802970886, -0.010134385898709297, 0.20340515673160553, -0.1298724263906479, 0.4923700988292694, -0.04010821878910065, -0.2758663296699524, -0.2192285805940628, -0.3464270532131195, 0.17859651148319244, -0.008020536042749882, 0.5385330319404602, 0.5283279418945312, 0.011746515519917011, 0.4922002851963043, 0.045098062604665756, -0.10305880010128021, -0.2824665307998657, -0.4460033178329468, -0.19963937997817993, 0.045424144715070724, 0.12376141548156738, -0.2252001017332077, -0.24200595915317535, -0.5144338607788086, -0.10970614105463028, 0.1445297747850418, -0.03499860689043999, 0.12034542113542557, 0.2705279588699341, -0.1861667037010193, 0.13806486129760742, -0.006836606189608574, -0.06195502355694771, -0.15893426537513733, 0.0721367672085762, -0.09341389685869217, -0.15120366215705872, -0.15034709870815277, 0.05138750001788139, 0.00914393924176693, -0.026123177260160446, -0.035869721323251724, -0.34018945693969727, -0.4658869206905365, 0.15529783070087433, 0.1391378492116928, 0.3482474088668823, -0.22888343036174774, 0.19969873130321503, 0.2338801622390747, 0.16123837232589722, -0.06107689067721367, -0.34528395533561707, 0.14796294271945953, 0.0475279875099659, 0.24498209357261658, 0.26928848028182983, 0.4195292890071869, -0.01958582177758217, 0.7755745649337769, 0.1433252990245819, -0.0958903357386589, -0.07831888645887375, -0.2671537399291992, 0.006282196380198002, -0.054247502237558365, -0.005567108280956745, 0.3108571171760559, -0.10074716806411743, -0.2828232944011688, 0.31856876611709595, -0.18947528302669525, 0.05149240046739578, -0.09044535458087921, 0.26561617851257324, -0.06809332221746445, -0.04419085383415222, 0.050073593854904175, -0.2819506525993347, 0.03933994099497795, -0.01265739556401968, -0.037520136684179306, -0.13320250809192657, -0.21860459446907043, 0.0995267927646637, 0.36235517263412476, -0.01867976039648056, 0.2903425395488739, -0.11583919078111649, 0.1568783074617386, -0.3423962891101837, 0.2570706903934479, 0.4411432147026062, 0.6185426115989685, 0.08319737762212753, 0.09134770184755325, 0.18326568603515625, 0.12001969665288925, 0.227323979139328, -0.03647739067673683, 0.23378126323223114, 0.06326079368591309, -0.23603500425815582, -0.017865899950265884, -0.22646179795265198, 0.0005943626747466624, 0.09605222195386887, -0.07572035491466522, 0.1736181676387787, -0.4774003028869629, -0.23207791149616241, 0.21772131323814392, 0.1053234413266182, -0.07665782421827316, -0.3893098831176758, -0.06505267322063446, 0.03828467056155205, -0.05286506935954094, -0.31292611360549927, -0.13613063097000122, -0.00522266048938036, -0.08902177214622498, 0.433249831199646, -0.10226177424192429, -0.14556972682476044, -0.005133001133799553, 0.10432876646518707, 0.044709961861371994, -0.19614440202713013, 0.20612120628356934, 0.19784873723983765, 0.17465130984783173, 0.21705131232738495, 0.18835365772247314, -0.08402637392282486, -0.1561625748872757, -0.03743548318743706, -0.05730273202061653, 0.3986187279224396, 0.36632469296455383, 0.10062605887651443, 0.042771678417921066, -0.21017621457576752, -0.1530296951532364, -0.3422596752643585, 0.22055666148662567, 0.28233087062835693, -0.16399046778678894, -0.05102985352277756, -0.28409048914909363, 0.6471728086471558, 0.14356403052806854, -0.2456750124692917, 0.4512069821357727, -0.15048614144325256, -0.30368807911872864, -0.025457872077822685, 0.12860232591629028, 0.9857165813446045, 0.043363623321056366, 0.22869525849819183, 0.3851467967033386, -0.3389439284801483, 0.42819687724113464, 0.14748504757881165, 0.0266636423766613, -0.5198192000389099, -0.05517934262752533, -0.05943072959780693, -0.16547302901744843, 0.27415212988853455, -0.15859279036521912, -0.20023567974567413, 0.3456081748008728, 0.087950199842453, 0.3736143410205841, 0.4137504994869232, 0.3992362916469574, -0.10544586926698685, -0.05788305774331093, -0.16087813675403595, 0.022065000608563423, -0.18493744730949402, 0.2843337953090668, -0.2157171070575714, -0.020653011277318, -0.18661019206047058, -0.02854021266102791, -0.24675104022026062, -0.2373766005039215, -0.12553027272224426, 0.017984716221690178, 0.14046046137809753, -0.10922040790319443, 0.05769878625869751, 0.16871440410614014, 0.18284763395786285, -0.09496679902076721, -0.1353597790002823, 0.025371327996253967, -0.2317042499780655, 0.2188095599412918, 0.304460734128952, 0.02827681228518486, 0.3999040424823761, -0.2815256714820862, -0.07076083868741989, 0.14869137108325958, -0.2799606919288635, -0.2452794462442398, -0.17162957787513733, -0.21621999144554138, 0.16090381145477295, -0.3544023334980011, -0.26525357365608215, -0.184625506401062, -0.02043423242866993, -0.03614290803670883, -0.038548678159713745, 0.23290947079658508, -0.330003947019577, -0.013232219964265823, -0.037422340363264084, -0.3219466209411621, 0.03071548417210579, 0.48213693499565125, -0.04637771472334862, -0.17012383043766022, 0.630793571472168, -0.241555854678154, 0.0289763193577528, -0.08974562585353851, 0.03222336620092392, -0.18571686744689941, -0.18621842563152313, 0.038523223251104355, -0.07007140666246414, -0.3210484981536865, -0.034937869757413864, 0.6218964457511902, 0.5002502202987671, 0.026228822767734528, -0.07093372195959091, -0.42394715547561646, -0.47686007618904114, -0.019826892763376236, -0.21596616506576538, 0.1438417136669159, -0.20705556869506836, 0.30472201108932495, 0.09505732357501984, 0.03632878139615059, -0.16911883652210236, -0.23322820663452148, 0.1270904392004013, 0.26113513112068176, -0.20231211185455322, 0.03588193282485008, -0.010399928316473961, 0.04883396252989769, -0.0775550976395607, 0.05452252924442291, -0.3443523347377777, -0.02274063602089882, -0.22121770679950714, 0.24034826457500458, 0.12901589274406433, 0.01588352955877781, 0.1009758859872818, -0.12922441959381104, -0.03020513243973255, 0.19713303446769714, -0.03883172199130058, 0.1435362845659256, -0.08869193494319916, 0.26557841897010803, 0.31917962431907654, 0.16784889996051788, -0.08941975235939026, 0.29042360186576843, 0.08135984092950821, 0.2633022367954254, -0.03736265376210213, 0.3029986619949341, -0.22787153720855713, 0.03696596622467041, 0.1037135198712349, 0.2465103566646576, -0.018953701481223106, 0.2433106005191803, 0.11851630359888077, -0.08313387632369995, 0.09952298551797867, 0.0027660417836159468, 0.3833668828010559, 0.776116669178009, -0.1842089742422104, -0.31482037901878357, 0.4969840347766876, 0.01780637353658676, -0.10703188180923462, 0.0064505706541240215, 0.2476186752319336, 0.16800203919410706, 0.025308435782790184, 0.36049529910087585, 0.030094314366579056, -0.035909656435251236, 0.09229408204555511, 0.1062341034412384, 0.0886678472161293, -0.2641746997833252, -0.2379089891910553, 0.5210921168327332, -0.22591142356395721, 0.24124954640865326, 0.3535764813423157, 0.13580672442913055, 0.29599884152412415, 0.037660591304302216, 0.07566310465335846, 0.5019500851631165, -0.010013422928750515, 0.04312722384929657, 0.14046485722064972, 0.13297906517982483, 0.014500528573989868, 0.2870315611362457, 0.0750100389122963, 0.013744967989623547, 0.33756136894226074, 0.590536892414093, -0.20217761397361755, -0.16447243094444275, -0.1727118045091629, -0.16748885810375214, 0.07695972174406052, -0.22806961834430695, -0.2735535204410553, 0.12884864211082458, -0.21248236298561096, -0.10302986949682236, -0.2302062064409256, -0.29267239570617676, 0.2116546630859375, 0.05275099724531174, -0.33862489461898804, -0.2714468538761139, -0.4014514684677124, 0.14914582669734955, 0.05067812278866768, -0.34592899680137634, 0.12932860851287842, -0.07099145650863647, 0.11562065780162811, 0.39561131596565247, 0.25407347083091736, 0.3397335112094879, 0.06733015179634094, -0.3097277879714966, -0.3134121894836426, -0.32432126998901367, 0.15070997178554535, 0.00874039251357317, 0.3977159559726715, 0.017715128138661385, -0.1132785975933075, 0.6163251996040344, 0.0042132106609642506, 0.10033039003610611, 0.26405972242355347, 0.3659212589263916, -0.5369970798492432, -0.1176464855670929, 0.7859678864479065, 0.07999031990766525, -0.19354212284088135, 0.07940370589494705, 0.22868436574935913, -0.19130268692970276, 0.1738029420375824, 0.046440400183200836, -0.11265207082033157, 0.18436257541179657, 0.11147193610668182, 0.015183850191533566, -0.052295438945293427, 0.34844350814819336, 0.419271856546402, -0.06455577909946442, -0.27449411153793335, -0.015578641556203365, -0.6457425355911255, 0.18133953213691711, -0.053281500935554504, -0.2521618902683258, 0.004350710194557905, -0.24237211048603058, 0.17777687311172485, 0.11598450690507889, -0.24688774347305298, 0.13935713469982147, -0.08880838751792908, 0.08666317164897919, -0.14721113443374634, -0.3679310381412506, -0.17883354425430298, 0.08051306754350662, -0.07546930015087128, -0.4362680912017822, 0.20022070407867432, -0.42061206698417664, -0.12240856885910034, 0.07628114521503448, 0.015164615586400032, 0.5335346460342407, -0.01931019313633442, 0.2926968038082123, -0.05495208501815796, 0.5333089828491211, 0.19510102272033691, 0.006028718315064907, -0.18519830703735352, 0.18206755816936493, -0.1992654949426651, 0.13084910809993744, -0.04895925894379616, 0.06954137980937958, -0.14772431552410126, 0.5281546115875244, -0.4214252829551697, 0.5098816156387329, -0.05077577009797096, 0.1615394502878189, -0.329012393951416, 0.10240171104669571, -0.2403387725353241, 0.08635559678077698, 0.2888401448726654, 0.21862724423408508, -0.11910432577133179, 0.37171006202697754, -0.2540442645549774, -0.20215825736522675, 0.4779549837112427, -0.1859750896692276, -0.10019843280315399, 0.15425172448158264, 0.30320876836776733, 0.19870367646217346, -0.1865212470293045, -0.5294919013977051, 0.15762847661972046, 0.3263660669326782, -0.10436941683292389, -0.04502718523144722, 0.22184570133686066, 0.20797045528888702, 0.060865383595228195, 0.007538842502981424, 0.9585251212120056, -0.13303996622562408, 0.08896082639694214, -0.0688256248831749, -0.24809671938419342 ]
https://github.com/huggingface/datasets/issues/161
Discussion on version identifier & MockDataLoaderManager for test data
I'm good with either putting it in description, adding it to the config, or loosening version formatting. I mostly don't have a full conceptual grasp of what each identifier ends up meaning in the datasets code so hard to evaluate the best approach. For background, the multiple formats is a consequence of: 1. Each example is one multi-sentence trivia question 2. For training, its better to treat each sentence as an example 3. For evaluation, should test on: (1) first sentence, (2) full question, and (3) partial questions (does the model get the question right having seen the first half) We use the date format for version since: (1) we expect some degree of updates since new questions come in every year and (2) the timestamp itself matches the Wikipedia dump that it is dependent on (so similar to the Wikipedia dataset). perhaps this is better discussed in https://github.com/huggingface/nlp/pull/169 or update title?
Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done.
152
Discussion on version identifier & MockDataLoaderManager for test data Hi, I'm working on adding a dataset and ran into an error due to `download` not being defined on `MockDataLoaderManager`, but being defined in `nlp/utils/download_manager.py`. The readme step running this: `RUN_SLOW=1 pytest tests/test_dataset_common.py::DatasetTest::test_load_real_dataset_localmydatasetname` triggers the error. If I can get something to work, I can include it in my data PR once I'm done. I'm good with either putting it in description, adding it to the config, or loosening version formatting. I mostly don't have a full conceptual grasp of what each identifier ends up meaning in the datasets code so hard to evaluate the best approach. For background, the multiple formats is a consequence of: 1. Each example is one multi-sentence trivia question 2. For training, its better to treat each sentence as an example 3. For evaluation, should test on: (1) first sentence, (2) full question, and (3) partial questions (does the model get the question right having seen the first half) We use the date format for version since: (1) we expect some degree of updates since new questions come in every year and (2) the timestamp itself matches the Wikipedia dump that it is dependent on (so similar to the Wikipedia dataset). perhaps this is better discussed in https://github.com/huggingface/nlp/pull/169 or update title?
[ 0.05760371312499046, 0.2424968183040619, -0.0005985943134874105, -0.016873564571142197, -0.026222307235002518, -0.10918852686882019, 0.37034478783607483, 0.27449795603752136, 0.0017394530586898327, -0.073363296687603, 0.19066652655601501, 0.10673809796571732, -0.259066104888916, 0.05271456390619278, 0.0636916235089302, -0.17159391939640045, 0.06112439185380936, 0.20525714755058289, -0.09032153338193893, 0.19936949014663696, -0.10669605433940887, -0.013935601338744164, -0.021723762154579163, 0.20252689719200134, -0.0918591096997261, -0.04485810548067093, -0.2902241051197052, 0.06356655806303024, -0.3563999831676483, -0.9518451690673828, 0.25599655508995056, 0.16132815182209015, 0.21488910913467407, 0.39023515582084656, -0.00012299601803533733, -0.1342220902442932, 0.5136099457740784, -0.030447743833065033, -0.7506928443908691, -0.19761860370635986, 0.32741332054138184, -0.43662112951278687, 0.39083224534988403, -0.08335205167531967, 0.2027074247598648, -0.26675987243652344, 0.3141607940196991, 0.20798015594482422, 0.14404089748859406, 0.21452759206295013, 0.07082144916057587, 0.25298404693603516, 0.029718058183789253, -0.09841064363718033, -0.04017804563045502, 0.11929906159639359, -0.034477852284908295, 0.5099806189537048, 0.3937709331512451, 0.12621353566646576, -0.16321903467178345, 0.11727821081876755, 0.1400747448205948, 0.3983491063117981, 0.22597938776016235, -0.01795371063053608, 0.5039623379707336, -0.047598760575056076, -0.20237988233566284, 0.054717592895030975, 0.8494493365287781, -0.5009228587150574, -0.4326483905315399, -0.14169466495513916, 0.03228963539004326, -0.16478879749774933, 0.2154713124036789, -0.10920257121324539, -0.15507301688194275, 0.057125356048345566, -0.24445419013500214, -0.11133256554603577, -0.15746952593326569, 0.3368920087814331, -0.22511975467205048, 0.29151421785354614, -0.010365284979343414, 0.19610385596752167, -0.03072475455701351, -0.18355828523635864, 0.05790716037154198, -0.2011938989162445, 0.07630230486392975, 0.1746194064617157, 0.016124768182635307, -0.2886020839214325, -0.2943032681941986, -0.18726474046707153, 0.12454863637685776, 0.014858272857964039, -0.4481174051761627, -0.1796247810125351, -0.03736454248428345, 0.21170829236507416, 0.2502317726612091, 0.3405125141143799, 0.39323899149894714, 0.1903860867023468, 0.15328200161457062, 0.0801822766661644, 0.15144819021224976, 0.15082012116909027, -0.04188370332121849, -0.22569634020328522, -0.011926066130399704, 0.025541841983795166, 0.1661314070224762, -0.28247174620628357, -0.051678966730833054, -0.12572993338108063, -0.5409556031227112, -0.07574262470006943, -0.13150514662265778, 0.2328939586877823, -0.1798987239599228, -0.04507238045334816, 0.03212318196892738, 0.41973361372947693, -0.17848749458789825, -0.1591353416442871, -0.01568848267197609, 0.07461448013782501, -0.2581368684768677, -0.0456654354929924, 0.17432135343551636, -0.09063173085451126, 0.45086634159088135, -0.21950899064540863, -0.3382495045661926, -0.17939545214176178, 0.06953320652246475, 0.01314233336597681, -0.09206803888082504, 0.23776037991046906, -0.19513824582099915, 0.06764157861471176, 0.08036749809980392, -0.41097769141197205, -0.27822065353393555, 0.010887356474995613, 0.05627705529332161, -0.3433369994163513, -0.027284223586320877, 0.0628063753247261, -0.48337218165397644, -0.010563365183770657, 0.12420593947172165, 0.09027833491563797, 0.1491735428571701, -0.31665071845054626, 0.10074428468942642, -0.4133778214454651, -0.23477114737033844, -0.09195555001497269, -0.1889970749616623, 0.30426037311553955, -0.2840297818183899, -0.2755051255226135, -0.16645795106887817, -0.3127400577068329, 0.09089331328868866, -0.1886778175830841, -0.07610435038805008, 0.27452632784843445, -0.19024670124053955, 0.030188512057065964, 0.5996150970458984, -0.5389531254768372, -0.2087564915418625, 0.5325264930725098, -0.27516698837280273, -0.052934467792510986, 0.2893846929073334, -0.07397907227277756, 0.11017477512359619, -0.3594629466533661, -0.10819559544324875, 0.05834265798330307, 0.003968278877437115, -0.1924615502357483, -0.21571455895900726, -0.26801061630249023, 0.18028438091278076, 0.2851307690143585, -0.19205887615680695, 0.018702734261751175, 0.06548096239566803, 0.185058131814003, 0.32735610008239746, 0.09518332779407501, 0.12832672894001007, -0.18868406116962433, -0.055135082453489304, -0.07846163213253021, -0.12363851070404053, -0.01617109775543213, -0.7373561859130859, 0.15254506468772888, -0.10349656641483307, 0.20631973445415497, 0.04400218278169632, 0.0035778533201664686, -0.4655619263648987, -0.04969049617648125, -0.07461147010326385, -0.3006821572780609, 0.03034364804625511, 0.23752450942993164, 0.07605625689029694, -0.003968603443354368, 0.008902412839233875, -0.012331622652709484, -0.30985939502716064, 0.2481217235326767, -0.049322061240673065, -0.18347029387950897, -0.08366874605417252, 0.14995962381362915, 0.20539923012256622, 0.25154659152030945, 0.033977754414081573, -0.2759888172149658, 0.10466917604207993, 0.32568368315696716, 0.23714804649353027, 0.24148467183113098, 0.171538844704628, 0.06900247931480408, 0.23371604084968567, 0.3315995931625366, 0.30703678727149963, 0.16558176279067993, 0.08676391839981079, -0.26936131715774536, -0.09193121641874313, 0.1549014449119568, 0.24279801547527313, 0.004586626309901476, -0.16184118390083313, -0.06739852577447891, 0.33853986859321594, 0.1212913915514946, -0.23096613585948944, -0.26559704542160034, 0.012635604478418827, 0.29515525698661804, 0.640271008014679, 0.0631796196103096, 0.0867367684841156, -0.034492626786231995, -0.1113731861114502, -0.4682067632675171, -0.02329222299158573, 0.09991318732500076, -0.2714598774909973, -0.16550683975219727, -0.019511865451931953, 0.25677624344825745, 0.7403602004051208, 0.014357608743011951, 0.2837745249271393, 0.04305139556527138, -0.26060429215431213, -0.21312560141086578, 0.1901429295539856, 