html_url
stringlengths
48
51
title
stringlengths
5
268
comment
stringlengths
70
51.8k
body
stringlengths
0
29.8k
comment_length
int64
16
1.52k
text
stringlengths
164
54.1k
embeddings
listlengths
768
768
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 ?
[ -0.1408683806657791, -0.1892591118812561, -0.03353653848171234, 0.46936744451522827, 0.4054456651210785, 0.112000472843647, 0.05603649839758873, 0.3544987142086029, -0.10948891192674637, 0.2927757501602173, -0.1404101848602295, 0.16212107241153717, 0.05022774264216423, -0.02804594300687313, 0.05054078996181488, -0.03532538563013077, -0.1691269874572754, 0.1510201394557953, 0.13207300007343292, 0.005323641002178192, -0.1707707941532135, 0.045582517981529236, -0.2690636217594147, 0.2587416470050812, -0.2867065370082855, -0.08528336137533188, 0.09144627302885056, 0.3859780728816986, -0.3081700801849365, -0.5349303483963013, 0.045082226395606995, -0.1443348228931427, 0.2407742589712143, 0.19576658308506012, -0.00011105801240773872, 0.049332816153764725, 0.2629140019416809, -0.05734889954328537, -0.5286890864372253, -0.14541827142238617, -0.35688459873199463, -0.12314393371343613, -0.014699259772896767, -0.31323131918907166, -0.12878188490867615, -0.22096027433872223, 0.40224409103393555, -0.3689236044883728, 0.2718048095703125, 0.31527605652809143, 0.20390814542770386, 0.20004352927207947, 0.12182676792144775, -0.054713886231184006, 0.04933757707476616, 0.13519367575645447, -0.13144762814044952, -0.013970177620649338, 0.23928497731685638, 0.1932624876499176, 0.16366681456565857, 0.4578830897808075, -0.17119532823562622, -0.11164619773626328, 0.25520724058151245, 0.243842214345932, -0.030298270285129547, -0.16049779951572418, 0.3209943473339081, 0.2331337183713913, 0.4437566101551056, -0.4332888126373291, -0.4334990680217743, -0.11148460954427719, -0.003818502649664879, -0.26497238874435425, 0.24886417388916016, 0.10019537061452866, -0.3177081346511841, 0.1554599553346634, -0.11993873119354248, -0.36654940247535706, -0.3681374490261078, 0.1454017162322998, 0.26279857754707336, 0.38023069500923157, -0.05968918278813362, 0.0670863687992096, -0.06119953840970993, -0.19583746790885925, 0.05281441658735275, -0.0894942656159401, -0.16413280367851257, 0.30266934633255005, -0.2269848883152008, 0.1276467889547348, 0.13528814911842346, 0.05715295672416687, 0.11984618753194809, 0.3628025949001312, 0.20324356853961945, -0.18401360511779785, -0.375539630651474, 0.06257759779691696, -0.26458486914634705, 0.4228105843067169, 0.295701265335083, 0.14088541269302368, 0.3827153444290161, -0.008944286033511162, 0.2218949943780899, -0.3649007976055145, -0.29457464814186096, -0.16017687320709229, -0.11438271403312683, 0.25096845626831055, 0.36158040165901184, -0.15953703224658966, -0.24764703214168549, -0.059575796127319336, -0.12639565765857697, 0.019348111003637314, 0.20292779803276062, 0.5645594000816345, 0.0536450631916523, 0.1894831657409668, 0.03513301908969879, 0.32177814841270447, 0.05370361730456352, -0.5332470536231995, -0.17430877685546875, 0.21654145419597626, -0.30879589915275574, 0.14569520950317383, 0.3105255663394928, -0.331324964761734, 0.3001636564731598, -0.1157546415925026, -0.30328547954559326, 0.0032351629342883825, 0.12947304546833038, -0.17491817474365234, 0.10203277319669724, 0.07796083390712738, 0.09318635612726212, 0.10305067896842957, 0.19042691588401794, -0.07996855676174164, 0.065503790974617, -0.03327649086713791, -0.04719267040491104, -0.48522597551345825, 0.025506697595119476, 0.13068559765815735, 0.07165276259183884, -0.009505514986813068, -0.08786728978157043, -0.05430344119668007, 0.19623053073883057, -0.22172126173973083, -0.11173630505800247, 0.11289944499731064, -0.02556457184255123, -0.271345853805542, 0.21264243125915527, 0.5540772676467896, -0.3136104941368103, 0.20926669239997864, -0.19852595031261444, -0.13489241898059845, 0.16316388547420502, 0.3628433346748352, -0.11672967672348022, 0.07758896797895432, -0.38488221168518066, 0.32663190364837646, 0.6447770595550537, -0.5202512145042419, -0.507689356803894, 0.2816506624221802, -0.222357839345932, -0.0715678259730339, 0.5067176818847656, 0.24071189761161804, -0.0645313635468483, -0.1131831631064415, 0.047000400722026825, 0.11700915545225143, 0.046720925718545914, -0.08850014209747314, -0.1114891916513443, -0.24508875608444214, 0.34023773670196533, 0.11301998794078827, -0.03068048134446144, 0.13209125399589539, -0.029109438881278038, 0.16131925582885742, 0.3716830313205719, -0.291082501411438, 0.15118570625782013, 0.284060537815094, 0.3005661368370056, 0.03720932453870773, -0.14257796108722687, -0.11960297077894211, -0.294660747051239, 0.23998583853244781, -0.0920315757393837, -0.020116664469242096, -0.3088814616203308, 0.0002528683398850262, -0.11309006810188293, -0.040919654071331024, -0.38762661814689636, -0.2783125042915344, 0.1904965192079544, 0.45270562171936035, 0.22734427452087402, -0.019652726128697395, -0.28994178771972656, 0.3173922300338745, -0.48267507553100586, 0.0027163613121956587, -0.6354675889015198, 0.12806390225887299, -0.09318916499614716, -0.14903521537780762, -0.1639638990163803, 0.16392529010772705, 0.023550231009721756, -0.002848584670573473, -0.09274846315383911, 0.21247054636478424, -0.23807621002197266, 0.08131276816129684, 0.2638706564903259, 0.10236118733882904, 0.0382031612098217, -0.5709120035171509, 0.33758243918418884, 0.4717598557472229, 0.22842155396938324, 0.08941130340099335, -0.26937147974967957, 0.11830674111843109, -0.08254852145910263, 0.09030044078826904, 0.0437278151512146, 0.01720273867249489, 0.4666252136230469, 0.01481084618717432, 0.06998182088136673, -0.21279394626617432, 0.30032503604888916, 0.21649004518985748, 0.4621780514717102, -0.08612556010484695, -0.5364863276481628, 0.08613045513629913, 0.27960124611854553, 0.06051981449127197, -0.09864352643489838, -0.01093621738255024, -0.35100457072257996, -0.04163489118218422, 0.05506303161382675, 0.22897903621196747, 0.2813912332057953, 0.22780871391296387, -0.06535414606332779, 0.0831112340092659, 0.09837401658296585, -0.2621798515319824, 0.18072955310344696, -0.2579272389411926, 0.10123036801815033, 0.2698712944984436, 0.19980134069919586, 0.03205374628305435, -0.03564492240548134, -0.5033629536628723, 0.07217034697532654, 0.445490300655365, -0.270343542098999, -0.17282283306121826, -0.13690432906150818, -0.6854812502861023, 0.060650162398815155, 0.07638508826494217, -0.11054059118032455, -0.3432241976261139, -0.022228943184018135, 0.18776096403598785, 0.21456211805343628, -0.07199499011039734, -0.09505226463079453, 0.0755409300327301, 0.06864802539348602, -0.1094084084033966, -0.4193322956562042, -0.31495967507362366, -0.20777533948421478, 0.04685627296566963, 0.27079546451568604, -0.022806797176599503, 0.246942937374115, -0.28027981519699097, 0.06392844021320343, -0.4144379198551178, -0.025489790365099907, 0.02347724698483944, 0.029297424480319023, -0.00047369685489684343, -0.034068841487169266, 0.4818882346153259, 0.4775172173976898, -0.06232468783855438, 0.3475897014141083, 0.04154755920171738, -0.17423026263713837, 0.2578030824661255, -0.21523959934711456, -0.23448705673217773, -0.04180873930454254, -0.5249141454696655, -0.35637015104293823, -0.3282744884490967, 0.08275269716978073, -0.13988064229488373, 0.08039943873882294, 0.3180902898311615, -0.14647722244262695, 0.20626114308834076, -0.03888370841741562, 0.19586797058582306, -0.07545895129442215, -0.22561410069465637, 0.3382495641708374, -0.16139712929725647, -0.21599528193473816, 0.23609322309494019, 0.06490513682365417, 0.21008872985839844, -0.041243430227041245, -0.2306007444858551, 0.17212629318237305, -0.22556982934474945, 0.284463495016098, -0.1305042803287506, -0.15192101895809174, 0.34998250007629395, -0.010504031553864479, -0.0014595107641071081, 0.0043805427849292755, 0.011881567537784576, 0.11780835688114166, 0.010026367381215096, 0.5716209411621094, 0.2039428949356079, 0.41344475746154785, 0.14451955258846283, 0.37235599756240845, 0.44866472482681274, 0.026759229600429535, 0.32839083671569824, -0.26213914155960083, 0.06690725684165955, -0.06588201969861984, -0.3156571388244629, 0.11945395171642303, 0.17475612461566925, -0.17332394421100616, -0.10797251760959625, 0.1386820524930954, 0.38894352316856384, -0.33151960372924805, -0.13302302360534668, -0.28740736842155457, -0.08548351377248764, -0.1463778018951416, 0.020249146968126297, 0.011118698865175247, 0.1291111260652542, 0.20937487483024597, -0.34255450963974, -0.3424656093120575, -0.11213227361440659, 0.46337810158729553, 0.0023359160404652357, 0.15231479704380035, -0.27443617582321167, -0.28418031334877014, -0.5618228912353516, 0.4806252717971802, 0.11800625175237656, 0.21206682920455933, -0.2758861780166626, 0.12500107288360596, 0.26634660363197327, 0.02257581427693367, 0.3283255994319916, -0.30048316717147827, 0.2563644349575043, 0.08078678697347641, -0.09892932325601578, -0.23587952554225922, 0.1696384698152542, -0.1501615196466446, 0.1985006481409073, 0.3358711302280426, 0.19395430386066437, -0.16084091365337372, -0.1110728532075882, -0.016944043338298798, 0.10838139802217484, -0.1749695986509323, -0.21759426593780518, -0.017784247174859047, -0.003704600501805544, -0.1634150892496109, -0.32848235964775085, -0.17156817018985748, 0.24727573990821838, 0.13296093046665192, 0.33589136600494385, 0.0468716099858284, -0.0742722675204277, 0.17847178876399994, 0.22041527926921844, 0.33928734064102173, -0.09613542258739471, -0.15330930054187775, 0.16526304185390472, 0.24983622133731842, 0.07175802439451218, 0.49867457151412964, 0.28984543681144714, -0.3622645139694214, -0.01410093903541565, -0.22840812802314758, 0.1272355318069458, 0.290335476398468, -0.2587491273880005, 0.10405406355857849, 0.03436676040291786, -0.00782658252865076, 0.061695646494627, -0.09590508788824081, 0.2513183653354645, 0.031543806195259094, -0.12594184279441833, -0.3986913859844208, 0.32729798555374146, 0.10116897523403168, 0.003973283804953098, 0.45400184392929077, -0.14631949365139008, -0.22761744260787964, 0.023735547438263893, 0.05723195523023605, 0.9369786381721497, 0.12992826104164124, 0.22923734784126282, 0.39308348298072815, 0.1683122217655182, 0.5259997844696045, -0.3035133183002472, 0.36880117654800415, -0.387155145406723, -0.08039590716362, -0.0029478250071406364, -0.11370435357093811, -0.011663719080388546, -0.1774902045726776, -0.08031540364027023, 0.23407749831676483, -0.15633413195610046, 0.006111164577305317, -0.09859154373407364, 0.26289814710617065, -0.33494704961776733, -0.09046962857246399, -0.46061480045318604, 0.0847524031996727, -0.20800425112247467, 0.3349803686141968, -0.10503843426704407, -0.1498464196920395, -0.38530316948890686, 0.11261560767889023, -0.4202083945274353, 0.3758963644504547, -0.08597502112388611, 0.12252431362867355, -0.07361695915460587, -0.3276987671852112, -0.05533349886536598, 0.6137102246284485, 0.5807363986968994, 0.050126899033784866, -0.33917200565338135, 0.007756744045764208, -0.21130061149597168, -0.023091385141015053, -0.09389332681894302, 0.15532897412776947, 0.33820486068725586, -0.040855325758457184, -0.3091779947280884, 0.5142861604690552, -0.11258623749017715, 0.016241922974586487, -0.23662200570106506, -0.2566821575164795, -0.2402956783771515, -0.043224044144153595, -0.12301512062549591, 0.019926153123378754, -0.1297467052936554, 0.025179626420140266, 0.09975019097328186, 0.01266563218086958, -0.23290440440177917, -0.2464844137430191, -0.13393035531044006, -0.14448674023151398, 0.1830042004585266, 0.3253815770149231, 0.04744619503617287, 0.17619532346725464, 0.6140016317367554, 0.22167596220970154, -0.05016925930976868, -0.24814768135547638, 0.06540388613939285, 0.15567611157894135, -0.2556733191013336, -0.21238122880458832, 0.10414333641529083, 0.17469236254692078, 0.09654462337493896, 0.2774951756000519, 0.1864689588546753, -0.00625629024580121, -0.1428343951702118, -0.4487023651599884, -0.5422471761703491, 0.15174885094165802, -0.1925469934940338, 0.1390155404806137, -0.1516934335231781, 0.2596281170845032, 0.11874712258577347, -0.26524877548217773, -0.337027907371521, -0.09570225328207016, -0.32341182231903076, -0.005745481699705124, 0.39344820380210876, 0.05905340611934662, 0.19687822461128235, 0.04367301985621452, 0.13710643351078033, -0.036699239164590836, -0.2849011719226837, -0.15982770919799805, 0.04887757450342178, 0.10002416372299194, 0.13717131316661835, -0.4027870297431946, -0.11658163368701935, -0.288262277841568, -0.08812706172466278, -0.16044308245182037, 0.050287120044231415, 0.29853373765945435, 0.03162457048892975, 0.2532196640968323, -0.313344806432724, 0.0580083467066288, -0.2136874794960022, 0.15737789869308472, -0.15283800661563873, 0.18410509824752808, 0.05986035615205765, 0.12612994015216827, 0.01767563819885254, 0.009883875027298927, -0.5507248640060425, -0.06257525086402893, -0.09079775959253311, 0.18297632038593292, 0.1152568981051445, -0.4182228744029999, 0.22805607318878174, -0.14199566841125488, 0.17141960561275482, 0.19626258313655853, -0.03331775963306427, 0.2832372784614563, 0.181142657995224, 0.18628522753715515, -0.2386668622493744, 0.20860807597637177, 0.21577955782413483, 0.09463855624198914, -0.09388963878154755, 0.0430794395506382, -0.04945429414510727, -0.025662260130047798, 0.2692062258720398, -0.1024496927857399, 0.3867934048175812, -0.40722113847732544, -0.05248969793319702, 0.11946595460176468, 0.0713922306895256, -0.052523281425237656, 0.1153058260679245, 0.3662636876106262, 0.16459231078624725, 0.7827573418617249, -0.2470327615737915, 0.14657162129878998, -0.18498484790325165, 0.3281096816062927, -0.24878251552581787, -0.33459779620170593, -0.33597877621650696, 0.11348117142915726, -0.29427027702331543, -0.12617415189743042, -0.18528464436531067, 0.2995419502258301, -0.10181436687707901, 0.1427348256111145, -0.23578010499477386, 0.3921545445919037, -0.148835688829422, -0.045395225286483765, -0.562584400177002, -0.22663293778896332, 0.10585535317659378, -0.1797059178352356, 0.025090906769037247, 0.01599462516605854, -0.009149172343313694, 0.13094381988048553, -0.26523810625076294, -0.40185314416885376, 0.3093723654747009, 0.4047764539718628, -0.016356606036424637, -0.27932968735694885, 0.41067445278167725, 0.2905891537666321, 0.0659504383802414, 0.2053302377462387, 0.319293349981308, 0.42760345339775085, 0.4022086262702942, 0.06650719046592712, 0.20000672340393066, -0.10694000124931335, -0.25286245346069336, -0.013881714083254337, 0.0525229349732399, 0.3014645576477051, 0.05845606327056885, 0.312530517578125, 0.14970752596855164, -0.0062347291968762875, 0.14302214980125427, 0.04569157585501671, 0.34620293974876404, 0.09222419559955597, -0.015020138584077358, -0.14654742181301117, -0.24552860856056213, -0.18946370482444763, -0.12856698036193848, -0.5581471920013428, 0.13780361413955688, 0.21282006800174713, 0.2150428146123886, 0.2549575865268707, -0.004209568258374929, 0.049009550362825394, -0.06656995415687561, 0.1328437179327011, 0.5250777006149292, -0.03401191532611847, -0.27033570408821106, -0.3209694027900696, -0.5026275515556335, 0.16537779569625854, -0.02528417855501175, -0.42086562514305115, -0.24853593111038208, 0.1211538165807724, -0.1755254566669464, -0.2355567365884781, 0.24101947247982025, 0.10140462964773178, 0.06959448009729385, 0.3392389118671417, -0.22689968347549438, -0.3327464163303375, 0.26964059472084045, 0.05312706530094147, -0.1709013134241104, -0.17286847531795502, -0.09464504569768906, -0.14867763221263885, 0.04900035262107849, -0.04603910446166992, 0.3904176354408264, 0.007114359177649021, 0.20572982728481293, 0.197795107960701, 0.199811652302742, 0.4811958074569702, 0.06806617975234985, -0.040438488125801086, -0.3731039762496948, -0.11173656582832336, -0.20125913619995117, -0.04157199710607529, 0.31112125515937805, 0.5081038475036621, -0.3583826422691345, -0.008344961330294609, -0.3815689980983734, 0.12783004343509674, -0.16659362614154816, -0.017484210431575775, -0.29559364914894104, 0.3703537583351135, -0.2835768163204193, 0.2372782975435257, 0.1641736626625061, -0.045551177114248276, -0.0290332343429327, 0.10304368287324905, -0.3331677317619324, 0.0036246071103960276, 0.5232220888137817, -0.21019794046878815, -0.2365408092737198, 0.09007932245731354, 0.1933176964521408, 0.1135040745139122, 0.010346866212785244, -0.3378690779209137, 0.1756683588027954, 0.18242180347442627, -0.005351877771317959, -0.2406204491853714, -0.0816911906003952, 0.0652153491973877, 0.11355292052030563, -0.06532721221446991, -0.14270712435245514, 0.15939894318580627, -0.0728515014052391, -0.013610941357910633, -0.08653884381055832 ]
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 ?
[ -0.1408683806657791, -0.1892591118812561, -0.03353653848171234, 0.46936744451522827, 0.4054456651210785, 0.112000472843647, 0.05603649839758873, 0.3544987142086029, -0.10948891192674637, 0.2927757501602173, -0.1404101848602295, 0.16212107241153717, 0.05022774264216423, -0.02804594300687313, 0.05054078996181488, -0.03532538563013077, -0.1691269874572754, 0.1510201394557953, 0.13207300007343292, 0.005323641002178192, -0.1707707941532135, 0.045582517981529236, -0.2690636217594147, 0.2587416470050812, -0.2867065370082855, -0.08528336137533188, 0.09144627302885056, 0.3859780728816986, -0.3081700801849365, -0.5349303483963013, 0.045082226395606995, -0.1443348228931427, 0.2407742589712143, 0.19576658308506012, -0.00011105801240773872, 0.049332816153764725, 0.2629140019416809, -0.05734889954328537, -0.5286890864372253, -0.14541827142238617, -0.35688459873199463, -0.12314393371343613, -0.014699259772896767, -0.31323131918907166, -0.12878188490867615, -0.22096027433872223, 0.40224409103393555, -0.3689236044883728, 0.2718048095703125, 0.31527605652809143, 0.20390814542770386, 0.20004352927207947, 0.12182676792144775, -0.054713886231184006, 0.04933757707476616, 0.13519367575645447, -0.13144762814044952, -0.013970177620649338, 0.23928497731685638, 0.1932624876499176, 0.16366681456565857, 0.4578830897808075, -0.17119532823562622, -0.11164619773626328, 0.25520724058151245, 0.243842214345932, -0.030298270285129547, -0.16049779951572418, 0.3209943473339081, 0.2331337183713913, 0.4437566101551056, -0.4332888126373291, -0.4334990680217743, -0.11148460954427719, -0.003818502649664879, -0.26497238874435425, 0.24886417388916016, 0.10019537061452866, -0.3177081346511841, 0.1554599553346634, -0.11993873119354248, -0.36654940247535706, -0.3681374490261078, 0.1454017162322998, 0.26279857754707336, 0.38023069500923157, -0.05968918278813362, 0.0670863687992096, -0.06119953840970993, -0.19583746790885925, 0.05281441658735275, -0.0894942656159401, -0.16413280367851257, 0.30266934633255005, -0.2269848883152008, 0.1276467889547348, 0.13528814911842346, 0.05715295672416687, 0.11984618753194809, 0.3628025949001312, 0.20324356853961945, -0.18401360511779785, -0.375539630651474, 0.06257759779691696, -0.26458486914634705, 0.4228105843067169, 0.295701265335083, 0.14088541269302368, 0.3827153444290161, -0.008944286033511162, 0.2218949943780899, -0.3649007976055145, -0.29457464814186096, -0.16017687320709229, -0.11438271403312683, 0.25096845626831055, 0.36158040165901184, -0.15953703224658966, -0.24764703214168549, -0.059575796127319336, -0.12639565765857697, 0.019348111003637314, 0.20292779803276062, 0.5645594000816345, 0.0536450631916523, 0.1894831657409668, 0.03513301908969879, 0.32177814841270447, 0.05370361730456352, -0.5332470536231995, -0.17430877685546875, 0.21654145419597626, -0.30879589915275574, 0.14569520950317383, 0.3105255663394928, -0.331324964761734, 0.3001636564731598, -0.1157546415925026, -0.30328547954559326, 0.0032351629342883825, 0.12947304546833038, -0.17491817474365234, 0.10203277319669724, 0.07796083390712738, 0.09318635612726212, 0.10305067896842957, 0.19042691588401794, -0.07996855676174164, 0.065503790974617, -0.03327649086713791, -0.04719267040491104, -0.48522597551345825, 0.025506697595119476, 0.13068559765815735, 0.07165276259183884, -0.009505514986813068, -0.08786728978157043, -0.05430344119668007, 0.19623053073883057, -0.22172126173973083, -0.11173630505800247, 0.11289944499731064, -0.02556457184255123, -0.271345853805542, 0.21264243125915527, 0.5540772676467896, -0.3136104941368103, 0.20926669239997864, -0.19852595031261444, -0.13489241898059845, 0.16316388547420502, 0.3628433346748352, -0.11672967672348022, 0.07758896797895432, -0.38488221168518066, 0.32663190364837646, 0.6447770595550537, -0.5202512145042419, -0.507689356803894, 0.2816506624221802, -0.222357839345932, -0.0715678259730339, 0.5067176818847656, 0.24071189761161804, -0.0645313635468483, -0.1131831631064415, 0.047000400722026825, 0.11700915545225143, 0.046720925718545914, -0.08850014209747314, -0.1114891916513443, -0.24508875608444214, 0.34023773670196533, 0.11301998794078827, -0.03068048134446144, 0.13209125399589539, -0.029109438881278038, 0.16131925582885742, 0.3716830313205719, -0.291082501411438, 0.15118570625782013, 0.284060537815094, 0.3005661368370056, 0.03720932453870773, -0.14257796108722687, -0.11960297077894211, -0.294660747051239, 0.23998583853244781, -0.0920315757393837, -0.020116664469242096, -0.3088814616203308, 0.0002528683398850262, -0.11309006810188293, -0.040919654071331024, -0.38762661814689636, -0.2783125042915344, 0.1904965192079544, 0.45270562171936035, 0.22734427452087402, -0.019652726128697395, -0.28994178771972656, 0.3173922300338745, -0.48267507553100586, 0.0027163613121956587, -0.6354675889015198, 0.12806390225887299, -0.09318916499614716, -0.14903521537780762, -0.1639638990163803, 0.16392529010772705, 0.023550231009721756, -0.002848584670573473, -0.09274846315383911, 0.21247054636478424, -0.23807621002197266, 0.08131276816129684, 0.2638706564903259, 0.10236118733882904, 0.0382031612098217, -0.5709120035171509, 0.33758243918418884, 0.4717598557472229, 0.22842155396938324, 0.08941130340099335, -0.26937147974967957, 0.11830674111843109, -0.08254852145910263, 0.09030044078826904, 0.0437278151512146, 0.01720273867249489, 0.4666252136230469, 0.01481084618717432, 0.06998182088136673, -0.21279394626617432, 0.30032503604888916, 0.21649004518985748, 0.4621780514717102, -0.08612556010484695, -0.5364863276481628, 0.08613045513629913, 0.27960124611854553, 0.06051981449127197, -0.09864352643489838, -0.01093621738255024, -0.35100457072257996, -0.04163489118218422, 0.05506303161382675, 0.22897903621196747, 0.2813912332057953, 0.22780871391296387, -0.06535414606332779, 0.0831112340092659, 0.09837401658296585, -0.2621798515319824, 0.18072955310344696, -0.2579272389411926, 0.10123036801815033, 0.2698712944984436, 0.19980134069919586, 0.03205374628305435, -0.03564492240548134, -0.5033629536628723, 0.07217034697532654, 0.445490300655365, -0.270343542098999, -0.17282283306121826, -0.13690432906150818, -0.6854812502861023, 0.060650162398815155, 0.07638508826494217, -0.11054059118032455, -0.3432241976261139, -0.022228943184018135, 0.18776096403598785, 0.21456211805343628, -0.07199499011039734, -0.09505226463079453, 0.0755409300327301, 0.06864802539348602, -0.1094084084033966, -0.4193322956562042, -0.31495967507362366, -0.20777533948421478, 0.04685627296566963, 0.27079546451568604, -0.022806797176599503, 0.246942937374115, -0.28027981519699097, 0.06392844021320343, -0.4144379198551178, -0.025489790365099907, 0.02347724698483944, 0.029297424480319023, -0.00047369685489684343, -0.034068841487169266, 0.4818882346153259, 0.4775172173976898, -0.06232468783855438, 0.3475897014141083, 0.04154755920171738, -0.17423026263713837, 0.2578030824661255, -0.21523959934711456, -0.23448705673217773, -0.04180873930454254, -0.5249141454696655, -0.35637015104293823, -0.3282744884490967, 0.08275269716978073, -0.13988064229488373, 0.08039943873882294, 0.3180902898311615, -0.14647722244262695, 0.20626114308834076, -0.03888370841741562, 0.19586797058582306, -0.07545895129442215, -0.22561410069465637, 0.3382495641708374, -0.16139712929725647, -0.21599528193473816, 0.23609322309494019, 0.06490513682365417, 0.21008872985839844, -0.041243430227041245, -0.2306007444858551, 0.17212629318237305, -0.22556982934474945, 0.284463495016098, -0.1305042803287506, -0.15192101895809174, 0.34998250007629395, -0.010504031553864479, -0.0014595107641071081, 0.0043805427849292755, 0.011881567537784576, 0.11780835688114166, 0.010026367381215096, 0.5716209411621094, 0.2039428949356079, 0.41344475746154785, 0.14451955258846283, 0.37235599756240845, 0.44866472482681274, 0.026759229600429535, 0.32839083671569824, -0.26213914155960083, 0.06690725684165955, -0.06588201969861984, -0.3156571388244629, 0.11945395171642303, 0.17475612461566925, -0.17332394421100616, -0.10797251760959625, 0.1386820524930954, 0.38894352316856384, -0.33151960372924805, -0.13302302360534668, -0.28740736842155457, -0.08548351377248764, -0.1463778018951416, 0.020249146968126297, 0.011118698865175247, 0.1291111260652542, 0.20937487483024597, -0.34255450963974, -0.3424656093120575, -0.11213227361440659, 0.46337810158729553, 0.0023359160404652357, 0.15231479704380035, -0.27443617582321167, -0.28418031334877014, -0.5618228912353516, 0.4806252717971802, 0.11800625175237656, 0.21206682920455933, -0.2758861780166626, 0.12500107288360596, 0.26634660363197327, 0.02257581427693367, 0.3283255994319916, -0.30048316717147827, 0.2563644349575043, 0.08078678697347641, -0.09892932325601578, -0.23587952554225922, 0.1696384698152542, -0.1501615196466446, 0.1985006481409073, 0.3358711302280426, 0.19395430386066437, -0.16084091365337372, -0.1110728532075882, -0.016944043338298798, 0.10838139802217484, -0.1749695986509323, -0.21759426593780518, -0.017784247174859047, -0.003704600501805544, -0.1634150892496109, -0.32848235964775085, -0.17156817018985748, 0.24727573990821838, 0.13296093046665192, 0.33589136600494385, 0.0468716099858284, -0.0742722675204277, 0.17847178876399994, 0.22041527926921844, 0.33928734064102173, -0.09613542258739471, -0.15330930054187775, 0.16526304185390472, 0.24983622133731842, 0.07175802439451218, 0.49867457151412964, 0.28984543681144714, -0.3622645139694214, -0.01410093903541565, -0.22840812802314758, 0.1272355318069458, 0.290335476398468, -0.2587491273880005, 0.10405406355857849, 0.03436676040291786, -0.00782658252865076, 0.061695646494627, -0.09590508788824081, 0.2513183653354645, 0.031543806195259094, -0.12594184279441833, -0.3986913859844208, 0.32729798555374146, 0.10116897523403168, 0.003973283804953098, 0.45400184392929077, -0.14631949365139008, -0.22761744260787964, 0.023735547438263893, 0.05723195523023605, 0.9369786381721497, 0.12992826104164124, 0.22923734784126282, 0.39308348298072815, 0.1683122217655182, 0.5259997844696045, -0.3035133183002472, 0.36880117654800415, -0.387155145406723, -0.08039590716362, -0.0029478250071406364, -0.11370435357093811, -0.011663719080388546, -0.1774902045726776, -0.08031540364027023, 0.23407749831676483, -0.15633413195610046, 0.006111164577305317, -0.09859154373407364, 0.26289814710617065, -0.33494704961776733, -0.09046962857246399, -0.46061480045318604, 0.0847524031996727, -0.20800425112247467, 0.3349803686141968, -0.10503843426704407, -0.1498464196920395, -0.38530316948890686, 0.11261560767889023, -0.4202083945274353, 0.3758963644504547, -0.08597502112388611, 0.12252431362867355, -0.07361695915460587, -0.3276987671852112, -0.05533349886536598, 0.6137102246284485, 0.5807363986968994, 0.050126899033784866, -0.33917200565338135, 0.007756744045764208, -0.21130061149597168, -0.023091385141015053, -0.09389332681894302, 0.15532897412776947, 0.33820486068725586, -0.040855325758457184, -0.3091779947280884, 0.5142861604690552, -0.11258623749017715, 0.016241922974586487, -0.23662200570106506, -0.2566821575164795, -0.2402956783771515, -0.043224044144153595, -0.12301512062549591, 0.019926153123378754, -0.1297467052936554, 0.025179626420140266, 0.09975019097328186, 0.01266563218086958, -0.23290440440177917, -0.2464844137430191, -0.13393035531044006, -0.14448674023151398, 0.1830042004585266, 0.3253815770149231, 0.04744619503617287, 0.17619532346725464, 0.6140016317367554, 0.22167596220970154, -0.05016925930976868, -0.24814768135547638, 0.06540388613939285, 0.15567611157894135, -0.2556733191013336, -0.21238122880458832, 0.10414333641529083, 0.17469236254692078, 0.09654462337493896, 0.2774951756000519, 0.1864689588546753, -0.00625629024580121, -0.1428343951702118, -0.4487023651599884, -0.5422471761703491, 0.15174885094165802, -0.1925469934940338, 0.1390155404806137, -0.1516934335231781, 0.2596281170845032, 0.11874712258577347, -0.26524877548217773, -0.337027907371521, -0.09570225328207016, -0.32341182231903076, -0.005745481699705124, 0.39344820380210876, 0.05905340611934662, 0.19687822461128235, 0.04367301985621452, 0.13710643351078033, -0.036699239164590836, -0.2849011719226837, -0.15982770919799805, 0.04887757450342178, 0.10002416372299194, 0.13717131316661835, -0.4027870297431946, -0.11658163368701935, -0.288262277841568, -0.08812706172466278, -0.16044308245182037, 0.050287120044231415, 0.29853373765945435, 0.03162457048892975, 0.2532196640968323, -0.313344806432724, 0.0580083467066288, -0.2136874794960022, 0.15737789869308472, -0.15283800661563873, 0.18410509824752808, 0.05986035615205765, 0.12612994015216827, 0.01767563819885254, 0.009883875027298927, -0.5507248640060425, -0.06257525086402893, -0.09079775959253311, 0.18297632038593292, 0.1152568981051445, -0.4182228744029999, 0.22805607318878174, -0.14199566841125488, 0.17141960561275482, 0.19626258313655853, -0.03331775963306427, 0.2832372784614563, 0.181142657995224, 0.18628522753715515, -0.2386668622493744, 0.20860807597637177, 0.21577955782413483, 0.09463855624198914, -0.09388963878154755, 0.0430794395506382, -0.04945429414510727, -0.025662260130047798, 0.2692062258720398, -0.1024496927857399, 0.3867934048175812, -0.40722113847732544, -0.05248969793319702, 0.11946595460176468, 0.0713922306895256, -0.052523281425237656, 0.1153058260679245, 0.3662636876106262, 0.16459231078624725, 0.7827573418617249, -0.2470327615737915, 0.14657162129878998, -0.18498484790325165, 0.3281096816062927, -0.24878251552581787, -0.33459779620170593, -0.33597877621650696, 0.11348117142915726, -0.29427027702331543, -0.12617415189743042, -0.18528464436531067, 0.2995419502258301, -0.10181436687707901, 0.1427348256111145, -0.23578010499477386, 0.3921545445919037, -0.148835688829422, -0.045395225286483765, -0.562584400177002, -0.22663293778896332, 0.10585535317659378, -0.1797059178352356, 0.025090906769037247, 0.01599462516605854, -0.009149172343313694, 0.13094381988048553, -0.26523810625076294, -0.40185314416885376, 0.3093723654747009, 0.4047764539718628, -0.016356606036424637, -0.27932968735694885, 0.41067445278167725, 0.2905891537666321, 0.0659504383802414, 0.2053302377462387, 0.319293349981308, 0.42760345339775085, 0.4022086262702942, 0.06650719046592712, 0.20000672340393066, -0.10694000124931335, -0.25286245346069336, -0.013881714083254337, 0.0525229349732399, 0.3014645576477051, 0.05845606327056885, 0.312530517578125, 0.14970752596855164, -0.0062347291968762875, 0.14302214980125427, 0.04569157585501671, 0.34620293974876404, 0.09222419559955597, -0.015020138584077358, -0.14654742181301117, -0.24552860856056213, -0.18946370482444763, -0.12856698036193848, -0.5581471920013428, 0.13780361413955688, 0.21282006800174713, 0.2150428146123886, 0.2549575865268707, -0.004209568258374929, 0.049009550362825394, -0.06656995415687561, 0.1328437179327011, 0.5250777006149292, -0.03401191532611847, -0.27033570408821106, -0.3209694027900696, -0.5026275515556335, 0.16537779569625854, -0.02528417855501175, -0.42086562514305115, -0.24853593111038208, 0.1211538165807724, -0.1755254566669464, -0.2355567365884781, 0.24101947247982025, 0.10140462964773178, 0.06959448009729385, 0.3392389118671417, -0.22689968347549438, -0.3327464163303375, 0.26964059472084045, 0.05312706530094147, -0.1709013134241104, -0.17286847531795502, -0.09464504569768906, -0.14867763221263885, 0.04900035262107849, -0.04603910446166992, 0.3904176354408264, 0.007114359177649021, 0.20572982728481293, 0.197795107960701, 0.199811652302742, 0.4811958074569702, 0.06806617975234985, -0.040438488125801086, -0.3731039762496948, -0.11173656582832336, -0.20125913619995117, -0.04157199710607529, 0.31112125515937805, 0.5081038475036621, -0.3583826422691345, -0.008344961330294609, -0.3815689980983734, 0.12783004343509674, -0.16659362614154816, -0.017484210431575775, -0.29559364914894104, 0.3703537583351135, -0.2835768163204193, 0.2372782975435257, 0.1641736626625061, -0.045551177114248276, -0.0290332343429327, 0.10304368287324905, -0.3331677317619324, 0.0036246071103960276, 0.5232220888137817, -0.21019794046878815, -0.2365408092737198, 0.09007932245731354, 0.1933176964521408, 0.1135040745139122, 0.010346866212785244, -0.3378690779209137, 0.1756683588027954, 0.18242180347442627, -0.005351877771317959, -0.2406204491853714, -0.0816911906003952, 0.0652153491973877, 0.11355292052030563, -0.06532721221446991, -0.14270712435245514, 0.15939894318580627, -0.0728515014052391, -0.013610941357910633, -0.08653884381055832 ]
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!
[ -0.14572685956954956, -0.16974395513534546, -0.19722990691661835, 0.3397465944290161, -0.11949790269136429, -0.42426037788391113, -0.42039990425109863, 0.14767372608184814, -0.2700442969799042, 0.1725064367055893, -0.08286793529987335, 0.1803882122039795, 0.12033049017190933, 0.008093875832855701, -0.10038977116346359, -0.0979752391576767, 0.16565439105033875, 0.03882868215441704, 0.09248064458370209, -0.3720678985118866, 0.06655950099229813, 0.5696007013320923, -0.03133246302604675, 0.21043486893177032, 0.05402210354804993, 0.2702385187149048, 0.28511330485343933, 0.30352962017059326, -0.14857949316501617, 0.028436794877052307, 0.12800218164920807, 0.03064112551510334, -0.02369813062250614, 0.22214946150779724, -0.00010886789095820859, -0.2347189337015152, 0.1026097908616066, -0.003166122827678919, -0.27572280168533325, -0.3026086986064911, -0.03455707058310509, 0.06436417996883392, -0.12292839586734772, -0.0703945979475975, 0.08584654331207275, -0.044892698526382446, 0.048772066831588745, 0.20954260230064392, 0.4293982982635498, 0.14757366478443146, 0.2262546569108963, -0.302436888217926, -0.18534860014915466, 0.024081973358988762, 0.5763346552848816, 0.16267527639865875, -0.11448317021131516, 0.44726458191871643, 0.005699192173779011, -0.3526369333267212, -0.24019046127796173, -0.008564254269003868, 0.34304091334342957, -0.17255942523479462, 0.13255897164344788, 0.19315510988235474, 0.14858552813529968, -0.04594070836901665, -0.0795355811715126, 0.24472832679748535, -0.27028146386146545, 0.03221140801906586, -0.22717103362083435, -0.27783745527267456, -0.0763155147433281, -0.19976061582565308, 0.11689168214797974, -0.08144566416740417, 0.164285346865654, 0.012247888371348381, 0.04413249343633652, 0.2954265773296356, 0.049639228731393814, 0.007163939066231251, -0.05405556410551071, -0.10784529894590378, 0.05084597319364548, 0.09846587479114532, 0.29527968168258667, -0.05054333806037903, -0.5163143873214722, 0.010771967470645905, -0.2593674659729004, 0.07496055960655212, -0.30228152871131897, -0.037855636328458786, 0.314496248960495, 0.465972900390625, -0.20087100565433502, 0.49962663650512695, 0.2691965103149414, -0.0934176966547966, -0.132206991314888, -0.03944694623351097, -0.1378244012594223, 0.10813269764184952, -0.028946537524461746, 0.020693540573120117, 0.0881183072924614, 0.2246806025505066, -0.13358275592327118, -0.012008275836706161, 0.3982228934764862, -0.5637695789337158, -0.47779834270477295, 0.10259093344211578, -0.26103419065475464, -0.4323970377445221, -0.6891929507255554, -0.06736095249652863, -0.255431205034256, -0.11924252659082413, 0.0989082008600235, 0.22720099985599518, -0.1748848259449005, 0.2922978103160858, 0.046788543462753296, 0.18339082598686218, -0.28950437903404236, 0.019662385806441307, -0.18399950861930847, 0.06357943266630173, -0.3143925368785858, 0.1345241218805313, 0.17056576907634735, 0.11274606734514236, 0.3180577754974365, 0.25605395436286926, 0.27351245284080505, 0.03864363580942154, 0.32986629009246826, -0.2839510440826416, -0.022635240107774734, -0.21111637353897095, -0.21139104664325714, 0.19745378196239471, 0.02712884731590748, 0.09184888750314713, -0.008701791986823082, -0.07489363849163055, -0.18084070086479187, 0.14423634111881256, 0.4666179120540619, 0.20658020675182343, 0.1496511995792389, -0.1105542704463005, 0.0406920462846756, 0.11693733930587769, -0.24016320705413818, -0.5936909914016724, 0.2040604203939438, -0.1365518420934677, -0.12701161205768585, -0.05817794427275658, 0.1098201721906662, -0.03458258509635925, 0.09312111884355545, -0.11743143945932388, 0.3731987178325653, 0.06358207762241364, 0.26031288504600525, 0.4534834325313568, -0.040831148624420166, 0.011864669620990753, -0.16285665333271027, -0.07088562846183777, 0.4665316343307495, -0.43853113055229187, -0.22229698300361633, 0.335500031709671, -0.1650703251361847, -0.06518622487783432, 0.23904655873775482, 0.039276134222745895, -0.3886536955833435, -0.16600681841373444, 0.17745044827461243, 0.1975248008966446, 0.07779905945062637, -0.04839687421917915, -0.2970069348812103, 0.13495881855487823, 0.3166254758834839, 0.06547731161117554, -0.23182781040668488, -0.2993497848510742, -0.1114048957824707, -0.0643918588757515, 0.5141969919204712, 0.13280728459358215, -0.27513790130615234, 0.10499178618192673, -0.27215927839279175, 0.2936934530735016, -0.05169530585408211, 0.11857300996780396, 0.42303702235221863, -0.023375390097498894, -0.7951251268386841, -0.16027696430683136, 0.4370330572128296, -0.05248728767037392, -0.2122921198606491, -0.4714087247848511, 0.24053314328193665, -0.23946455121040344, 0.22328291833400726, 0.05504946783185005, 0.18732212483882904, 0.030354375019669533, 0.02407137304544449, 0.09660714119672775, -0.23642654716968536, -0.15861904621124268, -0.17153574526309967, 0.20367777347564697, -0.22109949588775635, -0.007433130405843258, -0.01054032240062952, 0.3596079349517822, 0.31164130568504333, 0.087580107152462, 0.037304140627384186, 0.2634183466434479, 0.05015543848276138, 0.09804993867874146, 0.3672504425048828, 0.1448051780462265, -0.0007639059331268072, 0.1889609843492508, -0.12266047298908234, 0.6879804730415344, 0.016463015228509903, 0.03225129842758179, -0.34003958106040955, 0.4259154200553894, -0.21703895926475525, 0.20069143176078796, 0.4093107283115387, -0.07173748314380646, -0.05009740591049194, 0.05778007209300995, -0.006799567490816116, -0.017152758315205574, 0.7467244267463684, 0.195269376039505, 0.2502555847167969, 0.002956725424155593, -0.3220260441303253, 0.2596171498298645, 0.3991406261920929, -0.015342758037149906, -0.02085251547396183, -0.11995428800582886, 0.19371414184570312, -0.09338858723640442, -0.03672006353735924, 0.2507387101650238, 0.38106977939605713, 0.19513380527496338, 0.1557071954011917, -0.13119931519031525, -0.33823686838150024, -0.34380200505256653, 0.03513306751847267, -0.1821034550666809, 0.08744698762893677, 0.4963866174221039, 0.10160738229751587, 0.21628694236278534, -0.30207735300064087, -0.5422850847244263, -0.2536480724811554, -0.2223142683506012, 0.06132549047470093, 0.16164575517177582, -0.14391346275806427, -0.36783429980278015, -0.1536611169576645, -0.09190325438976288, -0.291888564825058, -0.27327659726142883, 0.23009710013866425, 0.020875034853816032, -0.07249940931797028, -0.21893742680549622, 0.023232491686940193, 0.14874732494354248, -0.3703942894935608, 0.17838124930858612, -0.017374062910676003, -0.02804275043308735, -0.2795387804508209, 0.2228180319070816, -0.056798651814460754, -0.019304219633340836, -0.031476203352212906, -0.27918684482574463, -0.45083627104759216, 0.16968810558319092, -0.37626785039901733, -0.010960167273879051, 0.06464443355798721, -0.0032362905330955982, 0.05967200919985771, -0.2506176829338074, -0.07057545334100723, 0.14640071988105774, 0.09220819175243378, -0.1741306632757187, -0.3507436215877533, -0.08965113759040833, 0.03678455203771591, 0.01421413291245699, -0.19140613079071045, -0.2775220274925232, 0.07937435805797577, -0.1433200091123581, 0.6197946667671204, 0.32701390981674194, 0.005701013375073671, 0.38587868213653564, -0.23695549368858337, 0.00001166912534245057, 0.1717425435781479, 0.5837677121162415, -0.2605839669704437, -0.04442860186100006, 0.19860906898975372, -0.2997126877307892, -0.44218897819519043, -0.27306148409843445, 0.18830041587352753, 0.032321929931640625, -0.008429384790360928, -0.17861713469028473, 0.03100421279668808, -0.130944162607193, -0.04899725317955017, -0.21304549276828766, -0.0558176226913929, 0.33782678842544556, -0.1836601197719574, -0.24219219386577606, -0.06447426229715347, 0.2167336493730545, 0.029940590262413025, 0.2865462005138397, 0.45993542671203613, -0.4840274751186371, 0.16461245715618134, -0.14009401202201843, 0.12755998969078064, 0.04408295452594757, 0.5117961168289185, 0.15338148176670074, 0.13323117792606354, 0.000875302474014461, -0.14067688584327698, 0.09818056970834732, 0.48969370126724243, 0.056188344955444336, -0.1449027806520462, 0.33084896206855774, 0.20873352885246277, 0.24980418384075165, -0.11164986342191696, -0.04191569611430168, -0.07632195204496384, -0.007912340573966503, 0.10350745916366577, 0.07728642970323563, -0.056000638753175735, 0.017135536298155785, 0.07414510101079941, -0.0036139548756182194, -0.1537918746471405, 0.21246279776096344, 0.07286876440048218, 0.3913016617298126, -0.11124490946531296, -0.2594479024410248, -0.005773448385298252, -0.46897637844085693, -0.005584189668297768, 0.03185326233506203, -0.02154267579317093, -0.15841053426265717, -0.1728692352771759, 0.05907154083251953, -0.34859004616737366, 0.26234257221221924, 0.16069407761096954, 0.23712922632694244, 0.1690763384103775, -0.13415345549583435, -0.5272521376609802, -0.29414820671081543, -0.33222100138664246, 0.1811346411705017, 0.5960766077041626, 0.11514391005039215, -0.13822145760059357, -0.1728227287530899, 0.10999253392219543, -0.051850780844688416, -0.1482481211423874, 0.100667804479599, -0.33127155900001526, 0.20241914689540863, 0.15132899582386017, -0.28061461448669434, 0.04583752155303955, 0.3753645420074463, -0.2438100129365921, 0.12480227649211884, 0.06615743041038513, -0.14999310672283173, 0.21232935786247253, 0.13495203852653503, 0.33930912613868713, -0.10402200371026993, -0.24397091567516327, -0.2931976616382599, 0.301355242729187, -0.35144299268722534, 0.3701416254043579, 0.12124454975128174, 0.0003356844827067107, 0.30029141902923584, -0.09358745068311691, 0.10684172809123993, 0.3101276159286499, -0.18164154887199402, 0.16640682518482208, -0.6221522092819214, 0.1127028614282608, -0.44617894291877747, -0.21020950376987457, 0.36847397685050964, -0.07535780966281891, 0.2850716710090637, -0.1410285234451294, 0.20845356583595276, 0.23064082860946655, -0.4770977795124054, -0.07949856668710709, -0.1620941162109375, -0.19412706792354584, 0.27743759751319885, -0.020417451858520508, 0.7596278786659241, 0.4354729950428009, 0.5633991360664368, 0.23021675646305084, -0.14735116064548492, 0.3893979489803314, -0.5955852270126343, 0.19127967953681946, -0.3415159583091736, -0.3702351450920105, 0.09574329853057861, 0.1044330894947052, -0.06511340290307999, 0.21454893052577972, 0.20764988660812378, 0.2953760325908661, 0.3931213319301605, -0.09911183267831802, -0.10828020423650742, 0.17655330896377563, 0.12343073636293411, 0.18754133582115173, -0.28709059953689575, 0.19913668930530548, 0.3464435636997223, 0.004576630890369415, -0.05969354137778282, 0.0959683358669281, -0.7420675754547119, 0.02251577004790306, 0.042491283267736435, 0.10044656693935394, -0.09559985250234604, 0.35474273562431335, 0.1195804700255394, -0.22432774305343628, -0.04871835187077522, 0.36723893880844116, 0.15987487137317657, -0.015912538394331932, 0.007196633145213127, 0.21002450585365295, -0.31940823793411255, 0.32633113861083984, 0.242310032248497, -0.03880343958735466, 0.031407810747623444, -0.22021400928497314, -0.07871436327695847, -0.07076071947813034, 0.15572457015514374, -0.16848225891590118, -0.5484740734100342, 0.18215550482273102, -0.3347029685974121, 0.00720702251419425, -0.02991711162030697, 0.10068954527378082, -0.09960561245679855, -0.053430404514074326, 0.1672593653202057, -0.05686331167817116, -0.06265366077423096, -0.0348849780857563, 0.09307186305522919, -0.33285799622535706, -0.05046362057328224, 0.47442546486854553, -0.1367378532886505, -0.046009741723537445, 0.06487458944320679, -0.3305658996105194, 0.03526678681373596, -0.2431761771440506, 0.03892475739121437, -0.08075731247663498, -0.19715368747711182, -0.10430093109607697, -0.007580780424177647, -0.22914928197860718, 0.25272732973098755, 0.2584366202354431, 0.24956689774990082, 0.21303744614124298, 0.032852526754140854, -0.3630850315093994, -0.0983254685997963, 0.08107906579971313, 0.16799966990947723, -0.034733593463897705, -0.01961834914982319, -0.15958867967128754, 0.25884079933166504, 0.2306322157382965, -0.35463547706604004, -0.15017247200012207, -0.08997651189565659, 0.12681721150875092, -0.22102321684360504, -0.31873923540115356, 0.022486506029963493, 0.14685355126857758, 0.14308586716651917, -0.008150095120072365, -0.057829126715660095, -0.24771755933761597, -0.2051662653684616, 0.08322573453187943, -0.04834730178117752, -0.09744089096784592, -0.13200174272060394, 0.2183990627527237, 0.1043625921010971, -0.14063476026058197, -0.23132243752479553, -0.13392587006092072, 0.3573988676071167, 0.12174534797668457, -0.08553808182477951, 0.18251514434814453, -0.14263035356998444, 0.10866622626781464, 0.01882336474955082, -0.025385407730937004, 0.07357721030712128, -0.026512552052736282, 0.10945802181959152, -0.20184217393398285, -0.2564111649990082, -0.07424166798591614, -0.2803894579410553, 0.1471734195947647, 0.10414636880159378, -0.05297306552529335, 0.00878276489675045, -0.18328934907913208, 0.2844216227531433, 0.3649750351905823, 0.11780503392219543, -0.09963679313659668, -0.01391767617315054, 0.2252829521894455, -0.2222018986940384, 0.0942535474896431, 0.4496176540851593, 0.04398674890398979, 0.06563549488782883, -0.02828432433307171, 0.15857844054698944, -0.20769768953323364, -0.018285106867551804, -0.05924857780337334, -0.2744238078594208, -0.01816912554204464, 0.17280738055706024, -0.19119606912136078, 0.25152310729026794, 0.07363126426935196, 0.15538185834884644, -0.37914833426475525, 0.18718236684799194, 0.35076987743377686, -0.002578590763732791, 0.1941521018743515, -0.2856433689594269, 0.14951296150684357, -0.023675812408328056, 0.09032037109136581, -0.2703600525856018, 0.34511232376098633, 0.08539851009845734, -0.02180582284927368, 0.08363573998212814, 0.16912148892879486, -0.10833372920751572, -0.33761313557624817, -0.0048825740814208984, -0.08809257298707962, -0.07047509402036667, -0.5444002151489258, -0.46642446517944336, -0.22647009789943695, -0.394583523273468, 0.07105033099651337, -0.16874061524868011, 0.38178667426109314, 0.029519328847527504, 0.2477216124534607, -0.09672555327415466, -0.18267722427845, -0.4617646038532257, 0.010246986523270607, 0.4208807647228241, -0.04675710201263428, 0.17743252217769623, 0.2167159467935562, 0.41431310772895813, 0.35717928409576416, 0.1616116464138031, 0.238745778799057, 0.18882735073566437, -0.2336145043373108, -0.22793073952198029, 0.032466839998960495, 0.08932045102119446, -0.1961580216884613, 0.17421601712703705, 0.2599774897098541, 0.04339025169610977, 0.30021801590919495, 0.13220681250095367, -0.24108189344406128, 0.07742052525281906, 0.2649872303009033, -0.0035745594650506973, -0.21412993967533112, 0.2947370409965515, -0.03916141763329506, 0.018681896850466728, -0.3314743936061859, 0.37974855303764343, -0.5077440738677979, -0.08658883720636368, -0.5439392924308777, 0.1514037698507309, 0.17864568531513214, -0.12435479462146759, 0.10393224656581879, 0.2688938081264496, 0.4173148274421692, -0.1272355616092682, 0.3387746810913086, -0.2841376066207886, -0.4613204300403595, -0.2653636336326599, -0.04584995284676552, 0.039508745074272156, 0.27052944898605347, -0.4309377372264862, 0.28873559832572937, -0.29741543531417847, 0.0514499731361866, 0.2745603024959564, 0.4283302426338196, 0.24691928923130035, 0.06649776548147202, -0.338229238986969, 0.1559208780527115, 0.19840243458747864, -0.08824510127305984, -0.11956949532032013, 0.1306329220533371, -0.1739061176776886, 0.019816575571894646, 0.10971532762050629, 0.08555539697408676, -0.34653621912002563, -0.2886534035205841, 0.4207552969455719, 0.3385227918624878, 0.2750665247440338, 0.26808667182922363, -0.3725818991661072, -0.38539445400238037, 0.031004171818494797, -0.12375252693891525, -0.062709741294384, 0.06228296458721161, -0.06584187597036362, 0.38764384388923645, 0.025063978508114815, 0.0747738853096962, 0.0738610103726387, 0.5041284561157227, 0.055955689400434494, 0.0017196714179590344, -0.2270296812057495, -0.02620655670762062, 0.0936463251709938, 0.06342288106679916, -0.6129947900772095, 0.4074437618255615, 0.0005169608048163354, 0.1871442049741745, -0.23370830714702606, -0.2538999319076538, 0.3750455975532532, -0.35842156410217285, 0.27204298973083496, -0.09970194101333618, 0.34275421500205994, 0.22176185250282288, 0.2968052625656128, 0.04679151251912117, -0.10589530318975449, 0.07175473868846893, 0.2957477271556854, 0.16224460303783417, 0.2303461879491806, -0.23342002928256989, -0.0838029682636261, -0.001676929765380919, 0.2927951216697693, 0.06492647528648376, -0.29197123646736145, 0.4134334921836853, -0.18410761654376984 ]
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 ^^
[ -0.14572685956954956, -0.16974395513534546, -0.19722990691661835, 0.3397465944290161, -0.11949790269136429, -0.42426037788391113, -0.42039990425109863, 0.14767372608184814, -0.2700442969799042, 0.1725064367055893, -0.08286793529987335, 0.1803882122039795, 0.12033049017190933, 0.008093875832855701, -0.10038977116346359, -0.0979752391576767, 0.16565439105033875, 0.03882868215441704, 0.09248064458370209, -0.3720678985118866, 0.06655950099229813, 0.5696007013320923, -0.03133246302604675, 0.21043486893177032, 0.05402210354804993, 0.2702385187149048, 0.28511330485343933, 0.30352962017059326, -0.14857949316501617, 0.028436794877052307, 0.12800218164920807, 0.03064112551510334, -0.02369813062250614, 0.22214946150779724, -0.00010886789095820859, -0.2347189337015152, 0.1026097908616066, -0.003166122827678919, -0.27572280168533325, -0.3026086986064911, -0.03455707058310509, 0.06436417996883392, -0.12292839586734772, -0.0703945979475975, 0.08584654331207275, -0.044892698526382446, 0.048772066831588745, 0.20954260230064392, 0.4293982982635498, 0.14757366478443146, 0.2262546569108963, -0.302436888217926, -0.18534860014915466, 0.024081973358988762, 0.5763346552848816, 0.16267527639865875, -0.11448317021131516, 0.44726458191871643, 0.005699192173779011, -0.3526369333267212, -0.24019046127796173, -0.008564254269003868, 0.34304091334342957, -0.17255942523479462, 0.13255897164344788, 0.19315510988235474, 0.14858552813529968, -0.04594070836901665, -0.0795355811715126, 0.24472832679748535, -0.27028146386146545, 0.03221140801906586, -0.22717103362083435, -0.27783745527267456, -0.0763155147433281, -0.19976061582565308, 0.11689168214797974, -0.08144566416740417, 0.164285346865654, 0.012247888371348381, 0.04413249343633652, 0.2954265773296356, 0.049639228731393814, 0.007163939066231251, -0.05405556410551071, -0.10784529894590378, 0.05084597319364548, 0.09846587479114532, 0.29527968168258667, -0.05054333806037903, -0.5163143873214722, 0.010771967470645905, -0.2593674659729004, 0.07496055960655212, -0.30228152871131897, -0.037855636328458786, 0.314496248960495, 0.465972900390625, -0.20087100565433502, 0.49962663650512695, 0.2691965103149414, -0.0934176966547966, -0.132206991314888, -0.03944694623351097, -0.1378244012594223, 0.10813269764184952, -0.028946537524461746, 0.020693540573120117, 0.0881183072924614, 0.2246806025505066, -0.13358275592327118, -0.012008275836706161, 0.3982228934764862, -0.5637695789337158, -0.47779834270477295, 0.10259093344211578, -0.26103419065475464, -0.4323970377445221, -0.6891929507255554, -0.06736095249652863, -0.255431205034256, -0.11924252659082413, 0.0989082008600235, 0.22720099985599518, -0.1748848259449005, 0.2922978103160858, 0.046788543462753296, 0.18339082598686218, -0.28950437903404236, 0.019662385806441307, -0.18399950861930847, 0.06357943266630173, -0.3143925368785858, 0.1345241218805313, 0.17056576907634735, 0.11274606734514236, 0.3180577754974365, 0.25605395436286926, 0.27351245284080505, 0.03864363580942154, 0.32986629009246826, -0.2839510440826416, -0.022635240107774734, -0.21111637353897095, -0.21139104664325714, 0.19745378196239471, 0.02712884731590748, 0.09184888750314713, -0.008701791986823082, -0.07489363849163055, -0.18084070086479187, 0.14423634111881256, 0.4666179120540619, 0.20658020675182343, 0.1496511995792389, -0.1105542704463005, 0.0406920462846756, 0.11693733930587769, -0.24016320705413818, -0.5936909914016724, 0.2040604203939438, -0.1365518420934677, -0.12701161205768585, -0.05817794427275658, 0.1098201721906662, -0.03458258509635925, 0.09312111884355545, -0.11743143945932388, 0.3731987178325653, 0.06358207762241364, 0.26031288504600525, 0.4534834325313568, -0.040831148624420166, 0.011864669620990753, -0.16285665333271027, -0.07088562846183777, 0.4665316343307495, -0.43853113055229187, -0.22229698300361633, 0.335500031709671, -0.1650703251361847, -0.06518622487783432, 0.23904655873775482, 0.039276134222745895, -0.3886536955833435, -0.16600681841373444, 0.17745044827461243, 0.1975248008966446, 0.07779905945062637, -0.04839687421917915, -0.2970069348812103, 0.13495881855487823, 0.3166254758834839, 0.06547731161117554, -0.23182781040668488, -0.2993497848510742, -0.1114048957824707, -0.0643918588757515, 0.5141969919204712, 0.13280728459358215, -0.27513790130615234, 0.10499178618192673, -0.27215927839279175, 0.2936934530735016, -0.05169530585408211, 0.11857300996780396, 0.42303702235221863, -0.023375390097498894, -0.7951251268386841, -0.16027696430683136, 0.4370330572128296, -0.05248728767037392, -0.2122921198606491, -0.4714087247848511, 0.24053314328193665, -0.23946455121040344, 0.22328291833400726, 0.05504946783185005, 0.18732212483882904, 0.030354375019669533, 0.02407137304544449, 0.09660714119672775, -0.23642654716968536, -0.15861904621124268, -0.17153574526309967, 0.20367777347564697, -0.22109949588775635, -0.007433130405843258, -0.01054032240062952, 0.3596079349517822, 0.31164130568504333, 0.087580107152462, 0.037304140627384186, 0.2634183466434479, 0.05015543848276138, 0.09804993867874146, 0.3672504425048828, 0.1448051780462265, -0.0007639059331268072, 0.1889609843492508, -0.12266047298908234, 0.6879804730415344, 0.016463015228509903, 0.03225129842758179, -0.34003958106040955, 0.4259154200553894, -0.21703895926475525, 0.20069143176078796, 0.4093107283115387, -0.07173748314380646, -0.05009740591049194, 0.05778007209300995, -0.006799567490816116, -0.017152758315205574, 0.7467244267463684, 0.195269376039505, 0.2502555847167969, 0.002956725424155593, -0.3220260441303253, 0.2596171498298645, 0.3991406261920929, -0.015342758037149906, -0.02085251547396183, -0.11995428800582886, 0.19371414184570312, -0.09338858723640442, -0.03672006353735924, 0.2507387101650238, 0.38106977939605713, 0.19513380527496338, 0.1557071954011917, -0.13119931519031525, -0.33823686838150024, -0.34380200505256653, 0.03513306751847267, -0.1821034550666809, 0.08744698762893677, 0.4963866174221039, 0.10160738229751587, 0.21628694236278534, -0.30207735300064087, -0.5422850847244263, -0.2536480724811554, -0.2223142683506012, 0.06132549047470093, 0.16164575517177582, -0.14391346275806427, -0.36783429980278015, -0.1536611169576645, -0.09190325438976288, -0.291888564825058, -0.27327659726142883, 0.23009710013866425, 0.020875034853816032, -0.07249940931797028, -0.21893742680549622, 0.023232491686940193, 0.14874732494354248, -0.3703942894935608, 0.17838124930858612, -0.017374062910676003, -0.02804275043308735, -0.2795387804508209, 0.2228180319070816, -0.056798651814460754, -0.019304219633340836, -0.031476203352212906, -0.27918684482574463, -0.45083627104759216, 0.16968810558319092, -0.37626785039901733, -0.010960167273879051, 0.06464443355798721, -0.0032362905330955982, 0.05967200919985771, -0.2506176829338074, -0.07057545334100723, 0.14640071988105774, 0.09220819175243378, -0.1741306632757187, -0.3507436215877533, -0.08965113759040833, 0.03678455203771591, 0.01421413291245699, -0.19140613079071045, -0.2775220274925232, 0.07937435805797577, -0.1433200091123581, 0.6197946667671204, 0.32701390981674194, 0.005701013375073671, 0.38587868213653564, -0.23695549368858337, 0.00001166912534245057, 0.1717425435781479, 0.5837677121162415, -0.2605839669704437, -0.04442860186100006, 0.19860906898975372, -0.2997126877307892, -0.44218897819519043, -0.27306148409843445, 0.18830041587352753, 0.032321929931640625, -0.008429384790360928, -0.17861713469028473, 0.03100421279668808, -0.130944162607193, -0.04899725317955017, -0.21304549276828766, -0.0558176226913929, 0.33782678842544556, -0.1836601197719574, -0.24219219386577606, -0.06447426229715347, 0.2167336493730545, 0.029940590262413025, 0.2865462005138397, 0.45993542671203613, -0.4840274751186371, 0.16461245715618134, -0.14009401202201843, 0.12755998969078064, 0.04408295452594757, 0.5117961168289185, 0.15338148176670074, 0.13323117792606354, 0.000875302474014461, -0.14067688584327698, 0.09818056970834732, 0.48969370126724243, 0.056188344955444336, -0.1449027806520462, 0.33084896206855774, 0.20873352885246277, 0.24980418384075165, -0.11164986342191696, -0.04191569611430168, -0.07632195204496384, -0.007912340573966503, 0.10350745916366577, 0.07728642970323563, -0.056000638753175735, 0.017135536298155785, 0.07414510101079941, -0.0036139548756182194, -0.1537918746471405, 0.21246279776096344, 0.07286876440048218, 0.3913016617298126, -0.11124490946531296, -0.2594479024410248, -0.005773448385298252, -0.46897637844085693, -0.005584189668297768, 0.03185326233506203, -0.02154267579317093, -0.15841053426265717, -0.1728692352771759, 0.05907154083251953, -0.34859004616737366, 0.26234257221221924, 0.16069407761096954, 0.23712922632694244, 0.1690763384103775, -0.13415345549583435, -0.5272521376609802, -0.29414820671081543, -0.33222100138664246, 0.1811346411705017, 0.5960766077041626, 0.11514391005039215, -0.13822145760059357, -0.1728227287530899, 0.10999253392219543, -0.051850780844688416, -0.1482481211423874, 0.100667804479599, -0.33127155900001526, 0.20241914689540863, 0.15132899582386017, -0.28061461448669434, 0.04583752155303955, 0.3753645420074463, -0.2438100129365921, 0.12480227649211884, 0.06615743041038513, -0.14999310672283173, 0.21232935786247253, 0.13495203852653503, 0.33930912613868713, -0.10402200371026993, -0.24397091567516327, -0.2931976616382599, 0.301355242729187, -0.35144299268722534, 0.3701416254043579, 0.12124454975128174, 0.0003356844827067107, 0.30029141902923584, -0.09358745068311691, 0.10684172809123993, 0.3101276159286499, -0.18164154887199402, 0.16640682518482208, -0.6221522092819214, 0.1127028614282608, -0.44617894291877747, -0.21020950376987457, 0.36847397685050964, -0.07535780966281891, 0.2850716710090637, -0.1410285234451294, 0.20845356583595276, 0.23064082860946655, -0.4770977795124054, -0.07949856668710709, -0.1620941162109375, -0.19412706792354584, 0.27743759751319885, -0.020417451858520508, 0.7596278786659241, 0.4354729950428009, 0.5633991360664368, 0.23021675646305084, -0.14735116064548492, 0.3893979489803314, -0.5955852270126343, 0.19127967953681946, -0.3415159583091736, -0.3702351450920105, 0.09574329853057861, 0.1044330894947052, -0.06511340290307999, 0.21454893052577972, 0.20764988660812378, 0.2953760325908661, 0.3931213319301605, -0.09911183267831802, -0.10828020423650742, 0.17655330896377563, 0.12343073636293411, 0.18754133582115173, -0.28709059953689575, 0.19913668930530548, 0.3464435636997223, 0.004576630890369415, -0.05969354137778282, 0.0959683358669281, -0.7420675754547119, 0.02251577004790306, 0.042491283267736435, 0.10044656693935394, -0.09559985250234604, 0.35474273562431335, 0.1195804700255394, -0.22432774305343628, -0.04871835187077522, 0.36723893880844116, 0.15987487137317657, -0.015912538394331932, 0.007196633145213127, 0.21002450585365295, -0.31940823793411255, 0.32633113861083984, 0.242310032248497, -0.03880343958735466, 0.031407810747623444, -0.22021400928497314, -0.07871436327695847, -0.07076071947813034, 0.15572457015514374, -0.16848225891590118, -0.5484740734100342, 0.18215550482273102, -0.3347029685974121, 0.00720702251419425, -0.02991711162030697, 0.10068954527378082, -0.09960561245679855, -0.053430404514074326, 0.1672593653202057, -0.05686331167817116, -0.06265366077423096, -0.0348849780857563, 0.09307186305522919, -0.33285799622535706, -0.05046362057328224, 0.47442546486854553, -0.1367378532886505, -0.046009741723537445, 0.06487458944320679, -0.3305658996105194, 0.03526678681373596, -0.2431761771440506, 0.03892475739121437, -0.08075731247663498, -0.19715368747711182, -0.10430093109607697, -0.007580780424177647, -0.22914928197860718, 0.25272732973098755, 0.2584366202354431, 0.24956689774990082, 0.21303744614124298, 0.032852526754140854, -0.3630850315093994, -0.0983254685997963, 0.08107906579971313, 0.16799966990947723, -0.034733593463897705, -0.01961834914982319, -0.15958867967128754, 0.25884079933166504, 0.2306322157382965, -0.35463547706604004, -0.15017247200012207, -0.08997651189565659, 0.12681721150875092, -0.22102321684360504, -0.31873923540115356, 0.022486506029963493, 0.14685355126857758, 0.14308586716651917, -0.008150095120072365, -0.057829126715660095, -0.24771755933761597, -0.2051662653684616, 0.08322573453187943, -0.04834730178117752, -0.09744089096784592, -0.13200174272060394, 0.2183990627527237, 0.1043625921010971, -0.14063476026058197, -0.23132243752479553, -0.13392587006092072, 0.3573988676071167, 0.12174534797668457, -0.08553808182477951, 0.18251514434814453, -0.14263035356998444, 0.10866622626781464, 0.01882336474955082, -0.025385407730937004, 0.07357721030712128, -0.026512552052736282, 0.10945802181959152, -0.20184217393398285, -0.2564111649990082, -0.07424166798591614, -0.2803894579410553, 0.1471734195947647, 0.10414636880159378, -0.05297306552529335, 0.00878276489675045, -0.18328934907913208, 0.2844216227531433, 0.3649750351905823, 0.11780503392219543, -0.09963679313659668, -0.01391767617315054, 0.2252829521894455, -0.2222018986940384, 0.0942535474896431, 0.4496176540851593, 0.04398674890398979, 0.06563549488782883, -0.02828432433307171, 0.15857844054698944, -0.20769768953323364, -0.018285106867551804, -0.05924857780337334, -0.2744238078594208, -0.01816912554204464, 0.17280738055706024, -0.19119606912136078, 0.25152310729026794, 0.07363126426935196, 0.15538185834884644, -0.37914833426475525, 0.18718236684799194, 0.35076987743377686, -0.002578590763732791, 0.1941521018743515, -0.2856433689594269, 0.14951296150684357, -0.023675812408328056, 0.09032037109136581, -0.2703600525856018, 0.34511232376098633, 0.08539851009845734, -0.02180582284927368, 0.08363573998212814, 0.16912148892879486, -0.10833372920751572, -0.33761313557624817, -0.0048825740814208984, -0.08809257298707962, -0.07047509402036667, -0.5444002151489258, -0.46642446517944336, -0.22647009789943695, -0.394583523273468, 0.07105033099651337, -0.16874061524868011, 0.38178667426109314, 0.029519328847527504, 0.2477216124534607, -0.09672555327415466, -0.18267722427845, -0.4617646038532257, 0.010246986523270607, 0.4208807647228241, -0.04675710201263428, 0.17743252217769623, 0.2167159467935562, 0.41431310772895813, 0.35717928409576416, 0.1616116464138031, 0.238745778799057, 0.18882735073566437, -0.2336145043373108, -0.22793073952198029, 0.032466839998960495, 0.08932045102119446, -0.1961580216884613, 0.17421601712703705, 0.2599774897098541, 0.04339025169610977, 0.30021801590919495, 0.13220681250095367, -0.24108189344406128, 0.07742052525281906, 0.2649872303009033, -0.0035745594650506973, -0.21412993967533112, 0.2947370409965515, -0.03916141763329506, 0.018681896850466728, -0.3314743936061859, 0.37974855303764343, -0.5077440738677979, -0.08658883720636368, -0.5439392924308777, 0.1514037698507309, 0.17864568531513214, -0.12435479462146759, 0.10393224656581879, 0.2688938081264496, 0.4173148274421692, -0.1272355616092682, 0.3387746810913086, -0.2841376066207886, -0.4613204300403595, -0.2653636336326599, -0.04584995284676552, 0.039508745074272156, 0.27052944898605347, -0.4309377372264862, 0.28873559832572937, -0.29741543531417847, 0.0514499731361866, 0.2745603024959564, 0.4283302426338196, 0.24691928923130035, 0.06649776548147202, -0.338229238986969, 0.1559208780527115, 0.19840243458747864, -0.08824510127305984, -0.11956949532032013, 0.1306329220533371, -0.1739061176776886, 0.019816575571894646, 0.10971532762050629, 0.08555539697408676, -0.34653621912002563, -0.2886534035205841, 0.4207552969455719, 0.3385227918624878, 0.2750665247440338, 0.26808667182922363, -0.3725818991661072, -0.38539445400238037, 0.031004171818494797, -0.12375252693891525, -0.062709741294384, 0.06228296458721161, -0.06584187597036362, 0.38764384388923645, 0.025063978508114815, 0.0747738853096962, 0.0738610103726387, 0.5041284561157227, 0.055955689400434494, 0.0017196714179590344, -0.2270296812057495, -0.02620655670762062, 0.0936463251709938, 0.06342288106679916, -0.6129947900772095, 0.4074437618255615, 0.0005169608048163354, 0.1871442049741745, -0.23370830714702606, -0.2538999319076538, 0.3750455975532532, -0.35842156410217285, 0.27204298973083496, -0.09970194101333618, 0.34275421500205994, 0.22176185250282288, 0.2968052625656128, 0.04679151251912117, -0.10589530318975449, 0.07175473868846893, 0.2957477271556854, 0.16224460303783417, 0.2303461879491806, -0.23342002928256989, -0.0838029682636261, -0.001676929765380919, 0.2927951216697693, 0.06492647528648376, -0.29197123646736145, 0.4134334921836853, -0.18410761654376984 ]
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!