0.2923334240913391, -0.07247483730316162, 0.14260812103748322, 0.05944402143359184, -0.22141170501708984, -0.011801526881754398, -0.14425219595432281, 0.3677423894405365, 0.32906296849250793, -0.4422621726989746, 0.011568572372198105, -0.2243393212556839, -0.44633749127388, -0.5092360377311707, -0.37744763493537903, -0.2606249451637268, -0.34826013445854187, -0.05424041301012039, 0.22306592762470245, -0.0988536924123764, 0.20850591361522675, -0.0444551520049572, 0.35148823261260986, -0.33391454815864563, -0.06189078837633133, 0.18121914565563202, -0.28744637966156006, -0.43042632937431335, -0.021277468651533127, 0.37744414806365967, 0.010299940593540668, 0.3699015974998474, -0.16296133399009705, -0.30679163336753845, -0.1718640774488449, -0.3577468991279602, 0.17577378451824188, -0.051336344331502914, 0.4902472198009491, 0.43741726875305176, 0.056805118918418884, 0.4220125377178192, -0.0797564685344696, -0.005828259978443384, -0.06160523742437363, -0.35395547747612, -0.19850435853004456, 0.008054148405790329, 0.01152966357767582, -0.22299246490001678, -0.46764811873435974, -0.5795462131500244, -0.06716489791870117, 0.23849786818027496, 0.023740096017718315, 0.13625745475292206, 0.35848936438560486, -0.23525892198085785, 0.23094338178634644, -0.11682175099849701, 0.06175219640135765, -0.13076823949813843, 0.04842391982674599, 0.03346915915608406, -0.14100679755210876, -0.14650902152061462, 0.08306866884231567, -0.06929145008325577, -0.056238994002342224, -0.125167116522789, -0.4092947840690613, -0.4603891968727112, 0.1925334632396698, -0.019878726452589035, 0.3854578137397766, -0.1197538822889328, 0.19413042068481445, 0.2285912185907364, 0.15268801152706146, -0.09221991896629333, -0.2539879381656647, 0.30851125717163086, 0.048168107867240906, 0.2682610750198364, 0.2667803466320038, 0.3099619746208191, -0.08249540627002716, 0.854885458946228, 0.23433127999305725, -0.2229713350534439, -0.17823928594589233, -0.028071917593479156, -0.07432776689529419, 0.008990504778921604, 0.03905468061566353, 0.2459833025932312, -0.19373981654644012, -0.25766199827194214, 0.4539284110069275, -0.11471637338399887, -0.0794103667140007, -0.012890145182609558, 0.2891296148300171, -0.10928501188755035, -0.00573522737249732, 0.19463367760181427, -0.15763522684574127, 0.02347104623913765, 0.010373669676482677, -0.03370597958564758, -0.15833452343940735, -0.06846433132886887, -0.033196888864040375, 0.341937392950058, -0.11140768975019455, 0.21972471475601196, -0.27268749475479126, -0.06407026201486588, -0.2533639669418335, 0.23277467489242554, 0.39788496494293213, 0.4647166132926941, 0.15090206265449524, -0.031669456511735916, 0.21203070878982544, 0.20424818992614746, 0.08792273700237274, -0.03190002590417862, 0.10637810826301575, 0.18112245202064514, -0.22042164206504822, -0.030371466651558876, -0.30279994010925293, -0.22541536390781403, 0.042755305767059326, 0.01996622607111931, 0.17918311059474945, -0.4684460461139679, -0.24010716378688812, 0.32111069560050964, 0.06214398145675659, -0.05001530796289444, -0.23638103902339935, 0.023509759455919266, 0.005604391451925039, -0.0498347170650959, -0.17327050864696503, -0.1989140510559082, -0.013889658264815807, -0.11765541881322861, 0.42405954003334045, -0.038367900997400284, 0.04531438276171684, -0.021975474432110786, 0.24745996296405792, 0.0965980738401413, -0.14922749996185303, 0.14238043129444122, 0.269252210855484, 0.14296676218509674, 0.2716630697250366, 0.1835293471813202, -0.13022008538246155, -0.22181782126426697, 0.05195346102118492, -0.06875786930322647, 0.37542372941970825, 0.4415145218372345, 0.20746338367462158, 0.03125409036874771, -0.24363282322883606, -0.17298327386379242, -0.3376980721950531, 0.27770349383354187, 0.3622957468032837, -0.12266840040683746, -0.018863610923290253, -0.3984583616256714, 0.7639316916465759, 0.11514201760292053, -0.31864961981773376, 0.31977638602256775, -0.17126977443695068, -0.27540990710258484, 0.06674157083034515, 0.04492628201842308, 0.9632917642593384, 0.18758003413677216, 0.26972997188568115, 0.3356110751628876, -0.3050232231616974, 0.5088142156600952, 0.0772995576262474, 0.08329905569553375, -0.5266805291175842, -0.004951165989041328, 0.011898696422576904, -0.0852636843919754, 0.2526690363883972, -0.09175092726945877, -0.29020753502845764, 0.3714599311351776, 0.15412455797195435, 0.33690568804740906, 0.2666241526603699, 0.3904850482940674, -0.1026519164443016, -0.0818486139178276, -0.31482914090156555, 0.032741546630859375, -0.18126191198825836, 0.29423072934150696, -0.2555637061595917, -0.05453578010201454, -0.194590225815773, -0.04891160503029823, -0.16635583341121674, -0.2321009784936905, -0.17980745434761047, 0.04404440522193909, 0.09738369286060333, -0.24454781413078308, 0.0070836846716701984, 0.24630339443683624, 0.13479027152061462, -0.3061778247356415, -0.08299575001001358, -0.0007388164522126317, -0.20389820635318756, 0.19736775755882263, 0.269948810338974, -0.05206357687711716, 0.45949187874794006, -0.26142480969429016, 0.0046785082668066025, 0.06200600787997246, -0.29994553327560425, -0.011491084471344948, -0.07748136669397354, -0.17925870418548584, 0.18972741067409515, -0.2973543703556061, -0.2757718861103058, 0.034674838185310364, 0.015729377046227455, -0.016935789957642555, -0.013132099062204361, 0.19917599856853485, -0.27070099115371704, 0.016309089958667755, -0.09080898761749268, -0.23815368115901947, -0.0127136604860425, 0.3433263599872589, -0.08797584474086761, -0.07147157937288284, 0.7205213308334351, -0.1423332691192627, -0.013047975488007069, -0.08775806427001953, 0.07892642170190811, -0.07721257954835892, -0.21021512150764465, -0.024667615070939064, -0.09398006647825241, -0.3097338080406189, -0.11627903580665588, 0.6664240956306458, 0.5130172967910767, 0.08737687766551971, -0.01276517566293478, -0.3916796147823334, -0.33617308735847473, 0.11024472862482071, -0.2133312225341797, 0.20463024079799652, -0.15378211438655853, 0.23011039197444916, 0.20599862933158875, 0.0917343869805336, -0.1989636868238449, -0.21286644041538239, -0.052418820559978485, 0.3219780921936035, -0.20810915529727936, 0.008620302192866802, -0.02283157780766487, 0.07475601881742477, -0.07202211767435074, -0.030391572043299675, -0.3601391613483429, -0.03611622005701065, -0.16067545115947723, 0.23085981607437134, 0.16787390410900116, -0.1009950339794159, 0.05119547247886658, -0.05666710436344147, 0.0830276757478714, 0.15836569666862488, -0.06596092879772186, -0.06922030448913574, -0.046982262283563614, 0.31562021374702454, 0.2211274355649948, 0.2653515040874481, -0.061260778456926346, 0.15333685278892517, 0.07793182134628296, 0.1418086141347885, 0.08331699669361115, 0.18333105742931366, -0.32301199436187744, 0.013775154948234558, 0.07649040967226028, 0.294942170381546, -0.11261388659477234, 0.23987098038196564, 0.11625204235315323, -0.03209979087114334, 0.14109596610069275, 0.08694107085466385, 0.284458190202713, 0.8438324928283691, -0.18474644422531128, -0.2881437838077545, 0.41060033440589905, 0.017603551968932152, -0.04218687117099762, 0.08298318833112717, 0.26608002185821533, 0.11460418254137039, 0.022272950038313866, 0.28549671173095703, 0.05585002526640892, -0.0035337205044925213, 0.039524324238300323, 0.16859540343284607, 0.19243654608726501, -0.2128521353006363, -0.06870882213115692, 0.4142081141471863, -0.2447902411222458, 0.26162704825401306, 0.37993934750556946, 0.05843663215637207, 0.21717189252376556, -0.09376568347215652, 0.06660223752260208, 0.4720221757888794, 0.10343360900878906, 0.07247643917798996, 0.1404721885919571, 0.10956500470638275, -0.05472450703382492, 0.46144673228263855, 0.027293935418128967, 0.019859852269291878, 0.28616175055503845, 0.6132852435112, -0.3343771994113922, -0.321731835603714, -0.17343159019947052, -0.07680520415306091, 0.10721207410097122, -0.18119484186172485, -0.5276594161987305, 0.014906727708876133, -0.2530691623687744, -0.07244831323623657, -0.2701846659183502, -0.35571739077568054, 0.3199388086795807, 0.041181087493896484, -0.3105318546295166, -0.41609326004981995, -0.3604917526245117, 0.08546431362628937, 0.12812121212482452, -0.32149264216423035, 0.2644841969013214, 0.23103854060173035, 0.06905652582645416, 0.4039294719696045, 0.3324469327926636, 0.4023068845272064, 0.09167810529470444, -0.38616636395454407, -0.30364349484443665, -0.25391867756843567, 0.10707513988018036, 0.12506240606307983, 0.27835798263549805, -0.042769718915224075, -0.21499723196029663, 0.5813985466957092, 0.015517053194344044, 0.07906582206487656, 0.2922721207141876, 0.409268856048584, -0.5363088846206665, -0.24184051156044006, 0.6709067225456238, -0.0029568402096629143, -0.24788813292980194, 0.10846399515867233, 0.15607382357120514, -0.36080697178840637, 0.08438004553318024, 0.0010098937200382352, 0.07150530815124512, 0.19846223294734955, 0.09360793232917786, 0.026088636368513107, -0.009924635291099548, 0.4939209520816803, 0.49144336581230164, -0.10784632712602615, -0.34578391909599304, -0.10265504568815231, -0.7086455225944519, 0.02320426143705845, -0.02492276206612587, -0.3069433271884918, -0.028651652857661247, -0.13132323324680328, 0.18050970137119293, 0.09593173861503601, -0.25932997465133667, 0.05915563926100731, -0.03458641469478607, 0.07855800539255142, -0.3168298006057739, -0.23085398972034454, -0.163457989692688, 0.011754205450415611, -0.11217889934778214, -0.24734269082546234, 0.11470548063516617, -0.3909788429737091, -0.14848601818084717, 0.22859300673007965, 0.0486631877720356, 0.4706299901008606, 0.0439799465239048, 0.1680421680212021, -0.014944995753467083, 0.5717008709907532, 0.27138978242874146, -0.12703844904899597, -0.22806523740291595, 0.21541117131710052, -0.14663124084472656, 0.1772405505180359, -0.12875743210315704, 0.3634103238582611, -0.1993662416934967, 0.5129457712173462, -0.20585589110851288, 0.4702889919281006, -0.044683605432510376, 0.05702648311853409, -0.33190709352493286, 0.08545436710119247, -0.22928445041179657, 0.04520855471491814, 0.3392931818962097, 0.3558691143989563, -0.06925652921199799, 0.29817667603492737, -0.260295033454895, -0.2253071665763855, 0.3200797736644745, -0.31884104013442993, -0.11840435117483139, 0.1429835855960846, 0.2893078327178955, 0.14046384394168854, 0.02453598380088806, -0.6277329325675964, 0.07381118834018707, 0.2311151623725891, -0.17052221298217773, -0.09328301250934601, 0.17623208463191986, 0.23813925683498383, -0.06259176880121231, -0.03592416271567345, 0.9138442277908325, -0.07508823275566101, -0.0677717849612236, -0.10770925879478455, -0.33352944254875183 ]
https://github.com/huggingface/datasets/issues/160
caching in map causes same result to be returned for train, validation and test
Hi @dpressel, thanks for posting your issue! Can you maybe add a complete code snippet that we can copy paste to reproduce the error? For example, I'm not sure where the variable `train_set` comes from in your code and it seems like you are loading multiple datasets at once?
hello, I am working on a program that uses the `nlp` library with the `SST2` dataset. The rough outline of the program is: ``` import nlp as nlp_datasets ... parser.add_argument('--dataset', help='HuggingFace Datasets id', default=['glue', 'sst2'], nargs='+') ... dataset = nlp_datasets.load_dataset(*args.dataset) ... # Create feature vocabs vocabs = create_vocabs(dataset.values(), vectorizers) ... # Create a function to vectorize based on vectorizers and vocabs: print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) # factory method to create a `convert_to_features` function based on vocabs convert_to_features = create_featurizer(vectorizers, vocabs) train_set = train_set.map(convert_to_features, batched=True) train_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) train_loader = torch.utils.data.DataLoader(train_set, batch_size=args.batchsz) valid_set = valid_set.map(convert_to_features, batched=True) valid_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) valid_loader = torch.utils.data.DataLoader(valid_set, batch_size=args.batchsz) test_set = test_set.map(convert_to_features, batched=True) test_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) test_loader = torch.utils.data.DataLoader(test_set, batch_size=args.batchsz) print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) ``` Im not sure if Im using it incorrectly, but the results are not what I expect. Namely, the `.map()` seems to grab the datset from the cache and then loses track of what the specific dataset is, instead using my training data for all datasets: ``` TS 67349 VS 872 ES 1821 TS 67349 VS 67349 ES 67349 ``` The behavior changes if I turn off the caching but then the results fail: ``` train_set = train_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... valid_set = valid_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... test_set = test_set.map(convert_to_features, batched=True, load_from_cache_file=False) ``` Now I get the right set of features back... ``` TS 67349 VS 872 ES 1821 100%|██████████| 68/68 [00:00<00:00, 92.78it/s] 100%|██████████| 1/1 [00:00<00:00, 75.47it/s] 0%| | 0/2 [00:00<?, ?it/s]TS 67349 VS 872 ES 1821 100%|██████████| 2/2 [00:00<00:00, 77.19it/s] ``` but I think its losing track of the original training set: ``` Traceback (most recent call last): File "/home/dpressel/dev/work/baseline/api-examples/layers-classify-hf-datasets.py", line 148, in <module> for x in train_loader: File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 345, in __next__ data = self._next_data() File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 385, in _next_data data = self._dataset_fetcher.fetch(index) # may raise StopIteration File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp> data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 338, in __getitem__ output_all_columns=self._output_all_columns, File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 294, in _getitem outputs = self._unnest(self._data.slice(key, 1).to_pydict()) File "pyarrow/table.pxi", line 1211, in pyarrow.lib.Table.slice File "pyarrow/public-api.pxi", line 390, in pyarrow.lib.pyarrow_wrap_table File "pyarrow/error.pxi", line 85, in pyarrow.lib.check_status pyarrow.lib.ArrowInvalid: Column 3: In chunk 0: Invalid: Length spanned by list offsets (15859698) larger than values array (length 100000) Process finished with exit code 1 ``` The full-example program (minus the print stmts) is here: https://github.com/dpressel/mead-baseline/pull/620/files
49
caching in map causes same result to be returned for train, validation and test hello, I am working on a program that uses the `nlp` library with the `SST2` dataset. The rough outline of the program is: ``` import nlp as nlp_datasets ... parser.add_argument('--dataset', help='HuggingFace Datasets id', default=['glue', 'sst2'], nargs='+') ... dataset = nlp_datasets.load_dataset(*args.dataset) ... # Create feature vocabs vocabs = create_vocabs(dataset.values(), vectorizers) ... # Create a function to vectorize based on vectorizers and vocabs: print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) # factory method to create a `convert_to_features` function based on vocabs convert_to_features = create_featurizer(vectorizers, vocabs) train_set = train_set.map(convert_to_features, batched=True) train_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) train_loader = torch.utils.data.DataLoader(train_set, batch_size=args.batchsz) valid_set = valid_set.map(convert_to_features, batched=True) valid_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) valid_loader = torch.utils.data.DataLoader(valid_set, batch_size=args.batchsz) test_set = test_set.map(convert_to_features, batched=True) test_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) test_loader = torch.utils.data.DataLoader(test_set, batch_size=args.batchsz) print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) ``` Im not sure if Im using it incorrectly, but the results are not what I expect. Namely, the `.map()` seems to grab the datset from the cache and then loses track of what the specific dataset is, instead using my training data for all datasets: ``` TS 67349 VS 872 ES 1821 TS 67349 VS 67349 ES 67349 ``` The behavior changes if I turn off the caching but then the results fail: ``` train_set = train_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... valid_set = valid_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... test_set = test_set.map(convert_to_features, batched=True, load_from_cache_file=False) ``` Now I get the right set of features back... ``` TS 67349 VS 872 ES 1821 100%|██████████| 68/68 [00:00<00:00, 92.78it/s] 100%|██████████| 1/1 [00:00<00:00, 75.47it/s] 0%| | 0/2 [00:00<?, ?it/s]TS 67349 VS 872 ES 1821 100%|██████████| 2/2 [00:00<00:00, 77.19it/s] ``` but I think its losing track of the original training set: ``` Traceback (most recent call last): File "/home/dpressel/dev/work/baseline/api-examples/layers-classify-hf-datasets.py", line 148, in <module> for x in train_loader: File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 345, in __next__ data = self._next_data() File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 385, in _next_data data = self._dataset_fetcher.fetch(index) # may raise StopIteration File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp> data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 338, in __getitem__ output_all_columns=self._output_all_columns, File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 294, in _getitem outputs = self._unnest(self._data.slice(key, 1).to_pydict()) File "pyarrow/table.pxi", line 1211, in pyarrow.lib.Table.slice File "pyarrow/public-api.pxi", line 390, in pyarrow.lib.pyarrow_wrap_table File "pyarrow/error.pxi", line 85, in pyarrow.lib.check_status pyarrow.lib.ArrowInvalid: Column 3: In chunk 0: Invalid: Length spanned by list offsets (15859698) larger than values array (length 100000) Process finished with exit code 1 ``` The full-example program (minus the print stmts) is here: https://github.com/dpressel/mead-baseline/pull/620/files Hi @dpressel, thanks for posting your issue! Can you maybe add a complete code snippet that we can copy paste to reproduce the error? For example, I'm not sure where the variable `train_set` comes from in your code and it seems like you are loading multiple datasets at once?