[ -0.14572685956954956, -0.16974395513534546, -0.19722990691661835, 0.3397465944290161, -0.11949790269136429, -0.42426037788391113, -0.42039990425109863, 0.14767372608184814, -0.2700442969799042, 0.1725064367055893, -0.08286793529987335, 0.1803882122039795, 0.12033049017190933, 0.008093875832855701, -0.10038977116346359, -0.0979752391576767, 0.16565439105033875, 0.03882868215441704, 0.09248064458370209, -0.3720678985118866, 0.06655950099229813, 0.5696007013320923, -0.03133246302604675, 0.21043486893177032, 0.05402210354804993, 0.2702385187149048, 0.28511330485343933, 0.30352962017059326, -0.14857949316501617, 0.028436794877052307, 0.12800218164920807, 0.03064112551510334, -0.02369813062250614, 0.22214946150779724, -0.00010886789095820859, -0.2347189337015152, 0.1026097908616066, -0.003166122827678919, -0.27572280168533325, -0.3026086986064911, -0.03455707058310509, 0.06436417996883392, -0.12292839586734772, -0.0703945979475975, 0.08584654331207275, -0.044892698526382446, 0.048772066831588745, 0.20954260230064392, 0.4293982982635498, 0.14757366478443146, 0.2262546569108963, -0.302436888217926, -0.18534860014915466, 0.024081973358988762, 0.5763346552848816, 0.16267527639865875, -0.11448317021131516, 0.44726458191871643, 0.005699192173779011, -0.3526369333267212, -0.24019046127796173, -0.008564254269003868, 0.34304091334342957, -0.17255942523479462, 0.13255897164344788, 0.19315510988235474, 0.14858552813529968, -0.04594070836901665, -0.0795355811715126, 0.24472832679748535, -0.27028146386146545, 0.03221140801906586, -0.22717103362083435, -0.27783745527267456, -0.0763155147433281, -0.19976061582565308, 0.11689168214797974, -0.08144566416740417, 0.164285346865654, 0.012247888371348381, 0.04413249343633652, 0.2954265773296356, 0.049639228731393814, 0.007163939066231251, -0.05405556410551071, -0.10784529894590378, 0.05084597319364548, 0.09846587479114532, 0.29527968168258667, -0.05054333806037903, -0.5163143873214722, 0.010771967470645905, -0.2593674659729004, 0.07496055960655212, -0.30228152871131897, -0.037855636328458786, 0.314496248960495, 0.465972900390625, -0.20087100565433502, 0.49962663650512695, 0.2691965103149414, -0.0934176966547966, -0.132206991314888, -0.03944694623351097, -0.1378244012594223, 0.10813269764184952, -0.028946537524461746, 0.020693540573120117, 0.0881183072924614, 0.2246806025505066, -0.13358275592327118, -0.012008275836706161, 0.3982228934764862, -0.5637695789337158, -0.47779834270477295, 0.10259093344211578, -0.26103419065475464, -0.4323970377445221, -0.6891929507255554, -0.06736095249652863, -0.255431205034256, -0.11924252659082413, 0.0989082008600235, 0.22720099985599518, -0.1748848259449005, 0.2922978103160858, 0.046788543462753296, 0.18339082598686218, -0.28950437903404236, 0.019662385806441307, -0.18399950861930847, 0.06357943266630173, -0.3143925368785858, 0.1345241218805313, 0.17056576907634735, 0.11274606734514236, 0.3180577754974365, 0.25605395436286926, 0.27351245284080505, 0.03864363580942154, 0.32986629009246826, -0.2839510440826416, -0.022635240107774734, -0.21111637353897095, -0.21139104664325714, 0.19745378196239471, 0.02712884731590748, 0.09184888750314713, -0.008701791986823082, -0.07489363849163055, -0.18084070086479187, 0.14423634111881256, 0.4666179120540619, 0.20658020675182343, 0.1496511995792389, -0.1105542704463005, 0.0406920462846756, 0.11693733930587769, -0.24016320705413818, -0.5936909914016724, 0.2040604203939438, -0.1365518420934677, -0.12701161205768585, -0.05817794427275658, 0.1098201721906662, -0.03458258509635925, 0.09312111884355545, -0.11743143945932388, 0.3731987178325653, 0.06358207762241364, 0.26031288504600525, 0.4534834325313568, -0.040831148624420166, 0.011864669620990753, -0.16285665333271027, -0.07088562846183777, 0.4665316343307495, -0.43853113055229187, -0.22229698300361633, 0.335500031709671, -0.1650703251361847, -0.06518622487783432, 0.23904655873775482, 0.039276134222745895, -0.3886536955833435, -0.16600681841373444, 0.17745044827461243, 0.1975248008966446, 0.07779905945062637, -0.04839687421917915, -0.2970069348812103, 0.13495881855487823, 0.3166254758834839, 0.06547731161117554, -0.23182781040668488, -0.2993497848510742, -0.1114048957824707, -0.0643918588757515, 0.5141969919204712, 0.13280728459358215, -0.27513790130615234, 0.10499178618192673, -0.27215927839279175, 0.2936934530735016, -0.05169530585408211, 0.11857300996780396, 0.42303702235221863, -0.023375390097498894, -0.7951251268386841, -0.16027696430683136, 0.4370330572128296, -0.05248728767037392, -0.2122921198606491, -0.4714087247848511, 0.24053314328193665, -0.23946455121040344, 0.22328291833400726, 0.05504946783185005, 0.18732212483882904, 0.030354375019669533, 0.02407137304544449, 0.09660714119672775, -0.23642654716968536, -0.15861904621124268, -0.17153574526309967, 0.20367777347564697, -0.22109949588775635, -0.007433130405843258, -0.01054032240062952, 0.3596079349517822, 0.31164130568504333, 0.087580107152462, 0.037304140627384186, 0.2634183466434479, 0.05015543848276138, 0.09804993867874146, 0.3672504425048828, 0.1448051780462265, -0.0007639059331268072, 0.1889609843492508, -0.12266047298908234, 0.6879804730415344, 0.016463015228509903, 0.03225129842758179, -0.34003958106040955, 0.4259154200553894, -0.21703895926475525, 0.20069143176078796, 0.4093107283115387, -0.07173748314380646, -0.05009740591049194, 0.05778007209300995, -0.006799567490816116, -0.017152758315205574, 0.7467244267463684, 0.195269376039505, 0.2502555847167969, 0.002956725424155593, -0.3220260441303253, 0.2596171498298645, 0.3991406261920929, -0.015342758037149906, -0.02085251547396183, -0.11995428800582886, 0.19371414184570312, -0.09338858723640442, -0.03672006353735924, 0.2507387101650238, 0.38106977939605713, 0.19513380527496338, 0.1557071954011917, -0.13119931519031525, -0.33823686838150024, -0.34380200505256653, 0.03513306751847267, -0.1821034550666809, 0.08744698762893677, 0.4963866174221039, 0.10160738229751587, 0.21628694236278534, -0.30207735300064087, -0.5422850847244263, -0.2536480724811554, -0.2223142683506012, 0.06132549047470093, 0.16164575517177582, -0.14391346275806427, -0.36783429980278015, -0.1536611169576645, -0.09190325438976288, -0.291888564825058, -0.27327659726142883, 0.23009710013866425, 0.020875034853816032, -0.07249940931797028, -0.21893742680549622, 0.023232491686940193, 0.14874732494354248, -0.3703942894935608, 0.17838124930858612, -0.017374062910676003, -0.02804275043308735, -0.2795387804508209, 0.2228180319070816, -0.056798651814460754, -0.019304219633340836, -0.031476203352212906, -0.27918684482574463, -0.45083627104759216, 0.16968810558319092, -0.37626785039901733, -0.010960167273879051, 0.06464443355798721, -0.0032362905330955982, 0.05967200919985771, -0.2506176829338074, -0.07057545334100723, 0.14640071988105774, 0.09220819175243378, -0.1741306632757187, -0.3507436215877533, -0.08965113759040833, 0.03678455203771591, 0.01421413291245699, -0.19140613079071045, -0.2775220274925232, 0.07937435805797577, -0.1433200091123581, 0.6197946667671204, 0.32701390981674194, 0.005701013375073671, 0.38587868213653564, -0.23695549368858337, 0.00001166912534245057, 0.1717425435781479, 0.5837677121162415, -0.2605839669704437, -0.04442860186100006, 0.19860906898975372, -0.2997126877307892, -0.44218897819519043, -0.27306148409843445, 0.18830041587352753, 0.032321929931640625, -0.008429384790360928, -0.17861713469028473, 0.03100421279668808, -0.130944162607193, -0.04899725317955017, -0.21304549276828766, -0.0558176226913929, 0.33782678842544556, -0.1836601197719574, -0.24219219386577606, -0.06447426229715347, 0.2167336493730545, 0.029940590262413025, 0.2865462005138397, 0.45993542671203613, -0.4840274751186371, 0.16461245715618134, -0.14009401202201843, 0.12755998969078064, 0.04408295452594757, 0.5117961168289185, 0.15338148176670074, 0.13323117792606354, 0.000875302474014461, -0.14067688584327698, 0.09818056970834732, 0.48969370126724243, 0.056188344955444336, -0.1449027806520462, 0.33084896206855774, 0.20873352885246277, 0.24980418384075165, -0.11164986342191696, -0.04191569611430168, -0.07632195204496384, -0.007912340573966503, 0.10350745916366577, 0.07728642970323563, -0.056000638753175735, 0.017135536298155785, 0.07414510101079941, -0.0036139548756182194, -0.1537918746471405, 0.21246279776096344, 0.07286876440048218, 0.3913016617298126, -0.11124490946531296, -0.2594479024410248, -0.005773448385298252, -0.46897637844085693, -0.005584189668297768, 0.03185326233506203, -0.02154267579317093, -0.15841053426265717, -0.1728692352771759, 0.05907154083251953, -0.34859004616737366, 0.26234257221221924, 0.16069407761096954, 0.23712922632694244, 0.1690763384103775, -0.13415345549583435, -0.5272521376609802, -0.29414820671081543, -0.33222100138664246, 0.1811346411705017, 0.5960766077041626, 0.11514391005039215, -0.13822145760059357, -0.1728227287530899, 0.10999253392219543, -0.051850780844688416, -0.1482481211423874, 0.100667804479599, -0.33127155900001526, 0.20241914689540863, 0.15132899582386017, -0.28061461448669434, 0.04583752155303955, 0.3753645420074463, -0.2438100129365921, 0.12480227649211884, 0.06615743041038513, -0.14999310672283173, 0.21232935786247253, 0.13495203852653503, 0.33930912613868713, -0.10402200371026993, -0.24397091567516327, -0.2931976616382599, 0.301355242729187, -0.35144299268722534, 0.3701416254043579, 0.12124454975128174, 0.0003356844827067107, 0.30029141902923584, -0.09358745068311691, 0.10684172809123993, 0.3101276159286499, -0.18164154887199402, 0.16640682518482208, -0.6221522092819214, 0.1127028614282608, -0.44617894291877747, -0.21020950376987457, 0.36847397685050964, -0.07535780966281891, 0.2850716710090637, -0.1410285234451294, 0.20845356583595276, 0.23064082860946655, -0.4770977795124054, -0.07949856668710709, -0.1620941162109375, -0.19412706792354584, 0.27743759751319885, -0.020417451858520508, 0.7596278786659241, 0.4354729950428009, 0.5633991360664368, 0.23021675646305084, -0.14735116064548492, 0.3893979489803314, -0.5955852270126343, 0.19127967953681946, -0.3415159583091736, -0.3702351450920105, 0.09574329853057861, 0.1044330894947052, -0.06511340290307999, 0.21454893052577972, 0.20764988660812378, 0.2953760325908661, 0.3931213319301605, -0.09911183267831802, -0.10828020423650742, 0.17655330896377563, 0.12343073636293411, 0.18754133582115173, -0.28709059953689575, 0.19913668930530548, 0.3464435636997223, 0.004576630890369415, -0.05969354137778282, 0.0959683358669281, -0.7420675754547119, 0.02251577004790306, 0.042491283267736435, 0.10044656693935394, -0.09559985250234604, 0.35474273562431335, 0.1195804700255394, -0.22432774305343628, -0.04871835187077522, 0.36723893880844116, 0.15987487137317657, -0.015912538394331932, 0.007196633145213127, 0.21002450585365295, -0.31940823793411255, 0.32633113861083984, 0.242310032248497, -0.03880343958735466, 0.031407810747623444, -0.22021400928497314, -0.07871436327695847, -0.07076071947813034, 0.15572457015514374, -0.16848225891590118, -0.5484740734100342, 0.18215550482273102, -0.3347029685974121, 0.00720702251419425, -0.02991711162030697, 0.10068954527378082, -0.09960561245679855, -0.053430404514074326, 0.1672593653202057, -0.05686331167817116, -0.06265366077423096, -0.0348849780857563, 0.09307186305522919, -0.33285799622535706, -0.05046362057328224, 0.47442546486854553, -0.1367378532886505, -0.046009741723537445, 0.06487458944320679, -0.3305658996105194, 0.03526678681373596, -0.2431761771440506, 0.03892475739121437, -0.08075731247663498, -0.19715368747711182, -0.10430093109607697, -0.007580780424177647, -0.22914928197860718, 0.25272732973098755, 0.2584366202354431, 0.24956689774990082, 0.21303744614124298, 0.032852526754140854, -0.3630850315093994, -0.0983254685997963, 0.08107906579971313, 0.16799966990947723, -0.034733593463897705, -0.01961834914982319, -0.15958867967128754, 0.25884079933166504, 0.2306322157382965, -0.35463547706604004, -0.15017247200012207, -0.08997651189565659, 0.12681721150875092, -0.22102321684360504, -0.31873923540115356, 0.022486506029963493, 0.14685355126857758, 0.14308586716651917, -0.008150095120072365, -0.057829126715660095, -0.24771755933761597, -0.2051662653684616, 0.08322573453187943, -0.04834730178117752, -0.09744089096784592, -0.13200174272060394, 0.2183990627527237, 0.1043625921010971, -0.14063476026058197, -0.23132243752479553, -0.13392587006092072, 0.3573988676071167, 0.12174534797668457, -0.08553808182477951, 0.18251514434814453, -0.14263035356998444, 0.10866622626781464, 0.01882336474955082, -0.025385407730937004, 0.07357721030712128, -0.026512552052736282, 0.10945802181959152, -0.20184217393398285, -0.2564111649990082, -0.07424166798591614, -0.2803894579410553, 0.1471734195947647, 0.10414636880159378, -0.05297306552529335, 0.00878276489675045, -0.18328934907913208, 0.2844216227531433, 0.3649750351905823, 0.11780503392219543, -0.09963679313659668, -0.01391767617315054, 0.2252829521894455, -0.2222018986940384, 0.0942535474896431, 0.4496176540851593, 0.04398674890398979, 0.06563549488782883, -0.02828432433307171, 0.15857844054698944, -0.20769768953323364, -0.018285106867551804, -0.05924857780337334, -0.2744238078594208, -0.01816912554204464, 0.17280738055706024, -0.19119606912136078, 0.25152310729026794, 0.07363126426935196, 0.15538185834884644, -0.37914833426475525, 0.18718236684799194, 0.35076987743377686, -0.002578590763732791, 0.1941521018743515, -0.2856433689594269, 0.14951296150684357, -0.023675812408328056, 0.09032037109136581, -0.2703600525856018, 0.34511232376098633, 0.08539851009845734, -0.02180582284927368, 0.08363573998212814, 0.16912148892879486, -0.10833372920751572, -0.33761313557624817, -0.0048825740814208984, -0.08809257298707962, -0.07047509402036667, -0.5444002151489258, -0.46642446517944336, -0.22647009789943695, -0.394583523273468, 0.07105033099651337, -0.16874061524868011, 0.38178667426109314, 0.029519328847527504, 0.2477216124534607, -0.09672555327415466, -0.18267722427845, -0.4617646038532257, 0.010246986523270607, 0.4208807647228241, -0.04675710201263428, 0.17743252217769623, 0.2167159467935562, 0.41431310772895813, 0.35717928409576416, 0.1616116464138031, 0.238745778799057, 0.18882735073566437, -0.2336145043373108, -0.22793073952198029, 0.032466839998960495, 0.08932045102119446, -0.1961580216884613, 0.17421601712703705, 0.2599774897098541, 0.04339025169610977, 0.30021801590919495, 0.13220681250095367, -0.24108189344406128, 0.07742052525281906, 0.2649872303009033, -0.0035745594650506973, -0.21412993967533112, 0.2947370409965515, -0.03916141763329506, 0.018681896850466728, -0.3314743936061859, 0.37974855303764343, -0.5077440738677979, -0.08658883720636368, -0.5439392924308777, 0.1514037698507309, 0.17864568531513214, -0.12435479462146759, 0.10393224656581879, 0.2688938081264496, 0.4173148274421692, -0.1272355616092682, 0.3387746810913086, -0.2841376066207886, -0.4613204300403595, -0.2653636336326599, -0.04584995284676552, 0.039508745074272156, 0.27052944898605347, -0.4309377372264862, 0.28873559832572937, -0.29741543531417847, 0.0514499731361866, 0.2745603024959564, 0.4283302426338196, 0.24691928923130035, 0.06649776548147202, -0.338229238986969, 0.1559208780527115, 0.19840243458747864, -0.08824510127305984, -0.11956949532032013, 0.1306329220533371, -0.1739061176776886, 0.019816575571894646, 0.10971532762050629, 0.08555539697408676, -0.34653621912002563, -0.2886534035205841, 0.4207552969455719, 0.3385227918624878, 0.2750665247440338, 0.26808667182922363, -0.3725818991661072, -0.38539445400238037, 0.031004171818494797, -0.12375252693891525, -0.062709741294384, 0.06228296458721161, -0.06584187597036362, 0.38764384388923645, 0.025063978508114815, 0.0747738853096962, 0.0738610103726387, 0.5041284561157227, 0.055955689400434494, 0.0017196714179590344, -0.2270296812057495, -0.02620655670762062, 0.0936463251709938, 0.06342288106679916, -0.6129947900772095, 0.4074437618255615, 0.0005169608048163354, 0.1871442049741745, -0.23370830714702606, -0.2538999319076538, 0.3750455975532532, -0.35842156410217285, 0.27204298973083496, -0.09970194101333618, 0.34275421500205994, 0.22176185250282288, 0.2968052625656128, 0.04679151251912117, -0.10589530318975449, 0.07175473868846893, 0.2957477271556854, 0.16224460303783417, 0.2303461879491806, -0.23342002928256989, -0.0838029682636261, -0.001676929765380919, 0.2927951216697693, 0.06492647528648376, -0.29197123646736145, 0.4134334921836853, -0.18410761654376984 ]
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?
[ 0.0004874137230217457, -0.1943998485803604, 0.0977800041437149, 0.14103536307811737, 0.28600648045539856, -0.21947555243968964, 0.298161119222641, 0.1757620871067047, 0.16847147047519684, 0.15842962265014648, -0.22391535341739655, 0.20470982789993286, -0.4715833365917206, 0.24365317821502686, -0.03945281356573105, -0.09374228864908218, -0.18991324305534363, -0.033631764352321625, 0.10520388931035995, -0.015585018321871758, -0.09069815278053284, 0.2365891933441162, 0.04565009847283363, -0.028711572289466858, -0.31746068596839905, 0.3504810333251953, 0.055219557136297226, 0.055927108973264694, -0.14533673226833344, -0.032764796167612076, 0.3463769555091858, -0.19246923923492432, 0.25576677918434143, 0.792607843875885, -0.00012355620856396854, -0.2130509614944458, 0.06719066947698593, -0.011289763264358044, -0.2755607068538666, -0.3058052957057953, 0.4654071033000946, -0.25233644247055054, 0.0014710916439071298, -0.2047472894191742, -0.02709215320646763, -0.062226615846157074, -0.2024286687374115, -0.08055958896875381, 0.008652371354401112, 0.31620800495147705, 0.04068655148148537, 0.2029191106557846, 0.0005702140624634922, 0.07062742114067078, 0.09595886617898941, -0.10697992891073227, -0.11937197297811508, 0.5899263620376587, 0.1905045509338379, -0.03901801258325577, -0.19934454560279846, 0.16511183977127075, 0.15574252605438232, 0.2218758761882782, 0.7811313271522522, 0.03963197022676468, -0.10514877736568451, 0.007861083373427391, 0.05174339935183525, 0.3505224287509918, -0.05701658874750137, -0.23309963941574097, -0.24215441942214966, 0.18319889903068542, 0.06828803569078445, -0.2815704941749573, -0.2522399425506592, 0.05219976603984833, 0.04698935151100159, -0.2774737477302551, -0.3826290965080261, -0.3548020124435425, -0.25620725750923157, 0.13442926108837128, -0.2618769705295563, 0.39998874068260193, -0.06487223505973816, -0.05895048752427101, 0.18283894658088684, 0.05480577424168587, 0.20971021056175232, -0.012297696433961391, 0.3575526177883148, 0.3185587525367737, 0.04500001668930054, -0.22733253240585327, 0.15663950145244598, -0.14627164602279663, 0.06553857028484344, -0.2430347353219986, -0.23485775291919708, 0.12916657328605652, -0.311217725276947, 0.00908262375742197, -0.005182185675948858, 0.4701381027698517, 0.03891579806804657, 0.3749934732913971, 0.3503911793231964, 0.2842976748943329, -0.029754530638456345, -0.07869080454111099, -0.06778591126203537, -0.2594401240348816, 0.2883804738521576, 0.16139474511146545, -0.21710924804210663, -0.3110697865486145, -0.14794088900089264, -0.05814843252301216, -0.18362608551979065, -0.21360044181346893, 0.3854408860206604, 0.2447403073310852, -0.3576725721359253, -0.11658520251512527, 0.5851033329963684, 0.058130327612161636, -0.2100522667169571, 0.08590393513441086, -0.09296077489852905, -0.09709055721759796, -0.2978731393814087, 0.0861726626753807, 0.206184521317482, -0.15004146099090576, 0.20931997895240784, -0.10786914080381393, 0.6836856007575989, 0.13581736385822296, 0.18622811138629913, 0.33343347907066345, 0.08329442143440247, -0.16647227108478546, -0.24997039139270782, -0.4027539789676666, 0.25848880410194397, 0.09364014118909836, -0.1719941794872284, 0.06458364427089691, -0.11431381106376648, -0.33401909470558167, -0.2537509500980377, 0.003277276176959276, -0.29372066259384155, -0.13937142491340637, 0.04673641920089722, 0.3141670525074005, 0.08527445793151855, -0.2999880015850067, 0.02187502197921276, -0.29904916882514954, -0.2519984543323517, 0.09707099944353104, 0.29968854784965515, -0.34366220235824585, -0.373259037733078, -0.42583414912223816, 0.11342974007129669, 0.3820706009864807, 0.2269531637430191, 0.1458987593650818, -0.150264173746109, -0.11249702423810959, 0.18048390746116638, -0.32851114869117737, 0.3918130099773407, 0.03860646113753319, -0.18129590153694153, -0.23008044064044952, 0.10167474299669266, -0.286239355802536, -0.21849626302719116, 0.06395970284938812, -0.18907836079597473, -0.12197033315896988, 0.3215974271297455, 0.26926326751708984, -0.1522381752729416, -0.11927387118339539, -0.23923268914222717, -0.2757667303085327, 0.23823872208595276, 0.10660859942436218, 0.38597172498703003, -0.01850191131234169, 0.15808381140232086, 0.519302487373352, 0.01872323267161846, -0.4093891680240631, -0.04577041417360306, -0.1325848251581192, 0.381832480430603, -0.2926695644855499, 0.3908911645412445, -0.32067644596099854, -0.12131321430206299, 0.2370034009218216, -0.22047311067581177, 0.3617454171180725, 0.19992296397686005, 0.11519987136125565, -0.10204402357339859, -0.04495709761977196, 0.3409543037414551, 0.03917735442519188, -0.1111278235912323, -0.14549942314624786, -0.2357196807861328, -0.19972234964370728, -0.20017968118190765, -0.33275482058525085, -0.2919325828552246, -0.054332464933395386, 0.42354896664619446, 0.20417210459709167, -0.03850064426660538, -0.08839354664087296, -0.08542566746473312, 0.4208584129810333, 0.13108499348163605, 0.21697232127189636, -0.1212124153971672, -0.20755894482135773, 0.13657227158546448, -0.2814384996891022, 0.24335330724716187, 0.28780117630958557, 0.34533363580703735, -0.08362200856208801, 0.170485258102417, 0.2630353271961212, -0.12292075157165527, -0.055307671427726746, -0.05954968184232712, 0.2311905026435852, 0.2989141047000885, 0.26676926016807556, 0.25323060154914856, 0.07388067245483398, -0.2816992402076721, 0.05441197007894516, 0.01285086665302515, 0.012367187067866325, 0.4185678958892822, -0.2691672742366791, -0.14407935738563538, -0.19352594017982483, 0.20795509219169617, 0.06539241224527359, 0.3745553195476532, -0.02603316307067871, 0.04296601936221123, 0.3506178557872772, -0.29047876596450806, -0.2164357602596283, -0.12244211882352829, -0.02304805815219879, 0.17220118641853333, 0.10130812972784042, 0.4216386079788208, 0.12463192641735077, -0.10538560897111893, 0.07629338651895523, -0.05406542122364044, 0.114288330078125, -0.19989147782325745, -0.10426609218120575, 0.04114818572998047, 0.08948372304439545, 0.16858124732971191, -0.01906030811369419, 0.06301798671483994, 0.12986329197883606, -0.2475425899028778, 0.3689173460006714, 0.033474236726760864, 0.3673945665359497, -0.1088714599609375, -0.30087125301361084, -0.2557470500469208, -0.1884489506483078, 0.008497416973114014, 0.25391167402267456, 0.09166291356086731, -0.14201843738555908, 0.05402166396379471, 0.4649899899959564, -0.07662482559680939, -0.14190170168876648, -0.029269954189658165, -0.3652302324771881, -0.34538912773132324, 0.009307602420449257, 0.041949305683374405, 0.04209766909480095, 0.09645857661962509, 0.08965097367763519, -0.031336281448602676, -0.23312802612781525, -0.7229450941085815, 0.004387649707496166, -0.006775722373276949, 0.28053513169288635, 0.2111174464225769, -0.08313453197479248, -0.3391124904155731, 0.00408897502347827, 0.052356548607349396, -0.500447690486908, -0.005801419727504253, 0.3323717415332794, -0.35640445351600647, 0.08894766867160797, -0.07864093780517578, -0.10567032545804977, -0.09344395250082016, -0.07754146307706833, -0.07092247158288956, 0.19844041764736176, 0.053707171231508255, -0.058387380093336105, -0.020465265959501266, 0.4024658501148224, 0.508685827255249, -0.04803289845585823, -0.16854079067707062, 0.2404119074344635, 0.34235894680023193, -0.3283787667751312, -0.3037557899951935, -0.028277186676859856, 0.06588014215230942, 0.029017774388194084, 0.005568660795688629, -0.3014433979988098, -0.4165250360965729, -0.1035078912973404, -0.075280100107193, 0.0514252632856369, 0.20570355653762817, 0.024882957339286804, 0.024754101410508156, 0.07889010012149811, -0.2982485592365265, -0.2002650797367096, -0.019337177276611328, 0.14835584163665771, 0.7090578079223633, -0.1582210212945938, -0.17756029963493347, 0.14850440621376038, 0.8652079701423645, 0.01294041145592928, -0.45980194211006165, 0.10806550085544586, 0.17440685629844666, 0.09269359707832336, 0.015705714002251625, -0.21939314901828766, 0.26923003792762756, -0.30339276790618896, -0.17277422547340393, 0.30442357063293457, 0.10612737387418747, -0.12558642029762268, -0.27443015575408936, -0.12407680600881577, 0.5246574878692627, -0.13078713417053223, 0.04327637329697609, 0.08086695522069931, 0.027793539687991142, -0.05913833528757095, -0.1332899034023285, -0.335997611284256, -0.1166520044207573, 0.2886207699775696, 0.08492159843444824, -0.016900058835744858, 0.04680139943957329, 0.46229124069213867, -0.11969231814146042, -0.5215778350830078, 0.477922648191452, 0.1486179381608963, -0.6269398927688599, 0.030343206599354744, -0.09077299386262894, -0.01497755479067564, -0.0043393271043896675, -0.30573007464408875, 0.08298984915018082, 0.022055432200431824, -0.02300391159951687, 0.04754522070288658, -0.263955295085907, 0.11107385158538818, -0.1353965401649475, -0.32031112909317017, 0.47493723034858704, 0.28359052538871765, -0.357986181974411, 0.0602521076798439, 0.2661556899547577, 0.3918268084526062, -0.16820651292800903, -0.044259704649448395, -0.3871450424194336, -0.2475210279226303, 0.05117447301745415, -0.16052789986133575, 0.05983024835586548, 0.2689397931098938, 0.06398992240428925, 0.15757085382938385, 0.14858362078666687, -0.05238524079322815, 0.15332137048244476, 0.0633864775300026, 0.06765418499708176, -0.03543097525835037, 0.03659825026988983, 0.15093488991260529, -0.053821172565221786, 0.1668815314769745, 0.3117910623550415, -0.4149140417575836, -0.49683678150177, 0.13977213203907013, 0.2661862373352051, 0.2174074351787567, 0.03490893170237541, 0.0777490958571434, -0.10712718963623047, 0.3252788782119751, 0.3136734962463379, 0.1727413684129715, 0.3225982189178467, 0.15361590683460236, 0.3155539929866791, 0.11419924348592758, -0.18386071920394897, 0.3391719460487366, -0.11920898407697678, 0.12232588976621628, 0.34898823499679565, 0.02994696982204914, -0.02753542736172676, 0.06459633260965347, 0.537132203578949, 0.9507373571395874, 0.07198940217494965, 0.10562899708747864, 0.34539180994033813, -0.2159128189086914, 0.3318790793418884, -0.1346592754125595, -0.05477147176861763, -0.20902931690216064, -0.2768647372722626, -0.0287370253354311, -0.009400887414813042, -0.416841059923172, -0.30249518156051636, -0.24513211846351624, 0.17041349411010742, -0.372273325920105, -0.2893790602684021, 0.2739179730415344, 0.2895250916481018, -0.01275047566741705, 0.02680167555809021, 0.2843710780143738, 0.07212066650390625, -0.24628672003746033, 0.3137531578540802, -0.0689193606376648, -0.13711783289909363, -0.12174919247627258, 0.11128251999616623, -0.3442816138267517, 0.1823110282421112, -0.14361518621444702, -0.04318654537200928, -0.13794083893299103, -0.2727838456630707, 0.04846414923667908, 0.04304115101695061, 0.7600719332695007, 0.019665652886033058, -0.3796805739402771, 0.22250501811504364, -0.06933462619781494, 0.18749545514583588, 0.12327995896339417, -0.11628541350364685, 0.36192786693573, -0.21870329976081848, -0.0835881233215332, -0.1212909147143364, 0.23607268929481506, -0.4464300572872162, 0.1687420904636383, -0.15053173899650574, -0.1622355729341507, 0.011514401063323021, 0.24560821056365967, 0.236531600356102, 0.11821109056472778, -0.10112278908491135, 0.058677975088357925, 0.14740242063999176, -0.23920926451683044, 0.03271002694964409, 0.04361787810921669, -0.31197062134742737, 0.02909916825592518, -0.041437868028879166, -0.07832878082990646, -0.009219412691891193, 0.22514605522155762, 0.18119817972183228, -0.08502614498138428, 0.23313811421394348, -0.5456887483596802, 0.4800317585468292, -0.39778536558151245, 0.16316461563110352, 0.08292625844478607, -0.033607326447963715, 0.0867537260055542, 0.7250292897224426, 0.44195863604545593, 0.13836199045181274, -0.4551561176776886, 0.10337227582931519, -0.3853459060192108, 0.05547995865345001, -0.18755154311656952, 0.3016722798347473, -0.21308256685733795, 0.1631283164024353, -0.05166437849402428, 0.19557633996009827, -0.1942232847213745, -0.019207892939448357, 0.12870492041110992, -0.13616466522216797, 0.05789302662014961, -0.3698287904262543, 0.18015708029270172, 0.07099298387765884, -0.2820512056350708, -0.12980441749095917, 0.07917610555887222, -0.0011406919220462441, -0.4950151741504669, 0.16216827929019928, 0.23005983233451843, -0.015350167639553547, -0.3231396973133087, -0.44553878903388977, -0.0574517697095871, -0.10607440024614334, 0.021053370088338852, 0.1417103111743927, 0.044789817184209824, 0.3292350172996521, -0.059124160557985306, -0.047191277146339417, -0.10356040298938751, -0.05455079302191734, 0.46705272793769836, 0.22028090059757233, 0.11532850563526154, 0.012021391652524471, 0.3437567353248596, 0.0014599289279431105, 0.13209125399589539, 0.029760928824543953, 0.006227472331374884, 0.060461387038230896, 0.2978818416595459, 0.14499564468860626, 0.19285285472869873, -0.36993908882141113, 0.22901393473148346, 0.1482079178094864, 0.12221937626600266, -0.16198109090328217, 0.17190240323543549, 0.03464831784367561, -0.2363877147436142, -0.20888008177280426, -0.3177421987056732, 0.08207522332668304, -0.1735856831073761, 0.13484255969524384, -0.280415803194046, 0.4191931188106537, 0.31389814615249634, -0.09825554490089417, -0.24227780103683472, -0.2245110720396042, 0.06813595443964005, 0.16164100170135498, -0.187010258436203, 0.3133469521999359, 0.40764638781547546, 0.11467580497264862, 0.20449280738830566, -0.40119579434394836, -0.14249010384082794, 0.0656728446483612, -0.42258650064468384, -0.0895749107003212, 0.32271742820739746, 0.1163296028971672, -0.053208041936159134, -0.13574326038360596, 0.1743607223033905, -0.07144583761692047, -0.19191360473632812, 0.3954492211341858, -0.31552600860595703, -0.21161691844463348, -0.08746606111526489, 0.35621926188468933, 0.1253102719783783, 0.09618904441595078, 0.10229542851448059, -0.08662339299917221, 0.20117172598838806, -0.055316824465990067, 0.03321731463074684, 0.15262722969055176, -0.1583860069513321, 0.12446359544992447, -0.04102521762251854, 0.26512816548347473, -0.7464032769203186, 0.29261288046836853, 0.28785282373428345, 0.14978185296058655, -0.21726618707180023, 0.11766471713781357, -0.013434738852083683, -0.053702212870121, 0.31686994433403015, 0.032993827015161514, 0.19358162581920624, -0.3775281608104706, -0.03685900568962097, 0.2186945676803589, 0.0019625150598585606, 0.1600337028503418, 0.16048282384872437, 0.49909549951553345, -0.18967556953430176, 0.05753696337342262, -0.10264880955219269, -0.14954733848571777, 0.41688287258148193, -0.39484134316444397, 0.1670207679271698, -0.06114399805665016, 0.8726223111152649, 0.2835286259651184, -0.15029390156269073, -0.269493043422699, -0.0804477408528328, -0.21827693283557892, 0.08427640795707703, 0.08718106150627136, -0.120140440762043, 0.07406143099069595, -0.05157263204455376, 0.006864543538540602, 0.18708549439907074, 0.1844523698091507, 0.24675054848194122, 0.3651886284351349, -0.2760647237300873, 0.31659793853759766, -0.3954797685146332, 0.15575818717479706, 0.17428447306156158, 0.06074608862400055, -0.048489369451999664, -0.14108151197433472, 0.023485952988266945, -0.11504442989826202, 0.264630526304245, -0.22524188458919525, 0.051122937351465225, -0.18277773261070251, -0.23931065201759338, -0.3419877886772156, -0.050036557018756866, 0.03890248388051987, -0.03173678740859032, -0.12461566179990768, -0.06375419348478317, -0.0963197723031044, -0.12807005643844604, -0.050185561180114746, -0.29143330454826355, 0.39763709902763367, 0.3708542585372925, 0.1504565328359604, -0.07252904772758484, 0.7684550285339355, 0.20320607721805573, 0.006716940551996231, 0.30119824409484863, 0.20140670239925385, -0.11478050798177719, -0.20222161710262299, -0.037507716566324234, 0.19382300972938538, -0.2518007457256317, -0.2619611322879791, -0.4261302053928375, 0.6186966300010681, 0.10939055681228638, -0.4776005446910858, -0.0888892114162445, -0.2265511155128479, -0.22605812549591064, -0.13082438707351685, 0.2619840204715729, 0.42206698656082153, 0.07908151298761368, 0.03222950920462608, -0.39357444643974304, 0.07538663595914841, 0.5604332089424133, -0.42825859785079956, -0.19135087728500366, -0.012697042897343636, 0.18095433712005615, 0.23919162154197693, -0.3084864616394043, -0.7673505544662476, -0.12440454214811325, 0.11889923363924026, 0.168977752327919, -0.16665223240852356, 0.23522734642028809, 0.132450133562088, 0.2679497003555298, -0.22290942072868347, 0.5535245537757874, -0.07835773378610611, -0.1384921669960022, -0.4126298427581787, -0.09996570646762848 ]
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
[ 0.0396035872399807, -0.2945489287376404, 0.06959506869316101, 0.1643863320350647, 0.2727983295917511, -0.281116247177124, 0.3806678354740143, 0.14022691547870636, 0.14548741281032562, 0.16574156284332275, -0.21055983006954193, 0.2934771776199341, -0.4552499055862427, 0.22415995597839355, 0.06959347426891327, -0.10988401621580124, -0.05097775161266327, 0.03715793043375015, 0.1063586100935936, -0.13577412068843842, -0.10876111686229706, 0.16499075293540955, 0.059349652379751205, -0.1525467485189438, -0.29161548614501953, 0.31761497259140015, 0.019248785451054573, 0.02570628561079502, -0.2094435840845108, -0.05774231627583504, 0.29480934143066406, -0.06041409447789192, 0.20583492517471313, 0.794573962688446, -0.0001224943553097546, -0.2677386403083801, 0.1325640082359314, -0.025316204875707626, -0.3406814932823181, -0.2211291342973709, 0.38395193219184875, -0.3060538172721863, 0.08232956379652023, -0.17544670403003693, 0.042788371443748474, -0.03406551480293274, -0.24324071407318115, -0.042655978351831436, 0.18749359250068665, 0.2166583240032196, 0.04224838316440582, 0.23043595254421234, 0.001225490472279489, 0.11258139461278915, 0.1391414850950241, -0.10558633506298065, -0.10437057912349701, 0.5303358435630798, 0.2758401334285736, -0.06240947172045708, -0.19226302206516266, 0.23816817998886108, 0.07783681154251099, 0.10142317414283752, 0.6716700792312622, 0.039060287177562714, -0.05559522286057472, -0.06814541667699814, 0.005765399429947138, 0.3251339793205261, -0.11480871587991714, -0.21514949202537537, -0.30670034885406494, 0.08728213608264923, 0.07104223966598511, -0.2673068046569824, -0.217970609664917, 0.10698831081390381, 0.10014529526233673, -0.26901212334632874, -0.41902419924736023, -0.3648318350315094, -0.19585184752941132, 0.07973774522542953, -0.27265122532844543, 0.3813117444515228, 0.009816629812121391, -0.15502573549747467, 0.1310458481311798, 0.14506621658802032, 0.3880736231803894, -0.07437721639871597, 0.3471340537071228, 0.27463990449905396, -0.016506904736161232, -0.3021690547466278, 0.2685924768447876, -0.13609124720096588, -0.005038934294134378, -0.2247716784477234, -0.16654247045516968, 0.20244863629341125, -0.26308169960975647, 0.07759600132703781, -0.0429927296936512, 0.40149927139282227, -0.00023947870067786425, 0.3728131949901581, 0.32136309146881104, 0.40310487151145935, 0.10458310693502426, -0.01962621510028839, -0.0739622637629509, -0.08638160675764084, 0.2585075795650482, 0.19309963285923004, -0.22396449744701385, -0.25219640135765076, -0.06776179373264313, -0.020372504368424416, -0.22098703682422638, -0.2402275651693344, 0.333380788564682, 0.19345824420452118, -0.4989130198955536, -0.023181911557912827, 0.500806450843811, 0.006021158304065466, -0.16556988656520844, 0.10502757877111435, -0.14187148213386536, -0.12604540586471558, -0.3094724416732788, 0.07590869069099426, 0.2528242766857147, -0.16018134355545044, 0.2812947630882263, -0.07687195390462875, 0.4981261193752289, 0.047503240406513214, 0.30355262756347656, 0.3707238435745239, 0.09624810516834259, -0.14412443339824677, -0.2020062506198883, -0.3485705852508545, 0.18558330833911896, 0.05072668194770813, -0.1315111517906189, -0.030971704050898552, -0.11490458250045776, -0.27944114804267883, -0.07502087950706482, 0.05097329244017601, -0.2576744258403778, -0.0576423779129982, 0.02539871074259281, 0.3534553349018097, 0.0978761687874794, -0.24719765782356262, 0.0006358660175465047, -0.2485276162624359, -0.16766895353794098, 0.06729400157928467, 0.39564624428749084, -0.18060393631458282, -0.31701815128326416, -0.5069207549095154, 0.1681724488735199, 0.4395362436771393, 0.18799079954624176, 0.1661761999130249, -0.24754561483860016, -0.03294210880994797, 0.2556147277355194, -0.34265580773353577, 0.2737848460674286, 0.1172076016664505, -0.30567196011543274, -0.23019126057624817, 0.08945833891630173, -0.3044487237930298, -0.21890732645988464, 0.08569762855768204, -0.14614693820476532, -0.06445048749446869, 0.2756138741970062, 0.2686769962310791, -0.17650923132896423, -0.012937339954078197, -0.2883862257003784, -0.2562452256679535, 0.2462272047996521, 0.17889513075351715, 0.37698817253112793, -0.03790093958377838, 0.1523507535457611, 0.5387816429138184, 0.06697940826416016, -0.3838602304458618, -0.030025245621800423, -0.1570001244544983, 0.5144122838973999, -0.46003031730651855, 0.3816399872303009, -0.21484921872615814, -0.11701159179210663, 0.17491988837718964, -0.3829174339771271, 0.3466677963733673, 0.15802472829818726, 0.05973023176193237, -0.08232434839010239, 0.004817782901227474, 0.3204120695590973, -0.04485797882080078, -0.07983558624982834, -0.19225353002548218, -0.25190114974975586, -0.15173840522766113, -0.23149128258228302, -0.2959873378276825, -0.18502190709114075, -0.040659550577402115, 0.2206055074930191, 0.36955735087394714, -0.021372586488723755, -0.14678053557872772, -0.08496764302253723, 0.36801233887672424, 0.12806807458400726, 0.16879601776599884, -0.03653530403971672, -0.20773835480213165, 0.16967612504959106, -0.18458296358585358, 0.258116215467453, 0.27523720264434814, 0.40712812542915344, 0.02619829587638378, 0.28724661469459534, 0.2546696960926056, -0.10023247450590134, -0.10496462136507034, -0.20701201260089874, 0.29886433482170105, 0.307541161775589, 0.23850318789482117, 0.2911815643310547, 0.0936935767531395, -0.2713990807533264, 0.0976061001420021, -0.06041191890835762, 0.1711326539516449, 0.39449188113212585, -0.13852427899837494, -0.0819869339466095, -0.16408881545066833, 0.1472579836845398, 0.14154712855815887, 0.36072495579719543, -0.052627649158239365, 0.017791729420423508, 0.23726026713848114, -0.41456809639930725, -0.28494343161582947, -0.08580286800861359, 0.0554221086204052, 0.142669215798378, 0.06415864825248718, 0.41493287682533264, 0.05140436068177223, -0.0188014917075634, 0.044747866690158844, -0.1211605817079544, 0.15478213131427765, -0.23463353514671326, -0.04626797139644623, 0.08961518108844757, 0.09873827546834946, 0.09015549719333649, 0.02514052577316761, 0.07456102222204208, 0.11431244015693665, -0.14357633888721466, 0.35184118151664734, 0.13265907764434814, 0.42570605874061584, -0.23266738653182983, -0.364344984292984, -0.25728514790534973, -0.21227845549583435, 0.03864741325378418, 0.3008294403553009, 0.07326650619506836, -0.0855230987071991, 0.2842051386833191, 0.5591573119163513, -0.049217674881219864, -0.10149630904197693, -0.06357827037572861, -0.29799947142601013, -0.2619439363479614, 0.01304942648857832, -0.04707183688879013, 0.021044062450528145, -0.012131846509873867, 0.1273248940706253, -0.07290883362293243, -0.1648215651512146, -0.7918720245361328, -0.0071388958021998405, 0.06195409595966339, 0.2949562668800354, 0.12773065268993378, -0.13784657418727875, -0.3270552158355713, -0.0012705267872661352, 0.0172598734498024, -0.40893101692199707, -0.023082295432686806, 0.3334006667137146, -0.2403264045715332, 0.020099619403481483, -0.03319770470261574, -0.19501307606697083, -0.12176147103309631, -0.03714713826775551, 0.03505808860063553, 0.17747154831886292, 0.037284594029188156, -0.12759685516357422, 0.11328962445259094, 0.3214522898197174, 0.4272088408470154, -0.15338578820228577, -0.038435958325862885, 0.17423082888126373, 0.4179563820362091, -0.2875511348247528, -0.3364487886428833, -0.11314099282026291, -0.042940691113471985, 0.09721875190734863, 0.020036116242408752, -0.33385929465293884, -0.3707353174686432, -0.13501006364822388, -0.15471304953098297, 0.167624369263649, 0.2113264799118042, 0.08064769208431244, 0.13571645319461823, 0.08123935014009476, -0.16434068977832794, -0.11949415504932404, -0.05121040344238281, 0.16528785228729248, 0.6772114038467407, -0.22687917947769165, -0.13615933060646057, 0.0998496487736702, 0.8651846051216125, 0.1475522667169571, -0.4560464322566986, 0.03625171631574631, 0.31710225343704224, 0.025822501629590988, -0.030955845490098, -0.29106995463371277, 0.1881336122751236, -0.2465847134590149, -0.2843209207057953, 0.28277838230133057, 0.10243093222379684, -0.29742857813835144, -0.303190678358078, -0.20264634490013123, 0.5068924427032471, -0.122345931828022, 0.073630191385746, 0.11307641118764877, 0.0693340003490448, -0.004520020913332701, -0.06540974974632263, -0.293837308883667, -0.12536653876304626, 0.24079279601573944, -0.007609843276441097, 0.0948038101196289, -0.03745770826935768, 0.44763633608818054, 0.0032576473895460367, -0.5484259724617004, 0.348766028881073, 0.05293456092476845, -0.5662439465522766, 0.1476331353187561, -0.07294177263975143, -0.03930995613336563, -0.046537838876247406, -0.3226350247859955, 0.07470078766345978, 0.036167729645967484, -0.14837059378623962, -0.06120389327406883, -0.25290024280548096, 0.09232489764690399, -0.2258274406194687, -0.473783016204834, 0.5314685106277466, 0.33735010027885437, -0.38272035121917725, 0.09321525692939758, 0.22271127998828888, 0.3154436945915222, -0.20287935435771942, -0.01469367928802967, -0.29760923981666565, -0.22090910375118256, 0.07640592008829117, -0.07465627789497375, -0.02643394097685814, 0.22570647299289703, -0.05120163410902023, -0.02730267122387886, 0.12710948288440704, -0.10163190960884094, 0.18122631311416626, 0.07489554584026337, 0.008387375622987747, -0.02671448513865471, 0.15502214431762695, 0.11128568649291992, -0.0781671404838562, 0.19248129427433014, 0.3351227045059204, -0.4193524122238159, -0.5140179991722107, 0.1123957484960556, 0.15022079646587372, 0.17336134612560272, 0.09620566666126251, 0.09976077824831009, -0.010332637466490269, 0.40235766768455505, 0.28301727771759033, 0.16237834095954895, 0.2686346471309662, 0.13129712641239166, 0.4590049684047699, 0.06577197462320328, -0.14670030772686005, 0.3369625210762024, -0.18430645763874054, 0.029224559664726257, 0.30647632479667664, -0.09339824318885803, -0.026042697951197624, 0.17510439455509186, 0.45825284719467163, 0.9547624588012695, 0.11388455331325531, 0.07662217319011688, 0.29241108894348145, -0.12001810222864151, 0.33727169036865234, -0.2470247596502304, 0.05289619415998459, -0.3119497001171112, -0.3774307668209076, -0.037566352635622025, -0.08929379284381866, -0.40337902307510376, -0.2799673080444336, -0.2035807967185974, 0.11634590476751328, -0.34473076462745667, -0.22656004130840302, 0.3043633699417114, 0.32432469725608826, 0.07537977397441864, 0.021569890901446342, 0.28700366616249084, 0.061904534697532654, -0.2093512862920761, 0.3146818280220032, -0.07799986004829407, -0.23030878603458405, -0.2361432909965515, 0.10131248831748962, -0.18740475177764893, 0.26554423570632935, -0.057189684361219406, -0.006722273770719767, 0.010055220685899258, -0.31991901993751526, 0.07066001743078232, 0.06933703273534775, 0.7521312832832336, 0.1196122094988823, -0.37206554412841797, 0.14824484288692474, -0.1703067421913147, 0.13719086349010468, 0.1343541145324707, -0.16840781271457672, 0.3660241961479187, -0.1432255059480667, -0.1668994128704071, -0.08256826549768448, 0.21665291488170624, -0.28432852029800415, 0.16792218387126923, -0.17392027378082275, -0.11232798546552658, -0.07144220918416977, 0.4080495536327362, 0.3213435411453247, 0.07981806993484497, -0.20158801972866058, 0.09625989198684692, 0.2228284329175949, -0.27825021743774414, 0.08793109655380249, 0.13651299476623535, -0.23687343299388885, -0.003059095237404108, 0.021061234176158905, -0.04544353857636452, -0.07609298080205917, 0.2099473923444748, 0.1896342635154724, -0.12245567888021469, 0.20275980234146118, -0.5506263375282288, 0.5572551488876343, -0.4292539954185486, 0.11070045083761215, -0.13716809451580048, -0.022318247705698013, 0.034150443971157074, 0.6142632961273193, 0.46294882893562317, 0.1775754988193512, -0.4233264923095703, 0.12258630245923996, -0.4270443320274353, 0.10487561672925949, -0.18003776669502258, 0.3510756194591522, -0.2739526331424713, -0.006761339958757162, 0.006709706038236618, 0.38980716466903687, -0.22701622545719147, -0.026360513642430305, 0.12607687711715698, -0.07171669602394104, -0.021651502698659897, -0.4830068349838257, 0.1117929071187973, 0.04559274762868881, -0.2318335771560669, -0.10764285176992416, 0.08288013935089111, -0.042953167110681534, -0.46715986728668213, 0.15198083221912384, 0.1879013031721115, 0.05123542621731758, -0.29135313630104065, -0.3445463478565216, -0.0923868715763092, -0.1816118061542511, -0.0012826431775465608, -0.03598837926983833, 0.07129799574613571, 0.3768666088581085, 0.16384753584861755, 0.03225522115826607, -0.20140355825424194, -0.05826187506318092, 0.3841693699359894, 0.11524109542369843, 0.06221020594239235, -0.08224427700042725, 0.3912833034992218, -0.02347264252603054, 0.026852721348404884, 0.050318408757448196, 0.027712631970643997, 0.0872533768415451, 0.42564713954925537, 0.09776279330253601, 0.15738631784915924, -0.41123929619789124, 0.2807863652706146, 0.17748744785785675, 0.11828829348087311, -0.13777071237564087, 0.1800997257232666, 0.10431301593780518, -0.22499199211597443, -0.23325379192829132, -0.19721490144729614, 0.12463856488466263, -0.11268652975559235, 0.1643534004688263, -0.11488211899995804, 0.3551521301269531, 0.36327412724494934, -0.009765919297933578, -0.2480154186487198, -0.30713966488838196, 0.04222582280635834, 0.054887883365154266, -0.18486298620700836, 0.27694791555404663, 0.4547150433063507, 0.06889292597770691, 0.10064514726400375, -0.4465752840042114, -0.09266716241836548, 0.08696968853473663, -0.38432323932647705, -0.08820286393165588, 0.3419310748577118, 0.19567809998989105, -0.15060892701148987, -0.20592065155506134, 0.1579793244600296, -0.1140969842672348, -0.13242536783218384, 0.36024945974349976, -0.24125301837921143, -0.141825333237648, -0.05407191440463066, 0.3252496123313904, 0.02627587504684925, 0.07756403833627701, 0.05787572264671326, -0.10660604387521744, 0.06521818786859512, -0.0064457617700099945, 0.10747765749692917, 0.06735975295305252, -0.012964901514351368, 0.12857095897197723, -0.1101795881986618, 0.20703834295272827, -0.6235060095787048, 0.35154879093170166, 0.34066593647003174, 0.0404871441423893, -0.23865219950675964, 0.1311081349849701, -0.06537601351737976, -0.26793190836906433, 0.33320537209510803, -0.07992033660411835, 0.20084735751152039, -0.34345167875289917, 0.08213343471288681, 0.28266212344169617, -0.05414741113781929, 0.06991566717624664, 0.2067139595746994, 0.401702880859375, -0.0886712521314621, -0.051762428134679794, -0.05980342626571655, -0.17698325216770172, 0.38699644804000854, -0.4427589178085327, 0.16624002158641815, -0.09579415619373322, 0.957119345664978, 0.3474620282649994, -0.2492387294769287, -0.24940216541290283, -0.1470511108636856, -0.3462855815887451, 0.10768388211727142, 0.17763425409793854, -0.038340773433446884, -0.003949018660932779, -0.06391064077615738, 0.002883234294131398, 0.05237787961959839, 0.15503710508346558, 0.22364652156829834, 0.5317480564117432, -0.35456880927085876, 0.2993886172771454, -0.22832487523555756, 0.10329069197177887, 0.2751069962978363, 0.13227307796478271, -0.1272495985031128, -0.017683954909443855, -0.009352344088256359, -0.049271050840616226, 0.02665274776518345, -0.32910609245300293, 0.0641801655292511, -0.14669616520404816, -0.26521891355514526, -0.2175392210483551, -0.049030851572752, 0.03659144416451454, 0.04084454104304314, -0.06513195484876633, 0.032620370388031006, -0.12107524275779724, -0.10738211870193481, -0.05225310102105141, -0.31600677967071533, 0.34854206442832947, 0.36672621965408325, 0.0898682102560997, -0.027841800823807716, 0.6427885890007019, 0.1550077348947525, -0.11144867539405823, 0.20141665637493134, 0.14983776211738586, -0.11773519963026047, -0.23597532510757446, 0.022950299084186554, 0.30525535345077515, -0.11438217759132385, -0.3438955247402191, -0.4947918653488159, 0.6355993747711182, 0.17985568940639496, -0.5693953633308411, -0.17015691101551056, -0.24646645784378052, -0.1722554862499237, -0.11563282459974289, 0.30473896861076355, 0.3367747366428375, 0.10700061917304993, -0.017186090350151062, -0.36785197257995605, 0.06726889312267303, 0.4848317801952362, -0.45635727047920227, -0.07332667708396912, -0.11960720270872116, 0.16673529148101807, 0.2647436857223511, -0.1948126256465912, -0.9240174889564514, -0.13740631937980652, 0.14410816133022308, -0.014073066413402557, -0.17872881889343262, 0.29899728298187256, 0.24600151181221008, 0.2937110662460327, -0.2554593086242676, 0.5315840840339661, -0.04953799769282341, -0.0996764674782753, -0.45906880497932434, -0.21204864978790283 ]
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
[ 0.031051943078637123, -0.21039588749408722, 0.11031077057123184, 0.24948249757289886, 0.28502869606018066, -0.3368866443634033, 0.29802533984184265, 0.14864586293697357, 0.09188979864120483, 0.1669934093952179, -0.3092392086982727, 0.3939819037914276, -0.4846564829349518, 0.25464630126953125, 0.04687966778874397, -0.08954013139009476, -0.06799955666065216, 0.05894642695784569, 0.10282021015882492, -0.17925669252872467, -0.11254620552062988, 0.15591709315776825, 0.11106303334236145, -0.10564200580120087, -0.19486534595489502, 0.3722076714038849, -0.066965252161026, 0.11051201075315475, -0.23483684659004211, -0.0558181069791317, 0.23994286358356476, -0.09313483536243439, 0.19716273248195648, 0.740787923336029, -0.00012301238893996924, -0.2967880070209503, 0.10562039166688919, -0.013818134553730488, -0.3956645131111145, -0.22906650602817535, 0.3934166431427002, -0.3064045310020447, 0.14753971993923187, -0.18821853399276733, -0.023235997185111046, -0.046829596161842346, -0.2829231023788452, -0.11125514656305313, 0.13092012703418732, 0.24891822040081024, 0.03568286821246147, 0.12069819122552872, -0.040036436170339584, 0.10717982798814774, 0.17364291846752167, -0.19217020273208618, -0.025240518152713776, 0.5683522820472717, 0.33517521619796753, -0.019421154633164406, -0.22494246065616608, 0.17774049937725067, 0.011492983438074589, 0.10073015093803406, 0.6617957353591919, 0.040606193244457245, -0.05076683312654495, -0.0463242307305336, 0.0426529198884964, 0.33016136288642883, -0.11374515295028687, -0.19089485704898834, -0.25507643818855286, 0.12239792197942734, 0.07248964160680771, -0.24157080054283142, -0.23092660307884216, 0.160721093416214, 0.10856509208679199, -0.24158060550689697, -0.3910420536994934, -0.3491133749485016, -0.1409151256084442, 0.13794498145580292, -0.2982495427131653, 0.3092897832393646, -0.06806904822587967, -0.12711909413337708, 0.15347883105278015, 0.08535538613796234, 0.34188732504844666, -0.09555453062057495, 0.3543400764465332, 0.19703887403011322, 0.08634345978498459, -0.18238548934459686, 0.22257326543331146, 0.042103834450244904, -0.06244343891739845, -0.3093152642250061, -0.12535391747951508, 0.2069181203842163, -0.28140294551849365, 0.011797353625297546, -0.027307026088237762, 0.3203352391719818, 0.040419381111860275, 0.35977378487586975, 0.28627580404281616, 0.3707205355167389, 0.07752574980258942, -0.04536353051662445, -0.08907117694616318, -0.15487778186798096, 0.2409738153219223, 0.19222857058048248, -0.2800595164299011, -0.32870179414749146, -0.03376989811658859, -0.054196979850530624, -0.217697411775589, -0.10590725392103195, 0.3486950993537903, 0.13636158406734467, -0.49230241775512695, -0.06439295411109924, 0.6397175788879395, 0.04687049612402916, -0.18193812668323517, 0.0798485204577446, -0.14149247109889984, -0.09508397430181503, -0.307386189699173, 0.067583367228508, 0.21250420808792114, -0.19385340809822083, 0.27882277965545654, 0.03200084716081619, 0.5187742710113525, 0.1252877414226532, 0.1617126613855362, 0.4335991442203522, 0.16154727339744568, -0.19240933656692505, -0.24166598916053772, -0.29619359970092773, 0.2262890487909317, 0.026455840095877647, -0.11240115016698837, 0.08938474953174591, -0.12872354686260223, -0.3263464570045471, -0.15209676325321198, 0.023959895595908165, -0.2871280908584595, -0.01918993517756462, -0.01095444057136774, 0.2780379056930542, 0.10987893491983414, -0.2487732321023941, -0.032197680324316025, -0.39510127902030945, -0.12442659586668015, 0.09461557865142822, 0.40959540009498596, -0.20389004051685333, -0.40325552225112915, -0.42737653851509094, 0.1044919565320015, 0.4528105556964874, 0.3346209228038788, 0.09447573870420456, -0.2226802110671997, -0.20602333545684814, 0.1822284609079361, -0.3569920063018799, 0.24966318905353546, 0.19206231832504272, -0.2746538817882538, -0.2299160659313202, 0.13340477645397186, -0.2530931532382965, -0.21538212895393372, 0.025944756343960762, -0.1508408635854721, -0.06364262849092484, 0.3645898103713989, 0.11044366657733917, -0.18229268491268158, -0.02314537949860096, -0.37692463397979736, -0.17071153223514557, 0.24087780714035034, 0.1452450454235077, 0.3112165629863739, 0.04421837627887726, 0.1397859752178192, 0.5452293157577515, 0.10949310660362244, -0.3767815828323364, 0.038649871945381165, -0.20870758593082428, 0.4394533932209015, -0.3888005316257477, 0.36291465163230896, -0.13229472935199738, -0.04362035170197487, 0.22862321138381958, -0.32526811957359314, 0.4498295783996582, 0.15172047913074493, 0.1373928040266037, -0.11641087383031845, -0.11465421319007874, 0.2857361137866974, -0.04155861586332321, -0.09858451038599014, -0.17538587749004364, -0.28409698605537415, -0.2204529494047165, -0.14794102311134338, -0.25469687581062317, -0.23914535343647003, -0.05173951014876366, 0.1921083778142929, 0.3260837495326996, -0.06349755823612213, -0.16392400860786438, -0.0386807806789875, 0.38624006509780884, 0.18508106470108032, 0.21768546104431152, -0.07578510791063309, -0.22854900360107422, 0.11393874138593674, -0.20823435485363007, 0.14065782725811005, 0.19605548679828644, 0.3851839601993561, 0.03162381425499916, 0.4031253159046173, 0.21448065340518951, -0.11146416515111923, -0.0912846103310585, -0.12201741337776184, 0.28122851252555847, 0.3246663212776184, 0.26691147685050964, 0.2754969000816345, 0.1210632398724556, -0.24934864044189453, 0.10694237053394318, -0.045374516397714615, 0.10155405849218369, 0.351559579372406, -0.16831733286380768, 0.06461978703737259, -0.16824567317962646, 0.16156993806362152, 0.29566341638565063, 0.2568989396095276, -0.05976753309369087, -0.04502646625041962, 0.2192200869321823, -0.46931278705596924, -0.272273987531662, 0.011140033602714539, 0.07040096819400787, 0.08245287835597992, 0.0731721892952919, 0.46005934476852417, 0.14472335577011108, -0.01683213748037815, 0.0277311559766531, -0.08601680397987366, 0.1478746086359024, -0.11266057938337326, -0.06314833462238312, 0.1250303089618683, 0.06865333020687103, 0.07814249396324158, 0.1039392501115799, 0.055234476923942566, 0.1424945890903473, -0.1808573454618454, 0.33856436610221863, -0.006266979034990072, 0.43459823727607727, -0.24935466051101685, -0.3888391852378845, -0.2261941134929657, -0.2483251541852951, -0.03489147871732712, 0.3237878680229187, -0.013397648930549622, -0.1362864077091217, 0.19925042986869812, 0.4045136868953705, -0.08564760535955429, -0.05584074184298515, -0.11347294598817825, -0.2637966275215149, -0.3660843074321747, -0.001267404411919415, -0.06414412707090378, 0.06512785702943802, -0.08705271780490875, 0.035879261791706085, -0.15460720658302307, -0.19300717115402222, -0.7795091867446899, 0.060213442891836166, 0.03927214443683624, 0.30001887679100037, 0.20597092807292938, -0.13490094244480133, -0.31092119216918945, 0.04444657266139984, 0.10181494802236557, -0.37420204281806946, -0.026990776881575584, 0.41768160462379456, -0.23497267067432404, 0.07670968770980835, -0.033229146152734756, -0.22106099128723145, -0.128867506980896, -0.02256789058446884, -0.055179256945848465, 0.12109815329313278, 0.0652792751789093, -0.20432518422603607, 0.13932454586029053, 0.3219203054904938, 0.4152192771434784, -0.09877269715070724, -0.06713995337486267, 0.22492273151874542, 0.4278005063533783, -0.25694629549980164, -0.3013521432876587, -0.1849202811717987, -0.00057788664707914, 0.017254387959837914, 0.00037643281393684447, -0.3237825334072113, -0.36015430092811584, -0.11460576206445694, -0.1512560248374939, 0.20219996571540833, 0.16915789246559143, 0.061578940600156784, 0.2242959886789322, 0.06467756628990173, -0.21677865087985992, -0.1627926379442215, -0.07023164629936218, 0.2430410236120224, 0.6987828612327576, -0.2315383404493332, -0.09995251893997192, 0.08970417827367783, 0.7524689435958862, 0.130575031042099, -0.4231867790222168, 0.06254180520772934, 0.260733425617218, 0.06086153909564018, -0.1139407530426979, -0.2671458423137665, 0.15565058588981628, -0.31453976035118103, -0.27640238404273987, 0.3156626522541046, 0.1146005243062973, -0.28042179346084595, -0.2611306607723236, -0.20049725472927094, 0.49796155095100403, -0.17563261091709137, 0.06963610649108887, 0.034685488790273666, -0.04997827485203743, 0.022821960970759392, -0.117080919444561, -0.2740621864795685, -0.1145453155040741, 0.20600098371505737, 0.022086935117840767, 0.0904163271188736, 0.036680519580841064, 0.45275938510894775, 0.01557352114468813, -0.7020946145057678, 0.4291398227214813, 0.09122669696807861, -0.6415271162986755, 0.20079074800014496, -0.15091966092586517, -0.08306712657213211, -0.053233593702316284, -0.36331906914711, 0.11488115042448044, 0.036584556102752686, -0.06340626627206802, -0.024901598691940308, -0.2808915078639984, 0.01993608847260475, -0.2606958746910095, -0.31099599599838257, 0.610831618309021, 0.3093990087509155, -0.3783528208732605, 0.05132340267300606, 0.23827791213989258, 0.44776082038879395, -0.16089579463005066, 0.03582229092717171, -0.3568879961967468, -0.21575215458869934, 0.09272737056016922, -0.11704529076814651, -0.03972282260656357, 0.2884736657142639, -0.12854881584644318, -0.07872049510478973, 0.2440057396888733, -0.08774784952402115, 0.19971905648708344, 0.04970068857073784, 0.10960303992033005, -0.09483002871274948, 0.17061886191368103, 0.05049654841423035, 0.06471309065818787, 0.2465963363647461, 0.41444966197013855, -0.5077789425849915, -0.4713492691516876, 0.2168162316083908, 0.16121579706668854, 0.17137642204761505, 0.10935492068529129, 0.10491935163736343, -0.1360112726688385, 0.43549802899360657, 0.272286593914032, 0.1856849640607834, 0.23380301892757416, 0.03917355090379715, 0.3682485520839691, 0.11878351867198944, -0.19488808512687683, 0.3924100399017334, -0.17067420482635498, 0.01442062295973301, 0.2270653396844864, 0.06649202108383179, -0.012271923944354057, 0.12193993479013443, 0.4338003695011139, 0.9371414184570312, 0.14221999049186707, 0.09444335103034973, 0.2991999685764313, -0.1356024444103241, 0.32738181948661804, -0.04595339298248291, 0.028853600844740868, -0.341989666223526, -0.2665644586086273, 0.043637387454509735, -0.049589820206165314, -0.44529542326927185, -0.16226738691329956, -0.19859671592712402, 0.14732979238033295, -0.3593902587890625, -0.25683721899986267, 0.28275713324546814, 0.2808369994163513, 0.10884816199541092, 0.030932080000638962, 0.27702629566192627, 0.08055364340543747, -0.1751105785369873, 0.2692444920539856, -0.08487454801797867, -0.18439915776252747, -0.26210474967956543, 0.13207878172397614, -0.18024908006191254, 0.3719135522842407, -0.10059577226638794, -0.04485533386468887, -0.06090130656957626, -0.34924784302711487, 0.07375746965408325, 0.032763514667749405, 0.7054666876792908, 0.09231698513031006, -0.3070914149284363, 0.08027984201908112, -0.062137726694345474, 0.17744185030460358, 0.1961665004491806, -0.1495552510023117, 0.3595605790615082, -0.1338011473417282, -0.16224025189876556, -0.16799667477607727, 0.14697037637233734, -0.3749425709247589, 0.2212456911802292, -0.1645987629890442, -0.045548681169748306, -0.11824869364500046, 0.40338510274887085, 0.33453622460365295, 0.15933924913406372, -0.25532108545303345, 0.07714975625276566, 0.20327968895435333, -0.26020246744155884, 0.04581991583108902, 0.11606162786483765, -0.23642198741436005, 0.040236227214336395, -0.009904127568006516, 0.021524865180253983, -0.0013381626922637224, 0.2349434792995453, 0.1742057204246521, -0.11456702649593353, 0.2359018623828888, -0.6396331787109375, 0.5488188862800598, -0.36841291189193726, 0.17571435868740082, -0.13607355952262878, -0.05517900735139847, 0.1222449317574501, 0.5733035802841187, 0.4357829988002777, 0.21459414064884186, -0.3986990451812744, 0.18546858429908752, -0.3889482319355011, 0.08772923052310944, -0.1864103376865387, 0.36033210158348083, -0.17637737095355988, 0.0029720463789999485, -0.1169334426522255, 0.2312013804912567, -0.20673416554927826, 0.03993795067071915, 0.19829189777374268, -0.030683914199471474, 0.009558908641338348, -0.44185376167297363, 0.02748606726527214, 0.035833414644002914, -0.23903542757034302, -0.06395754963159561, 0.11508995294570923, -0.022896090522408485, -0.43679139018058777, 0.15885204076766968, 0.23203137516975403, 0.009140901267528534, -0.3205917775630951, -0.3893791139125824, -0.13043992221355438, -0.18962839245796204, 0.06611350178718567, 0.027370113879442215, 0.06494367867708206, 0.21625950932502747, 0.20794491469860077, 0.02030588872730732, -0.16019946336746216, -0.04887622967362404, 0.41021987795829773, 0.16647982597351074, 0.009113364852964878, -0.03434068337082863, 0.42611536383628845, -0.02806236781179905, 0.19161000847816467, 0.024117473512887955, -0.025749124586582184, 0.16407015919685364, 0.41582247614860535, 0.13150374591350555, 0.15301193296909332, -0.39151984453201294, 0.3136754035949707, 0.25202059745788574, 0.20041455328464508, -0.20123547315597534, 0.21131369471549988, 0.08531982451677322, -0.24712762236595154, -0.17805717885494232, -0.20906847715377808, 0.05017305165529251, -0.1371205449104309, 0.17602622509002686, -0.1112411692738533, 0.36048266291618347, 0.38274678587913513, -0.06124007701873779, -0.23618614673614502, -0.2675482928752899, 0.050848402082920074, 0.04483162984251976, -0.2170991599559784, 0.31279847025871277, 0.5844204425811768, 0.04400445148348808, 0.05221611633896828, -0.38256514072418213, -0.11984270811080933, 0.0084651755169034, -0.2900795042514801, -0.10884638875722885, 0.35886478424072266, 0.20460042357444763, -0.10920129716396332, -0.0990251749753952, 0.15128178894519806, -0.07170303165912628, -0.1213434711098671, 0.39471161365509033, -0.40146952867507935, -0.16225996613502502, -0.03691478818655014, 0.3568078875541687, -0.026629630476236343, 0.06826355308294296, 0.10880731791257858, -0.21613861620426178, 0.07733306288719177, -0.015025364235043526, -0.018274657428264618, 0.09222204983234406, 0.07474473863840103, 0.10134068876504898, -0.09447605162858963, 0.1506626009941101, -0.7600218057632446, 0.3500681221485138, 0.2666798532009125, 0.07755821943283081, -0.19197624921798706, 0.1510920524597168, -0.11532806605100632, -0.23615211248397827, 0.38309964537620544, -0.016696056351065636, 0.25774863362312317, -0.3252435028553009, -0.04142884165048599, 0.4066010117530823, -0.02356448955833912, 0.04343340918421745, 0.11910440027713776, 0.4557111859321594, -0.13713513314723969, -0.0801689475774765, -0.08495016396045685, -0.17106112837791443, 0.3347727358341217, -0.39211300015449524, 0.22192469239234924, 0.05730530247092247, 0.8864076733589172, 0.3305409550666809, -0.20438869297504425, -0.2726922333240509, -0.09083286672830582, -0.29800698161125183, 0.046754319220781326, 0.10341706871986389, -0.08217348903417587, 0.027586013078689575, -0.02778368815779686, 0.006058186292648315, 0.12412014603614807, 0.1421758383512497, 0.26756617426872253, 0.2982494533061981, -0.3599909245967865, 0.2530735433101654, -0.2842274308204651, 0.1555223912000656, 0.18207071721553802, 0.06622332334518433, -0.2323409765958786, -0.007762580178678036, 0.07117688655853271, -0.10512129217386246, 0.05717304348945618, -0.23002095520496368, 0.07925743609666824, -0.2101910561323166, -0.23410522937774658, -0.2609880566596985, 0.06782663613557816, -0.014428561553359032, 0.004517441149801016, -0.0863475650548935, 0.023151999339461327, -0.11450634151697159, -0.0945780873298645, 0.02124127559363842, -0.28783372044563293, 0.37281301617622375, 0.3003574013710022, 0.12165669351816177, 0.0571562759578228, 0.5850409269332886, 0.12948976457118988, -0.17101655900478363, 0.2642616033554077, 0.13617169857025146, -0.10343566536903381, -0.18146231770515442, -0.0433797612786293, 0.26375114917755127, -0.05089734122157097, -0.2635940909385681, -0.5351039171218872, 0.5506893396377563, 0.11400321125984192, -0.5117973685264587, -0.2289525270462036, -0.16178423166275024, -0.1855894774198532, -0.10991441458463669, 0.24824479222297668, 0.38117244839668274, 0.13884250819683075, -0.028626197949051857, -0.4515787363052368, 0.004064416512846947, 0.4650702476501465, -0.47542059421539307, -0.12336898595094681, -0.12123144418001175, 0.1840684711933136, 0.2990708649158478, -0.12919795513153076, -0.8295279741287231, -0.15432953834533691, 0.0958305075764656, -0.03808797150850296, -0.13884228467941284, 0.23325446248054504, 0.11724262684583664, 0.25402483344078064, -0.29439640045166016, 0.5520851612091064, 0.04639894887804985, -0.1617395132780075, -0.47914397716522217, -0.19192788004875183 ]
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 :)
[ -0.09875010699033737, 0.2994913160800934, -0.19066587090492249, 0.08413106948137283, -0.09124992042779922, 0.1056409627199173, 0.45074713230133057, -0.14683809876441956, -0.21770669519901276, 0.00450033787637949, -0.2079361230134964, -0.17496582865715027, -0.15123465657234192, 0.2744065821170807, 0.07923603802919388, -0.36933499574661255, -0.0019668303430080414, 0.0335368812084198, 0.08688855171203613, -0.34479355812072754, -0.10768919438123703, 0.1783020794391632, -0.04975680634379387, 0.19358953833580017, -0.00017960203695110977, 0.09124086052179337, -0.22928372025489807, -0.1584288328886032, -0.15099920332431793, -0.022320590913295746, 0.13173730671405792, 0.24340614676475525, 0.07313697785139084, 0.3848913609981537, -0.00010290565842296928, -0.21260999143123627, 0.12683208286762238, -0.12616372108459473, 0.029008446261286736, 0.032224565744400024, 0.022003527730703354, -0.22676846385002136, -0.3399677574634552, 0.03778175637125969, -0.12805689871311188, -0.25360584259033203, -0.1875300258398056, -0.5613651871681213, -0.1343904733657837, 0.5067587494850159, 0.24277815222740173, -0.1785590499639511, 0.13484235107898712, 0.014460683800280094, 0.1603565812110901, 0.2705264687538147, -0.3232007622718811, -0.16572847962379456, 0.629185140132904, 0.14990390837192535, 0.03446435555815697, 0.2998957931995392, 0.22463485598564148, -0.3710680305957794, -0.20697104930877686, -0.13152335584163666, -0.05224642902612686, -0.36818763613700867, 0.027923500165343285, 0.3565877676010132, 0.6327369213104248, -0.08152002841234207, -0.5868776440620422, -0.09462013840675354, 0.24371454119682312, -0.0755494013428688, -0.10097456723451614, -0.1344938725233078, 0.2280341386795044, 0.1947890818119049, 0.30443185567855835, -0.4035974144935608, -0.06974370032548904, 0.09901441633701324, 0.08143585920333862, 0.009417956694960594, -0.05463438853621483, -0.01763049140572548, 0.11922620236873627, -0.16042019426822662, 0.010838862508535385, 0.16477160155773163, -0.060169171541929245, 0.3607107996940613, -0.21490657329559326, -0.32067200541496277, 0.06903423368930817, -0.05932866409420967, 0.3171025514602661, 0.07181316614151001, -0.6085307598114014, 0.03394044563174248, -0.47544702887535095, -0.13653278350830078, 0.23075486719608307, 0.14107879996299744, 0.11318842321634293, -0.2693188190460205, 0.16951429843902588, -0.3491484522819519, 0.12838312983512878, -0.008936537429690361, -0.006788275670260191, -0.02576979249715805, -0.28193026781082153, 0.06205953285098076, -0.0937892347574234, -0.3450583517551422, 0.009534739889204502, 0.02847064845263958, 0.19382160902023315, -0.21026396751403809, -0.10387547314167023, 0.19578389823436737, 0.08459504693746567, 0.31471410393714905, -0.1896243542432785, 0.19519759714603424, -0.18190443515777588, -0.6139769554138184, -0.15016721189022064, 0.046688444912433624, -0.2589319348335266, 0.033653706312179565, 0.11798001080751419, -0.3058915138244629, 0.04503726214170456, 0.10600249469280243, -0.03552212193608284, 0.2437448501586914, 0.18960942327976227, -0.33063432574272156, -0.14126469194889069, -0.23433740437030792, -0.2384474128484726, 0.08291608095169067, 0.006255893502384424, 0.2903527617454529, -0.2742964029312134, -0.1327151656150818, 0.0041040461510419846, -0.0777202695608139, -0.5493544936180115, 0.25537705421447754, 0.12873275578022003, -0.2849130630493164, -0.050306715071201324, 0.747109591960907, -0.18387587368488312, -0.2807389497756958, 0.06724195182323456, -0.05907431244850159, -0.09857071191072464, -0.0643858015537262, 0.03518526628613472, 0.3334689140319824, -0.5780577659606934, -0.1407889574766159, -0.21537351608276367, -0.304147332906723, 0.2633829116821289, 0.24018338322639465, -0.013686591759324074, 0.1278274804353714, -0.2747955322265625, -0.029189441353082657, 0.05475681647658348, -0.051042426377534866, -0.007746825460344553, 0.1915835440158844, -0.17146103084087372, 0.2518326938152313, 0.005483168642967939, 0.07924213260412216, -0.015349567867815495, -0.07730838656425476, 0.299966961145401, 0.35437026619911194, -0.26174798607826233, -0.0703958049416542, -0.0361998975276947, -0.42472758889198303, 0.16641631722450256, 0.18734188377857208, 0.07433459907770157, -0.0027771838940680027, 0.14446143805980682, -0.23539049923419952, 0.13004982471466064, -0.06374669820070267, 0.22386789321899414, 0.267032265663147, 0.13562443852424622, 0.5249234437942505, 0.060762837529182434, 0.048925772309303284, -0.159783273935318, -0.015663987025618553, 0.1047908142209053, 0.4020426571369171, 0.31528976559638977, -0.15735812485218048, -0.02523087151348591, -0.12310609221458435, -0.24866142868995667, -0.36729198694229126, 0.24655041098594666, 0.26383593678474426, 0.01991291157901287, 0.33392006158828735, -0.3577462434768677, 0.28669437766075134, -0.22801710665225983, -0.25668102502822876, -0.18812121450901031, 0.26827433705329895, -0.08029618114233017, 0.02375074103474617, -0.1958562135696411, 0.35900285840034485, -0.10002166777849197, -0.06508948653936386, -0.10964339971542358, -0.16664467751979828, 0.2140120565891266, 0.14497852325439453, 0.5165334939956665, 0.299884557723999, 0.25742027163505554, -0.24572432041168213, -0.13876990973949432, 0.18688997626304626, 0.0629616230726242, -0.2090667188167572, 0.1430256962776184, 0.2379070222377777, 0.18251900374889374, 0.051116690039634705, 0.0688580572605133, -0.024304788559675217, 0.3960515856742859, 0.21396589279174805, -0.07813747227191925, -0.14466844499111176, 0.3003721833229065, -0.06700188666582108, 0.28173163533210754, -0.07894668728113174, -0.3979152739048004, -0.021499378606677055, 0.3684980571269989, 0.022610601037740707, 0.11486054956912994, 0.13951663672924042, -0.1527700275182724, -0.10070941597223282, 0.2167852520942688, -0.4813975393772125, 0.12765103578567505, 0.26268675923347473, 0.2217150181531906, -0.1654065102338791, 0.2598194479942322, 0.03460797667503357, 0.04588092118501663, -0.2548021376132965, -0.10957498103380203, -0.039888620376586914, 0.15866932272911072, 0.07674156874418259, -0.3456180691719055, -0.4028260111808777, -0.002588552189990878, 0.08199041336774826, 0.04955485835671425, -0.19892194867134094, 0.07890421897172928, -0.054053664207458496, 0.23634302616119385, -0.3461320996284485, -0.2608320415019989, -0.09982588142156601, 0.24856777489185333, -0.3377683162689209, -0.0064278244972229, 0.04294532537460327, -0.19007036089897156, 0.3239136338233948, -0.03564954176545143, -0.0031525022350251675, -0.24567300081253052, 0.005889873020350933, -0.47467857599258423, 0.2933710813522339, -0.05381492152810097, -0.31592857837677, 0.38174858689308167, -0.07545981556177139, 0.14767485857009888, -0.33012983202934265, -0.5107157230377197, 0.16973018646240234, -0.12999433279037476, 0.5753858685493469, 0.12542322278022766, 0.025765102356672287, 0.013467631302773952, -0.12582725286483765, 0.023961924016475677, -0.054652877151966095, -0.16415315866470337, -0.11448458582162857, -0.08873176574707031, 0.4223625659942627, 0.038697391748428345, -0.45314261317253113, -0.42036640644073486, -0.17231105268001556, 0.2529330551624298, -0.16396492719650269, -0.00300893047824502, 0.22998610138893127, -0.13780371844768524, 0.1382645219564438, 0.009906692430377007, 0.1318792849779129, -0.26472750306129456, 0.027845686301589012, 0.2004287838935852, -0.31068697571754456, -0.27746403217315674, 0.06118796020746231, -0.09069789946079254, 0.08521270006895065, -0.24416667222976685, -0.36258023977279663, 0.25801458954811096, -0.09587215632200241, 0.2738504111766815, -0.19152085483074188, 0.07490342855453491, 0.19146399199962616, 0.16966590285301208, -0.0674571841955185, 0.11042961478233337, 0.007872006855905056, -0.13218480348587036, 0.010557425208389759, 0.08699038624763489, -0.10921656340360641, -0.08538514375686646, -0.011631516739726067, 0.29455995559692383, -0.011589882895350456, 0.0675022155046463, 0.2806147634983063, 0.026751616969704628, 0.2315535545349121, -0.37062862515449524, -0.44300025701522827, 0.3553465008735657, 0.040036000311374664, 0.048634693026542664, 0.419598251581192, 0.21355532109737396, -0.21425551176071167, -0.20989453792572021, 0.05257720500230789, 0.18030835688114166, -0.07134491205215454, 0.04076196253299713, -0.10820142179727554, 0.1665847897529602, -0.24493515491485596, -0.21670866012573242, -0.17305874824523926, -0.10115175694227219, 0.2547438144683838, 0.3276863992214203, 0.09362060576677322, -0.07407157868146896, -0.22788845002651215, 0.23817850649356842, -0.3003154397010803, -0.044358473271131516, 0.0840812399983406, 0.25461214780807495, -0.2716076076030731, 0.3353066146373749, 0.03003021702170372, 0.1542760580778122, 0.2186022400856018, -0.6904953122138977, -0.4939383864402771, 0.13585570454597473, 0.3060668706893921, 0.15567119419574738, 0.1767360121011734, -0.18371886014938354, -0.22654743492603302, 0.3084183633327484, 0.24551071226596832, -0.16246230900287628, 0.029782073572278023, 0.19417038559913635, 0.42701593041419983, -0.01719987392425537, -0.04692161828279495, 0.09551957249641418, -0.16586413979530334, -0.09738191217184067, 0.3946690261363983, -0.029312539845705032, 0.01717277429997921, 0.08575218170881271, 0.1370743364095688, -0.033564213663339615, 0.20169438421726227, 0.2797214984893799, -0.05939615145325661, -0.1606949269771576, 0.15213683247566223, 0.08683425933122635, -0.0403820164501667, -0.4850471615791321, 0.078610360622406, 0.19921673834323883, -0.020880453288555145, -0.5251501798629761, 0.08761709928512573, -0.23867139220237732, 0.46955016255378723, 0.16023920476436615, 0.2618004381656647, 0.3291913866996765, -0.2775519788265228, -0.2181495875120163, 0.030388616025447845, 0.2831471860408783, 0.4071330428123474, -0.038944922387599945, -0.43501949310302734, -0.776552677154541, -0.0014820249052718282, -0.09129543602466583, 0.016643624752759933, 0.4128483533859253, 0.4393833577632904, -0.1327076256275177, 0.2475263774394989, 0.0767604261636734, 0.91904217004776, 0.23129472136497498, 0.04154977947473526, 0.07719230651855469, -0.08021679520606995, 0.19294854998588562, -0.3251570165157318, -0.13492451608181, 0.04967108741402626, -0.23054392635822296, 0.005983916111290455, 0.2619851529598236, 0.3471667766571045, 0.24996373057365417, -0.2336074262857437, 0.10049985349178314, 0.09623256325721741, -0.19594305753707886, 0.2961096167564392, 0.3082047998905182, 0.10677410662174225, -0.43125566840171814, 0.033846043050289154, 0.23043569922447205, 0.04379279538989067, 0.20384731888771057, -0.07254520803689957, -0.05668376013636589, 0.10696385055780411, 0.13140138983726501, -0.45485877990722656, -0.07609514892101288, -0.5070961117744446, 0.12641838192939758, -0.4729514420032501, 0.08941693603992462, 0.2029143124818802, -0.08584137260913849, 0.4214234948158264, 0.4717184603214264, 0.040915824472904205, 0.02292502112686634, 0.4762389063835144, 0.06283982843160629, -0.0877065509557724, -0.23555326461791992, 0.2755263149738312, -0.05053972080349922, -0.26363590359687805, 0.32308346033096313, -0.20167528092861176, -0.13621482253074646, -0.3825468122959137, -0.11631955206394196, -0.10659096390008926, -0.045728325843811035, -0.17525960505008698, 0.4103653132915497, 0.17232656478881836, -0.19534620642662048, 0.14802271127700806, 0.34975072741508484, -0.23560038208961487, -0.19416089355945587, 0.10732937604188919, -0.09747310727834702, -0.3136826157569885, -0.22666020691394806, 0.2490634024143219, -0.6465165019035339, 0.2945075035095215, 0.43899527192115784, -0.15036290884017944, -0.23534387350082397, 0.08640585094690323, 0.44002705812454224, -0.4701535403728485, -0.38490691781044006, 0.005976917687803507, -0.14805954694747925, -0.04269758239388466, 0.2869291305541992, 0.06660960614681244, -0.2061111479997635, 0.15076743066310883, -0.09458732604980469, -0.3692907989025116, -0.02894095703959465, 0.14872434735298157, 0.1253797709941864, 0.2711099088191986, -0.22384405136108398, 0.035785261541604996, 0.0790962353348732, -0.3972316384315491, -0.10924946516752243, -0.10658746212720871, 0.003047830890864134, 0.24276462197303772, 0.055687982589006424, 0.014138046652078629, 0.03401077166199684, 0.19847609102725983, -0.34704360365867615, -0.07309703528881073, -0.18337328732013702, -0.1146322712302208, 0.07452739775180817, 0.13493070006370544, 0.12282229214906693, -0.07794784009456635, -0.03505668416619301, -0.10303802788257599, -0.14566443860530853, 0.11736613512039185, 0.07653312385082245, -0.07971978187561035, 0.2564327120780945, -0.09329617023468018, 0.011540493927896023, -0.16124559938907623, 0.09220390021800995, 0.28333544731140137, 0.04205448925495148, 0.2887311577796936, 0.1680704951286316, 0.18502826988697052, -0.13068316876888275, 0.08269290626049042, 0.29423755407333374, -0.004716711118817329, 0.349841833114624, -0.13883322477340698, 0.03794851899147034, 0.28712016344070435, 0.304423987865448, -0.0796227753162384, 0.0239539984613657, -0.10387271642684937, -0.03874022886157036, 0.3412598669528961, 0.28051406145095825, -0.32523155212402344, 0.16337797045707703, 0.10603347420692444, -0.14815397560596466, 0.19855040311813354, 0.21697553992271423, 0.3297956585884094, 0.2921772003173828, 0.42173564434051514, -0.15993689000606537, -0.13561323285102844, 0.010873337276279926, 0.40319764614105225, 0.3240829110145569, -0.03742901235818863, -0.1339365690946579, 0.09704184532165527, 0.26424267888069153, 0.010727872140705585, 0.2740620970726013, -0.13993851840496063, 0.32852235436439514, 0.19472761452198029, 0.21688242256641388, -0.012067843228578568, -0.23323240876197815, 0.44506603479385376, 0.30710768699645996, 0.28177210688591003, 0.2235705554485321, 0.10950888693332672, -0.1401674449443817, -0.17300277948379517, -0.22349730134010315, -0.046356450766325, 0.08016791939735413, -0.20201201736927032, -0.001011449028737843, 0.4020219147205353, -0.3433103859424591, -0.18070563673973083, -0.22531422972679138, -0.05336703732609749, -0.0010697635589167476, -0.2768159806728363, 0.11281485110521317, 0.13130395114421844, -0.3100360035896301, 0.06996675580739975, 0.02971101924777031, 0.02664848417043686, -0.06110566109418869, 0.2780028283596039, 0.010399439372122288, 0.1770157366991043, 0.04772002622485161, 0.4197746515274048, -0.13973698019981384, 0.08330439031124115, 0.016874631866812706, 0.318022221326828, -0.20493687689304352, -0.018013136461377144, 0.0879959687590599, 0.32789117097854614, 0.06277633458375931, 0.24595969915390015, 0.3205832839012146, 0.1203443631529808, -0.1926157921552658, -0.015043003484606743, 0.171802818775177, -0.1916513293981552, -0.4136286973953247, 0.012700074352324009, -0.0384085476398468, 0.08607937395572662, -0.3797043263912201, -0.031050771474838257, -0.12854589521884918, -0.13488519191741943, 0.3256210386753082, -0.25858423113822937, -0.10838145017623901, -0.07899264991283417, 0.13846109807491302, 0.1661756932735443, -0.20290391147136688, 0.4855581820011139, 0.23405610024929047, -0.1668653041124344, -0.217442587018013, -0.4381392300128937, 0.05126534029841423, 0.08015064150094986, 0.17446169257164001, -0.12456180155277252, 0.0625835508108139, 0.38319626450538635, 0.11679388582706451, -0.042772069573402405, -0.06421233713626862, 0.16320033371448517, 0.32329174876213074, -0.3479471802711487, 0.06293922662734985, 0.28593379259109497, 0.408440500497818, -0.03020792454481125, -0.018504060804843903, -0.07329060137271881, 0.41259080171585083, 0.1505437046289444, -0.13348501920700073, -0.24621184170246124, 0.13039077818393707, 0.12726649641990662, 0.3557191491127014, 0.226509228348732, 0.5275259613990784, -0.22152625024318695, -0.03797096014022827, -0.26853349804878235, -0.223232701420784, -0.06925425678491592, 0.17556272447109222, 0.028544750064611435, 0.06362190842628479, -0.11477155238389969, -0.31324559450149536, -0.2461608350276947, 0.20301179587841034, 0.17591342329978943, 0.2730839252471924, -0.5236431360244751, 0.21069715917110443, 0.09567064046859741, -0.374582439661026, -0.182957261800766, 0.145992711186409, 0.04192346706986427, -0.0368623211979866, -0.09746722877025604, -0.05345086008310318, 0.043697744607925415, -0.17672082781791687, -0.23924511671066284, -0.11105264723300934, 0.13255837559700012, 0.3138341009616852, 0.11591632664203644, -0.3667922914028168, 0.17310667037963867, 0.2874094843864441, 0.2281910479068756, 0.283444881439209, -0.02660801075398922, -0.020219577476382256, 0.03794602304697037, -0.2808719575405121, 0.30048614740371704, 0.11556483805179596, 0.09652005136013031, 0.19395141303539276, -0.22159993648529053 ]
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`
[ -0.20377333462238312, 0.10567417740821838, -0.03363502025604248, 0.06406789273023605, -0.10075332969427109, -0.16913917660713196, 0.13653074204921722, 0.10918204486370087, -0.3450913727283478, 0.12806248664855957, -0.1643531173467636, 0.518744707107544, 0.1809387058019638, 0.08213916420936584, 0.16695693135261536, 0.07651326060295105, 0.06449414789676666, 0.46056920289993286, 0.046868856996297836, 0.15668267011642456, -0.21310803294181824, 0.4624914228916168, -0.40795883536338806, 0.15370409190654755, -0.23863641917705536, -0.16774123907089233, -0.20612679421901703, 0.27909156680107117, -0.5138828754425049, -0.2758505940437317, 0.2979663610458374, 0.14172407984733582, -0.008438868448138237, 0.18113596737384796, -0.00011696519504766911, -0.23673886060714722, 0.12079593539237976, -0.051851607859134674, -0.3485305607318878, -0.3542289137840271, -0.008578434586524963, -0.443655788898468, 0.1595691740512848, -0.34196069836616516, 0.14092856645584106, 0.39214691519737244, 0.3511864244937897, 0.3706689774990082, 0.34925946593284607, 0.38748082518577576, 0.15486441552639008, 0.31376323103904724, 0.215810164809227, -0.05447586998343468, 0.16369736194610596, -0.2806617319583893, -0.29767870903015137, 0.32439979910850525, 0.582219123840332, -0.16857078671455383, 0.0035514789633452892, 0.10590348392724991, -0.09550746530294418, 0.05590515211224556, 0.12202007323503494, 0.1638886034488678, -0.2535194456577301, -0.42911747097969055, 0.29944539070129395, 0.1212935820221901, 0.0607747845351696, -0.2001693993806839, -0.018151961266994476, 0.0014597601257264614, 0.2585045099258423, -0.34449490904808044, 0.21898558735847473, 0.3701315224170685, -0.3430837094783783, 0.07960153371095657, 0.03470706567168236, -0.03276298567652702, -0.21706078946590424, 0.23085390031337738, 0.05665188655257225, 0.36470404267311096, -0.06477499008178711, 0.01026010699570179, 0.25631555914878845, 0.07807118445634842, 0.22043858468532562, 0.23245769739151, -0.1993969827890396, -0.012577180750668049, -0.020614491775631905, -0.19665183126926422, -0.1525166630744934, 0.3759247958660126, 0.23538963496685028, -0.353580117225647, 0.19688460230827332, -0.06267084926366806, 0.19996117055416107, 0.1969323605298996, 0.02321113832294941, -0.012032562866806984, 0.3583293557167053, -0.041698385030031204, 0.3990333676338196, 0.40184175968170166, -0.02484765835106373, 0.001664825715124607, 0.04790098965167999, -0.17749693989753723, -0.2790902256965637, -0.12938421964645386, 0.06468920409679413, -0.18971864879131317, -0.21751658618450165, -0.016829630360007286, -0.15009929239749908, 0.2782166004180908, 0.03317290544509888, 0.19462893903255463, -0.06817463785409927, -0.23874235153198242, 0.10872018337249756, -0.02255195379257202, -0.36746838688850403, -0.1743016391992569, 0.00994944665580988, -0.0019466696539893746, -0.18763971328735352, 0.1096838116645813, 0.26192912459373474, 0.18368849158287048, 0.477835088968277, -0.15345945954322815, 0.09960541874170303, -0.11781337857246399, 0.05239604786038399, -0.26532357931137085, -0.10134172439575195, 0.31440988183021545, 0.3155170679092407, -0.22006887197494507, -0.27172353863716125, -0.5757361054420471, 0.04511934518814087, 0.19760020077228546, -0.03203390911221504, -0.012889780104160309, -0.12502741813659668, 0.10841795802116394, -0.284467488527298, -0.07026511430740356, -0.3243394196033478, 0.05498479679226875, -0.21550442278385162, -0.2589314579963684, 0.04896043241024017, -0.15393133461475372, -0.24565055966377258, -0.0019287567120045424, -0.05847945809364319, 0.1869661957025528, -0.032391589134931564, -0.09881692379713058, 0.16054755449295044, 0.008412638679146767, -0.0025030761025846004, 0.25573214888572693, -0.14600396156311035, -0.08713574707508087, 0.06068763509392738, 0.19247466325759888, 0.3825962245464325, -0.3142409026622772, -0.71578049659729, 0.12331640720367432, -0.07620734721422195, -0.28646764159202576, -0.09852335602045059, 0.1813640147447586, 0.420303612947464, -0.18550941348075867, 0.07004635781049728, 0.4979858696460724, -0.1739974319934845, 0.10833082348108292, -0.26977047324180603, -0.17382082343101501, -0.1288990080356598, 0.05015194043517113, 0.1142122820019722, -0.01719146966934204, -0.03910653665661812, 0.7044432163238525, 0.16681408882141113, -0.09890887141227722, -0.2014876753091812, 0.05992189422249794, 0.32618895173072815, -0.30106380581855774, -0.0882210060954094, -0.1463492065668106, -0.5062059760093689, 0.010046889074146748, -0.3460908830165863, 0.2132529765367508, 0.09683457762002945, -0.21113012731075287, -0.059183478355407715, -0.06821373850107193, -0.0121840201318264, -0.1781017780303955, 0.10006791353225708, 0.020487025380134583, 0.04887041822075844, -0.048605479300022125, 0.09019957482814789, 0.3979104161262512, -0.19552576541900635, 0.4248863160610199, -0.20517778396606445, 0.18992042541503906, -0.3483103811740875, -0.3624229431152344, 0.06926017999649048, 0.3132469952106476, 0.003431719494983554, -0.1705968827009201, -0.10006584972143173, 0.2572609782218933, 0.08105365931987762, -0.0746907889842987, 0.10122884809970856, -0.3666819930076599, 0.39606189727783203, 0.06066787987947464, -0.04035742208361626, -0.06082822009921074, 0.012146716937422752, -0.03536980226635933, -0.028800826519727707, 0.22457678616046906, -0.18602292239665985, -0.04142359644174576, 0.17305617034435272, 0.20941875874996185, 0.03926591947674751, -0.11988578736782074, -0.2044532746076584, 0.13204501569271088, 0.3231660723686218, 0.23681381344795227, 0.041179731488227844, 0.010915272869169712, -0.36299949884414673, 0.21057844161987305, 0.4938599169254303, 0.027466069906949997, 0.2480994611978531, 0.019872769713401794, 0.04852671921253204, 0.002645436441525817, 0.34254059195518494, 0.08806080371141434, 0.3751702606678009, 0.11413431912660599, -0.11013109982013702, -0.0038268412463366985, -0.185357466340065, -0.3078567385673523, 0.13142111897468567, 0.08304540812969208, 0.12990714609622955, -0.1697394996881485, 0.30302730202674866, 0.014845985919237137, -0.36467909812927246, -0.11107063293457031, -0.07071489840745926, 0.2350340634584427, 0.10903169214725494, 0.30596423149108887, -0.14340347051620483, -0.6106245517730713, -0.2032153159379959, -0.11518286168575287, -0.18614642322063446, -0.18630239367485046, -0.054775483906269073, 0.39907002449035645, 0.39377912878990173, 0.13219666481018066, 0.3591202199459076, 0.20930393040180206, -0.11759832501411438, 0.18299813568592072, 0.17196933925151825, -0.04389527812600136, -0.4294462203979492, 0.03136968985199928, 0.3862448036670685, 0.2979986369609833, 0.10854143649339676, -0.23654918372631073, -0.2492026686668396, 0.015471947379410267, -0.42980432510375977, 0.2603360414505005, -0.33625662326812744, 0.2794797718524933, 0.09976357966661453, 0.2478223443031311, -0.5951435565948486, -0.10310754925012589, 0.16499005258083344, -0.11449181288480759, -0.12290869653224945, -0.0929282158613205, -0.03688306361436844, -0.40554603934288025, -0.26714780926704407, -0.20519249141216278, -0.43597277998924255, -0.2693943679332733, 0.33582836389541626, 0.01575186289846897, 0.21157875657081604, 0.3739762604236603, 0.09294811636209488, 0.3100523054599762, 0.04953606054186821, -0.20536307990550995, 0.12083636969327927, 0.03579309210181236, 0.08852703124284744, -0.2505458891391754, -0.49854183197021484, 0.09977533668279648, 0.05173468217253685, 0.0376451313495636, -0.24499201774597168, -0.4161093831062317, -0.5225143432617188, 0.12612803280353546, 0.18824495375156403, -0.0012071693781763315, -0.07886310666799545, 0.2069428265094757, -0.19732947647571564, 0.10955823957920074, -0.013077055104076862, -0.24611100554466248, -0.12782588601112366, 0.08309172838926315, 0.3247509002685547, -0.15073609352111816, 0.1058800145983696, 0.14839686453342438, 0.19341640174388885, 0.3145641088485718, -0.0725046768784523, 0.11436934024095535, 0.12013459950685501, 0.27393531799316406, -0.21050916612148285, -0.3405759930610657, 0.34247997403144836, 0.010135017335414886, -0.009618915617465973, 0.34545621275901794, -0.05972890555858612, -0.5129147171974182, -0.1813620626926422, 0.1022452861070633, -0.25864043831825256, -0.356607049703598, 0.10800468176603317, 0.4365466833114624, -0.04296324402093887, 0.2192896157503128, 0.17693935334682465, -0.006431665271520615, -0.2671230733394623, -0.39389073848724365, 0.2829703986644745, -0.2330522984266281, 0.10045421123504639, 0.18662990629673004, -0.4540291130542755, -0.6995072364807129, 0.1291135549545288, 0.02011226676404476, 0.29739704728126526, -0.008754913695156574, 0.23194865882396698, 0.13225023448467255, -0.028030572459101677, 0.24217186868190765, 0.08653072267770767, -0.12931200861930847, 0.0555366612970829, 0.024041907861828804, -0.25207942724227905, 0.14389953017234802, -0.35901403427124023, 0.5358403325080872, 0.8802663087844849, 0.3770969808101654, -0.16434849798679352, 0.05778135359287262, 0.5292019844055176, -0.03279132768511772, 0.18958865106105804, 0.10803993046283722, -0.30443379282951355, -0.3213100731372833, -0.14403432607650757, 0.05696246400475502, -0.04829372838139534, -0.007251468487083912, 0.09713461995124817, -0.10260345786809921, 0.2945188283920288, -0.1304941475391388, -0.1803799569606781, 0.4900842010974884, -0.04314813390374184, 0.06898374855518341, 0.034631695598363876, -0.20050613582134247, -0.14851181209087372, 0.23588134348392487, 0.27820128202438354, -0.21470890939235687, -0.2510850727558136, 0.19106832146644592, -0.29935526847839355, 0.1263989508152008, 0.5574198961257935, -0.012705020606517792, -0.08767621964216232, -0.14137090742588043, -0.010357537306845188, 0.37668395042419434, 0.008204529993236065, 0.34758460521698, 0.13797667622566223, -0.08299773931503296, 0.15213726460933685, 0.2743896245956421, -0.010234457440674305, 0.060648757964372635, -0.12669670581817627, -0.133026123046875, -0.26653575897216797, 0.4317963421344757, 0.031324151903390884, 0.8313062787055969, -0.19805559515953064, 0.18259058892726898, 0.2165958285331726, 0.03458157926797867, 0.7005779147148132, 0.09818464517593384, 0.29737433791160583, -0.1605037897825241, -0.2612920105457306, -0.02691725455224514, 0.034172751009464264, 0.2771380543708801, 0.0035683668684214354, -0.6976839303970337, 0.3543379008769989, 0.5069890022277832, -0.06665947288274765, 0.14088714122772217, 0.20467044413089752, -0.20412836968898773, -0.33093953132629395, -0.6929690837860107, 0.14991536736488342, -0.24720895290374756, 0.21208223700523376, -0.23182708024978638, -0.01660643145442009, -0.22074243426322937, -0.05585189536213875, -0.39808890223503113, 0.09784265607595444, -0.49289390444755554, -0.1491033136844635, 0.3097960948944092, -0.2716408669948578, -0.015284744091331959, 0.31808724999427795, 0.2980523109436035, 0.21503806114196777, -0.041175954043865204, 0.05580109357833862, -0.00948933232575655, -0.428806334733963, 0.06961552053689957, 0.1207268014550209, 0.07078715413808823, -0.23157918453216553, -0.011699261143803596, -0.1561686396598816, 0.19199419021606445, -0.14730076491832733, 0.3778710961341858, 0.26534608006477356, 0.11379934102296829, 0.15317007899284363, -0.11881118267774582, -0.04288330301642418, -0.18436682224273682, -0.26054757833480835, 0.10958399623632431, -0.10283760726451874, -0.10809368640184402, 0.17686621844768524, 0.17040115594863892, -0.16747762262821198, -0.018127942457795143, 0.24444781243801117, -0.10914558917284012, 0.05315781757235527, 0.21790297329425812, 0.2029287964105606, -0.04807593673467636, -0.21867182850837708, 0.16416017711162567, -0.4710911810398102, -0.20756691694259644, 0.01929772086441517, 0.004306074231863022, -0.17338906228542328, -0.05642121285200119, 0.6419910788536072, 0.29731157422065735, 0.032893646508455276, -0.12059910595417023, -0.3671327233314514, -0.2201475352048874, 0.20411580801010132, 0.22043627500534058, -0.12178989499807358, 0.33465564250946045, 0.334052175283432, 0.28580617904663086, 0.09651746600866318, -0.2278681844472885, 0.2888354957103729, -0.3742857873439789, 0.17683298885822296, 0.03506309911608696, -0.24676944315433502, 0.16113603115081787, -0.13581202924251556, -0.05651438981294632, 0.5478607416152954, 0.06339485943317413, -0.21986696124076843, -0.13317610323429108, 0.06871259957551956, 0.11896975338459015, -0.10746383666992188, -0.31719622015953064, -0.06813294440507889, -0.09852124005556107, -0.04035225138068199, 0.07765185832977295, -0.06821086257696152, 0.10933523625135422, -0.06905239820480347, 0.26677951216697693, 0.2946944534778595, 0.2482580691576004, 0.07181031256914139, -0.054368313401937485, 0.21100717782974243, 0.12927596271038055, 0.25017642974853516, 0.21401149034500122, 0.03418020159006119, 0.14100207388401031, -0.00031043743365444243, -0.07852867245674133, 0.15759479999542236, 0.26674503087997437, -0.40233513712882996, 0.10853849351406097, -0.06369982659816742, -0.04671726003289223, 0.3891506493091583, 0.0595790259540081, -0.16276581585407257, 0.2553040683269501, 0.07318652421236038, -0.1452825516462326, 0.06295525282621384, 0.5381841659545898, 0.16475968062877655, 0.07420125603675842, 0.4867253303527832, 0.11547701060771942, -0.0481443926692009, 0.1442146599292755, 0.10595782101154327, 0.23524439334869385, 0.2533343434333801, 0.01160761620849371, -0.08355822414159775, 0.013890893198549747, 0.11324355751276016, 0.08847470581531525, 0.05775384604930878, 0.19474150240421295, 0.008706499822437763, -0.0025568294804543257, 0.18441611528396606, -0.1320842057466507, 0.09944771230220795, 0.2976967692375183, -0.03657418116927147, -0.2163548618555069, 0.04410652443766594, 0.2792077958583832, 0.062182363122701645, 0.13555817306041718, 0.8394277691841125, -0.5352563261985779, -0.25631365180015564, -0.11296939104795456, 0.13962054252624512, -0.361601859331131, -0.00691243214532733, -0.30979692935943604, 0.15970873832702637, -0.13192443549633026, -0.018767856061458588, -0.1546420156955719, -0.04683646932244301, 0.3026387393474579, 0.29672521352767944, -0.1764538735151291, -0.11941342055797577, -0.21040692925453186, 0.09632755070924759, -0.15803033113479614, -0.04321964085102081, 0.5650871992111206, -0.46155306696891785, 0.15587051212787628, -0.11572068929672241, 0.5128490328788757, 0.19595365226268768, 0.23445618152618408, -0.17214354872703552, -0.16993483901023865, -0.040043558925390244, 0.021636435762047768, -0.09688597917556763, -0.019116587936878204, 0.4978407323360443, 0.24874092638492584, 0.18865656852722168, 0.03008071519434452, -0.1156952977180481, 0.19563601911067963, 0.15852321684360504, -0.2628212869167328, 0.05865899473428726, -0.10240840911865234, -0.10741771757602692, -0.14664141833782196, -0.27449938654899597, -0.043056290596723557, -0.34732404351234436, 0.5429145693778992, -0.29940932989120483, -0.10698157548904419, 0.04497741535305977, -0.09464633464813232, 0.0484454482793808, -0.004604566376656294, 0.5350642204284668, 0.40321576595306396, 0.025475945323705673, -0.21520185470581055, -0.4277661144733429, -0.38895678520202637, 0.21612271666526794, -0.25636357069015503, 0.011096659116446972, 0.059635333716869354, 0.18237660825252533, -0.10124043375253677, 0.4139775335788727, -0.07893026620149612, 0.04184617102146149, -0.12649869918823242, -0.06532090902328491, -0.4724360406398773, -0.11600548774003983, 0.28529056906700134, -0.2760683298110962, -0.0696648359298706, -0.22366350889205933, 0.1521492749452591, -0.20586013793945312, -0.07175619900226593, -0.007133260369300842, -0.3104132115840912, -0.29001399874687195, -0.10448233038187027, 0.07206236571073532, 0.17017735540866852, 0.10107703506946564, 0.019608527421951294, -0.32109400629997253, -0.046575166285037994, -0.22404088079929352, -0.27195897698402405, 0.07557500898838043, 0.2256833165884018, 0.297549307346344, -0.05231308937072754, 0.3451104760169983, -0.16596205532550812, 0.5020351409912109, 0.10082203149795532, 0.03670184314250946, -0.012219477444887161, 0.0015542861074209213, -0.12299047410488129, 0.08524475991725922, -0.14766035974025726, 0.45889899134635925, -0.2785677909851074, 0.17531108856201172, -0.4899502694606781, -0.31665003299713135, 0.3832271695137024, -0.4803462326526642, -0.09981118142604828, 0.0017296771984547377, 0.20958420634269714, 0.13941428065299988, -0.3002651333808899, -0.39764609932899475, -0.025793958455324173, -0.15764190256595612, -0.1750725656747818, -0.4597729444503784, 0.1397780030965805, -0.20215070247650146, 0.051266517490148544, -0.08261647820472717, 0.41824233531951904, 0.07982710748910904, -0.25584185123443604, 0.015247363597154617, -0.15567602217197418 ]
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
[ -0.1622944474220276, 0.2886502146720886, -0.016385279595851898, 0.08907564729452133, -0.09513932466506958, -0.135273739695549, 0.24156036972999573, 0.04239345341920853, -0.44447287917137146, 0.05577056482434273, -0.342538982629776, 0.5006805062294006, 0.19029949605464935, 0.05172169581055641, 0.23175160586833954, 0.12949398159980774, 0.05272139608860016, 0.4562416970729828, -0.005170074291527271, 0.15157265961170197, -0.13718920946121216, 0.4322417080402374, -0.3422397971153259, 0.19298423826694489, -0.28372982144355774, -0.2066504806280136, -0.3387555480003357, 0.3366292715072632, -0.3838268518447876, -0.36793965101242065, 0.30076920986175537, 0.1165231391787529, -0.010650725103914738, 0.24211180210113525, -0.00011587276094360277, -0.2633942663669586, 0.0717015340924263, 0.05097358673810959, -0.3235510289669037, -0.1854398250579834, 0.15792886912822723, -0.2767644226551056, 0.3587699830532074, -0.36968275904655457, 0.23197363317012787, 0.2630348205566406, 0.19682088494300842, 0.4131794273853302, 0.37842145562171936, 0.25508952140808105, 0.15432631969451904, 0.3623245060443878, 0.15149953961372375, -0.04990313574671745, 0.3280055522918701, -0.24015356600284576, -0.24223308265209198, 0.5541878342628479, 0.5191040635108948, -0.17601540684700012, -0.06431873887777328, -0.023654667660593987, -0.15008217096328735, 0.10818654298782349, 0.09065628051757812, 0.2510840892791748, -0.30428019165992737, -0.3595670759677887, 0.14779295027256012, 0.07521241158246994, 0.010056709870696068, -0.2550986409187317, 0.08777579665184021, -0.08877250552177429, 0.2548275887966156, -0.46778032183647156, 0.22016596794128418, 0.2952529191970825, -0.33332523703575134, 0.04739920422434807, 0.06930744647979736, -0.09359442442655563, -0.15576404333114624, 0.1831718236207962, 0.1144595593214035, 0.358079731464386, -0.056636568158864975, 0.01170644536614418, 0.2579827606678009, 0.11046654731035233, 0.2554870545864105, 0.26135581731796265, -0.1273604929447174, -0.1209302470088005, -0.048167504370212555, -0.10802473872900009, -0.11227384954690933, 0.35357770323753357, 0.17000281810760498, -0.2848019301891327, 0.18919838964939117, -0.0828457772731781, 0.18041135370731354, 0.18949386477470398, -0.02586636319756508, 0.011831952258944511, 0.17808938026428223, -0.22734493017196655, 0.4025892913341522, 0.2651008069515228, -0.13250122964382172, -0.019605988636612892, 0.06968577951192856, -0.08283204585313797, -0.1812390387058258, -0.2002732902765274, 0.11146126687526703, -0.17330388724803925, -0.21198901534080505, -0.06052529066801071, -0.21012085676193237, 0.10313345491886139, 0.030671700835227966, 0.22637037932872772, -0.04862547293305397, -0.32488805055618286, -0.008808953687548637, 0.009723497554659843, -0.3067668676376343, -0.16587617993354797, -0.02070273831486702, -0.015813102945685387, -0.3158724308013916, 0.29279252886772156, 0.2561110556125641, 0.014169808477163315, 0.43691664934158325, -0.06674637645483017, 0.11256450414657593, -0.06157002970576286, -0.036542005836963654, -0.25050821900367737, -0.13767027854919434, 0.19454529881477356, 0.28504157066345215, -0.13704144954681396, -0.27462565898895264, -0.5182225108146667, -0.029475698247551918, 0.13417832553386688, -0.09137144684791565, -0.10038183629512787, -0.14188046753406525, 0.11342832446098328, -0.2824955880641937, -0.0845004990696907, -0.24448542296886444, -0.011239090003073215, -0.2775987684726715, -0.32704606652259827, 0.06488121300935745, -0.11334031075239182, -0.11958073824644089, -0.17336563766002655, -0.05139106139540672, 0.227773979306221, -0.1550368219614029, -0.06079227849841118, 0.1266472041606903, 0.08688858151435852, 0.056462887674570084, 0.3691752254962921, -0.22387829422950745, 0.05881858244538307, -0.03029719926416874, 0.2738121449947357, 0.47548314929008484, -0.13133664429187775, -0.5511143207550049, 0.022248219698667526, -0.0927392765879631, -0.220012366771698, -0.17768485844135284, 0.18534114956855774, 0.4359276294708252, -0.1713544875383377, 0.1205330565571785, 0.4776642620563507, -0.12799134850502014, 0.08927559107542038, -0.25258955359458923, -0.20198693871498108, 0.02106165699660778, 0.0683598667383194, 0.10308259725570679, -0.047923922538757324, 0.01084762904793024, 0.6572158336639404, 0.15238067507743835, -0.12794318795204163, -0.2924889922142029, 0.012023353017866611, 0.26373186707496643, -0.15930821001529694, -0.08614973723888397, -0.10207735747098923, -0.4976183772087097, -0.018914876505732536, -0.38266491889953613, 0.22172366082668304, 0.11234721541404724, -0.1261185258626938, -0.01968996785581112, -0.029196418821811676, 0.013084876351058483, -0.10262606292963028, 0.11551791429519653, 0.09121495485305786, 0.03127628192305565, 0.03552224859595299, 0.011038724333047867, 0.30791762471199036, -0.21081528067588806, 0.4070407748222351, -0.1642877459526062, 0.33948540687561035, -0.3726191818714142, -0.4543435871601105, 0.03883105888962746, 0.31756559014320374, -0.0054068975150585175, -0.18586817383766174, -0.08476999402046204, 0.3038400113582611, -0.03310534358024597, -0.0015199062181636691, -0.04197310656309128, -0.21679909527301788, 0.41755738854408264, -0.03388036787509918, 0.040236372500658035, -0.1533370167016983, 0.011028807610273361, -0.11775396764278412, 0.12128815054893494, 0.15278561413288116, -0.27263882756233215, -0.04674266278743744, 0.1193268820643425, 0.08081517368555069, 0.000818429165519774, -0.12340366095304489, -0.12160487473011017, 0.20139898359775543, 0.25048863887786865, 0.2321152538061142, 0.055614423006772995, 0.0039876350201666355, -0.34483620524406433, 0.16886043548583984, 0.638894259929657, 0.009611155837774277, 0.1933998018503189, 0.0026634298264980316, -0.055356092751026154, -0.09187642484903336, 0.36545616388320923, -0.1382659673690796, 0.44734692573547363, 0.08535990118980408, -0.11062635481357574, 0.04611583426594734, -0.2001611590385437, -0.3595978617668152, 0.08434481173753738, 0.1504208892583847, 0.3393532931804657, -0.16112470626831055, 0.32813262939453125, 0.18927927315235138, -0.41650494933128357, -0.19667762517929077, -0.11323964595794678, 0.16031935811042786, 0.04637620598077774, 0.31053581833839417, -0.22582709789276123, -0.6363138556480408, -0.21384750306606293, -0.2246876358985901, -0.17365430295467377, -0.21213825047016144, -0.06543602049350739, 0.417105495929718, 0.3795740008354187, 0.10483027994632721, 0.3908367455005646, 0.29194122552871704, -0.14079336822032928, 0.1514676958322525, 0.07684756815433502, -0.11653347313404083, -0.44519689679145813, 0.035596154630184174, 0.44034042954444885, 0.3027225732803345, 0.1866854578256607, -0.1826227456331253, -0.14730459451675415, 0.04165022447705269, -0.47399434447288513, 0.2552635371685028, -0.28294238448143005, 0.250566691160202, 0.005636426154524088, 0.21463942527770996, -0.591575026512146, -0.29663047194480896, 0.17837540805339813, -0.19174475967884064, -0.06382712721824646, -0.07141577452421188, -0.09075511246919632, -0.3686997592449188, -0.23214149475097656, -0.19074290990829468, -0.3892240524291992, -0.2519962191581726, 0.30632469058036804, 0.04709029942750931, 0.12468479573726654, 0.3451833724975586, -0.05784667283296585, 0.42036640644073486, -0.006661339197307825, -0.044965509325265884, 0.06837756186723709, 0.03224069997668266, 0.0629432201385498, -0.2506498098373413, -0.39018309116363525, 0.08632998168468475, 0.08140074461698532, 0.07147706300020218, -0.32992246747016907, -0.44783154129981995, -0.45164912939071655, 0.10525135695934296, 0.25135788321495056, 0.009099363349378109, -0.09579433500766754, 0.2081671804189682, -0.1809079647064209, 0.052147477865219116, 0.12245682626962662, -0.35239294171333313, -0.20721884071826935, 0.1554120033979416, 0.3015429377555847, -0.1551506072282791, 0.14002063870429993, -0.036580611020326614, 0.24720659852027893, 0.12030626088380814, -0.16269691288471222, 0.24349384009838104, 0.04703788086771965, 0.2754901349544525, -0.26920464634895325, -0.32646575570106506, 0.35097163915634155, -0.13879022002220154, -0.07192061096429825, 0.35272225737571716, -0.173997163772583, -0.5435379147529602, -0.17621365189552307, 0.08835022151470184, -0.3520292043685913, -0.3512006103992462, 0.14087724685668945, 0.48909199237823486, -0.12838980555534363, 0.16154561936855316, 0.10040249675512314, 0.1366298794746399, -0.31285780668258667, -0.2801971435546875, 0.14688041806221008, -0.3128335773944855, 0.015016581863164902, 0.08403666317462921, -0.5422757267951965, -0.6152955293655396, 0.09041255712509155, 0.02460482530295849, 0.1647317111492157, 0.010637062601745129, 0.16953398287296295, 0.1697925329208374, -0.07759537547826767, 0.18881717324256897, 0.239374041557312, -0.21169346570968628, 0.10137899219989777, 0.10286177694797516, -0.24931305646896362, 0.0044770622625947, -0.44667425751686096, 0.439561128616333, 0.8527019023895264, 0.21628935635089874, -0.16088272631168365, 0.14705875515937805, 0.4967115521430969, 0.08200754970312119, 0.1896100789308548, 0.13085345923900604, -0.2746921181678772, -0.3481546938419342, -0.20279738306999207, 0.027922648936510086, -0.20913070440292358, 0.035677287727594376, -0.04792356863617897, -0.057375725358724594, 0.3363128900527954, -0.05814094468951225, -0.117659792304039, 0.5659453868865967, -0.05071735754609108, 0.1383466273546219, -0.03656580299139023, -0.16751012206077576, -0.05538635328412056, 0.20562466979026794, 0.31446439027786255, -0.1491706371307373, -0.11714652925729752, 0.2581256628036499, -0.17916467785835266, 0.1546858251094818, 0.5725120902061462, 0.03370540216565132, -0.015726691111922264, -0.254978209733963, -0.02728019282221794, 0.35344305634498596, 0.038460612297058105, 0.3919437527656555, 0.02410220541059971, -0.09691449999809265, 0.22327600419521332, 0.3603271543979645, 0.006348146591335535, 0.005739233922213316, -0.05591968446969986, -0.13157889246940613, -0.1634044498205185, 0.5269419550895691, 0.09646144509315491, 0.9199798703193665, -0.13518963754177094, 0.23280033469200134, 0.3955855667591095, 0.005457582417875528, 0.8213494420051575, 0.036442406475543976, 0.29021501541137695, -0.22001130878925323, -0.10888121277093887, 0.03861958533525467, -0.02745904214680195, 0.14550799131393433, -0.1564488708972931, -0.6769136190414429, 0.24205613136291504, 0.4510635733604431, 0.08140232414007187, 0.05928407609462738, 0.2267390638589859, -0.04052646458148956, -0.27934908866882324, -0.6700119972229004, 0.16516231000423431, -0.23910801112651825, 0.2571815252304077, -0.31535494327545166, 0.0061792898923158646, -0.12149979174137115, -0.019443737342953682, -0.41948962211608887, 0.1596042513847351, -0.394016295671463, -0.10973095148801804, 0.2596362233161926, -0.24909433722496033, -0.0036672339774668217, 0.19052882492542267, 0.3236689865589142, 0.18673241138458252, -0.03592139482498169, -0.02719729393720627, -0.0723293274641037, -0.3444035053253174, 0.1388995200395584, 0.07803376764059067, 0.13850203156471252, -0.23670168220996857, -0.08687041699886322, 0.02865392528474331, 0.2405574917793274, -0.1788836568593979, 0.2510041892528534, 0.26750999689102173, 0.09652287513017654, 0.038775645196437836, 0.021772440522909164, 0.04931177198886871, -0.12440268695354462, -0.2563909590244293, 0.08975748717784882, -0.10483003407716751, -0.046367745846509933, 0.2597416341304779, 0.16537612676620483, -0.15297962725162506, -0.0932944267988205, 0.16011492908000946, -0.12814603745937347, 0.09165734052658081, 0.18983609974384308, 0.261301726102829, -0.07720839977264404, -0.1752098649740219, 0.045474275946617126, -0.5221548676490784, -0.30002251267433167, -0.07522942125797272, 0.02601170912384987, -0.054492294788360596, 0.07419304549694061, 0.6213569641113281, 0.34520575404167175, 0.13891111314296722, -0.0982711911201477, -0.3008251488208771, -0.13825131952762604, 0.22200608253479004, 0.31040093302726746, -0.10538048297166824, 0.433083176612854, 0.42102208733558655, 0.23336480557918549, 0.1701848804950714, -0.23408553004264832, 0.30826258659362793, -0.34560009837150574, 0.13437233865261078, -0.04723072424530983, -0.27821364998817444, 0.18948937952518463, -0.06866835057735443, -0.05093332380056381, 0.5490370392799377, 0.13024336099624634, -0.19484561681747437, -0.055499155074357986, 0.064859539270401, 0.08955330401659012, -0.1063515841960907, -0.355571985244751, -0.08047809451818466, 0.07174216210842133, 0.007463629823178053, 0.11996128410100937, -0.13209033012390137, 0.1513804793357849, -0.06826148927211761, 0.3728260099887848, 0.24534691870212555, 0.13897669315338135, 0.11990583688020706, 0.018168145790696144, 0.2314881682395935, 0.16493672132492065, 0.26410728693008423, 0.11129946261644363, -0.008376375772058964, 0.10891080647706985, 0.012166950851678848, -0.04408545792102814, 0.06559466570615768, 0.1964224874973297, -0.38780471682548523, 0.07796408981084824, -0.23517337441444397, -0.03739144280552864, 0.48941275477409363, 0.13260647654533386, -0.18248452246189117, 0.24811676144599915, 0.06972338259220123, -0.11593446135520935, 0.02227625623345375, 0.44858938455581665, 0.18295279145240784, -0.019731316715478897, 0.4950127899646759, 0.15736478567123413, -0.0033345401752740145, 0.1262979954481125, 0.04812315106391907, 0.26875609159469604, 0.3688947558403015, 0.0004858215106651187, -0.06775560975074768, -0.06217390298843384, 0.035238198935985565, 0.17615745961666107, 0.06021856144070625, 0.1521509438753128, -0.0674014464020729, -0.12788823246955872, 0.15209056437015533, -0.28901126980781555, 0.06382209807634354, 0.3066234290599823, -0.05420032888650894, -0.25578391551971436, 0.11049550026655197, 0.3738560378551483, -0.02909071184694767, 0.08440814167261124, 0.8333594799041748, -0.3544100224971771, -0.3209327459335327, -0.10203970968723297, 0.25673288106918335, -0.34500885009765625, 0.0990147814154625, -0.2724241614341736, 0.10394688695669174, -0.08362744748592377, 0.08783464133739471, -0.1785786896944046, -0.13201043009757996, 0.3258434534072876, 0.273580938577652, -0.25387582182884216, -0.1599980741739273, -0.16739694774150848, 0.054495375603437424, -0.18168999254703522, 0.07494218647480011, 0.5572720766067505, -0.33135420083999634, 0.09935422986745834, -0.2330409437417984, 0.5280354619026184, 0.2896026372909546, 0.20487375557422638, -0.1909983605146408, -0.15233838558197021, -0.040742870420217514, 0.04505087807774544, -0.014735610224306583, -0.06881915777921677, 0.4951959252357483, 0.10088463127613068, 0.18965785205364227, 0.04977738484740257, -0.13975860178470612, 0.28492459654808044, 0.1317388415336609, -0.21514558792114258, 0.037306077778339386, -0.10045886784791946, 0.003753671422600746, -0.17565694451332092, -0.30538421869277954, -0.050677571445703506, -0.31464943289756775, 0.5148184299468994, -0.14559154212474823, -0.11733537912368774, -0.04240341857075691, -0.13935598731040955, 0.05286145582795143, 0.09525218605995178, 0.6466183066368103, 0.5057281851768494, 0.1590743064880371, -0.19428087770938873, -0.45061251521110535, -0.4104739725589752, 0.2689818739891052, -0.30991461873054504, 0.11366620659828186, 0.05045697093009949, 0.19246648252010345, -0.16570338606834412, 0.4790000915527344, -0.02666253224015236, 0.11698699742555618, -0.24810898303985596, -0.15506230294704437, -0.5336964130401611, -0.015051678754389286, 0.11213735491037369, -0.3295288681983948, -0.1642407923936844, -0.043305665254592896, 0.2182345986366272, -0.3421745002269745, -0.07783358544111252, -0.06188169866800308, -0.3067426085472107, -0.2668483555316925, -0.0754542350769043, 0.2547646164894104, 0.10352200269699097, 0.13317327201366425, 0.07666167616844177, -0.3111726641654968, 0.029302917420864105, -0.38363736867904663, -0.1451123207807541, 0.047343019396066666, 0.17323517799377441, 0.19779284298419952, 0.0003887346538249403, 0.3683888018131256, -0.19359786808490753, 0.504595160484314, 0.0966079831123352, 0.06460700929164886, -0.002981429686769843, -0.0904897153377533, -0.0720776617527008, 0.08612070232629776, -0.08781006187200546, 0.4204275906085968, -0.23913690447807312, 0.25775471329689026, -0.5477001070976257, -0.263095885515213, 0.41291242837905884, -0.45274847745895386, -0.04186301678419113, -0.004356177058070898, 0.19462774693965912, 0.11911728233098984, -0.1933591663837433, -0.4065576493740082, -0.09077245742082596, -0.1022675409913063, -0.12823081016540527, -0.44656166434288025, 0.1807403862476349, -0.12650547921657562, -0.009618566371500492, -0.11334364116191864, 0.20604681968688965, 0.04433004930615425, -0.27548086643218994, -0.0735136866569519, -0.1813737004995346 ]
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 :)