[ -0.12632250785827637, -0.22070901095867157, -0.10063198953866959, 0.27801084518432617, 0.24547572433948517, -0.20522341132164001, 0.137210875749588, 0.5128593444824219, 0.18828663229942322, -0.06233531981706619, 0.12781980633735657, 0.38913458585739136, 0.003776297438889742, -0.1606176495552063, -0.010743561200797558, 0.024212030693888664, 0.09608013927936554, 0.23505231738090515, -0.14556236565113068, -0.1648235321044922, 0.08189409971237183, 0.2129841297864914, -0.042409978806972504, 0.11652691662311554, -0.39034950733184814, 0.041289906948804855, 0.21235668659210205, -0.2308902144432068, 0.06845776736736298, -0.25706538558006287, 0.1678507924079895, -0.09324328601360321, 0.02583368308842182, 0.16631756722927094, -0.00010973814642056823, 0.0067980014719069, -0.01538778841495514, -0.13856948912143707, -0.007274831645190716, 0.04019518941640854, 0.11813383549451828, -0.21172897517681122, -0.10918274521827698, -0.3121316730976105, -0.3320115804672241, -0.03766085207462311, 0.01780932955443859, -0.35565948486328125, 0.31285759806632996, 0.14745205640792847, 0.26071032881736755, 0.22490893304347992, -0.2538125216960907, 0.21113626658916473, 0.14512759447097778, -0.029327336698770523, -0.17419742047786713, -0.04230547323822975, -0.04397892951965332, -0.33446553349494934, -0.37676504254341125, 0.5254571437835693, -0.07658161967992783, 0.17014139890670776, 0.21466633677482605, 0.22505579888820648, 0.17260590195655823, 0.035937704145908356, 0.1429363638162613, -0.11396253854036331, 0.011094827204942703, -0.22691738605499268, -0.027116434648633003, -0.3733721077442169, -0.4186401665210724, -0.07852599024772644, 0.13762502372264862, 0.18104472756385803, -0.1363932490348816, -0.06371653825044632, -0.43193766474723816, 0.2392898052930832, 0.0652494877576828, -0.012788408435881138, 0.11952681839466095, 0.29516080021858215, 0.04380622133612633, 0.19522294402122498, 0.12944649159908295, -0.06973281502723694, -0.022558962926268578, -0.21451696753501892, -0.05897140875458717, 0.22640138864517212, -0.4359849691390991, 0.07208239287137985, 0.4315200448036194, -0.1404140740633011, -0.053242750465869904, 0.03063775785267353, 0.3919311761856079, 0.07921352982521057, 0.11681784689426422, -0.008491764776408672, -0.10416068881750107, 0.5898676514625549, 0.04432177543640137, 0.17118391394615173, -0.014969921670854092, -0.19827847182750702, -0.3560168445110321, 0.004987859632819891, 0.2880178987979889, -0.28660091757774353, 0.42841145396232605, 0.40236055850982666, -0.04226037487387657, -0.12152973562479019, -0.030102189630270004, 0.09402408450841904, -0.33085379004478455, 0.05566990002989769, 0.058729998767375946, 0.31395798921585083, -0.22545045614242554, 0.14926567673683167, -0.17376750707626343, -0.10335483402013779, -0.3849693238735199, 0.12735125422477722, -0.33784618973731995, -0.033323898911476135, -0.4962189197540283, 0.0015165578806772828, 0.28538474440574646, -0.002859933767467737, 0.28535225987434387, -0.038445454090833664, -0.13722018897533417, -0.3019891679286957, 0.10317365825176239, -0.39851170778274536, 0.5702314376831055, -0.1072295531630516, -0.19701702892780304, 0.24102117121219635, 0.13037829101085663, 0.08526421338319778, -0.29768794775009155, -0.19837751984596252, -0.1698075532913208, -0.20906901359558105, 0.2598607540130615, 0.24224542081356049, -0.07942765206098557, 0.14139588177204132, 0.09248226881027222, 0.0776788741350174, 0.3432585299015045, -0.232854962348938, 0.10284040868282318, -0.20148923993110657, -0.28198012709617615, -0.2043590545654297, 0.16256873309612274, 0.2623099982738495, 0.06677649915218353, -0.13737238943576813, 0.4017971158027649, 0.2371826320886612, 0.14729078114032745, 0.4138067364692688, -0.1965339034795761, 0.30949410796165466, -0.43349170684814453, 0.3486190438270569, 0.506476879119873, -0.32673147320747375, -0.49634483456611633, 0.09944828599691391, -0.15631580352783203, 0.1002880334854126, 0.10244903713464737, 0.15137287974357605, 0.029087670147418976, 0.014419378712773323, 0.1285371035337448, 0.018702296540141106, -0.0005146907060407102, 0.208420529961586, -0.31692788004875183, -0.021933916956186295, 0.5500646233558655, -0.2524673640727997, 0.02847237139940262, -0.13123734295368195, -0.05504271760582924, 0.021933000534772873, 0.006768169347196817, -0.11219003051519394, 0.12185932695865631, 0.11460470408201218, -0.056878868490457535, -0.1581844687461853, -0.10643851011991501, -0.024692509323358536, -0.28826257586479187, 0.23830120265483856, -0.14981895685195923, -0.0999680832028389, 0.06172538921236992, -0.09704917669296265, -0.16200260818004608, -0.12927502393722534, -0.18091638386249542, -0.27306458353996277, 0.20130178332328796, 0.23052407801151276, 0.3202398419380188, -0.07290733605623245, 0.2252090722322464, 0.3680817782878876, 0.048363614827394485, -0.21607212722301483, -0.4300347566604614, 0.08568026125431061, -0.17773471772670746, -0.2188279777765274, -0.0748467668890953, 0.22306986153125763, 0.2090262770652771, -0.04792642965912819, -0.14242812991142273, 0.14827579259872437, -0.2220856100320816, 0.4189179241657257, -0.13702799379825592, 0.269452840089798, -0.06965683400630951, 0.029306238517165184, 0.016667358577251434, 0.25400519371032715, 0.05881987884640694, -0.08980134129524231, -0.07270598411560059, 0.47423407435417175, 0.18140216171741486, 0.17011605203151703, -0.048070792108774185, -0.21626794338226318, 0.051984209567308426, -0.17831753194332123, -0.12091574817895889, -0.11881643533706665, 0.13640747964382172, -0.08057200163602829, 0.1488880068063736, 0.15232868492603302, -0.10698743164539337, 0.3120083510875702, 0.7600833773612976, 0.19691871106624603, -0.011016919277608395, 0.015740947797894478, -0.057213034480810165, -0.21779173612594604, 0.3338564336299896, -0.05868269130587578, 0.31038790941238403, 0.10473166406154633, 0.12716487050056458, -0.04620952904224396, -0.21965020895004272, -0.15443570911884308, 0.12464798986911774, -0.21436883509159088, -0.1919763833284378, 0.308034211397171, 0.28208687901496887, -0.15836724638938904, -0.35378316044807434, 0.0940157026052475, -0.01276430580765009, 0.026191623881459236, -0.15800140798091888, 0.21723663806915283, -0.1665293276309967, -0.2780851721763611, -0.28641536831855774, 0.13615694642066956, -0.12220374494791031, -0.0921654924750328, -0.04126313701272011, 0.42542964220046997, -0.036654673516750336, 0.19515161216259003, -0.3149721920490265, 0.08951952308416367, 0.226604163646698, -0.2663276791572571, -0.07168538868427277, -0.1902879923582077, -0.39492008090019226, 0.04223187640309334, -0.09523041546344757, -0.0929960310459137, 0.3783297538757324, -0.04731621965765953, -0.06796029210090637, -0.2881457805633545, -0.1725403368473053, 0.10717813670635223, -0.09261170029640198, -0.15536807477474213, 0.16724860668182373, -0.03133843466639519, -0.10115880519151688, -0.2987455725669861, 0.29379722476005554, -0.3053131401538849, -0.29914355278015137, -0.05198654159903526, 0.09890630096197128, 0.03010939620435238, -0.28034764528274536, -0.42862412333488464, 0.27357834577560425, -0.31724390387535095, 0.07546782493591309, -0.14892923831939697, 0.18622653186321259, 0.33093470335006714, 0.008953339420258999, -0.014561006799340248, -0.06404032558202744, 0.16439402103424072, -0.46741002798080444, -0.1409541219472885, 0.32257840037345886, 0.01876099780201912, -0.21630428731441498, -0.32459139823913574, -0.27572473883628845, 0.347724050283432, 0.19238898158073425, -0.5803960561752319, -0.21442314982414246, -0.04890046268701553, 0.24803197383880615, -0.05125873163342476, -0.04649535194039345, 0.5745421648025513, 0.01972527615725994, -0.17946361005306244, -0.16727842390537262, -0.33184051513671875, 0.35973140597343445, 0.41926297545433044, 0.2494877278804779, -0.011159147135913372, 0.17781507968902588, 0.20092034339904785, 0.6707466244697571, 0.20191961526870728, -0.25905755162239075, 0.3317307233810425, 0.10723208636045456, 0.07499753683805466, -0.2034970372915268, -0.3130224943161011, 0.18576058745384216, -0.18160304427146912, 0.019732147455215454, 0.2719840407371521, 0.006251134444028139, -0.2801845371723175, 0.01774226687848568, 0.21881961822509766, -0.2253446877002716, -0.3280579447746277, 0.2597459852695465, -0.34242793917655945, 0.24094845354557037, 0.03637058660387993, 0.31580042839050293, -0.5820142030715942, -0.24272412061691284, 0.19462814927101135, -0.33594635128974915, 0.2798454165458679, 0.04449770972132683, -0.5836008787155151, 0.12147048860788345, -0.538021981716156, 0.18194131553173065, 0.3146284818649292, 0.2724076807498932, -0.09202561527490616, -0.2465258091688156, 0.04334738850593567, -0.14713139832019806, 0.5711411833763123, 0.040972813963890076, 0.4085797369480133, 0.1699264794588089, -0.1302468627691269, -0.1994732767343521, -0.2247127890586853, 0.017234763130545616, 0.2801370620727539, 0.32649919390678406, 0.40084776282310486, -0.3452445864677429, -0.36327362060546875, -0.26576584577560425, 0.007364223711192608, -0.04618128761649132, 0.007107758894562721, -0.12583346664905548, 0.18938525021076202, 0.01648680679500103, 0.17434526979923248, -0.01498047262430191, 0.1183028295636177, 0.016039855778217316, -0.13654641807079315, -0.03913364186882973, -0.21942690014839172, 0.1282409280538559, 0.2264377325773239, 0.4735944867134094, 0.15032504498958588, 0.03246142715215683, 0.033214494585990906, 0.21515744924545288, -0.5671088099479675, 0.26595762372016907, -0.08884964138269424, -0.004131742753088474, 0.07464420050382614, 0.14443816244602203, -0.03030155971646309, 0.3378978669643402, -0.04252500832080841, 0.14571364223957062, -0.12862949073314667, 0.008956843055784702, -0.5539686679840088, 0.4854544401168823, 0.3874591886997223, 0.13246646523475647, -0.04488928243517876, -0.24685749411582947, 0.17346294224262238, 0.07412006705999374, -0.2765309810638428, 0.1322597712278366, -0.4306926131248474, -0.43199479579925537, 0.1857973337173462, 0.20948611199855804, 0.9232466816902161, 0.332022100687027, 0.3719581067562103, 0.1007041484117508, 0.10358098894357681, 0.2642512321472168, -0.23411183059215546, 0.40925082564353943, -0.3693642020225525, -0.0463179275393486, 0.024538656696677208, -0.16673317551612854, -0.22738394141197205, -0.010426265187561512, -0.01099333819001913, 0.30259495973587036, 0.32405051589012146, 0.5999365448951721, -0.11211955547332764, 0.16669917106628418, 0.21569165587425232, -0.15811017155647278, -0.2492315024137497, 0.2010732889175415, -0.29230374097824097, 0.3152405023574829, -0.11240831762552261, -0.07410863041877747, -0.01650959439575672, -0.35134196281433105, 0.06714027374982834, 0.06094216927886009, -0.3365170657634735, 0.3048412501811981, 0.13346198201179504, -0.27266666293144226, -0.2969881594181061, 0.2550773024559021, 0.11198677867650986, 0.08252744376659393, -0.048600129783153534, -0.03518514335155487, 0.40064483880996704, 0.3504585027694702, -0.006891036406159401, -0.20158861577510834, 0.1864127367734909, 0.06387528032064438, -0.012894121930003166, 0.03024262934923172, -0.05640564113855362, -0.3015829622745514, -0.25251179933547974, 0.14820435643196106, 0.003372181672602892, -0.41296184062957764, -0.05531826987862587, 0.061445046216249466, -0.066788449883461, -0.09413714706897736, 0.15088026225566864, -0.026068205013871193, -0.10794040560722351, 0.3151542544364929, 0.1173383817076683, -0.24304209649562836, 0.05235802382230759, 0.47483208775520325, -0.040397144854068756, 0.03181783854961395, 0.41542795300483704, -0.15281881392002106, -0.10380645841360092, -0.2553911507129669, -0.28417274355888367, -0.10205967724323273, -0.44884082674980164, 0.11220790445804596, -0.2194061428308487, -0.4019114077091217, 0.0442325733602047, 0.280818372964859, 0.014706877060234547, 0.31849586963653564, -0.011618148535490036, -0.12495850026607513, -0.37697234749794006, -0.10379048436880112, -0.1065821498632431, 0.2671777307987213, 0.2795186936855316, 0.2843484878540039, -0.29586702585220337, 0.2492261826992035, -0.38310956954956055, -0.1071421355009079, -0.18716979026794434, 0.4379133880138397, -0.03913804888725281, -0.3655827045440674, -0.009147044271230698, 0.00185489805880934, 0.12997117638587952, 0.17754493653774261, -0.6386589407920837, -0.2926190495491028, -0.09099180996417999, 0.08186071366071701, -0.007587700616568327, -0.159822016954422, -0.23928211629390717, -0.017213808372616768, 0.18238461017608643, -0.42740321159362793, 0.09215689450502396, 0.15121674537658691, 0.047031741589307785, 0.23718656599521637, -0.16102980077266693, 0.12145575135946274, 0.2881677746772766, -0.18652726709842682, -0.18414875864982605, -0.07622820883989334, 0.08048128336668015, 0.2570165991783142, -0.069468192756176, 0.02047928236424923, -0.21040309965610504, 0.03047284670174122, 0.08927395939826965, 0.1427948772907257, 0.23692886531352997, -0.024752512574195862, -0.2156609743833542, 0.0005510254995897412, 0.1827995628118515, 0.11411970853805542, -0.07855338603258133, -0.32897529006004333, 0.2802230417728424, 0.27836501598358154, -0.11751806735992432, -0.010236726142466068, 0.16770757734775543, -0.09938807040452957, 0.2528513967990875, 0.2071538120508194, 0.10758750885725021, 0.052611079066991806, 0.26961180567741394, 0.12649419903755188, 0.40504035353660583, -0.34832870960235596, 0.08942142128944397, 0.06441634148359299, 0.10164941847324371, 0.0377780981361866, 0.3812146782875061, -0.15324392914772034, 0.08992191404104233, 0.14838334918022156, 0.18833690881729126, 0.20253953337669373, 0.2628898620605469, 0.21354152262210846, -0.1878644973039627, -0.03412109240889549, -0.24108682572841644, 0.14363647997379303, -0.14272600412368774, 0.4156598746776581, 0.15059241652488708, 0.19285538792610168, -0.11485260725021362, -0.4619389772415161, -0.24626794457435608, 0.23536370694637299, -0.3245418071746826, -0.30920886993408203, -0.04867832362651825, -0.08018530905246735, 0.05592920631170273, 0.1945614516735077, -0.07929019629955292, 0.20385901629924774, 0.34283849596977234, -0.008660116232931614, -0.013750203885138035, -0.08435358107089996, -0.17061741650104523, 0.11397870630025864, 0.27486518025398254, -0.08173874765634537, 0.44693678617477417, 0.01080083753913641, -0.05156949535012245, -0.2344711422920227, 0.41141751408576965, 0.28612861037254333, 0.1949474811553955, -0.168862447142601, 0.06395307183265686, 0.20064708590507507, 0.0306134894490242, -0.3546890914440155, 0.3286518454551697, 0.1511448621749878, -0.2930276095867157, 0.36800575256347656, 0.12379536032676697, -0.23215578496456146, 0.2764037549495697, -0.10419601202011108, -0.20146721601486206, -0.04988892748951912, 0.3224368989467621, 0.0997074767947197, -0.0399320125579834, -0.19496309757232666, 0.11112021654844284, -0.14457158744335175, -0.28533950448036194, 0.15934592485427856, -0.02389831282198429, 0.46766093373298645, -0.275675505399704, 0.11720515042543411, -0.06182502955198288, 0.6220859885215759, -0.02922753430902958, 0.12432888895273209, -0.350137859582901, 0.10264351218938828, -0.33316680788993835, 0.1984093189239502, -0.08381783962249756, 0.15374279022216797, -0.3975840210914612, 0.20426496863365173, -0.11549855023622513, -0.023466093465685844, -0.11804244667291641, -0.00788638461381197, 0.14740723371505737, 0.17893213033676147, -0.3869176506996155, 0.1554814726114273, -0.0862593725323677, -0.00626726308837533, 0.2917737364768982, -0.37667545676231384, 0.22199058532714844, 0.04247966781258583, 0.20691059529781342, 0.010986385866999626, -0.02172369882464409, -0.054806627333164215, 0.33657434582710266, 0.6232408285140991, 0.34702974557876587, 0.14255116879940033, -0.10102774947881699, 0.08155330270528793, -0.34743764996528625, 0.014104106463491917, -0.06088816747069359, -0.20964983105659485, 0.09756835550069809, 0.5258304476737976, -0.22562745213508606, 0.13194629549980164, -0.17804564535617828, 0.2366943657398224, 0.38227659463882446, -0.15272878110408783, 0.06179971620440483, 0.048225898295640945, -0.25984057784080505, -0.005199328530579805, -0.12586632370948792, 0.45379582047462463, 0.007182890549302101, 0.32383018732070923, -0.3592907786369324, -0.5417624711990356, 0.4329099953174591, -0.4511263966560364, -0.3025250732898712, 0.12327846884727478, 0.23878563940525055, 0.34956884384155273, -0.06000060588121414, -0.5436676144599915, -0.1724908947944641, 0.26911357045173645, -0.030661094933748245, -0.15704523026943207, 0.20768482983112335, -0.1486440747976303, -0.04107564687728882, -0.11675458401441574, 0.17646899819374084, 0.10152711719274521, -0.29141318798065186, 0.2436991184949875, -0.37010371685028076 ]
https://github.com/huggingface/datasets/issues/160
caching in map causes same result to be returned for train, validation and test
Hi, the full example was listed in the PR above, but here is the exact link: https://github.com/dpressel/mead-baseline/blob/3c1aa3ca062cb23f303ca98ac40b6652b37ee971/api-examples/layers-classify-hf-datasets.py The problem is coming from ``` if cache_file_name is None: # we create a unique hash from the function, current dataset file and the mapping args cache_kwargs = { "with_indices": with_indices, "batched": batched, "batch_size": batch_size, "remove_columns": remove_columns, "keep_in_memory": keep_in_memory, "load_from_cache_file": load_from_cache_file, "cache_file_name": cache_file_name, "writer_batch_size": writer_batch_size, "arrow_schema": arrow_schema, "disable_nullable": disable_nullable, } cache_file_name = self._get_cache_file_path(function, cache_kwargs) ``` The cached value is always the same, but I was able to change that by just renaming the function each time which seems to fix the issue.