[ -0.20812608301639557, 0.08529721200466156, -0.08818323165178299, 0.16447682678699493, -0.023115744814276695, -0.10333465784788132, 0.32870593667030334, 0.08260271698236465, -0.31269553303718567, 0.21142789721488953, -0.39940759539604187, 0.3362332284450531, 0.2900559902191162, 0.16409511864185333, 0.15483738481998444, 0.1645144373178482, 0.09830833971500397, 0.3727589249610901, -0.06697466969490051, 0.14713601768016815, -0.2285146862268448, 0.4517592489719391, -0.3685057759284973, 0.0600338950753212, -0.2734024226665497, -0.20931944251060486, -0.2376539409160614, 0.273295134305954, -0.42368990182876587, -0.21809497475624084, 0.28793686628341675, 0.18390342593193054, -0.11174599826335907, 0.2599034309387207, -0.00011433531471993774, -0.24122953414916992, -0.020322391763329506, -0.0653364434838295, -0.3073401153087616, -0.1606721431016922, -0.08061980456113815, -0.2332576960325241, 0.22001704573631287, -0.40985584259033203, 0.13190801441669464, 0.41221481561660767, 0.22275255620479584, 0.2862118184566498, 0.2939855754375458, 0.2952931821346283, 0.1804482787847519, 0.4040350914001465, 0.12714529037475586, -0.17765124142169952, 0.32549697160720825, -0.2722628116607666, -0.24315333366394043, 0.41001975536346436, 0.6274051070213318, -0.08831335604190826, -0.06562691181898117, 0.13427627086639404, -0.12808924913406372, 0.12494764477014542, 0.08908439427614212, 0.24771198630332947, -0.28943344950675964, -0.48387375473976135, 0.3436744809150696, 0.1162119209766388, 0.15104767680168152, -0.12315115332603455, 0.05892603099346161, 0.058995891362428665, 0.17138376832008362, -0.2932753562927246, 0.2615859806537628, 0.33505716919898987, -0.22270548343658447, 0.1343463808298111, 0.006289522163569927, 0.08987287431955338, -0.1818217635154724, 0.13155531883239746, -0.027878981083631516, 0.16018426418304443, -0.14802958071231842, -0.0957619696855545, 0.2256309539079666, 0.011602788232266903, 0.246577188372612, 0.0990665853023529, -0.16215813159942627, -0.09483560174703598, -0.10899645835161209, -0.24236765503883362, -0.24434909224510193, 0.24999402463436127, 0.23728686571121216, -0.24156518280506134, 0.31874603033065796, -0.029161177575588226, 0.24501104652881622, 0.18477676808834076, -0.03199952840805054, -0.14745505154132843, 0.25958165526390076, -0.10221575200557709, 0.4240212142467499, 0.41627925634384155, -0.18827465176582336, -0.09184850007295609, -0.024419013410806656, -0.1793900430202484, -0.22493116557598114, -0.0915117934346199, 0.04073407128453255, -0.24053426086902618, -0.2889053523540497, 0.07782181352376938, -0.1368495523929596, 0.27898064255714417, 0.0691424161195755, 0.24263978004455566, -0.0014099943218752742, -0.41014164686203003, 0.03285716101527214, -0.01016711350530386, -0.2695295214653015, -0.2202635258436203, -0.030173353850841522, -0.09110909700393677, -0.1644536554813385, 0.2610427439212799, 0.2496694028377533, 0.136254221200943, 0.5002056360244751, -0.09450648725032806, 0.23768046498298645, -0.20367813110351562, 0.03617643937468529, -0.42523887753486633, -0.09555720537900925, 0.35759851336479187, 0.3498214781284332, -0.2039550393819809, -0.22149425745010376, -0.5597003698348999, 0.10842231661081314, 0.07068201899528503, -0.03556102514266968, -0.02277120016515255, -0.03724556788802147, 0.1511412411928177, -0.20522919297218323, 0.024308882653713226, -0.31279510259628296, 0.04790034890174866, -0.28003251552581787, -0.14217346906661987, 0.013601741753518581, -0.03400617092847824, -0.215387225151062, -0.031969793140888214, -0.01668485812842846, 0.26053547859191895, -0.1001935750246048, 0.008026586845517159, 0.2057425081729889, -0.04610027000308037, -0.040377046912908554, 0.312823623418808, -0.15650898218154907, -0.009523753076791763, -0.01001940481364727, 0.2053346335887909, 0.3080560863018036, -0.25814899802207947, -0.7119433283805847, 0.05824523791670799, 0.005247350782155991, -0.26256272196769714, -0.13132767379283905, 0.23769721388816833, 0.5601802468299866, -0.19879047572612762, 0.1145845577120781, 0.463534414768219, -0.18903324007987976, 0.09863808006048203, -0.150082066655159, -0.15123167634010315, -0.127501979470253, 0.1236405000090599, 0.20790447294712067, -0.006194116082042456, 0.07666584104299545, 0.5572870373725891, 0.16624771058559418, -0.1265098750591278, -0.35303959250450134, 0.07210410386323929, 0.27843520045280457, -0.3214607536792755, -0.09803184866905212, -0.108583465218544, -0.5646517276763916, 0.0625721663236618, -0.2864474356174469, 0.08351265639066696, 0.13649024069309235, -0.23094585537910461, -0.05809517204761505, -0.06265649199485779, -0.07724550366401672, -0.036541711539030075, 0.13375915586948395, -0.09502670913934708, -0.03345921263098717, 0.02000877819955349, 0.1139894649386406, 0.34533604979515076, -0.29028013348579407, 0.5099155306816101, 0.00863154698163271, 0.32485470175743103, -0.275591105222702, -0.3000114858150482, 0.13371877372264862, 0.09635912626981735, 0.08106764405965805, -0.16778746247291565, -0.14541804790496826, 0.2911043167114258, 0.19663816690444946, 0.031059395521879196, 0.14562590420246124, -0.2731468975543976, 0.4598880112171173, 0.058973051607608795, -0.03577909618616104, -0.13375790417194366, 0.08443677425384521, -0.1913497895002365, -0.04629846289753914, 0.23205755650997162, -0.24287986755371094, -0.04818802326917648, 0.11958298087120056, 0.1858397126197815, 0.04129062965512276, -0.13352659344673157, -0.15656790137290955, 0.09857823699712753, 0.23880945146083832, 0.20704039931297302, -0.0013628584565594792, -0.04337590932846069, -0.3433440625667572, 0.19563497602939606, 0.4620441794395447, -0.00548150110989809, 0.20487985014915466, -0.06295567750930786, -0.0692543163895607, 0.0003483508771751076, 0.4217230975627899, 0.04121643304824829, 0.4216999113559723, 0.08651913702487946, -0.09068877249956131, 0.000798987690359354, -0.03490288183093071, -0.3570864796638489, 0.16904479265213013, 0.1428687572479248, 0.2106133997440338, -0.1635540872812271, 0.22982488572597504, -0.0365557000041008, -0.39952921867370605, -0.10759526491165161, -0.08823571354150772, 0.1315503865480423, 0.1349104344844818, 0.32000792026519775, -0.11699079722166061, -0.5118412971496582, -0.07171622663736343, -0.18816904723644257, -0.09109190851449966, -0.2565024793148041, -0.04709184542298317, 0.5061329007148743, 0.3253306448459625, 0.013789165765047073, 0.34854328632354736, 0.34163278341293335, -0.0021227174438536167, 0.3129969537258148, 0.1184987723827362, -0.04346618056297302, -0.2894243896007538, 0.0792035162448883, 0.42273062467575073, 0.24148361384868622, 0.2128828465938568, -0.30292388796806335, -0.15141020715236664, -0.14659331738948822, -0.2868514060974121, 0.29411807656288147, -0.19982562959194183, 0.2345816045999527, -0.023697596043348312, 0.19926981627941132, -0.536819338798523, -0.1262858361005783, 0.15906767547130585, -0.010400800034403801, -0.08413990586996078, -0.05235574021935463, -0.01666237600147724, -0.4458804726600647, -0.20343227684497833, -0.1740431785583496, -0.39175131916999817, -0.24149520695209503, 0.1878705471754074, -0.11721907556056976, 0.18275432288646698, 0.23959949612617493, 0.06552783399820328, 0.40171727538108826, -0.03590897470712662, -0.24859125912189484, 0.01844647154211998, -0.12903721630573273, 0.2076355367898941, -0.323977530002594, -0.5115103125572205, 0.10641659051179886, 0.18794196844100952, 0.0032007477711886168, -0.2690906524658203, -0.4355989992618561, -0.37786349654197693, 0.049906183034181595, 0.1191340982913971, 0.02384091541171074, -0.2214813232421875, 0.005934369750320911, -0.2621510624885559, 0.04465482011437416, -0.010199579410254955, -0.27398818731307983, -0.111051045358181, 0.07669293880462646, 0.3078034520149231, -0.19712603092193604, 0.11028701066970825, 0.03152871131896973, 0.2529439926147461, 0.3045102655887604, -0.11898024380207062, 0.18053948879241943, 0.10017790645360947, 0.2179146111011505, -0.28155317902565, -0.4495452344417572, 0.33795154094696045, -0.16195033490657806, -0.06860882043838501, 0.3591616451740265, -0.09100805222988129, -0.5044241547584534, -0.17165997624397278, 0.07266724109649658, -0.2678212821483612, -0.4227607548236847, 0.07086485624313354, 0.44430267810821533, -0.14769341051578522, 0.1125858873128891, 0.21931485831737518, 0.019851921126246452, -0.22919774055480957, -0.31253305077552795, 0.33609867095947266, -0.28678032755851746, 0.0014983726432546973, 0.1557372510433197, -0.41713231801986694, -0.6907583475112915, 0.16188707947731018, -0.05036396160721779, 0.11516433209180832, -0.11067390441894531, 0.29069948196411133, 0.10505305975675583, 0.03925445303320885, 0.314164400100708, 0.11675331741571426, -0.15859423577785492, 0.04357824847102165, -0.07495138794183731, -0.19444723427295685, 0.03445037081837654, -0.4641056954860687, 0.5842929482460022, 0.7916570901870728, 0.39100542664527893, -0.04272054135799408, 0.14425696432590485, 0.5085096955299377, -0.032958224415779114, 0.15193390846252441, 0.08288563042879105, -0.14068742096424103, -0.30309998989105225, -0.12402485311031342, 0.059478890150785446, -0.02739955298602581, -0.060650475323200226, 0.03376597538590431, -0.2005617916584015, 0.30888575315475464, -0.07161397486925125, -0.026788007467985153, 0.5213570594787598, -0.019524237141013145, 0.08624424785375595, 0.10341477394104004, -0.10577686876058578, -0.11154290288686752, 0.4208712577819824, 0.30464038252830505, -0.20926688611507416, -0.267480731010437, 0.2845477759838104, -0.4900810718536377, 0.09635598957538605, 0.5848202109336853, -0.010691389441490173, -0.031012993305921555, -0.1739789843559265, 0.00506605813279748, 0.3778678774833679, 0.053109608590602875, 0.4014870226383209, 0.17706401646137238, -0.02953922003507614, 0.17760571837425232, 0.25741779804229736, -0.09426883608102798, 0.01886991783976555, 0.014035893604159355, -0.1412167251110077, -0.23207971453666687, 0.4488864839076996, 0.07712186127901077, 0.8232885599136353, -0.15118785202503204, 0.06643514335155487, 0.177730992436409, 0.005393131636083126, 0.555837094783783, -0.01587972417473793, 0.22724463045597076, -0.13967925310134888, -0.44501420855522156, -0.022555768489837646, -0.10461319237947464, 0.28905588388442993, -0.02663339301943779, -0.7058285474777222, 0.21495874226093292, 0.2657940089702606, 0.1526275873184204, 0.020415760576725006, 0.05576958879828453, -0.14507775008678436, -0.24285975098609924, -0.5719937682151794, 0.22100278735160828, -0.17519980669021606, 0.23505225777626038, -0.28548693656921387, 0.030615681782364845, -0.2551962435245514, 0.04343568906188011, -0.45206621289253235, 0.06690510362386703, -0.48229384422302246, 0.013023583218455315, 0.14497776329517365, -0.250901460647583, 0.04381668567657471, 0.34479543566703796, 0.3875301480293274, 0.17426273226737976, -0.043433044105768204, -0.018648680299520493, -0.035572800785303116, -0.30340999364852905, -0.06309986114501953, -0.016941165551543236, 0.005116963293403387, -0.26457956433296204, -0.06806784868240356, -0.07060003280639648, 0.30341196060180664, -0.149381622672081, 0.3579692840576172, 0.1762150675058365, 0.1282223016023636, -0.06678539514541626, -0.0537990964949131, -0.11600923538208008, -0.17481091618537903, -0.29133138060569763, 0.1321408599615097, -0.12050136923789978, -0.005070179235190153, 0.297807514667511, 0.18357324600219727, -0.12276753783226013, -0.03555615618824959, 0.2913230061531067, -0.2046392261981964, 0.10926888138055801, 0.21874330937862396, 0.25724443793296814, -0.12845036387443542, -0.2801275849342346, 0.3003304898738861, -0.49323737621307373, -0.17429231107234955, 0.11607915908098221, -0.17401978373527527, -0.12756574153900146, -0.13076519966125488, 0.6403629779815674, 0.4149584174156189, -0.010212050750851631, -0.06361083686351776, -0.36319300532341003, -0.20702576637268066, 0.20271053910255432, 0.25006306171417236, 0.025946833193302155, 0.22087973356246948, 0.19487763941287994, 0.2190665900707245, 0.0669330582022667, -0.270333468914032, 0.3732134997844696, -0.4460946023464203, 0.05523926764726639, -0.030021270737051964, -0.2469729334115982, 0.23904527723789215, -0.15189619362354279, -0.03373752906918526, 0.5021682977676392, 0.06566664576530457, -0.26115086674690247, -0.06453012675046921, 0.06806578487157822, 0.008530446328222752, -0.1281813681125641, -0.26094144582748413, 0.004728189669549465, -0.030353570356965065, -0.06656630337238312, 0.17197634279727936, -0.05886872857809067, 0.2700214982032776, -0.1260623335838318, 0.33104407787323, 0.2087761014699936, 0.15549935400485992, 0.084695003926754, 0.013366826809942722, 0.20232266187667847, 0.12882846593856812, 0.12136552482843399, 0.15800122916698456, -0.07078985869884491, 0.2570585608482361, 0.024877695366740227, -0.07617675513029099, -0.007841233164072037, 0.3335098326206207, -0.3386331796646118, 0.15942125022411346, 0.002658279612660408, -0.16646401584148407, 0.35182181000709534, -0.012282237410545349, -0.12231257557868958, 0.10254056006669998, 0.13919833302497864, -0.10381678491830826, -0.05152599886059761, 0.5680356621742249, 0.2006264626979828, 0.030121371150016785, 0.5001778602600098, 0.08568546921014786, -0.040030475705862045, 0.2109011560678482, 0.04401499778032303, 0.31992053985595703, 0.4771452844142914, 0.07485044002532959, -0.11321049183607101, 0.025738954544067383, 0.23483607172966003, 0.11907744407653809, 0.056440867483615875, 0.16241033375263214, -0.04504365101456642, -0.05025982856750488, 0.11100412905216217, -0.17791953682899475, 0.08935986459255219, 0.4992905557155609, -0.10340705513954163, -0.22849811613559723, 0.04294434189796448, 0.10159790515899658, 0.027173684909939766, 0.12443159520626068, 0.7389979362487793, -0.48798155784606934, -0.2675541043281555, -0.1957969069480896, 0.14382854104042053, -0.4215509295463562, 0.08674003183841705, -0.31011253595352173, 0.06441132724285126, -0.09468212723731995, -0.0019607411231845617, -0.1672721803188324, -0.0573422834277153, 0.395906537771225, 0.2621210217475891, -0.19068604707717896, 0.03612058237195015, -0.1225874274969101, 0.10783533751964569, -0.1399857997894287, -0.012021651491522789, 0.4424281418323517, -0.30636027455329895, 0.2628047466278076, 0.014340170659124851, 0.6072162985801697, 0.18633610010147095, 0.20774538815021515, -0.10181611776351929, -0.22457700967788696, 0.06042502447962761, -0.08757283538579941, 0.012050033546984196, -0.05506788566708565, 0.44311845302581787, 0.08779517561197281, 0.26832979917526245, 0.08101984858512878, -0.1402062326669693, 0.22144781053066254, 0.12705063819885254, -0.18602299690246582, 0.03443322703242302, -0.298171728849411, -0.09205526858568192, -0.1460096836090088, -0.2923966348171234, -0.17628230154514313, -0.304012268781662, 0.4630804657936096, -0.23748011887073517, -0.15882493555545807, 0.09287761151790619, 0.026171907782554626, 0.07179026305675507, 0.0003693557227961719, 0.5908597111701965, 0.3205853998661041, -0.001303468132391572, -0.2988016903400421, -0.5059593915939331, -0.3325134813785553, 0.19119055569171906, -0.16133591532707214, 0.13938651978969574, 0.08501020073890686, 0.18678277730941772, -0.02311491034924984, 0.45860755443573, 0.16402624547481537, -0.08083488792181015, -0.21005937457084656, -0.06281007826328278, -0.5517159700393677, -0.13538366556167603, 0.2651442587375641, -0.2606630325317383, -0.16066497564315796, -0.16178695857524872, 0.289763867855072, -0.15786464512348175, 0.008336842991411686, -0.04479442536830902, -0.27315452694892883, -0.33197641372680664, -0.24862368404865265, 0.15242041647434235, 0.129592627286911, 0.12297084927558899, 0.007584733422845602, -0.33314165472984314, 0.12094765901565552, -0.28892138600349426, -0.10682787746191025, 0.01142647210508585, 0.2162543386220932, 0.2503010332584381, 0.028534255921840668, 0.2635626196861267, -0.09871041029691696, 0.4951653778553009, 0.15883506834506989, 0.028559766709804535, -0.16628409922122955, -0.08736904710531235, -0.07266148179769516, 0.05997529998421669, -0.04567220062017441, 0.40819427371025085, -0.32333138585090637, 0.26766663789749146, -0.4084075391292572, -0.27963072061538696, 0.3924650251865387, -0.3721626400947571, -0.07345755398273468, -0.017433881759643555, 0.21171970665454865, 0.13282810151576996, -0.1360815316438675, -0.32911038398742676, -0.07494758069515228, -0.11623644083738327, -0.29002681374549866, -0.45486101508140564, 0.26036709547042847, -0.07690222561359406, 0.06409205496311188, -0.10680917650461197, 0.340888112783432, 0.12940239906311035, -0.28295955061912537, -0.1672976016998291, -0.2179582267999649 ]
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) ```
[ -0.21085909008979797, 0.31306004524230957, -0.04938048496842384, 0.03972192853689194, -0.0872686505317688, -0.1855650693178177, 0.26312482357025146, 0.1595553308725357, -0.44758906960487366, 0.1087360829114914, -0.3607417345046997, 0.5733001828193665, 0.19588503241539001, 0.08197695016860962, 0.17324359714984894, 0.1950279176235199, -0.008884613402187824, 0.39630308747291565, -0.05746747553348541, 0.16443295776844025, -0.2231675684452057, 0.34456780552864075, -0.3933256268501282, 0.10321464389562607, -0.2219560295343399, -0.19702449440956116, -0.1591430902481079, 0.29760807752609253, -0.39194455742836, -0.21676357090473175, 0.2760631740093231, 0.1715928018093109, -0.0599934384226799, 0.1792944073677063, -0.00011473386257421225, -0.20161768794059753, -0.030263284221291542, -0.07803510129451752, -0.3729618787765503, -0.2322017401456833, 0.03527377173304558, -0.33603227138519287, 0.3040671944618225, -0.44802847504615784, 0.13219548761844635, 0.4575708508491516, 0.2886238396167755, 0.3576168119907379, 0.29694080352783203, 0.27599725127220154, 0.164218470454216, 0.32642704248428345, 0.13450884819030762, 0.005115279462188482, 0.2324996143579483, -0.25311291217803955, -0.26933562755584717, 0.42635348439216614, 0.5949572920799255, -0.26348528265953064, -0.13266949355602264, 0.10558642446994781, -0.16910988092422485, 0.15007275342941284, 0.110907644033432, 0.2283693552017212, -0.34415674209594727, -0.4259018003940582, 0.22296039760112762, 0.06068377569317818, 0.026528431102633476, -0.26247650384902954, 0.054827816784381866, 0.005878469906747341, 0.2763141691684723, -0.30670443177223206, 0.15786677598953247, 0.35337477922439575, -0.37504321336746216, 0.11998174339532852, 0.024275759235024452, 0.03381401672959328, -0.1993512660264969, 0.185835063457489, 0.058839619159698486, 0.3432292342185974, -0.024360572919249535, -0.013874203898012638, 0.21716898679733276, 0.08212466537952423, 0.3446107506752014, 0.2156180888414383, -0.03924698755145073, -0.0030003166757524014, -0.1363028585910797, -0.20195701718330383, -0.14123892784118652, 0.2344246804714203, 0.23353028297424316, -0.32346656918525696, 0.2888829708099365, 0.006387042813003063, 0.29112687706947327, 0.19484849274158478, -0.058986078947782516, -0.11289205402135849, 0.21247805655002594, -0.13682161271572113, 0.396963506937027, 0.3778127431869507, -0.10544463992118835, 0.0028239355888217688, -0.04434072598814964, -0.16905811429023743, -0.15089064836502075, -0.10000812262296677, 0.12246556580066681, -0.17493703961372375, -0.2805671691894531, 0.02378266118466854, -0.22518065571784973, 0.29884621500968933, 0.0426323376595974, 0.24172687530517578, -0.061569150537252426, -0.376302033662796, 0.066953144967556, -0.025022434070706367, -0.34266889095306396, -0.20411549508571625, -0.002302655251696706, -0.0190372746437788, -0.16797620058059692, 0.09720920771360397, 0.30007508397102356, 0.1617031693458557, 0.4099462628364563, -0.118857741355896, 0.17195527255535126, -0.10903356224298477, 0.03236829489469528, -0.2970986068248749, -0.09300358593463898, 0.2763931155204773, 0.3880445659160614, -0.2414582520723343, -0.22902102768421173, -0.5571025609970093, 0.04567006975412369, 0.16808457672595978, -0.012525117956101894, -0.008397676981985569, -0.15670840442180634, 0.1424696445465088, -0.2702634036540985, -0.04810592532157898, -0.3073248267173767, 0.010949683375656605, -0.2560480237007141, -0.18128342926502228, 0.031656838953495026, -0.07667215168476105, -0.1764993965625763, -0.07915304601192474, 0.0010232838103547692, 0.19804725050926208, -0.1631060242652893, -0.03464571386575699, 0.12622450292110443, 0.03138294070959091, 0.047350894659757614, 0.23515042662620544, -0.18265117704868317, 0.05419299751520157, 0.028839074075222015, 0.22627660632133484, 0.4111820161342621, -0.2042841613292694, -0.6995334029197693, -0.007170289289206266, -0.08847208321094513, -0.24175748229026794, -0.13286654651165009, 0.24760930240154266, 0.4465601444244385, -0.14351347088813782, 0.21046534180641174, 0.5739886164665222, -0.16754508018493652, 0.14042896032333374, -0.2227223813533783, -0.1723984330892563, -0.06703267246484756, 0.18465381860733032, 0.14324398338794708, -0.10967383533716202, 0.01193225383758545, 0.6371188759803772, 0.23053859174251556, -0.11815842986106873, -0.2321152538061142, -0.004396307282149792, 0.26636746525764465, -0.30782264471054077, -0.10001222789287567, -0.0542004257440567, -0.5955289006233215, 0.03019573725759983, -0.27251413464546204, 0.19959956407546997, -0.012548868544399738, -0.19998207688331604, -0.03226301446557045, 0.0027679954655468464, -0.021475598216056824, -0.04150375351309776, 0.14903375506401062, -0.10393498092889786, 0.03776169568300247, 0.01641099527478218, 0.13712650537490845, 0.3391304016113281, -0.23257866501808167, 0.4468238055706024, -0.003963662311434746, 0.257414847612381, -0.28839805722236633, -0.32532405853271484, 0.15580102801322937, 0.2438877671957016, 0.030585313215851784, -0.1248808354139328, -0.06957949697971344, 0.21655496954917908, 0.1029074564576149, -0.08505726605653763, 0.06705223768949509, -0.33819201588630676, 0.37754976749420166, 0.08620857447385788, -0.052933190017938614, -0.0791063979268074, 0.076689213514328, -0.07673192769289017, 0.01853305846452713, 0.22335053980350494, -0.31230780482292175, -0.07031110674142838, 0.10400642454624176, 0.2196219563484192, 0.06820761412382126, -0.13731424510478973, -0.12774144113063812, 0.21088212728500366, 0.22927634418010712, 0.24448968470096588, 0.015721047297120094, 0.03844160586595535, -0.36530575156211853, 0.27148905396461487, 0.4710395634174347, 0.009045137092471123, 0.2805318832397461, -0.027613872662186623, -0.0678357258439064, -0.017648344859480858, 0.3719349801540375, 0.02318975329399109, 0.3294188976287842, 0.11665693670511246, -0.11657088994979858, 0.01113898679614067, -0.10210783779621124, -0.2971956729888916, 0.11165916919708252, 0.12786349654197693, 0.317874550819397, -0.10101085156202316, 0.3305141031742096, -0.03437204286456108, -0.343092143535614, -0.20008864998817444, -0.09329670667648315, 0.24215541779994965, 0.1284601390361786, 0.2591361403465271, -0.13541024923324585, -0.6098312735557556, -0.055080004036426544, -0.16726812720298767, -0.09036044031381607, -0.22800865769386292, -0.023899495601654053, 0.45616260170936584, 0.2761538028717041, 0.07358157634735107, 0.29388219118118286, 0.21407940983772278, 0.03674884885549545, 0.08619970083236694, 0.15598152577877045, -0.11963656544685364, -0.31850942969322205, 0.07188040018081665, 0.39522692561149597, 0.24352456629276276, 0.08686234056949615, -0.24151521921157837, -0.1873188316822052, -0.0984908863902092, -0.3265637457370758, 0.2032727301120758, -0.23567259311676025, 0.3059912323951721, -0.017746977508068085, 0.23887324333190918, -0.6264254450798035, -0.21012365818023682, 0.10280899703502655, -0.12800590693950653, -0.11295338720083237, -0.03887337073683739, -0.005858792923390865, -0.36101672053337097, -0.24185678362846375, -0.14751116931438446, -0.5018798112869263, -0.24258743226528168, 0.25749287009239197, -0.003788007888942957, 0.19984537363052368, 0.27991941571235657, 0.07124165445566177, 0.31753045320510864, 0.09950869530439377, -0.23881185054779053, 0.06216580048203468, 0.02206621691584587, 0.23596236109733582, -0.2466728538274765, -0.5061877369880676, 0.0659509226679802, 0.05682864040136337, 0.0324612520635128, -0.2706144154071808, -0.4002764821052551, -0.4651819169521332, 0.1409192681312561, 0.2356443703174591, -0.04908060282468796, -0.13650183379650116, 0.1378428190946579, -0.1899326890707016, 0.07168599218130112, 0.015008172020316124, -0.23039887845516205, -0.17598670721054077, 0.13017921149730682, 0.28328385949134827, -0.2543368935585022, 0.045959871262311935, 0.1570533961057663, 0.22293902933597565, 0.11176837980747223, -0.08313310891389847, 0.15639114379882812, 0.09553362429141998, 0.19741588830947876, -0.18021893501281738, -0.4202902019023895, 0.33444973826408386, -0.11916851997375488, -0.09323758631944656, 0.3436354994773865, -0.1598416417837143, -0.6042139530181885, -0.1298762559890747, 0.06717195361852646, -0.3866879642009735, -0.46388742327690125, 0.16150760650634766, 0.5007525086402893, -0.13982725143432617, 0.2503451704978943, 0.18707333505153656, 0.008182629942893982, -0.3484296202659607, -0.37130990624427795, 0.22920942306518555, -0.26665541529655457, 0.04256104305386543, 0.20368079841136932, -0.4245848059654236, -0.6710538864135742, 0.18562451004981995, 0.07730108499526978, 0.12946687638759613, 0.033733587712049484, 0.16176477074623108, 0.14359435439109802, 0.010684258304536343, 0.20230995118618011, 0.23272603750228882, -0.043824534863233566, 0.08600259572267532, 0.05781108886003494, -0.2841726243495941, 0.027078626677393913, -0.39344489574432373, 0.5534499287605286, 0.8214974999427795, 0.217286616563797, -0.12232992053031921, 0.16169331967830658, 0.5125232338905334, -0.025665180757641792, 0.16802284121513367, 0.13292109966278076, -0.22827956080436707, -0.3713435232639313, -0.14193885028362274, 0.10894772410392761, -0.058419711887836456, 0.0799488052725792, 0.09687243402004242, -0.06940150260925293, 0.2451847344636917, -0.07987092435359955, -0.05815887823700905, 0.5733699202537537, -0.043096788227558136, 0.17280882596969604, 0.0742165744304657, -0.18566344678401947, -0.1351664513349533, 0.3057752847671509, 0.25499993562698364, -0.04839923977851868, -0.2348625361919403, 0.22553174197673798, -0.317595899105072, 0.1691054254770279, 0.5100568532943726, -0.028746088966727257, -0.008707733824849129, -0.2693239450454712, 0.009692060761153698, 0.3701845705509186, 0.08202012628316879, 0.25667110085487366, 0.07617249339818954, -0.08871470391750336, 0.20067395269870758, 0.23740257322788239, -0.12386061996221542, 0.024681853130459785, -0.15152516961097717, -0.19545289874076843, -0.2053944617509842, 0.43531274795532227, 0.09283728897571564, 0.7878350019454956, -0.23235712945461273, 0.16138100624084473, 0.27821919322013855, 0.029196254909038544, 0.6408745050430298, -0.03151531517505646, 0.33673030138015747, -0.18474219739437103, -0.2963075637817383, 0.05368510261178017, -0.021155748516321182, 0.1532129943370819, 0.01419666688889265, -0.605322539806366, 0.23526039719581604, 0.3834209144115448, 0.029827147722244263, 0.08889066427946091, 0.1362038403749466, -0.1128024011850357, -0.3289937376976013, -0.638590395450592, 0.17903031408786774, -0.3138793706893921, 0.19821514189243317, -0.30542832612991333, 0.0033898784313350916, -0.2572369873523712, 0.05614398047327995, -0.3699459135532379, 0.15100403130054474, -0.5142828822135925, -0.048113662749528885, 0.20631751418113708, -0.3006668984889984, -0.10240323841571808, 0.3006402552127838, 0.1783236861228943, 0.2046705186367035, 0.060395367443561554, 0.06155688315629959, -0.05859052762389183, -0.30445829033851624, 0.08371353149414062, 0.037178412079811096, 0.07622481137514114, -0.20884211361408234, -0.08287835121154785, -0.0731535255908966, 0.27861684560775757, -0.23776979744434357, 0.38632020354270935, 0.27965182065963745, 0.1820627599954605, 0.020237941294908524, -0.06534849852323532, -0.03768235072493553, -0.1724325716495514, -0.280821830034256, 0.10838425904512405, -0.10882590711116791, -0.06522040069103241, 0.2896272540092468, 0.14065712690353394, -0.1736132651567459, -0.003592650406062603, 0.2703544795513153, -0.17901000380516052, 0.10638611763715744, 0.2071494609117508, 0.2732165455818176, -0.11991041898727417, -0.21892866492271423, 0.19247572124004364, -0.5636967420578003, -0.17945489287376404, 0.029461203143000603, -0.07871950417757034, -0.02636132389307022, -0.08183694630861282, 0.631582498550415, 0.30250146985054016, 0.02565673366189003, -0.07841389626264572, -0.40588268637657166, -0.1844434142112732, 0.2054409682750702, 0.1786111444234848, -0.1519635021686554, 0.376731276512146, 0.3331975042819977, 0.19334203004837036, 0.048748679459095, -0.23890846967697144, 0.29427945613861084, -0.4090560972690582, 0.16265273094177246, -0.05745488032698631, -0.31670475006103516, 0.18075703084468842, -0.16345804929733276, -0.00683080218732357, 0.5193843841552734, 0.12761375308036804, -0.23762638866901398, -0.0832374095916748, 0.06300914287567139, 0.020876117050647736, -0.10651829838752747, -0.3156396746635437, -0.07924643158912659, -0.03797794133424759, -0.0308688972145319, 0.09227924793958664, -0.11090970784425735, 0.2121519297361374, -0.04508564993739128, 0.25605419278144836, 0.22376346588134766, 0.18538798391819, 0.03360747918486595, -0.02662513591349125, 0.25719955563545227, 0.10353595018386841, 0.20876501500606537, 0.17264549434185028, 0.038589537143707275, 0.22618533670902252, -0.04137703403830528, -0.16636620461940765, 0.07988866418600082, 0.3123323321342468, -0.3684277832508087, 0.06631990522146225, -0.12071608752012253, -0.0833493024110794, 0.3914594352245331, 0.09454575926065445, -0.22146707773208618, 0.24221889674663544, 0.12057970464229584, -0.20341162383556366, 0.05944071337580681, 0.5620883703231812, 0.19853384792804718, 0.11759084463119507, 0.5118356347084045, 0.0033111339434981346, -0.00012543363845907152, 0.14947061240673065, 0.11183508485555649, 0.290325403213501, 0.4005495309829712, -0.05944674462080002, -0.10892600566148758, -0.04077716916799545, 0.10463778674602509, 0.12247251719236374, 0.10486507415771484, 0.18972018361091614, -0.003135223174467683, -0.08938495814800262, 0.09117861092090607, -0.21281328797340393, 0.07399356365203857, 0.34565040469169617, -0.07070349156856537, -0.27605220675468445, 0.06119923293590546, 0.22912633419036865, 0.062462467700242996, 0.07917772233486176, 0.7251801490783691, -0.34647995233535767, -0.22814308106899261, -0.02409788779914379, 0.1677074283361435, -0.39005324244499207, 0.11122935265302658, -0.25787630677223206, 0.10365495830774307, -0.07326430082321167, 0.06669112294912338, -0.15407663583755493, -0.13330839574337006, 0.35017356276512146, 0.29399871826171875, -0.17318908870220184, -0.024732787162065506, -0.10646174848079681, 0.16021224856376648, -0.10827101767063141, -0.023033805191516876, 0.3853769302368164, -0.34098491072654724, 0.19518731534481049, -0.18485867977142334, 0.5654686689376831, 0.17510510981082916, 0.18691027164459229, -0.25897037982940674, -0.19222985208034515, -0.046545252203941345, -0.009189996868371964, -0.129123255610466, -0.10915730893611908, 0.48110443353652954, 0.17956112325191498, 0.25972631573677063, 0.0710274800658226, -0.13908205926418304, 0.2577846646308899, 0.19828620553016663, -0.2809127867221832, 0.1496541053056717, -0.2773095667362213, 0.04808780178427696, -0.17036636173725128, -0.25108009576797485, -0.10992149263620377, -0.2670920193195343, 0.4196455180644989, -0.26302942633628845, -0.08880860358476639, 0.08305783569812775, -0.06837176531553268, 0.06529147922992706, 0.03694207966327667, 0.5547950267791748, 0.3388826847076416, 0.017961187288165092, -0.25608721375465393, -0.40430763363838196, -0.3963001072406769, 0.22965607047080994, -0.22898238897323608, -0.07774484157562256, 0.08669256418943405, 0.29644253849983215, -0.050080690532922745, 0.4934329390525818, 0.014594274573028088, 0.044201549142599106, -0.1833358108997345, -0.009463395923376083, -0.5380406975746155, -0.06271477043628693, 0.2135578691959381, -0.33248093724250793, -0.0747075304389, -0.2555917799472809, 0.22630348801612854, -0.34457799792289734, 0.011229979805648327, -0.022472137585282326, -0.37018170952796936, -0.24126379191875458, -0.08207998424768448, 0.17296914756298065, 0.07968711853027344, 0.06808541715145111, 0.012299931608140469, -0.233347088098526, 0.03424585238099098, -0.25865328311920166, -0.20849579572677612, 0.040847815573215485, 0.20121583342552185, 0.22902457416057587, 0.08602122962474823, 0.36334484815597534, -0.09533344209194183, 0.515275239944458, 0.09685730934143066, 0.04516338184475899, -0.04169933870434761, -0.024246882647275925, -0.12852972745895386, 0.1383090764284134, -0.2299615889787674, 0.44079646468162537, -0.3113742768764496, 0.2767089307308197, -0.485502153635025, -0.3400692045688629, 0.363351970911026, -0.5322355628013611, -0.1876486986875534, 0.04002273082733154, 0.2281879335641861, 0.08981304615736008, -0.17658738791942596, -0.2977266013622284, -0.024630488827824593, -0.11291182041168213, -0.23598921298980713, -0.41634902358055115, 0.23685212433338165, -0.18495163321495056, 0.04134868085384369, -0.09677941352128983, 0.2848266363143921, 0.038823384791612625, -0.24422140419483185, -0.2131585031747818, -0.21907654404640198 ]
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 ?
[ -0.06062937527894974, -0.4335717558860779, -0.05662928894162178, -0.015485304407775402, -0.16559097170829773, 0.04524773731827736, 0.18558689951896667, 0.18404969573020935, 0.20415976643562317, -0.14907827973365784, -0.24861840903759003, 0.2844122648239136, -0.1800074279308319, 0.4715519845485687, 0.3802827298641205, -0.43187224864959717, 0.07262536138296127, -0.14835748076438904, -0.3279542028903961, 0.3465223014354706, 0.0434320867061615, 0.12348020076751709, -0.18072368204593658, 0.029902249574661255, -0.351310670375824, -0.16163283586502075, -0.3850715756416321, 0.07486454397439957, 0.014117834158241749, 0.03146935626864433, 0.06988033652305603, 0.3947068452835083, -0.041643787175416946, 0.1824844926595688, -0.00011498481035232544, -0.10626960545778275, 0.11025400459766388, -0.10874622315168381, 0.09373383224010468, -0.03407788276672363, 0.13381801545619965, -0.37906506657600403, 0.05808848887681961, -0.1654161959886551, 0.2026762217283249, 0.14574675261974335, 0.19343502819538116, 0.05130370333790779, 0.36705654859542847, -0.21853332221508026, 0.08402936160564423, 0.2644318640232086, -0.13484986126422882, 0.3218310475349426, -0.06517389416694641, 0.02047051675617695, 0.04488581046462059, 0.23994137346744537, 0.6195874214172363, -0.47857505083084106, -0.02777704782783985, 0.22522099316120148, -0.011841692961752415, -0.012316335923969746, 0.12501248717308044, -0.0638716071844101, -0.2911248505115509, -0.23895962536334991, -0.3664330542087555, 0.24455377459526062, -0.04346802830696106, -0.2190152108669281, -0.22686222195625305, -0.5487810373306274, -0.025223135948181152, -0.016629589721560478, -0.32687970995903015, 0.06366670876741409, -0.16956277191638947, -0.05313924700021744, -0.16527920961380005, -0.07771022617816925, -0.06440581381320953, 0.13073144853115082, 0.29616212844848633, 0.5442550778388977, 0.038709983229637146, 0.05356621369719505, 0.29034164547920227, 0.036765795201063156, -0.11274223029613495, -0.2751784026622772, 0.30769824981689453, 0.07872707396745682, -0.3209538757801056, -0.34821996092796326, 0.06954455375671387, -0.3125927448272705, 0.3218032121658325, 0.08210963010787964, 0.13770167529582977, 0.06355176120996475, -0.1306171864271164, 0.217574805021286, 0.3832522928714752, 0.008566949516534805, -0.07508260756731033, -0.16860586404800415, 0.007640562951564789, -0.14267075061798096, 0.16614334285259247, 0.1513281911611557, 0.11848489940166473, -0.004000337794423103, -0.44975146651268005, -0.062116339802742004, -0.07101401686668396, 0.05218591168522835, -0.12578363716602325, -0.5080181956291199, -0.11419264227151871, -0.09674479067325592, -0.11398360878229141, -0.1760413497686386, -0.26717719435691833, 0.17692410945892334, 0.0036280399654060602, 0.2718092203140259, -0.16389109194278717, 0.03908415138721466, -0.0379803329706192, 0.01682925783097744, -0.4351593852043152, -0.08235824853181839, 0.28147652745246887, 0.0876089558005333, 0.0532539039850235, 0.16166582703590393, -0.04158475250005722, 0.06549903005361557, 0.33225399255752563, 0.028684379532933235, 0.1381813883781433, -0.045800529420375824, 0.1786007434129715, -0.2670106589794159, -0.010549972765147686, 0.15709353983402252, -0.31846240162849426, -0.12240981310606003, -0.1524713784456253, -0.208932563662529, 0.05303369462490082, 0.010931088589131832, -0.1749030500650406, -0.28919899463653564, -0.4556143879890442, 0.7248806357383728, 0.015413815155625343, -0.08319570124149323, -0.19418630003929138, 0.1386072039604187, -0.15149904787540436, -0.0932871624827385, 0.22360371053218842, 0.11550801992416382, -0.0781726986169815, -0.2775520980358124, 0.019101450219750404, -0.08028506487607956, 0.012738131918013096, 0.3045518696308136, -0.06466130912303925, 0.2548964023590088, 0.06150007247924805, 0.20962834358215332, 0.30092260241508484, -0.3996352553367615, -0.0024378541857004166, 0.3445945084095001, -0.055489394813776016, 0.3846747577190399, 0.10268251597881317, 0.1290329545736313, 0.0022140536457300186, -0.19203856587409973, 0.2654208540916443, 0.9147470593452454, -0.3811211884021759, 0.06711137294769287, -0.19821710884571075, -0.3537558913230896, 0.5127942562103271, 0.2841840982437134, 0.011132043786346912, -0.34456250071525574, -0.13308796286582947, -0.004995118826627731, 0.07843330502510071, 0.0030070978682488203, 0.037370674312114716, -0.02651924453675747, -0.3109758496284485, 0.07546742260456085, -0.19495417177677155, -0.11524131149053574, -0.42474639415740967, -0.1003589779138565, 0.04577907547354698, 0.321023553609848, 0.26057329773902893, -0.15339210629463196, 0.22681349515914917, -0.30863356590270996, -0.15567195415496826, -0.2565404176712036, 0.13309966027736664, -0.026752345263957977, -0.014473693445324898, -0.20608538389205933, -0.09645254164934158, 0.32367759943008423, 0.05779195949435234, -0.075899139046669, -0.4959027171134949, 0.24824398756027222, -0.07776227593421936, 0.02005738765001297, 0.07954811304807663, 0.6802392601966858, -0.18799695372581482, -0.07588975131511688, 0.18547269701957703, 0.26534152030944824, -0.3162192702293396, 0.25385814905166626, 0.22005975246429443, 0.33087339997291565, 0.21403628587722778, -0.06979615986347198, 0.1053735762834549, -0.06378693133592606, -0.041656702756881714, -0.22341929376125336, -0.19158971309661865, 0.43757519125938416, -0.17595274746418, 0.46365219354629517, -0.03217601776123047, 0.02915242128074169, -0.22143889963626862, -0.01690278761088848, -0.2342645823955536, 0.3847416341304779, 0.3131430447101593, -0.20771299302577972, 0.3308095335960388, -0.04697221890091896, -0.4412024915218353, 0.2577657997608185, 0.24474403262138367, 0.06818308681249619, 0.09573809802532196, -0.00955189112573862, -0.045534323900938034, -0.33494138717651367, -0.06029945984482765, 0.3420468866825104, 0.566644012928009, 0.2699572443962097, 0.19837486743927002, 0.06839548796415329, -0.2664467990398407, -0.11241088062524796, 0.06929805129766464, 0.16281481087207794, 0.012806328944861889, 0.27648890018463135, 0.035906799137592316, 0.07019905745983124, -0.016358444467186928, -0.09048362076282501, 0.1398392915725708, -0.1848675012588501, -0.15955159068107605, 0.013511033728718758, -0.23286251723766327, -0.6612973213195801, -0.26553869247436523, 0.05956566333770752, -0.3062056005001068, -0.319520503282547, 0.12936240434646606, 0.017231881618499756, -0.14742326736450195, 0.17125511169433594, 0.23219703137874603, 0.6074469089508057, -0.3907059133052826, -0.1505921632051468, -0.0688081607222557, -0.33188149333000183, -0.14265243709087372, 0.07787401229143143, 0.549889862537384, 0.36540573835372925, 0.43545612692832947, 0.17507435381412506, 0.0016455785371363163, 0.040315236896276474, -0.41975587606430054, 0.09459847956895828, -0.3344561457633972, 0.16005411744117737, 0.06284262239933014, -0.16257420182228088, 0.14567060768604279, -0.42311936616897583, -0.08997105062007904, 0.024390358477830887, -0.09786451607942581, 0.017521440982818604, -0.03967594727873802, 0.027871422469615936, -0.2234395146369934, -0.30245500802993774, -0.5716315507888794, -0.45202386379241943, 0.3431040048599243, -0.2798544764518738, 0.14475853741168976, 0.06941840797662735, -0.22567540407180786, 0.14136306941509247, -0.17480282485485077, 0.07033861428499222, -0.11356338113546371, -0.10608287900686264, 0.04219140112400055, -0.3436841070652008, -0.24148795008659363, -0.38526850938796997, -0.026993097737431526, 0.17016872763633728, -0.0015303750988095999, -0.06010867655277252, -0.23663610219955444, 0.0734778568148613, -0.21928179264068604, 0.11132663488388062, 0.29573217034339905, 0.13768965005874634, -0.16666589677333832, 0.1337781846523285, -0.04997451975941658, -0.18687556684017181, 0.2078094333410263, 0.3281923830509186, 0.38636574149131775, 0.15270426869392395, 0.001718635787256062, 0.2775654196739197, 0.6705095171928406, 0.3482151925563812, 0.04389117285609245, 0.09031028300523758, 0.25889766216278076, 0.06046077609062195, 0.12866388261318207, -0.1806710958480835, 0.2668643295764923, -0.20002710819244385, 0.5133736729621887, 0.15122029185295105, 0.18066485226154327, -0.10541104525327682, -0.21315091848373413, 0.12126965075731277, -0.35108867287635803, -0.19965636730194092, 0.3020295798778534, -0.31928253173828125, 0.02158222161233425, -0.054490428417921066, -0.3992008566856384, -0.25336796045303345, -0.04955427721142769, 0.01742720790207386, 0.35206514596939087, -0.05508112162351608, 0.19885975122451782, -0.4190731346607208, -0.09170432388782501, -0.46005377173423767, 0.19697122275829315, 0.08974100649356842, 0.33363214135169983, -0.06856129318475723, -0.11177212744951248, -0.08517727255821228, 0.13111209869384766, 0.5710866451263428, -0.5576233267784119, -0.22170650959014893, 0.17702369391918182, 0.15061673521995544, -0.08356267958879471, -0.12241511046886444, -0.3805668354034424, -0.026714833453297615, 0.4663636386394501, 0.425428569316864, -0.4082852005958557, -0.3596429228782654, 0.2669551372528076, -0.16865256428718567, -0.02068937197327614, -0.13011783361434937, -0.20354966819286346, -0.21533091366291046, -0.17324471473693848, 0.02658178098499775, 0.2245217114686966, 0.16896402835845947, -0.15442802011966705, 0.1507239192724228, -0.046021848917007446, 0.2372143268585205, 0.21118460595607758, -0.0128335515037179, -0.05038948357105255, 0.2638845443725586, 0.0650329738855362, 0.0724661648273468, -0.03061080165207386, -0.2702558636665344, 0.10697241127490997, -0.13292692601680756, 0.06653118133544922, 0.27353909611701965, -0.07887066155672073, 0.7340446710586548, 0.1507011204957962, 0.38636553287506104, 0.17446652054786682, -0.10423657298088074, 0.02026847004890442, -0.24745479226112366, 0.21436716616153717, 0.20877261459827423, 0.24122527241706848, -0.1877690851688385, -0.19011333584785461, 0.5577905178070068, 0.33030596375465393, -0.015693049877882004, 0.4144617021083832, -0.4715425670146942, -0.2955450415611267, 0.31731587648391724, 0.03746155649423599, 0.7747875452041626, -0.05594339594244957, 0.3117319345474243, 0.08896902203559875, -0.4208650290966034, 0.5706253051757812, -0.4346333146095276, 0.15796616673469543, -0.09969218075275421, -0.18400438129901886, 0.030260618776082993, -0.10041109472513199, -0.026498878374695778, 0.25339803099632263, -0.38214874267578125, 0.3142700493335724, 0.42039427161216736, -0.2569933831691742, -0.03189197927713394, 0.5367090106010437, -0.05316834896802902, -0.39282992482185364, -0.056313179433345795, 0.041811503469944, -0.32738611102104187, 0.08863341063261032, -0.12786121666431427, -0.3133982717990875, 0.23929095268249512, 0.15863993763923645, -0.3130236566066742, 0.10464876145124435, -0.06307264417409897, 0.14268921315670013, 0.22421936690807343, -0.035015080124139786, 0.24764733016490936, -0.13931336998939514, 0.042993586510419846, -0.13076253235340118, -0.06671719998121262, 0.03300977870821953, -0.10206122696399689, 0.07684879004955292, 0.0985972210764885, -0.16361360251903534, 0.5141544342041016, -0.00334109365940094, 0.008826040662825108, -0.2976304292678833, 0.2612650692462921, -0.38956311345100403, -0.27695080637931824, -0.02467651478946209, 0.18370650708675385, 0.09997136890888214, -0.2911476194858551, 0.6143853068351746, 0.08731158077716827, 0.03576843440532684, 0.040252119302749634, 0.13858170807361603, -0.23400595784187317, 0.5710297226905823, -0.2432594895362854, -0.014290206134319305, 0.08371669799089432, -0.01370136346668005, -0.11303955316543579, -0.4529513716697693, -0.020737536251544952, 0.3857307732105255, 0.07245339453220367, -0.022204987704753876, 0.06766756623983383, 0.13055846095085144, -0.06083383038640022, -0.017283093184232712, -0.2517121136188507, -0.2992801070213318, 0.102488212287426, 0.24880841374397278, 0.24665343761444092, 0.15516231954097748, -0.04194396361708641, -0.24314746260643005, -0.13975271582603455, 0.5461681485176086, -0.030902022495865822, 0.26203489303588867, -0.05144912004470825, 0.3808259069919586, 0.38158249855041504, 0.5050756931304932, -0.25075966119766235, 0.07350720465183258, 0.06648756563663483, 0.19243542850017548, 0.4419762194156647, -0.2678069770336151, -0.032334040850400925, 0.2462569922208786, -0.13508741557598114, 0.10860057920217514, -0.3245002031326294, -0.15105053782463074, -0.20546084642410278, 0.10554119199514389, 0.19726668298244476, 0.029151109978556633, 0.04750947654247284, -0.03353991359472275, -0.13116854429244995, -0.10678403824567795, 0.02167419157922268, -0.3718712627887726, -0.12796536087989807, 0.23094752430915833, -0.118099145591259, 0.03677048161625862, 0.0795663520693779, 0.13500705361366272, 0.004463936667889357, -0.09252690523862839, 0.31753668189048767, 0.09242815524339676, 0.06634542346000671, -0.18915191292762756, -0.18550194799900055, 0.04238720238208771, 0.23217514157295227, 0.18918229639530182, 0.002319952240213752, 0.0865117534995079, -0.17059379816055298, -0.16673637926578522, 0.2126782387495041, 0.2336454838514328, 0.010325543582439423, -0.09136775135993958, 0.1436551809310913, 0.09549890458583832, -0.054430946707725525, -0.005216358695179224, 0.2259073555469513, 0.059953462332487106, -0.00721340486779809, 0.06645400822162628, 0.1982610821723938, 0.39348670840263367, -0.1817292422056198, 0.045478738844394684, 0.035260383039712906, -0.3973195254802704, 0.15198002755641937, -0.0401003398001194, 0.00020162499276921153, 0.005286709871143103, 0.06006823107600212, 0.037787873297929764, 0.12491367757320404, -0.321554958820343, 0.328936904668808, 0.26297372579574585, 0.02591048739850521, 0.03557631000876427, 0.1799466460943222, -0.028497537598013878, -0.15546958148479462, 0.2796451449394226, 0.07556171715259552, 0.43249401450157166, 0.05860915035009384, 0.8388558030128479, -0.2928156554698944, -0.5137995481491089, 0.11348997801542282, 0.4244186580181122, -0.0020872417371720076, -0.19419708847999573, 0.054137084633111954, 0.11889070272445679, -0.3490162193775177, 0.07172693312168121, -0.06716286391019821, -0.1238756999373436, 0.5954534411430359, -0.00362043222412467, 0.1662236750125885, -0.2218685895204544, -0.1740971803665161, 0.2818501889705658, 0.23936423659324646, -0.1145237609744072, 0.11490724235773087, 0.08922520279884338, -0.12978625297546387, -0.1515514999628067, 0.5569120645523071, 0.2587621212005615, 0.05257069319486618, -0.42290493845939636, 0.01423298753798008, 0.09556449949741364, 0.10881700366735458, 0.07581429183483124, 0.08942343294620514, 0.4531347453594208, -0.13934747874736786, 0.2778194546699524, -0.011475451290607452, -0.011696717701852322, -0.15413135290145874, 0.03352433443069458, -0.48072105646133423, -0.3438223898410797, 0.39191511273384094, 0.14652587473392487, -0.16903084516525269, 0.10381103307008743, 0.26207026839256287, -0.7270406484603882, -0.2905237674713135, 0.4198151230812073, -0.02890263870358467, -0.03651430457830429, -0.20693431794643402, 0.06619681417942047, 0.18844379484653473, 0.5728142261505127, 0.27638599276542664, 0.3275373876094818, -0.3885544240474701, -0.1176982894539833, -0.5355231761932373, -0.17271265387535095, 0.1998899281024933, -0.04911494255065918, -0.1868184208869934, 0.147975355386734, -0.021433210000395775, 0.32468077540397644, -0.26161468029022217, 0.10892421007156372, 0.16829942166805267, 0.07349079102277756, -0.5057327747344971, 0.11960247159004211, 0.20687858760356903, 0.04922628030180931, 0.19890856742858887, -0.057058364152908325, -0.08498957008123398, 0.13170498609542847, -0.07632112503051758, -0.08576222509145737, 0.005866042338311672, 0.2451677918434143, 0.16403944790363312, -0.026462871581315994, 0.08232583850622177, 0.4630804657936096, -0.013872461393475533, -0.19936424493789673, -0.019803276285529137, 0.04457425698637962, -0.4071345031261444, -0.07029040902853012, -0.09318266063928604, 0.1952257603406906, -0.3953467905521393, 0.2937234938144684, 0.008449839428067207, -0.11666034162044525, 0.0764622613787651, 0.028759140521287918, -0.15167884528636932, -0.04983152449131012, 0.08785522729158401, 0.19569392502307892, 0.2983799874782562, 0.740128219127655, -0.005244027823209763, 0.12361196428537369, -0.17261381447315216, -0.06902063637971878, 0.027409885078668594, -0.05322320759296417, -0.23431597650051117, -0.10208181291818619, 0.061302609741687775, 0.16821056604385376, 0.21723411977291107, -0.6970352530479431, -0.37758487462997437, 0.03720242902636528, -0.05564779415726662, 0.013424010016024113, 0.7778046727180481, 0.13142290711402893, -0.16813668608665466, -0.16086885333061218, -0.30782753229141235, 0.011136838234961033, -0.2971327602863312, 0.048108506947755814, -0.4513052701950073 ]
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.
[ -0.07244551181793213, -0.4672185480594635, -0.040159910917282104, -0.05503133684396744, -0.30456021428108215, 0.07474423944950104, 0.17443379759788513, 0.0672951489686966, 0.2653239965438843, -0.11934695392847061, -0.25975650548934937, 0.3054884374141693, -0.16949790716171265, 0.3780534863471985, 0.35094738006591797, -0.4499829113483429, -0.02154986746609211, -0.1463443487882614, -0.2921193540096283, 0.32420727610588074, 0.054316531866788864, 0.11644264310598373, -0.154354065656662, 0.0661999061703682, -0.27521762251853943, -0.17669448256492615, -0.34399914741516113, 0.12340904772281647, 0.08483222872018814, 0.07976431399583817, 0.13584454357624054, 0.3954450786113739, -0.018593018874526024, 0.12798623740673065, -0.00011141274444526061, -0.09476050734519958, 0.09925226867198944, -0.09712209552526474, 0.15263740718364716, 0.039992883801460266, 0.17799268662929535, -0.3429226875305176, 0.06340260803699493, -0.14008668065071106, 0.13159699738025665, 0.24719160795211792, 0.19724272191524506, 0.08876565098762512, 0.45008978247642517, -0.3533526062965393, 0.10489164292812347, 0.3131386637687683, -0.24382872879505157, 0.3010331988334656, -0.0856856182217598, -0.15280266106128693, 0.011506377719342709, 0.2888505756855011, 0.6218157410621643, -0.5280038714408875, -0.015142939984798431, 0.10488267242908478, -0.046287644654512405, 0.04999201372265816, 0.13104332983493805, -0.06663844734430313, -0.28239771723747253, -0.2535238265991211, -0.36275118589401245, 0.27751314640045166, -0.12239676713943481, -0.06288986653089523, -0.1925867646932602, -0.5656943321228027, -0.0037443567998707294, 0.05627410486340523, -0.3290136754512787, 0.0878172442317009, -0.21605637669563293, 0.015035570599138737, -0.12497936189174652, -0.12686865031719208, -0.007691535633057356, 0.11148105561733246, 0.3081289231777191, 0.5256883502006531, 0.07436870783567429, 0.1346540004014969, 0.35248711705207825, 0.10151558369398117, -0.14089259505271912, -0.2651450037956238, 0.2907777428627014, 0.13547112047672272, -0.328878253698349, -0.3770976960659027, 0.15168584883213043, -0.2695584297180176, 0.28027069568634033, 0.0967128574848175, 0.14681364595890045, 0.052291110157966614, -0.08265092223882675, 0.2643829882144928, 0.3891296982765198, 0.012631959281861782, -0.09676513075828552, -0.1197434663772583, 0.023959586396813393, -0.10591067373752594, 0.19155773520469666, 0.1394885629415512, 0.1446608155965805, 0.009004338644444942, -0.4349183738231659, -0.1123366579413414, -0.19096286594867706, 0.09654742479324341, -0.18051853775978088, -0.551717221736908, -0.08570043742656708, -0.10183706134557724, -0.0204154159873724, -0.12436327338218689, -0.288607120513916, 0.15308277308940887, -0.07474670559167862, 0.15042005479335785, -0.1804398149251938, 0.020788347348570824, -0.07014710456132889, 0.06636684387922287, -0.4629806578159332, -0.034968357533216476, 0.2592131197452545, 0.1055930033326149, 0.07144858688116074, 0.11740311235189438, -0.020503384992480278, 0.16793933510780334, 0.4261094331741333, 0.030805343762040138, 0.02872658707201481, -0.031046610325574875, 0.1476186066865921, -0.30910634994506836, -0.031393345445394516, 0.23604990541934967, -0.35183149576187134, -0.09727901965379715, -0.14454935491085052, -0.17756305634975433, 0.13741569221019745, 0.04549838975071907, -0.21582342684268951, -0.2692708969116211, -0.4415574073791504, 0.7260331511497498, -0.013283724896609783, -0.023931266739964485, -0.17177021503448486, 0.2594797611236572, -0.20922614634037018, -0.11802922189235687, 0.18940530717372894, 0.0955994725227356, -0.021369731053709984, -0.3190784156322479, -0.1060374528169632, -0.06091993302106857, -0.01561582088470459, 0.32130637764930725, -0.15178388357162476, 0.19926674664020538, 0.10687888413667679, 0.1746222823858261, 0.2618972659111023, -0.42945659160614014, -0.0568961538374424, 0.35273489356040955, -0.06147141382098198, 0.38411831855773926, 0.05087504908442497, 0.17250460386276245, -0.009614992886781693, -0.21683654189109802, 0.24459484219551086, 0.9837129712104797, -0.404123991727829, 0.0024757408536970615, -0.1859397143125534, -0.42440998554229736, 0.5637449622154236, 0.3482626974582672, 0.06482559442520142, -0.2935226559638977, -0.18867339193820953, 0.03436826169490814, 0.046717770397663116, 0.0923098623752594, 0.07490981370210648, -0.0802316889166832, -0.26448923349380493, 0.01035953126847744, -0.22413423657417297, -0.16984015703201294, -0.4216528832912445, -0.008736278861761093, -0.03696330636739731, 0.37871110439300537, 0.3727782368659973, -0.14035223424434662, 0.237596333026886, -0.3064366579055786, -0.11149434745311737, -0.31171661615371704, 0.14455874264240265, -0.06806720793247223, 0.012566061690449715, -0.2394191324710846, -0.06527674943208694, 0.2053755223751068, 0.09740225970745087, -0.1050233468413353, -0.42300787568092346, 0.2914789617061615, -0.04073400795459747, 0.02150779962539673, -0.013076732866466045, 0.6146170496940613, -0.21612341701984406, -0.11658573895692825, 0.20798185467720032, 0.291594535112381, -0.3138335943222046, 0.21099942922592163, 0.3093210756778717, 0.3545563220977783, 0.19584311544895172, -0.06140455976128578, 0.09926992654800415, -0.012210969813168049, -0.08991384506225586, -0.17108185589313507, -0.08095517009496689, 0.4797084629535675, -0.20128488540649414, 0.35754501819610596, -0.021910270676016808, 0.03489585965871811, -0.20772428810596466, -0.05692203715443611, -0.19069354236125946, 0.3881007134914398, 0.3106744587421417, -0.227995365858078, 0.334780216217041, -0.07678507268428802, -0.40030932426452637, 0.24344071745872498, 0.21397241950035095, 0.013479459099471569, 0.09026622027158737, -0.02306794933974743, 0.031240906566381454, -0.3101980686187744, -0.060503147542476654, 0.337543785572052, 0.5514169335365295, 0.2894958555698395, 0.17737025022506714, 0.01672918163239956, -0.21758221089839935, -0.07269225269556046, 0.04261382296681404, 0.16537678241729736, -0.06318721920251846, 0.14198455214500427, -0.04109911620616913, 0.09890107810497284, 0.026658032089471817, -0.10134842246770859, 0.12281981110572815, -0.14453920722007751, -0.04268140345811844, -0.037505004554986954, -0.15475647151470184, -0.5833336114883423, -0.19117645919322968, 0.052576012909412384, -0.351531058549881, -0.2960585951805115, 0.20875602960586548, 0.05685426667332649, -0.17757244408130646, 0.16798114776611328, 0.19975115358829498, 0.63985675573349, -0.39235299825668335, -0.17517179250717163, -0.05248158797621727, -0.29990020394325256, -0.11523349583148956, 0.13736394047737122, 0.43582582473754883, 0.3023829162120819, 0.49836045503616333, 0.23070253431797028, -0.032879896461963654, 0.12016379833221436, -0.38628512620925903, 0.11135121434926987, -0.35678213834762573, 0.29216980934143066, 0.09976053237915039, -0.10114531219005585, 0.09706719219684601, -0.366112619638443, -0.07395420968532562, -0.09430315345525742, -0.06166933476924896, -0.04039405658841133, 0.008098612539470196, 0.026680627837777138, -0.22710449993610382, -0.22857160866260529, -0.59697425365448, -0.49180635809898376, 0.2876456379890442, -0.2523239850997925, 0.09606634080410004, 0.04323148727416992, -0.28764471411705017, 0.2085765153169632, -0.14133451879024506, 0.03815574198961258, -0.18621955811977386, -0.07839325815439224, -0.02086154744029045, -0.3604213297367096, -0.17928364872932434, -0.422829806804657, -0.012583612464368343, 0.19307276606559753, -0.07771403342485428, -0.06733384728431702, -0.28919151425361633, 0.032976020127534866, -0.10656095296144485, 0.08936824649572372, 0.2331584244966507, 0.14413031935691833, -0.20325005054473877, 0.0736004039645195, -0.02718196250498295, -0.1418774276971817, 0.273416668176651, 0.39811545610427856, 0.3655805289745331, 0.1455877423286438, -0.025387810543179512, 0.21974128484725952, 0.7495046257972717, 0.28150779008865356, 0.08701787143945694, 0.10910996794700623, 0.29363617300987244, -0.03916477411985397, 0.09498471021652222, -0.17914023995399475, 0.25715702772140503, -0.11236836761236191, 0.49735674262046814, 0.17449797689914703, 0.1585705429315567, -0.20294512808322906, -0.27151724696159363, 0.01982324942946434, -0.29721835255622864, -0.21401141583919525, 0.2666807174682617, -0.2744176387786865, -0.002879852196201682, -0.10535697638988495, -0.41696521639823914, -0.2343723326921463, -0.08540158718824387, -0.020184984430670738, 0.2863893210887909, 0.04233941808342934, 0.17477644979953766, -0.37227708101272583, -0.02491079829633236, -0.4501192271709442, 0.1350296139717102, 0.0535183921456337, 0.3062891960144043, -0.12003651261329651, -0.10522747784852982, -0.09981342405080795, 0.1517457813024521, 0.5571805834770203, -0.5467592477798462, -0.15157894790172577, 0.08499225229024887, 0.0808279886841774, -0.13787342607975006, -0.07683363556861877, -0.34837594628334045, -0.013061238452792168, 0.450073778629303, 0.31701233983039856, -0.43303239345550537, -0.3620246648788452, 0.22607114911079407, -0.19920864701271057, -0.0786072313785553, -0.15291936695575714, -0.20009426772594452, -0.18559376895427704, -0.12708871066570282, 0.014104990288615227, 0.2638651430606842, 0.20318429172039032, -0.09490799903869629, 0.1841098517179489, -0.022126752883195877, 0.2680429220199585, 0.22520704567432404, 0.033784590661525726, -0.08329090476036072, 0.23447133600711823, 0.14794915914535522, 0.07315341383218765, -0.027881739661097527, -0.38055920600891113, 0.0787949413061142, -0.08251587301492691, 0.11951682716608047, 0.3015066087245941, 0.006075012963265181, 0.6325334906578064, 0.08441029489040375, 0.3359812796115875, 0.28056028485298157, -0.10613873600959778, 0.08464619517326355, -0.21934019029140472, 0.13918524980545044, 0.15871581435203552, 0.25185438990592957, -0.09183438122272491, -0.23387913405895233, 0.5279425382614136, 0.2838379442691803, -0.0503707192838192, 0.42784854769706726, -0.4882168769836426, -0.2839680314064026, 0.2870754599571228, 0.09380444139242172, 0.7551208734512329, -0.02168121002614498, 0.3053111433982849, 0.11571749299764633, -0.354065477848053, 0.6180205345153809, -0.35709407925605774, 0.10530248284339905, -0.0797438770532608, -0.1983344554901123, 0.03314230218529701, -0.0539892241358757, -0.1225515604019165, 0.2684195637702942, -0.34920957684516907, 0.33544158935546875, 0.4309313893318176, -0.2644553780555725, -0.04509816691279411, 0.5335496664047241, -0.02422487363219261, -0.32665401697158813, -0.020113598555326462, 0.07147341221570969, -0.3004836440086365, 0.06917271018028259, -0.14425431191921234, -0.24988873302936554, 0.21163959801197052, 0.17763075232505798, -0.2500084638595581, 0.05973878502845764, -0.035751257091760635, 0.15424509346485138, 0.21908295154571533, 0.020609421655535698, 0.2647172212600708, -0.23860932886600494, -0.07471196353435516, -0.16828158497810364, -0.13552433252334595, 0.0720595270395279, -0.19097580015659332, 0.04476761817932129, 0.11568400263786316, -0.0988849624991417, 0.5525389313697815, -0.07500901073217392, -0.10550446063280106, -0.26024293899536133, 0.25004443526268005, -0.40987783670425415, -0.21422822773456573, 0.020000416785478592, 0.17416387796401978, 0.15157103538513184, -0.3563391864299774, 0.5705785155296326, 0.06015294790267944, 0.05220317095518112, 0.052962128072977066, 0.1652722805738449, -0.2551437020301819, 0.4723825752735138, -0.24417747557163239, 0.006948040332645178, 0.057827480137348175, 0.009747170843183994, -0.10201285779476166, -0.35238391160964966, -0.06964641064405441, 0.31335723400115967, 0.09305272251367569, -0.042641039937734604, 0.03194184973835945, 0.13272316753864288, -0.1125108003616333, 0.019939253106713295, -0.17063000798225403, -0.36914998292922974, 0.14089328050613403, 0.24945884943008423, 0.2711484134197235, 0.1825062334537506, -0.11808469146490097, -0.20866374671459198, -0.10900405794382095, 0.5103972554206848, 0.0016740297432988882, 0.2612941265106201, -0.1301582008600235, 0.3655267655849457, 0.42578959465026855, 0.49408164620399475, -0.2902446687221527, 0.03717302531003952, 0.037440188229084015, 0.09641529619693756, 0.4827970266342163, -0.2636580765247345, 0.021779926493763924, 0.2817065119743347, -0.13145855069160461, 0.05857224762439728, -0.3447950482368469, -0.16829527914524078, -0.22006316483020782, 0.11335429549217224, 0.10334274172782898, -0.020654132589697838, 0.07225688546895981, -0.043053414672613144, -0.14204132556915283, -0.06888572871685028, 0.04575029015541077, -0.49925705790519714, -0.08053183555603027, 0.22027695178985596, -0.13717958331108093, -0.01160928700119257, 0.02034965343773365, 0.2394140064716339, -0.008112073875963688, -0.12586094439029694, 0.2913137376308441, 0.1330929845571518, 0.12393414229154587, -0.1924993246793747, -0.16767150163650513, 0.01987883821129799, 0.21255379915237427, 0.17324024438858032, -0.056935857981443405, 0.19535772502422333, -0.1702473908662796, -0.14223438501358032, 0.24023394286632538, 0.1274132877588272, 0.04383225366473198, -0.17915266752243042, 0.1771993339061737, 0.14164851605892181, -0.09756153076887131, 0.06519671529531479, 0.22627557814121246, 0.08455797284841537, 0.12077272683382034, 0.09019730985164642, 0.1219424158334732, 0.43579477071762085, -0.25440528988838196, -0.017903031781315804, 0.14192314445972443, -0.3514215052127838, 0.1340951770544052, -0.04542085900902748, -0.07775241881608963, 0.003096309956163168, 0.11714646965265274, -0.07418843358755112, 0.12427607923746109, -0.35630735754966736, 0.3523234724998474, 0.27313119173049927, -0.01798127591609955, 0.060443200170993805, 0.17766690254211426, 0.005221745930612087, -0.13915109634399414, 0.35451409220695496, 0.10290100425481796, 0.40711647272109985, 0.03838276118040085, 0.9214754700660706, -0.28261420130729675, -0.49107712507247925, 0.07427015155553818, 0.4407669007778168, 0.013904755935072899, -0.2088600993156433, 0.06338045001029968, 0.07670655846595764, -0.3490920960903168, 0.0910915806889534, 0.028142906725406647, -0.16176584362983704, 0.6003192663192749, -0.09266366809606552, 0.19700394570827484, -0.12051007151603699, -0.18638814985752106, 0.30416539311408997, 0.21934549510478973, -0.08482000231742859, 0.038662996143102646, 0.08103376626968384, -0.09659158438444138, -0.20828434824943542, 0.5394877195358276, 0.13788525760173798, -0.030670106410980225, -0.48136815428733826, 0.026542292907834053, 0.1303126960992813, 0.16194480657577515, -0.04226177558302879, 0.169663205742836, 0.4215087592601776, -0.14176730811595917, 0.2907409965991974, 0.0038775065913796425, -0.09474685788154602, -0.18484866619110107, 0.07130373269319534, -0.4407300353050232, -0.3587898910045624, 0.3806774914264679, 0.1489291936159134, -0.18169757723808289, 0.12254278361797333, 0.27388161420822144, -0.6083030104637146, -0.22651779651641846, 0.2887651324272156, -0.1661260426044464, -0.0840119868516922, -0.22621017694473267, 0.08732540905475616, 0.050029851496219635, 0.6215311288833618, 0.2386113405227661, 0.3424453139305115, -0.312441885471344, -0.11013562977313995, -0.5452613830566406, -0.12234259396791458, 0.15473422408103943, -0.1268235743045807, -0.16537794470787048, 0.1752581149339676, -0.02540072426199913, 0.3293079137802124, -0.26487937569618225, 0.18954193592071533, 0.13836880028247833, 0.08010133355855942, -0.4509575664997101, 0.14279329776763916, 0.08015456795692444, 0.06859400123357773, 0.2718910276889801, -0.10980740934610367, -0.17113752663135529, 0.08312541246414185, -0.06507851928472519, -0.06606286764144897, -0.06892774254083633, 0.2686034142971039, 0.16191942989826202, 0.03138841316103935, 0.07718027383089066, 0.43555670976638794, 0.004979684483259916, -0.2062925100326538, 0.0015182332135736942, 0.0014965086011216044, -0.35154399275779724, -0.19137006998062134, -0.07274571061134338, 0.16833795607089996, -0.33567512035369873, 0.29710105061531067, -0.029855601489543915, -0.11843478679656982, 0.12275101989507675, 0.06191490590572357, -0.194886714220047, -0.09245681762695312, 0.10430525243282318, 0.15145070850849152, 0.21517956256866455, 0.7912339568138123, -0.03391668573021889, 0.05124376341700554, -0.10069800168275833, -0.010347864590585232, 0.04032714664936066, -0.09624388813972473, -0.2264849990606308, -0.07546060532331467, 0.034431666135787964, 0.08696582913398743, 0.10204707831144333, -0.6406651139259338, -0.37819692492485046, 0.03704727068543434, -0.04062320291996002, -0.0003267158754169941, 0.742240309715271, 0.09145685285329819, -0.19132637977600098, -0.07553967833518982, -0.32336390018463135, 0.00820229109376669, -0.29208531975746155, 0.16921190917491913, -0.3407362401485443 ]
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.
[ -0.09163220971822739, -0.421159952878952, -0.04513245075941086, -0.08009861409664154, -0.24046726524829865, 0.059044960886240005, 0.1587897092103958, 0.05346061661839485, 0.3097863793373108, -0.1581772267818451, -0.23331521451473236, 0.31752070784568787, -0.19275808334350586, 0.3524588942527771, 0.42011559009552, -0.4721606969833374, 0.027789244428277016, -0.20416471362113953, -0.30799007415771484, 0.3535073399543762, 0.0078004393726587296, 0.11438111215829849, -0.17745055258274078, 0.08323553949594498, -0.29803037643432617, -0.16597460210323334, -0.35576876997947693, 0.1336842179298401, 0.09495052695274353, 0.019252978265285492, 0.17773200571537018, 0.37621423602104187, -0.0357888825237751, 0.06485924124717712, -0.00011105541489087045, -0.11337058991193771, 0.2117232233285904, -0.082444928586483, 0.0959489494562149, -0.0414968803524971, 0.14164210855960846, -0.3843333125114441, 0.0004230434715282172, -0.09866788983345032, 0.1814693808555603, 0.16644397377967834, 0.19007344543933868, 0.04961876943707466, 0.3385031819343567, -0.29403936862945557, 0.12200932204723358, 0.3003104627132416, -0.16042430698871613, 0.2594372630119324, -0.0790456235408783, -0.07974832504987717, 0.04912272095680237, 0.22365310788154602, 0.6150230169296265, -0.549124538898468, -0.0011969185434281826, 0.12823133170604706, -0.0015523391775786877, -0.024031948298215866, 0.18237632513046265, -0.0457616001367569, -0.35870039463043213, -0.2435755431652069, -0.3750145435333252, 0.2361428290605545, -0.11858711391687393, -0.1537858247756958, -0.2185506671667099, -0.5235365629196167, 0.02771846018731594, 0.016797054558992386, -0.2812063694000244, 0.029973551630973816, -0.1675618290901184, -0.04844870790839195, -0.165602907538414, -0.12260466068983078, 0.0131073584780097, 0.15340234339237213, 0.3461078703403473, 0.568138837814331, 0.06152698025107384, 0.1679220199584961, 0.4003540873527527, 0.0016654805513098836, -0.15272557735443115, -0.3198026716709137, 0.29201024770736694, 0.09851135313510895, -0.38790518045425415, -0.33713626861572266, 0.14897942543029785, -0.1886470764875412, 0.24584217369556427, 0.1551017314195633, 0.10689501464366913, 0.08710204064846039, -0.11532720923423767, 0.2748214602470398, 0.3207058608531952, 0.06586848944425583, -0.06847183406352997, -0.11937646567821503, 0.02774709463119507, -0.08491221070289612, 0.21637924015522003, 0.12413559854030609, 0.2167028784751892, -0.027859853580594063, -0.44677767157554626, -0.1466728150844574, -0.17592649161815643, 0.12826530635356903, -0.17640073597431183, -0.5241800546646118, -0.05227205529808998, -0.10353737324476242, 0.08943966776132584, -0.048955608159303665, -0.28327587246894836, 0.1021498292684555, -0.07685298472642899, 0.2082911729812622, -0.0627482533454895, 0.0749339684844017, -0.07943776249885559, 0.12596236169338226, -0.43758025765419006, -0.01497544813901186, 0.3017202913761139, 0.11206887662410736, 0.11711075156927109, 0.1599152535200119, -0.028330354019999504, 0.10989683121442795, 0.37951841950416565, -0.016750724986195564, 0.0509767085313797, -0.007164106238633394, 0.1398807018995285, -0.32601627707481384, -0.048829592764377594, 0.1761283427476883, -0.3429519832134247, -0.146550253033638, -0.13508251309394836, -0.2072174847126007, 0.14267078042030334, 0.05246463790535927, -0.19705785810947418, -0.2726626992225647, -0.4697835445404053, 0.7273024916648865, -0.06460848450660706, -0.0431518591940403, -0.1394859403371811, 0.2776137888431549, -0.24227072298526764, -0.09580088406801224, 0.19651950895786285, 0.0900479257106781, -0.008326757699251175, -0.32873407006263733, -0.09379424154758453, -0.11021610349416733, -0.03254237771034241, 0.26917803287506104, -0.17343449592590332, 0.2014375627040863, 0.04872145131230354, 0.20125922560691833, 0.2725920081138611, -0.46744662523269653, -0.06602431833744049, 0.3148565888404846, -0.02159803733229637, 0.37245580554008484, 0.04715697839856148, 0.22150470316410065, -0.00911981612443924, -0.19109755754470825, 0.27087271213531494, 0.9994991421699524, -0.39096599817276, -0.0052986787632107735, -0.1448642760515213, -0.3675379455089569, 0.5181838870048523, 0.3499695360660553, 0.06550057232379913, -0.3326673209667206, -0.14244580268859863, -0.023778392001986504, 0.08096596598625183, 0.07707786560058594, 0.08081141859292984, -0.02339562401175499, -0.3496009409427643, 0.025720752775669098, -0.21601691842079163, -0.13175392150878906, -0.39827898144721985, -0.013193050399422646, -0.011377648450434208, 0.3362920582294464, 0.26290634274482727, -0.11397728323936462, 0.23644916713237762, -0.2684760093688965, -0.1313517689704895, -0.34611213207244873, 0.1502705216407776, -0.06126665696501732, 0.05107953026890755, -0.2666863203048706, -0.03675666078925133, 0.32183605432510376, 0.11384937167167664, -0.11466069519519806, -0.5048949718475342, 0.27984175086021423, -0.08512662351131439, 0.03201058506965637, -0.015921637415885925, 0.6123453974723816, -0.13566909730434418, -0.09030592441558838, 0.18309856951236725, 0.3067478835582733, -0.3286169469356537, 0.2714979946613312, 0.32483503222465515, 0.2580251693725586, 0.18337470293045044, -0.023683473467826843, 0.11495953798294067, -0.03480035439133644, -0.031369488686323166, -0.20113025605678558, -0.10553114116191864, 0.4779738485813141, -0.17939847707748413, 0.4034448564052582, -0.02794782444834709, 0.027526522055268288, -0.19954317808151245, -0.0360027514398098, -0.22701187431812286, 0.36756691336631775, 0.3015570044517517, -0.19315436482429504, 0.3859136402606964, -0.09095414727926254, -0.34884005784988403, 0.17814305424690247, 0.20084048807621002, 0.013799051754176617, 0.09303056448698044, -0.016065264120697975, -0.060315538197755814, -0.31057435274124146, -0.06443575769662857, 0.41596800088882446, 0.646736204624176, 0.2788432538509369, 0.14651699364185333, 0.05391909182071686, -0.2370406985282898, -0.07217655330896378, 0.02709747478365898, 0.17971579730510712, -0.022058062255382538, 0.13218265771865845, 0.028972577303647995, 0.053668953478336334, -0.019798049703240395, -0.14596883952617645, 0.13913629949092865, -0.18290124833583832, -0.026788391172885895, 0.035045113414525986, -0.19006143510341644, -0.5519412755966187, -0.231194406747818, 0.04605128988623619, -0.3759758174419403, -0.30405309796333313, 0.12783874571323395, 0.0250947754830122, -0.19186358153820038, 0.1835845708847046, 0.2177858203649521, 0.5667122006416321, -0.42838552594184875, -0.25661709904670715, -0.10000106692314148, -0.25852105021476746, -0.10058113187551498, 0.12403924018144608, 0.4180923104286194, 0.300885945558548, 0.496792733669281, 0.28931286931037903, -0.014469991438090801, 0.020839178934693336, -0.3807591199874878, 0.04763730987906456, -0.28813809156417847, 0.26607802510261536, 0.10908009111881256, -0.23446263372898102, 0.1544487029314041, -0.41126134991645813, -0.02723083831369877, -0.06777586042881012, -0.051392801105976105, -0.04911769554018974, -0.015199271962046623, 0.020708436146378517, -0.176865816116333, -0.22578813135623932, -0.6239804625511169, -0.4909672439098358, 0.28933680057525635, -0.28130969405174255, 0.06138415262103081, 0.01653517596423626, -0.2601395845413208, 0.19389963150024414, -0.15522344410419464, 0.037599556148052216, -0.17455723881721497, -0.10408563911914825, 0.04657382518053055, -0.36165890097618103, -0.16067302227020264, -0.41546836495399475, -0.1143825426697731, 0.2324679046869278, -0.08507466316223145, -0.04257410392165184, -0.23878790438175201, 0.048094697296619415, -0.1275748610496521, 0.0941581055521965, 0.22289183735847473, 0.17515943944454193, -0.2228708416223526, 0.09907606244087219, -0.03677714616060257, -0.264628529548645, 0.1948915272951126, 0.33807265758514404, 0.42725709080696106, 0.1778106838464737, 0.016580823808908463, 0.17279300093650818, 0.7245492339134216, 0.25919124484062195, 0.05576372891664505, 0.1505172699689865, 0.30706268548965454, 0.033568985760211945, 0.14654755592346191, -0.23375561833381653, 0.24164292216300964, -0.16472771763801575, 0.5056430697441101, 0.09843328595161438, 0.17190134525299072, -0.20434823632240295, -0.28388193249702454, -0.012782332487404346, -0.25792184472084045, -0.1879773885011673, 0.26838403940200806, -0.22223521769046783, 0.020278388634324074, -0.05537978559732437, -0.3724348247051239, -0.25288161635398865, -0.042795002460479736, 0.017948506399989128, 0.31589120626449585, -0.030992668122053146, 0.15221893787384033, -0.34515348076820374, -0.04119982197880745, -0.4830808639526367, 0.15105336904525757, 0.08815188705921173, 0.27570009231567383, -0.08164788037538528, -0.07314686477184296, -0.1406986266374588, 0.10453657060861588, 0.5943602919578552, -0.5221481919288635, -0.16565899550914764, 0.09090547263622284, 0.08042621612548828, -0.09018899500370026, -0.08775421231985092, -0.272024005651474, -0.043630603700876236, 0.44831690192222595, 0.36958014965057373, -0.3692058026790619, -0.31362178921699524, 0.2565196454524994, -0.18059848248958588, -0.03166769817471504, -0.19901448488235474, -0.18238738179206848, -0.189231738448143, -0.14206203818321228, 0.032154630869627, 0.18435092270374298, 0.21136435866355896, -0.10405109822750092, 0.17735280096530914, -0.0519656278192997, 0.25296080112457275, 0.20239831507205963, 0.013077511452138424, -0.040300339460372925, 0.22695456445217133, 0.15439783036708832, 0.03186015412211418, -0.12105237692594528, -0.29360535740852356, 0.09633071720600128, -0.06996091455221176, 0.11004825681447983, 0.2714960277080536, -0.0030555585399270058, 0.6472854614257812, 0.05717649683356285, 0.31057438254356384, 0.28897690773010254, -0.12562018632888794, 0.05900802090764046, -0.23012983798980713, 0.1417718529701233, 0.22950921952724457, 0.25832298398017883, -0.13930946588516235, -0.19662609696388245, 0.5935573577880859, 0.31629839539527893, -0.07326269894838333, 0.38970163464546204, -0.46200183033943176, -0.2712933123111725, 0.25234121084213257, 0.02791888453066349, 0.6917436122894287, -0.02594774030148983, 0.2671623229980469, 0.06503253430128098, -0.3845383822917938, 0.6288004517555237, -0.3909117579460144, 0.1575741469860077, -0.07772619277238846, -0.23540404438972473, 0.032601770013570786, -0.09457410871982574, -0.0697193592786789, 0.268308162689209, -0.37171638011932373, 0.3496938645839691, 0.47733667492866516, -0.2044108808040619, -0.05118018016219139, 0.5518016219139099, -0.029192328453063965, -0.40027952194213867, -0.10723665356636047, 0.08906088024377823, -0.28121334314346313, 0.10603450238704681, -0.16336657106876373, -0.2763764262199402, 0.248642697930336, 0.11350058764219284, -0.27409282326698303, 0.11336162686347961, -0.035164475440979004, 0.18557000160217285, 0.2648399770259857, 0.01631402038037777, 0.25120624899864197, -0.17889147996902466, -0.06662190705537796, -0.15494687855243683, -0.08863594383001328, 0.13129130005836487, -0.1466124951839447, 0.050258368253707886, 0.030554814264178276, -0.2045278400182724, 0.5105486512184143, -0.07615149021148682, -0.1078912615776062, -0.24378012120723724, 0.24476532638072968, -0.4367164075374603, -0.15425078570842743, 0.1101064383983612, 0.1682257354259491, 0.21321603655815125, -0.3162393271923065, 0.600138247013092, 0.08641385287046432, -0.009584211744368076, 0.05580231547355652, 0.1773013174533844, -0.2368270605802536, 0.4851785898208618, -0.2578957974910736, -0.0046104546636343, 0.052440665662288666, 0.010878507047891617, -0.1622539609670639, -0.37275442481040955, -0.033533114939928055, 0.29813796281814575, 0.08636652678251266, -0.01616482436656952, -0.0029818566981703043, 0.10210420191287994, -0.07975629717111588, -0.0030083481688052416, -0.1554485261440277, -0.3322066366672516, 0.11653729528188705, 0.14724498987197876, 0.19470562040805817, 0.16701367497444153, -0.05395301431417465, -0.24001413583755493, -0.18599310517311096, 0.4692336618900299, 0.040620140731334686, 0.223419651389122, -0.1361560821533203, 0.34052085876464844, 0.42906028032302856, 0.5344386100769043, -0.2816094756126404, 0.07217343151569366, 0.03279007226228714, 0.08643418550491333, 0.48312050104141235, -0.2349262535572052, 0.07466652989387512, 0.27867913246154785, -0.11924928426742554, 0.04919930920004845, -0.34023579955101013, -0.1727285385131836, -0.2547360062599182, 0.09835316985845566, 0.11223388463258743, -0.0018387713935226202, 0.005412950646132231, -0.06127319112420082, -0.0791017934679985, -0.08158954232931137, 0.016571901738643646, -0.3894308805465698, -0.0693221464753151, 0.15564653277397156, -0.11541076004505157, 0.015998009592294693, 0.05923793092370033, 0.20302154123783112, 0.009794154204428196, -0.05242888629436493, 0.29116445779800415, 0.12709859013557434, 0.11118432879447937, -0.19097991287708282, -0.17631736397743225, 0.004691425245255232, 0.21382759511470795, 0.14815166592597961, -0.029078425839543343, 0.1604178100824356, -0.23436836898326874, -0.14009538292884827, 0.23429255187511444, 0.1486867517232895, 0.020620545372366905, -0.12043074518442154, 0.10565609484910965, 0.14062274992465973, -0.12489892542362213, -0.028531573712825775, 0.1781453788280487, 0.11084804683923721, 0.1366320699453354, 0.11059831827878952, 0.15870623290538788, 0.4499151110649109, -0.21001622080802917, 0.025973431766033173, 0.11526613682508469, -0.36859607696533203, 0.12917540967464447, -0.1154741570353508, -0.0404563806951046, 0.005308446474373341, 0.14748342335224152, -0.011129364371299744, 0.13959290087223053, -0.32164719700813293, 0.30607354640960693, 0.3096957802772522, -0.023712854832410812, 0.05864633619785309, 0.23911485075950623, 0.014807814732193947, -0.13310983777046204, 0.30467912554740906, 0.01857585459947586, 0.31923994421958923, 0.042772937566041946, 0.8754273653030396, -0.27222710847854614, -0.45505988597869873, 0.10648228973150253, 0.43658551573753357, 0.03211098164319992, -0.2155291885137558, 0.15521635115146637, 0.09118607640266418, -0.40597599744796753, 0.12097867578268051, 0.05810915306210518, -0.16374599933624268, 0.6574729681015015, -0.116494320333004, 0.16466790437698364, -0.15811505913734436, -0.14994098246097565, 0.3523969054222107, 0.19673553109169006, -0.0954383835196495, 0.010715988464653492, 0.11992733180522919, -0.11069277673959732, -0.16985256969928741, 0.5139833092689514, 0.2285723239183426, 0.03479301929473877, -0.41247546672821045, 0.037352774292230606, 0.1071360856294632, 0.14290346205234528, -0.07321170717477798, 0.13803686201572418, 0.4449816048145294, -0.18126770853996277, 0.25427672266960144, 0.03678707405924797, -0.07882070541381836, -0.1354019194841385, 0.04338379576802254, -0.3884734809398651, -0.39733776450157166, 0.45656606554985046, 0.10967676341533661, -0.11010193824768066, 0.04526510462164879, 0.32069969177246094, -0.6264175772666931, -0.23748116195201874, 0.36148080229759216, -0.13629017770290375, -0.050093408674001694, -0.24483510851860046, 0.08715888857841492, 0.08407646417617798, 0.6122648119926453, 0.2146085649728775, 0.353494256734848, -0.4017142951488495, -0.14038483798503876, -0.5742549896240234, -0.21603292226791382, 0.17194177210330963, -0.07661565393209457, -0.25840139389038086, 0.1467667818069458, -0.06859239935874939, 0.2895059585571289, -0.21401166915893555, 0.11786486953496933, 0.17773331701755524, 0.027057595551013947, -0.43053901195526123, 0.11578288674354553, 0.058791350573301315, 0.004421781748533249, 0.31976133584976196, -0.11052565276622772, -0.1742171347141266, 0.08242561668157578, -0.056964170187711716, -0.0908903107047081, -0.06792813539505005, 0.23781512677669525, 0.16779714822769165, 0.09648948907852173, 0.08136257529258728, 0.46605291962623596, -0.049324795603752136, -0.20165373384952545, -0.00878460705280304, 0.012513473629951477, -0.37709832191467285, -0.1661679893732071, -0.0771833062171936, 0.18014609813690186, -0.3348352313041687, 0.254970520734787, -0.01663435623049736, -0.08482299000024796, 0.11921888589859009, 0.018418315798044205, -0.1893487423658371, -0.049458231776952744, 0.07420572638511658, 0.19506895542144775, 0.22280599176883698, 0.7319667339324951, -0.010562296025454998, 0.08575521409511566, -0.09298611432313919, -0.03421400114893913, 0.07086595892906189, -0.0521119087934494, -0.20773281157016754, -0.13613834977149963, 0.024237249046564102, 0.09727298468351364, 0.13568149507045746, -0.6967557072639465, -0.33582890033721924, 0.014331517741084099, -0.022863827645778656, 0.031776681542396545, 0.7183234691619873, 0.06665455549955368, -0.14486509561538696, -0.11753136664628983, -0.2986989915370941, 0.05342422425746918, -0.2820819616317749, 0.18920624256134033, -0.347708523273468 ]
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).
[ -0.07760575413703918, -0.430137038230896, -0.05128100886940956, -0.007647973019629717, -0.24152901768684387, 0.1148412823677063, 0.17433670163154602, 0.0923096314072609, 0.28521785140037537, -0.11796169728040695, -0.24373742938041687, 0.2735089361667633, -0.19464358687400818, 0.416962593793869, 0.39819997549057007, -0.4560350775718689, 0.007996948435902596, -0.21966706216335297, -0.260622501373291, 0.3799697756767273, 0.04182815179228783, 0.09555917233228683, -0.14734548330307007, 0.0627831369638443, -0.2902204990386963, -0.13717210292816162, -0.343586802482605, 0.08050666004419327, 0.10446936637163162, 0.0340154655277729, 0.19900573790073395, 0.3737786114215851, 0.008997002616524696, 0.10891919583082199, -0.0001125285416492261, -0.09093976765871048, 0.16050821542739868, -0.07122167944908142, 0.130285382270813, -0.023051893338561058, 0.1236005425453186, -0.35964640974998474, 0.06394299119710922, -0.10716304928064346, 0.18151775002479553, 0.18675684928894043, 0.2017907202243805, 0.08393161743879318, 0.31810232996940613, -0.31550025939941406, 0.09998162090778351, 0.2903844118118286, -0.18832506239414215, 0.28077980875968933, -0.054062098264694214, -0.06420376151800156, 0.030681487172842026, 0.22856436669826508, 0.6673145890235901, -0.5304580330848694, 0.044611699879169464, 0.068392314016819, -0.02713390812277794, 0.041563984006643295, 0.117887943983078, -0.08518798649311066, -0.3070952594280243, -0.23946751654148102, -0.373504102230072, 0.2802632749080658, -0.07192564755678177, -0.10643937438726425, -0.23320840299129486, -0.5333684086799622, 0.01573350839316845, 0.0769621729850769, -0.2952457070350647, 0.05960951745510101, -0.16279664635658264, -0.019170673564076424, -0.12913021445274353, -0.12187950313091278, -0.0014562945580109954, 0.17361846566200256, 0.3086414933204651, 0.5555931329727173, 0.06791824847459793, 0.1660681664943695, 0.33545374870300293, 0.0697007104754448, -0.09848671406507492, -0.2893274128437042, 0.283674418926239, 0.06580860912799835, -0.4028255343437195, -0.35675525665283203, 0.13010726869106293, -0.2654321491718292, 0.2804775536060333, 0.13964110612869263, 0.1625583916902542, 0.09564376622438431, -0.16685350239276886, 0.2518715560436249, 0.33976057171821594, -0.022479485720396042, -0.1228717565536499, -0.13603012263774872, 0.017504794523119926, -0.10887762904167175, 0.16654378175735474, 0.13609996438026428, 0.17564202845096588, -0.018663041293621063, -0.42686501145362854, -0.11984273046255112, -0.14647670090198517, 0.09820448607206345, -0.18795838952064514, -0.603138267993927, -0.09075374901294708, -0.10779083520174026, 0.031006427481770515, -0.1576593965291977, -0.2877179682254791, 0.09265287965536118, -0.10346347093582153, 0.2439785897731781, -0.131792351603508, 0.008540341630578041, -0.04857705906033516, 0.06551400572061539, -0.42392849922180176, -0.04849156364798546, 0.29413992166519165, 0.09935430437326431, 0.04831107333302498, 0.16498585045337677, -0.03427765145897865, 0.12868382036685944, 0.42059463262557983, 0.029048139229416847, 0.014121653512120247, -0.03673773631453514, 0.1730395406484604, -0.31929782032966614, -0.011867186054587364, 0.2226768285036087, -0.37173980474472046, -0.11837541311979294, -0.1698479950428009, -0.21736475825309753, 0.09760751575231552, 0.033433765172958374, -0.21831919252872467, -0.31849947571754456, -0.4913886785507202, 0.7129409909248352, -0.0447411835193634, -0.04646187648177147, -0.16467784345149994, 0.26663944125175476, -0.23226149380207062, -0.09799231588840485, 0.1632637232542038, 0.08497171849012375, 0.013150009326636791, -0.2949586510658264, -0.09312962740659714, -0.10383573919534683, -0.06481073796749115, 0.3017258048057556, -0.16893531382083893, 0.21227426826953888, 0.06585875898599625, 0.13697408139705658, 0.2494572550058365, -0.4536500871181488, -0.04889878258109093, 0.33244839310646057, -0.024968693032860756, 0.357054203748703, 0.07683289051055908, 0.21231050789356232, -0.0030693835578858852, -0.20406708121299744, 0.19492653012275696, 1.0177950859069824, -0.395271360874176, -0.021733781322836876, -0.11022840440273285, -0.3906151056289673, 0.542577862739563, 0.3610701560974121, 0.06735077500343323, -0.3310127556324005, -0.13180625438690186, -0.00870626512914896, 0.09087156504392624, 0.07975571602582932, 0.0483211986720562, -0.0628935918211937, -0.34346047043800354, 0.04504150152206421, -0.2458025962114334, -0.19732028245925903, -0.4192027151584625, -0.02027890458703041, 0.016956960782408714, 0.29230326414108276, 0.3340204954147339, -0.10002611577510834, 0.27135223150253296, -0.3039945960044861, -0.15376226603984833, -0.31332987546920776, 0.11975879967212677, -0.0056192921474576, 0.03799644112586975, -0.2483402043581009, -0.030674457550048828, 0.31918036937713623, 0.09926754236221313, -0.13511964678764343, -0.4337625503540039, 0.3398185968399048, -0.047643307596445084, 0.046760641038417816, -0.04639912024140358, 0.6206656694412231, -0.19886364042758942, -0.09158608317375183, 0.17018060386180878, 0.24870337545871735, -0.3234957456588745, 0.256033718585968, 0.324140340089798, 0.350034236907959, 0.19196605682373047, -0.08836964517831802, 0.08618832379579544, -0.012060914188623428, -0.05773845314979553, -0.19454269111156464, -0.08406422287225723, 0.4658960998058319, -0.18547140061855316, 0.4102384150028229, -0.05356261506676674, 0.014754711650311947, -0.19410152733325958, -0.055069323629140854, -0.18381230533123016, 0.4107012450695038, 0.3079739511013031, -0.19726142287254333, 0.33659157156944275, -0.11086127161979675, -0.3630487620830536, 0.1967424601316452, 0.1851433664560318, 0.05677727609872818, 0.07615280151367188, -0.032488856464624405, -0.019166333600878716, -0.2904629707336426, -0.11108557879924774, 0.4045723080635071, 0.632398784160614, 0.28050950169563293, 0.11706018447875977, -0.021244820207357407, -0.2500744163990021, -0.10282347351312637, 0.011356580071151257, 0.20606470108032227, -0.01684357225894928, 0.1502842754125595, 0.04689301922917366, 0.07360987365245819, 0.043971799314022064, -0.14593525230884552, 0.19349101185798645, -0.17940066754817963, -0.04956895112991333, 0.02461143583059311, -0.19285309314727783, -0.5652399063110352, -0.21911512315273285, 0.043787989765405655, -0.3279033303260803, -0.30137425661087036, 0.15034152567386627, 0.0533406026661396, -0.17695507407188416, 0.15528978407382965, 0.17466291785240173, 0.5224667191505432, -0.444394588470459, -0.23384366929531097, -0.06413109600543976, -0.2828476130962372, -0.08073674887418747, 0.11607666313648224, 0.49301406741142273, 0.2858504354953766, 0.4589037299156189, 0.26435229182243347, -0.045532457530498505, 0.0037953294813632965, -0.393947958946228, 0.02901114523410797, -0.30691760778427124, 0.2791736125946045, 0.08302601426839828, -0.2171398401260376, 0.1505548506975174, -0.39499631524086, -0.06443796306848526, -0.048821937292814255, -0.07792937755584717, -0.028692856431007385, -0.025047777220606804, 0.08972236514091492, -0.17804250121116638, -0.22694063186645508, -0.6084063053131104, -0.4707541763782501, 0.2826777994632721, -0.28210893273353577, 0.0727405920624733, 0.038764454424381256, -0.2779342234134674, 0.15081927180290222, -0.15593425929546356, 0.02802894450724125, -0.1545216292142868, -0.11299071460962296, -0.018639694899320602, -0.3668549060821533, -0.15962427854537964, -0.412315309047699, -0.07378344237804413, 0.22635796666145325, -0.07291723787784576, -0.049834955483675, -0.22019940614700317, 0.06068311631679535, -0.10324534773826599, 0.05648268759250641, 0.17424625158309937, 0.17185261845588684, -0.22455734014511108, 0.11065278947353363, -0.04123273119330406, -0.20303359627723694, 0.2085089236497879, 0.36925259232521057, 0.3985559642314911, 0.17247536778450012, -0.04534249007701874, 0.158677339553833, 0.6988728046417236, 0.22570140659809113, 0.07469477504491806, 0.1957734078168869, 0.30035749077796936, -0.008465267717838287, 0.14122039079666138, -0.23328648507595062, 0.21620109677314758, -0.13694699108600616, 0.49532994627952576, 0.11535315215587616, 0.15646058320999146, -0.17848844826221466, -0.2667303681373596, 0.053566522896289825, -0.2868150472640991, -0.18034504354000092, 0.2325090616941452, -0.2959994971752167, 0.001816321862861514, -0.06489180028438568, -0.4627402424812317, -0.28313717246055603, -0.049076057970523834, -0.02859243005514145, 0.30236193537712097, -0.018328504636883736, 0.1623070240020752, -0.3587011992931366, -0.02396813966333866, -0.3594933748245239, 0.1668870598077774, 0.09561163187026978, 0.25671496987342834, -0.08165300637483597, -0.045284830033779144, -0.0754171758890152, 0.13249599933624268, 0.5941113233566284, -0.5726910829544067, -0.11468035727739334, 0.11226192116737366, 0.06943970173597336, -0.10702887922525406, -0.11246585845947266, -0.25311192870140076, -0.02277880162000656, 0.4108993709087372, 0.3724682927131653, -0.3512175679206848, -0.3359483778476715, 0.27706095576286316, -0.17802581191062927, -0.06660396605730057, -0.17734864354133606, -0.1805456578731537, -0.19922912120819092, -0.11174323409795761, -0.015011969022452831, 0.2250315248966217, 0.23601853847503662, -0.09397964179515839, 0.16912855207920074, -0.11902878433465958, 0.23430877923965454, 0.17777664959430695, -0.02001618780195713, -0.08604385703802109, 0.21967251598834991, 0.19012925028800964, 0.043740447610616684, -0.12046506255865097, -0.28548485040664673, 0.09027731418609619, -0.08531797677278519, 0.0410669706761837, 0.32314565777778625, -0.01695917546749115, 0.6543565392494202, 0.0035460330545902252, 0.3540613353252411, 0.2970188558101654, -0.1518300622701645, 0.035119932144880295, -0.22525554895401, 0.15921913087368011, 0.19328387081623077, 0.2645816206932068, -0.16709131002426147, -0.22239504754543304, 0.5615097880363464, 0.3239040970802307, -0.04171910136938095, 0.3951147198677063, -0.5345007181167603, -0.2746390998363495, 0.27251505851745605, 0.029301969334483147, 0.6709373593330383, -0.085677370429039, 0.28159984946250916, 0.012769298627972603, -0.4185328781604767, 0.6093618869781494, -0.3563195765018463, 0.14552362263202667, -0.07994051277637482, -0.23156191408634186, 0.021089831367135048, -0.07193604111671448, -0.09989535808563232, 0.2743527591228485, -0.3496197462081909, 0.30141234397888184, 0.443901926279068, -0.1805112212896347, -0.03987335041165352, 0.5531436204910278, -0.01531270332634449, -0.3570755124092102, -0.03860880807042122, 0.06670353561639786, -0.29009079933166504, 0.07109356671571732, -0.15476755797863007, -0.25153273344039917, 0.24413546919822693, 0.19273704290390015, -0.34432053565979004, 0.06983192265033722, 0.018976962193846703, 0.1855357438325882, 0.22446338832378387, -0.00212018471211195, 0.2966082990169525, -0.16361577808856964, -0.031092140823602676, -0.1591513454914093, -0.11198891699314117, 0.12094850093126297, -0.18845385313034058, 0.004920545034110546, 0.042632684111595154, -0.1961911916732788, 0.5082730650901794, -0.05026670917868614, -0.17390176653862, -0.24324491620063782, 0.2771671414375305, -0.4623686373233795, -0.23436491191387177, 0.12246381491422653, 0.194098100066185, 0.2181559056043625, -0.31469640135765076, 0.5614801645278931, 0.12799696624279022, -0.011074633337557316, 0.04455452412366867, 0.13490189611911774, -0.20838187634944916, 0.4753076136112213, -0.2853442132472992, -0.025862855836749077, 0.06688477843999863, -0.004448235034942627, -0.09673922508955002, -0.37176477909088135, -0.04464820772409439, 0.317708283662796, 0.16569334268569946, -0.02841034345328808, 0.07400878518819809, 0.1382140964269638, -0.09192948788404465, -0.007477794773876667, -0.1643103063106537, -0.29819852113723755, 0.17177735269069672, 0.18262071907520294, 0.22253349423408508, 0.18226590752601624, -0.071593277156353, -0.23599857091903687, -0.12220046669244766, 0.47690534591674805, -0.008985565043985844, 0.24603380262851715, -0.13987475633621216, 0.3845987617969513, 0.4304969012737274, 0.5406715273857117, -0.2570675313472748, 0.12530675530433655, 0.04334172233939171, 0.08136963844299316, 0.5019499063491821, -0.22554343938827515, 0.038004450500011444, 0.2741834223270416, -0.14212453365325928, 0.04171757027506828, -0.28885871171951294, -0.16853711009025574, -0.2163938730955124, 0.11692984402179718, 0.12696067988872528, 0.024143794551491737, 0.03174252063035965, -0.05489109829068184, -0.1292056143283844, -0.058138225227594376, -0.010787956416606903, -0.3781210780143738, -0.0466112457215786, 0.13476991653442383, -0.06214899569749832, 0.04813925549387932, 0.045553773641586304, 0.24881422519683838, -0.005483593791723251, -0.04891285300254822, 0.3026910722255707, 0.11371723562479019, 0.1313488781452179, -0.1924312859773636, -0.22585037350654602, -0.02276783622801304, 0.27811652421951294, 0.16005299985408783, -0.08049102872610092, 0.15094345808029175, -0.2103303223848343, -0.14571766555309296, 0.19737188518047333, 0.1054658368229866, 0.06881043314933777, -0.0883580669760704, 0.14663836359977722, 0.10758323222398758, -0.0755477026104927, 0.013402139768004417, 0.19421455264091492, 0.0746554583311081, 0.0725075826048851, 0.10837575048208237, 0.1365596354007721, 0.4735901951789856, -0.22609297931194305, 0.003198835765942931, 0.12325024604797363, -0.3594178557395935, 0.15418651700019836, -0.08749885112047195, -0.06692653149366379, -0.031022153794765472, 0.1314779371023178, -0.059005215764045715, 0.13142898678779602, -0.3177984952926636, 0.3143520951271057, 0.29027891159057617, -0.0166154932230711, -0.023868560791015625, 0.24449287354946136, 0.03816725313663483, -0.14794552326202393, 0.29756519198417664, 0.04241351783275604, 0.3625636696815491, 0.031823113560676575, 0.8982049822807312, -0.28706395626068115, -0.44902175664901733, 0.08506789803504944, 0.4442313015460968, 0.0931386947631836, -0.21840958297252655, 0.12272963672876358, 0.1058456301689148, -0.3336717486381531, 0.12666085362434387, 0.027430497109889984, -0.16793394088745117, 0.6969410181045532, -0.08871027827262878, 0.13946987688541412, -0.1556522697210312, -0.1600608229637146, 0.37912964820861816, 0.1530623733997345, -0.08720098435878754, 0.005124739371240139, 0.11803250759840012, -0.11495554447174072, -0.17130598425865173, 0.5400869846343994, 0.19663314521312714, -0.01698659546673298, -0.441108763217926, -0.002918966580182314, 0.0928228348493576, 0.18584544956684113, -0.0031086925882846117, 0.11664816737174988, 0.42755255103111267, -0.17765945196151733, 0.2757043242454529, 0.016501229256391525, -0.042451679706573486, -0.12721553444862366, 0.041943926364183426, -0.3996351361274719, -0.4405321478843689, 0.43349653482437134, 0.126633420586586, -0.1013311818242073, 0.122391477227211, 0.3422187268733978, -0.6254706978797913, -0.21663402020931244, 0.328392893075943, -0.1351538598537445, -0.09978500753641129, -0.21278639137744904, 0.08299475908279419, 0.05206337943673134, 0.6268002390861511, 0.26694655418395996, 0.3559402823448181, -0.36786210536956787, -0.10077344626188278, -0.5913138389587402, -0.15383894741535187, 0.1531483381986618, -0.0713004395365715, -0.22856460511684418, 0.15664465725421906, -0.045735523104667664, 0.2795035243034363, -0.16623640060424805, 0.15578573942184448, 0.22006630897521973, 0.0632854700088501, -0.4629289507865906, 0.05816913768649101, 0.15672552585601807, 0.0008714653085917234, 0.2697930634021759, -0.09562154859304428, -0.18361014127731323, 0.10721495002508163, -0.08659328520298004, -0.13303780555725098, -0.083083376288414, 0.2785084843635559, 0.14434191584587097, 0.12579524517059326, 0.06052194535732269, 0.4831184446811676, -0.013349966146051884, -0.1953566074371338, 0.00032382557401433587, -0.017219046130776405, -0.3659600019454956, -0.13347701728343964, -0.11219020187854767, 0.16162464022636414, -0.40695860981941223, 0.3112415075302124, 0.02026781067252159, -0.11749887466430664, 0.07889237254858017, 0.08205991983413696, -0.19863149523735046, -0.10772024840116501, 0.0925135537981987, 0.17402318120002747, 0.22926753759384155, 0.706404447555542, -0.009655658155679703, 0.0815725326538086, -0.08162675052881241, -0.02464240975677967, 0.01380555797368288, 0.001977638341486454, -0.23659111559391022, -0.09722092747688293, 0.03598747402429581, 0.047070883214473724, 0.1454034298658371, -0.6879174113273621, -0.33900102972984314, 0.07124518603086472, 0.008100294508039951, 0.06026773154735565, 0.7683095932006836, 0.13407109677791595, -0.14332081377506256, -0.10919398069381714, -0.306508332490921, 0.028837328776717186, -0.2758439779281616, 0.1856124997138977, -0.3361104726791382 ]
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.
[ -0.08985186368227005, -0.39693301916122437, -0.03166532516479492, -0.03735155984759331, -0.22644886374473572, 0.08738168329000473, 0.19695472717285156, 0.0660226121544838, 0.3327304422855377, -0.13962098956108093, -0.22614960372447968, 0.33182400465011597, -0.20250491797924042, 0.36451777815818787, 0.40155574679374695, -0.5019195079803467, -0.00903155468404293, -0.17320945858955383, -0.3404366970062256, 0.34348368644714355, 0.015904966741800308, 0.0897597223520279, -0.12996907532215118, 0.05399265140295029, -0.2659047245979309, -0.14809370040893555, -0.34803786873817444, 0.10398362576961517, 0.09025166183710098, 0.0002099891717080027, 0.2654462158679962, 0.3834899663925171, -0.004184700548648834, 0.11468706279993057, -0.00011099899711553007, -0.10409428924322128, 0.17799219489097595, -0.06188741326332092, 0.13401709496974945, -0.022732706740498543, 0.2059175968170166, -0.3741091191768646, 0.0412948876619339, -0.14299578964710236, 0.09510323405265808, 0.186512291431427, 0.14864684641361237, 0.06131574138998985, 0.43896666169166565, -0.3271811604499817, 0.11371627449989319, 0.2715320587158203, -0.21278303861618042, 0.21605993807315826, -0.10353539884090424, -0.0817161351442337, 0.03618595376610756, 0.2948400378227234, 0.6407874822616577, -0.525549590587616, -0.026187710464000702, 0.09548535943031311, -0.027362259104847908, -0.026917099952697754, 0.16530896723270416, -0.0557938851416111, -0.353110134601593, -0.2346816211938858, -0.39337554574012756, 0.26181405782699585, -0.19633176922798157, -0.11107049882411957, -0.22825069725513458, -0.5546966791152954, 0.04327027499675751, 0.05650639906525612, -0.32799801230430603, 0.04984396696090698, -0.16306129097938538, 0.015066009946167469, -0.1603158563375473, -0.11377076804637909, -0.03513530641794205, 0.15094421803951263, 0.3342220187187195, 0.5391837954521179, 0.07172612845897675, 0.18103161454200745, 0.3762328624725342, 0.066431425511837, -0.15596945583820343, -0.25584715604782104, 0.2604007124900818, 0.14659355580806732, -0.30014482140541077, -0.38151082396507263, 0.10290294140577316, -0.20207516849040985, 0.2648243010044098, 0.16467316448688507, 0.0934794694185257, 0.12698109447956085, -0.11747134476900101, 0.28201183676719666, 0.3259006440639496, -0.0005844696424901485, -0.058437347412109375, -0.08932642638683319, 0.03309013321995735, -0.08139172941446304, 0.16326521337032318, 0.12141107767820358, 0.17396456003189087, 0.061764802783727646, -0.4309345483779907, -0.12424059212207794, -0.20832177996635437, 0.14269237220287323, -0.12683868408203125, -0.5166333913803101, -0.067426897585392, -0.08390925079584122, 0.06903011351823807, -0.08924651890993118, -0.30100777745246887, 0.11446746438741684, -0.02293545939028263, 0.21466882526874542, -0.06969939172267914, 0.014183021150529385, -0.054956987500190735, 0.11815524101257324, -0.42165589332580566, -0.016975687816739082, 0.3122231066226959, 0.09955603629350662, 0.0971217229962349, 0.17393353581428528, -0.07775566726922989, 0.12298347800970078, 0.42270323634147644, 0.029997050762176514, 0.02881641313433647, -0.0377974659204483, 0.14484566450119019, -0.32560494542121887, -0.006402771454304457, 0.1831968128681183, -0.3624861538410187, -0.15413980185985565, -0.09662583470344543, -0.16510269045829773, 0.12378998845815659, 0.0594375878572464, -0.22519682347774506, -0.27898308634757996, -0.37818220257759094, 0.7292918562889099, -0.06950251013040543, -0.061431583017110825, -0.17215366661548615, 0.2976945638656616, -0.25143396854400635, -0.07137691974639893, 0.1812436729669571, 0.07105004787445068, -0.032820843160152435, -0.38990509510040283, -0.10039552301168442, -0.10813401639461517, -0.07206139713525772, 0.26874998211860657, -0.20015184581279755, 0.21656769514083862, 0.04004440829157829, 0.1675095111131668, 0.2588686943054199, -0.4050792455673218, -0.10159745067358017, 0.2870189845561981, -0.037553559988737106, 0.3858805298805237, 0.07683325558900833, 0.24547551572322845, 0.08919310569763184, -0.17639215290546417, 0.2526645362377167, 1.0071951150894165, -0.39769551157951355, -0.06931670755147934, -0.10972336679697037, -0.38633719086647034, 0.5040083527565002, 0.3668293058872223, 0.10830044001340866, -0.23684105277061462, -0.14913038909435272, -0.05921345576643944, 0.06534450501203537, 0.045004915446043015, 0.10629872232675552, -0.08196774870157242, -0.3427743911743164, 0.030671704560518265, -0.21699616312980652, -0.20223870873451233, -0.36724185943603516, -0.002257948974147439, -0.06509006768465042, 0.3331247866153717, 0.37277644872665405, -0.19357994198799133, 0.18135853111743927, -0.2651146352291107, -0.0785321518778801, -0.2893359661102295, 0.13996616005897522, -0.07206795364618301, 0.05796969681978226, -0.2934381067752838, -0.04770268499851227, 0.3240184783935547, 0.06630038470029831, -0.09036249667406082, -0.4725522994995117, 0.31648769974708557, -0.06607350707054138, 0.041039615869522095, -0.06410323828458786, 0.6213017702102661, -0.16827833652496338, -0.13204902410507202, 0.2150787115097046, 0.22514663636684418, -0.32856523990631104, 0.21186082065105438, 0.3367438018321991, 0.24053174257278442, 0.17107193171977997, -0.05437030270695686, 0.10397021472454071, 0.019159287214279175, -0.012394275516271591, -0.22969655692577362, -0.10859386622905731, 0.5694275498390198, -0.0680292546749115, 0.4326314628124237, -0.026005800813436508, 0.042701683938503265, -0.18314501643180847, -0.11107352375984192, -0.2428102195262909, 0.3811073303222656, 0.29904961585998535, -0.2120334655046463, 0.35236018896102905, -0.05134245753288269, -0.35533979535102844, 0.13639573752880096, 0.13285283744335175, 0.0007289279019460082, 0.08132331073284149, 0.01811354234814644, -0.06393161416053772, -0.2471839189529419, -0.09759613126516342, 0.4116668701171875, 0.5995680689811707, 0.29787203669548035, 0.14136427640914917, 0.0798589289188385, -0.25959107279777527, -0.07501405477523804, 0.03891154006123543, 0.19106236100196838, -0.07945382595062256, 0.1542113572359085, 0.010218797251582146, 0.05711279809474945, -0.015598952770233154, -0.09665276110172272, 0.12611663341522217, -0.22449806332588196, -0.02574712224304676, 0.04047052934765816, -0.15040050446987152, -0.5861158967018127, -0.23188872635364532, 0.00516112707555294, -0.31230685114860535, -0.26452410221099854, 0.14035670459270477, 0.039329200983047485, -0.17840607464313507, 0.15653905272483826, 0.18521945178508759, 0.6711624264717102, -0.4588549733161926, -0.22821266949176788, -0.05708075687289238, -0.2689541280269623, -0.11032219976186752, 0.12571710348129272, 0.42610594630241394, 0.27763333916664124, 0.5188314914703369, 0.30895689129829407, -0.0348958894610405, 0.06895484030246735, -0.39623183012008667, 0.05782638490200043, -0.2660854756832123, 0.3105849623680115, 0.11344971507787704, -0.22008086740970612, 0.17031775414943695, -0.37428948283195496, 0.0022702880669385195, -0.10265398025512695, -0.0340799055993557, -0.07405850291252136, -0.04390918090939522, 0.08226983994245529, -0.20276083052158356, -0.1946839690208435, -0.638298511505127, -0.473905473947525, 0.3052135705947876, -0.21334078907966614, 0.06977508217096329, 0.019458618015050888, -0.25893694162368774, 0.1785203218460083, -0.13915324211120605, -0.01368755754083395, -0.1945847123861313, -0.1289626955986023, 0.01676172763109207, -0.33759644627571106, -0.16410107910633087, -0.4051864743232727, -0.20086058974266052, 0.22619272768497467, -0.09001284092664719, -0.06369994580745697, -0.25243034958839417, 0.07718147337436676, -0.12796752154827118, 0.1112930104136467, 0.2199832648038864, 0.1511763483285904, -0.2682066857814789, 0.09364637732505798, -0.02915474772453308, -0.20805364847183228, 0.2571520209312439, 0.2949787378311157, 0.37840792536735535, 0.15610384941101074, 0.03820465877652168, 0.13332583010196686, 0.7063825726509094, 0.24469418823719025, 0.03195074573159218, 0.14577065408229828, 0.3078962564468384, 0.017253490164875984, 0.15771785378456116, -0.22609084844589233, 0.19193771481513977, -0.17036138474941254, 0.4489412307739258, 0.10616744309663773, 0.19418956339359283, -0.20405061542987823, -0.2649410367012024, 0.05006043612957001, -0.22742311656475067, -0.23034416139125824, 0.26681947708129883, -0.18870267271995544, 0.011721186339855194, -0.09040232747793198, -0.36917421221733093, -0.246987447142601, -0.02260042540729046, 0.016349250450730324, 0.26587480306625366, -0.023195816203951836, 0.1255914866924286, -0.4113297760486603, -0.036341991275548935, -0.43079227209091187, 0.18226012587547302, 0.07882989197969437, 0.26000627875328064, -0.06049290671944618, -0.07115327566862106, -0.08950971066951752, 0.10068026930093765, 0.5856121778488159, -0.5367429256439209, -0.1742345094680786, 0.11988387256860733, 0.07640209048986435, -0.14059610664844513, -0.06939814239740372, -0.2734324634075165, -0.04439018666744232, 0.4505394697189331, 0.32974350452423096, -0.3563489317893982, -0.2867955267429352, 0.29957589507102966, -0.17227564752101898, -0.09658415615558624, -0.20668788254261017, -0.15361686050891876, -0.1994931548833847, -0.13871710002422333, 0.05471839755773544, 0.18321536481380463, 0.15969330072402954, -0.10423213988542557, 0.12230970710515976, -0.11102587729692459, 0.21563135087490082, 0.19780385494232178, 0.009142747148871422, -0.04508138447999954, 0.21428565680980682, 0.17898501455783844, 0.03236948326230049, -0.14142341911792755, -0.3413660526275635, 0.12961746752262115, -0.0628696009516716, 0.12924633920192719, 0.2377246618270874, -0.015323272906243801, 0.6900308728218079, 0.042731888592243195, 0.352876752614975, 0.3185281753540039, -0.10639480501413345, 0.009463772177696228, -0.3053721487522125, 0.13097627460956573, 0.16529573500156403, 0.24443571269512177, -0.11793571710586548, -0.22836296260356903, 0.5872114896774292, 0.3297060430049896, -0.08609979599714279, 0.3757953345775604, -0.47376933693885803, -0.3104906380176544, 0.32061371207237244, 0.0482926145195961, 0.7352598309516907, -0.020149782299995422, 0.2903290390968323, 0.06005857512354851, -0.3805789649486542, 0.6396000385284424, -0.3495900630950928, 0.12659981846809387, -0.13993500173091888, -0.20601853728294373, 0.008966943249106407, -0.08151328563690186, -0.061738163232803345, 0.27457961440086365, -0.3608982264995575, 0.306580513715744, 0.3756426274776459, -0.22479882836341858, -0.02693840116262436, 0.5579148530960083, -0.028485896065831184, -0.40201911330223083, -0.07268843799829483, 0.10170155763626099, -0.2750024199485779, 0.06954970210790634, -0.1666456162929535, -0.23241932690143585, 0.2085064947605133, 0.1353931576013565, -0.2649179995059967, 0.10901223123073578, -0.021995963528752327, 0.15653365850448608, 0.27810898423194885, 0.021803904324769974, 0.274069607257843, -0.2461932748556137, -0.0806761085987091, -0.10604246705770493, -0.08823467791080475, 0.13408750295639038, -0.22274012863636017, 0.0591328926384449, 0.07656443864107132, -0.1661917120218277, 0.5294389128684998, -0.029330365359783173, -0.1059572845697403, -0.20204512774944305, 0.23767612874507904, -0.42896097898483276, -0.1612631231546402, 0.10507817566394806, 0.194949209690094, 0.12262522429227829, -0.3510892987251282, 0.5494359731674194, 0.09299591928720474, -0.0307429488748312, 0.059492334723472595, 0.16084745526313782, -0.28504326939582825, 0.46693897247314453, -0.20441311597824097, 0.014504993334412575, 0.06675849854946136, -0.034968674182891846, -0.0981704518198967, -0.3632012605667114, -0.03582318127155304, 0.28807690739631653, 0.1427469700574875, -0.045033153146505356, 0.008256928063929081, 0.1737511157989502, -0.0649222657084465, -0.0009059262229129672, -0.10859963297843933, -0.36780935525894165, 0.13817249238491058, 0.15505193173885345, 0.281816303730011, 0.2077016532421112, -0.10849197208881378, -0.24853521585464478, -0.22846628725528717, 0.45401978492736816, 0.08564350008964539, 0.2493201494216919, -0.13096792995929718, 0.3653296232223511, 0.4037599563598633, 0.5555815100669861, -0.28828123211860657, 0.047473832964897156, 0.030073637142777443, 0.0926191508769989, 0.4883192181587219, -0.24070203304290771, 0.0004643139836844057, 0.27189964056015015, -0.12661147117614746, 0.06990136206150055, -0.335458904504776, -0.1801498383283615, -0.27896952629089355, 0.10328447818756104, 0.12159030884504318, 0.004378818906843662, 0.019434254616498947, -0.025179723277688026, -0.09327100217342377, -0.07621198892593384, -0.03240201622247696, -0.4200456440448761, -0.05962028726935387, 0.13473346829414368, -0.07899182289838791, 0.04574650526046753, -0.004300149157643318, 0.21682432293891907, -0.005198086146265268, -0.04506367817521095, 0.27724534273147583, 0.08747972548007965, 0.1442188173532486, -0.1765517294406891, -0.14229907095432281, 0.006133261136710644, 0.3090972304344177, 0.10410712659358978, -0.02817574515938759, 0.17523446679115295, -0.20415666699409485, -0.07347269356250763, 0.26419728994369507, 0.19137431681156158, 0.028277669101953506, -0.06573779881000519, 0.12197975069284439, 0.15415164828300476, -0.13383083045482635, 0.03742095082998276, 0.24652691185474396, 0.052805472165346146, 0.12569411098957062, 0.1320367157459259, 0.15263526141643524, 0.4616326689720154, -0.2277931421995163, -0.009581931866705418, 0.1340147852897644, -0.3001982569694519, 0.1663927584886551, -0.14242252707481384, -0.0036705981474369764, -0.018842993304133415, 0.14484095573425293, -0.030460627749562263, 0.1771898716688156, -0.3630976974964142, 0.2931552529335022, 0.3141862750053406, -0.012024542316794395, 0.04972730204463005, 0.21571557223796844, 0.011386317200958729, -0.1301044076681137, 0.309308260679245, 0.09537197649478912, 0.3369997441768646, 0.0031128544360399246, 0.8959373831748962, -0.3085148334503174, -0.49877575039863586, 0.08208055794239044, 0.4228001832962036, 0.03482484444975853, -0.1993444412946701, 0.08230797946453094, 0.12727196514606476, -0.36981824040412903, 0.07597450166940689, 0.07843571901321411, -0.18659216165542603, 0.6474287509918213, -0.11679477989673615, 0.16359639167785645, -0.17023111879825592, -0.20496292412281036, 0.34298175573349, 0.15966221690177917, -0.13731500506401062, 0.027703117579221725, 0.11107561737298965, -0.06848157942295074, -0.16262051463127136, 0.5162652134895325, 0.17436501383781433, 0.01610332541167736, -0.44874680042266846, 0.041226577013731, 0.050035975873470306, 0.15217261016368866, -0.08710531145334244, 0.1192086786031723, 0.44169026613235474, -0.16612090170383453, 0.24096070230007172, 0.03840499743819237, -0.10613936930894852, -0.15496870875358582, 0.014971940778195858, -0.3957519233226776, -0.4279156029224396, 0.4407229423522949, 0.12134905159473419, -0.1510057896375656, 0.018220704048871994, 0.2517421543598175, -0.612326979637146, -0.2271735668182373, 0.35456332564353943, -0.17337359488010406, -0.07675124704837799, -0.2256258726119995, 0.08382951468229294, 0.03794966638088226, 0.6115516424179077, 0.2173786759376526, 0.35958367586135864, -0.3842931389808655, -0.17158129811286926, -0.5617933869361877, -0.18688637018203735, 0.14032027125358582, -0.10462841391563416, -0.24724498391151428, 0.16251885890960693, -0.07911871373653412, 0.29674139618873596, -0.22634288668632507, 0.1232830211520195, 0.2228541374206543, -0.02459866739809513, -0.41087034344673157, 0.12129537016153336, 0.12449537217617035, 0.021148057654500008, 0.3134693503379822, -0.08089785277843475, -0.21032847464084625, 0.07324440777301788, -0.07079581916332245, -0.03564117103815079, -0.08734451234340668, 0.26318785548210144, 0.14747077226638794, 0.11267495155334473, 0.0629214271903038, 0.5086536407470703, -0.026272781193256378, -0.16295543313026428, 0.020117634907364845, -0.0493784174323082, -0.3885861039161682, -0.12324897944927216, -0.08058823645114899, 0.2154364138841629, -0.3562215268611908, 0.25851935148239136, 0.004246820695698261, -0.15498511493206024, 0.14422044157981873, 0.032606132328510284, -0.2562313973903656, -0.058259088546037674, -0.012389660812914371, 0.21576137840747833, 0.21925494074821472, 0.7426396608352661, 0.02795436978340149, 0.08706792443990707, -0.03915257751941681, -0.009825468063354492, 0.08792974054813385, -0.04098895192146301, -0.22142860293388367, -0.16324186325073242, 0.010918770916759968, 0.015441704541444778, 0.09817137569189072, -0.6646649241447449, -0.3680718243122101, 0.0423702634871006, -0.0076064737513661385, 0.09818343818187714, 0.7144815921783447, 0.08214782923460007, -0.1468360424041748, -0.1454651653766632, -0.27579137682914734, 0.03483705222606659, -0.24518166482448578, 0.19327954947948456, -0.3910033106803894 ]
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.
[ -0.029288018122315407, -0.3940451145172119, -0.06102925166487694, -0.02566205896437168, -0.2608688473701477, 0.08258087188005447, 0.18105271458625793, 0.046955324709415436, 0.28684258460998535, -0.1594536453485489, -0.23304539918899536, 0.3090500831604004, -0.2074708193540573, 0.38534125685691833, 0.37004509568214417, -0.4468323886394501, -0.007145140785723925, -0.20719198882579803, -0.26221925020217896, 0.39257070422172546, 0.03699754923582077, 0.1119900792837143, -0.14147144556045532, 0.10980471968650818, -0.30104711651802063, -0.13675542175769806, -0.33219119906425476, 0.10305801779031754, 0.11648951470851898, 0.018128694966435432, 0.21229079365730286, 0.36170390248298645, -0.047550592571496964, 0.14389604330062866, -0.00011497207742650062, -0.04456184059381485, 0.14054247736930847, -0.07454655319452286, 0.12798768281936646, -0.03853219002485275, 0.172773540019989, -0.3879152536392212, 0.06748507916927338, -0.13466723263263702, 0.21631959080696106, 0.15920691192150116, 0.18093286454677582, 0.06926141679286957, 0.4056938588619232, -0.32851308584213257, 0.0875927284359932, 0.28492650389671326, -0.19006867706775665, 0.34371379017829895, -0.10634822398424149, -0.1040676012635231, 0.03765297681093216, 0.25891444087028503, 0.5864796042442322, -0.591979444026947, 0.053165026009082794, 0.08587003499269485, 0.012978083454072475, -0.0354301817715168, 0.158354714512825, -0.028579382225871086, -0.31108126044273376, -0.2244444340467453, -0.41868317127227783, 0.18872618675231934, -0.10443771630525589, -0.15518957376480103, -0.2215287834405899, -0.5562808513641357, -0.017073987051844597, -0.0037985090166330338, -0.30292555689811707, -0.002131415531039238, -0.17310242354869843, -0.039497632533311844, -0.14767669141292572, -0.10621426999568939, -0.05107506364583969, 0.15497079491615295, 0.3348459303379059, 0.5709914565086365, 0.12959901988506317, 0.17109747231006622, 0.3198665380477905, 0.053875960409641266, -0.1331726312637329, -0.29221659898757935, 0.30810120701789856, 0.164288729429245, -0.36470887064933777, -0.35979118943214417, 0.19177576899528503, -0.25898078083992004, 0.23242567479610443, 0.09343162924051285, 0.127032071352005, 0.08136390149593353, -0.14478690922260284, 0.2692345976829529, 0.33163177967071533, 0.036866798996925354, -0.1052708774805069, -0.12733294069766998, 0.02876291424036026, -0.1599457710981369, 0.23521602153778076, 0.09670251607894897, 0.18593545258045197, -0.022179437801241875, -0.3954426944255829, -0.16337120532989502, -0.134634867310524, 0.14888425171375275, -0.19198042154312134, -0.5355587601661682, -0.1626533418893814, -0.13995642960071564, 0.047498367726802826, -0.13046735525131226, -0.3142300546169281, 0.07659867405891418, -0.030082112178206444, 0.18955890834331512, -0.14268597960472107, 0.12083221226930618, -0.052640508860349655, 0.07124767452478409, -0.4480918347835541, -0.04607677459716797, 0.2776445150375366, 0.10952528566122055, 0.08483005315065384, 0.13384391367435455, 0.014841342344880104, 0.12548010051250458, 0.4019384980201721, 0.03929176926612854, -0.01993747241795063, -0.03314130753278732, 0.16819174587726593, -0.3317156732082367, -0.00626253942027688, 0.1757427155971527, -0.3530603349208832, -0.14903011918067932, -0.15961135923862457, -0.18683291971683502, 0.12111599743366241, 0.024497414007782936, -0.220840185880661, -0.3201463222503662, -0.47823530435562134, 0.6768115162849426, -0.027634093537926674, -0.0647311583161354, -0.16472403705120087, 0.2757859230041504, -0.2691700756549835, -0.12027686834335327, 0.17447838187217712, 0.0458286851644516, 0.0011193964164704084, -0.3560561537742615, -0.0667320266366005, -0.07920414954423904, -0.07760483771562576, 0.25847142934799194, -0.17129398882389069, 0.2623891830444336, 0.09088419377803802, 0.18576352298259735, 0.3408809006214142, -0.47314998507499695, -0.02268316224217415, 0.2970295250415802, -0.055632494390010834, 0.41089123487472534, 0.07046733796596527, 0.227961003780365, -0.08120615035295486, -0.20891094207763672, 0.27190226316452026, 1.0307588577270508, -0.37644118070602417, -0.02222248539328575, -0.09570226818323135, -0.3904015123844147, 0.5535186529159546, 0.3494327962398529, 0.0771406888961792, -0.30797070264816284, -0.10218051075935364, -0.00963363703340292, 0.0667359009385109, 0.0841098502278328, 0.07473359256982803, -0.08781099319458008, -0.3393523097038269, 0.11391221731901169, -0.2102668136358261, -0.15189921855926514, -0.3530142307281494, 0.03392348438501358, 0.024166803807020187, 0.31401923298835754, 0.25735363364219666, -0.13017821311950684, 0.20869873464107513, -0.28740164637565613, -0.14865370094776154, -0.2948306202888489, 0.10739147663116455, -0.07125716656446457, 0.0002877805382013321, -0.23797887563705444, -0.03177556395530701, 0.3077722489833832, 0.1399240344762802, -0.1362244337797165, -0.477207750082016, 0.29331859946250916, -0.06322244554758072, 0.05785408243536949, -0.05632902681827545, 0.6482880115509033, -0.1621541529893875, -0.06727930903434753, 0.19220846891403198, 0.3254878520965576, -0.3306892514228821, 0.2363167256116867, 0.31374433636665344, 0.3130965232849121, 0.16561447083950043, 0.006151277106255293, 0.11799031496047974, 0.011639831587672234, -0.0012402519350871444, -0.21448662877082825, -0.0947565883398056, 0.5063757300376892, -0.1947738081216812, 0.42482978105545044, -0.023951217532157898, -0.0024470328353345394, -0.2053104043006897, -0.01521735917776823, -0.1752176433801651, 0.41539132595062256, 0.29060009121894836, -0.23537136614322662, 0.2970910370349884, -0.16056793928146362, -0.3739416003227234, 0.17211981117725372, 0.17921607196331024, 0.018867259845137596, 0.034404728561639786, -0.04453393444418907, 0.007732796482741833, -0.27543824911117554, -0.10600510984659195, 0.31868237257003784, 0.657733142375946, 0.2544577419757843, 0.13760989904403687, 0.06570456922054291, -0.2021636813879013, -0.059559114277362823, 0.0485229566693306, 0.21359142661094666, -0.0488939993083477, 0.1590915024280548, 0.03624701499938965, 0.08083909004926682, 0.062116190791130066, -0.16772831976413727, 0.1392095983028412, -0.20449788868427277, -0.11621738970279694, 0.0541704036295414, -0.18401747941970825, -0.6338427662849426, -0.1893036812543869, -0.009177280589938164, -0.3316514194011688, -0.2957560122013092, 0.1372014582157135, 0.05582486838102341, -0.11464421451091766, 0.13877370953559875, 0.1970457285642624, 0.5815643072128296, -0.43081381916999817, -0.2768557071685791, -0.03663836419582367, -0.3137177526950836, -0.10173459351062775, 0.10424809157848358, 0.4355732798576355, 0.29597726464271545, 0.5036947727203369, 0.2990516424179077, -0.06889355927705765, 0.09869717061519623, -0.40217214822769165, 0.06541761010885239, -0.3533985912799835, 0.25853002071380615, 0.1358286589384079, -0.22173333168029785, 0.14582949876785278, -0.3897811770439148, -0.0530332513153553, -0.025669774040579796, -0.07492628693580627, -0.05133776366710663, 0.05406815931200981, 0.06931345164775848, -0.19855491816997528, -0.1903681457042694, -0.6149934530258179, -0.44191795587539673, 0.30627501010894775, -0.22799819707870483, 0.08272524923086166, 0.062167923897504807, -0.2821543514728546, 0.23458458483219147, -0.08945964276790619, 0.02536354400217533, -0.17027702927589417, -0.05233968794345856, 0.010179697535932064, -0.38480690121650696, -0.14478427171707153, -0.33935216069221497, -0.03989516198635101, 0.22139795124530792, -0.06605792790651321, -0.029407303780317307, -0.27509206533432007, 0.08280228078365326, -0.08778975903987885, 0.07219218462705612, 0.26224303245544434, 0.16565285623073578, -0.20603437721729279, 0.13177162408828735, -0.08111485093832016, -0.27193498611450195, 0.2103467434644699, 0.330236554145813, 0.3737384080886841, 0.18449372053146362, 0.03631793335080147, 0.1870741844177246, 0.7639585733413696, 0.24814310669898987, 0.0630791038274765, 0.14840330183506012, 0.2611877918243408, -0.012645253911614418, 0.1614735722541809, -0.1882694810628891, 0.18401655554771423, -0.1733156144618988, 0.5277930498123169, 0.08122697472572327, 0.16958020627498627, -0.14921656250953674, -0.33210939168930054, 0.018026968464255333, -0.31423598527908325, -0.18452827632427216, 0.1901613175868988, -0.2771724760532379, 0.017156418412923813, -0.09102130681276321, -0.436222106218338, -0.2857382893562317, -0.037379562854766846, -0.017000535503029823, 0.33567842841148376, -0.006728521082550287, 0.17077063024044037, -0.33541014790534973, -0.04341873526573181, -0.421749085187912, 0.14420609176158905, 0.05830398574471474, 0.23728244006633759, -0.08073709160089493, -0.0721387192606926, -0.10103421658277512, 0.11182695627212524, 0.5621090531349182, -0.5473765730857849, -0.10901261866092682, 0.09505578130483627, 0.07976459711790085, -0.12696301937103271, -0.116548553109169, -0.29187896847724915, 0.020434163510799408, 0.42763015627861023, 0.37687984108924866, -0.37812379002571106, -0.2943046987056732, 0.2408275604248047, -0.253957599401474, -0.07215578854084015, -0.19330410659313202, -0.2405911087989807, -0.21128340065479279, -0.12438025325536728, -0.005925411358475685, 0.2536102533340454, 0.23441122472286224, -0.1205020546913147, 0.1836339384317398, -0.08123071491718292, 0.2604241967201233, 0.24369798600673676, 0.010381501168012619, -0.11040745675563812, 0.2441926896572113, 0.17174625396728516, 0.038996465504169464, -0.12102340161800385, -0.3591463267803192, 0.09985440969467163, -0.02167346701025963, 0.06650105863809586, 0.25530391931533813, -0.0018159348983317614, 0.6958091855049133, 0.04607256129384041, 0.32781824469566345, 0.27770766615867615, -0.14805646240711212, -0.002945331856608391, -0.2239086925983429, 0.1858537644147873, 0.23404711484909058, 0.24906140565872192, -0.14151251316070557, -0.23254236578941345, 0.5836414694786072, 0.3070313632488251, -0.08686497807502747, 0.4385268986225128, -0.4941880702972412, -0.33374685049057007, 0.21709123253822327, 0.02576623670756817, 0.7134888768196106, -0.022532032802700996, 0.3297562003135681, 0.11808132380247116, -0.41327378153800964, 0.5890750288963318, -0.3362225294113159, 0.12408442050218582, -0.05483768880367279, -0.17275108397006989, -0.011037030257284641, -0.07591129094362259, -0.10295789688825607, 0.20575565099716187, -0.3488755524158478, 0.33698567748069763, 0.403509259223938, -0.23023754358291626, -0.04739336296916008, 0.5434771180152893, -0.014471079222857952, -0.34057825803756714, -0.07095809280872345, 0.05130472406744957, -0.2821178138256073, 0.06947988271713257, -0.21265527606010437, -0.25520142912864685, 0.21050380170345306, 0.15928414463996887, -0.2744199335575104, 0.06679590791463852, 0.008197980932891369, 0.20083540678024292, 0.22477668523788452, 0.03559299558401108, 0.2640020251274109, -0.17304953932762146, -0.10373122245073318, -0.17674151062965393, -0.06488712131977081, 0.10578858852386475, -0.23131319880485535, 0.03633549064397812, 0.09533291310071945, -0.15070922672748566, 0.49384957551956177, -0.07441750913858414, -0.10552697628736496, -0.22589978575706482, 0.2618072032928467, -0.44568172097206116, -0.2548682391643524, 0.1317417323589325, 0.16105273365974426, 0.18171559274196625, -0.3218861520290375, 0.5945082306861877, 0.08666704595088959, 0.05346429720520973, 0.01800571195781231, 0.17048412561416626, -0.23972730338573456, 0.4645397961139679, -0.3223522901535034, 0.00029868874116800725, 0.09918472915887833, 0.00014896398351993412, -0.15182992815971375, -0.418944776058197, -0.05358614772558212, 0.2748265564441681, 0.12449437379837036, 0.05534154921770096, -0.016994718462228775, 0.11429882049560547, -0.03799332305788994, -0.05551133304834366, -0.11189225316047668, -0.2893208861351013, 0.1342899203300476, 0.24401691555976868, 0.1765126734972, 0.1978478729724884, -0.06045486778020859, -0.2944515645503998, -0.13006706535816193, 0.4992499053478241, -0.016898609697818756, 0.23208345472812653, -0.1513768434524536, 0.37086325883865356, 0.41062816977500916, 0.490372896194458, -0.24310001730918884, 0.05704471841454506, 0.09853258728981018, 0.10004331171512604, 0.46761199831962585, -0.22680526971817017, 0.04738743603229523, 0.28212013840675354, -0.14611603319644928, 0.031245769932866096, -0.3519592881202698, -0.13588055968284607, -0.24969306588172913, 0.12456842511892319, 0.1288156658411026, -0.02847566455602646, 0.02841692790389061, -0.11128661781549454, -0.10461035370826721, -0.057755421847105026, 0.04444063827395439, -0.36330124735832214, -0.0975726842880249, 0.1614966243505478, -0.09412816911935806, 0.08671195805072784, 0.05705138295888901, 0.20702826976776123, 0.06947349011898041, -0.07310902327299118, 0.33872029185295105, 0.12007112056016922, 0.08686445653438568, -0.16815894842147827, -0.14083050191402435, 0.01643958128988743, 0.2955598831176758, 0.18031568825244904, -0.0689324364066124, 0.14224345982074738, -0.2093847095966339, -0.1997295320034027, 0.20770074427127838, 0.20243293046951294, 0.03528780862689018, -0.1302381008863449, 0.10691755264997482, 0.09648487716913223, -0.060015637427568436, -0.012907352298498154, 0.23015180230140686, 0.08779218792915344, 0.07244808226823807, 0.11680933833122253, 0.11368706822395325, 0.4800722599029541, -0.2822709381580353, 0.03274901211261749, 0.09447211772203445, -0.3594554662704468, 0.11638288199901581, -0.08243805170059204, -0.08555764704942703, -0.014490846544504166, 0.12477201223373413, 0.00008948001777753234, 0.11933604627847672, -0.33420175313949585, 0.32404956221580505, 0.2672713100910187, 0.015397962182760239, 0.01500164158642292, 0.13875192403793335, 0.04229080677032471, -0.17318104207515717, 0.2679607570171356, 0.03501957282423973, 0.3751721680164337, 0.06041334196925163, 0.8366438150405884, -0.21346957981586456, -0.3915119171142578, 0.15799497067928314, 0.42227041721343994, 0.08762931078672409, -0.21886204183101654, 0.13186530768871307, 0.09811235219240189, -0.3754848837852478, 0.069796122610569, 0.05699251964688301, -0.18208125233650208, 0.6210578083992004, -0.08598873764276505, 0.16826127469539642, -0.11792013794183731, -0.18515902757644653, 0.35806751251220703, 0.21875640749931335, -0.0734490379691124, -0.03960030525922775, 0.1282282918691635, -0.061942171305418015, -0.19880656898021698, 0.4743296205997467, 0.21330268681049347, -0.00240953266620636, -0.48168671131134033, 0.03337971121072769, 0.10145290195941925, 0.17900334298610687, 0.0025961576029658318, 0.13696521520614624, 0.4460620880126953, -0.15026040375232697, 0.20349405705928802, -0.007044783793389797, -0.03401824086904526, -0.1344565600156784, 0.06418363749980927, -0.40295514464378357, -0.42371538281440735, 0.46206051111221313, 0.11025841534137726, -0.14952687919139862, 0.11755193024873734, 0.3762097954750061, -0.6295855045318604, -0.26949021220207214, 0.3343755900859833, -0.05823202058672905, -0.08502041548490524, -0.19609695672988892, 0.06549713015556335, 0.07512515783309937, 0.6126105189323425, 0.20807671546936035, 0.4082013964653015, -0.3459770679473877, -0.09671197086572647, -0.5682132244110107, -0.20369435846805573, 0.17174029350280762, -0.15297995507717133, -0.2241836041212082, 0.11838730424642563, -0.06933396309614182, 0.27628377079963684, -0.18905538320541382, 0.10309771448373795, 0.24528980255126953, 0.04981963336467743, -0.43662115931510925, 0.08494255691766739, 0.19068247079849243, 0.040402915328741074, 0.2738912105560303, -0.0639430582523346, -0.22662174701690674, 0.0473250150680542, -0.06836356222629547, -0.09690742939710617, -0.11419856548309326, 0.2681552767753601, 0.18392759561538696, 0.09595850110054016, 0.05707346275448799, 0.5114500522613525, 0.010969377122819424, -0.15749339759349823, 0.0035763652995228767, 0.043719690293073654, -0.3633820116519928, -0.13569672405719757, -0.09348126500844955, 0.13298876583576202, -0.3916036784648895, 0.3226492404937744, 0.0057075135409832, -0.0780506283044815, 0.09187621623277664, 0.0003342302516102791, -0.12278403341770172, -0.11219867318868637, 0.043778251856565475, 0.22140483558177948, 0.18611280620098114, 0.7520401477813721, -0.012685087509453297, 0.13772818446159363, -0.09494926035404205, -0.008783330209553242, 0.02726873755455017, -0.06434711068868637, -0.2542928159236908, -0.09406179189682007, 0.06744574010372162, 0.06809119135141373, 0.14320766925811768, -0.6984320878982544, -0.3541291356086731, 0.05046319216489792, 0.016152385622262955, 0.015999432653188705, 0.7800260782241821, 0.09473782777786255, -0.13276280462741852, -0.08945008367300034, -0.3622458279132843, 0.0005289989057928324, -0.22626972198486328, 0.16598643362522125, -0.374137282371521 ]
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!