hello, I am working on a program that uses the `nlp` library with the `SST2` dataset. The rough outline of the program is: ``` import nlp as nlp_datasets ... parser.add_argument('--dataset', help='HuggingFace Datasets id', default=['glue', 'sst2'], nargs='+') ... dataset = nlp_datasets.load_dataset(*args.dataset) ... # Create feature vocabs vocabs = create_vocabs(dataset.values(), vectorizers) ... # Create a function to vectorize based on vectorizers and vocabs: print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) # factory method to create a `convert_to_features` function based on vocabs convert_to_features = create_featurizer(vectorizers, vocabs) train_set = train_set.map(convert_to_features, batched=True) train_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) train_loader = torch.utils.data.DataLoader(train_set, batch_size=args.batchsz) valid_set = valid_set.map(convert_to_features, batched=True) valid_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) valid_loader = torch.utils.data.DataLoader(valid_set, batch_size=args.batchsz) test_set = test_set.map(convert_to_features, batched=True) test_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) test_loader = torch.utils.data.DataLoader(test_set, batch_size=args.batchsz) print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) ``` Im not sure if Im using it incorrectly, but the results are not what I expect. Namely, the `.map()` seems to grab the datset from the cache and then loses track of what the specific dataset is, instead using my training data for all datasets: ``` TS 67349 VS 872 ES 1821 TS 67349 VS 67349 ES 67349 ``` The behavior changes if I turn off the caching but then the results fail: ``` train_set = train_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... valid_set = valid_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... test_set = test_set.map(convert_to_features, batched=True, load_from_cache_file=False) ``` Now I get the right set of features back... ``` TS 67349 VS 872 ES 1821 100%|██████████| 68/68 [00:00<00:00, 92.78it/s] 100%|██████████| 1/1 [00:00<00:00, 75.47it/s] 0%| | 0/2 [00:00<?, ?it/s]TS 67349 VS 872 ES 1821 100%|██████████| 2/2 [00:00<00:00, 77.19it/s] ``` but I think its losing track of the original training set: ``` Traceback (most recent call last): File "/home/dpressel/dev/work/baseline/api-examples/layers-classify-hf-datasets.py", line 148, in <module> for x in train_loader: File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 345, in __next__ data = self._next_data() File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 385, in _next_data data = self._dataset_fetcher.fetch(index) # may raise StopIteration File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp> data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 338, in __getitem__ output_all_columns=self._output_all_columns, File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 294, in _getitem outputs = self._unnest(self._data.slice(key, 1).to_pydict()) File "pyarrow/table.pxi", line 1211, in pyarrow.lib.Table.slice File "pyarrow/public-api.pxi", line 390, in pyarrow.lib.pyarrow_wrap_table File "pyarrow/error.pxi", line 85, in pyarrow.lib.check_status pyarrow.lib.ArrowInvalid: Column 3: In chunk 0: Invalid: Length spanned by list offsets (15859698) larger than values array (length 100000) Process finished with exit code 1 ``` The full-example program (minus the print stmts) is here: https://github.com/dpressel/mead-baseline/pull/620/files
99
caching in map causes same result to be returned for train, validation and test hello, I am working on a program that uses the `nlp` library with the `SST2` dataset. The rough outline of the program is: ``` import nlp as nlp_datasets ... parser.add_argument('--dataset', help='HuggingFace Datasets id', default=['glue', 'sst2'], nargs='+') ... dataset = nlp_datasets.load_dataset(*args.dataset) ... # Create feature vocabs vocabs = create_vocabs(dataset.values(), vectorizers) ... # Create a function to vectorize based on vectorizers and vocabs: print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) # factory method to create a `convert_to_features` function based on vocabs convert_to_features = create_featurizer(vectorizers, vocabs) train_set = train_set.map(convert_to_features, batched=True) train_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) train_loader = torch.utils.data.DataLoader(train_set, batch_size=args.batchsz) valid_set = valid_set.map(convert_to_features, batched=True) valid_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) valid_loader = torch.utils.data.DataLoader(valid_set, batch_size=args.batchsz) test_set = test_set.map(convert_to_features, batched=True) test_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) test_loader = torch.utils.data.DataLoader(test_set, batch_size=args.batchsz) print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) ``` Im not sure if Im using it incorrectly, but the results are not what I expect. Namely, the `.map()` seems to grab the datset from the cache and then loses track of what the specific dataset is, instead using my training data for all datasets: ``` TS 67349 VS 872 ES 1821 TS 67349 VS 67349 ES 67349 ``` The behavior changes if I turn off the caching but then the results fail: ``` train_set = train_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... valid_set = valid_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... test_set = test_set.map(convert_to_features, batched=True, load_from_cache_file=False) ``` Now I get the right set of features back... ``` TS 67349 VS 872 ES 1821 100%|██████████| 68/68 [00:00<00:00, 92.78it/s] 100%|██████████| 1/1 [00:00<00:00, 75.47it/s] 0%| | 0/2 [00:00<?, ?it/s]TS 67349 VS 872 ES 1821 100%|██████████| 2/2 [00:00<00:00, 77.19it/s] ``` but I think its losing track of the original training set: ``` Traceback (most recent call last): File "/home/dpressel/dev/work/baseline/api-examples/layers-classify-hf-datasets.py", line 148, in <module> for x in train_loader: File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 345, in __next__ data = self._next_data() File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 385, in _next_data data = self._dataset_fetcher.fetch(index) # may raise StopIteration File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp> data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 338, in __getitem__ output_all_columns=self._output_all_columns, File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 294, in _getitem outputs = self._unnest(self._data.slice(key, 1).to_pydict()) File "pyarrow/table.pxi", line 1211, in pyarrow.lib.Table.slice File "pyarrow/public-api.pxi", line 390, in pyarrow.lib.pyarrow_wrap_table File "pyarrow/error.pxi", line 85, in pyarrow.lib.check_status pyarrow.lib.ArrowInvalid: Column 3: In chunk 0: Invalid: Length spanned by list offsets (15859698) larger than values array (length 100000) Process finished with exit code 1 ``` The full-example program (minus the print stmts) is here: https://github.com/dpressel/mead-baseline/pull/620/files Hi, the full example was listed in the PR above, but here is the exact link: https://github.com/dpressel/mead-baseline/blob/3c1aa3ca062cb23f303ca98ac40b6652b37ee971/api-examples/layers-classify-hf-datasets.py The problem is coming from ``` if cache_file_name is None: # we create a unique hash from the function, current dataset file and the mapping args cache_kwargs = { "with_indices": with_indices, "batched": batched, "batch_size": batch_size, "remove_columns": remove_columns, "keep_in_memory": keep_in_memory, "load_from_cache_file": load_from_cache_file, "cache_file_name": cache_file_name, "writer_batch_size": writer_batch_size, "arrow_schema": arrow_schema, "disable_nullable": disable_nullable, } cache_file_name = self._get_cache_file_path(function, cache_kwargs) ``` The cached value is always the same, but I was able to change that by just renaming the function each time which seems to fix the issue.
[ -0.12632250785827637, -0.22070901095867157, -0.10063198953866959, 0.27801084518432617, 0.24547572433948517, -0.20522341132164001, 0.137210875749588, 0.5128593444824219, 0.18828663229942322, -0.06233531981706619, 0.12781980633735657, 0.38913458585739136, 0.003776297438889742, -0.1606176495552063, -0.010743561200797558, 0.024212030693888664, 0.09608013927936554, 0.23505231738090515, -0.14556236565113068, -0.1648235321044922, 0.08189409971237183, 0.2129841297864914, -0.042409978806972504, 0.11652691662311554, -0.39034950733184814, 0.041289906948804855, 0.21235668659210205, -0.2308902144432068, 0.06845776736736298, -0.25706538558006287, 0.1678507924079895, -0.09324328601360321, 0.02583368308842182, 0.16631756722927094, -0.00010973814642056823, 0.0067980014719069, -0.01538778841495514, -0.13856948912143707, -0.007274831645190716, 0.04019518941640854, 0.11813383549451828, -0.21172897517681122, -0.10918274521827698, -0.3121316730976105, -0.3320115804672241, -0.03766085207462311, 0.01780932955443859, -0.35565948486328125, 0.31285759806632996, 0.14745205640792847, 0.26071032881736755, 0.22490893304347992, -0.2538125216960907, 0.21113626658916473, 0.14512759447097778, -0.029327336698770523, -0.17419742047786713, -0.04230547323822975, -0.04397892951965332, -0.33446553349494934, -0.37676504254341125, 0.5254571437835693, -0.07658161967992783, 0.17014139890670776, 0.21466633677482605, 0.22505579888820648, 0.17260590195655823, 0.035937704145908356, 0.1429363638162613, -0.11396253854036331, 0.011094827204942703, -0.22691738605499268, -0.027116434648633003, -0.3733721077442169, -0.4186401665210724, -0.07852599024772644, 0.13762502372264862, 0.18104472756385803, -0.1363932490348816, -0.06371653825044632, -0.43193766474723816, 0.2392898052930832, 0.0652494877576828, -0.012788408435881138, 0.11952681839466095, 0.29516080021858215, 0.04380622133612633, 0.19522294402122498, 0.12944649159908295, -0.06973281502723694, -0.022558962926268578, -0.21451696753501892, -0.05897140875458717, 0.22640138864517212, -0.4359849691390991, 0.07208239287137985, 0.4315200448036194, -0.1404140740633011, -0.053242750465869904, 0.03063775785267353, 0.3919311761856079, 0.07921352982521057, 0.11681784689426422, -0.008491764776408672, -0.10416068881750107, 0.5898676514625549, 0.04432177543640137, 0.17118391394615173, -0.014969921670854092, -0.19827847182750702, -0.3560168445110321, 0.004987859632819891, 0.2880178987979889, -0.28660091757774353, 0.42841145396232605, 0.40236055850982666, -0.04226037487387657, -0.12152973562479019, -0.030102189630270004, 0.09402408450841904, -0.33085379004478455, 0.05566990002989769, 0.058729998767375946, 0.31395798921585083, -0.22545045614242554, 0.14926567673683167, -0.17376750707626343, -0.10335483402013779, -0.3849693238735199, 0.12735125422477722, -0.33784618973731995, -0.033323898911476135, -0.4962189197540283, 0.0015165578806772828, 0.28538474440574646, -0.002859933767467737, 0.28535225987434387, -0.038445454090833664, -0.13722018897533417, -0.3019891679286957, 0.10317365825176239, -0.39851170778274536, 0.5702314376831055, -0.1072295531630516, -0.19701702892780304, 0.24102117121219635, 0.13037829101085663, 0.08526421338319778, -0.29768794775009155, -0.19837751984596252, -0.1698075532913208, -0.20906901359558105, 0.2598607540130615, 0.24224542081356049, -0.07942765206098557, 0.14139588177204132, 0.09248226881027222, 0.0776788741350174, 0.3432585299015045, -0.232854962348938, 0.10284040868282318, -0.20148923993110657, -0.28198012709617615, -0.2043590545654297, 0.16256873309612274, 0.2623099982738495, 0.06677649915218353, -0.13737238943576813, 0.4017971158027649, 0.2371826320886612, 0.14729078114032745, 0.4138067364692688, -0.1965339034795761, 0.30949410796165466, -0.43349170684814453, 0.3486190438270569, 0.506476879119873, -0.32673147320747375, -0.49634483456611633, 0.09944828599691391, -0.15631580352783203, 0.1002880334854126, 0.10244903713464737, 0.15137287974357605, 0.029087670147418976, 0.014419378712773323, 0.1285371035337448, 0.018702296540141106, -0.0005146907060407102, 0.208420529961586, -0.31692788004875183, -0.021933916956186295, 0.5500646233558655, -0.2524673640727997, 0.02847237139940262, -0.13123734295368195, -0.05504271760582924, 0.021933000534772873, 0.006768169347196817, -0.11219003051519394, 0.12185932695865631, 0.11460470408201218, -0.056878868490457535, -0.1581844687461853, -0.10643851011991501, -0.024692509323358536, -0.28826257586479187, 0.23830120265483856, -0.14981895685195923, -0.0999680832028389, 0.06172538921236992, -0.09704917669296265, -0.16200260818004608, -0.12927502393722534, -0.18091638386249542, -0.27306458353996277, 0.20130178332328796, 0.23052407801151276, 0.3202398419380188, -0.07290733605623245, 0.2252090722322464, 0.3680817782878876, 0.048363614827394485, -0.21607212722301483, -0.4300347566604614, 0.08568026125431061, -0.17773471772670746, -0.2188279777765274, -0.0748467668890953, 0.22306986153125763, 0.2090262770652771, -0.04792642965912819, -0.14242812991142273, 0.14827579259872437, -0.2220856100320816, 0.4189179241657257, -0.13702799379825592, 0.269452840089798, -0.06965683400630951, 0.029306238517165184, 0.016667358577251434, 0.25400519371032715, 0.05881987884640694, -0.08980134129524231, -0.07270598411560059, 0.47423407435417175, 0.18140216171741486, 0.17011605203151703, -0.048070792108774185, -0.21626794338226318, 0.051984209567308426, -0.17831753194332123, -0.12091574817895889, -0.11881643533706665, 0.13640747964382172, -0.08057200163602829, 0.1488880068063736, 0.15232868492603302, -0.10698743164539337, 0.3120083510875702, 0.7600833773612976, 0.19691871106624603, -0.011016919277608395, 0.015740947797894478, -0.057213034480810165, -0.21779173612594604, 0.3338564336299896, -0.05868269130587578, 0.31038790941238403, 0.10473166406154633, 0.12716487050056458, -0.04620952904224396, -0.21965020895004272, -0.15443570911884308, 0.12464798986911774, -0.21436883509159088, -0.1919763833284378, 0.308034211397171, 0.28208687901496887, -0.15836724638938904, -0.35378316044807434, 0.0940157026052475, -0.01276430580765009, 0.026191623881459236, -0.15800140798091888, 0.21723663806915283, -0.1665293276309967, -0.2780851721763611, -0.28641536831855774, 0.13615694642066956, -0.12220374494791031, -0.0921654924750328, -0.04126313701272011, 0.42542964220046997, -0.036654673516750336, 0.19515161216259003, -0.3149721920490265, 0.08951952308416367, 0.226604163646698, -0.2663276791572571, -0.07168538868427277, -0.1902879923582077, -0.39492008090019226, 0.04223187640309334, -0.09523041546344757, -0.0929960310459137, 0.3783297538757324, -0.04731621965765953, -0.06796029210090637, -0.2881457805633545, -0.1725403368473053, 0.10717813670635223, -0.09261170029640198, -0.15536807477474213, 0.16724860668182373, -0.03133843466639519, -0.10115880519151688, -0.2987455725669861, 0.29379722476005554, -0.3053131401538849, -0.29914355278015137, -0.05198654159903526, 0.09890630096197128, 0.03010939620435238, -0.28034764528274536, -0.42862412333488464, 0.27357834577560425, -0.31724390387535095, 0.07546782493591309, -0.14892923831939697, 0.18622653186321259, 0.33093470335006714, 0.008953339420258999, -0.014561006799340248, -0.06404032558202744, 0.16439402103424072, -0.46741002798080444, -0.1409541219472885, 0.32257840037345886, 0.01876099780201912, -0.21630428731441498, -0.32459139823913574, -0.27572473883628845, 0.347724050283432, 0.19238898158073425, -0.5803960561752319, -0.21442314982414246, -0.04890046268701553, 0.24803197383880615, -0.05125873163342476, -0.04649535194039345, 0.5745421648025513, 0.01972527615725994, -0.17946361005306244, -0.16727842390537262, -0.33184051513671875, 0.35973140597343445, 0.41926297545433044, 0.2494877278804779, -0.011159147135913372, 0.17781507968902588, 0.20092034339904785, 0.6707466244697571, 0.20191961526870728, -0.25905755162239075, 0.3317307233810425, 0.10723208636045456, 0.07499753683805466, -0.2034970372915268, -0.3130224943161011, 0.18576058745384216, -0.18160304427146912, 0.019732147455215454, 0.2719840407371521, 0.006251134444028139, -0.2801845371723175, 0.01774226687848568, 0.21881961822509766, -0.2253446877002716, -0.3280579447746277, 0.2597459852695465, -0.34242793917655945, 0.24094845354557037, 0.03637058660387993, 0.31580042839050293, -0.5820142030715942, -0.24272412061691284, 0.19462814927101135, -0.33594635128974915, 0.2798454165458679, 0.04449770972132683, -0.5836008787155151, 0.12147048860788345, -0.538021981716156, 0.18194131553173065, 0.3146284818649292, 0.2724076807498932, -0.09202561527490616, -0.2465258091688156, 0.04334738850593567, -0.14713139832019806, 0.5711411833763123, 0.040972813963890076, 0.4085797369480133, 0.1699264794588089, -0.1302468627691269, -0.1994732767343521, -0.2247127890586853, 0.017234763130545616, 0.2801370620727539, 0.32649919390678406, 0.40084776282310486, -0.3452445864677429, -0.36327362060546875, -0.26576584577560425, 0.007364223711192608, -0.04618128761649132, 0.007107758894562721, -0.12583346664905548, 0.18938525021076202, 0.01648680679500103, 0.17434526979923248, -0.01498047262430191, 0.1183028295636177, 0.016039855778217316, -0.13654641807079315, -0.03913364186882973, -0.21942690014839172, 0.1282409280538559, 0.2264377325773239, 0.4735944867134094, 0.15032504498958588, 0.03246142715215683, 0.033214494585990906, 0.21515744924545288, -0.5671088099479675, 0.26595762372016907, -0.08884964138269424, -0.004131742753088474, 0.07464420050382614, 0.14443816244602203, -0.03030155971646309, 0.3378978669643402, -0.04252500832080841, 0.14571364223957062, -0.12862949073314667, 0.008956843055784702, -0.5539686679840088, 0.4854544401168823, 0.3874591886997223, 0.13246646523475647, -0.04488928243517876, -0.24685749411582947, 0.17346294224262238, 0.07412006705999374, -0.2765309810638428, 0.1322597712278366, -0.4306926131248474, -0.43199479579925537, 0.1857973337173462, 0.20948611199855804, 0.9232466816902161, 0.332022100687027, 0.3719581067562103, 0.1007041484117508, 0.10358098894357681, 0.2642512321472168, -0.23411183059215546, 0.40925082564353943, -0.3693642020225525, -0.0463179275393486, 0.024538656696677208, -0.16673317551612854, -0.22738394141197205, -0.010426265187561512, -0.01099333819001913, 0.30259495973587036, 0.32405051589012146, 0.5999365448951721, -0.11211955547332764, 0.16669917106628418, 0.21569165587425232, -0.15811017155647278, -0.2492315024137497, 0.2010732889175415, -0.29230374097824097, 0.3152405023574829, -0.11240831762552261, -0.07410863041877747, -0.01650959439575672, -0.35134196281433105, 0.06714027374982834, 0.06094216927886009, -0.3365170657634735, 0.3048412501811981, 0.13346198201179504, -0.27266666293144226, -0.2969881594181061, 0.2550773024559021, 0.11198677867650986, 0.08252744376659393, -0.048600129783153534, -0.03518514335155487, 0.40064483880996704, 0.3504585027694702, -0.006891036406159401, -0.20158861577510834, 0.1864127367734909, 0.06387528032064438, -0.012894121930003166, 0.03024262934923172, -0.05640564113855362, -0.3015829622745514, -0.25251179933547974, 0.14820435643196106, 0.003372181672602892, -0.41296184062957764, -0.05531826987862587, 0.061445046216249466, -0.066788449883461, -0.09413714706897736, 0.15088026225566864, -0.026068205013871193, -0.10794040560722351, 0.3151542544364929, 0.1173383817076683, -0.24304209649562836, 0.05235802382230759, 0.47483208775520325, -0.040397144854068756, 0.03181783854961395, 0.41542795300483704, -0.15281881392002106, -0.10380645841360092, -0.2553911507129669, -0.28417274355888367, -0.10205967724323273, -0.44884082674980164, 0.11220790445804596, -0.2194061428308487, -0.4019114077091217, 0.0442325733602047, 0.280818372964859, 0.014706877060234547, 0.31849586963653564, -0.011618148535490036, -0.12495850026607513, -0.37697234749794006, -0.10379048436880112, -0.1065821498632431, 0.2671777307987213, 0.2795186936855316, 0.2843484878540039, -0.29586702585220337, 0.2492261826992035, -0.38310956954956055, -0.1071421355009079, -0.18716979026794434, 