[ -0.041029609739780426, -0.38001424074172974, -0.056135889142751694, -0.05533311888575554, -0.22696135938167572, 0.038144826889038086, 0.19116364419460297, 0.028541211038827896, 0.3140307366847992, -0.16932393610477448, -0.2113306075334549, 0.3226812779903412, -0.21177799999713898, 0.36092787981033325, 0.4108109772205353, -0.4081553518772125, 0.04961150884628296, -0.238549143075943, -0.30185481905937195, 0.3816818594932556, 0.029433391988277435, 0.11054188013076782, -0.18156097829341888, 0.11165957152843475, -0.27534905076026917, -0.1470182240009308, -0.3376018702983856, 0.105767160654068, 0.12588639557361603, 0.00012279996008146554, 0.25411534309387207, 0.4050157070159912, -0.06453400105237961, 0.1640157401561737, -0.00011387034464860335, -0.028973232954740524, 0.23698237538337708, -0.06454761326313019, 0.11490747332572937, -0.07505486905574799, 0.20829954743385315, -0.4271637797355652, 0.05915158614516258, -0.1675683706998825, 0.17296102643013, 0.15671952068805695, 0.15836820006370544, 0.04002639651298523, 0.39363452792167664, -0.2916635274887085, 0.10325055569410324, 0.28750285506248474, -0.15446849167346954, 0.2761387825012207, -0.12511217594146729, -0.04768775776028633, 0.03673860803246498, 0.2582833766937256, 0.5602096319198608, -0.6138644218444824, 0.0037750480696558952, 0.11586388200521469, 0.0011097949463874102, 0.004889057949185371, 0.15319648385047913, -0.020795093849301338, -0.3521726429462433, -0.21768818795681, -0.403907835483551, 0.18628624081611633, -0.14515520632266998, -0.20172210037708282, -0.23753738403320312, -0.5448266267776489, 0.004670035094022751, 0.006173086818307638, -0.26987841725349426, -0.018358584493398666, -0.20179778337478638, -0.06324595957994461, -0.12326818704605103, -0.0896952748298645, -0.03366657346487045, 0.15525327622890472, 0.31824907660484314, 0.5842542052268982, 0.09058943390846252, 0.16732610762119293, 0.36797162890434265, 0.04139328375458717, -0.15085747838020325, -0.2804940342903137, 0.32753613591194153, 0.1458226442337036, -0.3698505163192749, -0.30306077003479004, 0.20416361093521118, -0.2367648333311081, 0.22982420027256012, 0.1045718789100647, 0.09047479927539825, 0.1172202005982399, -0.14798341691493988, 0.26190418004989624, 0.3045746684074402, 0.046121951192617416, -0.09614865481853485, -0.09544767439365387, 0.03467637673020363, -0.08594397455453873, 0.192735493183136, 0.11217812448740005, 0.18549804389476776, -0.012385286390781403, -0.3950408101081848, -0.1339057832956314, -0.1293439269065857, 0.16474471986293793, -0.21528463065624237, -0.48490285873413086, -0.10857389867305756, -0.11952351033687592, 0.11771432310342789, -0.1045864149928093, -0.309135764837265, 0.09875436127185822, -0.03895292431116104, 0.2612784504890442, -0.13749749958515167, 0.1306164562702179, -0.07104401290416718, 0.0840979591012001, -0.44123634696006775, -0.07710482180118561, 0.3059290051460266, 0.1518343985080719, 0.0807022675871849, 0.16790179908275604, -0.0014317177701741457, 0.12360255420207977, 0.38938185572624207, 0.015577934682369232, 0.0013970992295071483, -0.0009416155517101288, 0.13964267075061798, -0.3161430358886719, 0.020423496142029762, 0.1589081585407257, -0.3474521338939667, -0.19029997289180756, -0.11934830248355865, -0.168540820479393, 0.14057384431362152, 0.03215436637401581, -0.20059600472450256, -0.23849967122077942, -0.49818819761276245, 0.7015706300735474, -0.05721471831202507, -0.09345905482769012, -0.17591449618339539, 0.3037513494491577, -0.24646389484405518, -0.08007054030895233, 0.21196672320365906, 0.05764036253094673, 0.011350923217833042, -0.3442133069038391, -0.10541749000549316, -0.09868180751800537, -0.02154751494526863, 0.25007927417755127, -0.16200292110443115, 0.27927157282829285, 0.0527336522936821, 0.20901410281658173, 0.2798389494419098, -0.48212286829948425, -0.06264190375804901, 0.2596437335014343, -0.044113755226135254, 0.42299884557724, 0.05807698145508766, 0.1961359977722168, -0.02601371519267559, -0.17014889419078827, 0.29014715552330017, 0.9980718493461609, -0.3438389301300049, -0.019431564956903458, -0.10914512723684311, -0.3598765730857849, 0.5201606154441833, 0.32512760162353516, 0.10994375497102737, -0.2907424569129944, -0.13588453829288483, -0.01521302480250597, 0.12174713611602783, 0.07488252967596054, 0.08309365808963776, -0.0717741847038269, -0.34809333086013794, 0.06251143664121628, -0.21467038989067078, -0.169268399477005, -0.34028947353363037, 0.02337685227394104, 0.053740937262773514, 0.3371373414993286, 0.21158944070339203, -0.14518490433692932, 0.19584694504737854, -0.25534290075302124, -0.14328454434871674, -0.3022669553756714, 0.14626282453536987, -0.06994934380054474, 0.013663792982697487, -0.24087925255298615, -0.03189757466316223, 0.3690549433231354, 0.11622259020805359, -0.12677867710590363, -0.4973998963832855, 0.30051323771476746, -0.0742245465517044, 0.06661347299814224, -0.032590899616479874, 0.6585506796836853, -0.15063244104385376, -0.08514773100614548, 0.1811935156583786, 0.281255841255188, -0.29423224925994873, 0.2012968808412552, 0.30363044142723083, 0.19760753214359283, 0.15173542499542236, 0.004687630571424961, 0.07329735904932022, 0.05224241316318512, 0.02479650266468525, -0.23360532522201538, -0.09858046472072601, 0.5100671052932739, -0.16134783625602722, 0.45229896903038025, -0.028002306818962097, 0.025157831609249115, -0.14601369202136993, -0.029682928696274757, -0.20386548340320587, 0.3801974952220917, 0.28941038250923157, -0.18660040199756622, 0.3278830349445343, -0.1623670905828476, -0.3643941581249237, 0.15504808723926544, 0.13120122253894806, 0.01262622606009245, 0.05463072657585144, -0.04349236190319061, -0.013247961178421974, -0.2791139781475067, -0.10775347799062729, 0.35112524032592773, 0.6806017160415649, 0.23463591933250427, 0.19186381995677948, 0.06999508291482925, -0.28555455803871155, -0.07115595042705536, 0.001360893016681075, 0.22168393433094025, -0.040608227252960205, 0.16736531257629395, 0.06225534528493881, 0.04244372621178627, 0.021964620798826218, -0.15282662212848663, 0.1556808054447174, -0.1852090060710907, -0.09366289526224136, 0.1074046939611435, -0.2216828167438507, -0.5938882231712341, -0.1774962842464447, 0.002236221916973591, -0.287817120552063, -0.30850082635879517, 0.10392731428146362, 0.05332008749246597, -0.14587941765785217, 0.16206082701683044, 0.2049497663974762, 0.5456447005271912, -0.42458274960517883, -0.2756490111351013, -0.041723012924194336, -0.31425362825393677, -0.11091046035289764, 0.10897098481655121, 0.45676758885383606, 0.33519771695137024, 0.4890904128551483, 0.31006768345832825, -0.08167870342731476, 0.04969377443194389, -0.40164682269096375, 0.0475735105574131, -0.32548999786376953, 0.26705148816108704, 0.1173536404967308, -0.2622724175453186, 0.1621399074792862, -0.3722046911716461, -0.011826642788946629, -0.08734847605228424, -0.08597201108932495, -0.021902628242969513, -0.01953175663948059, 0.05914955586194992, -0.210007444024086, -0.1561511754989624, -0.6618928909301758, -0.45373156666755676, 0.3190035820007324, -0.23834210634231567, 0.08222033828496933, 0.0593155100941658, -0.2954932451248169, 0.22237063944339752, -0.06498783826828003, 0.01636137254536152, -0.1773642897605896, -0.04884563013911247, 0.02931772544980049, -0.3534868657588959, -0.17807655036449432, -0.37236475944519043, -0.13820205628871918, 0.2802264988422394, -0.06061980128288269, -0.007123746909201145, -0.25036337971687317, 0.09986147284507751, -0.07900950312614441, 0.10570479929447174, 0.24656325578689575, 0.1785147786140442, -0.23915870487689972, 0.1344883292913437, -0.06015944480895996, -0.2647286653518677, 0.24683848023414612, 0.28179121017456055, 0.3843088448047638, 0.18339279294013977, 0.05993510037660599, 0.17080524563789368, 0.6879075169563293, 0.22980080544948578, 0.06567370891571045, 0.14727607369422913, 0.2549448311328888, 0.019310563802719116, 0.1698748916387558, -0.2620755732059479, 0.20821624994277954, -0.21381087601184845, 0.46087658405303955, 0.045081447809934616, 0.16560129821300507, -0.23663628101348877, -0.2956610918045044, 0.02418254315853119, -0.3138536512851715, -0.21501648426055908, 0.1974210888147354, -0.22645534574985504, 0.0024597179144620895, -0.07362707704305649, -0.3792168200016022, -0.2847122251987457, -0.03315737843513489, 0.004512267652899027, 0.3330302834510803, -0.09604185819625854, 0.11804552376270294, -0.3570633828639984, -0.031005915254354477, -0.4697401225566864, 0.14145803451538086, 0.008327007293701172, 0.25842148065567017, -0.050975169986486435, -0.07873965799808502, -0.08971891552209854, 0.07503261417150497, 0.6008836627006531, -0.4771213233470917, -0.12468115985393524, 0.15265114605426788, 0.11117684096097946, -0.13729427754878998, -0.12630073726177216, -0.29125574231147766, 0.0468968003988266, 0.4684965908527374, 0.37232542037963867, -0.38754743337631226, -0.26343581080436707, 0.22910930216312408, -0.21606242656707764, -0.04498942568898201, -0.190890371799469, -0.2358568161725998, -0.25600332021713257, -0.0991714671254158, 0.04424198344349861, 0.21777841448783875, 0.2320205122232437, -0.12347624450922012, 0.21100080013275146, -0.10897145420312881, 0.2612607479095459, 0.2202199399471283, 0.004205711185932159, -0.027149217203259468, 0.20108662545681, 0.17098389565944672, 0.03842540457844734, -0.13986201584339142, -0.37141889333724976, 0.13281750679016113, 0.004647309426218271, 0.07882429659366608, 0.24406203627586365, -0.00765709625557065, 0.6334801316261292, 0.07391273230314255, 0.29640623927116394, 0.2896258234977722, -0.14868690073490143, 0.01774541102349758, -0.25444164872169495, 0.19163651764392853, 0.24283257126808167, 0.2626425623893738, -0.14039893448352814, -0.25643882155418396, 0.590033233165741, 0.3069305717945099, -0.09948910772800446, 0.436851441860199, -0.49968448281288147, -0.35054877400398254, 0.2613230049610138, 0.02210117131471634, 0.6753667593002319, 0.0059888409450650215, 0.31732290983200073, 0.1614203006029129, -0.3751915991306305, 0.5796289443969727, -0.37775230407714844, 0.1610085666179657, -0.07037851214408875, -0.14921042323112488, 0.00480886735022068, -0.08049365878105164, -0.03711891919374466, 0.22904320061206818, -0.34711650013923645, 0.33152318000793457, 0.44647932052612305, -0.2043331265449524, -0.06859045475721359, 0.5156773328781128, -0.023617414757609367, -0.3742784857749939, -0.12388629466295242, 0.07793343812227249, -0.26072412729263306, 0.09877177327871323, -0.15557415783405304, -0.24807989597320557, 0.19329673051834106, 0.14641085267066956, -0.2879009544849396, 0.10499010235071182, 0.030853934586048126, 0.2383008897304535, 0.25692272186279297, 0.009897232986986637, 0.21698255836963654, -0.19437433779239655, -0.05444125831127167, -0.15986986458301544, -0.060079652816057205, 0.1227310299873352, -0.20815379917621613, 0.08337399363517761, 0.0815267488360405, -0.22250576317310333, 0.5000626444816589, -0.053083039820194244, -0.10866217315196991, -0.2400420904159546, 0.2608030438423157, -0.4327811896800995, -0.18399746716022491, 0.16753675043582916, 0.16328512132167816, 0.2320406287908554, -0.32190215587615967, 0.6011402010917664, 0.11713370680809021, -0.009043478406965733, 0.03524141013622284, 0.18379534780979156, -0.27487972378730774, 0.510238766670227, -0.3566158413887024, 0.0119086392223835, 0.08584187924861908, 0.012221011333167553, -0.14734864234924316, -0.3761586844921112, -0.032249245792627335, 0.2993479371070862, 0.08523877710103989, 0.03198697790503502, -0.010904082097113132, 0.10516070574522018, -0.03416304290294647, -0.04484129697084427, -0.09733857214450836, -0.3260810673236847, 0.12282872945070267, 0.18378713726997375, 0.17996643483638763, 0.17217352986335754, -0.08546765893697739, -0.27731460332870483, -0.1805696189403534, 0.4803910553455353, 0.023020168766379356, 0.20212797820568085, -0.15253108739852905, 0.3534679114818573, 0.41968733072280884, 0.5000770092010498, -0.2531452775001526, 0.055512405931949615, 0.07240799069404602, 0.07157938182353973, 0.42429590225219727, -0.2460528463125229, 0.0602252259850502, 0.24506016075611115, -0.12159862369298935, 0.04352053254842758, -0.3518277406692505, -0.14629200100898743, -0.2637702226638794, 0.10850553214550018, 0.10938441753387451, 0.0003896795096807182, 0.05136268213391304, -0.09670080244541168, -0.09724648296833038, -0.06615390628576279, -0.012981077656149864, -0.3419175148010254, -0.07169050723314285, 0.12081168591976166, -0.06952273845672607, 0.043189555406570435, 0.07352245599031448, 0.18630006909370422, 0.03954514488577843, -0.07555221766233444, 0.31191936135292053, 0.09127143770456314, 0.12354875355958939, -0.16061893105506897, -0.13676081597805023, 0.020297810435295105, 0.2227650284767151, 0.155967116355896, -0.04287665709853172, 0.11895954608917236, -0.20203465223312378, -0.1555565893650055, 0.2734251320362091, 0.22079375386238098, 0.02745634689927101, -0.13336579501628876, 0.13120363652706146, 0.11210894584655762, -0.11988219618797302, -0.025684421882033348, 0.23964141309261322, 0.0806819498538971, 0.10649408400058746, 0.12201157957315445, 0.11496564745903015, 0.46063432097435, -0.2895885407924652, 0.04949919879436493, 0.11247320473194122, -0.3702649176120758, 0.11843666434288025, -0.1373874843120575, -0.010500995442271233, 0.026158485561609268, 0.14512372016906738, -0.004540349822491407, 0.11540006846189499, -0.24399952590465546, 0.3238441050052643, 0.3000834286212921, -0.01018952764570713, 0.03903517872095108, 0.13944236934185028, 0.021483182907104492, -0.17022685706615448, 0.22123783826828003, -0.0121525339782238, 0.352103590965271, 0.03193247318267822, 0.842972993850708, -0.30105870962142944, -0.39672011137008667, 0.12651820480823517, 0.40005433559417725, 0.06075846776366234, -0.23886534571647644, 0.13217167556285858, 0.10595054924488068, -0.3579234480857849, 0.06070015951991081, 0.06416977941989899, -0.1542830765247345, 0.6550421118736267, -0.14000150561332703, 0.16470615565776825, -0.1756533682346344, -0.21198035776615143, 0.3319481611251831, 0.20278340578079224, -0.07796867191791534, -0.011605493724346161, 0.12744788825511932, -0.0609828419983387, -0.19573868811130524, 0.4920666217803955, 0.2548990845680237, 0.05156726762652397, -0.4656883478164673, 0.0072833518497645855, 0.09706339985132217, 0.13736294209957123, -0.09356130659580231, 0.15308305621147156, 0.4537774920463562, -0.1514500081539154, 0.20709162950515747, 0.004273141268640757, -0.05843536555767059, -0.09915964305400848, 0.0827237069606781, -0.3991205096244812, -0.40854737162590027, 0.5045662522315979, 0.060877036303281784, -0.11606744676828384, 0.08093591779470444, 0.33707666397094727, -0.6575289368629456, -0.26047512888908386, 0.3494884669780731, -0.08489353954792023, -0.03731124848127365, -0.22380788624286652, 0.07407914102077484, 0.07273787260055542, 0.6086329221725464, 0.1636587232351303, 0.3771527111530304, -0.3970237076282501, -0.10719940811395645, -0.6464778184890747, -0.213205948472023, 0.09976083040237427, -0.16706182062625885, -0.2602814733982086, 0.13774143159389496, -0.038988128304481506, 0.2751610577106476, -0.20754221081733704, 0.05163250491023064, 0.253215491771698, 0.02692127786576748, -0.4563053250312805, 0.0790511965751648, 0.13707683980464935, 0.02010967768728733, 0.283748984336853, -0.07789776474237442, -0.221319317817688, 0.06585615128278732, -0.056703973561525345, -0.05049363896250725, -0.08613935112953186, 0.2779480516910553, 0.17560884356498718, 0.03599395230412483, 0.04806514084339142, 0.5123880505561829, -0.02241557091474533, -0.17422179877758026, -0.026028305292129517, 0.028573568910360336, -0.37258651852607727, -0.11181635409593582, -0.12267294526100159, 0.21131880581378937, -0.37687674164772034, 0.270047664642334, 0.016957273706793785, -0.10174857825040817, 0.12268909066915512, -0.0095326853916049, -0.17501020431518555, -0.02108294516801834, 0.04411758854985237, 0.22659963369369507, 0.21104510128498077, 0.7624980211257935, -0.0001504886313341558, 0.1709391176700592, -0.07620475441217422, -0.06651191413402557, 0.07553735375404358, -0.11749504506587982, -0.26458311080932617, -0.15456542372703552, 0.06485731899738312, 0.04752332717180252, 0.1106644868850708, -0.7056499123573303, -0.34685927629470825, 0.016169235110282898, -0.019177056849002838, 0.05395838990807533, 0.7361040115356445, 0.07266198098659515, -0.14398157596588135, -0.131094291806221, -0.323185533285141, 0.020540250465273857, -0.23995505273342133, 0.18913593888282776, -0.36765506863594055 ]
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).
[ -0.030960535630583763, -0.41867437958717346, -0.03965788707137108, -0.046870362013578415, -0.22470258176326752, 0.0937633141875267, 0.16783763468265533, 0.03758598864078522, 0.27953657507896423, -0.1670985370874405, -0.22088010609149933, 0.33422231674194336, -0.21486340463161469, 0.3291552662849426, 0.41964229941368103, -0.4405977725982666, 0.024054480716586113, -0.21915344893932343, -0.2579807937145233, 0.37987807393074036, 0.06556902825832367, 0.10199673473834991, -0.16362528502941132, 0.05979224294424057, -0.2679941952228546, -0.1647077351808548, -0.36144745349884033, 0.12241499871015549, 0.13698257505893707, -0.010515522211790085, 0.2192515730857849, 0.3853779435157776, -0.0239703506231308, 0.11684758961200714, -0.00011339464981574565, -0.07660692185163498, 0.15948884189128876, -0.12002439796924591, 0.0910901129245758, 0.006583901587873697, 0.1546204686164856, -0.42770907282829285, 0.07293397188186646, -0.15843740105628967, 0.14493303000926971, 0.1565854847431183, 0.14506763219833374, -0.006900622509419918, 0.4177412986755371, -0.25564852356910706, 0.09768414497375488, 0.3037808835506439, -0.16508667171001434, 0.2707972526550293, -0.1420087367296219, -0.06699220836162567, 0.027930326759815216, 0.2208583652973175, 0.6148959398269653, -0.5899092555046082, 0.004168082028627396, 0.07856695353984833, -0.014349767938256264, 0.007405638229101896, 0.155482679605484, -0.02988366410136223, -0.33571839332580566, -0.2630695104598999, -0.3932986557483673, 0.21611784398555756, -0.1300581395626068, -0.17696857452392578, -0.2505919337272644, -0.5384415984153748, 0.017066923901438713, 0.03541625291109085, -0.2903543710708618, 0.031573448330163956, -0.2046491950750351, -0.054591212421655655, -0.12335369735956192, -0.10292111337184906, -0.006522923707962036, 0.15645062923431396, 0.3066307604312897, 0.5719147324562073, 0.1277766078710556, 0.17128871381282806, 0.35259145498275757, 0.023896045982837677, -0.14905200898647308, -0.287600040435791, 0.2940402925014496, 0.16211552917957306, -0.4089534878730774, -0.3507792353630066, 0.1309260129928589, -0.2493543028831482, 0.28444722294807434, 0.08810096234083176, 0.11836832016706467, 0.05591220036149025, -0.13708586990833282, 0.2888893783092499, 0.30111178755760193, 0.035575393587350845, -0.07668817788362503, -0.12673501670360565, 0.021894441917538643, -0.13308849930763245, 0.18895916640758514, 0.12626677751541138, 0.1817830502986908, -0.008661767467856407, -0.37493789196014404, -0.13579267263412476, -0.1604587733745575, 0.17387224733829498, -0.16110000014305115, -0.5166130065917969, -0.11007850617170334, -0.09324755519628525, 0.07216586917638779, -0.092467300593853, -0.29434117674827576, 0.07609979808330536, -0.04885454103350639, 0.22475813329219818, -0.08821700513362885, 0.05215867981314659, -0.06363728642463684, 0.08743636310100555, -0.42712727189064026, -0.031100375577807426, 0.3111230731010437, 0.09531347453594208, 0.07913783192634583, 0.16832081973552704, -0.009065116755664349, 0.1078636422753334, 0.42135122418403625, 0.020736493170261383, -0.005292032845318317, -0.02572023496031761, 0.2052459865808487, -0.2978519797325134, 0.013126312755048275, 0.13944882154464722, -0.3443312346935272, -0.12336929142475128, -0.14080345630645752, -0.16177324950695038, 0.10516698658466339, 0.02616794779896736, -0.22706963121891022, -0.2634855806827545, -0.47871336340904236, 0.709678590297699, -0.016317375004291534, -0.04073476791381836, -0.18003684282302856, 0.28762444853782654, -0.2646004259586334, -0.11316069215536118, 0.22921276092529297, 0.10030777752399445, -0.04390285164117813, -0.2998162806034088, -0.07132042199373245, -0.0835307389497757, -0.08975832164287567, 0.2788398265838623, -0.1998571753501892, 0.22274678945541382, 0.06801255792379379, 0.1548837274312973, 0.30274030566215515, -0.46299365162849426, -0.036578938364982605, 0.26916882395744324, -0.03577674552798271, 0.38466599583625793, 0.06560057401657104, 0.22492589056491852, 0.019382279366254807, -0.20013093948364258, 0.2639881372451782, 1.042127013206482, -0.4086362421512604, 0.0009426839533261955, -0.09630411863327026, -0.35686028003692627, 0.5225275158882141, 0.3448227047920227, 0.05289361998438835, -0.28935670852661133, -0.14739537239074707, -0.0054203225299716, 0.07239212095737457, 0.058151740580797195, 0.09516765922307968, -0.03698749095201492, -0.3416343033313751, 0.06954710930585861, -0.22003591060638428, -0.1500299870967865, -0.44301116466522217, 0.019452305510640144, 0.04457835108041763, 0.34341761469841003, 0.23521655797958374, -0.1388448029756546, 0.22767780721187592, -0.29061415791511536, -0.13999676704406738, -0.3124927580356598, 0.13172070682048798, -0.050935517996549606, 0.031070679426193237, -0.22075983881950378, -0.03535984084010124, 0.36213162541389465, 0.14933735132217407, -0.12701575458049774, -0.48026588559150696, 0.32783156633377075, -0.07594720274209976, 0.03664091229438782, -0.027286309748888016, 0.6227849721908569, -0.15069268643856049, -0.07301945984363556, 0.19060412049293518, 0.266891747713089, -0.28597140312194824, 0.21143212914466858, 0.36122795939445496, 0.2709766626358032, 0.17123514413833618, -0.03409281373023987, 0.09529756009578705, 0.023809364065527916, -0.005174318328499794, -0.2202402651309967, -0.08570580184459686, 0.5233914852142334, -0.17810910940170288, 0.4187559187412262, -0.045078497380018234, 0.036600202322006226, -0.13553042709827423, -0.018147611990571022, -0.19881539046764374, 0.41928941011428833, 0.27263545989990234, -0.17454440891742706, 0.3594205975532532, -0.1392039954662323, -0.34282904863357544, 0.16008315980434418, 0.15589235723018646, 0.038506560027599335, 0.09078485518693924, -0.003108135424554348, -0.04850026220083237, -0.26149341464042664, -0.10891595482826233, 0.3499148190021515, 0.6467790603637695, 0.25869444012641907, 0.13757582008838654, 0.013220415450632572, -0.25058895349502563, -0.06948146969079971, 0.03184835985302925, 0.1693359613418579, -0.028486303985118866, 0.14386524260044098, 0.06159089133143425, 0.04908038303256035, 0.02933247573673725, -0.17154929041862488, 0.18374301493167877, -0.16221767663955688, -0.07556942850351334, 0.07238426804542542, -0.1664850413799286, -0.5669123530387878, -0.22831566631793976, 0.02773306332528591, -0.38150718808174133, -0.2889532148838043, 0.10477866232395172, 0.051413118839263916, -0.12023962289094925, 0.18662790954113007, 0.2017742097377777, 0.5722042918205261, -0.3877086341381073, -0.2930241525173187, -0.02914554439485073, -0.3389904797077179, -0.11456481367349625, 0.11243943870067596, 0.45572441816329956, 0.26035943627357483, 0.48422324657440186, 0.32592520117759705, -0.03397271782159805, 0.04637593403458595, -0.39842474460601807, 0.017644455656409264, -0.3012900650501251, 0.26761648058891296, 0.0737738236784935, -0.18586866557598114, 0.1533937305212021, -0.40689215064048767, -0.0056120408698916435, -0.021065853536128998, -0.07087575644254684, -0.0528932549059391, 0.0044157602824270725, 0.09541946649551392, -0.1834779679775238, -0.19025984406471252, -0.6412385702133179, -0.4301629364490509, 0.26739317178726196, -0.2601638436317444, 0.08659783750772476, 0.01404163334518671, -0.2555755376815796, 0.17967838048934937, -0.10786699503660202, -0.002607900183647871, -0.16347718238830566, -0.03895358741283417, 0.046613115817308426, -0.3563413918018341, -0.17569804191589355, -0.36977866291999817, -0.1142725944519043, 0.24605385959148407, -0.06947685778141022, -0.0248477254062891, -0.2876039743423462, 0.09452743083238602, -0.11205733567476273, 0.05305669084191322, 0.20907104015350342, 0.19394831359386444, -0.22071023285388947, 0.13459794223308563, -0.07369869202375412, -0.29403576254844666, 0.23027537763118744, 0.2727072238922119, 0.38758084177970886, 0.15295566618442535, 0.07122116535902023, 0.17559170722961426, 0.7356224656105042, 0.2542088031768799, 0.01822599023580551, 0.15264931321144104, 0.2519245743751526, -0.008517536334693432, 0.17827638983726501, -0.24685527384281158, 0.17666272819042206, -0.15473131835460663, 0.4875797629356384, 0.09875667095184326, 0.15080209076404572, -0.24501655995845795, -0.30624011158943176, 0.058926112949848175, -0.2685535252094269, -0.21715977787971497, 0.2424306422472, -0.22875800728797913, 0.014071307145059109, -0.08966036885976791, -0.4162387549877167, -0.29063764214515686, -0.053883686661720276, -0.02230987325310707, 0.3469155430793762, -0.07768463343381882, 0.15731759369373322, -0.34133991599082947, -0.04604516178369522, -0.4354322552680969, 0.14782826602458954, 0.07338906079530716, 0.28789252042770386, -0.06971289962530136, -0.024439597502350807, -0.11077502369880676, 0.10292783379554749, 0.615239143371582, -0.518580973148346, -0.12251307815313339, 0.10406070202589035, 0.08391129970550537, -0.13758157193660736, -0.10161548107862473, -0.26018208265304565, 0.014867396093904972, 0.4297337830066681, 0.33571797609329224, -0.3773830831050873, -0.2934282720088959, 0.17445775866508484, -0.20981262624263763, -0.08447665721178055, -0.16246342658996582, -0.25328436493873596, -0.22844794392585754, -0.12505652010440826, 0.04177669808268547, 0.229720339179039, 0.22009968757629395, -0.06440141797065735, 0.16534996032714844, -0.10738150030374527, 0.201708123087883, 0.21851898729801178, 0.023033006116747856, -0.040012605488300323, 0.25569409132003784, 0.19773410260677338, 0.0395459309220314, -0.1268981248140335, -0.3682376742362976, 0.06594010442495346, -0.0586392879486084, 0.06726764887571335, 0.25432851910591125, -0.03158373013138771, 0.6494701504707336, 0.042327750474214554, 0.33929696679115295, 0.2956755459308624, -0.17103946208953857, 0.0038371910341084003, -0.22791872918605804, 0.1694696694612503, 0.22283950448036194, 0.22637008130550385, -0.15907864272594452, -0.2618798017501831, 0.5629382133483887, 0.2996862530708313, -0.07139333337545395, 0.4516775906085968, -0.489803671836853, -0.330451101064682, 0.2624640166759491, 0.07081340998411179, 0.6956494450569153, -0.06268087029457092, 0.2947652041912079, 0.07930438965559006, -0.3640850782394409, 0.6168702840805054, -0.3644067645072937, 0.15062376856803894, -0.1062219962477684, -0.19690917432308197, 0.012675678357481956, -0.09358713775873184, -0.07310056686401367, 0.2685149610042572, -0.38254690170288086, 0.3077969551086426, 0.4287252724170685, -0.20226317644119263, -0.025360634550452232, 0.5634381175041199, 0.05623327195644379, -0.3985058665275574, -0.158392995595932, 0.068796306848526, -0.30815497040748596, 0.05436553806066513, -0.1665421426296234, -0.2580890655517578, 0.2257060408592224, 0.1558324247598648, -0.3013504147529602, 0.06678995490074158, -0.03523050248622894, 0.21778292953968048, 0.21286410093307495, 0.015265408903360367, 0.2679018974304199, -0.11605888605117798, -0.03785024583339691, -0.1535063236951828, -0.08397116512060165, 0.13552790880203247, -0.1933331936597824, 0.03781479597091675, 0.018267739564180374, -0.18674054741859436, 0.49888867139816284, -0.06247984617948532, -0.11852221190929413, -0.21425196528434753, 0.28020742535591125, -0.4461521506309509, -0.19157017767429352, 0.14757519960403442, 0.1756839156150818, 0.19827403128147125, -0.30494898557662964, 0.5695933103561401, 0.10945103317499161, 0.011156485415995121, 0.03410443663597107, 0.210119366645813, -0.27115941047668457, 0.5002744793891907, -0.2811298966407776, 0.009963184595108032, 0.10484027117490768, 0.025052182376384735, -0.11265305429697037, -0.38574883341789246, -0.030786246061325073, 0.31705954670906067, 0.13859917223453522, 0.008726587519049644, 0.028529353439807892, 0.11237522214651108, -0.06501581519842148, -0.027069566771388054, -0.14490538835525513, -0.3513543903827667, 0.13187620043754578, 0.18006359040737152, 0.18757522106170654, 0.23299042880535126, -0.02897365763783455, -0.2715376615524292, -0.16603974997997284, 0.44810593128204346, 0.03526170551776886, 0.20998859405517578, -0.1294512301683426, 0.346599280834198, 0.4091411530971527, 0.5483716130256653, -0.25959473848342896, 0.0745583102107048, 0.055365968495607376, 0.11098016053438187, 0.4592452049255371, -0.25116145610809326, 0.07607986032962799, 0.2771492898464203, -0.12474248558282852, 0.03135427460074425, -0.3204190135002136, -0.15355819463729858, -0.2497408092021942, 0.11093162000179291, 0.10442154109477997, 0.004201697651296854, 0.004018060863018036, -0.10565076768398285, -0.11311241239309311, -0.08425319194793701, -0.0002986578329000622, -0.3816279470920563, -0.059590794146060944, 0.12420288473367691, -0.038065847009420395, 0.06506039202213287, 0.036879073828458786, 0.18779079616069794, 0.023435935378074646, -0.1025390550494194, 0.34939658641815186, 0.11723900586366653, 0.10813338309526443, -0.1778067648410797, -0.16322234272956848, 0.023343486711382866, 0.24289695918560028, 0.1519823521375656, -0.0641242116689682, 0.12868034839630127, -0.24157719314098358, -0.12066948413848877, 0.23631301522254944, 0.16919119656085968, 0.01744801551103592, -0.10271259397268295, 0.13327613472938538, 0.11498124152421951, -0.09520787000656128, -0.03510050103068352, 0.27450183033943176, 0.07827574014663696, 0.13403868675231934, 0.14364208281040192, 0.09875930845737457, 0.4728015959262848, -0.2388041913509369, 0.009598851203918457, 0.14305877685546875, -0.35201001167297363, 0.12352603673934937, -0.13037140667438507, -0.03644505888223648, -0.011556075885891914, 0.1032375767827034, -0.024382054805755615, 0.14119140803813934, -0.2700304687023163, 0.2807750701904297, 0.3052045404911041, 0.0004437522729858756, -0.033913396298885345, 0.2237345576286316, 0.05606955289840698, -0.14373500645160675, 0.246437206864357, 0.04788030683994293, 0.3264901340007782, 0.02148454450070858, 0.8898254632949829, -0.2593218684196472, -0.44537046551704407, 0.10914938151836395, 0.4276832640171051, 0.043728578835725784, -0.20927917957305908, 0.14286820590496063, 0.10250189900398254, -0.35283249616622925, 0.07324332743883133, 0.06681504845619202, -0.21987560391426086, 0.6501639485359192, -0.10448974370956421, 0.21174606680870056, -0.1332382708787918, -0.16710616648197174, 0.39403846859931946, 0.17511959373950958, -0.0650777667760849, 0.015708088874816895, 0.14367762207984924, -0.1233622282743454, -0.17897967994213104, 0.43864336609840393, 0.21286189556121826, 0.027809210121631622, -0.4619538187980652, 0.06281710416078568, 0.09056851267814636, 0.14693017303943634, -0.03955456614494324, 0.10659759491682053, 0.43498721718788147, -0.15175822377204895, 0.23860198259353638, 0.013740134425461292, -0.053497958928346634, -0.194731205701828, 0.07981104403734207, -0.4298861622810364, -0.3788125216960907, 0.47887882590293884, 0.06758502870798111, -0.11324051022529602, 0.059435490518808365, 0.33388441801071167, -0.6175369620323181, -0.2591392993927002, 0.3267904222011566, -0.0954337790608406, -0.0662723109126091, -0.19924667477607727, 0.07281958311796188, 0.06321084499359131, 0.6397792100906372, 0.17726364731788635, 0.36059924960136414, -0.3750837743282318, -0.12910091876983643, -0.6346372961997986, -0.156051367521286, 0.11392442882061005, -0.1761634349822998, -0.2531370222568512, 0.17074880003929138, -0.04700508713722229, 0.32274842262268066, -0.2095467746257782, 0.09758211672306061, 0.2715470790863037, 0.02710817940533161, -0.4724886417388916, 0.04005105420947075, 0.09500543773174286, 0.04796021804213524, 0.2817944884300232, -0.07571340352296829, -0.19103217124938965, 0.06382116675376892, -0.06736116856336594, -0.08385756611824036, -0.11887865513563156, 0.24940581619739532, 0.17716878652572632, 0.06206151843070984, 0.03445276990532875, 0.5532029867172241, 0.010818389244377613, -0.1616266667842865, 0.031095534563064575, 0.0285956934094429, -0.377071738243103, -0.13372254371643066, -0.08216700702905655, 0.20839731395244598, -0.39294588565826416, 0.252955824136734, -0.007655260153114796, -0.02523563615977764, 0.1193835586309433, 0.011496702209115028, -0.2017444521188736, -0.024046877399086952, 0.08408021181821823, 0.24017983675003052, 0.2028205394744873, 0.743805468082428, -0.014433643780648708, 0.17315654456615448, -0.08187689632177353, -0.037452757358551025, 0.06602861732244492, -0.13763947784900665, -0.24784846603870392, -0.13568443059921265, 0.03555169329047203, 0.062083419412374496, 0.12089323252439499, -0.6768571138381958, -0.32936862111091614, 0.04510824754834175, -0.0008667914080433547, 0.004000001586973667, 0.74090576171875, 0.06504146754741669, -0.11601366102695465, -0.11704596132040024, -0.35686421394348145, 0.03341832756996155, -0.2515329420566559, 0.18173997104167938, -0.3713236451148987 ]
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`.
[ -0.08343654125928879, -0.3772912621498108, -0.054012227803468704, -0.08327734470367432, -0.21224689483642578, 0.016944464296102524, 0.13349920511245728, 0.06816377490758896, 0.3124588131904602, -0.02526000328361988, -0.23778598010540009, 0.35998135805130005, -0.1587040275335312, 0.34977900981903076, 0.3823589086532593, -0.4798983335494995, -0.030744347721338272, -0.13304609060287476, -0.3688013553619385, 0.38237297534942627, 0.040902189910411835, 0.09933459758758545, -0.12265491485595703, 0.007462496403604746, -0.23627233505249023, -0.1522180140018463, -0.343296617269516, 0.09290041774511337, 0.1845618486404419, -0.055031392723321915, 0.19634634256362915, 0.449955016374588, -0.017280209809541702, 0.0523214191198349, -0.00011072369670728222, -0.0873187929391861, 0.1317596286535263, -0.120020292699337, 0.049835316836833954, 0.0378308966755867, 0.21656617522239685, -0.4117274284362793, -0.06971539556980133, -0.18803778290748596, 0.1087673231959343, 0.21512386202812195, 0.1830584704875946, 0.0244120005518198, 0.41552141308784485, -0.321086049079895, 0.12937010824680328, 0.326847642660141, -0.22198709845542908, 0.23453138768672943, -0.09464359283447266, -0.1533048152923584, 0.09412947297096252, 0.3127891719341278, 0.49168404936790466, -0.5477128624916077, -0.06117920204997063, 0.14151141047477722, 0.0077215698547661304, 0.05961751565337181, 0.1025722399353981, -0.051516611129045486, -0.32518428564071655, -0.2065102756023407, -0.35482069849967957, 0.294082373380661, -0.18114687502384186, -0.14917246997356415, -0.19877667725086212, -0.5856706500053406, 0.06005866453051567, -0.010826733894646168, -0.29971036314964294, 0.020461412146687508, -0.14967505633831024, -0.012792973779141903, -0.1295461356639862, -0.09831440448760986, 0.07478516548871994, 0.09907766431570053, 0.32749244570732117, 0.5603528022766113, 0.09386061877012253, 0.18363536894321442, 0.39189472794532776, 0.07738848775625229, -0.21224841475486755, -0.24602682888507843, 0.28512752056121826, 0.09630578011274338, -0.28267204761505127, -0.3462059199810028, 0.16300103068351746, -0.26397940516471863, 0.23295806348323822, 0.05784006044268608, 0.16775771975517273, 0.11276701092720032, -0.13788194954395294, 0.2494640201330185, 0.28214171528816223, 0.05985141545534134, 0.010651602409780025, -0.11803432554006577, 0.01498411688953638, -0.0573374480009079, 0.15855038166046143, 0.09276798367500305, 0.21617808938026428, -0.044970300048589706, -0.40501418709754944, -0.08261926472187042, -0.2754175066947937, 0.19549566507339478, -0.20601990818977356, -0.5297423601150513, -0.01862676441669464, -0.11480357497930527, 0.023905016481876373, -0.10028745234012604, -0.3187437355518341, 0.1601436734199524, -0.056538861244916916, 0.1746997982263565, -0.11744694411754608, 0.07287586480379105, -0.0856357142329216, 0.0885186493396759, -0.4986364543437958, 0.025194423273205757, 0.25986453890800476, 0.08004935830831528, 0.16220234334468842, 0.16258685290813446, -0.026162488386034966, 0.15069812536239624, 0.3910186290740967, 0.05960116162896156, 0.001835769391618669, -0.05813036859035492, 0.15901078283786774, -0.294969767332077, -0.04215236380696297, 0.1823384016752243, -0.424641877412796, -0.15480472147464752, -0.14691780507564545, -0.07604146748781204, 0.16089080274105072, 0.04554299637675285, -0.24469679594039917, -0.2981190085411072, -0.4465664029121399, 0.7699315547943115, -0.009562777355313301, -0.11291082203388214, -0.10955849289894104, 0.2079467922449112, -0.3065619170665741, -0.08008914440870285, 0.21946780383586884, 0.03209251910448074, -0.019059041514992714, -0.4033975601196289, -0.04226965457201004, -0.037179771810770035, -0.10395503789186478, 0.26446592807769775, -0.21608881652355194, 0.18137210607528687, 0.07405588030815125, 0.17394685745239258, 0.29999297857284546, -0.4679257273674011, -0.08489701896905899, 0.23441082239151, -0.036087516695261, 0.328967422246933, 0.08498451113700867, 0.12413081526756287, 0.05196549743413925, -0.17159172892570496, 0.17609842121601105, 0.9709119200706482, -0.35481807589530945, 0.024371763691306114, -0.10637897998094559, -0.3863515555858612, 0.5253021121025085, 0.29128971695899963, 0.0707944706082344, -0.2695535123348236, -0.1268913596868515, 0.004664760082960129, 0.09227943420410156, 0.018563248217105865, 0.08767064660787582, -0.0491231270134449, -0.3308200240135193, 0.00919767189770937, -0.2163791060447693, -0.15521517395973206, -0.3733970820903778, 0.053826265037059784, -0.061897192150354385, 0.38791772723197937, 0.3986712396144867, -0.11829815804958344, 0.2000267207622528, -0.35789164900779724, -0.03352377191185951, -0.2823464870452881, 0.14898425340652466, -0.013988743536174297, 0.048435408622026443, -0.22091205418109894, -0.04951193556189537, 0.23565387725830078, 0.1350269317626953, -0.15551771223545074, -0.40747809410095215, 0.30450934171676636, 0.020174415782094002, -0.0015047539491206408, -0.05420180782675743, 0.583307147026062, -0.16488303244113922, -0.10873425006866455, 0.20871888101100922, 0.30474478006362915, -0.3626171350479126, 0.2941078245639801, 0.31045374274253845, 0.19204077124595642, 0.15806695818901062, 0.019926100969314575, 0.15771086513996124, -0.014536937698721886, -0.06476866453886032, -0.16084513068199158, -0.04156317561864853, 0.6416842341423035, -0.04830149561166763, 0.3482072949409485, -0.029241634532809258, -0.03110128827393055, -0.17479853332042694, -0.16735848784446716, -0.25965815782546997, 0.3309374451637268, 0.2930155396461487, -0.27841469645500183, 0.3624514937400818, -0.12749841809272766, -0.40907424688339233, 0.15859641134738922, 0.19830994307994843, 0.05338192358613014, 0.12208299338817596, 0.011676467955112457, -0.0540282242000103, -0.30141013860702515, -0.10330083221197128, 0.3447790741920471, 0.6520267724990845, 0.28751131892204285, 0.16832181811332703, 0.021604420617222786, -0.2816524803638458, -0.09253242611885071, 0.002398239215835929, 0.1340305060148239, -0.029896754771471024, 0.18200230598449707, 0.10243816673755646, 0.03147490695118904, -0.08563186228275299, -0.06431347876787186, 0.16443897783756256, -0.2171371579170227, -0.032316964119672775, 0.03986583650112152, -0.17864565551280975, -0.5128586888313293, -0.29099464416503906, 0.035595815628767014, -0.30082470178604126, -0.3023705780506134, 0.18047155439853668, 0.09110571444034576, -0.2213294357061386, 0.21164461970329285, 0.23787370324134827, 0.6204199194908142, -0.39749443531036377, -0.12988033890724182, -0.10952050983905792, -0.346415251493454, -0.10274596512317657, 0.12991194427013397, 0.35552778840065, 0.27457255125045776, 0.47900816798210144, 0.2562705874443054, -0.05841083452105522, 0.10199447721242905, -0.4040747284889221, 0.026286330074071884, -0.3197815716266632, 0.25025174021720886, 0.11853539943695068, -0.1891646683216095, 0.15250933170318604, -0.3981904983520508, 0.011865864507853985, -0.1402013897895813, -0.029884299263358116, -0.05930731073021889, 0.011886288411915302, 0.14863142371177673, -0.20273487269878387, -0.2027556151151657, -0.6083054542541504, -0.42396360635757446, 0.24561406672000885, -0.2116805613040924, 0.04542778059840202, 0.05185730382800102, -0.21421317756175995, 0.1482132375240326, -0.16540071368217468, 0.015459360554814339, -0.18778468668460846, -0.10263346135616302, -0.006060393992811441, -0.3906424939632416, -0.10505235195159912, -0.5042768716812134, -0.12487222999334335, 0.2693502902984619, -0.10490911453962326, -0.10202059894800186, -0.3205792307853699, 0.057099584490060806, -0.09570690989494324, 0.016935022547841072, 0.2777614891529083, 0.10743659734725952, -0.22773589193820953, 0.03465279936790466, -0.06954736262559891, -0.21417143940925598, 0.21847182512283325, 0.4109444320201874, 0.38754042983055115, 0.1112000048160553, 0.003796773264184594, 0.14250242710113525, 0.7288059592247009, 0.20422545075416565, -0.006215881556272507, 0.24172459542751312, 0.22308897972106934, 0.05900600180029869, 0.11693597584962845, -0.14880706369876862, 0.21649539470672607, -0.13906526565551758, 0.5181333422660828, 0.13408218324184418, 0.144903764128685, -0.2363603115081787, -0.26169708371162415, 0.001547743915580213, -0.20625898241996765, -0.24861374497413635, 0.2672934830188751, -0.23298229277133942, -0.03897743299603462, -0.043080251663923264, -0.35451385378837585, -0.3271293640136719, -0.04302390292286873, 0.021407082676887512, 0.16934067010879517, -0.009690881706774235, 0.13491593301296234, -0.446654736995697, 0.007309668697416782, -0.4875246286392212, 0.17486968636512756, 0.03552919626235962, 0.2063131034374237, -0.014624481089413166, -0.09552441537380219, -0.1373225748538971, 0.10531484335660934, 0.4926847815513611, -0.511674702167511, -0.07899968326091766, 0.0813731774687767, 0.06042468547821045, -0.1340131014585495, -0.13115094602108002, -0.3227556645870209, -0.07703307271003723, 0.44155240058898926, 0.377454936504364, -0.42304712533950806, -0.3025461733341217, 0.22415246069431305, -0.1646210253238678, -0.0954803079366684, -0.17159678041934967, -0.2111213356256485, -0.1963445097208023, -0.15463075041770935, 0.07413265854120255, 0.14330698549747467, 0.26712173223495483, -0.1253158450126648, 0.19032762944698334, -0.10813796520233154, 0.18458867073059082, 0.17987218499183655, 0.01582762598991394, -0.04419182613492012, 0.35392290353775024, 0.11021851003170013, 0.07063795626163483, -0.13125525414943695, -0.3319583237171173, 0.10548827797174454, -0.13519999384880066, 0.205081969499588, 0.22697044909000397, 0.09067486226558685, 0.5530037879943848, 0.11436181515455246, 0.3576850891113281, 0.30341100692749023, -0.0634743794798851, 0.12543807923793793, -0.3138459622859955, 0.2028873711824417, 0.1758410632610321, 0.2550445795059204, -0.10684067010879517, -0.21424804627895355, 0.5952863097190857, 0.2874774932861328, -0.08879178762435913, 0.3823322355747223, -0.509010374546051, -0.3641587197780609, 0.28972434997558594, 0.08487677574157715, 0.656855583190918, -0.012666179798543453, 0.33726656436920166, 0.059178680181503296, -0.3040441870689392, 0.6199076771736145, -0.275953471660614, 0.14735539257526398, -0.13500449061393738, -0.2456529289484024, 0.036676570773124695, -0.04864168539643288, -0.08241485804319382, 0.26778170466423035, -0.35127565264701843, 0.3676716685295105, 0.4438670575618744, -0.19981546700000763, -0.12936092913150787, 0.5204609036445618, -0.08692397177219391, -0.32492905855178833, -0.11418379843235016, 0.09938494116067886, -0.2334737479686737, 0.116575226187706, -0.12016190588474274, -0.25054502487182617, 0.224079892039299, 0.13571174442768097, -0.22700902819633484, 0.04554487392306328, -0.06588099896907806, 0.13693292438983917, 0.2686856985092163, 0.03022916242480278, 0.19252634048461914, -0.22001151740550995, -0.025526138022542, -0.10796055942773819, -0.10625410079956055, 0.12955433130264282, -0.16364239156246185, 0.07055354118347168, 0.13920389115810394, -0.14759686589241028, 0.519082248210907, 0.012029817327857018, -0.11477285623550415, -0.19632446765899658, 0.2693732976913452, -0.3825451731681824, -0.1779465675354004, 0.024098174646496773, 0.09247919917106628, 0.10347146540880203, -0.3090541958808899, 0.5627152323722839, 0.12510062754154205, -0.02834976278245449, 0.06166974455118179, 0.10210751742124557, -0.22989551723003387, 0.5123823285102844, -0.18565791845321655, -0.03955167531967163, 0.054152488708496094, 0.09490847587585449, -0.1626531034708023, -0.3041134178638458, -0.040406327694654465, 0.2412853091955185, 0.13462132215499878, -0.01445312425494194, 0.01438663899898529, 0.16485249996185303, -0.17030933499336243, 0.04330522194504738, -0.07534753531217575, -0.3425027132034302, 0.15653543174266815, 0.22472837567329407, 0.21894192695617676, 0.2962503135204315, -0.11109838634729385, -0.1854034960269928, -0.12255635857582092, 0.4910307228565216, 0.09181759506464005, 0.2689321041107178, -0.11989658325910568, 0.32257208228111267, 0.35474762320518494, 0.5900278687477112, -0.3123570680618286, 0.07766614854335785, 0.06387634575366974, 0.13484272360801697, 0.5099005699157715, -0.2374129295349121, 0.008969716727733612, 0.33056220412254333, -0.11035566031932831, 0.07590001076459885, -0.3295682668685913, -0.17685581743717194, -0.2944237291812897, 0.08765855431556702, 0.07303455471992493, 0.012625141069293022, 0.02389492653310299, -0.034263644367456436, -0.045068491250276566, -0.16708727180957794, 0.005586816929280758, -0.4168510437011719, -0.10029634833335876, 0.20000974833965302, -0.06102343276143074, 0.06438522040843964, 0.1422734409570694, 0.2284623086452484, 0.009681571274995804, -0.15692287683486938, 0.2438974231481552, 0.16898371279239655, 0.10425101220607758, -0.1232447624206543, -0.1688486784696579, -0.0028999370988458395, 0.21621911227703094, 0.08830451220273972, -0.12145921587944031, 0.19059354066848755, -0.20523490011692047, -0.09061135351657867, 0.3178473711013794, 0.18736673891544342, 0.0489361435174942, -0.18051376938819885, 0.1527685672044754, 0.16645264625549316, -0.11254394054412842, 0.0003928378864657134, 0.15967319905757904, 0.11689894646406174, 0.1586347073316574, 0.14670738577842712, 0.14573141932487488, 0.3819451630115509, -0.19101984798908234, -0.04100756347179413, 0.08764418214559555, -0.3214346468448639, 0.11080725491046906, -0.060433126986026764, -0.020613647997379303, -0.032123979181051254, 0.18058431148529053, -0.0632687509059906, 0.20519575476646423, -0.4029553532600403, 0.2911440134048462, 0.338180273771286, -0.008856232278048992, 0.030189383774995804, 0.12258748710155487, 0.011403169482946396, -0.13385051488876343, 0.3777519166469574, 0.05748794600367546, 0.33230826258659363, 0.051880449056625366, 0.9268810749053955, -0.26295775175094604, -0.4481850862503052, 0.06358562409877777, 0.5313957929611206, 0.029138417914509773, -0.32142001390457153, 0.06138947978615761, 0.06478053331375122, -0.37453266978263855, 0.16029538214206696, 0.01161224115639925, -0.154006227850914, 0.664172887802124, -0.1534094661474228, 0.19332948327064514, -0.16817308962345123, -0.22600269317626953, 0.2835279107093811, 0.18811441957950592, -0.14975041151046753, 0.044861193746328354, 0.0780448392033577, -0.0641850158572197, -0.12437926232814789, 0.5293444395065308, 0.17358115315437317, -0.013322700746357441, -0.41422805190086365, -0.0032767748925834894, 0.14429625868797302, 0.16323970258235931, -0.1211547926068306, 0.12707531452178955, 0.4776848256587982, -0.1713278591632843, 0.29197296500205994, 0.05136556923389435, -0.1416742205619812, -0.1650524139404297, 0.05372526869177818, -0.4065130054950714, -0.41112086176872253, 0.4177230894565582, 0.1373796910047531, -0.041479453444480896, 0.0670824721455574, 0.285561203956604, -0.549807071685791, -0.22149793803691864, 0.2549920678138733, -0.17315123975276947, -0.07097944617271423, -0.22309289872646332, 0.09759368002414703, 0.044759031385183334, 0.6479759812355042, 0.2121247947216034, 0.25204381346702576, -0.3549044728279114, -0.0999402329325676, -0.5510978698730469, -0.13988251984119415, 0.17897436022758484, -0.10730110108852386, -0.2773844301700592, 0.16404388844966888, -0.03695598617196083, 0.278695672750473, -0.18398253619670868, 0.18127013742923737, 0.21143770217895508, 0.009155184030532837, -0.4085698127746582, 0.15160100162029266, 0.01877455599606037, 0.04160570353269577, 0.3125752806663513, -0.11525025218725204, -0.15051117539405823, 0.021374734118580818, -0.026635710150003433, -0.04524421691894531, -0.08165759593248367, 0.24790723621845245, 0.18512995541095734, 0.16567839682102203, 0.013480407185852528, 0.4840943217277527, -0.07819242775440216, -0.12652701139450073, -0.020787717774510384, -0.07461028546094894, -0.41930708289146423, -0.20675158500671387, -0.0996447429060936, 0.21812203526496887, -0.2795775234699249, 0.31740817427635193, -0.009172819554805756, -0.11152952164411545, 0.1520828753709793, 0.06763429194688797, -0.10436657816171646, -0.05300794169306755, 0.11525581032037735, 0.09159121662378311, 0.19210445880889893, 0.7487506866455078, 0.06614501029253006, 0.09840891510248184, -0.08251792192459106, -0.11334842443466187, 0.0889340192079544, -0.06935074180364609, -0.18787802755832672, -0.16191552579402924, 0.012331225909292698, 0.1088346540927887, 0.0035944688133895397, -0.6421046257019043, -0.30900105834007263, 0.028568970039486885, 0.04078551381826401, 0.036155816167593, 0.7093765735626221, 0.05208497866988182, -0.16649803519248962, -0.12391164898872375, -0.34871959686279297, 0.08365865796804428, -0.33575522899627686, 0.06608190387487411, -0.3758978247642517 ]
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.
[ -0.08342080563306808, -0.3881210386753082, -0.05889221280813217, -0.05794796347618103, -0.2545417249202728, 0.06940440088510513, 0.19493092596530914, 0.051409509032964706, 0.34056010842323303, -0.148264542222023, -0.20243649184703827, 0.3717307150363922, -0.23944693803787231, 0.34754082560539246, 0.3872973918914795, -0.45999622344970703, 0.02413763664662838, -0.2675703167915344, -0.2698787450790405, 0.3728725016117096, 0.06227116659283638, 0.08812961727380753, -0.1811579167842865, 0.08036638796329498, -0.2539784610271454, -0.16344736516475677, -0.3493884205818176, 0.10462110489606857, 0.15728838741779327, -0.013106824830174446, 0.2120376080274582, 0.3812630772590637, -0.0404963418841362, 0.13177944719791412, -0.00011317195458104834, -0.07457608729600906, 0.21052533388137817, -0.07623796164989471, 0.08656477928161621, -0.0652443915605545, 0.22962722182273865, -0.43176504969596863, 0.03671237826347351, -0.1433313935995102, 0.14059677720069885, 0.15172156691551208, 0.1642495095729828, -0.007431411650031805, 0.38448163866996765, -0.27362796664237976, 0.11178205162286758, 0.28741052746772766, -0.159559428691864, 0.2645362317562103, -0.10869436711072922, -0.04743966832756996, 0.04901299625635147, 0.2619546055793762, 0.6061145067214966, -0.6340845227241516, -0.01412485633045435, 0.10082333534955978, -0.04980574920773506, 0.03194373473525047, 0.12348435074090958, -0.08227285742759705, -0.38277167081832886, -0.23529942333698273, -0.39225122332572937, 0.24548806250095367, -0.1774669736623764, -0.19453614950180054, -0.22744812071323395, -0.5580010414123535, 0.006252025719732046, 0.041994377970695496, -0.3048388957977295, 0.05022207275032997, -0.19981548190116882, -0.0673157200217247, -0.13756804168224335, -0.08848807960748672, 0.011830097064375877, 0.1732160747051239, 0.3156490623950958, 0.5342772006988525, 0.10947060585021973, 0.1746341735124588, 0.3422694206237793, 0.042830947786569595, -0.13867340981960297, -0.28696703910827637, 0.3420206308364868, 0.12645919620990753, -0.37125059962272644, -0.30758628249168396, 0.1824488639831543, -0.261148601770401, 0.2541569471359253, 0.11465057730674744, 0.1163625717163086, 0.09271923452615738, -0.12476155906915665, 0.28519806265830994, 0.30204400420188904, 0.015004616230726242, -0.081357441842556, -0.0808272436261177, 0.02355457842350006, -0.13282261788845062, 0.22979402542114258, 0.12041144073009491, 0.16450059413909912, -0.030592767521739006, -0.4087803065776825, -0.06945730000734329, -0.16123728454113007, 0.197640523314476, -0.23830971121788025, -0.4846022427082062, -0.14205887913703918, -0.08315503597259521, 0.07923740148544312, -0.09192672371864319, -0.2979578971862793, 0.1001800149679184, -0.07304757088422775, 0.24686886370182037, -0.1000252440571785, 0.07183851301670074, -0.05634528771042824, 0.09835023432970047, -0.43943965435028076, -0.06991677731275558, 0.3233094811439514, 0.1422460377216339, 0.11645476520061493, 0.17511360347270966, -0.016917649656534195, 0.1301395148038864, 0.44203129410743713, -0.001940420945174992, 0.020957816392183304, -0.014058977365493774, 0.19523239135742188, -0.3251470923423767, 0.004884007386863232, 0.22755244374275208, -0.3726581931114197, -0.14289040863513947, -0.12809108197689056, -0.15609094500541687, 0.11269741505384445, 0.03565211594104767, -0.23599129915237427, -0.258800208568573, -0.4986611604690552, 0.6829744577407837, -0.02900308556854725, -0.04187587648630142, -0.15797121822834015, 0.25221335887908936, -0.24844473600387573, -0.12514981627464294, 0.21692104637622833, 0.0657944530248642, -0.0024428146425634623, -0.3119383156299591, -0.12970834970474243, -0.08177658170461655, -0.03710116818547249, 0.22530722618103027, -0.16007855534553528, 0.27307724952697754, 0.04683222994208336, 0.20835661888122559, 0.26415368914604187, -0.5000429749488831, -0.04386585205793381, 0.26692017912864685, -0.04217663034796715, 0.4183535873889923, 0.04843825474381447, 0.21280823647975922, 0.03480438515543938, -0.16171269118785858, 0.2985629141330719, 1.0328638553619385, -0.3860523998737335, 0.004557004664093256, -0.09491439163684845, -0.3651702404022217, 0.5233174562454224, 0.33920565247535706, 0.0907391831278801, -0.3195147216320038, -0.11596634984016418, -0.05300163850188255, 0.10058601945638657, 0.08589228242635727, 0.07345283031463623, -0.07229653000831604, -0.3650394082069397, 0.05978412553668022, -0.2487354576587677, -0.19382649660110474, -0.34214386343955994, 0.030358649790287018, 0.1119062528014183, 0.2797662615776062, 0.24866753816604614, -0.13200019299983978, 0.2229708433151245, -0.25494563579559326, -0.12981575727462769, -0.3390694260597229, 0.14184050261974335, -0.05979146063327789, 0.03747927397489548, -0.2510698437690735, 0.0015006177127361298, 0.3411407172679901, 0.1274174451828003, -0.15057171881198883, -0.47374773025512695, 0.28402578830718994, -0.055938731878995895, 0.039368581026792526, -0.022019030526280403, 0.6338947415351868, -0.16466929018497467, -0.08810757100582123, 0.19178521633148193, 0.28774401545524597, -0.26854732632637024, 0.21729809045791626, 0.3182615339756012, 0.16927018761634827, 0.14662021398544312, -0.022832021117210388, 0.09416674077510834, 0.06210275739431381, -0.0240916945040226, -0.2137686014175415, -0.08747951686382294, 0.5368931889533997, -0.15059474110603333, 0.4393784701824188, -0.06569959968328476, 0.06121241673827171, -0.1688978374004364, -0.017659086734056473, -0.1878749132156372, 0.41319045424461365, 0.30525219440460205, -0.1949423998594284, 0.3491773009300232, -0.18313749134540558, -0.3764328062534332, 0.15709860622882843, 0.0892985388636589, 0.029829982668161392, 0.09827033430337906, -0.04359465464949608, -0.05164363980293274, -0.2501387894153595, -0.06704875826835632, 0.34258052706718445, 0.6638851165771484, 0.262909859418869, 0.18624204397201538, 0.044392216950654984, -0.2409892976284027, -0.061128951609134674, 0.008607822470366955, 0.17471590638160706, -0.08582835644483566, 0.16375643014907837, 0.05545478314161301, 0.03326443210244179, 0.04951437935233116, -0.1321863830089569, 0.1540715992450714, -0.18271614611148834, -0.0768347829580307, 0.09348665922880173, -0.20606394112110138, -0.543978750705719, -0.19669069349765778, 0.022410798817873, -0.28677839040756226, -0.28189733624458313, 0.09111472219228745, 0.09463925659656525, -0.13861556351184845, 0.15519261360168457, 0.21347928047180176, 0.5341396331787109, -0.3972025513648987, -0.2968823313713074, -0.021574750542640686, -0.3439865708351135, -0.13003486394882202, 0.1255967915058136, 0.43872174620628357, 0.3332569897174835, 0.5198893547058105, 0.3033437430858612, -0.08945299685001373, 0.04793314263224602, -0.33727192878723145, 0.010944824665784836, -0.2858090400695801, 0.2852506935596466, 0.13329175114631653, -0.24219301342964172, 0.16811177134513855, -0.3485582172870636, -0.025232886895537376, -0.07369435578584671, -0.08946512639522552, -0.03885037451982498, -0.002678406424820423, 0.07000185549259186, -0.17970308661460876, -0.1755024641752243, -0.6369867920875549, -0.4389747679233551, 0.2991117537021637, -0.2640557885169983, 0.09133279323577881, -0.016800837591290474, -0.24855028092861176, 0.18541935086250305, -0.11016453802585602, 0.005768994335085154, -0.17779739201068878, -0.09301435947418213, 0.02673407457768917, -0.383955180644989, -0.182639017701149, -0.38308316469192505, -0.11380984634160995, 0.2173304557800293, -0.019879089668393135, -0.023588698357343674, -0.25342902541160583, 0.10554022341966629, -0.06678219139575958, 0.060140080749988556, 0.21710610389709473, 0.15147024393081665, -0.23825187981128693, 0.13472534716129303, -0.05888691917061806, -0.23463888466358185, 0.23617619276046753, 0.27068647742271423, 0.4174256920814514, 0.16836753487586975, 0.04018954932689667, 0.207238107919693, 0.6981123685836792, 0.2486543357372284, 0.02843017689883709, 0.14254838228225708, 0.23719801008701324, -0.0051187267526984215, 0.18883711099624634, -0.24631153047084808, 0.21866931021213531, -0.2162001132965088, 0.4830581843852997, 0.028078660368919373, 0.1475924849510193, -0.21976684033870697, -0.27514877915382385, 0.005366137716919184, -0.29938754439353943, -0.19954891502857208, 0.22123683989048004, -0.2730291485786438, 0.011601646430790424, -0.11817163228988647, -0.3471134305000305, -0.2714877426624298, -0.036767952144145966, 0.013553179800510406, 0.30513423681259155, -0.08174072206020355, 0.15701262652873993, -0.3513631224632263, -0.04900920391082764, -0.4496108889579773, 0.1521969437599182, 0.04184611886739731, 0.2807386815547943, -0.06174745783209801, -0.060967057943344116, -0.10397997498512268, 0.09561655670404434, 0.5847181081771851, -0.5166592597961426, -0.09553287923336029, 0.1287488043308258, 0.1254887580871582, -0.13120098412036896, -0.11914452165365219, -0.2858883738517761, 0.07362537831068039, 0.4597930312156677, 0.34093743562698364, -0.3444135785102844, -0.2527545392513275, 0.2195499688386917, -0.17277470231056213, -0.05759383738040924, -0.18763354420661926, -0.23618870973587036, -0.23544250428676605, -0.08206900954246521, 0.05536193028092384, 0.22892630100250244, 0.22814008593559265, -0.11155769228935242, 0.1948900818824768, -0.12732535600662231, 0.2405482828617096, 0.21754448115825653, -0.0025135488249361515, -0.024932803586125374, 0.18107397854328156, 0.1787150651216507, 0.044978030025959015, -0.14879080653190613, -0.30222949385643005, 0.09064803272485733, -0.05727596580982208, 0.11080890148878098, 0.22470276057720184, -0.020743446424603462, 0.6292215585708618, 0.04769947752356529, 0.3185666501522064, 0.277864933013916, -0.1944337785243988, -0.002981046447530389, -0.2579375207424164, 0.2018369734287262, 0.22035075724124908, 0.2621212303638458, -0.11640959233045578, -0.2639237344264984, 0.5500637292861938, 0.3211117088794708, -0.06593163311481476, 0.44205009937286377, -0.5858775973320007, -0.3123212158679962, 0.243300661444664, 0.06730468571186066, 0.6210483312606812, -0.011136769317090511, 0.28696224093437195, 0.08895474672317505, -0.38303837180137634, 0.5410963892936707, -0.3828495442867279, 0.16390204429626465, -0.08226892352104187, -0.15365390479564667, 0.015216918662190437, -0.09603598713874817, -0.04071798920631409, 0.252658873796463, -0.3295802175998688, 0.35041874647140503, 0.41206979751586914, -0.20682275295257568, -0.06753406673669815, 0.5238789916038513, -0.01396350096911192, -0.3526766300201416, -0.042019255459308624, 0.08234072476625443, -0.2836204767227173, 0.10621761530637741, -0.14440612494945526, -0.2439008355140686, 0.2170105278491974, 0.15858161449432373, -0.3457740545272827, 0.08260703831911087, -0.03591290861368179, 0.2272757887840271, 0.21372631192207336, -0.03624603524804115, 0.20517517626285553, -0.17723719775676727, -0.07493139058351517, -0.151271790266037, -0.03667148947715759, 0.12637098133563995, -0.14632144570350647, 0.04019339382648468, 0.08086957037448883, -0.21810753643512726, 0.45674657821655273, -0.046431493014097214, -0.12222272902727127, -0.19975242018699646, 0.28687265515327454, -0.4222885072231293, -0.18035933375358582, 0.15953998267650604, 0.17278656363487244, 0.2526027262210846, -0.3115990459918976, 0.5563430190086365, 0.14838626980781555, -0.010223702527582645, 0.03905484080314636, 0.18057692050933838, -0.2604200541973114, 0.507905125617981, -0.3668680489063263, 0.014650353230535984, 0.08976063132286072, 0.020912496373057365, -0.13653318583965302, -0.3640403747558594, -0.04824347421526909, 0.3149685561656952, 0.11398127675056458, 0.004376925062388182, 0.06601303815841675, 0.12075281143188477, 0.013668324798345566, -0.04037727043032646, -0.16214582324028015, -0.343129962682724, 0.1165359765291214, 0.16043223440647125, 0.17776018381118774, 0.18723304569721222, -0.04996248334646225, -0.2939015030860901, -0.16337950527668, 0.430541455745697, 0.04407862573862076, 0.20243118703365326, -0.13496190309524536, 0.35182154178619385, 0.38145264983177185, 0.49481630325317383, -0.25801005959510803, 0.0770564153790474, 0.09505412727594376, 0.09604030847549438, 0.44298040866851807, -0.2144613265991211, 0.05208286643028259, 0.2533467710018158, -0.11877475678920746, 0.029874805361032486, -0.32196882367134094, -0.1579003483057022, -0.2378179281949997, 0.09774888306856155, 0.07003478705883026, 0.0175649281591177, 0.041437216103076935, -0.10240130126476288, -0.08863116800785065, -0.07675343006849289, -0.011172679252922535, -0.3248330056667328, -0.06518680602312088, 0.1471288502216339, -0.10908201336860657, 0.04781101644039154, 0.07037895917892456, 0.19740547239780426, -0.0022583967074751854, -0.08733033388853073, 0.2813519239425659, 0.12359527498483658, 0.18367764353752136, -0.16559192538261414, -0.14358779788017273, -0.025252966210246086, 0.21271580457687378, 0.1461201310157776, -0.07139429450035095, 0.1613970696926117, -0.2196010947227478, -0.15503615140914917, 0.2535494863986969, 0.22577455639839172, 0.024593807756900787, -0.10969045758247375, 0.14707544445991516, 0.11975523829460144, -0.12432166934013367, -0.0021177141461521387, 0.20850633084774017, 0.09803565591573715, 0.1171838715672493, 0.10346046090126038, 0.1028519943356514, 0.46599671244621277, -0.3051428198814392, 0.031375400722026825, 0.10533084720373154, -0.34837907552719116, 0.11580237001180649, -0.08272487670183182, -0.002272938145324588, 0.002444850979372859, 0.1037568598985672, -0.002350101014599204, 0.1406654566526413, -0.23288774490356445, 0.3303033709526062, 0.3737263083457947, 0.06398878246545792, 0.007039976306259632, 0.17101411521434784, 0.0021042642183601856, -0.11631634831428528, 0.23376406729221344, 0.020215541124343872, 0.3528512120246887, 0.036387719213962555, 0.867721438407898, -0.3164049983024597, -0.44532155990600586, 0.10563308745622635, 0.39686673879623413, 0.056119028478860855, -0.23129111528396606, 0.1447221338748932, 0.10362928360700607, -0.3610670566558838, 0.08872585743665695, 0.06926027685403824, -0.17860956490039825, 0.6680487394332886, -0.12029635161161423, 0.16749346256256104, -0.18092672526836395, -0.22236600518226624, 0.3528715968132019, 0.1628759503364563, -0.10737733542919159, -0.03565680980682373, 0.10721438378095627, -0.10292459279298782, -0.1998308300971985, 0.4940882921218872, 0.2465532273054123, 0.04875333234667778, -0.513351321220398, 0.01197284460067749, 0.06946771591901779, 0.15743614733219147, -0.0794646143913269, 0.1021793931722641, 0.4472751021385193, -0.17720083892345428, 0.2422115057706833, 0.015447254292666912, -0.05776253715157509, -0.10079403221607208, 0.09602150321006775, -0.4450972080230713, -0.3861229419708252, 0.4605051577091217, 0.07121424376964569, -0.10437771677970886, 0.06537523120641708, 0.3186512291431427, -0.6325953602790833, -0.2643193304538727, 0.35300213098526, -0.08635837584733963, -0.05638343095779419, -0.20915165543556213, 0.08163756877183914, 0.04211239889264107, 0.6012604832649231, 0.1831810176372528, 0.3498595654964447, -0.3703470826148987, -0.11620751768350601, -0.6050468683242798, -0.19771045446395874, 0.051925212144851685, -0.1589607149362564, -0.2622523903846741, 0.1496829092502594, -0.06800590455532074, 0.32621118426322937, -0.23237261176109314, 0.09375456720590591, 0.2632236182689667, 0.10629775375127792, -0.4595515727996826, 0.056498222053050995, 0.10483268648386002, 0.03072602115571499, 0.30466949939727783, -0.10057753324508667, -0.22762712836265564, 0.04652253910899162, -0.055756133049726486, -0.0615883432328701, -0.14107660949230194, 0.29113465547561646, 0.18577246367931366, 0.08580721169710159, 0.025451652705669403, 0.5246702432632446, -0.003438395680859685, -0.14572010934352875, -0.025782369077205658, 0.054996151477098465, -0.37645646929740906, -0.12250717729330063, -0.0945223942399025, 0.16025792062282562, -0.3817416727542877, 0.2723006010055542, 0.01171000488102436, -0.04529961943626404, 0.11852068454027176, 0.0029891745653003454, -0.19630928337574005, -0.03083917684853077, 0.06354299932718277, 0.23153404891490936, 0.17898140847682953, 0.7314295768737793, -0.02548297308385372, 0.1557295024394989, -0.07545027881860733, -0.03531167656183243, 0.05345027893781662, -0.13339723646640778, -0.236647367477417, -0.20428404211997986, 0.06664784997701645, 0.05969933047890663, 0.14906315505504608, -0.6794449090957642, -0.33675822615623474, 0.03388068452477455, -0.002848314354196191, 0.060768354684114456, 0.7086074352264404, 0.07913322001695633, -0.12908051908016205, -0.11463309824466705, -0.3271324038505554, -0.0003095499996561557, -0.2582993507385254, 0.2029402256011963, -0.3788246810436249 ]
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.
[ -0.005270857829600573, -0.45644351840019226, -0.04090088605880737, -0.03315361961722374, -0.2421444207429886, 0.08998844027519226, 0.22626757621765137, 0.03379290550947189, 0.3455828130245209, -0.18324881792068481, -0.17462481558322906, 0.278237521648407, -0.21712945401668549, 0.31477978825569153, 0.39542341232299805, -0.4474583864212036, 0.05297606810927391, -0.2687872052192688, -0.24235081672668457, 0.37680450081825256, 0.023829219862818718, 0.13764406740665436, -0.16504845023155212, 0.10277912765741348, -0.34901726245880127, -0.14932645857334137, -0.29104575514793396, 0.10548999160528183, 0.1436440348625183, 0.049433790147304535, 0.24418771266937256, 0.34175875782966614, -0.03197263926267624, 0.1957813799381256, -0.00011775487655540928, -0.026304088532924652, 0.2295907437801361, -0.05339080095291138, 0.13802766799926758, -0.05143071711063385, 0.13814197480678558, -0.37766188383102417, 0.06772840023040771, -0.12424960732460022, 0.19982798397541046, 0.26353883743286133, 0.1815098226070404, 0.13065916299819946, 0.34446820616722107, -0.3399110734462738, 0.058766745030879974, 0.279428094625473, -0.20459184050559998, 0.3041357100009918, -0.08194995671510696, 0.04751146212220192, 0.0006846009055152535, 0.2406253069639206, 0.6316859126091003, -0.6122636795043945, 0.009547127410769463, 0.11814545840024948, 0.02063160017132759, -0.07559294998645782, 0.21316522359848022, -0.07582748681306839, -0.3081117868423462, -0.2063165009021759, -0.4016377925872803, 0.19827382266521454, -0.13543865084648132, -0.12603086233139038, -0.175089493393898, -0.5726633071899414, -0.09653264284133911, 0.07819848507642746, -0.28448036313056946, 0.015646815299987793, -0.19182898104190826, -0.04815204441547394, -0.13264083862304688, -0.12773172557353973, -0.03177593648433685, 0.13176469504833221, 0.35107341408729553, 0.6222615242004395, 0.14799049496650696, 0.15523608028888702, 0.3822128176689148, 0.10093443840742111, -0.08788488805294037, -0.24048122763633728, 0.3139003813266754, 0.11480094492435455, -0.3939477801322937, -0.3623136878013611, 0.20420047640800476, -0.2805655896663666, 0.22935117781162262, 0.06943336129188538, 0.16647140681743622, 0.08447276055812836, -0.12309883534908295, 0.261205792427063, 0.31291455030441284, 0.004651171155273914, -0.1349528431892395, -0.09768782556056976, 0.052231524139642715, -0.07296262681484222, 0.23544877767562866, 0.10745204985141754, 0.1639312505722046, -0.016648542135953903, -0.4331243634223938, -0.081639863550663, -0.13500434160232544, 0.1605202555656433, -0.12132038921117783, -0.5605202913284302, -0.03214409202337265, -0.14209027588367462, 0.1316175013780594, -0.0990327000617981, -0.30580538511276245, 0.1391567587852478, -0.08490342646837234, 0.268111914396286, -0.09449823200702667, 0.13290254771709442, -0.03960471972823143, 0.04850223660469055, -0.44950026273727417, -0.06490986794233322, 0.32811111211776733, 0.2135877013206482, 0.05287647619843483, 0.13712139427661896, -0.0014476589858531952, 0.07443194091320038, 0.35362884402275085, 0.05483119934797287, -0.010001715272665024, 0.04295634478330612, 0.07790771871805191, -0.3095370829105377, -0.02714591845870018, 0.21304713189601898, -0.3466987609863281, -0.135694682598114, -0.1648988276720047, -0.2356417030096054, 0.1343442052602768, -0.014896386303007603, -0.18980446457862854, -0.25104671716690063, -0.5263707637786865, 0.659267246723175, -0.030560890212655067, -0.05707307532429695, -0.12782686948776245, 0.32526859641075134, -0.24029715359210968, -0.07828091830015182, 0.1297537237405777, 0.04584129899740219, 0.0346592515707016, -0.3742160499095917, -0.09624477475881577, -0.08568520843982697, -0.040357138961553574, 0.301182359457016, -0.1886323243379593, 0.2752722203731537, 0.09550362825393677, 0.13118743896484375, 0.21981312334537506, -0.48723843693733215, -0.03061850182712078, 0.29354098439216614, -0.026290513575077057, 0.37811198830604553, 0.07428213953971863, 0.24824491143226624, -0.04977632686495781, -0.1843896359205246, 0.2528437376022339, 0.9806958436965942, -0.34456777572631836, 0.007813059724867344, -0.1259721964597702, -0.3848104774951935, 0.5938196778297424, 0.35658639669418335, 0.12478767335414886, -0.28234654664993286, -0.13511602580547333, 0.08289165049791336, 0.09046827256679535, 0.11593074351549149, 0.07905609905719757, -0.09205939620733261, -0.32606303691864014, 0.02391454018652439, -0.16235794126987457, -0.1164996474981308, -0.4359433650970459, 0.010673820041120052, -0.029216088354587555, 0.3597767949104309, 0.21428419649600983, -0.10928675532341003, 0.27148178219795227, -0.27914270758628845, -0.12076322734355927, -0.3025945723056793, 0.0706901103258133, -0.12522870302200317, -0.033662475645542145, -0.2842167913913727, -0.014253447763621807, 0.28782278299331665, 0.17413702607154846, -0.09134791791439056, -0.4625680446624756, 0.3106144368648529, -0.11350979655981064, 0.07947513461112976, -0.07837436348199844, 0.671086311340332, -0.1330435574054718, -0.05170908942818642, 0.18674395978450775, 0.25456011295318604, -0.30537259578704834, 0.2602538466453552, 0.3174130916595459, 0.30435237288475037, 0.196237713098526, -0.0506783090531826, 0.08027413487434387, 0.07100631296634674, -0.006333427503705025, -0.2542901933193207, -0.16883884370326996, 0.45542818307876587, -0.17983196675777435, 0.4642220437526703, -0.07356217503547668, -0.006369756069034338, -0.20803199708461761, -0.044365786015987396, -0.21304573118686676, 0.44610482454299927, 0.29147428274154663, -0.16969944536685944, 0.21452762186527252, -0.10940790176391602, -0.35217344760894775, 0.1489381194114685, 0.1390201896429062, 0.03116179071366787, 0.035136643797159195, 0.01258457638323307, 0.00668760621920228, -0.2945304214954376, -0.11442484706640244, 0.29003724455833435, 0.5758761167526245, 0.21885429322719574, 0.13711349666118622, 0.017956439405679703, -0.20882217586040497, -0.09033525735139847, 0.013693984597921371, 0.27527928352355957, 0.016922984272241592, 0.19390952587127686, 0.015642015263438225, 0.06455523520708084, 0.04245781525969505, -0.20732960104942322, 0.15761098265647888, -0.2026110291481018, -0.059820737689733505, 0.03191621974110603, -0.18620601296424866, -0.5582314133644104, -0.25548043847084045, -0.010924152098596096, -0.3671205937862396, -0.2729373276233673, 0.15407222509384155, 0.043434999883174896, -0.16207264363765717, 0.13760487735271454, 0.2474062293767929, 0.637561559677124, -0.46454840898513794, -0.23034009337425232, -0.140639066696167, -0.3725029230117798, -0.07144732028245926, 0.07383936643600464, 0.45979830622673035, 0.2861061990261078, 0.3831220865249634, 0.33520060777664185, -0.07496841996908188, 0.10937522351741791, -0.40889304876327515, 0.010390426963567734, -0.2810267210006714, 0.24833287298679352, 0.08054861426353455, -0.21447692811489105, 0.17429640889167786, -0.39816129207611084, -0.06309576332569122, -0.027325188741087914, -0.06673932820558548, -0.004362777341157198, -0.006471790838986635, 0.08041532337665558, -0.19725628197193146, -0.1686619073152542, -0.6268008351325989, -0.47573336958885193, 0.32473596930503845, -0.28774306178092957, 0.06258343160152435, 0.09646841883659363, -0.2704404890537262, 0.12282247096300125, -0.07755087316036224, -0.0060857380740344524, -0.12584716081619263, -0.042075395584106445, 0.027647577226161957, -0.3368200659751892, -0.14505471289157867, -0.37486058473587036, -0.09668299555778503, 0.22548221051692963, -0.07548369467258453, -0.014169171452522278, -0.1489139199256897, 0.10138888657093048, -0.162775456905365, 0.1472625583410263, 0.13594411313533783, 0.17427384853363037, -0.2007131576538086, 0.16263286769390106, -0.0187072791159153, -0.32956138253211975, 0.22921353578567505, 0.37306100130081177, 0.45728498697280884, 0.17798157036304474, 0.054322533309459686, 0.16411496698856354, 0.7202290296554565, 0.2120826691389084, 0.11296975612640381, 0.12070498615503311, 0.33972054719924927, -0.09869568794965744, 0.16566058993339539, -0.23940828442573547, 0.25505945086479187, -0.17682714760303497, 0.4767412841320038, 0.04971015825867653, 0.2018958479166031, -0.20101484656333923, -0.28190481662750244, -0.030092600733041763, -0.2980353534221649, -0.21822111308574677, 0.26773801445961, -0.20839957892894745, 0.04012434929609299, -0.0738258883357048, -0.4009011685848236, -0.28047263622283936, -0.06336963176727295, 0.05634463578462601, 0.4238351583480835, -0.0578906312584877, 0.17249934375286102, -0.33453822135925293, -0.06981935352087021, -0.4045552909374237, 0.0591115728020668, 0.02873890846967697, 0.23816978931427002, -0.09146812558174133, -0.06206570938229561, -0.0657404214143753, 0.16588737070560455, 0.6495795249938965, -0.47983983159065247, -0.1041218712925911, 0.08365973830223083, 0.08293932676315308, -0.043583307415246964, -0.14425043761730194, -0.2400396317243576, -0.045977991074323654, 0.43813788890838623, 0.40390878915786743, -0.39408445358276367, -0.2811403274536133, 0.19087883830070496, -0.2568361163139343, -0.1257813572883606, -0.23165152966976166, -0.21141894161701202, -0.17475788295269012, -0.14961431920528412, 0.05115102604031563, 0.2645762264728546, 0.21153363585472107, -0.10403916984796524, 0.13656333088874817, -0.01972055248916149, 0.23842531442642212, 0.21176309883594513, -0.014663736335933208, -0.0997283086180687, 0.24854591488838196, 0.12894082069396973, 0.03632763400673866, -0.1606728583574295, -0.3427600860595703, 0.07137361913919449, 0.041647572070360184, 0.13564063608646393, 0.2838731110095978, -0.008410525508224964, 0.6613307595252991, 0.07134005427360535, 0.38200536370277405, 0.295071005821228, -0.03617741912603378, -0.006973037961870432, -0.27927422523498535, 0.1381276696920395, 0.18902555108070374, 0.24390991032123566, -0.19070494174957275, -0.17326834797859192, 0.5538422465324402, 0.2548624277114868, -0.10853835195302963, 0.3889467716217041, -0.5794235467910767, -0.28470343351364136, 0.2820960581302643, 0.06647180020809174, 0.7568140625953674, -0.02275473065674305, 0.29468193650245667, 0.051071349531412125, -0.46020111441612244, 0.5372833013534546, -0.424101859331131, 0.15177728235721588, -0.04898108169436455, -0.11968103051185608, -0.01085193082690239, -0.10564862936735153, -0.10179813951253891, 0.30793988704681396, -0.33810189366340637, 0.30630987882614136, 0.4919751286506653, -0.23215807974338531, -0.0793590247631073, 0.5174074769020081, 0.09096969664096832, -0.3882366716861725, 0.008939210325479507, 0.00892338715493679, -0.25484147667884827, 0.16929422318935394, -0.19708047807216644, -0.2969512343406677, 0.19138285517692566, 0.12744630873203278, -0.29909518361091614, 0.0804286077618599, 0.08365577459335327, 0.26139864325523376, 0.3219776749610901, 0.0013862814521417022, 0.3056204617023468, -0.21544674038887024, -0.08355698734521866, -0.1875605285167694, -0.101717509329319, 0.16740193963050842, -0.22281651198863983, 0.07013490796089172, 0.04552357271313667, -0.181261345744133, 0.5122027397155762, -0.07783938944339752, -0.12757042050361633, -0.2751886546611786, 0.25064870715141296, -0.45339518785476685, -0.23373855650424957, 0.14939138293266296, 0.14604662358760834, 0.23558342456817627, -0.3322303295135498, 0.5851311683654785, 0.08234502375125885, -0.005585568491369486, 0.012265518307685852, 0.18602240085601807, -0.20392738282680511, 0.5055810809135437, -0.317449688911438, 0.04855097830295563, 0.08217552304267883, -0.04395449906587601, -0.11986508965492249, -0.40590375661849976, -0.09712430834770203, 0.3543567359447479, 0.09507563710212708, 0.025454986840486526, -0.05688733235001564, 0.13590478897094727, -0.12698401510715485, -0.059183355420827866, -0.13868795335292816, -0.39263221621513367, 0.10445525497198105, 0.21058306097984314, 0.2017633467912674, 0.1646537035703659, -0.07526357471942902, -0.25805091857910156, -0.21936972439289093, 0.5150887370109558, -0.026737695559859276, 0.20104719698429108, -0.16419155895709991, 0.375638484954834, 0.42429491877555847, 0.541885495185852, -0.2086775302886963, 0.0648793950676918, 0.061778824776411057, 0.08104384690523148, 0.37772050499916077, -0.2797585129737854, 0.04444720596075058, 0.22707787156105042, -0.18765461444854736, 0.006583500187844038, -0.27863770723342896, -0.10919922590255737, -0.27293941378593445, 0.13855049014091492, 0.1478899121284485, 0.035220667719841, -0.003477861639112234, -0.08391073346138, -0.04549406096339226, -0.018514081835746765, -0.061204127967357635, -0.3870375156402588, -0.02604975365102291, 0.18790316581726074, -0.04360800236463547, 0.055761609226465225, 0.017927607521414757, 0.24481485784053802, 0.04288804158568382, -0.06855230778455734, 0.36694201827049255, 0.09157974272966385, 0.15586112439632416, -0.14030112326145172, -0.2409391701221466, 0.020427944138646126, 0.23353593051433563, 0.13942907750606537, -0.00918225385248661, 0.19073867797851562, -0.23085373640060425, -0.1371113806962967, 0.16963449120521545, 0.12699565291404724, 0.07510603964328766, -0.10606863349676132, 0.09852853417396545, 0.07580411434173584, -0.11297927796840668, -0.02344462275505066, 0.21591246128082275, 0.09666841477155685, 0.043264858424663544, 0.11526874452829361, 0.14180324971675873, 0.49411511421203613, -0.2662070393562317, 0.08396894484758377, 0.108830526471138, -0.3563464283943176, 0.1345183551311493, -0.15543559193611145, -0.0202246755361557, 0.029280880466103554, 0.15330573916435242, -0.01826348528265953, 0.11684282124042511, -0.35832107067108154, 0.3812861442565918, 0.2211316078901291, -0.003973078913986683, 0.03103141486644745, 0.2010093778371811, 0.05039381980895996, -0.23988503217697144, 0.20653463900089264, 0.027465201914310455, 0.2873486876487732, 0.06632660329341888, 0.8376128673553467, -0.21268777549266815, -0.3841237425804138, 0.14744488894939423, 0.4088926911354065, 0.0464993491768837, -0.20879651606082916, 0.1675235480070114, 0.1362488567829132, -0.37386560440063477, 0.06661023944616318, 0.049195922911167145, -0.14546629786491394, 0.6858696937561035, -0.10418247431516647, 0.1188836619257927, -0.0746171623468399, -0.21313539147377014, 0.38438987731933594, 0.16913041472434998, -0.054651305079460144, -0.06694035977125168, 0.1378999650478363, -0.11123909056186676, -0.17631064355373383, 0.4905514419078827, 0.20801791548728943, 0.00905772764235735, -0.45847088098526, 0.02022913284599781, 0.12952636182308197, 0.13868200778961182, -0.05805095657706261, 0.17673492431640625, 0.47014424204826355, -0.07212401181459427, 0.1568211168050766, -0.0337037518620491, -0.010842585936188698, -0.12366799265146255, 0.04220794141292572, -0.41784200072288513, -0.3098796606063843, 0.5140645503997803, 0.09148647636175156, -0.17002198100090027, 0.09728062152862549, 0.3429413437843323, -0.6166443228721619, -0.25685298442840576, 0.299437552690506, -0.11552105098962784, -0.09028030931949615, -0.23393939435482025, 0.0465732142329216, 0.10411286354064941, 0.621881902217865, 0.14806897938251495, 0.39067041873931885, -0.3981970548629761, -0.10698498040437698, -0.5736942291259766, -0.23161475360393524, 0.13995260000228882, -0.05170753225684166, -0.1787451058626175, 0.09441894292831421, -0.06492915749549866, 0.2246154248714447, -0.2561451494693756, -0.0194271057844162, 0.2132716029882431, 0.09001773595809937, -0.40948227047920227, 0.16877323389053345, 0.18962787091732025, 0.06828110665082932, 0.30649787187576294, -0.03220447525382042, -0.15833806991577148, 0.20729203522205353, -0.13092073798179626, -0.13873687386512756, -0.08773798495531082, 0.24361652135849, 0.18980851769447327, 0.04684355482459068, 0.02114308625459671, 0.4844204783439636, -0.004028086084872484, -0.21207097172737122, 0.010847743600606918, -0.0026887801941484213, -0.32564735412597656, -0.1787254959344864, -0.07713396847248077, 0.2080409675836563, -0.3600689470767975, 0.2944645285606384, -0.009957676753401756, -0.1363239735364914, 0.023876627907156944, -0.008793107233941555, -0.21922771632671356, -0.02552374079823494, 0.033671118319034576, 0.2317085713148117, 0.24931323528289795, 0.6675165295600891, 0.016650762408971786, 0.09321209043264389, -0.08999080955982208, -0.050489552319049835, 0.026074852794408798, -0.015021594241261482, -0.2413497418165207, -0.1449449509382248, 0.033026762306690216, 0.014686644077301025, 0.18479909002780914, -0.7851415276527405, -0.3945764899253845, -0.021979307755827904, 0.03141121193766594, 0.0848197191953659, 0.7013662457466125, 0.12302414327859879, -0.17507024109363556, -0.09732476621866226, -0.34183695912361145, 0.02017504721879959, -0.22876346111297607, 0.17516574263572693, -0.36070138216018677 ]
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.
[ -0.06561239063739777, -0.37235602736473083, -0.06819300353527069, -0.04926763474941254, -0.2852720618247986, 0.13376595079898834, 0.21648041903972626, 0.12065185606479645, 0.30216771364212036, -0.14211788773536682, -0.16877050697803497, 0.32036516070365906, -0.16379883885383606, 0.34869423508644104, 0.4185851812362671, -0.41218045353889465, 0.03168852627277374, -0.21076588332653046, -0.2145611196756363, 0.424690306186676, 0.06257966160774231, 0.08338547497987747, -0.1656297743320465, 0.10172658413648605, -0.29676905274391174, -0.18775083124637604, -0.33034709095954895, 0.12289679795503616, 0.18027503788471222, -0.0067453994415700436, 0.2167428880929947, 0.33866459131240845, -0.005247809924185276, 0.07626168429851532, -0.00011275351425865665, -0.07976377755403519, 0.21804796159267426, -0.05337473005056381, 0.10312972962856293, -0.01396236103028059, 0.19306249916553497, -0.36390161514282227, 0.046674441546201706, -0.10680338740348816, 0.16875791549682617, 0.19356901943683624, 0.17223428189754486, -0.07640407234430313, 0.3412652611732483, -0.2545018494129181, 0.10234149545431137, 0.31566929817199707, -0.23991075158119202, 0.2976696789264679, -0.11138317734003067, -0.08681914210319519, 0.011975612491369247, 0.24389208853244781, 0.6718512177467346, -0.6273880004882812, -0.060072023421525955, 0.057241588830947876, -0.03238610178232193, 0.0497773252427578, 0.09267549961805344, -0.13015498220920563, -0.3021053969860077, -0.21089528501033783, -0.33772170543670654, 0.25109928846359253, -0.15973101556301117, -0.13706086575984955, -0.22889862954616547, -0.5552749037742615, -0.013059502467513084, 0.051291678100824356, -0.30808961391448975, 0.03183313086628914, -0.17259536683559418, -0.006247353740036488, -0.10114547610282898, -0.12405447661876678, 0.05755392089486122, 0.08909878134727478, 0.3357066214084625, 0.558486819267273, 0.1207938939332962, 0.1740795522928238, 0.36989516019821167, 0.07315333187580109, -0.08478505909442902, -0.2591191530227661, 0.27419862151145935, 0.10016487538814545, -0.41580915451049805, -0.3103591799736023, 0.13953205943107605, -0.23301182687282562, 0.30230602622032166, 0.09700164198875427, 0.1351519227027893, 0.11098425835371017, -0.12362212687730789, 0.27369511127471924, 0.340284526348114, 0.03425591066479683, -0.07769946753978729, -0.16922663152217865, 0.010786578059196472, -0.14165529608726501, 0.18027420341968536, 0.11355339735746384, 0.20747894048690796, 0.025803472846746445, -0.43926817178726196, -0.13691213726997375, -0.2284802496433258, 0.163488507270813, -0.25956201553344727, -0.5688908100128174, -0.11211659014225006, -0.09568822383880615, 0.062081966549158096, -0.13887587189674377, -0.26140210032463074, 0.1391804963350296, -0.09216353297233582, 0.26953399181365967, -0.06885502487421036, 0.0062883710488677025, -0.05109152942895889, 0.08796748518943787, -0.45554202795028687, -0.01012904942035675, 0.3323618471622467, 0.10064942389726639, 0.08620402216911316, 0.1457323133945465, -0.01978299580514431, 0.06223139539361, 0.3870139718055725, 0.011976806446909904, -0.00445981090888381, -0.012135080993175507, 0.18673940002918243, -0.31405529379844666, 0.029219094663858414, 0.18315310776233673, -0.3368547856807709, -0.08135084062814713, -0.14553304016590118, -0.14867335557937622, 0.14257395267486572, 0.03213292360305786, -0.2253686934709549, -0.2860385775566101, -0.5482082366943359, 0.6673356294631958, -0.011244331486523151, -0.06299509853124619, -0.16312554478645325, 0.3023530840873718, -0.30237793922424316, -0.09308991581201553, 0.17095287144184113, 0.01316557265818119, 0.021748431026935577, -0.27748405933380127, -0.16175632178783417, -0.0704328790307045, -0.10190112888813019, 0.3153512477874756, -0.19219160079956055, 0.308442085981369, 0.04201124235987663, 0.17138786613941193, 0.23467035591602325, -0.48465973138809204, -0.03291637450456619, 0.3313533067703247, -0.07133058458566666, 0.40189462900161743, 0.05024699866771698, 0.22615432739257812, 0.05564398318529129, -0.1782502830028534, 0.2328605055809021, 0.9699900150299072, -0.3880667984485626, 0.00952666811645031, -0.0859294906258583, -0.39205971360206604, 0.5433834791183472, 0.35699063539505005, 0.05959539860486984, -0.24156568944454193, -0.14711663126945496, -0.012278015725314617, 0.08712275326251984, 0.10982902348041534, 0.09951715171337128, -0.07516489177942276, -0.42072439193725586, 0.023374568670988083, -0.22361040115356445, -0.1986405849456787, -0.43184399604797363, 0.018178649246692657, 0.021138926967978477, 0.2296023666858673, 0.3293682634830475, -0.1370195895433426, 0.24782592058181763, -0.2794992923736572, -0.04655064642429352, -0.375577837228775, 0.12504078447818756, -0.051529619842767715, 0.00891315471380949, -0.20238706469535828, -0.048160869628190994, 0.35677456855773926, 0.17522495985031128, -0.12952227890491486, -0.391213059425354, 0.30711808800697327, -0.05294463038444519, 0.038124844431877136, -0.04055549204349518, 0.6141847968101501, -0.21523286402225494, -0.05715174227952957, 0.18918974697589874, 0.24514207243919373, -0.27274128794670105, 0.2884800434112549, 0.3536679148674011, 0.27539175748825073, 0.19267572462558746, -0.035916686058044434, 0.0541769415140152, 0.060281235724687576, -0.032950155436992645, -0.2539370059967041, -0.05172340199351311, 0.4938867688179016, -0.1278960257768631, 0.4670571982860565, -0.06947563588619232, 0.053518570959568024, -0.14168478548526764, -0.03650074079632759, -0.19595518708229065, 0.4411695897579193, 0.2900508642196655, -0.21228137612342834, 0.38620397448539734, -0.10933702439069748, -0.31891024112701416, 0.17448732256889343, 0.1380329728126526, 0.051245685666799545, 0.06468282639980316, -0.009072651155292988, -0.03657042235136032, -0.27271610498428345, -0.07160297781229019, 0.37076735496520996, 0.6577779650688171, 0.2386048138141632, 0.17744721472263336, 0.014817694202065468, -0.2293141782283783, -0.08080164343118668, -0.031206989660859108, 0.19398140907287598, 0.026519592851400375, 0.09888984262943268, 0.013474402017891407, 0.02231728844344616, 0.048615068197250366, -0.14007854461669922, 0.1509895771741867, -0.16983994841575623, -0.009785219095647335, 0.041374992579221725, -0.22050835192203522, -0.5566983222961426, -0.21707718074321747, 0.02917323261499405, -0.3283436894416809, -0.2532227635383606, 0.10271662473678589, 0.12225035578012466, -0.13139241933822632, 0.15123412013053894, 0.2321767508983612, 0.6160978674888611, -0.44601523876190186, -0.2688649296760559, -0.0835350826382637, -0.3008217513561249, -0.11323416978120804, 0.1290014386177063, 0.458760142326355, 0.3014472723007202, 0.38529685139656067, 0.33292949199676514, -0.009069636464118958, -0.021664781495928764, -0.3423929512500763, -0.038963329046964645, -0.2504136264324188, 0.2589898407459259, 0.03648275136947632, -0.24169915914535522, 0.14477291703224182, -0.3871200382709503, -0.07622846961021423, -0.07680205255746841, -0.07060393691062927, -0.08701089024543762, 0.03128528222441673, 0.1460472196340561, -0.18527090549468994, -0.15247833728790283, -0.5852788686752319, -0.4743439555168152, 0.2974983751773834, -0.29475656151771545, 0.05387389659881592, -0.10809840261936188, -0.3133610785007477, 0.19335028529167175, -0.12611325085163116, -0.019432177767157555, -0.13752523064613342, -0.09073217213153839, -0.055493876338005066, -0.3396884799003601, -0.15745007991790771, -0.3975778818130493, -0.08304990082979202, 0.2783389687538147, -0.04936175420880318, -0.04247850924730301, -0.15597306191921234, 0.13113829493522644, -0.15483495593070984, 0.060068752616643906, 0.09263867884874344, 0.17417506873607635, -0.23634226620197296, 0.12928369641304016, -0.028271034359931946, -0.2620972692966461, 0.2368292510509491, 0.2830851972103119, 0.3746199905872345, 0.17052830755710602, 0.022374175488948822, 0.2380325198173523, 0.7015316486358643, 0.23946109414100647, 0.054874029010534286, 0.1664234697818756, 0.2817079722881317, -0.023641420528292656, 0.11926253885030746, -0.2199317216873169, 0.15194280445575714, -0.1884620040655136, 0.47817325592041016, 0.10397263616323471, 0.15760919451713562, -0.2565964162349701, -0.2707090377807617, 0.03013981506228447, -0.2531300485134125, -0.17043852806091309, 0.21320022642612457, -0.3159544765949249, 0.033423587679862976, -0.0661194771528244, -0.39819642901420593, -0.22174987196922302, -0.018919290974736214, -0.014532205648720264, 0.28800392150878906, -0.04299427196383476, 0.16064612567424774, -0.3352033197879791, -0.031863562762737274, -0.4085686504840851, 0.11785247921943665, 0.09698819369077682, 0.3318116068840027, -0.0865965485572815, -0.045031994581222534, -0.08494319766759872, 0.11186127364635468, 0.5522235631942749, -0.46722403168678284, -0.0927446186542511, 0.1084989383816719, 0.06989071518182755, -0.09652107208967209, -0.08832478523254395, -0.22060008347034454, 0.045805808156728745, 0.42899081110954285, 0.36070704460144043, -0.2990398406982422, -0.2886457145214081, 0.15508940815925598, -0.2195773869752884, -0.04865579679608345, -0.1847698986530304, -0.21510973572731018, -0.17987044155597687, -0.1340266466140747, 0.0656886175274849, 0.23183831572532654, 0.21855084598064423, -0.08076169341802597, 0.15293246507644653, -0.11033450067043304, 0.15788625180721283, 0.18605409562587738, -0.043711256235837936, -0.09453719109296799, 0.15242980420589447, 0.2028803825378418, 0.013336152769625187, -0.1537550687789917, -0.307027667760849, 0.02104790508747101, 0.013778558932244778, 0.1322273313999176, 0.24650104343891144, -0.017967501655220985, 0.6386030912399292, -0.0434408113360405, 0.3078649938106537, 0.3434126079082489, -0.15040110051631927, -0.06010923907160759, -0.28928840160369873, 0.16420391201972961, 0.18684615194797516, 0.25749462842941284, -0.12326858937740326, -0.18717509508132935, 0.5725805759429932, 0.33587220311164856, -0.06958134472370148, 0.45397284626960754, -0.5722922086715698, -0.312608540058136, 0.24356558918952942, 0.05923960357904434, 0.5774693489074707, -0.083980031311512, 0.2616598308086395, -0.019969524815678596, -0.410173624753952, 0.5315449833869934, -0.37374407052993774, 0.15340806543827057, -0.05188096687197685, -0.2438555508852005, 0.02060128003358841, -0.0840628370642662, -0.007088164333254099, 0.3127877116203308, -0.3229180872440338, 0.324655145406723, 0.4892105758190155, -0.2282966673374176, -0.001217604847624898, 0.5655279755592346, 0.02586468495428562, -0.3701806962490082, 0.016943439841270447, 0.0685834288597107, -0.29029542207717896, 0.08192066848278046, -0.1532072126865387, -0.2107270061969757, 0.220350444316864, 0.17975354194641113, -0.35201990604400635, 0.05883757397532463, 0.02680566906929016, 0.21444052457809448, 0.20536558330059052, 0.035369012504816055, 0.2878016233444214, -0.1741962730884552, -0.06096969172358513, -0.1578620970249176, -0.08689051866531372, 0.12871673703193665, -0.19305962324142456, 0.03640551492571831, 0.09449710696935654, -0.20959553122520447, 0.4564691185951233, -0.06033067777752876, -0.14519846439361572, -0.22870053350925446, 0.27023714780807495, -0.509292483329773, -0.1495707482099533, 0.19817128777503967, 0.1473064422607422, 0.28635290265083313, -0.29994630813598633, 0.5590229630470276, 0.1000288650393486, -0.015114393085241318, 0.04153582453727722, 0.19969822466373444, -0.23633213341236115, 0.5029982924461365, -0.3028043508529663, -0.01523120142519474, 0.06073189154267311, 0.05679795891046524, -0.14584946632385254, -0.3981848657131195, -0.07032731920480728, 0.2884272038936615, 0.14057910442352295, -0.004566958174109459, 0.05541457608342171, 0.09924278408288956, 0.05065205693244934, 0.019722912460565567, -0.20058390498161316, -0.32429277896881104, 0.13359783589839935, 0.14717690646648407, 0.20381352305412292, 0.2274576723575592, -0.09523484110832214, -0.30265894532203674, -0.19763076305389404, 0.45944756269454956, -0.029067935422062874, 0.17826801538467407, -0.12853685021400452, 0.34765657782554626, 0.4074634611606598, 0.5269021987915039, -0.24591678380966187, 0.11764302104711533, 0.04560841619968414, 0.12695489823818207, 0.5051094889640808, -0.21541808545589447, -0.0042850561439991, 0.31251057982444763, -0.13643281161785126, 0.06286732852458954, -0.26634323596954346, -0.1617911159992218, -0.19228853285312653, 0.10798425227403641, 0.12295633554458618, 0.025694089010357857, 0.029705006629228592, -0.1108931452035904, -0.09539099037647247, -0.0409354493021965, -0.05580602213740349, -0.3989924490451813, -0.06124718487262726, 0.09844255447387695, -0.06748095154762268, 0.07659320533275604, 0.024723637849092484, 0.21679329872131348, -0.0350380502641201, -0.12325635552406311, 0.24483495950698853, 0.11760397255420685, 0.17357215285301208, -0.15246358513832092, -0.16086600720882416, -0.03669581189751625, 0.22497503459453583, 0.14430280029773712, -0.050422463566064835, 0.19050456583499908, -0.23372283577919006, -0.07541351020336151, 0.24634653329849243, 0.10478643327951431, 0.09123034030199051, -0.08847781270742416, 0.17272529006004333, 0.103700652718544, -0.08446688205003738, -0.020609810948371887, 0.23100978136062622, 0.05959812551736832, 0.11645418405532837, 0.13184750080108643, 0.19638776779174805, 0.46792513132095337, -0.1857157200574875, -0.003102899296209216, 0.11268313974142075, -0.3847610354423523, 0.12622162699699402, -0.18793362379074097, -0.07245909422636032, -0.028666682541370392, 0.13434651494026184, -0.01793181151151657, 0.13674642145633698, -0.26214122772216797, 0.3380008041858673, 0.38897019624710083, 0.07167445123195648, -0.0074673304334282875, 0.22169548273086548, 0.06501396000385284, -0.15316630899906158, 0.25695475935935974, 0.05426369979977608, 0.30022212862968445, 0.012245891615748405, 0.8360198140144348, -0.32402536273002625, -0.4433020055294037, 0.11316003650426865, 0.3656121492385864, 0.05183563381433487, -0.21638937294483185, 0.2377731204032898, 0.13523754477500916, -0.3363128900527954, 0.08761171251535416, 0.06379135698080063, -0.1809615045785904, 0.7053231596946716, -0.13340453803539276, 0.14903192222118378, -0.08333373069763184, -0.24597689509391785, 0.3666488230228424, 0.11514948308467865, -0.07119174301624298, -0.0004823503259103745, 0.05708678811788559, -0.1151062548160553, -0.17619872093200684, 0.5325777530670166, 0.180943101644516, 0.018203886225819588, -0.4425995349884033, 0.010071493685245514, 0.01492972020059824, 0.1756376475095749, -0.07289610058069229, 0.17742714285850525, 0.44115403294563293, -0.18200574815273285, 0.24563750624656677, 0.0059403604827821255, -0.046177931129932404, -0.13982555270195007, 0.04684886708855629, -0.41081345081329346, -0.44855642318725586, 0.4289398491382599, 0.07663720101118088, -0.12908519804477692, 0.08759177476167679, 0.3628625273704529, -0.5635618567466736, -0.25134050846099854, 0.3186553716659546, -0.2019166797399521, -0.13163183629512787, -0.20967739820480347, 0.08235722035169601, 0.049965329468250275, 0.6422716379165649, 0.2010980099439621, 0.3915328085422516, -0.3583630621433258, -0.17929251492023468, -0.5960273146629333, -0.17861220240592957, 0.08641806989908218, -0.08518479764461517, -0.22825580835342407, 0.15490569174289703, -0.11009515821933746, 0.3391738533973694, -0.17584402859210968, 0.0758240595459938, 0.24115340411663055, 0.000038015685277059674, -0.41481858491897583, 0.09212309122085571, 0.12537768483161926, -0.017390994355082512, 0.2854137718677521, -0.10768584907054901, -0.14416518807411194, 0.12176597863435745, -0.0711134821176529, -0.07582589983940125, -0.11642737686634064, 0.2526831030845642, 0.15655411779880524, 0.10559040307998657, 0.010084067471325397, 0.4512919485569, 0.017851974815130234, -0.14859499037265778, -0.05152134969830513, 0.03749116137623787, -0.35272160172462463, -0.18147289752960205, -0.10226466506719589, 0.1490994691848755, -0.4346386194229126, 0.2957685887813568, 0.021365875378251076, -0.061808280646800995, 0.1016729399561882, 0.005676898639649153, -0.2547983229160309, -0.058562908321619034, 0.031436480581760406, 0.19942903518676758, 0.25435224175453186, 0.7323935031890869, -0.0023266253992915154, 0.1040797010064125, -0.056921057403087616, 0.030946219339966774, -0.028117982670664787, -0.08862271904945374, -0.25010132789611816, -0.13463178277015686, 0.05298806354403496, 0.07657415419816971, 0.08411658555269241, -0.6738632321357727, -0.40565258264541626, 0.0536893792450428, 0.02241557277739048, 0.05789879709482193, 0.7152572870254517, 0.027556542307138443, -0.1257440596818924, -0.11397719383239746, -0.38882240653038025, 0.00857969094067812, -0.22798463702201843, 0.2192232757806778, -0.4018597900867462 ]
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.
[ -0.002133985748514533, -0.3933590352535248, -0.04490889981389046, 0.011329795233905315, -0.22390785813331604, 0.021760975942015648, 0.1978980004787445, 0.027536053210496902, 0.414302259683609, -0.19782714545726776, -0.23707208037376404, 0.39821064472198486, -0.19453240931034088, 0.3028287887573242, 0.35159602761268616, -0.40193265676498413, -0.022123213857412338, -0.2419828176498413, -0.29496026039123535, 0.35135412216186523, 0.06044180318713188, 0.19776852428913116, -0.21249958872795105, 0.0691925510764122, -0.21596968173980713, -0.1789105385541916, -0.35519078373908997, 0.1598132699728012, 0.08680374920368195, 0.008535301312804222, 0.25258201360702515, 0.42071616649627686, -0.047911256551742554, 0.17200236022472382, -0.00011538730905158445, -0.040628209710121155, 0.22244951128959656, -0.12262418121099472, 0.053343016654253006, -0.06257640570402145, 0.17279647290706635, -0.4180443286895752, 0.0764208510518074, -0.17792527377605438, 0.1683042347431183, 0.1507716327905655, 0.12851442396640778, 0.018031053245067596, 0.3956318199634552, -0.2608158588409424, 0.10167452692985535, 0.31338444352149963, -0.15474563837051392, 0.3200140595436096, -0.13475893437862396, -0.021238192915916443, 0.02505510486662388, 0.2562732398509979, 0.6043383479118347, -0.5478257536888123, -0.03986029326915741, 0.09073146432638168, 0.02038217894732952, -0.08250818401575089, 0.12837639451026917, -0.02715415507555008, -0.3430268466472626, -0.25179335474967957, -0.38982969522476196, 0.22086523473262787, -0.05872553586959839, -0.18009454011917114, -0.22417065501213074, -0.5440153479576111, -0.057422444224357605, 0.0014682061737403274, -0.20001620054244995, -0.09796568751335144, -0.20692288875579834, -0.07884544879198074, -0.08055776357650757, -0.10031552612781525, -0.0552615188062191, 0.12278597056865692, 0.2862268388271332, 0.5722649097442627, 0.09168433398008347, 0.16959600150585175, 0.41128894686698914, -0.003770094830542803, -0.1038937121629715, -0.2763197124004364, 0.28899946808815, 0.10878864675760269, -0.4652673602104187, -0.3517707288265228, 0.20772288739681244, -0.27356281876564026, 0.2420552521944046, 0.11815439909696579, 0.12676702439785004, 0.1409202367067337, -0.17259535193443298, 0.23502503335475922, 0.3101992905139923, 0.16164416074752808, -0.0854417160153389, -0.1401854157447815, -0.0408443920314312, -0.10049773752689362, 0.26056742668151855, 0.09310340881347656, 0.13976195454597473, -0.024156706407666206, -0.43835675716400146, -0.11291632801294327, -0.19020868837833405, 0.1750694215297699, -0.23685738444328308, -0.4651109278202057, -0.11562246829271317, -0.13347426056861877, 0.16438159346580505, -0.08931572735309601, -0.32635945081710815, 0.09385589510202408, -0.06360592693090439, 0.2285139560699463, -0.1643097996711731, 0.09540653973817825, -0.0509239099919796, 0.07936469465494156, -0.42001891136169434, -0.1034962609410286, 0.3041563332080841, 0.2007603943347931, 0.08877848833799362, 0.10183306038379669, 0.04135569557547569, 0.16144326329231262, 0.3386404514312744, -0.01859002187848091, 0.07116793096065521, -0.006750204600393772, 0.17005880177021027, -0.24186673760414124, 0.023634061217308044, 0.08572126179933548, -0.327629029750824, -0.14327847957611084, -0.11624554544687271, -0.17121268808841705, 0.09555090963840485, 0.0015717269852757454, -0.24700190126895905, -0.21890227496623993, -0.529689610004425, 0.692986011505127, 0.00901387445628643, -0.07522506266832352, -0.17376214265823364, 0.21700380742549896, -0.22496739029884338, -0.11346253752708435, 0.18577608466148376, 0.13063417375087738, -0.019459569826722145, -0.23661847412586212, -0.08683834224939346, -0.11565877497196198, 0.007798844948410988, 0.2511883080005646, -0.17226925492286682, 0.31111007928848267, 0.0009650007705204189, 0.2511010766029358, 0.26969873905181885, -0.4574740529060364, -0.08002300560474396, 0.299605131149292, -0.07061264663934708, 0.4915619194507599, 0.06509639322757721, 0.20209118723869324, 0.01994786597788334, -0.20112179219722748, 0.27416425943374634, 0.9725068807601929, -0.3398754298686981, -0.04546654596924782, -0.13535140454769135, -0.3082438111305237, 0.584370493888855, 0.29365116357803345, 0.04606017470359802, -0.23651717603206635, -0.1625547707080841, 0.025135979056358337, 0.1369394212961197, 0.11461347341537476, 0.1020970419049263, -0.04339587315917015, -0.31103479862213135, 0.06801377981901169, -0.16123920679092407, -0.1333121359348297, -0.4370241165161133, 0.037815019488334656, 0.0461084358394146, 0.344215989112854, 0.18612156808376312, -0.042499128729104996, 0.2013804018497467, -0.24291658401489258, -0.20070359110832214, -0.32318753004074097, 0.12822403013706207, -0.03407769277691841, -0.006764433346688747, -0.3257838189601898, -0.0405891127884388, 0.39110180735588074, 0.11225410550832748, -0.14964769780635834, -0.5206970572471619, 0.2977626323699951, -0.08318765461444855, 0.07015016674995422, 0.0002397678908891976, 0.6380752921104431, -0.09240636229515076, -0.13536231219768524, 0.18318213522434235, 0.3113526701927185, -0.24773862957954407, 0.14653021097183228, 0.36833930015563965, 0.1787833422422409, 0.17248529195785522, -0.0537874698638916, 0.04117799922823906, 0.01477865967899561, 0.06409096717834473, -0.2540859282016754, -0.041745029389858246, 0.4979395866394043, -0.20704546570777893, 0.4949685335159302, -0.05947205796837807, 0.09615936130285263, -0.07772042602300644, 0.0036060153506696224, -0.21134968101978302, 0.4170468747615814, 0.2573398947715759, -0.1638273447751999, 0.3197740912437439, -0.17490935325622559, -0.38928574323654175, 0.21816878020763397, 0.14777952432632446, 0.05071191117167473, -0.023939412087202072, -0.0040272874757647514, -0.004173790570348501, -0.2835710346698761, -0.08406412601470947, 0.2592909634113312, 0.6727410554885864, 0.21467217803001404, 0.17579346895217896, -0.00355359329842031, -0.2128886729478836, -0.0756315141916275, 0.017034776508808136, 0.21315494179725647, -0.01838492974638939, 0.1383446305990219, 0.06322667747735977, 0.05501282587647438, 0.027259821072220802, -0.1688719093799591, 0.16363950073719025, -0.15025171637535095, -0.11828187853097916, 0.11910361051559448, -0.24008609354496002, -0.644916832447052, -0.19026373326778412, -0.005381218623369932, -0.3374342918395996, -0.337982714176178, 0.061976782977581024, 0.09609249234199524, -0.09279245138168335, 0.17458879947662354, 0.3390880525112152, 0.5370781421661377, -0.4226085841655731, -0.27869999408721924, -0.07020215690135956, -0.313002347946167, -0.2287057340145111, 0.10595954954624176, 0.47610998153686523, 0.314726322889328, 0.43627259135246277, 0.3334589898586273, -0.07831178605556488, -0.011106652207672596, -0.40816837549209595, 0.07382792234420776, -0.3157108724117279, 0.2591187357902527, 0.10779192298650742, -0.1619817167520523, 0.15699610114097595, -0.36640802025794983, -0.054459813982248306, 0.025672685354948044, -0.10825207084417343, 0.005735535640269518, -0.02132072113454342, 0.06178317964076996, -0.22494050860404968, -0.20574085414409637, -0.60491544008255, -0.3986761271953583, 0.3317519426345825, -0.268457293510437, 0.15955212712287903, 0.1113506481051445, -0.29019051790237427, 0.21595756709575653, -0.06657843291759491, -0.03549867123365402, -0.19768507778644562, -0.026613177731633186, 0.048638150095939636, -0.3335722088813782, -0.22930161654949188, -0.33348003029823303, -0.08374890685081482, 0.28511911630630493, -0.06061443313956261, -0.019240260124206543, -0.2144840955734253, 0.13713258504867554, -0.09409278631210327, 0.11316592991352081, 0.18199382722377777, 0.26239264011383057, -0.2219083160161972, 0.1516181379556656, -0.13297529518604279, -0.33784160017967224, 0.2892357110977173, 0.30653196573257446, 0.46678003668785095, 0.2628142833709717, 0.059581268578767776, 0.19852980971336365, 0.7363941073417664, 0.24672763049602509, 0.030771151185035706, 0.1628226637840271, 0.24470633268356323, -0.00857924111187458, 0.12864327430725098, -0.2955617904663086, 0.24368076026439667, -0.2057005763053894, 0.4891052544116974, 0.04684237390756607, 0.12142480164766312, -0.27849727869033813, -0.31981512904167175, 0.09729410707950592, -0.29605528712272644, -0.24099448323249817, 0.2288609743118286, -0.22867752611637115, -0.029668258503079414, -0.04629221558570862, -0.37965261936187744, -0.3028072714805603, -0.037758756428956985, 0.024953819811344147, 0.3650670349597931, -0.08353598415851593, 0.14516791701316833, -0.3331812620162964, -0.058830875903367996, -0.4737011194229126, 0.11649056524038315, 0.02723420225083828, 0.3008979558944702, -0.0765068307518959, -0.10251381248235703, -0.09614036977291107, 0.09850391000509262, 0.7110231518745422, -0.4442330002784729, -0.14522933959960938, 0.14357201755046844, 0.0976986438035965, -0.09330235421657562, -0.12175718694925308, -0.28652945160865784, 0.11032082140445709, 0.5207927823066711, 0.3619166910648346, -0.4547298550605774, -0.2635485529899597, 0.14256997406482697, -0.2180936485528946, -0.04638513922691345, -0.18992513418197632, -0.2505960166454315, -0.2670263350009918, -0.09710869938135147, 0.04224786534905434, 0.22509320080280304, 0.2796405851840973, -0.1162927970290184, 0.2169065624475479, -0.07204917073249817, 0.17009682953357697, 0.21983082592487335, 0.0008598408312536776, -0.02683493122458458, 0.2566704750061035, 0.14444376528263092, 0.08391502499580383, -0.09641490876674652, -0.35581180453300476, 0.1244938001036644, -0.02867574244737625, 0.13755032420158386, 0.3451252281665802, -0.06954687833786011, 0.5981546640396118, 0.15671098232269287, 0.300218790769577, 0.22123467922210693, -0.13670647144317627, -0.013893657363951206, -0.18889369070529938, 0.2365732342004776, 0.21810728311538696, 0.2083166390657425, -0.1431369036436081, -0.24706323444843292, 0.48490580916404724, 0.2541721761226654, -0.13027392327785492, 0.4766065776348114, -0.4629567563533783, -0.3232702314853668, 0.2342958152294159, 0.08237845450639725, 0.7584081292152405, -0.02576613798737526, 0.2482210248708725, 0.2218547761440277, -0.38729333877563477, 0.5620597004890442, -0.29611584544181824, 0.15221050381660461, -0.060211021453142166, -0.1560712605714798, -0.0034725721925497055, -0.10052987933158875, -0.06659519672393799, 0.1920006424188614, -0.30052220821380615, 0.31709545850753784, 0.47413089871406555, -0.2545061409473419, -0.025425471365451813, 0.5520508289337158, 0.07409635186195374, -0.44490015506744385, -0.06325893849134445, 0.059810902923345566, -0.2775411605834961, 0.12796059250831604, -0.12317351996898651, -0.32199662923812866, 0.19604870676994324, 0.14368177950382233, -0.33403676748275757, 0.08418521285057068, 0.002532614627853036, 0.2179049849510193, 0.30240121483802795, -0.018724041059613228, 0.24353761970996857, -0.15538783371448517, -0.06911821663379669, -0.20241819322109222, -0.08242573589086533, 0.12338777631521225, -0.15316860377788544, 0.1167895719408989, 0.09096920490264893, -0.23128840327262878, 0.5451807975769043, -0.09264512360095978, -0.07114855200052261, -0.24564051628112793, 0.22469691932201385, -0.40141767263412476, -0.20007513463497162, 0.200541689991951, 0.16138695180416107, 0.195503368973732, -0.2907209098339081, 0.5905135869979858, 0.12673091888427734, -0.003993513062596321, 0.01494220457971096, 0.22409634292125702, -0.23790514469146729, 0.49399155378341675, -0.33471807837486267, 0.03277846425771713, 0.1014421135187149, 0.041583351790905, -0.10776840150356293, -0.4338434934616089, -0.0012157423188909888, 0.33424776792526245, 0.07582011073827744, 0.051431961357593536, -0.08145014941692352, 0.07533995062112808, -0.04821412265300751, -0.09353341162204742, -0.18528561294078827, -0.3433988690376282, 0.08340638130903244, 0.26250818371772766, 0.14033231139183044, 0.2258596271276474, -0.06235167756676674, -0.30122560262680054, -0.18288245797157288, 0.4282069802284241, -0.09055987745523453, 0.17181062698364258, -0.16094064712524414, 0.31565922498703003, 0.41203197836875916, 0.46285831928253174, -0.24442195892333984, 0.016906926408410072, 0.02253974974155426, 0.09322283416986465, 0.3734447956085205, -0.23166672885417938, 0.07709405571222305, 0.24294087290763855, -0.13898321986198425, 0.04263553023338318, -0.34411025047302246, -0.12457109242677689, -0.2696315050125122, 0.11243967711925507, 0.13527865707874298, -0.04004857689142227, 0.004309911746531725, -0.13928286731243134, -0.13679300248622894, -0.06827160716056824, -0.03434356302022934, -0.4005202054977417, -0.10314907133579254, 0.08782892674207687, -0.047226082533597946, 0.06675806641578674, 0.061760712414979935, 0.1681150496006012, 0.05270753428339958, -0.08132509142160416, 0.3830561339855194, 0.09156350046396255, 0.07138922065496445, -0.18555377423763275, -0.1679844707250595, 0.054718393832445145, 0.2784917950630188, 0.14243048429489136, -0.048964064568281174, 0.0697530135512352, -0.21316052973270416, -0.13101544976234436, 0.2193908989429474, 0.22825919091701508, 0.05144965276122093, -0.19707226753234863, 0.13877631723880768, 0.08696112036705017, -0.07299170643091202, -0.02421007677912712, 0.2587135434150696, 0.0611557811498642, 0.05810309946537018, 0.1557799130678177, 0.1339227259159088, 0.43407899141311646, -0.1839955747127533, 0.047186147421598434, 0.16860894858837128, -0.3724971115589142, 0.05300131440162659, -0.19252854585647583, -0.12392836064100266, 0.043469976633787155, 0.1779734492301941, -0.020713742822408676, 0.11280082166194916, -0.1853485256433487, 0.2707878351211548, 0.3122725486755371, 0.009716646745800972, 0.04766671359539032, 0.1371765434741974, -0.059301573783159256, -0.18108196556568146, 0.20019406080245972, 0.057559527456760406, 0.34439483284950256, 0.018698006868362427, 0.8508671522140503, -0.2761971056461334, -0.38287946581840515, 0.17217451333999634, 0.40964174270629883, 0.031729090958833694, -0.18231993913650513, 0.0848931223154068, 0.07741569727659225, -0.34510061144828796, 0.018537797033786774, 0.038473378866910934, -0.1703852117061615, 0.6866496205329895, -0.1312553882598877, 0.11843214184045792, -0.22235526144504547, -0.2760984003543854, 0.3105561435222626, 0.17144033312797546, -0.05652584135532379, 0.06068389490246773, 0.13091380894184113, -0.09606010466814041, -0.16307979822158813, 0.42816364765167236, 0.20537397265434265, 0.07103615254163742, -0.45058774948120117, -0.055378858000040054, 0.1022796556353569, 0.11845844984054565, -0.0857544094324112, 0.15496812760829926, 0.4437686800956726, -0.12202119082212448, 0.2705669403076172, -0.02374356798827648, -0.02321956865489483, -0.13216722011566162, 0.16326656937599182, -0.3313332796096802, -0.373579740524292, 0.45984187722206116, 0.08283696323633194, -0.11625505238771439, 0.09675152599811554, 0.29102224111557007, -0.6527737975120544, -0.2447599470615387, 0.25609275698661804, -0.06182912737131119, -0.0006645118119195104, -0.18027791380882263, 0.06906040757894516, 0.08660934120416641, 0.6433354616165161, 0.16453365981578827, 0.3060009479522705, -0.4014545679092407, -0.12584920227527618, -0.63165682554245, -0.21747539937496185, 0.17937450110912323, -0.13840076327323914, -0.1703629046678543, 0.18058347702026367, -0.007043853402137756, 0.24776330590248108, -0.2609245181083679, -0.006971169263124466, 0.3292715847492218, 0.04296658933162689, -0.5192553997039795, 0.039342112839221954, 0.1607201099395752, 0.11304469406604767, 0.26529285311698914, -0.020683884620666504, -0.20508524775505066, 0.10758034139871597, -0.06017394736409187, -0.08420147001743317, -0.11822494864463806, 0.23890474438667297, 0.2081911861896515, 0.005505942273885012, 0.06280221045017242, 0.49198225140571594, 0.006552556995302439, -0.17013537883758545, -0.004909828770905733, 0.09330538660287857, -0.3385925590991974, -0.16125258803367615, -0.10824749618768692, 0.22235342860221863, -0.3420107960700989, 0.3595680594444275, -0.017937852069735527, -0.08411511778831482, 0.13189373910427094, 0.02573179267346859, -0.2636851668357849, 0.028533324599266052, -0.008091013878583908, 0.27323275804519653, 0.20005375146865845, 0.7247751355171204, 0.03976431488990784, 0.17260222136974335, -0.16154935956001282, -0.04141344875097275, 0.10445645451545715, -0.17149542272090912, -0.28233760595321655, -0.11068662256002426, 0.07003866136074066, 0.03175509348511696, 0.13340836763381958, -0.7072267532348633, -0.3839821219444275, 0.006639174185693264, -0.03267104923725128, 0.024425430223345757, 0.6573565602302551, 0.045901063829660416, -0.10252996534109116, -0.13903044164180756, -0.32310813665390015, 0.09402628988027573, -0.23133806884288788, 0.1839509755373001, -0.3546333909034729 ]
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?
[ -0.08960597962141037, -0.3705194592475891, -0.044123996049165726, -0.06419700384140015, -0.22869303822517395, 0.02895054779946804, 0.26768994331359863, 0.023359600454568863, 0.3950949013233185, -0.10933436453342438, -0.170592799782753, 0.32033514976501465, -0.20159706473350525, 0.3754856586456299, 0.39844685792922974, -0.4378182590007782, -0.002576937200501561, -0.2555522620677948, -0.3273438513278961, 0.3867937922477722, 0.003052113112062216, 0.11846144497394562, -0.14524997770786285, 0.056555841118097305, -0.28478389978408813, -0.12906812131404877, -0.3320092260837555, 0.10755112022161484, 0.0970587208867073, -0.011513102799654007, 0.23895089328289032, 0.3638800382614136, -0.0029584234580397606, 0.12613274157047272, -0.00011568747868295759, -0.041392624378204346, 0.22455735504627228, -0.08851224184036255, 0.06489231437444687, -0.045699119567871094, 0.17900130152702332, -0.37617799639701843, 0.05878779664635658, -0.14258639514446259, 0.20260295271873474, 0.17821507155895233, 0.17752520740032196, 0.0471881739795208, 0.34591445326805115, -0.32543402910232544, 0.08241244405508041, 0.32058292627334595, -0.2082771211862564, 0.2640053331851959, -0.11156582832336426, -0.05021273344755173, 0.0747942253947258, 0.31859833002090454, 0.6831017136573792, -0.569352388381958, -0.07522591203451157, 0.10948429256677628, 0.018864545971155167, -0.06389612704515457, 0.15395696461200714, -0.08444517105817795, -0.30900222063064575, -0.24644418060779572, -0.4362654983997345, 0.21254341304302216, -0.11863143742084503, -0.1288325935602188, -0.23607242107391357, -0.6045427918434143, -0.0291279386729002, 0.07235991209745407, -0.27344730496406555, -0.009662264958024025, -0.21529707312583923, -0.04984074458479881, -0.173860564827919, -0.10561857372522354, -0.031104685738682747, 0.14210361242294312, 0.3534386456012726, 0.548431932926178, 0.07717567682266235, 0.21021907031536102, 0.4161783456802368, 0.08392110466957092, -0.11613134294748306, -0.27954211831092834, 0.317326158285141, 0.12179882824420929, -0.3628465235233307, -0.34425222873687744, 0.17837630212306976, -0.2727920413017273, 0.26535099744796753, 0.14319463074207306, 0.1328376680612564, 0.05661186948418617, -0.12779051065444946, 0.24213258922100067, 0.358488529920578, 0.08159496635198593, -0.09061963856220245, -0.11054698377847672, 0.01107055600732565, -0.10448640584945679, 0.2219531238079071, 0.10372350364923477, 0.1453361064195633, -0.015399918891489506, -0.4716424345970154, -0.08467719703912735, -0.23086188733577728, 0.11420803517103195, -0.18811652064323425, -0.5452075600624084, -0.07727617025375366, -0.14436057209968567, 0.11034158617258072, -0.08995658904314041, -0.2681082785129547, 0.1173442006111145, -0.05604632571339607, 0.2535907030105591, -0.13841114938259125, 0.048750411719083786, -0.04957455396652222, 0.061328109353780746, -0.45848971605300903, -0.09332165122032166, 0.28703635931015015, 0.17207880318164825, 0.12903068959712982, 0.12023800611495972, -0.019157465547323227, 0.14410123229026794, 0.3373575806617737, 0.028595510870218277, 0.013050233013927937, 0.010993055067956448, 0.18654201924800873, -0.3049395680427551, 0.0114055834710598, 0.23146328330039978, -0.3490493893623352, -0.10899238288402557, -0.20152635872364044, -0.19108161330223083, 0.16736052930355072, 0.01393071748316288, -0.21046431362628937, -0.26141491532325745, -0.5597922801971436, 0.6944104433059692, -0.03270474821329117, -0.10934752225875854, -0.14054320752620697, 0.2590899169445038, -0.16008687019348145, -0.08462855219841003, 0.15515010058879852, 0.04823920130729675, -0.04864722117781639, -0.36557191610336304, -0.07868277281522751, -0.12028509378433228, 0.050715308636426926, 0.26793622970581055, -0.22832445800304413, 0.27010637521743774, 0.018788792192935944, 0.19740833342075348, 0.24765585362911224, -0.476846843957901, -0.03297765180468559, 0.3645946979522705, -0.04434576630592346, 0.4032135009765625, 0.10897734016180038, 0.20473609864711761, 0.04271746426820755, -0.20092681050300598, 0.24048037827014923, 1.0242563486099243, -0.3771040141582489, 0.00882321409881115, -0.1226193830370903, -0.3946358561515808, 0.5419129133224487, 0.33365949988365173, 0.0873924195766449, -0.26467573642730713, -0.11319318413734436, 0.02093152143061161, 0.14002995193004608, 0.06807214766740799, 0.09150015562772751, -0.06338712573051453, -0.37286943197250366, 0.045083921402692795, -0.21324527263641357, -0.09416282176971436, -0.3640451431274414, 0.020041728392243385, -0.07089473307132721, 0.33006343245506287, 0.27682816982269287, -0.10196194797754288, 0.2901107966899872, -0.25997981429100037, -0.13552559912204742, -0.2989262342453003, 0.08099530637264252, -0.09198132157325745, 0.019625572487711906, -0.30178651213645935, -0.04352948069572449, 0.3114902079105377, 0.09314757585525513, -0.1114654541015625, -0.5201098322868347, 0.2931293249130249, -0.03599867597222328, 0.06156448647379875, -0.05629592016339302, 0.592667818069458, -0.16117380559444427, -0.12305732071399689, 0.15402783453464508, 0.30346107482910156, -0.26907244324684143, 0.2271578162908554, 0.2872878611087799, 0.18800120055675507, 0.16485348343849182, -0.01652204990386963, 0.13337042927742004, 0.005181204527616501, -0.012540076859295368, -0.22247470915317535, -0.07112568616867065, 0.49432921409606934, -0.17993170022964478, 0.4873957633972168, -0.04466511681675911, 0.01573972962796688, -0.18632067739963531, -0.04753471165895462, -0.23269319534301758, 0.40750476717948914, 0.3192884624004364, -0.17949748039245605, 0.372585266828537, -0.08623787015676498, -0.35132768750190735, 0.17856428027153015, 0.09627661854028702, 0.00019783173047471792, 0.021255679428577423, -0.012915055267512798, -0.00913463905453682, -0.3338387906551361, -0.11009979993104935, 0.33554238080978394, 0.6644570827484131, 0.24936817586421967, 0.17727512121200562, -0.02865048125386238, -0.19699719548225403, -0.09434279799461365, 0.03606768324971199, 0.21235433220863342, -0.02482217736542225, 0.1570023000240326, -0.010826549492776394, 0.036738961935043335, -0.0019101903308182955, -0.14514297246932983, 0.14484189450740814, -0.20404052734375, -0.08474646508693695, 0.08304198831319809, -0.2019582986831665, -0.6090878844261169, -0.24757298827171326, -0.017396852374076843, -0.3290186822414398, -0.2885229289531708, 0.10992710292339325, 0.08244902640581131, -0.15863138437271118, 0.14156250655651093, 0.22592277824878693, 0.6497547030448914, -0.44584596157073975, -0.2383548766374588, -0.0980750173330307, -0.27364373207092285, -0.11788420379161835, 0.08979290723800659, 0.43891480565071106, 0.29879069328308105, 0.467877060174942, 0.33576500415802, -0.07294604927301407, 0.09495661407709122, -0.351706862449646, 0.035530783236026764, -0.2930784225463867, 0.3098824918270111, 0.13900688290596008, -0.18240360915660858, 0.19600307941436768, -0.3770197927951813, -0.06796387583017349, -0.04905347898602486, -0.09277120977640152, -0.06871553510427475, -0.026185184717178345, 0.07875309139490128, -0.19790850579738617, -0.14304089546203613, -0.6475340723991394, -0.42761489748954773, 0.26240992546081543, -0.2563133239746094, 0.053811248391866684, 0.05618860945105553, -0.2789652645587921, 0.16724105179309845, -0.11408746987581253, -0.030198959633708, -0.16675272583961487, -0.11565957963466644, -0.02231578528881073, -0.3414050042629242, -0.1668616235256195, -0.4368031919002533, -0.08814214915037155, 0.20312482118606567, -0.019464341923594475, -0.019430048763751984, -0.18683083355426788, 0.15789036452770233, -0.12860524654388428, 0.13385537266731262, 0.2255501002073288, 0.14002403616905212, -0.2490691989660263, 0.14498333632946014, -0.0692463219165802, -0.2159966379404068, 0.2266508787870407, 0.31023111939430237, 0.4672711193561554, 0.20428186655044556, 0.021374976262450218, 0.1477930098772049, 0.7359362840652466, 0.22510264813899994, 0.06165274232625961, 0.1580505669116974, 0.2971978783607483, -0.011522057466208935, 0.10848701745271683, -0.20904207229614258, 0.2638307511806488, -0.19775906205177307, 0.48094674944877625, 0.0856478288769722, 0.14452961087226868, -0.18779556453227997, -0.2808537781238556, 0.008959200233221054, -0.2754814028739929, -0.22052478790283203, 0.2475835382938385, -0.2749193608760834, 0.008521344512701035, -0.0770050659775734, -0.3628954589366913, -0.28628265857696533, -0.0392611138522625, -0.0071999807842075825, 0.31482070684432983, 0.03027891181409359, 0.16325701773166656, -0.399127721786499, -0.06987353414297104, -0.4792782962322235, 0.12462065368890762, 0.050041597336530685, 0.27237868309020996, -0.07456653565168381, -0.10680686682462692, -0.08096478134393692, 0.13911312818527222, 0.6718361973762512, -0.46807408332824707, -0.12010300904512405, 0.04675665497779846, 0.08207759261131287, -0.05977208539843559, -0.12703652679920197, -0.23704156279563904, 0.020004434511065483, 0.4522353410720825, 0.3554953634738922, -0.4107634127140045, -0.2772553265094757, 0.22630614042282104, -0.16873964667320251, -0.08126750588417053, -0.22157131135463715, -0.2137247771024704, -0.17879270017147064, -0.12051842361688614, 0.02292405441403389, 0.19871015846729279, 0.2567110061645508, -0.1758771687746048, 0.18028047680854797, -0.09640824049711227, 0.27499058842658997, 0.1885717213153839, -0.004854794591665268, -0.06122702732682228, 0.19523635506629944, 0.15556973218917847, 0.07697052508592606, -0.11399282515048981, -0.28886157274246216, 0.11771632730960846, -0.0944041758775711, 0.0906444787979126, 0.279668390750885, -0.027800949290394783, 0.6893083453178406, 0.08675216883420944, 0.3350839912891388, 0.2698395848274231, -0.0843534767627716, -0.01583132892847061, -0.25507763028144836, 0.1784936934709549, 0.15700088441371918, 0.25420084595680237, -0.15993034839630127, -0.26166707277297974, 0.5813794732093811, 0.30217254161834717, -0.07488245517015457, 0.41239237785339355, -0.5658516883850098, -0.3181344270706177, 0.32106226682662964, 0.00823485478758812, 0.6767162680625916, -0.011694858781993389, 0.2966559827327728, 0.09642019867897034, -0.29173052310943604, 0.5918289422988892, -0.3068067729473114, 0.13567274808883667, -0.08629514276981354, -0.15701107680797577, -0.000716497132088989, -0.10220589488744736, -0.06825960427522659, 0.2856674790382385, -0.2693990468978882, 0.3406903147697449, 0.45153963565826416, -0.25932613015174866, -0.06246368959546089, 0.4970199465751648, 0.04026089608669281, -0.3902256190776825, -0.05338802933692932, 0.03806683048605919, -0.2636736333370209, 0.1580262929201126, -0.13398073613643646, -0.27603286504745483, 0.24240407347679138, 0.09275724738836288, -0.2599030137062073, 0.1079481765627861, 0.002680608071386814, 0.21478161215782166, 0.31649455428123474, 0.024474071338772774, 0.24765878915786743, -0.18959146738052368, -0.15023551881313324, -0.17084519565105438, -0.06086103245615959, 0.13157859444618225, -0.1499565690755844, 0.07499967515468597, 0.10225096344947815, -0.18703819811344147, 0.5334511399269104, -0.06037314236164093, -0.08038946241140366, -0.25751838088035583, 0.2207837551832199, -0.4532485604286194, -0.15675638616085052, 0.15572863817214966, 0.17293326556682587, 0.1708429604768753, -0.3983144462108612, 0.5807893872261047, 0.13632534444332123, 0.016066959127783775, -0.0007993189501576126, 0.16640521585941315, -0.22039064764976501, 0.5007694363594055, -0.31078213453292847, 0.009038083255290985, 0.07070206105709076, 0.006287815514951944, -0.12925170361995697, -0.4176296293735504, -0.04808899760246277, 0.29771339893341064, 0.09809528291225433, 0.050419941544532776, -0.009111926890909672, 0.18022210896015167, -0.11553855985403061, -0.030411778017878532, -0.11281515657901764, -0.2936980724334717, 0.13766291737556458, 0.1897755265235901, 0.20098933577537537, 0.19345176219940186, -0.08014602214097977, -0.3293381631374359, -0.1923343688249588, 0.49070754647254944, -0.03517523407936096, 0.2011379450559616, -0.08926881849765778, 0.35346752405166626, 0.39224258065223694, 0.4573080837726593, -0.23124609887599945, 0.03419901803135872, 0.07259207218885422, 0.09041838347911835, 0.4238545596599579, -0.21112443506717682, 0.05492478609085083, 0.299670934677124, -0.15968701243400574, 0.032893575727939606, -0.2968650758266449, -0.12285884469747543, -0.24775628745555878, 0.11770709604024887, 0.1244916170835495, -0.019673913717269897, 0.0004342930915299803, -0.10363553464412689, -0.05940980091691017, -0.056695520877838135, 0.016158530488610268, -0.3881985545158386, -0.07074007391929626, 0.08002299815416336, -0.056610189378261566, 0.059175360947847366, 0.01889616996049881, 0.3053603172302246, 0.0377407968044281, -0.0479910634458065, 0.31572553515434265, 0.12572351098060608, 0.08717872202396393, -0.15319713950157166, -0.19503740966320038, 0.017244890332221985, 0.30177679657936096, 0.17275947332382202, -0.03927145153284073, 0.16133426129817963, -0.1671971082687378, -0.15903231501579285, 0.22609056532382965, 0.2231980711221695, 0.040098417550325394, -0.16032077372074127, 0.11157473921775818, 0.08390437811613083, -0.0976882129907608, 0.012813261710107327, 0.1506921947002411, 0.11321607232093811, 0.06102839857339859, 0.14053042232990265, 0.17595894634723663, 0.45911073684692383, -0.2538209557533264, -0.002814921783283353, 0.14726115763187408, -0.24171823263168335, 0.09182554483413696, -0.0745333805680275, -0.04242345690727234, 0.022125395014882088, 0.17669150233268738, 0.016910085454583168, 0.12707729637622833, -0.3183574974536896, 0.25807932019233704, 0.31017225980758667, 0.005788812879472971, 0.014650939963757992, 0.09429789334535599, -0.06749285757541656, -0.1866593062877655, 0.31709596514701843, 0.009613134898245335, 0.34642523527145386, 0.04529726132750511, 0.8753758668899536, -0.33376312255859375, -0.42751362919807434, 0.15399393439292908, 0.44894060492515564, 0.056862711906433105, -0.222113698720932, 0.0373581238090992, 0.12828105688095093, -0.368175745010376, 0.11881013959646225, 0.016469545662403107, -0.14899735152721405, 0.6261696815490723, -0.11506856232881546, 0.1415538489818573, -0.1920897215604782, -0.2517254054546356, 0.308824360370636, 0.15109413862228394, -0.10871945321559906, 0.003981576766818762, 0.08623914420604706, -0.031953852623701096, -0.08976542949676514, 0.4948278069496155, 0.24426427483558655, 0.04460235312581062, -0.42176422476768494, -0.05275587737560272, 0.06980505585670471, 0.15835516154766083, -0.09271056205034256, 0.1416843980550766, 0.4574715495109558, -0.14659762382507324, 0.259817510843277, -0.012058144435286522, -0.028776664286851883, -0.10949544608592987, 0.10569129139184952, -0.4016031324863434, -0.36630138754844666, 0.4545687139034271, 0.09633439034223557, -0.10901317745447159, 0.05089859664440155, 0.2974778413772583, -0.6373803019523621, -0.23752887547016144, 0.3041929304599762, -0.1298777014017105, -0.07157357782125473, -0.18063604831695557, 0.06053190305829048, 0.05796726047992706, 0.6411348581314087, 0.24478384852409363, 0.35875195264816284, -0.4191928505897522, -0.12837056815624237, -0.5469262599945068, -0.25038883090019226, 0.14156474173069, -0.07870970666408539, -0.26345717906951904, 0.1379135549068451, -0.04050283879041672, 0.25394561886787415, -0.27015817165374756, 0.10107441246509552, 0.23565636575222015, 0.09575539827346802, -0.4365881085395813, 0.08838320523500443, 0.13122279942035675, 0.10667721927165985, 0.25308334827423096, -0.09535457193851471, -0.1709667146205902, 0.10993640869855881, -0.12099023908376694, -0.08666328340768814, -0.08444841206073761, 0.3115096986293793, 0.22627824544906616, 0.04814795032143593, 0.05102430284023285, 0.5658938884735107, 0.00874767079949379, -0.17838306725025177, 0.019667858257889748, 0.009437787346541882, -0.3435698449611664, -0.16456297039985657, -0.0811554566025734, 0.15427018702030182, -0.3199593722820282, 0.34175774455070496, -0.01973627135157585, -0.11823628097772598, 0.1042819619178772, 0.0029659217689186335, -0.22956091165542603, -0.050969503819942474, 0.005004292353987694, 0.182532399892807, 0.25139591097831726, 0.7283112406730652, 0.028254276141524315, 0.09515015780925751, -0.07251230627298355, 0.0025232979096472263, 0.031027665361762047, -0.10433545708656311, -0.2157593071460724, -0.1757686883211136, 0.044877760112285614, 0.013303973712027073, 0.12388890981674194, -0.7102763056755066, -0.36426377296447754, 0.019617676734924316, 0.013225133530795574, 0.08402039110660553, 0.7129794955253601, 0.07473521679639816, -0.14252954721450806, -0.13840141892433167, -0.3290697932243347, 0.051101792603731155, -0.2399122267961502, 0.19118718802928925, -0.3383631408214569 ]
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!