0.4379133880138397, -0.03913804888725281, -0.3655827045440674, -0.009147044271230698, 0.00185489805880934, 0.12997117638587952, 0.17754493653774261, -0.6386589407920837, -0.2926190495491028, -0.09099180996417999, 0.08186071366071701, -0.007587700616568327, -0.159822016954422, -0.23928211629390717, -0.017213808372616768, 0.18238461017608643, -0.42740321159362793, 0.09215689450502396, 0.15121674537658691, 0.047031741589307785, 0.23718656599521637, -0.16102980077266693, 0.12145575135946274, 0.2881677746772766, -0.18652726709842682, -0.18414875864982605, -0.07622820883989334, 0.08048128336668015, 0.2570165991783142, -0.069468192756176, 0.02047928236424923, -0.21040309965610504, 0.03047284670174122, 0.08927395939826965, 0.1427948772907257, 0.23692886531352997, -0.024752512574195862, -0.2156609743833542, 0.0005510254995897412, 0.1827995628118515, 0.11411970853805542, -0.07855338603258133, -0.32897529006004333, 0.2802230417728424, 0.27836501598358154, -0.11751806735992432, -0.010236726142466068, 0.16770757734775543, -0.09938807040452957, 0.2528513967990875, 0.2071538120508194, 0.10758750885725021, 0.052611079066991806, 0.26961180567741394, 0.12649419903755188, 0.40504035353660583, -0.34832870960235596, 0.08942142128944397, 0.06441634148359299, 0.10164941847324371, 0.0377780981361866, 0.3812146782875061, -0.15324392914772034, 0.08992191404104233, 0.14838334918022156, 0.18833690881729126, 0.20253953337669373, 0.2628898620605469, 0.21354152262210846, -0.1878644973039627, -0.03412109240889549, -0.24108682572841644, 0.14363647997379303, -0.14272600412368774, 0.4156598746776581, 0.15059241652488708, 0.19285538792610168, -0.11485260725021362, -0.4619389772415161, -0.24626794457435608, 0.23536370694637299, -0.3245418071746826, -0.30920886993408203, -0.04867832362651825, -0.08018530905246735, 0.05592920631170273, 0.1945614516735077, -0.07929019629955292, 0.20385901629924774, 0.34283849596977234, -0.008660116232931614, -0.013750203885138035, -0.08435358107089996, -0.17061741650104523, 0.11397870630025864, 0.27486518025398254, -0.08173874765634537, 0.44693678617477417, 0.01080083753913641, -0.05156949535012245, -0.2344711422920227, 0.41141751408576965, 0.28612861037254333, 0.1949474811553955, -0.168862447142601, 0.06395307183265686, 0.20064708590507507, 0.0306134894490242, -0.3546890914440155, 0.3286518454551697, 0.1511448621749878, -0.2930276095867157, 0.36800575256347656, 0.12379536032676697, -0.23215578496456146, 0.2764037549495697, -0.10419601202011108, -0.20146721601486206, -0.04988892748951912, 0.3224368989467621, 0.0997074767947197, -0.0399320125579834, -0.19496309757232666, 0.11112021654844284, -0.14457158744335175, -0.28533950448036194, 0.15934592485427856, -0.02389831282198429, 0.46766093373298645, -0.275675505399704, 0.11720515042543411, -0.06182502955198288, 0.6220859885215759, -0.02922753430902958, 0.12432888895273209, -0.350137859582901, 0.10264351218938828, -0.33316680788993835, 0.1984093189239502, -0.08381783962249756, 0.15374279022216797, -0.3975840210914612, 0.20426496863365173, -0.11549855023622513, -0.023466093465685844, -0.11804244667291641, -0.00788638461381197, 0.14740723371505737, 0.17893213033676147, -0.3869176506996155, 0.1554814726114273, -0.0862593725323677, -0.00626726308837533, 0.2917737364768982, -0.37667545676231384, 0.22199058532714844, 0.04247966781258583, 0.20691059529781342, 0.010986385866999626, -0.02172369882464409, -0.054806627333164215, 0.33657434582710266, 0.6232408285140991, 0.34702974557876587, 0.14255116879940033, -0.10102774947881699, 0.08155330270528793, -0.34743764996528625, 0.014104106463491917, -0.06088816747069359, -0.20964983105659485, 0.09756835550069809, 0.5258304476737976, -0.22562745213508606, 0.13194629549980164, -0.17804564535617828, 0.2366943657398224, 0.38227659463882446, -0.15272878110408783, 0.06179971620440483, 0.048225898295640945, -0.25984057784080505, -0.005199328530579805, -0.12586632370948792, 0.45379582047462463, 0.007182890549302101, 0.32383018732070923, -0.3592907786369324, -0.5417624711990356, 0.4329099953174591, -0.4511263966560364, -0.3025250732898712, 0.12327846884727478, 0.23878563940525055, 0.34956884384155273, -0.06000060588121414, -0.5436676144599915, -0.1724908947944641, 0.26911357045173645, -0.030661094933748245, -0.15704523026943207, 0.20768482983112335, -0.1486440747976303, -0.04107564687728882, -0.11675458401441574, 0.17646899819374084, 0.10152711719274521, -0.29141318798065186, 0.2436991184949875, -0.37010371685028076 ]
https://github.com/huggingface/datasets/issues/160
caching in map causes same result to be returned for train, validation and test
Ok, I think @lhoestq has already found a solution :-) Maybe you can chime in @lhoestq
hello, I am working on a program that uses the `nlp` library with the `SST2` dataset. The rough outline of the program is: ``` import nlp as nlp_datasets ... parser.add_argument('--dataset', help='HuggingFace Datasets id', default=['glue', 'sst2'], nargs='+') ... dataset = nlp_datasets.load_dataset(*args.dataset) ... # Create feature vocabs vocabs = create_vocabs(dataset.values(), vectorizers) ... # Create a function to vectorize based on vectorizers and vocabs: print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) # factory method to create a `convert_to_features` function based on vocabs convert_to_features = create_featurizer(vectorizers, vocabs) train_set = train_set.map(convert_to_features, batched=True) train_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) train_loader = torch.utils.data.DataLoader(train_set, batch_size=args.batchsz) valid_set = valid_set.map(convert_to_features, batched=True) valid_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) valid_loader = torch.utils.data.DataLoader(valid_set, batch_size=args.batchsz) test_set = test_set.map(convert_to_features, batched=True) test_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) test_loader = torch.utils.data.DataLoader(test_set, batch_size=args.batchsz) print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) ``` Im not sure if Im using it incorrectly, but the results are not what I expect. Namely, the `.map()` seems to grab the datset from the cache and then loses track of what the specific dataset is, instead using my training data for all datasets: ``` TS 67349 VS 872 ES 1821 TS 67349 VS 67349 ES 67349 ``` The behavior changes if I turn off the caching but then the results fail: ``` train_set = train_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... valid_set = valid_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... test_set = test_set.map(convert_to_features, batched=True, load_from_cache_file=False) ``` Now I get the right set of features back... ``` TS 67349 VS 872 ES 1821 100%|██████████| 68/68 [00:00<00:00, 92.78it/s] 100%|██████████| 1/1 [00:00<00:00, 75.47it/s] 0%| | 0/2 [00:00<?, ?it/s]TS 67349 VS 872 ES 1821 100%|██████████| 2/2 [00:00<00:00, 77.19it/s] ``` but I think its losing track of the original training set: ``` Traceback (most recent call last): File "/home/dpressel/dev/work/baseline/api-examples/layers-classify-hf-datasets.py", line 148, in <module> for x in train_loader: File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 345, in __next__ data = self._next_data() File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 385, in _next_data data = self._dataset_fetcher.fetch(index) # may raise StopIteration File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp> data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 338, in __getitem__ output_all_columns=self._output_all_columns, File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 294, in _getitem outputs = self._unnest(self._data.slice(key, 1).to_pydict()) File "pyarrow/table.pxi", line 1211, in pyarrow.lib.Table.slice File "pyarrow/public-api.pxi", line 390, in pyarrow.lib.pyarrow_wrap_table File "pyarrow/error.pxi", line 85, in pyarrow.lib.check_status pyarrow.lib.ArrowInvalid: Column 3: In chunk 0: Invalid: Length spanned by list offsets (15859698) larger than values array (length 100000) Process finished with exit code 1 ``` The full-example program (minus the print stmts) is here: https://github.com/dpressel/mead-baseline/pull/620/files
16
caching in map causes same result to be returned for train, validation and test hello, I am working on a program that uses the `nlp` library with the `SST2` dataset. The rough outline of the program is: ``` import nlp as nlp_datasets ... parser.add_argument('--dataset', help='HuggingFace Datasets id', default=['glue', 'sst2'], nargs='+') ... dataset = nlp_datasets.load_dataset(*args.dataset) ... # Create feature vocabs vocabs = create_vocabs(dataset.values(), vectorizers) ... # Create a function to vectorize based on vectorizers and vocabs: print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) # factory method to create a `convert_to_features` function based on vocabs convert_to_features = create_featurizer(vectorizers, vocabs) train_set = train_set.map(convert_to_features, batched=True) train_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) train_loader = torch.utils.data.DataLoader(train_set, batch_size=args.batchsz) valid_set = valid_set.map(convert_to_features, batched=True) valid_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) valid_loader = torch.utils.data.DataLoader(valid_set, batch_size=args.batchsz) test_set = test_set.map(convert_to_features, batched=True) test_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) test_loader = torch.utils.data.DataLoader(test_set, batch_size=args.batchsz) print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) ``` Im not sure if Im using it incorrectly, but the results are not what I expect. Namely, the `.map()` seems to grab the datset from the cache and then loses track of what the specific dataset is, instead using my training data for all datasets: ``` TS 67349 VS 872 ES 1821 TS 67349 VS 67349 ES 67349 ``` The behavior changes if I turn off the caching but then the results fail: ``` train_set = train_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... valid_set = valid_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... test_set = test_set.map(convert_to_features, batched=True, load_from_cache_file=False) ``` Now I get the right set of features back... ``` TS 67349 VS 872 ES 1821 100%|██████████| 68/68 [00:00<00:00, 92.78it/s] 100%|██████████| 1/1 [00:00<00:00, 75.47it/s] 0%| | 0/2 [00:00<?, ?it/s]TS 67349 VS 872 ES 1821 100%|██████████| 2/2 [00:00<00:00, 77.19it/s] ``` but I think its losing track of the original training set: ``` Traceback (most recent call last): File "/home/dpressel/dev/work/baseline/api-examples/layers-classify-hf-datasets.py", line 148, in <module> for x in train_loader: File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 345, in __next__ data = self._next_data() File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 385, in _next_data data = self._dataset_fetcher.fetch(index) # may raise StopIteration File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp> data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 338, in __getitem__ output_all_columns=self._output_all_columns, File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 294, in _getitem outputs = self._unnest(self._data.slice(key, 1).to_pydict()) File "pyarrow/table.pxi", line 1211, in pyarrow.lib.Table.slice File "pyarrow/public-api.pxi", line 390, in pyarrow.lib.pyarrow_wrap_table File "pyarrow/error.pxi", line 85, in pyarrow.lib.check_status pyarrow.lib.ArrowInvalid: Column 3: In chunk 0: Invalid: Length spanned by list offsets (15859698) larger than values array (length 100000) Process finished with exit code 1 ``` The full-example program (minus the print stmts) is here: https://github.com/dpressel/mead-baseline/pull/620/files Ok, I think @lhoestq has already found a solution :-) Maybe you can chime in @lhoestq
[ -0.12632250785827637, -0.22070901095867157, -0.10063198953866959, 0.27801084518432617, 0.24547572433948517, -0.20522341132164001, 0.137210875749588, 0.5128593444824219, 0.18828663229942322, -0.06233531981706619, 0.12781980633735657, 0.38913458585739136, 0.003776297438889742, -0.1606176495552063, -0.010743561200797558, 0.024212030693888664, 0.09608013927936554, 0.23505231738090515, -0.14556236565113068, -0.1648235321044922, 0.08189409971237183, 0.2129841297864914, -0.042409978806972504, 0.11652691662311554, -0.39034950733184814, 0.041289906948804855, 0.21235668659210205, -0.2308902144432068, 0.06845776736736298, -0.25706538558006287, 0.1678507924079895, -0.09324328601360321, 0.02583368308842182, 0.16631756722927094, -0.00010973814642056823, 0.0067980014719069, -0.01538778841495514, -0.13856948912143707, -0.007274831645190716, 0.04019518941640854, 0.11813383549451828, -0.21172897517681122, -0.10918274521827698, -0.3121316730976105, -0.3320115804672241, -0.03766085207462311, 0.01780932955443859, -0.35565948486328125, 0.31285759806632996, 0.14745205640792847, 0.26071032881736755, 0.22490893304347992, -0.2538125216960907, 0.21113626658916473, 0.14512759447097778, -0.029327336698770523, -0.17419742047786713, -0.04230547323822975, -0.04397892951965332, -0.33446553349494934, -0.37676504254341125, 0.5254571437835693, -0.07658161967992783, 0.17014139890670776, 0.21466633677482605, 0.22505579888820648, 0.17260590195655823, 0.035937704145908356, 0.1429363638162613, -0.11396253854036331, 0.011094827204942703, -0.22691738605499268, -0.027116434648633003, -0.3733721077442169, -0.4186401665210724, -0.07852599024772644, 0.13762502372264862, 0.18104472756385803, -0.1363932490348816, -0.06371653825044632, -0.43193766474723816, 0.2392898052930832, 0.0652494877576828, -0.012788408435881138, 0.11952681839466095, 0.29516080021858215, 0.04380622133612633, 0.19522294402122498, 0.12944649159908295, -0.06973281502723694, -0.022558962926268578, -0.21451696753501892, -0.05897140875458717, 0.22640138864517212, -0.4359849691390991, 0.07208239287137985, 0.4315200448036194, -0.1404140740633011, -0.053242750465869904, 0.03063775785267353, 0.3919311761856079, 0.07921352982521057, 0.11681784689426422, -0.008491764776408672, -0.10416068881750107, 0.5898676514625549, 0.04432177543640137, 0.17118391394615173, -0.014969921670854092, -0.19827847182750702, -0.3560168445110321, 0.004987859632819891, 0.2880178987979889, -0.28660091757774353, 0.42841145396232605, 0.40236055850982666, -0.04226037487387657, -0.12152973562479019, -0.030102189630270004, 0.09402408450841904, -0.33085379004478455, 0.05566990002989769, 0.058729998767375946, 0.31395798921585083, -0.22545045614242554, 0.14926567673683167, -0.17376750707626343, -0.10335483402013779, -0.3849693238735199, 0.12735125422477722, -0.33784618973731995, -0.033323898911476135, -0.4962189197540283, 0.0015165578806772828, 0.28538474440574646, -0.002859933767467737, 0.28535225987434387, -0.038445454090833664, -0.13722018897533417, -0.3019891679286957, 0.10317365825176239, -0.39851170778274536, 0.5702314376831055, -0.1072295531630516, -0.19701702892780304, 0.24102117121219635, 0.13037829101085663, 0.08526421338319778, -0.29768794775009155, -0.19837751984596252, -0.1698075532913208, -0.20906901359558105, 0.2598607540130615, 0.24224542081356049, -0.07942765206098557, 0.14139588177204132, 0.09248226881027222, 0.0776788741350174, 0.3432585299015045, -0.232854962348938, 0.10284040868282318, -0.20148923993110657, -0.28198012709617615, -0.2043590545654297, 0.16256873309612274, 0.2623099982738495, 0.06677649915218353, -0.13737238943576813, 0.4017971158027649, 0.2371826320886612, 0.14729078114032745, 0.4138067364692688, -0.1965339034795761, 0.30949410796165466, -0.43349170684814453, 0.3486190438270569, 0.506476879119873, -0.32673147320747375, -0.49634483456611633, 0.09944828599691391, -0.15631580352783203, 0.1002880334854126, 0.10244903713464737, 0.15137287974357605, 0.029087670147418976, 0.014419378712773323, 0.1285371035337448, 0.018702296540141106, -0.0005146907060407102, 0.208420529961586, -0.31692788004875183, -0.021933916956186295, 0.5500646233558655, -0.2524673640727997, 0.02847237139940262, -0.13123734295368195, -0.05504271760582924, 0.021933000534772873, 0.006768169347196817, -0.11219003051519394, 0.12185932695865631, 0.11460470408201218, -0.056878868490457535, -0.1581844687461853, -0.10643851011991501, -0.024692509323358536, -0.28826257586479187, 0.23830120265483856, -0.14981895685195923, -0.0999680832028389, 0.06172538921236992, -0.09704917669296265, -0.16200260818004608, -0.12927502393722534, -0.18091638386249542, -0.27306458353996277, 0.20130178332328796, 0.23052407801151276, 0.3202398419380188, -0.07290733605623245, 0.2252090722322464, 0.3680817782878876, 0.048363614827394485, -0.21607212722301483, -0.4300347566604614, 0.08568026125431061, -0.17773471772670746, -0.2188279777765274, -0.0748467668890953, 0.22306986153125763, 0.2090262770652771, -0.04792642965912819, -0.14242812991142273, 0.14827579259872437, -0.2220856100320816, 0.4189179241657257, -0.13702799379825592, 0.269452840089798, -0.06965683400630951, 0.029306238517165184, 0.016667358577251434, 0.25400519371032715, 0.05881987884640694, -0.08980134129524231, -0.07270598411560059, 0.47423407435417175, 0.18140216171741486, 0.17011605203151703, -0.048070792108774185, -0.21626794338226318, 0.051984209567308426, -0.17831753194332123, -0.12091574817895889, -0.11881643533706665, 0.13640747964382172, -0.08057200163602829, 0.1488880068063736, 0.15232868492603302, -0.10698743164539337, 0.3120083510875702, 0.7600833773612976, 0.19691871106624603, -0.011016919277608395, 0.015740947797894478, -0.057213034480810165, -0.21779173612594604, 0.3338564336299896, -0.05868269130587578, 0.31038790941238403, 0.10473166406154633, 0.12716487050056458, -0.04620952904224396, -0.21965020895004272, -0.15443570911884308, 0.12464798986911774, -0.21436883509159088, -0.1919763833284378, 0.308034211397171, 0.28208687901496887, -0.15836724638938904, -0.35378316044807434, 0.0940157026052475, -0.01276430580765009, 0.026191623881459236, -0.15800140798091888, 0.21723663806915283, -0.1665293276309967, -0.2780851721763611, -0.28641536831855774, 0.13615694642066956, -0.12220374494791031, -0.0921654924750328, -0.04126313701272011, 0.42542964220046997, -0.036654673516750336, 0.19515161216259003, -0.3149721920490265, 0.08951952308416367, 0.226604163646698, -0.2663276791572571, -0.07168538868427277, -0.1902879923582077, -0.39492008090019226, 0.04223187640309334, -0.09523041546344757, -0.0929960310459137, 0.3783297538757324, -0.04731621965765953, -0.06796029210090637, -0.2881457805633545, -0.1725403368473053, 0.10717813670635223, -0.09261170029640198, -0.15536807477474213, 0.16724860668182373, -0.03133843466639519, -0.10115880519151688, -0.2987455725669861, 0.29379722476005554, -0.3053131401538849, -0.29914355278015137, -0.05198654159903526, 0.09890630096197128, 0.03010939620435238, -0.28034764528274536, -0.42862412333488464, 0.27357834577560425, -0.31724390387535095, 0.07546782493591309, -0.14892923831939697, 0.18622653186321259, 0.33093470335006714, 0.008953339420258999, -0.014561006799340248, -0.06404032558202744, 0.16439402103424072, -0.46741002798080444, -0.1409541219472885, 0.32257840037345886, 0.01876099780201912, -0.21630428731441498, -0.32459139823913574, -0.27572473883628845, 0.347724050283432, 0.19238898158073425, -0.5803960561752319, -0.21442314982414246, -0.04890046268701553, 0.24803197383880615, -0.05125873163342476, -0.04649535194039345, 0.5745421648025513, 0.01972527615725994, -0.17946361005306244, -0.16727842390537262, -0.33184051513671875, 0.35973140597343445, 0.41926297545433044, 0.2494877278804779, -0.011159147135913372, 0.17781507968902588, 0.20092034339904785, 0.6707466244697571, 0.20191961526870728, -0.25905755162239075, 0.3317307233810425, 0.10723208636045456, 0.07499753683805466, -0.2034970372915268, -0.3130224943161011, 0.18576058745384216, -0.18160304427146912, 0.019732147455215454, 0.2719840407371521, 0.006251134444028139, -0.2801845371723175, 0.01774226687848568, 0.21881961822509766, -0.2253446877002716, -0.3280579447746277, 0.2597459852695465, -0.34242793917655945, 0.24094845354557037, 0.03637058660387993, 0.31580042839050293, -0.5820142030715942, -0.24272412061691284, 0.19462814927101135, -0.33594635128974915, 0.2798454165458679, 0.04449770972132683, -0.5836008787155151, 0.12147048860788345, -0.538021981716156, 0.18194131553173065, 0.3146284818649292, 0.2724076807498932, -0.09202561527490616, -0.2465258091688156, 0.04334738850593567, -0.14713139832019806, 0.5711411833763123, 0.040972813963890076, 0.4085797369480133, 0.1699264794588089, -0.1302468627691269, -0.1994732767343521, -0.2247127890586853, 0.017234763130545616, 0.2801370620727539, 0.32649919390678406, 0.40084776282310486, -0.3452445864677429, -0.36327362060546875, -0.26576584577560425, 0.007364223711192608, -0.04618128761649132, 0.007107758894562721, -0.12583346664905548, 0.18938525021076202, 0.01648680679500103, 0.17434526979923248, -0.01498047262430191, 0.1183028295636177, 0.016039855778217316, -0.13654641807079315, -0.03913364186882973, -0.21942690014839172, 0.1282409280538559, 0.2264377325773239, 0.4735944867134094, 0.15032504498958588, 0.03246142715215683, 0.033214494585990906, 0.21515744924545288, -0.5671088099479675, 0.26595762372016907, -0.08884964138269424, -0.004131742753088474, 0.07464420050382614, 0.14443816244602203, -0.03030155971646309, 0.3378978669643402, -0.04252500832080841, 0.14571364223957062, -0.12862949073314667, 0.008956843055784702, -0.5539686679840088, 0.4854544401168823, 0.3874591886997223, 0.13246646523475647, -0.04488928243517876, -0.24685749411582947, 0.17346294224262238, 0.07412006705999374, -0.2765309810638428, 0.1322597712278366, -0.4306926131248474, -0.43199479579925537, 0.1857973337173462, 0.20948611199855804, 0.9232466816902161, 0.332022100687027, 0.3719581067562103, 0.1007041484117508, 0.10358098894357681, 0.2642512321472168, -0.23411183059215546, 0.40925082564353943, -0.3693642020225525, -0.0463179275393486, 0.024538656696677208, -0.16673317551612854, -0.22738394141197205, -0.010426265187561512, -0.01099333819001913, 0.30259495973587036, 0.32405051589012146, 0.5999365448951721, -0.11211955547332764, 0.16669917106628418, 0.21569165587425232, -0.15811017155647278, -0.2492315024137497, 0.2010732889175415, -0.29230374097824097, 0.3152405023574829, -0.11240831762552261, -0.07410863041877747, -0.01650959439575672, -0.35134196281433105, 0.06714027374982834, 0.06094216927886009, -0.3365170657634735, 0.3048412501811981, 0.13346198201179504, -0.27266666293144226, -0.2969881594181061, 0.2550773024559021, 0.11198677867650986, 0.08252744376659393, -0.048600129783153534, -0.03518514335155487, 0.40064483880996704, 0.3504585027694702, -0.006891036406159401, -0.20158861577510834, 0.1864127367734909, 0.06387528032064438, -0.012894121930003166, 0.03024262934923172, -0.05640564113855362, -0.3015829622745514, -0.25251179933547974, 0.14820435643196106, 0.003372181672602892, -0.41296184062957764, -0.05531826987862587, 0.061445046216249466, -0.066788449883461, -0.09413714706897736, 0.15088026225566864, -0.026068205013871193, -0.10794040560722351, 0.3151542544364929, 0.1173383817076683, -0.24304209649562836, 0.05235802382230759, 0.47483208775520325, -0.040397144854068756, 0.03181783854961395, 0.41542795300483704, -0.15281881392002106, -0.10380645841360092, -0.2553911507129669, -0.28417274355888367, -0.10205967724323273, -0.44884082674980164, 0.11220790445804596, -0.2194061428308487, -0.4019114077091217, 0.0442325733602047, 0.280818372964859, 0.014706877060234547, 0.31849586963653564, -0.011618148535490036, -0.12495850026607513, -0.37697234749794006, -0.10379048436880112, -0.1065821498632431, 0.2671777307987213, 0.2795186936855316, 0.2843484878540039, -0.29586702585220337, 0.2492261826992035, -0.38310956954956055, -0.1071421355009079, -0.18716979026794434, 0.4379133880138397, -0.03913804888725281, -0.3655827045440674, -0.009147044271230698, 0.00185489805880934, 0.12997117638587952, 0.17754493653774261, -0.6386589407920837, -0.2926190495491028, -0.09099180996417999, 0.08186071366071701, -0.007587700616568327, -0.159822016954422, -0.23928211629390717, -0.017213808372616768, 0.18238461017608643, -0.42740321159362793, 0.09215689450502396, 0.15121674537658691, 0.047031741589307785, 0.23718656599521637, -0.16102980077266693, 0.12145575135946274, 0.2881677746772766, -0.18652726709842682, -0.18414875864982605, -0.07622820883989334, 0.08048128336668015, 0.2570165991783142, -0.069468192756176, 0.02047928236424923, -0.21040309965610504, 0.03047284670174122, 0.08927395939826965, 0.1427948772907257, 0.23692886531352997, -0.024752512574195862, -0.2156609743833542, 0.0005510254995897412, 0.1827995628118515, 0.11411970853805542, -0.07855338603258133, -0.32897529006004333, 0.2802230417728424, 0.27836501598358154, -0.11751806735992432, -0.010236726142466068, 0.16770757734775543, -0.09938807040452957, 0.2528513967990875, 0.2071538120508194, 0.10758750885725021, 0.052611079066991806, 0.26961180567741394, 0.12649419903755188, 0.40504035353660583, -0.34832870960235596, 0.08942142128944397, 0.06441634148359299, 0.10164941847324371, 0.0377780981361866, 0.3812146782875061, -0.15324392914772034, 0.08992191404104233, 0.14838334918022156, 0.18833690881729126, 0.20253953337669373, 0.2628898620605469, 0.21354152262210846, -0.1878644973039627, -0.03412109240889549, -0.24108682572841644, 0.14363647997379303, -0.14272600412368774, 0.4156598746776581, 0.15059241652488708, 0.19285538792610168, -0.11485260725021362, -0.4619389772415161, -0.24626794457435608, 0.23536370694637299, -0.3245418071746826, -0.30920886993408203, -0.04867832362651825, -0.08018530905246735, 0.05592920631170273, 0.1945614516735077, -0.07929019629955292, 0.20385901629924774, 0.34283849596977234, -0.008660116232931614, -0.013750203885138035, -0.08435358107089996, -0.17061741650104523, 0.11397870630025864, 0.27486518025398254, -0.08173874765634537, 0.44693678617477417, 0.01080083753913641, -0.05156949535012245, -0.2344711422920227, 0.41141751408576965, 0.28612861037254333, 0.1949474811553955, -0.168862447142601, 0.06395307183265686, 0.20064708590507507, 0.0306134894490242, -0.3546890914440155, 0.3286518454551697, 0.1511448621749878, -0.2930276095867157, 0.36800575256347656, 0.12379536032676697, -0.23215578496456146, 0.2764037549495697, -0.10419601202011108, -0.20146721601486206, -0.04988892748951912, 0.3224368989467621, 0.0997074767947197, -0.0399320125579834, -0.19496309757232666, 0.11112021654844284, -0.14457158744335175, -0.28533950448036194, 0.15934592485427856, -0.02389831282198429, 0.46766093373298645, -0.275675505399704, 0.11720515042543411, -0.06182502955198288, 0.6220859885215759, -0.02922753430902958, 0.12432888895273209, -0.350137859582901, 0.10264351218938828, -0.33316680788993835, 0.1984093189239502, -0.08381783962249756, 0.15374279022216797, -0.3975840210914612, 0.20426496863365173, -0.11549855023622513, -0.023466093465685844, -0.11804244667291641, -0.00788638461381197, 0.14740723371505737, 0.17893213033676147, -0.3869176506996155, 0.1554814726114273, -0.0862593725323677, -0.00626726308837533, 0.2917737364768982, -0.37667545676231384, 0.22199058532714844, 0.04247966781258583, 0.20691059529781342, 0.010986385866999626, -0.02172369882464409, -0.054806627333164215, 0.33657434582710266, 0.6232408285140991, 0.34702974557876587, 0.14255116879940033, -0.10102774947881699, 0.08155330270528793, -0.34743764996528625, 0.014104106463491917, -0.06088816747069359, -0.20964983105659485, 0.09756835550069809, 0.5258304476737976, -0.22562745213508606, 0.13194629549980164, -0.17804564535617828, 0.2366943657398224, 0.38227659463882446, -0.15272878110408783, 0.06179971620440483, 0.048225898295640945, -0.25984057784080505, -0.005199328530579805, -0.12586632370948792, 0.45379582047462463, 0.007182890549302101, 0.32383018732070923, -0.3592907786369324, -0.5417624711990356, 0.4329099953174591, -0.4511263966560364, -0.3025250732898712, 0.12327846884727478, 0.23878563940525055, 0.34956884384155273, -0.06000060588121414, -0.5436676144599915, -0.1724908947944641, 0.26911357045173645, -0.030661094933748245, -0.15704523026943207, 0.20768482983112335, -0.1486440747976303, -0.04107564687728882, -0.11675458401441574, 0.17646899819374084, 0.10152711719274521, -0.29141318798065186, 0.2436991184949875, -0.37010371685028076 ]
https://github.com/huggingface/datasets/issues/160
caching in map causes same result to be returned for train, validation and test
> Ok, I think @lhoestq has already found a solution :-) Maybe you can chime in @lhoestq Oh, awesome! I see the PR, Ill check it out
hello, I am working on a program that uses the `nlp` library with the `SST2` dataset. The rough outline of the program is: ``` import nlp as nlp_datasets ... parser.add_argument('--dataset', help='HuggingFace Datasets id', default=['glue', 'sst2'], nargs='+') ... dataset = nlp_datasets.load_dataset(*args.dataset) ... # Create feature vocabs vocabs = create_vocabs(dataset.values(), vectorizers) ... # Create a function to vectorize based on vectorizers and vocabs: print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) # factory method to create a `convert_to_features` function based on vocabs convert_to_features = create_featurizer(vectorizers, vocabs) train_set = train_set.map(convert_to_features, batched=True) train_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) train_loader = torch.utils.data.DataLoader(train_set, batch_size=args.batchsz) valid_set = valid_set.map(convert_to_features, batched=True) valid_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) valid_loader = torch.utils.data.DataLoader(valid_set, batch_size=args.batchsz) test_set = test_set.map(convert_to_features, batched=True) test_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) test_loader = torch.utils.data.DataLoader(test_set, batch_size=args.batchsz) print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) ``` Im not sure if Im using it incorrectly, but the results are not what I expect. Namely, the `.map()` seems to grab the datset from the cache and then loses track of what the specific dataset is, instead using my training data for all datasets: ``` TS 67349 VS 872 ES 1821 TS 67349 VS 67349 ES 67349 ``` The behavior changes if I turn off the caching but then the results fail: ``` train_set = train_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... valid_set = valid_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... test_set = test_set.map(convert_to_features, batched=True, load_from_cache_file=False) ``` Now I get the right set of features back... ``` TS 67349 VS 872 ES 1821 100%|██████████| 68/68 [00:00<00:00, 92.78it/s] 100%|██████████| 1/1 [00:00<00:00, 75.47it/s] 0%| | 0/2 [00:00<?, ?it/s]TS 67349 VS 872 ES 1821 100%|██████████| 2/2 [00:00<00:00, 77.19it/s] ``` but I think its losing track of the original training set: ``` Traceback (most recent call last): File "/home/dpressel/dev/work/baseline/api-examples/layers-classify-hf-datasets.py", line 148, in <module> for x in train_loader: File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 345, in __next__ data = self._next_data() File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 385, in _next_data data = self._dataset_fetcher.fetch(index) # may raise StopIteration File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp> data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 338, in __getitem__ output_all_columns=self._output_all_columns, File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 294, in _getitem outputs = self._unnest(self._data.slice(key, 1).to_pydict()) File "pyarrow/table.pxi", line 1211, in pyarrow.lib.Table.slice File "pyarrow/public-api.pxi", line 390, in pyarrow.lib.pyarrow_wrap_table File "pyarrow/error.pxi", line 85, in pyarrow.lib.check_status pyarrow.lib.ArrowInvalid: Column 3: In chunk 0: Invalid: Length spanned by list offsets (15859698) larger than values array (length 100000) Process finished with exit code 1 ``` The full-example program (minus the print stmts) is here: https://github.com/dpressel/mead-baseline/pull/620/files
27
caching in map causes same result to be returned for train, validation and test hello, I am working on a program that uses the `nlp` library with the `SST2` dataset. The rough outline of the program is: ``` import nlp as nlp_datasets ... parser.add_argument('--dataset', help='HuggingFace Datasets id', default=['glue', 'sst2'], nargs='+') ... dataset = nlp_datasets.load_dataset(*args.dataset) ... # Create feature vocabs vocabs = create_vocabs(dataset.values(), vectorizers) ... # Create a function to vectorize based on vectorizers and vocabs: print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) # factory method to create a `convert_to_features` function based on vocabs convert_to_features = create_featurizer(vectorizers, vocabs) train_set = train_set.map(convert_to_features, batched=True) train_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) train_loader = torch.utils.data.DataLoader(train_set, batch_size=args.batchsz) valid_set = valid_set.map(convert_to_features, batched=True) valid_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) valid_loader = torch.utils.data.DataLoader(valid_set, batch_size=args.batchsz) test_set = test_set.map(convert_to_features, batched=True) test_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) test_loader = torch.utils.data.DataLoader(test_set, batch_size=args.batchsz) print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) ``` Im not sure if Im using it incorrectly, but the results are not what I expect. Namely, the `.map()` seems to grab the datset from the cache and then loses track of what the specific dataset is, instead using my training data for all datasets: ``` TS 67349 VS 872 ES 1821 TS 67349 VS 67349 ES 67349 ``` The behavior changes if I turn off the caching but then the results fail: ``` train_set = train_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... valid_set = valid_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... test_set = test_set.map(convert_to_features, batched=True, load_from_cache_file=False) ``` Now I get the right set of features back... ``` TS 67349 VS 872 ES 1821 100%|██████████| 68/68 [00:00<00:00, 92.78it/s] 100%|██████████| 1/1 [00:00<00:00, 75.47it/s] 0%| | 0/2 [00:00<?, ?it/s]TS 67349 VS 872 ES 1821 100%|██████████| 2/2 [00:00<00:00, 77.19it/s] ``` but I think its losing track of the original training set: ``` Traceback (most recent call last): File "/home/dpressel/dev/work/baseline/api-examples/layers-classify-hf-datasets.py", line 148, in <module> for x in train_loader: File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 345, in __next__ data = self._next_data() File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 385, in _next_data data = self._dataset_fetcher.fetch(index) # may raise StopIteration File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp> data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 338, in __getitem__ output_all_columns=self._output_all_columns, File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 294, in _getitem outputs = self._unnest(self._data.slice(key, 1).to_pydict()) File "pyarrow/table.pxi", line 1211, in pyarrow.lib.Table.slice File "pyarrow/public-api.pxi", line 390, in pyarrow.lib.pyarrow_wrap_table File "pyarrow/error.pxi", line 85, in pyarrow.lib.check_status pyarrow.lib.ArrowInvalid: Column 3: In chunk 0: Invalid: Length spanned by list offsets (15859698) larger than values array (length 100000) Process finished with exit code 1 ``` The full-example program (minus the print stmts) is here: https://github.com/dpressel/mead-baseline/pull/620/files > Ok, I think @lhoestq has already found a solution :-) Maybe you can chime in @lhoestq Oh, awesome! I see the PR, Ill check it out
[ -0.12632250785827637, -0.22070901095867157, -0.10063198953866959, 0.27801084518432617, 0.24547572433948517, -0.20522341132164001, 0.137210875749588, 0.5128593444824219, 0.18828663229942322, -0.06233531981706619, 0.12781980633735657, 0.38913458585739136, 0.003776297438889742, -0.1606176495552063, -0.010743561200797558, 0.024212030693888664, 0.09608013927936554, 0.23505231738090515, -0.14556236565113068, -0.1648235321044922, 0.08189409971237183, 0.2129841297864914, -0.042409978806972504, 0.11652691662311554, -0.39034950733184814, 0.041289906948804855, 0.21235668659210205, -0.2308902144432068, 0.06845776736736298, -0.25706538558006287, 0.1678507924079895, -0.09324328601360321, 0.02583368308842182, 0.16631756722927094, -0.00010973814642056823, 0.0067980014719069, -0.01538778841495514, -0.13856948912143707, -0.007274831645190716, 0.04019518941640854, 0.11813383549451828, -0.21172897517681122, -0.10918274521827698, -0.3121316730976105, -0.3320115804672241, -0.03766085207462311, 0.01780932955443859, -0.35565948486328125, 0.31285759806632996, 0.14745205640792847, 0.26071032881736755, 0.22490893304347992, -0.2538125216960907, 0.21113626658916473, 0.14512759447097778, -0.029327336698770523, -0.17419742047786713, -0.04230547323822975, -0.04397892951965332, -0.33446553349494934, -0.37676504254341125, 0.5254571437835693, -0.07658161967992783, 0.17014139890670776, 0.21466633677482605, 0.22505579888820648, 0.17260590195655823, 0.035937704145908356, 0.1429363638162613, -0.11396253854036331, 0.011094827204942703, -0.22691738605499268, -0.027116434648633003, -0.3733721077442169, -0.4186401665210724, -0.07852599024772644, 0.13762502372264862, 0.18104472756385803, -0.1363932490348816, -0.06371653825044632, -0.43193766474723816, 0.2392898052930832, 0.0652494877576828, -0.012788408435881138, 0.11952681839466095, 0.29516080021858215, 0.04380622133612633, 0.19522294402122498, 0.12944649159908295, -0.06973281502723694, -0.022558962926268578, -0.21451696753501892, -0.05897140875458717, 0.22640138864517212, -0.4359849691390991, 0.07208239287137985, 0.4315200448036194, -0.1404140740633011, -0.053242750465869904, 0.03063775785267353, 0.3919311761856079, 0.07921352982521057, 0.11681784689426422, -0.008491764776408672, -0.10416068881750107, 0.5898676514625549, 0.04432177543640137, 0.17118391394615173, -0.014969921670854092, -0.19827847182750702, -0.3560168445110321, 0.004987859632819891, 0.2880178987979889, -0.28660091757774353, 0.42841145396232605, 0.40236055850982666, -0.04226037487387657, -0.12152973562479019, -0.030102189630270004, 0.09402408450841904, -0.33085379004478455, 0.05566990002989769, 0.058729998767375946, 0.31395798921585083, -0.22545045614242554, 0.14926567673683167, -0.17376750707626343, -0.10335483402013779, -0.3849693238735199, 0.12735125422477722, -0.33784618973731995, -0.033323898911476135, -0.4962189197540283, 0.0015165578806772828, 0.28538474440574646, -0.002859933767467737, 0.28535225987434387, -0.038445454090833664, -0.13722018897533417, -0.3019891679286957, 0.10317365825176239, -0.39851170778274536, 0.5702314376831055, -0.1072295531630516, -0.19701702892780304, 0.24102117121219635, 0.13037829101085663, 0.08526421338319778, -0.29768794775009155, -0.19837751984596252, -0.1698075532913208, -0.20906901359558105, 0.2598607540130615, 0.24224542081356049, -0.07942765206098557, 0.14139588177204132, 0.09248226881027222, 0.0776788741350174, 0.3432585299015045, -0.232854962348938, 0.10284040868282318, -0.20148923993110657, -0.28198012709617615, -0.2043590545654297, 0.16256873309612274, 0.2623099982738495, 0.06677649915218353, -0.13737238943576813, 0.4017971158027649, 0.2371826320886612, 0.14729078114032745, 0.4138067364692688, -0.1965339034795761, 0.30949410796165466, -0.43349170684814453, 0.3486190438270569, 0.506476879119873, -0.32673147320747375, -0.49634483456611633, 0.09944828599691391, -0.15631580352783203, 0.1002880334854126, 0.10244903713464737, 0.15137287974357605, 0.029087670147418976, 0.014419378712773323, 0.1285371035337448, 0.018702296540141106, -0.0005146907060407102, 0.208420529961586, -0.31692788004875183, -0.021933916956186295, 0.5500646233558655, -0.2524673640727997, 0.02847237139940262, -0.13123734295368195, -0.05504271760582924, 0.021933000534772873, 0.006768169347196817, -0.11219003051519394, 0.12185932695865631, 0.11460470408201218, -0.056878868490457535, -0.1581844687461853, -0.10643851011991501, -0.024692509323358536, -0.28826257586479187, 0.23830120265483856, -0.14981895685195923, -0.0999680832028389, 0.06172538921236992, -0.09704917669296265, -0.16200260818004608, -0.12927502393722534, -0.18091638386249542, -0.27306458353996277, 0.20130178332328796, 0.23052407801151276, 0.3202398419380188, -0.07290733605623245, 0.2252090722322464, 0.3680817782878876, 0.048363614827394485, -0.21607212722301483, -0.4300347566604614, 0.08568026125431061, -0.17773471772670746, -0.2188279777765274, -0.0748467668890953, 0.22306986153125763, 0.2090262770652771, -0.04792642965912819, -0.14242812991142273, 0.14827579259872437, -0.2220856100320816, 0.4189179241657257, -0.13702799379825592, 0.269452840089798, -0.06965683400630951, 0.029306238517165184, 0.016667358577251434, 0.25400519371032715, 0.05881987884640694, -0.08980134129524231, -0.07270598411560059, 0.47423407435417175, 0.18140216171741486, 0.17011605203151703, -0.048070792108774185, -0.21626794338226318, 0.051984209567308426, -0.17831753194332123, -0.12091574817895889, -0.11881643533706665, 0.13640747964382172, -0.08057200163602829, 0.1488880068063736, 0.15232868492603302, -0.10698743164539337, 0.3120083510875702, 0.7600833773612976, 0.19691871106624603, -0.011016919277608395, 0.015740947797894478, -0.057213034480810165, -0.21779173612594604, 0.3338564336299896, -0.05868269130587578, 0.31038790941238403, 0.10473166406154633, 0.12716487050056458, -0.04620952904224396, -0.21965020895004272, -0.15443570911884308, 0.12464798986911774, -0.21436883509159088, -0.1919763833284378, 0.308034211397171, 0.28208687901496887, -0.15836724638938904, -0.35378316044807434, 0.0940157026052475, -0.01276430580765009, 0.026191623881459236, -0.15800140798091888, 0.21723663806915283, -0.1665293276309967, -0.2780851721763611, -0.28641536831855774, 0.13615694642066956, -0.12220374494791031, -0.0921654924750328, -0.04126313701272011, 0.42542964220046997, -0.036654673516750336, 0.19515161216259003, -0.3149721920490265, 0.08951952308416367, 0.226604163646698, -0.2663276791572571, -0.07168538868427277, -0.1902879923582077, -0.39492008090019226, 0.04223187640309334, -0.09523041546344757, -0.0929960310459137, 0.3783297538757324, -0.04731621965765953, -0.06796029210090637, -0.2881457805633545, -0.1725403368473053, 0.10717813670635223, -0.09261170029640198, -0.15536807477474213, 0.16724860668182373, -0.03133843466639519, -0.10115880519151688, -0.2987455725669861, 0.29379722476005554, -0.3053131401538849, -0.29914355278015137, -0.05198654159903526, 0.09890630096197128, 0.03010939620435238, -0.28034764528274536, -0.42862412333488464, 0.27357834577560425, -0.31724390387535095, 0.07546782493591309, -0.14892923831939697, 0.18622653186321259, 0.33093470335006714, 0.008953339420258999, -0.014561006799340248, -0.06404032558202744, 0.16439402103424072, -0.46741002798080444, -0.1409541219472885, 0.32257840037345886, 0.01876099780201912, -0.21630428731441498, -0.32459139823913574, -0.27572473883628845, 0.347724050283432, 0.19238898158073425, -0.5803960561752319, -0.21442314982414246, -0.04890046268701553, 0.24803197383880615, -0.05125873163342476, -0.04649535194039345, 0.5745421648025513, 0.01972527615725994, -0.17946361005306244, -0.16727842390537262, -0.33184051513671875, 0.35973140597343445, 0.41926297545433044, 0.2494877278804779, -0.011159147135913372, 0.17781507968902588, 0.20092034339904785, 0.6707466244697571, 0.20191961526870728, -0.25905755162239075, 0.3317307233810425, 0.10723208636045456, 0.07499753683805466, -0.2034970372915268, -0.3130224943161011, 0.18576058745384216, -0.18160304427146912, 0.019732147455215454, 0.2719840407371521, 0.006251134444028139, -0.2801845371723175, 0.01774226687848568, 0.21881961822509766, -0.2253446877002716, -0.3280579447746277, 0.2597459852695465, -0.34242793917655945, 0.24094845354557037, 0.03637058660387993, 0.31580042839050293, -0.5820142030715942, -0.24272412061691284, 0.19462814927101135, -0.33594635128974915, 0.2798454165458679, 0.04449770972132683, -0.5836008787155151, 0.12147048860788345, -0.538021981716156, 0.18194131553173065, 0.3146284818649292, 0.2724076807498932, -0.09202561527490616, -0.2465258091688156, 0.04334738850593567, -0.14713139832019806, 0.5711411833763123, 0.040972813963890076, 0.4085797369480133, 0.1699264794588089, -0.1302468627691269, -0.1994732767343521, -0.2247127890586853, 0.017234763130545616, 0.2801370620727539, 0.32649919390678406, 0.40084776282310486, -0.3452445864677429, -0.36327362060546875, -0.26576584577560425, 0.007364223711192608, -0.04618128761649132, 0.007107758894562721, -0.12583346664905548, 0.18938525021076202, 0.01648680679500103, 0.17434526979923248, -0.01498047262430191, 0.1183028295636177, 0.016039855778217316, -0.13654641807079315, -0.03913364186882973, -0.21942690014839172, 0.1282409280538559, 0.2264377325773239, 0.4735944867134094, 0.15032504498958588, 0.03246142715215683, 0.033214494585990906, 0.21515744924545288, -0.5671088099479675, 0.26595762372016907, -0.08884964138269424, -0.004131742753088474, 0.07464420050382614, 0.14443816244602203, -0.03030155971646309, 0.3378978669643402, -0.04252500832080841, 0.14571364223957062, -0.12862949073314667, 0.008956843055784702, -0.5539686679840088, 0.4854544401168823, 0.3874591886997223, 0.13246646523475647, -0.04488928243517876, -0.24685749411582947, 0.17346294224262238, 0.07412006705999374, -0.2765309810638428, 0.1322597712278366, -0.4306926131248474, -0.43199479579925537, 0.1857973337173462, 0.20948611199855804, 0.9232466816902161, 0.332022100687027, 0.3719581067562103, 0.1007041484117508, 0.10358098894357681, 0.2642512321472168, -0.23411183059215546, 0.40925082564353943, -0.3693642020225525, -0.0463179275393486, 0.024538656696677208, -0.16673317551612854, -0.22738394141197205, -0.010426265187561512, -0.01099333819001913, 0.30259495973587036, 0.32405051589012146, 0.5999365448951721, -0.11211955547332764, 0.16669917106628418, 0.21569165587425232, -0.15811017155647278, -0.2492315024137497, 0.2010732889175415, -0.29230374097824097, 0.3152405023574829, -0.11240831762552261, -0.07410863041877747, -0.01650959439575672, -0.35134196281433105, 0.06714027374982834, 0.06094216927886009, -0.3365170657634735, 0.3048412501811981, 0.13346198201179504, -0.27266666293144226, -0.2969881594181061, 0.2550773024559021, 0.11198677867650986, 0.08252744376659393, -0.048600129783153534, -0.03518514335155487, 0.40064483880996704, 0.3504585027694702, -0.006891036406159401, -0.20158861577510834, 0.1864127367734909, 0.06387528032064438, -0.012894121930003166, 0.03024262934923172, -0.05640564113855362, -0.3015829622745514, -0.25251179933547974, 0.14820435643196106, 0.003372181672602892, -0.41296184062957764, -0.05531826987862587, 0.061445046216249466, -0.066788449883461, -0.09413714706897736, 0.15088026225566864, -0.026068205013871193, -0.10794040560722351, 0.3151542544364929, 0.1173383817076683, -0.24304209649562836, 0.05235802382230759, 0.47483208775520325, -0.040397144854068756, 0.03181783854961395, 0.41542795300483704, -0.15281881392002106, -0.10380645841360092, -0.2553911507129669, -0.28417274355888367, -0.10205967724323273, -0.44884082674980164, 0.11220790445804596, -0.2194061428308487, -0.4019114077091217, 0.0442325733602047, 0.280818372964859, 0.014706877060234547, 0.31849586963653564, -0.011618148535490036, -0.12495850026607513, -0.37697234749794006, -0.10379048436880112, -0.1065821498632431, 0.2671777307987213, 0.2795186936855316, 0.2843484878540039, -0.29586702585220337, 0.2492261826992035, -0.38310956954956055, -0.1071421355009079, -0.18716979026794434, 0.4379133880138397, -0.03913804888725281, -0.3655827045440674, -0.009147044271230698, 0.00185489805880934, 0.12997117638587952, 0.17754493653774261, -0.6386589407920837, -0.2926190495491028, -0.09099180996417999, 0.08186071366071701, -0.007587700616568327, -0.159822016954422, -0.23928211629390717, -0.017213808372616768, 0.18238461017608643, -0.42740321159362793, 0.09215689450502396, 0.15121674537658691, 0.047031741589307785, 0.23718656599521637, -0.16102980077266693, 0.12145575135946274, 0.2881677746772766, -0.18652726709842682, -0.18414875864982605, -0.07622820883989334, 0.08048128336668015, 0.2570165991783142, -0.069468192756176, 0.02047928236424923, -0.21040309965610504, 0.03047284670174122, 0.08927395939826965, 0.1427948772907257, 0.23692886531352997, -0.024752512574195862, -0.2156609743833542, 0.0005510254995897412, 0.1827995628118515, 0.11411970853805542, -0.07855338603258133, -0.32897529006004333, 0.2802230417728424, 0.27836501598358154, -0.11751806735992432, -0.010236726142466068, 0.16770757734775543, -0.09938807040452957, 0.2528513967990875, 0.2071538120508194, 0.10758750885725021, 0.052611079066991806, 0.26961180567741394, 0.12649419903755188, 0.40504035353660583, -0.34832870960235596, 0.08942142128944397, 0.06441634148359299, 0.10164941847324371, 0.0377780981361866, 0.3812146782875061, -0.15324392914772034, 0.08992191404104233, 0.14838334918022156, 0.18833690881729126, 0.20253953337669373, 0.2628898620605469, 0.21354152262210846, -0.1878644973039627, -0.03412109240889549, -0.24108682572841644, 0.14363647997379303, -0.14272600412368774, 0.4156598746776581, 0.15059241652488708, 0.19285538792610168, -0.11485260725021362, -0.4619389772415161, -0.24626794457435608, 0.23536370694637299, -0.3245418071746826, -0.30920886993408203, -0.04867832362651825, -0.08018530905246735, 0.05592920631170273, 0.1945614516735077, -0.07929019629955292, 0.20385901629924774, 0.34283849596977234, -0.008660116232931614, -0.013750203885138035, -0.08435358107089996, -0.17061741650104523, 0.11397870630025864, 0.27486518025398254, -0.08173874765634537, 0.44693678617477417, 0.01080083753913641, -0.05156949535012245, -0.2344711422920227, 0.41141751408576965, 0.28612861037254333, 0.1949474811553955, -0.168862447142601, 0.06395307183265686, 0.20064708590507507, 0.0306134894490242, -0.3546890914440155, 0.3286518454551697, 0.1511448621749878, -0.2930276095867157, 0.36800575256347656, 0.12379536032676697, -0.23215578496456146, 0.2764037549495697, -0.10419601202011108, -0.20146721601486206, -0.04988892748951912, 0.3224368989467621, 0.0997074767947197, -0.0399320125579834, -0.19496309757232666, 0.11112021654844284, -0.14457158744335175, -0.28533950448036194, 0.15934592485427856, -0.02389831282198429, 0.46766093373298645, -0.275675505399704, 0.11720515042543411, -0.06182502955198288, 0.6220859885215759, -0.02922753430902958, 0.12432888895273209, -0.350137859582901, 0.10264351218938828, -0.33316680788993835, 0.1984093189239502, -0.08381783962249756, 0.15374279022216797, -0.3975840210914612, 0.20426496863365173, -0.11549855023622513, -0.023466093465685844, -0.11804244667291641, -0.00788638461381197, 0.14740723371505737, 0.17893213033676147, -0.3869176506996155, 0.1554814726114273, -0.0862593725323677, -0.00626726308837533, 0.2917737364768982, -0.37667545676231384, 0.22199058532714844, 0.04247966781258583, 0.20691059529781342, 0.010986385866999626, -0.02172369882464409, -0.054806627333164215, 0.33657434582710266, 0.6232408285140991, 0.34702974557876587, 0.14255116879940033, -0.10102774947881699, 0.08155330270528793, -0.34743764996528625, 0.014104106463491917, -0.06088816747069359, -0.20964983105659485, 0.09756835550069809, 0.5258304476737976, -0.22562745213508606, 0.13194629549980164, -0.17804564535617828, 0.2366943657398224, 0.38227659463882446, -0.15272878110408783, 0.06179971620440483, 0.048225898295640945, -0.25984057784080505, -0.005199328530579805, -0.12586632370948792, 0.45379582047462463, 0.007182890549302101, 0.32383018732070923, -0.3592907786369324, -0.5417624711990356, 0.4329099953174591, -0.4511263966560364, -0.3025250732898712, 0.12327846884727478, 0.23878563940525055, 0.34956884384155273, -0.06000060588121414, -0.5436676144599915, -0.1724908947944641, 0.26911357045173645, -0.030661094933748245, -0.15704523026943207, 0.20768482983112335, -0.1486440747976303, -0.04107564687728882, -0.11675458401441574, 0.17646899819374084, 0.10152711719274521, -0.29141318798065186, 0.2436991184949875, -0.37010371685028076 ]
https://github.com/huggingface/datasets/issues/160
caching in map causes same result to be returned for train, validation and test
The PR should prevent the cache from losing track of the of the dataset type (based on the location of its data). Not sure about your second problem though (cache off).
hello, I am working on a program that uses the `nlp` library with the `SST2` dataset. The rough outline of the program is: ``` import nlp as nlp_datasets ... parser.add_argument('--dataset', help='HuggingFace Datasets id', default=['glue', 'sst2'], nargs='+') ... dataset = nlp_datasets.load_dataset(*args.dataset) ... # Create feature vocabs vocabs = create_vocabs(dataset.values(), vectorizers) ... # Create a function to vectorize based on vectorizers and vocabs: print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) # factory method to create a `convert_to_features` function based on vocabs convert_to_features = create_featurizer(vectorizers, vocabs) train_set = train_set.map(convert_to_features, batched=True) train_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) train_loader = torch.utils.data.DataLoader(train_set, batch_size=args.batchsz) valid_set = valid_set.map(convert_to_features, batched=True) valid_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) valid_loader = torch.utils.data.DataLoader(valid_set, batch_size=args.batchsz) test_set = test_set.map(convert_to_features, batched=True) test_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) test_loader = torch.utils.data.DataLoader(test_set, batch_size=args.batchsz) print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) ``` Im not sure if Im using it incorrectly, but the results are not what I expect. Namely, the `.map()` seems to grab the datset from the cache and then loses track of what the specific dataset is, instead using my training data for all datasets: ``` TS 67349 VS 872 ES 1821 TS 67349 VS 67349 ES 67349 ``` The behavior changes if I turn off the caching but then the results fail: ``` train_set = train_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... valid_set = valid_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... test_set = test_set.map(convert_to_features, batched=True, load_from_cache_file=False) ``` Now I get the right set of features back... ``` TS 67349 VS 872 ES 1821 100%|██████████| 68/68 [00:00<00:00, 92.78it/s] 100%|██████████| 1/1 [00:00<00:00, 75.47it/s] 0%| | 0/2 [00:00<?, ?it/s]TS 67349 VS 872 ES 1821 100%|██████████| 2/2 [00:00<00:00, 77.19it/s] ``` but I think its losing track of the original training set: ``` Traceback (most recent call last): File "/home/dpressel/dev/work/baseline/api-examples/layers-classify-hf-datasets.py", line 148, in <module> for x in train_loader: File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 345, in __next__ data = self._next_data() File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 385, in _next_data data = self._dataset_fetcher.fetch(index) # may raise StopIteration File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp> data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 338, in __getitem__ output_all_columns=self._output_all_columns, File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 294, in _getitem outputs = self._unnest(self._data.slice(key, 1).to_pydict()) File "pyarrow/table.pxi", line 1211, in pyarrow.lib.Table.slice File "pyarrow/public-api.pxi", line 390, in pyarrow.lib.pyarrow_wrap_table File "pyarrow/error.pxi", line 85, in pyarrow.lib.check_status pyarrow.lib.ArrowInvalid: Column 3: In chunk 0: Invalid: Length spanned by list offsets (15859698) larger than values array (length 100000) Process finished with exit code 1 ``` The full-example program (minus the print stmts) is here: https://github.com/dpressel/mead-baseline/pull/620/files
31
caching in map causes same result to be returned for train, validation and test hello, I am working on a program that uses the `nlp` library with the `SST2` dataset. The rough outline of the program is: ``` import nlp as nlp_datasets ... parser.add_argument('--dataset', help='HuggingFace Datasets id', default=['glue', 'sst2'], nargs='+') ... dataset = nlp_datasets.load_dataset(*args.dataset) ... # Create feature vocabs vocabs = create_vocabs(dataset.values(), vectorizers) ... # Create a function to vectorize based on vectorizers and vocabs: print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) # factory method to create a `convert_to_features` function based on vocabs convert_to_features = create_featurizer(vectorizers, vocabs) train_set = train_set.map(convert_to_features, batched=True) train_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) train_loader = torch.utils.data.DataLoader(train_set, batch_size=args.batchsz) valid_set = valid_set.map(convert_to_features, batched=True) valid_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) valid_loader = torch.utils.data.DataLoader(valid_set, batch_size=args.batchsz) test_set = test_set.map(convert_to_features, batched=True) test_set.set_format(type='torch', columns=list(vectorizers.keys()) + ['y', 'lengths']) test_loader = torch.utils.data.DataLoader(test_set, batch_size=args.batchsz) print('TS', train_set.num_rows) print('VS', valid_set.num_rows) print('ES', test_set.num_rows) ``` Im not sure if Im using it incorrectly, but the results are not what I expect. Namely, the `.map()` seems to grab the datset from the cache and then loses track of what the specific dataset is, instead using my training data for all datasets: ``` TS 67349 VS 872 ES 1821 TS 67349 VS 67349 ES 67349 ``` The behavior changes if I turn off the caching but then the results fail: ``` train_set = train_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... valid_set = valid_set.map(convert_to_features, batched=True, load_from_cache_file=False) ... test_set = test_set.map(convert_to_features, batched=True, load_from_cache_file=False) ``` Now I get the right set of features back... ``` TS 67349 VS 872 ES 1821 100%|██████████| 68/68 [00:00<00:00, 92.78it/s] 100%|██████████| 1/1 [00:00<00:00, 75.47it/s] 0%| | 0/2 [00:00<?, ?it/s]TS 67349 VS 872 ES 1821 100%|██████████| 2/2 [00:00<00:00, 77.19it/s] ``` but I think its losing track of the original training set: ``` Traceback (most recent call last): File "/home/dpressel/dev/work/baseline/api-examples/layers-classify-hf-datasets.py", line 148, in <module> for x in train_loader: File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 345, in __next__ data = self._next_data() File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 385, in _next_data data = self._dataset_fetcher.fetch(index) # may raise StopIteration File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp> data = [self.dataset[idx] for idx in possibly_batched_index] File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 338, in __getitem__ output_all_columns=self._output_all_columns, File "/home/dpressel/anaconda3/lib/python3.7/site-packages/nlp/arrow_dataset.py", line 294, in _getitem outputs = self._unnest(self._data.slice(key, 1).to_pydict()) File "pyarrow/table.pxi", line 1211, in pyarrow.lib.Table.slice File "pyarrow/public-api.pxi", line 390, in pyarrow.lib.pyarrow_wrap_table File "pyarrow/error.pxi", line 85, in pyarrow.lib.check_status pyarrow.lib.ArrowInvalid: Column 3: In chunk 0: Invalid: Length spanned by list offsets (15859698) larger than values array (length 100000) Process finished with exit code 1 ``` The full-example program (minus the print stmts) is here: https://github.com/dpressel/mead-baseline/pull/620/files The PR should prevent the cache from losing track of the of the dataset type (based on the location of its data). Not sure about your second problem though (cache off).
[ -0.12632250785827637, -0.22070901095867157, -0.10063198953866959, 0.27801084518432617, 0.24547572433948517, -0.20522341132164001, 0.137210875749588, 0.5128593444824219, 0.18828663229942322, -0.06233531981706619, 0.12781980633735657, 0.38913458585739136, 0.003776297438889742, -0.1606176495552063, -0.010743561200797558, 0.024212030693888664, 0.09608013927936554, 0.23505231738090515, -0.14556236565113068, -0.1648235321044922, 0.08189409971237183, 0.2129841297864914, -0.042409978806972504, 0.11652691662311554, -0.39034950733184814, 0.041289906948804855, 0.21235668659210205, -0.2308902144432068, 0.06845776736736298, -0.25706538558006287, 0.1678507924079895, -0.09324328601360321, 0.02583368308842182, 0.16631756722927094, -0.00010973814642056823, 0.0067980014719069, -0.01538778841495514, -0.13856948912143707, -0.007274831645190716, 0.04019518941640854, 0.11813383549451828, -0.21172897517681122, -0.10918274521827698, -0.3121316730976105, -0.3320115804672241, -0.03766085207462311, 0.01780932955443859, -0.35565948486328125, 0.31285759806632996, 0.14745205640792847, 0.26071032881736755, 0.22490893304347992, -0.2538125216960907, 0.21113626658916473, 0.14512759447097778, -0.029327336698770523, -0.17419742047786713, -0.04230547323822975, -0.04397892951965332, -0.33446553349494934, -0.37676504254341125, 0.5254571437835693, -0.07658161967992783, 0.17014139890670776, 0.21466633677482605, 0.22505579888820648, 0.17260590195655823, 0.035937704145908356, 0.1429363638162613, -0.11396253854036331, 0.011094827204942703, -0.22691738605499268, -0.027116434648633003, -0.3733721077442169, -0.4186401665210724, -0.07852599024772644, 0.13762502372264862, 0.18104472756385803, -0.1363932490348816, -0.06371653825044632, -0.43193766474723816, 0.2392898052930832, 0.0652494877576828, -0.012788408435881138, 0.11952681839466095, 0.29516080021858215, 0.04380622133612633, 0.19522294402122498, 0.12944649159908295, -0.06973281502723694, -0.022558962926268578, -0.21451696753501892, -0.05897140875458717, 0.22640138864517212, -0.4359849691390991, 0.07208239287137985, 0.4315200448036194, -0.1404140740633011, -0.053242750465869904, 0.03063775785267353, 0.3919311761856079, 0.07921352982521057, 0.11681784689426422, -0.008491764776408672, -0.10416068881750107, 0.5898676514625549, 0.04432177543640137, 0.17118391394615173, -0.014969921670854092, -0.19827847182750702, -0.3560168445110321, 0.004987859632819891, 0.2880178987979889, -0.28660091757774353, 0.42841145396232605, 0.40236055850982666, -0.04226037487387657, -0.12152973562479019, -0.030102189630270004, 0.09402408450841904, -0.33085379004478455, 0.05566990002989769, 0.058729998767375946, 0.31395798921585083, -0.22545045614242554, 0.14926567673683167, -0.17376750707626343, -0.10335483402013779, -0.3849693238735199, 0.12735125422477722, -0.33784618973731995, -0.033323898911476135, -0.4962189197540283, 0.0015165578806772828, 0.28538474440574646, -0.002859933767467737, 0.28535225987434387, -0.038445454090833664, -0.13722018897533417, -0.3019891679286957, 0.10317365825176239, -0.39851170778274536, 0.5702314376831055, -0.1072295531630516, -0.19701702892780304, 0.24102117121219635, 0.13037829101085663, 0.08526421338319778, -0.29768794775009155, -0.19837751984596252, -0.1698075532913208, -0.20906901359558105, 0.2598607540130615, 0.24224542081356049, -0.07942765206098557, 0.14139588177204132, 0.09248226881027222, 0.0776788741350174, 0.3432585299015045, -0.232854962348938, 0.10284040868282318, -0.20148923993110657, -0.28198012709617615, -0.2043590545654297, 0.16256873309612274, 0.2623099982738495, 0.06677649915218353, -0.13737238943576813, 0.4017971158027649, 0.2371826320886612, 0.14729078114032745, 0.4138067364692688, -0.1965339034795761, 0.30949410796165466, -0.43349170684814453, 0.3486190438270569, 0.506476879119873, -0.32673147320747375, -0.49634483456611633, 0.09944828599691391, -0.15631580352783203, 0.1002880334854126, 0.10244903713464737, 0.15137287974357605, 0.029087670147418976, 0.014419378712773323, 0.1285371035337448, 0.018702296540141106, -0.0005146907060407102, 0.208420529961586, -0.31692788004875183, -0.021933916956186295, 0.5500646233558655, -0.2524673640727997, 0.02847237139940262, -0.13123734295368195, -0.05504271760582924, 0.021933000534772873, 0.006768169347196817, -0.11219003051519394, 0.12185932695865631, 0.11460470408201218, -0.056878868490457535, -0.1581844687461853, -0.10643851011991501, -0.024692509323358536, -0.28826257586479187, 0.23830120265483856, -0.14981895685195923, -0.0999680832028389, 0.06172538921236992, -0.09704917669296265, -0.16200260818004608, -0.12927502393722534, -0.18091638386249542, -0.27306458353996277, 0.20130178332328796, 0.23052407801151276, 0.3202398419380188, -0.07290733605623245, 0.2252090722322464, 0.3680817782878876, 0.048363614827394485, -0.21607212722301483, -0.4300347566604614, 0.08568026125431061, -0.17773471772670746, -0.2188279777765274, -0.0748467668890953, 0.22306986153125763, 0.2090262770652771, -0.04792642965912819, -0.14242812991142273, 0.14827579259872437, -0.2220856100320816, 0.4189179241657257, -0.13702799379825592, 0.269452840089798, -0.06965683400630951, 0.029306238517165184, 0.016667358577251434, 0.25400519371032715, 0.05881987884640694, -0.08980134129524231, -0.07270598411560059, 0.47423407435417175, 0.18140216171741486, 0.17011605203151703, -0.048070792108774185, -0.21626794338226318, 0.051984209567308426, -0.17831753194332123, -0.12091574817895889, -0.11881643533706665, 0.13640747964382172, -0.08057200163602829, 0.1488880068063736, 0.15232868492603302, -0.10698743164539337, 0.3120083510875702, 0.7600833773612976, 0.19691871106624603, -0.011016919277608395, 0.015740947797894478, -0.057213034480810165, -0.21779173612594604, 0.3338564336299896, -0.05868269130587578, 0.31038790941238403, 0.10473166406154633, 0.12716487050056458, -0.04620952904224396, -0.21965020895004272, -0.15443570911884308, 0.12464798986911774, -0.21436883509159088, -0.1919763833284378, 0.308034211397171, 0.28208687901496887, -0.15836724638938904, -0.35378316044807434, 0.0940157026052475, -0.01276430580765009, 0.026191623881459236, -0.15800140798091888, 0.21723663806915283, -0.1665293276309967, -0.2780851721763611, -0.28641536831855774, 0.13615694642066956, -0.12220374494791031, -0.0921654924750328, -0.04126313701272011, 0.42542964220046997, -0.036654673516750336, 0.19515161216259003, -0.3149721920490265, 0.08951952308416367, 0.226604163646698, -0.2663276791572571, -0.07168538868427277, -0.1902879923582077, -0.39492008090019226, 0.04223187640309334, -0.09523041546344757, -0.0929960310459137, 0.3783297538757324, -0.04731621965765953, -0.06796029210090637, -0.2881457805633545, -0.1725403368473053, 0.10717813670635223, -0.09261170029640198, -0.15536807477474213, 0.16724860668182373, -0.03133843466639519, -0.10115880519151688, -0.2987455725669861, 0.29379722476005554, -0.3053131401538849, -0.29914355278015137, -0.05198654159903526, 0.09890630096197128, 0.03010939620435238, -0.28034764528274536, -0.42862412333488464, 0.27357834577560425, -0.31724390387535095, 0.07546782493591309, -0.14892923831939697, 0.18622653186321259, 0.33093470335006714, 0.008953339420258999, -0.014561006799340248, -0.06404032558202744, 0.16439402103424072, -0.46741002798080444, -0.1409541219472885, 0.32257840037345886, 0.01876099780201912, -0.21630428731441498, -0.32459139823913574, -0.27572473883628845, 0.347724050283432, 0.19238898158073425, -0.5803960561752319, -0.21442314982414246, -0.04890046268701553, 0.24803197383880615, -0.05125873163342476, -0.04649535194039345, 0.5745421648025513, 0.01972527615725994, -0.17946361005306244, -0.16727842390537262, -0.33184051513671875, 0.35973140597343445, 0.41926297545433044, 0.2494877278804779, -0.011159147135913372, 0.17781507968902588, 0.20092034339904785, 0.6707466244697571, 0.20191961526870728, -0.25905755162239075, 0.3317307233810425, 0.10723208636045456, 0.07499753683805466, -0.2034970372915268, -0.3130224943161011, 0.18576058745384216, -0.18160304427146912, 0.019732147455215454, 0.2719840407371521, 0.006251134444028139, -0.2801845371723175, 0.01774226687848568, 0.21881961822509766, -0.2253446877002716, -0.3280579447746277, 0.2597459852695465, -0.34242793917655945, 0.24094845354557037, 0.03637058660387993, 0.31580042839050293, -0.5820142030715942, -0.24272412061691284, 0.19462814927101135, -0.33594635128974915, 0.2798454165458679, 0.04449770972132683, -0.5836008787155151, 0.12147048860788345, -0.538021981716156, 0.18194131553173065, 0.3146284818649292, 0.2724076807498932, -0.09202561527490616, -0.2465258091688156, 0.04334738850593567, -0.14713139832019806, 0.5711411833763123, 0.040972813963890076, 0.4085797369480133, 0.1699264794588089, -0.1302468627691269, -0.1994732767343521, -0.2247127890586853, 0.017234763130545616, 0.2801370620727539, 0.32649919390678406, 0.40084776282310486, -0.3452445864677429, -0.36327362060546875, -0.26576584577560425, 0.007364223711192608, -0.04618128761649132, 0.007107758894562721, -0.12583346664905548, 0.18938525021076202, 0.01648680679500103, 0.17434526979923248, -0.01498047262430191, 0.1183028295636177, 0.016039855778217316, -0.13654641807079315, -0.03913364186882973, -0.21942690014839172, 0.1282409280538559, 0.2264377325773239, 0.4735944867134094, 0.15032504498958588, 0.03246142715215683, 0.033214494585990906, 0.21515744924545288, -0.5671088099479675, 0.26595762372016907, -0.08884964138269424, -0.004131742753088474, 0.07464420050382614, 0.14443816244602203, -0.03030155971646309, 0.3378978669643402, -0.04252500832080841, 0.14571364223957062, -0.12862949073314667, 0.008956843055784702, -0.5539686679840088, 0.4854544401168823, 0.3874591886997223, 0.13246646523475647, -0.04488928243517876, -0.24685749411582947, 0.17346294224262238, 0.07412006705999374, -0.2765309810638428, 0.1322597712278366, -0.4306926131248474, -0.43199479579925537, 0.1857973337173462, 0.20948611199855804, 0.9232466816902161, 0.332022100687027, 0.3719581067562103, 0.1007041484117508, 0.10358098894357681, 0.2642512321472168, -0.23411183059215546, 0.40925082564353943, -0.3693642020225525, -0.0463179275393486, 0.024538656696677208, -0.16673317551612854, -0.22738394141197205, -0.010426265187561512, -0.01099333819001913, 0.30259495973587036, 0.32405051589012146, 0.5999365448951721, -0.11211955547332764, 0.16669917106628418, 0.21569165587425232, -0.15811017155647278, -0.2492315024137497, 0.2010732889175415, -0.29230374097824097, 0.3152405023574829, -0.11240831762552261, -0.07410863041877747, -0.01650959439575672, -0.35134196281433105, 0.06714027374982834, 0.06094216927886009, -0.3365170657634735, 0.3048412501811981, 0.13346198201179504, -0.27266666293144226, -0.2969881594181061, 0.2550773024559021, 0.11198677867650986, 0.08252744376659393, -0.048600129783153534, -0.03518514335155487, 0.40064483880996704, 0.3504585027694702, -0.006891036406159401, -0.20158861577510834, 0.1864127367734909, 0.06387528032064438, -0.012894121930003166, 0.03024262934923172, -0.05640564113855362, -0.3015829622745514, -0.25251179933547974, 0.14820435643196106, 0.003372181672602892, -0.41296184062957764, -0.05531826987862587, 0.061445046216249466, -0.066788449883461, -0.09413714706897736, 0.15088026225566864, -0.026068205013871193, -0.10794040560722351, 0.3151542544364929, 0.1173383817076683, -0.24304209649562836, 0.05235802382230759, 0.47483208775520325, -0.040397144854068756, 0.03181783854961395, 0.41542795300483704, -0.15281881392002106, -0.10380645841360092, -0.2553911507129669, -0.28417274355888367, -0.10205967724323273, -0.44884082674980164, 0.11220790445804596, -0.2194061428308487, -0.4019114077091217, 0.0442325733602047, 0.280818372964859, 0.014706877060234547, 0.31849586963653564, -0.011618148535490036, -0.12495850026607513, -0.37697234749794006, -0.10379048436880112, -0.1065821498632431, 0.2671777307987213, 0.2795186936855316, 0.2843484878540039, -0.29586702585220337, 0.2492261826992035, -0.38310956954956055, -0.1071421355009079, -0.18716979026794434, 0.4379133880138397, -0.03913804888725281, -0.3655827045440674, -0.009147044271230698, 0.00185489805880934, 0.12997117638587952, 0.17754493653774261, -0.6386589407920837, -0.2926190495491028, -0.09099180996417999, 0.08186071366071701, -0.007587700616568327, -0.159822016954422, -0.23928211629390717, -0.017213808372616768, 0.18238461017608643, -0.42740321159362793, 0.09215689450502396, 0.15121674537658691, 0.047031741589307785, 0.23718656599521637, -0.16102980077266693, 0.12145575135946274, 0.2881677746772766, -0.18652726709842682, -0.18414875864982605, -0.07622820883989334, 0.08048128336668015, 0.2570165991783142, -0.069468192756176, 0.02047928236424923, -0.21040309965610504, 0.03047284670174122, 0.08927395939826965, 0.1427948772907257, 0.23692886531352997, -0.024752512574195862, -0.2156609743833542, 0.0005510254995897412, 0.1827995628118515, 0.11411970853805542, -0.07855338603258133, -0.32897529006004333, 0.2802230417728424, 0.27836501598358154, -0.11751806735992432, -0.010236726142466068, 0.16770757734775543, -0.09938807040452957, 0.2528513967990875, 0.2071538120508194, 0.10758750885725021, 0.052611079066991806, 0.26961180567741394, 0.12649419903755188, 0.40504035353660583, -0.34832870960235596, 0.08942142128944397, 0.06441634148359299, 0.10164941847324371, 0.0377780981361866, 0.3812146782875061, -0.15324392914772034, 0.08992191404104233, 0.14838334918022156, 0.18833690881729126, 0.20253953337669373, 0.2628898620605469, 0.21354152262210846, -0.1878644973039627, -0.03412109240889549, -0.24108682572841644, 0.14363647997379303, -0.14272600412368774, 0.4156598746776581, 0.15059241652488708, 0.19285538792610168, -0.11485260725021362, -0.4619389772415161, -0.24626794457435608, 0.23536370694637299, -0.3245418071746826, -0.30920886993408203, -0.04867832362651825, -0.08018530905246735, 0.05592920631170273, 0.1945614516735077, -0.07929019629955292, 0.20385901629924774, 0.34283849596977234, -0.008660116232931614, -0.013750203885138035, -0.08435358107089996, -0.17061741650104523, 0.11397870630025864, 0.27486518025398254, -0.08173874765634537, 0.44693678617477417, 0.01080083753913641, -0.05156949535012245, -0.2344711422920227, 0.41141751408576965, 0.28612861037254333, 0.1949474811553955, -0.168862447142601, 0.06395307183265686, 0.20064708590507507, 0.0306134894490242, -0.3546890914440155, 0.3286518454551697, 0.1511448621749878, -0.2930276095867157, 0.36800575256347656, 0.12379536032676697, -0.23215578496456146, 0.2764037549495697, -0.10419601202011108, -0.20146721601486206, -0.04988892748951912, 0.3224368989467621, 0.0997074767947197, -0.0399320125579834, -0.19496309757232666, 0.11112021654844284, -0.14457158744335175, -0.28533950448036194, 0.15934592485427856, -0.02389831282198429, 0.46766093373298645, -0.275675505399704, 0.11720515042543411, -0.06182502955198288, 0.6220859885215759, -0.02922753430902958, 0.12432888895273209, -0.350137859582901, 0.10264351218938828, -0.33316680788993835, 0.1984093189239502, -0.08381783962249756, 0.15374279022216797, -0.3975840210914612, 0.20426496863365173, -0.11549855023622513, -0.023466093465685844, -0.11804244667291641, -0.00788638461381197, 0.14740723371505737, 0.17893213033676147, -0.3869176506996155, 0.1554814726114273, -0.0862593725323677, -0.00626726308837533, 0.2917737364768982, -0.37667545676231384, 0.22199058532714844, 0.04247966781258583, 0.20691059529781342, 0.010986385866999626, -0.02172369882464409, -0.054806627333164215, 0.33657434582710266, 0.6232408285140991, 0.34702974557876587, 0.14255116879940033, -0.10102774947881699, 0.08155330270528793, -0.34743764996528625, 0.014104106463491917, -0.06088816747069359, -0.20964983105659485, 0.09756835550069809, 0.5258304476737976, -0.22562745213508606, 0.13194629549980164, -0.17804564535617828, 0.2366943657398224, 0.38227659463882446, -0.15272878110408783, 0.06179971620440483, 0.048225898295640945, -0.25984057784080505, -0.005199328530579805, -0.12586632370948792, 0.45379582047462463, 0.007182890549302101, 0.32383018732070923, -0.3592907786369324, -0.5417624711990356, 0.4329099953174591, -0.4511263966560364, -0.3025250732898712, 0.12327846884727478, 0.23878563940525055, 0.34956884384155273, -0.06000060588121414, -0.5436676144599915, -0.1724908947944641, 0.26911357045173645, -0.030661094933748245, -0.15704523026943207, 0.20768482983112335, -0.1486440747976303, -0.04107564687728882, -0.11675458401441574, 0.17646899819374084, 0.10152711719274521, -0.29141318798065186, 0.2436991184949875, -0.37010371685028076 ]