[ -0.09565513581037521, -0.43038487434387207, -0.0462937094271183, -0.06745029985904694, -0.2578941881656647, 0.009947205893695354, 0.20224884152412415, 0.07624092698097229, 0.29039692878723145, -0.11092492938041687, -0.1859137862920761, 0.29462289810180664, -0.23009242117404938, 0.3789542615413666, 0.4100237190723419, -0.4520007073879242, 0.024441109970211983, -0.17720668017864227, -0.2945508360862732, 0.3627449870109558, 0.015423659235239029, 0.14288900792598724, -0.15470294654369354, 0.0810639038681984, -0.31288430094718933, -0.16926127672195435, -0.3484940528869629, 0.10086256265640259, 0.13701598346233368, -0.024620462208986282, 0.2090308964252472, 0.37785276770591736, 0.034009162336587906, 0.10139449685811996, -0.00011142912990180776, -0.09728644043207169, 0.20507892966270447, -0.09062550216913223, 0.10297652333974838, -0.05900316312909126, 0.16401736438274384, -0.3679225444793701, -0.0027357409708201885, -0.1269502490758896, 0.15940672159194946, 0.225113645195961, 0.2023809403181076, 0.08568336069583893, 0.39888110756874084, -0.3097245991230011, 0.11813075095415115, 0.32869458198547363, -0.18170078098773956, 0.23797312378883362, -0.08357023447751999, -0.06251078099012375, 0.060989443212747574, 0.23452846705913544, 0.6071351766586304, -0.5887544751167297, -0.06404906511306763, 0.1405182033777237, 0.03854504972696304, 0.021420918405056, 0.1444118618965149, -0.038735970854759216, -0.29622364044189453, -0.2327663004398346, -0.4362355172634125, 0.28065410256385803, -0.16127392649650574, -0.1306261569261551, -0.2634526491165161, -0.563300371170044, -0.0008529634797014296, 0.0014620971633121371, -0.3300986588001251, 0.04478824883699417, -0.21435777842998505, -0.03346634283661842, -0.15243948996067047, -0.060267411172389984, -0.040601182729005814, 0.11201071739196777, 0.36312389373779297, 0.5616534948348999, 0.08171797543764114, 0.13045653700828552, 0.37812280654907227, 0.05841723084449768, -0.19902954995632172, -0.2562062442302704, 0.28373706340789795, 0.11374134570360184, -0.3466758728027344, -0.3419148623943329, 0.1750624179840088, -0.1930897980928421, 0.3228089511394501, 0.10108854621648788, 0.09349624812602997, 0.09113761782646179, -0.10029949247837067, 0.2920592427253723, 0.3476564884185791, 0.12236906588077545, -0.05660543218255043, -0.12424643337726593, 0.039759568870067596, -0.06336764246225357, 0.20475304126739502, 0.1504274308681488, 0.19862747192382812, -0.007035157643258572, -0.42395004630088806, -0.11359032988548279, -0.18267913162708282, 0.14201082289218903, -0.2142997831106186, -0.5234177112579346, -0.02995622344315052, -0.1304442435503006, 0.08226396888494492, -0.05673924461007118, -0.2853320837020874, 0.0973641499876976, -0.08805633336305618, 0.1967228800058365, -0.12846854329109192, 0.07483359426259995, -0.08253398537635803, 0.07176250219345093, -0.4690386652946472, -0.048599973320961, 0.30152595043182373, 0.10744784772396088, 0.13133546710014343, 0.13816606998443604, -0.03774357587099075, 0.13288015127182007, 0.3394286632537842, 0.03031231462955475, -0.00033135770354419947, -0.03616666793823242, 0.15753677487373352, -0.33455848693847656, -0.04376891627907753, 0.15608325600624084, -0.40617045760154724, -0.1570379137992859, -0.1264195293188095, -0.21524950861930847, 0.13307417929172516, 0.05488704517483711, -0.23455175757408142, -0.2543679177761078, -0.4572625458240509, 0.7054927349090576, -0.046238988637924194, -0.03340635076165199, -0.13866697251796722, 0.28012052178382874, -0.2386479377746582, -0.07164378464221954, 0.15914325416088104, 0.09545491635799408, 0.000973437970969826, -0.37805646657943726, -0.03246137127280235, -0.10533326864242554, -0.06803692132234573, 0.3193642497062683, -0.17716696858406067, 0.18924590945243835, 0.05113634467124939, 0.20222032070159912, 0.24152317643165588, -0.49662354588508606, -0.033441998064517975, 0.3052603304386139, -0.03650665283203125, 0.3519679605960846, 0.07787419855594635, 0.18235014379024506, 0.09818737953901291, -0.19245630502700806, 0.26532718539237976, 1.0035322904586792, -0.3828693628311157, 0.022055748850107193, -0.12745924293994904, -0.40109652280807495, 0.48556265234947205, 0.32760190963745117, 0.023420998826622963, -0.3136386275291443, -0.1300782859325409, -0.025605084374547005, 0.09931141883134842, 0.06988126039505005, 0.07652926445007324, -0.06977460533380508, -0.3187344968318939, 0.013977989554405212, -0.15631644427776337, -0.13527335226535797, -0.4236288070678711, 0.034150730818510056, -0.06946315616369247, 0.3329487442970276, 0.32506075501441956, -0.1376507580280304, 0.21762597560882568, -0.26340481638908386, -0.10048346221446991, -0.3129212260246277, 0.1418483406305313, -0.10447493940591812, 0.05392621085047722, -0.2799593210220337, -0.07492871582508087, 0.3310253918170929, 0.08667047321796417, -0.12241590023040771, -0.46758076548576355, 0.2426159828901291, -0.10650418698787689, 0.01754617504775524, 0.007299565710127354, 0.6171312928199768, -0.14959514141082764, -0.07000712305307388, 0.2050333172082901, 0.2820175290107727, -0.28210899233818054, 0.25474727153778076, 0.28865140676498413, 0.22653202712535858, 0.18389944732189178, 0.012913758866488934, 0.10476882755756378, -0.01823929324746132, -0.03309952840209007, -0.18438971042633057, -0.12761735916137695, 0.533774733543396, -0.1134045198559761, 0.42374786734580994, 0.010280597023665905, -0.0016262957360595465, -0.1857483983039856, -0.08734390139579773, -0.2569581866264343, 0.35327187180519104, 0.3242207169532776, -0.21554189920425415, 0.3575746417045593, -0.09278342872858047, -0.38738495111465454, 0.15766696631908417, 0.18133831024169922, 0.02245199866592884, 0.08836910873651505, 0.022484732791781425, -0.043993666768074036, -0.2892811596393585, -0.0826866626739502, 0.36099016666412354, 0.6006553769111633, 0.30525362491607666, 0.17336371541023254, 0.027060024440288544, -0.3042919337749481, -0.11205681413412094, 0.015227694064378738, 0.18314236402511597, -0.030197197571396828, 0.15285532176494598, 0.04158676788210869, 0.055999789386987686, -0.048957306891679764, -0.1661144196987152, 0.09252739697694778, -0.17580822110176086, -0.08189018070697784, 0.0395759679377079, -0.16714338958263397, -0.5780365467071533, -0.24609337747097015, 0.04412269964814186, -0.32625436782836914, -0.255086213350296, 0.15884137153625488, 0.02943447232246399, -0.16992664337158203, 0.17299672961235046, 0.2292766273021698, 0.6286777257919312, -0.43886110186576843, -0.18500061333179474, -0.10761895030736923, -0.3222370147705078, -0.15262551605701447, 0.12096062302589417, 0.41602060198783875, 0.29050716757774353, 0.4932442009449005, 0.2830921709537506, -0.026001228019595146, 0.10474321991205215, -0.3773646354675293, 0.05445162579417229, -0.27542635798454285, 0.28514420986175537, 0.1103922575712204, -0.22095783054828644, 0.157856285572052, -0.39220547676086426, -0.050060104578733444, -0.12310833483934402, -0.09991301596164703, -0.08157988637685776, -0.026580797508358955, 0.042091283947229385, -0.19186252355575562, -0.16757245361804962, -0.6334232687950134, -0.4681072533130646, 0.3418899476528168, -0.2122172862291336, 0.10471151024103165, 0.057700224220752716, -0.2606828510761261, 0.16616958379745483, -0.15556247532367706, 0.02795887365937233, -0.16115058958530426, -0.08060231059789658, 0.027160409837961197, -0.36503589153289795, -0.16097736358642578, -0.42607787251472473, -0.13156571984291077, 0.21172524988651276, -0.08573530614376068, -0.018160639330744743, -0.25038689374923706, 0.07165055721998215, -0.13314278423786163, 0.1169314980506897, 0.26750218868255615, 0.19170324504375458, -0.2375509887933731, 0.09199517965316772, -0.030880078673362732, -0.23764267563819885, 0.2405964583158493, 0.3334428071975708, 0.3925416171550751, 0.14052048325538635, 0.030540570616722107, 0.19585321843624115, 0.7353605628013611, 0.2564299702644348, 0.023825863376259804, 0.18662692606449127, 0.2621115446090698, 0.024636389687657356, 0.16013890504837036, -0.16981853544712067, 0.2595054805278778, -0.17571216821670532, 0.44925418496131897, 0.1285616159439087, 0.18453162908554077, -0.23330718278884888, -0.24578748643398285, -0.01603531278669834, -0.2677397131919861, -0.2062060534954071, 0.24365246295928955, -0.2004663199186325, -0.0054403324611485004, -0.09856344014406204, -0.38829532265663147, -0.2598910331726074, -0.03290402889251709, 0.017790794372558594, 0.287910521030426, 0.002848853589966893, 0.16430732607841492, -0.31946754455566406, -0.055637177079916, -0.4845144748687744, 0.14956335723400116, 0.07221463322639465, 0.26892682909965515, -0.06599939614534378, -0.10260530561208725, -0.11336361616849899, 0.08600833266973495, 0.6084621548652649, -0.46098792552948, -0.12899261713027954, 0.02967524528503418, 0.09475580602884293, -0.12481434643268585, -0.10218691825866699, -0.26836177706718445, -0.06343214213848114, 0.46247410774230957, 0.37933430075645447, -0.36836323142051697, -0.3086761236190796, 0.2527795732021332, -0.2167629599571228, -0.05535298213362694, -0.23541651666164398, -0.21979965269565582, -0.18149441480636597, -0.15289431810379028, 0.09308314323425293, 0.2238735407590866, 0.17905877530574799, -0.10410721600055695, 0.1205001100897789, -0.07052141427993774, 0.20699220895767212, 0.23102588951587677, 0.03242015838623047, -0.0248913262039423, 0.202123761177063, 0.12999317049980164, 0.07291962206363678, -0.12812310457229614, -0.34359559416770935, 0.05205593258142471, -0.07871641218662262, 0.16747893393039703, 0.2720284163951874, 0.038946427404880524, 0.6369622349739075, 0.09319540113210678, 0.348348468542099, 0.2464488297700882, -0.10003010183572769, 0.028285088017582893, -0.2644093930721283, 0.11702138185501099, 0.19253753125667572, 0.24081197381019592, -0.13324132561683655, -0.18242208659648895, 0.6144341230392456, 0.32102540135383606, -0.07534237951040268, 0.39156875014305115, -0.4802115559577942, -0.2988836169242859, 0.30478939414024353, 0.04767739772796631, 0.6780331134796143, -0.008658544160425663, 0.2885778844356537, 0.10952293127775192, -0.35490503907203674, 0.566443920135498, -0.3462434411048889, 0.15785494446754456, -0.10396985709667206, -0.19105486571788788, 0.007184064947068691, -0.08794993162155151, -0.04873119294643402, 0.2688724100589752, -0.34670865535736084, 0.3435438275337219, 0.45668265223503113, -0.2356591373682022, -0.08080559223890305, 0.5609845519065857, -0.04154331609606743, -0.40955638885498047, -0.08284204453229904, 0.0758703425526619, -0.28342723846435547, 0.1413346379995346, -0.17460080981254578, -0.2745564877986908, 0.2578778564929962, 0.10646635293960571, -0.2738930881023407, 0.09171771258115768, -0.0500282384455204, 0.17283356189727783, 0.2756238877773285, 0.03175930306315422, 0.19174419343471527, -0.20281732082366943, -0.06302546709775925, -0.1964411586523056, -0.07966642081737518, 0.15450164675712585, -0.14524778723716736, 0.10553432255983353, 0.08784643560647964, -0.20984165370464325, 0.5108202695846558, -0.05306711420416832, -0.1016976609826088, -0.2674710750579834, 0.2604370713233948, -0.4042937159538269, -0.17751367390155792, 0.10693458467721939, 0.13696090877056122, 0.20796985924243927, -0.292481005191803, 0.5627292394638062, 0.09027541428804398, -0.005863124039024115, 0.05533440783619881, 0.15226757526397705, -0.27232465147972107, 0.5453826189041138, -0.25501900911331177, 0.01121588610112667, 0.05190276354551315, 0.011895972304046154, -0.14888282120227814, -0.36376768350601196, -0.08748192340135574, 0.27418458461761475, 0.09738308936357498, -0.00017540162662044168, -0.013708110898733139, 0.148003488779068, -0.11333443224430084, -0.02041776292026043, -0.12666338682174683, -0.3535959720611572, 0.17925092577934265, 0.2160957008600235, 0.2557377219200134, 0.21601098775863647, -0.07465723901987076, -0.24117819964885712, -0.19481374323368073, 0.4819282591342926, 0.027669662609696388, 0.26395002007484436, -0.1327856332063675, 0.3500429689884186, 0.4172840416431427, 0.5425981283187866, -0.2901419401168823, 0.01948871649801731, 0.04544682428240776, 0.11256454885005951, 0.4598182737827301, -0.24196526408195496, 0.0767640471458435, 0.28588107228279114, -0.1301109790802002, 0.07015662640333176, -0.3621214032173157, -0.17070724070072174, -0.26530221104621887, 0.10153096914291382, 0.10833405703306198, 0.04718959704041481, 0.00317596853710711, -0.019687555730342865, -0.07955937832593918, -0.09573731571435928, 0.018483707681298256, -0.37602153420448303, -0.09828220307826996, 0.1537202149629593, -0.06953918188810349, 0.04239388555288315, 0.12043507397174835, 0.20509037375450134, 0.046698737889528275, -0.06932928413152695, 0.3236750066280365, 0.12583915889263153, 0.07482609897851944, -0.15220272541046143, -0.1602603793144226, 0.007074384950101376, 0.22868555784225464, 0.13647712767124176, -0.061574384570121765, 0.16175289452075958, -0.2716625928878784, -0.1360899955034256, 0.24081094563007355, 0.14597181975841522, 0.07060441374778748, -0.15793447196483612, 0.11631926149129868, 0.14452940225601196, -0.11241637915372849, 0.0010165201965719461, 0.20374658703804016, 0.056546300649642944, 0.12797650694847107, 0.15175612270832062, 0.16456303000450134, 0.45382392406463623, -0.2142961025238037, 0.0364837683737278, 0.14220288395881653, -0.37099379301071167, 0.11530398577451706, -0.1337980180978775, -0.025103125721216202, 0.024295218288898468, 0.1567724049091339, -0.009453579783439636, 0.2167474925518036, -0.3467017412185669, 0.28668469190597534, 0.31906282901763916, -0.03136863559484482, 0.055137019604444504, 0.19727467000484467, -0.002018862869590521, -0.16688482463359833, 0.27943533658981323, 0.07129979133605957, 0.3188147246837616, 0.05534595623612404, 0.8951664566993713, -0.24943041801452637, -0.43798673152923584, 0.08847947418689728, 0.39543741941452026, 0.02461969293653965, -0.2290686070919037, 0.10322840511798859, 0.08457633852958679, -0.3302845060825348, 0.09990277141332626, 0.041745010763406754, -0.1267833411693573, 0.6448745727539062, -0.08233329653739929, 0.14766822755336761, -0.15032047033309937, -0.23305928707122803, 0.3292798697948456, 0.17817793786525726, -0.11302322894334793, 0.013509207405149937, 0.0990215390920639, -0.10855206102132797, -0.19895850121974945, 0.5276768803596497, 0.24399076402187347, 0.055927544832229614, -0.4059011936187744, 0.007024949416518211, 0.11586363613605499, 0.14389587938785553, -0.09145306795835495, 0.10737704485654831, 0.4503690302371979, -0.14932763576507568, 0.21223944425582886, 0.03401568531990051, -0.06872903555631638, -0.09852887690067291, 0.03311290591955185, -0.4560566246509552, -0.3647691309452057, 0.456890344619751, 0.1024249866604805, -0.08989399671554565, 0.042917583137750626, 0.3086777329444885, -0.6284579634666443, -0.18853949010372162, 0.29875877499580383, -0.13009721040725708, -0.07465940713882446, -0.22611458599567413, 0.0887196734547615, 0.07111997157335281, 0.5859490036964417, 0.2188568413257599, 0.35041671991348267, -0.40086081624031067, -0.09598866105079651, -0.6091050505638123, -0.20071952044963837, 0.1193348839879036, -0.08799014985561371, -0.21377204358577728, 0.1967511773109436, -0.10631431639194489, 0.27466580271720886, -0.1899932622909546, 0.11173823475837708, 0.18231917917728424, 0.019365742802619934, -0.4103381633758545, 0.15257972478866577, 0.09875063598155975, 0.011352363042533398, 0.30674153566360474, -0.09572204947471619, -0.170171320438385, 0.08457701653242111, -0.0756978988647461, -0.08605217188596725, -0.051421359181404114, 0.27112579345703125, 0.1778280884027481, 0.04539419338107109, 0.0555259995162487, 0.4602873623371124, -0.04975084960460663, -0.14651736617088318, -0.02097514271736145, -0.021388906985521317, -0.39996102452278137, -0.17749810218811035, -0.08906062692403793, 0.17003114521503448, -0.3348320722579956, 0.24290971457958221, -0.04528430476784706, -0.10608500242233276, 0.1206902265548706, 0.01244246307760477, -0.13162101805210114, -0.03240455687046051, 0.03830355033278465, 0.18737410008907318, 0.228465735912323, 0.7360839247703552, 0.037493377923965454, 0.08689627051353455, -0.0959966778755188, -0.054913122206926346, 0.05435870215296745, -0.09335880726575851, -0.23394685983657837, -0.17539483308792114, 0.07947846502065659, 0.0916745588183403, 0.07787437736988068, -0.6713749170303345, -0.35351109504699707, 0.01052528154104948, 0.010180986486375332, 0.07832980155944824, 0.7260479927062988, 0.07258140295743942, -0.15513920783996582, -0.15334582328796387, -0.3561753034591675, 0.0600898303091526, -0.2716899812221527, 0.1818741410970688, -0.4164086878299713 ]
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.
[ -0.0969296246767044, -0.46223971247673035, -0.07604546844959259, 0.017250584438443184, -0.3047633767127991, 0.05286393687129021, 0.22632122039794922, 0.06489531695842743, 0.276897132396698, -0.11790060251951218, -0.172514408826828, 0.20421773195266724, -0.2584689259529114, 0.44447678327560425, 0.4080318808555603, -0.41051334142684937, 0.014127161353826523, -0.27331629395484924, -0.2500177323818207, 0.3967762589454651, 0.05055217072367668, 0.08481966704130173, -0.17099376022815704, 0.06342436373233795, -0.2406720519065857, -0.13419732451438904, -0.33941221237182617, 0.08957741409540176, 0.1340947300195694, 0.027393314987421036, 0.14990706741809845, 0.30573007464408875, 0.027069417759776115, 0.14519363641738892, -0.0001170386458397843, -0.10317879170179367, 0.2246209979057312, -0.054976098239421844, 0.1232357993721962, -0.03875347226858139, 0.11764194071292877, -0.46906694769859314, 0.001323521020822227, -0.09075804799795151, 0.2729416787624359, 0.19905033707618713, 0.24085493385791779, 0.05113839730620384, 0.30715298652648926, -0.30362704396247864, 0.08462031185626984, 0.27169832587242126, -0.23924680054187775, 0.3131471872329712, -0.0409318208694458, -0.03336234390735626, 0.06543564051389694, 0.2598799765110016, 0.7037147283554077, -0.5388641953468323, 0.0035531341563910246, 0.11596080660820007, 0.02810378186404705, -0.02197118103504181, 0.1327914595603943, -0.07223230600357056, -0.21833215653896332, -0.24540360271930695, -0.34902581572532654, 0.20940156280994415, -0.05419573187828064, -0.08673984557390213, -0.21434029936790466, -0.5842260122299194, -0.018568459898233414, 0.0522705502808094, -0.31707558035850525, 0.13627324998378754, -0.20855441689491272, -0.07745726406574249, -0.1931692212820053, -0.10364774614572525, -0.019546782597899437, 0.10086795687675476, 0.37002745270729065, 0.5515719056129456, 0.05214418098330498, 0.16430217027664185, 0.34780994057655334, 0.022836819291114807, -0.1069188043475151, -0.24575971066951752, 0.29284900426864624, 0.12592433393001556, -0.35657161474227905, -0.31134799122810364, 0.18628355860710144, -0.23718038201332092, 0.2858080565929413, 0.06318112462759018, 0.13804411888122559, 0.03951588645577431, -0.10145273804664612, 0.20296591520309448, 0.38573843240737915, 0.0993378758430481, -0.08045083284378052, -0.14554138481616974, 0.026658881455659866, -0.16947408020496368, 0.19106580317020416, 0.15759561955928802, 0.15331661701202393, -0.08121543377637863, -0.4698126018047333, -0.13959702849388123, -0.24632923305034637, 0.12415030598640442, -0.1801917850971222, -0.5548356175422668, -0.06913362443447113, -0.20435242354869843, 0.027278870344161987, -0.12810049951076508, -0.2890940010547638, 0.13472481071949005, -0.08762960135936737, 0.21490168571472168, -0.13607242703437805, 0.07361121475696564, -0.042918067425489426, 0.06124962493777275, -0.47026845812797546, -0.10851026326417923, 0.27243974804878235, 0.18824897706508636, 0.07662249356508255, 0.11885004490613937, -0.04511087015271187, 0.11725407093763351, 0.32785674929618835, 0.015721598640084267, -0.005909494590014219, 0.01859528198838234, 0.21519234776496887, -0.30850884318351746, -0.09372915327548981, 0.26277387142181396, -0.3823625147342682, -0.049495190382003784, -0.2228010892868042, -0.17501138150691986, 0.15861168503761292, 0.013053184375166893, -0.13337022066116333, -0.24674822390079498, -0.5358604192733765, 0.6971760392189026, 0.029377087950706482, -0.06125351041555405, -0.15396423637866974, 0.24740655720233917, -0.21105247735977173, -0.14691412448883057, 0.08802919089794159, 0.06152327358722687, -0.07157877832651138, -0.3473767042160034, -0.03984362259507179, -0.0877721756696701, 0.016239864751696587, 0.33794912695884705, -0.19569744169712067, 0.2697061598300934, 0.07623138278722763, 0.12101094424724579, 0.2956905663013458, -0.47611987590789795, -0.042139794677495956, 0.39082029461860657, -0.029752973467111588, 0.35634085536003113, 0.12997442483901978, 0.2487475723028183, -0.019198760390281677, -0.17189984023571014, 0.27164506912231445, 0.9539657831192017, -0.4012833833694458, 0.03243048116564751, -0.1284986287355423, -0.4258192777633667, 0.5435494780540466, 0.28876742720603943, 0.09507416188716888, -0.2958891987800598, -0.08115893602371216, 0.0535379983484745, 0.14724496006965637, 0.11203369498252869, 0.1212654784321785, -0.06787078827619553, -0.39558103680610657, 0.023652980104088783, -0.191520094871521, -0.10706013441085815, -0.41904231905937195, -0.0004108141001779586, -0.09667254239320755, 0.2867056727409363, 0.34507930278778076, -0.07108807563781738, 0.30363729596138, -0.31333687901496887, -0.08020200580358505, -0.33918488025665283, 0.07497363537549973, -0.11293680965900421, 0.008267786353826523, -0.3027856647968292, -0.06041329354047775, 0.32144978642463684, 0.13272064924240112, -0.15883862972259521, -0.4637385606765747, 0.21984802186489105, -0.0665508583188057, 0.07297081500291824, -0.006661270745098591, 0.6013067364692688, -0.24156390130519867, -0.09500334411859512, 0.1452827900648117, 0.26792505383491516, -0.29937973618507385, 0.23697517812252045, 0.2770885229110718, 0.2627180814743042, 0.21460817754268646, 0.04104548692703247, 0.18638214468955994, -0.10739422589540482, -0.06513882428407669, -0.21716971695423126, -0.07372843474149704, 0.46423763036727905, -0.13604457676410675, 0.44314372539520264, -0.041300561279058456, -0.05664613097906113, -0.238515242934227, -0.042692068964242935, -0.20221085846424103, 0.36900603771209717, 0.346292644739151, -0.18958203494548798, 0.29735490679740906, -0.10137443989515305, -0.33274316787719727, 0.15534883737564087, 0.11895060539245605, 0.09320922940969467, 0.055571913719177246, -0.012828449718654156, -0.03845522552728653, -0.2931990325450897, -0.14500264823436737, 0.3609066903591156, 0.6997566223144531, 0.21742123365402222, 0.18000756204128265, 0.027521979063749313, -0.15481913089752197, -0.13634808361530304, 0.01572864130139351, 0.17999443411827087, -0.043210674077272415, 0.17419426143169403, 0.020082462579011917, 0.04398123547434807, 0.058782532811164856, -0.13080252707004547, 0.14341144263744354, -0.20149385929107666, -0.04728261008858681, 0.007521140389144421, -0.22752924263477325, -0.6512365341186523, -0.31002604961395264, 0.02434348315000534, -0.3236197233200073, -0.2421455681324005, 0.0769893229007721, -0.004847008269280195, -0.1467948853969574, 0.09195525199174881, 0.19153906404972076, 0.6404306292533875, -0.43182942271232605, -0.17510604858398438, -0.024588339030742645, -0.25083741545677185, -0.1563904732465744, 0.09369319677352905, 0.4699900448322296, 0.24401961266994476, 0.44433027505874634, 0.2816868722438812, -0.06684499979019165, 0.10976620018482208, -0.3688215911388397, 0.022475827485322952, -0.23586627840995789, 0.2514667510986328, 0.13616541028022766, -0.25158610939979553, 0.191814124584198, -0.3306293487548828, -0.05950037017464638, -0.1071176826953888, -0.08023755252361298, -0.14460334181785583, -0.017944319173693657, 0.060918401926755905, -0.21953286230564117, -0.16286633908748627, -0.6563555002212524, -0.42742404341697693, 0.2951759994029999, -0.23542194068431854, 0.09205672889947891, 0.0431002676486969, -0.27662041783332825, 0.1610891968011856, -0.11539866775274277, 0.005506027955561876, -0.11043919622898102, -0.12340127676725388, -0.06430670619010925, -0.38307780027389526, -0.14868482947349548, -0.37940171360969543, -0.03276270627975464, 0.2191522717475891, -0.013218484818935394, -0.06567750126123428, -0.13860073685646057, 0.09972243010997772, -0.194048210978508, 0.0751049742102623, 0.2888384461402893, 0.07945172488689423, -0.3194802403450012, 0.13742376863956451, -0.04164218157529831, -0.24689260125160217, 0.16785049438476562, 0.34957778453826904, 0.4843297302722931, 0.18409185111522675, 0.02253076620399952, 0.16180525720119476, 0.7138544917106628, 0.23926445841789246, 0.07667465507984161, 0.12391604483127594, 0.27780681848526, -0.02764454111456871, 0.15621839463710785, -0.1513795405626297, 0.34679922461509705, -0.13601288199424744, 0.4500523805618286, 0.12394820153713226, 0.21397538483142853, -0.19466909766197205, -0.26240041851997375, -0.00846942514181137, -0.26845112442970276, -0.1703488975763321, 0.255932092666626, -0.25556179881095886, 0.04936544969677925, -0.06720235198736191, -0.4713047444820404, -0.2685267925262451, -0.04071000590920448, -0.04501009359955788, 0.31340155005455017, -0.07623089104890823, 0.19009259343147278, -0.33631759881973267, -0.1390690803527832, -0.4542549252510071, 0.14053933322429657, 0.06395024061203003, 0.3106068968772888, -0.04748241603374481, -0.06474961340427399, -0.08007971197366714, 0.1626017838716507, 0.6325612664222717, -0.4828878939151764, -0.1499166339635849, 0.014306802302598953, 0.1030578762292862, -0.08693000674247742, -0.1750023365020752, -0.24498909711837769, 0.004539045970886946, 0.47975441813468933, 0.37007397413253784, -0.3253917396068573, -0.2802451550960541, 0.22733959555625916, -0.20513910055160522, -0.05097856745123863, -0.22064831852912903, -0.22033725678920746, -0.12599389255046844, -0.09524475783109665, 0.07316480576992035, 0.17764106392860413, 0.24161860346794128, -0.15876401960849762, 0.1878805011510849, -0.02065841294825077, 0.25508955121040344, 0.19459600746631622, -0.011927513405680656, -0.09128208458423615, 0.2523723542690277, 0.11457358300685883, -0.0004394821007736027, -0.07985694706439972, -0.24603697657585144, 0.07207031548023224, -0.10971225798130035, 0.19002871215343475, 0.31174829602241516, -0.05491630360484123, 0.6821685433387756, 0.07290805131196976, 0.3269190788269043, 0.23021677136421204, -0.09951416403055191, 0.04348277300596237, -0.19248589873313904, 0.13781531155109406, 0.17686712741851807, 0.2216978669166565, -0.18922299146652222, -0.2239363044500351, 0.5282307267189026, 0.36822307109832764, -0.021432986482977867, 0.3635031580924988, -0.6384041905403137, -0.2950291335582733, 0.24399636685848236, 0.03697993606328964, 0.6201881766319275, 0.0012556964065879583, 0.33115464448928833, -0.028299055993556976, -0.34045732021331787, 0.5145267844200134, -0.34301191568374634, 0.17477671802043915, -0.06714178621768951, -0.13017219305038452, 0.0227897260338068, -0.105239637196064, -0.0344005711376667, 0.3134414255619049, -0.27450647950172424, 0.33596789836883545, 0.4565870761871338, -0.2524260878562927, -0.12036941200494766, 0.5019967555999756, -0.0001567630679346621, -0.35710838437080383, -0.026803310960531235, 0.025746474042534828, -0.24538476765155792, 0.13652150332927704, -0.22022807598114014, -0.22735558450222015, 0.22728446125984192, 0.1529853641986847, -0.3146008849143982, 0.043000705540180206, -0.02222428284585476, 0.19116713106632233, 0.26745083928108215, 0.009535134769976139, 0.2880608141422272, -0.20601823925971985, -0.11121512204408646, -0.15486101806163788, -0.06647586822509766, 0.09129508584737778, -0.11013209819793701, 0.0908157080411911, 0.08554118126630783, -0.18792076408863068, 0.46513181924819946, -0.08558326959609985, -0.04987729713320732, -0.23923173546791077, 0.24239268898963928, -0.3915677070617676, -0.22097733616828918, 0.20412573218345642, 0.19909945130348206, 0.234390988945961, -0.3496169149875641, 0.5967293977737427, 0.11620333045721054, 0.06519228219985962, 0.001303176861256361, 0.17796744406223297, -0.17881353199481964, 0.5910233855247498, -0.3143088221549988, -0.02504739724099636, 0.04015875235199928, -0.06175951659679413, -0.03130727261304855, -0.4336969256401062, -0.07607483118772507, 0.2946576476097107, 0.14731962978839874, 0.04011423885822296, 0.07178995758295059, 0.06344661861658096, -0.024861272424459457, 0.006633317563682795, -0.08219362795352936, -0.2764517366886139, 0.21565298736095428, 0.2148270606994629, 0.17682822048664093, 0.24704410135746002, -0.0632069930434227, -0.2921832799911499, -0.11573338508605957, 0.4678472876548767, -0.030524753034114838, 0.2143302857875824, -0.14435702562332153, 0.4293505847454071, 0.4802252948284149, 0.530646562576294, -0.2060042917728424, 0.08634337782859802, 0.1195465698838234, 0.09168412536382675, 0.41203537583351135, -0.11529967188835144, 0.01623423956334591, 0.34856125712394714, -0.20053067803382874, 0.06389856338500977, -0.347944438457489, -0.11550615727901459, -0.23809652030467987, 0.12806591391563416, 0.13986648619174957, 0.024180836975574493, 0.040377601981163025, -0.10755091160535812, -0.05919542908668518, -0.0253166314214468, 0.0172389205545187, -0.3293403685092926, -0.06795450299978256, 0.10453052073717117, -0.025513997301459312, 0.015602486208081245, 0.0902869924902916, 0.3568156361579895, 0.07297897338867188, -0.09809201210737228, 0.34946921467781067, 0.1336919516324997, 0.12953081727027893, -0.15184399485588074, -0.21585918962955475, -0.042044252157211304, 0.259137898683548, 0.17298951745033264, -0.14142587780952454, 0.12014441937208176, -0.23301859200000763, -0.12351976335048676, 0.20526504516601562, 0.2321891039609909, 0.07684148848056793, -0.12201127409934998, 0.13692423701286316, 0.06485425680875778, -0.044327594339847565, -0.004679684992879629, 0.18947124481201172, 0.09242010861635208, 0.02123921737074852, 0.16045860946178436, 0.1898418813943863, 0.491852343082428, -0.26601579785346985, 0.02571583352982998, 0.08390656113624573, -0.34789326786994934, 0.10506010800600052, -0.15809723734855652, -0.05164262652397156, 0.019299743697047234, 0.19063273072242737, -0.05377305671572685, 0.13644815981388092, -0.29196637868881226, 0.358884334564209, 0.2559456527233124, 0.05868395417928696, -0.013341056182980537, 0.13964229822158813, -0.0393080972135067, -0.20437981188297272, 0.33291301131248474, 0.0015003990847617388, 0.3294958770275116, 0.02272752858698368, 0.8239639401435852, -0.22451303899288177, -0.4310363829135895, 0.1677999496459961, 0.40355440974235535, 0.022630371153354645, -0.24922966957092285, 0.12385435402393341, 0.10800230503082275, -0.31046244502067566, 0.15777739882469177, 0.009082464501261711, -0.11470036208629608, 0.6107504963874817, -0.08435101807117462, 0.16867579519748688, -0.18480518460273743, -0.23177842795848846, 0.36075064539909363, 0.06150577962398529, -0.11912205070257187, 0.03143632411956787, 0.014175094664096832, -0.08980632573366165, -0.13866424560546875, 0.5110016465187073, 0.19786587357521057, 0.0815284475684166, -0.36255207657814026, -0.08625250309705734, 0.12035297602415085, 0.2306906282901764, -0.1393301784992218, 0.12428097426891327, 0.5147859454154968, -0.18943656980991364, 0.22483009099960327, -0.03141593188047409, -0.027002260088920593, -0.07785665988922119, 0.05451776087284088, -0.46020668745040894, -0.43992283940315247, 0.43461495637893677, 0.04109225794672966, -0.11985764652490616, 0.0570865198969841, 0.3916841745376587, -0.6262602210044861, -0.21965046226978302, 0.33607345819473267, -0.15085987746715546, -0.08783327043056488, -0.2478642761707306, 0.05479941889643669, 0.06175940856337547, 0.6547305583953857, 0.2686644494533539, 0.3616762161254883, -0.36749401688575745, -0.13172826170921326, -0.57748943567276, -0.18853671848773956, 0.09687361866235733, -0.026180434972047806, -0.2875889539718628, 0.23825044929981232, -0.11827699095010757, 0.2193920910358429, -0.12811708450317383, 0.10893557965755463, 0.17497092485427856, 0.07805264741182327, -0.4262302815914154, 0.11324369162321091, 0.09243191033601761, 0.08165443688631058, 0.2756972908973694, -0.0810507982969284, -0.14506010711193085, 0.03609960153698921, -0.1280028223991394, -0.1489008069038391, -0.09801279753446579, 0.32596883177757263, 0.2252078801393509, 0.09227746725082397, 0.03513896465301514, 0.5028654932975769, 0.028904680162668228, -0.19265474379062653, -0.07568537443876266, 0.0043245041742920876, -0.35453489422798157, -0.1635534167289734, -0.05845780670642853, 0.16321392357349396, -0.4496609568595886, 0.36051541566848755, 0.04926587641239166, -0.11139609664678574, 0.09741541743278503, 0.0219487976282835, -0.09885629266500473, -0.09097614139318466, 0.08899535983800888, 0.1587652862071991, 0.28199055790901184, 0.7396944761276245, 0.01803213730454445, 0.04638221114873886, -0.0954442247748375, 0.010541694238781929, -0.019518863409757614, -0.07139844447374344, -0.21920375525951385, -0.20084401965141296, 0.07530627399682999, -0.03850241377949715, 0.08397182077169418, -0.6680577993392944, -0.4091363847255707, 0.03187590092420578, 0.04847072437405586, 0.07243107259273529, 0.7247986197471619, 0.054000113159418106, -0.18400409817695618, -0.11854542791843414, -0.36722370982170105, 0.031231079250574112, -0.2750069797039032, 0.16170822083950043, -0.3264084458351135 ]
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.
[ -0.15286675095558167, -0.4851095974445343, -0.08828967064619064, 0.043889760971069336, -0.3023051917552948, 0.02592059224843979, 0.26860541105270386, 0.09124496579170227, 0.272230327129364, -0.07411915063858032, -0.16744044423103333, 0.12118710577487946, -0.2572724223136902, 0.4499773681163788, 0.36345237493515015, -0.4062100350856781, 0.030634425580501556, -0.2730121612548828, -0.2287900596857071, 0.40668490529060364, 0.03692089021205902, 0.051403168588876724, -0.13749796152114868, 0.04864603653550148, -0.2384946644306183, -0.15247507393360138, -0.31967318058013916, 0.05557970330119133, 0.1975456029176712, 0.04465103894472122, 0.11467431485652924, 0.3015516698360443, 0.04731551557779312, 0.15700913965702057, -0.00011640706361504272, -0.08212149143218994, 0.22991494834423065, -0.020129891112446785, 0.07947728037834167, -0.019034579396247864, 0.1309007853269577, -0.5154120326042175, -0.03750040382146835, -0.10425843298435211, 0.27347880601882935, 0.2135601043701172, 0.26597291231155396, 0.03253541514277458, 0.25440898537635803, -0.2947162985801697, 0.09862411767244339, 0.3221185505390167, -0.2970360815525055, 0.29969286918640137, -0.010349784046411514, -0.01474220771342516, 0.053282249718904495, 0.26293057203292847, 0.6999953985214233, -0.4993130564689636, 0.001270017120987177, 0.16061930358409882, 0.027350783348083496, 0.02736654505133629, 0.1041141077876091, -0.057549796998500824, -0.14754274487495422, -0.22595560550689697, -0.30797430872917175, 0.24689792096614838, -0.08192530274391174, -0.06607373803853989, -0.23695695400238037, -0.6513186097145081, -0.00732409069314599, 0.04993263632059097, -0.3508644104003906, 0.18694578111171722, -0.18846476078033447, -0.09649697691202164, -0.15846094489097595, -0.0536109134554863, -0.007652950473129749, 0.06648006290197372, 0.3757088780403137, 0.5052574276924133, 0.013429484330117702, 0.17369231581687927, 0.34670841693878174, 0.031212646514177322, -0.10651388019323349, -0.24471555650234222, 0.28535640239715576, 0.10458504408597946, -0.37025922536849976, -0.2722153067588806, 0.16172164678573608, -0.264308899641037, 0.2712613046169281, 0.023663951084017754, 0.17093297839164734, 0.021655254065990448, -0.06886740028858185, 0.1701311320066452, 0.409762442111969, 0.11251690983772278, -0.09519115835428238, -0.14553847908973694, 0.041094109416007996, -0.17762106657028198, 0.183108389377594, 0.20799759030342102, 0.148593008518219, -0.13576675951480865, -0.4656370282173157, -0.08216269314289093, -0.25629374384880066, 0.14932136237621307, -0.20054586231708527, -0.5421443581581116, -0.046839389950037, -0.23184815049171448, -0.00907486118376255, -0.11373674869537354, -0.2909092903137207, 0.14404307305812836, -0.11999654769897461, 0.2271474450826645, -0.16093449294567108, 0.07435347139835358, -0.05789923667907715, 0.0046850950457155704, -0.4462979733943939, -0.11514412611722946, 0.2692190110683441, 0.16773714125156403, 0.08348636329174042, 0.07069877535104752, -0.06737244129180908, 0.11798066645860672, 0.28367066383361816, -0.004396381322294474, -0.032980844378471375, 0.03960235044360161, 0.2211727648973465, -0.27689027786254883, -0.13525162637233734, 0.30507463216781616, -0.3997879922389984, -0.022525206208229065, -0.23434129357337952, -0.19428651034832, 0.18709298968315125, 0.02074490673840046, -0.1286899298429489, -0.22317001223564148, -0.5973213911056519, 0.6830008625984192, 0.06751839816570282, -0.0465230792760849, -0.15667150914669037, 0.17821703851222992, -0.19402453303337097, -0.17894059419631958, 0.011592624709010124, 0.0831245481967926, -0.061652570962905884, -0.34613537788391113, 0.0005171322845853865, -0.11840672791004181, 0.015426085330545902, 0.3973841071128845, -0.19246156513690948, 0.2865578234195709, 0.0944521352648735, 0.12006449699401855, 0.26640915870666504, -0.47417497634887695, -0.04938265308737755, 0.45079970359802246, -0.019094353541731834, 0.3506059944629669, 0.1672840416431427, 0.260222464799881, 0.002020171843469143, -0.17475317418575287, 0.23725645244121552, 0.8971966505050659, -0.4060542583465576, 0.08358108997344971, -0.13938410580158234, -0.42281419038772583, 0.5240344405174255, 0.3101251721382141, 0.10595036298036575, -0.3104817271232605, -0.06986049562692642, 0.046227287501096725, 0.13682259619235992, 0.09810630232095718, 0.12557436525821686, -0.05776258558034897, -0.4325706660747528, -0.0024138104636222124, -0.19192521274089813, -0.06833688914775848, -0.49648308753967285, 0.0032335480209439993, -0.15541492402553558, 0.26533636450767517, 0.3736354112625122, -0.05858921259641647, 0.3415442407131195, -0.3247835338115692, -0.0682685598731041, -0.31966671347618103, 0.0744401216506958, -0.09567278623580933, 0.03928658738732338, -0.31979095935821533, -0.062062617391347885, 0.3237813115119934, 0.12032078951597214, -0.14445091784000397, -0.4255874752998352, 0.19232188165187836, -0.03772478550672531, 0.08093035221099854, 0.01669709011912346, 0.5754867792129517, -0.29112866520881653, -0.12075131386518478, 0.10930075496435165, 0.24802981317043304, -0.2801348567008972, 0.21860845386981964, 0.2109755128622055, 0.26685547828674316, 0.25490429997444153, 0.12753544747829437, 0.20218472182750702, -0.12013930827379227, -0.09720608592033386, -0.2197658270597458, -0.03214003145694733, 0.44920313358306885, -0.09584518522024155, 0.4679339826107025, -0.0414217934012413, -0.10826475918292999, -0.2286720871925354, -0.04206325113773346, -0.20951545238494873, 0.3518671691417694, 0.3388046324253082, -0.15275807678699493, 0.3212239444255829, -0.08753753453493118, -0.3210725486278534, 0.15290340781211853, 0.09710418432950974, 0.07538940012454987, 0.05987544357776642, -0.015522033907473087, -0.04942814260721207, -0.29519227147102356, -0.16740113496780396, 0.34958791732788086, 0.7306102514266968, 0.2229880392551422, 0.19981910288333893, -0.0016514703165739775, -0.16649743914604187, -0.1618088036775589, -0.007260357029736042, 0.1925978660583496, -0.019021518528461456, 0.20632952451705933, 0.026077833026647568, 0.01070361863821745, 0.047556810081005096, -0.1425853967666626, 0.1281285285949707, -0.19433903694152832, -0.03902203589677811, 0.0012269613798707724, -0.21815888583660126, -0.6630980372428894, -0.3449258506298065, 0.0396687351167202, -0.2795931398868561, -0.22961638867855072, 0.09637446701526642, 0.006200673058629036, -0.20221714675426483, 0.09744233638048172, 0.25448665022850037, 0.6557927131652832, -0.4205455482006073, -0.12200149148702621, -0.011811735108494759, -0.24652469158172607, -0.18224358558654785, 0.0804419070482254, 0.47870874404907227, 0.19717256724834442, 0.42263197898864746, 0.25549381971359253, -0.07075680792331696, 0.08091460913419724, -0.3518061935901642, 0.025702068582177162, -0.2472161054611206, 0.24243247509002686, 0.11621170490980148, -0.23505264520645142, 0.19698330760002136, -0.33835023641586304, -0.03175106644630432, -0.11529546976089478, -0.07352647185325623, -0.18286475539207458, -0.022766729816794395, 0.04582692310214043, -0.21434496343135834, -0.14710207283496857, -0.6761796474456787, -0.42367273569107056, 0.3288804590702057, -0.25891825556755066, 0.08312951028347015, 0.06865046918392181, -0.24256187677383423, 0.1101025715470314, -0.12082325667142868, 0.020177919417619705, -0.10749740898609161, -0.13586559891700745, -0.10516980290412903, -0.3834722340106964, -0.16408580541610718, -0.44173663854599, -0.0516643226146698, 0.22958336770534515, 0.027340129017829895, -0.10455966740846634, -0.13729694485664368, 0.09618262946605682, -0.22930769622325897, 0.0718233659863472, 0.3079036474227905, 0.05160653591156006, -0.3560313284397125, 0.1333221048116684, -0.012046038173139095, -0.2183775156736374, 0.12953421473503113, 0.3778133690357208, 0.4913676977157593, 0.14920911192893982, 0.024773715063929558, 0.12455194443464279, 0.7539142370223999, 0.20179404318332672, 0.0599362887442112, 0.1063677966594696, 0.24882152676582336, -0.016944868490099907, 0.1358458697795868, -0.11611083894968033, 0.39773041009902954, -0.13664871454238892, 0.4263167381286621, 0.1709872931241989, 0.19975216686725616, -0.20308136940002441, -0.19826620817184448, -0.044721875339746475, -0.2616236209869385, -0.1858825385570526, 0.28222334384918213, -0.22426606714725494, 0.052601318806409836, -0.0805116817355156, -0.43985602259635925, -0.28615254163742065, -0.030419832095503807, -0.06067322567105293, 0.2741159498691559, -0.09199514985084534, 0.1816534847021103, -0.32036030292510986, -0.15978290140628815, -0.4608958959579468, 0.1459829956293106, 0.0887589082121849, 0.3017205595970154, -0.012457003816962242, -0.08777285367250443, -0.07358941435813904, 0.20257051289081573, 0.637381374835968, -0.44876304268836975, -0.16017574071884155, -0.0056172567419707775, 0.06069774553179741, -0.13317374885082245, -0.2124774307012558, -0.25151216983795166, -0.002701704390347004, 0.46255654096603394, 0.3910317122936249, -0.35132142901420593, -0.2920810580253601, 0.1962273269891739, -0.16269084811210632, -0.05591331794857979, -0.23227863013744354, -0.24113252758979797, -0.10254251211881638, -0.10813823342323303, 0.11773529648780823, 0.15848417580127716, 0.2345425933599472, -0.16833291947841644, 0.16883882880210876, -0.03416438773274422, 0.21754111349582672, 0.17560705542564392, 0.0023398499470204115, -0.08446909487247467, 0.2803269028663635, 0.11757080256938934, -0.018795203417539597, -0.05235309526324272, -0.18998847901821136, 0.08903001248836517, -0.16537916660308838, 0.19147637486457825, 0.321917861700058, -0.037185199558734894, 0.6493062376976013, 0.09719554334878922, 0.33552634716033936, 0.19747218489646912, -0.0571751669049263, 0.10656417906284332, -0.20644064247608185, 0.16744469106197357, 0.16320186853408813, 0.19573619961738586, -0.20392559468746185, -0.25002920627593994, 0.5089166164398193, 0.4079137444496155, -0.005740012973546982, 0.3332827389240265, -0.6443976163864136, -0.2598767876625061, 0.23563899099826813, 0.007103616371750832, 0.5431055426597595, 0.005898771341890097, 0.37823083996772766, -0.04366764798760414, -0.3120470941066742, 0.48079124093055725, -0.3299480676651001, 0.16691377758979797, -0.049857258796691895, -0.14280010759830475, 0.0768364816904068, -0.09651870280504227, -0.05870378389954567, 0.3432525098323822, -0.21262772381305695, 0.34454938769340515, 0.49613073468208313, -0.2248227298259735, -0.17989474534988403, 0.4790879786014557, -0.012489977292716503, -0.3797692358493805, 0.002641389612108469, 0.029900550842285156, -0.23055484890937805, 0.1625853031873703, -0.2158859670162201, -0.2061106115579605, 0.22479017078876495, 0.1379055678844452, -0.3079604506492615, -0.02638571709394455, -0.05604434385895729, 0.19251886010169983, 0.26101332902908325, -0.017162883654236794, 0.2552073895931244, -0.24698878824710846, -0.11747197806835175, -0.17436352372169495, -0.05939347669482231, 0.053465064615011215, -0.0917481929063797, 0.11164918541908264, 0.07159192115068436, -0.19615232944488525, 0.46401089429855347, -0.0757192000746727, -0.02119438536465168, -0.22428274154663086, 0.23087917268276215, -0.3635615110397339, -0.22265051305294037, 0.209263414144516, 0.14877144992351532, 0.23195630311965942, -0.3144288957118988, 0.574362576007843, 0.08713379502296448, 0.07795120030641556, 0.01097925752401352, 0.1568933129310608, -0.12796789407730103, 0.6329975724220276, -0.2941189706325531, -0.07013081759214401, -0.006362566724419594, -0.08990344405174255, 0.015857292339205742, -0.4475608170032501, -0.06189993396401405, 0.2750473618507385, 0.13799354434013367, 0.0288335382938385, 0.09607625007629395, 0.04632258415222168, -0.07321403920650482, 0.02442176267504692, -0.05657602474093437, -0.2628394663333893, 0.21956154704093933, 0.24102383852005005, 0.187313511967659, 0.2717297375202179, -0.06187177821993828, -0.2633689045906067, -0.10183091461658478, 0.46383535861968994, -0.01353496965020895, 0.24274973571300507, -0.09981811046600342, 0.4236372709274292, 0.4712509214878082, 0.5466606616973877, -0.20703797042369843, 0.07370541989803314, 0.1323155015707016, 0.09527555853128433, 0.39160585403442383, -0.05915417894721031, 0.0413052961230278, 0.4014023542404175, -0.1892261803150177, 0.08920954167842865, -0.3487747013568878, -0.12751232087612152, -0.229723259806633, 0.12747307121753693, 0.15703806281089783, 0.03752836585044861, 0.058767180889844894, -0.08815512806177139, -0.013388597406446934, -0.04495707526803017, 0.06431730091571808, -0.290706992149353, -0.07052236795425415, 0.08662550896406174, 0.01915234513580799, 0.014970428310334682, 0.12828856706619263, 0.4145224392414093, 0.07437929511070251, -0.08982572704553604, 0.32783427834510803, 0.13821066915988922, 0.15724590420722961, -0.1297171711921692, -0.21711233258247375, -0.08144588768482208, 0.2716141939163208, 0.1716533750295639, -0.16723214089870453, 0.09358753263950348, -0.21850980818271637, -0.08033720403909683, 0.18466350436210632, 0.2425161600112915, 0.07674838602542877, -0.1286870390176773, 0.17664794623851776, 0.07307993620634079, -0.023402484133839607, -0.0032811579294502735, 0.13657677173614502, 0.09947998076677322, 0.03915521129965782, 0.20603717863559723, 0.2117108702659607, 0.5047975778579712, -0.23348022997379303, 0.016525501385331154, 0.06762969493865967, -0.31354820728302, 0.09969226270914078, -0.13878470659255981, -0.016195017844438553, 0.02821831777691841, 0.22626429796218872, -0.08678179979324341, 0.15465791523456573, -0.30596089363098145, 0.32733526825904846, 0.24166035652160645, 0.035661324858665466, -0.02388034015893936, 0.12478064000606537, -0.03723413124680519, -0.2066289186477661, 0.3375938832759857, -0.029000358656048775, 0.31513121724128723, 0.009087351150810719, 0.7983555197715759, -0.2380809187889099, -0.39609867334365845, 0.15436695516109467, 0.4071565270423889, 0.008456393145024776, -0.2650139033794403, 0.10717979073524475, 0.13453198969364166, -0.26112693548202515, 0.20869117975234985, -0.04789560288190842, -0.09806633740663528, 0.5933933854103088, -0.06555313616991043, 0.17357710003852844, -0.161214679479599, -0.26063644886016846, 0.3475779891014099, 0.008399967104196548, -0.12483527511358261, 0.05821120738983154, -0.02960263006389141, -0.09513760358095169, -0.10379630327224731, 0.5061762928962708, 0.18733717501163483, 0.09599246829748154, -0.34532931447029114, -0.12983374297618866, 0.1432611346244812, 0.24537979066371918, -0.18993669748306274, 0.11646682769060135, 0.49969810247421265, -0.1996125876903534, 0.269757479429245, -0.025633104145526886, -0.0519576258957386, -0.04092944413423538, 0.04421588405966759, -0.4889703691005707, -0.41631603240966797, 0.4167273938655853, 0.033291857689619064, -0.09751678258180618, 0.05186871439218521, 0.3810831308364868, -0.6517298817634583, -0.19317041337490082, 0.30628693103790283, -0.18151625990867615, -0.08955486863851547, -0.24177540838718414, 0.059558261185884476, 0.03920280188322067, 0.7016071081161499, 0.29243525862693787, 0.32732295989990234, -0.3832719624042511, -0.08002867549657822, -0.5715482831001282, -0.17670080065727234, 0.01820162869989872, 0.00006076315185055137, -0.2943784296512604, 0.26411956548690796, -0.12518638372421265, 0.22162359952926636, -0.09406103938817978, 0.08814511448144913, 0.1416034698486328, 0.07060171663761139, -0.4327945411205292, 0.09113311022520065, 0.05096104368567467, 0.08235737681388855, 0.30425775051116943, -0.07882446050643921, -0.1336902678012848, 0.05160797759890556, -0.13522711396217346, -0.17467185854911804, -0.08309830725193024, 0.31040331721305847, 0.21442069113254547, 0.10870138555765152, 0.014889259822666645, 0.4607817232608795, 0.05151567980647087, -0.18575258553028107, -0.12474764138460159, -0.022355860099196434, -0.3448334336280823, -0.1339934766292572, -0.05004814267158508, 0.17086781561374664, -0.44279080629348755, 0.3610151410102844, 0.04574107751250267, -0.12132138758897781, 0.11556056141853333, 0.0341140553355217, -0.050387680530548096, -0.0671820119023323, 0.12176003307104111, 0.12575604021549225, 0.27909255027770996, 0.6992524266242981, 0.026893766596913338, 0.02020050212740898, -0.0898650735616684, -0.02438388764858246, -0.033063728362321854, -0.05258088558912277, -0.20456334948539734, -0.2189653366804123, 0.0958389937877655, -0.06080857664346695, 0.10704454779624939, -0.641090452671051, -0.38680359721183777, 0.05448317900300026, 0.06030590832233429, 0.058844730257987976, 0.6827636957168579, 0.07146316021680832, -0.2238551378250122, -0.11899236589670181, -0.37256866693496704, 0.041019272059202194, -0.29618898034095764, 0.15026085078716278, -0.3555028736591339 ]
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.
[ -0.10336733609437943, -0.43319427967071533, -0.02786404825747013, -0.014036222361028194, -0.27077317237854004, 0.08221625536680222, 0.1549767553806305, 0.0747741237282753, 0.32949647307395935, -0.09032976627349854, -0.1661020815372467, 0.26055216789245605, -0.197214737534523, 0.4019602835178375, 0.3530554175376892, -0.4648926556110382, -0.053618840873241425, -0.20055337250232697, -0.3664041757583618, 0.3340202569961548, 0.015646392479538918, 0.08674690127372742, -0.17946544289588928, 0.019072720780968666, -0.27196642756462097, -0.1105162501335144, -0.3949353098869324, 0.10839381814002991, 0.11043496429920197, -0.01392314676195383, 0.18760856986045837, 0.3442354202270508, 0.06541890650987625, 0.05401728302240372, -0.00011230398376937956, -0.12137140333652496, 0.2175675928592682, -0.09191815555095673, 0.13066019117832184, -0.0005818564095534384, 0.07844523340463638, -0.47050103545188904, -0.007732009515166283, -0.13589663803577423, 0.20232845842838287, 0.1704602837562561, 0.2904416024684906, 0.008269130252301693, 0.39299479126930237, -0.3308924436569214, 0.09726564586162567, 0.25307172536849976, -0.15323780477046967, 0.262188196182251, -0.13102519512176514, -0.020230945199728012, 0.10178595781326294, 0.2531667649745941, 0.635272204875946, -0.5347983837127686, -0.03683255985379219, 0.10696148127317429, 0.02565453201532364, 0.030744923278689384, 0.16349442303180695, -0.08338321000337601, -0.3191305696964264, -0.33498692512512207, -0.4021604061126709, 0.23699520528316498, -0.14883266389369965, -0.08393002301454544, -0.24520665407180786, -0.5854562520980835, 0.0016671117627993226, 0.01647736132144928, -0.3153265118598938, 0.11600285023450851, -0.20372840762138367, -0.034372635185718536, -0.25125688314437866, -0.1469614952802658, -0.043442998081445694, 0.1449488401412964, 0.3343847095966339, 0.5547493100166321, 0.08374859392642975, 0.16105502843856812, 0.36956119537353516, 0.057474516332149506, -0.12632232904434204, -0.31179746985435486, 0.29279690980911255, 0.10796769708395004, -0.36916112899780273, -0.355026513338089, 0.1392046958208084, -0.20298588275909424, 0.2949889004230499, 0.10299031436443329, 0.025053691118955612, 0.08368668705224991, -0.07621680200099945, 0.256097674369812, 0.37171226739883423, 0.09638246148824692, -0.08300160616636276, -0.06298986822366714, 0.002955428324639797, -0.06848466396331787, 0.14341950416564941, 0.17668943107128143, 0.12384472042322159, 0.003416585735976696, -0.47943544387817383, -0.12967407703399658, -0.25868988037109375, 0.07781004905700684, -0.1598018854856491, -0.4763876497745514, -0.0689307153224945, -0.11731711775064468, 0.02938370406627655, -0.08602409809827805, -0.2983449697494507, 0.11831442266702652, -0.09277258813381195, 0.2307395339012146, -0.08128512650728226, 0.037010543048381805, -0.04833849146962166, 0.07137133926153183, -0.43581312894821167, -0.058910783380270004, 0.26673299074172974, 0.12061816453933716, 0.10626685619354248, 0.16640625894069672, 0.013974243775010109, 0.15166285634040833, 0.3776470124721527, 0.016560202464461327, 0.0059494031593203545, -0.06492945551872253, 0.21329855918884277, -0.30310770869255066, 0.0006737092626281083, 0.17076903581619263, -0.36183303594589233, -0.145702064037323, -0.19035513699054718, -0.16295656561851501, 0.09470005333423615, 0.01727861911058426, -0.19424989819526672, -0.26328516006469727, -0.4468875527381897, 0.7332508563995361, 0.02684798650443554, -0.01713932491838932, -0.18028567731380463, 0.2606482207775116, -0.22355850040912628, -0.06396752595901489, 0.20860576629638672, 0.12475055456161499, -0.01690930873155594, -0.3514384627342224, -0.046797145158052444, -0.014810271561145782, 0.034520529210567474, 0.36155635118484497, -0.2216624766588211, 0.1277340203523636, 0.06172614544630051, 0.14611048996448517, 0.2180098295211792, -0.43226194381713867, -0.09815125167369843, 0.406034380197525, -0.016480790451169014, 0.3736610412597656, 0.11828519403934479, 0.19480210542678833, 0.12559884786605835, -0.19569654762744904, 0.2478085309267044, 1.0400981903076172, -0.4225594103336334, 0.0221931841224432, -0.16494359076023102, -0.3881131708621979, 0.5052052140235901, 0.362834095954895, 0.09461911767721176, -0.3206782341003418, -0.16311682760715485, -0.026616262272000313, 0.11877454072237015, 0.041435450315475464, 0.10451116412878036, -0.00923471711575985, -0.40524640679359436, -0.022486835718154907, -0.22144313156604767, -0.07816143333911896, -0.33089834451675415, 0.015293239615857601, -0.08741426467895508, 0.3274635672569275, 0.3405655324459076, -0.10581529140472412, 0.2896714508533478, -0.2984875738620758, -0.13058176636695862, -0.33861082792282104, 0.1144704520702362, -0.03951878473162651, 0.07346747070550919, -0.23656024038791656, -0.06718046218156815, 0.3598692715167999, 0.0890434980392456, -0.13852788507938385, -0.5096865892410278, 0.3090386390686035, -0.05740106850862503, 0.06364454329013824, -0.025107499212026596, 0.596556544303894, -0.19375938177108765, -0.1289076954126358, 0.14488095045089722, 0.22485223412513733, -0.3346215784549713, 0.19492986798286438, 0.33768755197525024, 0.17342931032180786, 0.2078784555196762, -0.08267901092767715, 0.17195716500282288, -0.029911773279309273, -0.09459754824638367, -0.14048907160758972, -0.03039674647152424, 0.5003798007965088, -0.11950568109750748, 0.4435095191001892, -0.018836060538887978, 0.01526191271841526, -0.2615099251270294, -0.09255741536617279, -0.21653777360916138, 0.35943976044654846, 0.3366411030292511, -0.1929524689912796, 0.36073410511016846, -0.05211378261446953, -0.359465628862381, 0.1487421989440918, 0.20372147858142853, 0.08420227468013763, 0.04603233188390732, 0.020731423050165176, -0.10165806114673615, -0.27896034717559814, -0.1018223911523819, 0.46157968044281006, 0.5798782706260681, 0.2789746820926666, 0.1910255253314972, 0.002939151832833886, -0.2511986792087555, -0.0801590159535408, 0.0353563018143177, 0.21586552262306213, -0.0584859699010849, 0.14054608345031738, 0.02464822120964527, 0.009310408495366573, 0.010679331608116627, -0.11853764206171036, 0.1468302458524704, -0.18777887523174286, -0.01490307692438364, 0.06336428970098495, -0.19838857650756836, -0.6163857579231262, -0.3052639663219452, -0.015071968547999859, -0.3515803813934326, -0.28699883818626404, 0.13720381259918213, -0.003792539704591036, -0.13886840641498566, 0.16575410962104797, 0.126996710896492, 0.5480082035064697, -0.4202442467212677, -0.1703033149242401, -0.023589329794049263, -0.23749899864196777, -0.12512168288230896, 0.10783178359270096, 0.4728633165359497, 0.26112523674964905, 0.43125179409980774, 0.2812897861003876, -0.036243412643671036, 0.12175816297531128, -0.3444490134716034, 0.040847983211278915, -0.293626070022583, 0.25403326749801636, 0.01180786918848753, -0.2061462253332138, 0.13463842868804932, -0.42947566509246826, -0.060518696904182434, -0.019378487020730972, -0.053349461406469345, -0.028743073344230652, -0.028221162036061287, 0.03259546309709549, -0.17166346311569214, -0.2008933573961258, -0.6588298678398132, -0.45629560947418213, 0.28352153301239014, -0.20028935372829437, 0.09491460025310516, 0.025160210207104683, -0.23395320773124695, 0.0898786187171936, -0.16715116798877716, -0.001951000769622624, -0.1276065856218338, -0.12098366767168045, 0.08369618654251099, -0.3582075536251068, -0.16718415915966034, -0.47265321016311646, -0.12426881492137909, 0.2418014407157898, -0.07113945484161377, -0.07343459874391556, -0.18778806924819946, 0.07646546512842178, -0.19489237666130066, 0.07409638166427612, 0.2653912305831909, 0.10548759251832962, -0.2660581171512604, 0.1051522046327591, 0.0029673236422240734, -0.13249745965003967, 0.2176908701658249, 0.3457123339176178, 0.4796205163002014, 0.15022993087768555, 0.11335109919309616, 0.1909075677394867, 0.6792360544204712, 0.32960832118988037, 0.007964294403791428, 0.2135126292705536, 0.3103102147579193, 0.024073606356978416, 0.19754570722579956, -0.2108502984046936, 0.3423008918762207, -0.11590617150068283, 0.45432111620903015, 0.10699512809515, 0.15114682912826538, -0.24436314404010773, -0.24342337250709534, 0.024328602477908134, -0.2064073383808136, -0.17803651094436646, 0.25939589738845825, -0.22649136185646057, 0.02016138657927513, -0.09770278632640839, -0.3649142384529114, -0.235571026802063, -0.05439330264925957, -0.03128239884972572, 0.22833727300167084, 0.011614075861871243, 0.17223306000232697, -0.3531283140182495, -0.10165928304195404, -0.4612160325050354, 0.20466767251491547, 0.057742889970541, 0.30177536606788635, -0.019986417144536972, -0.08469288796186447, -0.1292436569929123, 0.12709110975265503, 0.5680524706840515, -0.47282925248146057, -0.22094778716564178, -0.025382887572050095, 0.09025246649980545, -0.09211969375610352, -0.10572797060012817, -0.2197125107049942, -0.025454401969909668, 0.4199270009994507, 0.30624184012413025, -0.36397266387939453, -0.27792608737945557, 0.2682924270629883, -0.16463151574134827, -0.07220473140478134, -0.14969685673713684, -0.21267953515052795, -0.13832351565361023, -0.17088153958320618, 0.07758443057537079, 0.12965182960033417, 0.1933685690164566, -0.13969528675079346, 0.16703538596630096, -0.10889361053705215, 0.260311484336853, 0.18591710925102234, 0.05234652757644653, -0.027580788359045982, 0.19997915625572205, 0.17190340161323547, 0.053513213992118835, -0.10656803846359253, -0.23571929335594177, 0.12169519811868668, -0.03250058740377426, 0.07200001925230026, 0.2976096272468567, 0.039041463285684586, 0.6938976049423218, 0.04260185360908508, 0.36705583333969116, 0.225030317902565, -0.06475920975208282, 0.08747003227472305, -0.21818473935127258, 0.1336088925600052, 0.22997228801250458, 0.20788529515266418, -0.16687694191932678, -0.22111362218856812, 0.5431410074234009, 0.29854562878608704, -0.009611457586288452, 0.3315662741661072, -0.5390050411224365, -0.25116589665412903, 0.2967572510242462, 0.0401647575199604, 0.6300409436225891, -0.007101005874574184, 0.2938927114009857, -0.03298943489789963, -0.33714497089385986, 0.5710723996162415, -0.33072757720947266, 0.19199493527412415, -0.14019998908042908, -0.24169902503490448, -0.00279090809635818, -0.08570387959480286, -0.07936576008796692, 0.276381254196167, -0.3371981382369995, 0.36917927861213684, 0.4814932942390442, -0.24805855751037598, -0.10458763688802719, 0.5050843954086304, -0.05645809695124626, -0.3726613223552704, -0.11116617918014526, 0.06115476042032242, -0.27586671710014343, 0.0950934961438179, -0.11868719011545181, -0.3008018136024475, 0.2708379924297333, 0.12162400037050247, -0.24231120944023132, 0.16438506543636322, -0.05043192207813263, 0.1845230907201767, 0.25303915143013, 0.008703055791556835, 0.2706410884857178, -0.16247984766960144, -0.08966227620840073, -0.11695247888565063, -0.07923763990402222, 0.11518257111310959, -0.14977097511291504, 0.01284440141171217, 0.11990488320589066, -0.20245912671089172, 0.48927128314971924, -0.00957292690873146, -0.1359957605600357, -0.21550925076007843, 0.31131091713905334, -0.3876410126686096, -0.17851808667182922, 0.10874956846237183, 0.18113833665847778, 0.22955574095249176, -0.3597268760204315, 0.5780022144317627, 0.13180939853191376, 0.027300750836730003, 0.030787160620093346, 0.1418282836675644, -0.25124698877334595, 0.5130022168159485, -0.2759886384010315, -0.03809617459774017, 0.042088303714990616, 0.0073547326028347015, -0.08804506808519363, -0.3510380685329437, -0.0318845696747303, 0.2833532691001892, 0.12923067808151245, 0.021556643769145012, 0.06551950424909592, 0.17462100088596344, -0.07396021485328674, -0.06540028750896454, -0.13226749002933502, -0.3194485306739807, 0.15650281310081482, 0.07593650370836258, 0.19611234962940216, 0.16132000088691711, -0.0762888565659523, -0.2127532660961151, -0.21801777184009552, 0.5112913250923157, 0.021172985434532166, 0.2578124701976776, -0.18163959681987762, 0.30822038650512695, 0.3887071907520294, 0.5697491765022278, -0.25568702816963196, 0.11726099252700806, 0.055237434804439545, 0.06974649429321289, 0.46103331446647644, -0.2023731768131256, 0.043634917587041855, 0.2892819046974182, -0.1484197974205017, 0.06235464662313461, -0.3408699035644531, -0.1428142935037613, -0.23465710878372192, 0.0988052487373352, 0.15117916464805603, -0.019356878474354744, 0.009457581676542759, -0.06798810511827469, -0.09655863046646118, -0.11638103425502777, 0.013011409901082516, -0.4335089325904846, -0.06818936765193939, 0.11059032380580902, -0.03718027472496033, -0.04546325281262398, 0.07459371536970139, 0.23765970766544342, 0.001246116473339498, -0.020071759819984436, 0.3473859429359436, 0.13781437277793884, 0.13269740343093872, -0.1667909175157547, -0.27622753381729126, -0.01261456124484539, 0.27358487248420715, 0.16343863308429718, -0.07122226059436798, 0.10352068394422531, -0.24374085664749146, -0.08736567944288254, 0.2628614902496338, 0.2063620537519455, 0.07711703330278397, -0.17260369658470154, 0.15144763886928558, 0.112718366086483, -0.07894748449325562, -0.0010419770842418075, 0.20892125368118286, 0.07174254953861237, 0.0864514634013176, 0.1702318638563156, 0.15795141458511353, 0.4282064139842987, -0.24424396455287933, -0.007555025164037943, 0.10386925935745239, -0.3436979651451111, 0.14428013563156128, -0.07468446344137192, -0.016898952424526215, -0.05834311991930008, 0.11644052714109421, -0.042675282806158066, 0.22477839887142181, -0.2821715474128723, 0.2502272129058838, 0.30582138895988464, 0.009968416765332222, -0.042123593389987946, 0.20314425230026245, 0.001803387189283967, -0.2245405912399292, 0.3329145312309265, 0.08519972860813141, 0.38552263379096985, 0.04895404353737831, 0.8560096025466919, -0.24788802862167358, -0.465786874294281, 0.037296075373888016, 0.4859018921852112, 0.04150897637009621, -0.2708488404750824, 0.14337965846061707, 0.09598027914762497, -0.4118724763393402, 0.16141976416110992, 0.07498534768819809, -0.18442367017269135, 0.6622331142425537, -0.058488182723522186, 0.16207745671272278, -0.19061194360256195, -0.17011423408985138, 0.3675386309623718, 0.14315122365951538, -0.136328786611557, 0.07419107109308243, 0.13879211246967316, -0.10969682782888412, -0.14287380874156952, 0.5203856229782104, 0.2531953752040863, 0.03317766636610031, -0.4345838129520416, -0.022072376683354378, 0.1602121889591217, 0.16314110159873962, -0.10989035665988922, 0.12353618443012238, 0.4092104434967041, -0.17104357481002808, 0.26817190647125244, 0.01345693226903677, -0.06439215689897537, -0.17000062763690948, -0.0010199063690379262, -0.43804988265037537, -0.3881176710128784, 0.4196392893791199, 0.11488617211580276, -0.05141700059175491, 0.019021103158593178, 0.2460399568080902, -0.6877139806747437, -0.23156841099262238, 0.3309122622013092, -0.13534867763519287, -0.10280811041593552, -0.24208922684192657, 0.07230068743228912, -0.01539039146155119, 0.6274809837341309, 0.2935217618942261, 0.33659449219703674, -0.4087730050086975, -0.15738630294799805, -0.5566298365592957, -0.14260302484035492, 0.10983393341302872, -0.15623551607131958, -0.28286755084991455, 0.2381749302148819, -0.09586019068956375, 0.23567363619804382, -0.21898680925369263, 0.133131742477417, 0.21304932236671448, 0.09838525950908661, -0.4491216242313385, 0.1221679151058197, 0.10001187771558762, 0.044281255453825, 0.27925246953964233, -0.09820941090583801, -0.17155346274375916, 0.053720872849226, -0.10322346538305283, -0.00814896635711193, -0.01687905564904213, 0.24552489817142487, 0.09913580864667892, 0.12277598679065704, 0.06725391745567322, 0.523567795753479, -0.03352418914437294, -0.21914944052696228, 0.005622683092951775, -0.04023357108235359, -0.4122084677219391, -0.13063295185565948, -0.07771560549736023, 0.24383801221847534, -0.34682729840278625, 0.26382026076316833, -0.04028180614113808, -0.12499427050352097, 0.12921619415283203, 0.0319637805223465, -0.12255164980888367, -0.05809992179274559, 0.05785607174038887, 0.21735639870166779, 0.2185802012681961, 0.7121257781982422, 0.038261447101831436, 0.10666997730731964, -0.028916291892528534, -0.024187935516238213, 0.0697459951043129, -0.08190598338842392, -0.2243690937757492, -0.20493775606155396, 0.02550661563873291, 0.06479815393686295, 0.07541289180517197, -0.6441549062728882, -0.32651713490486145, 0.011607208289206028, 0.0044605485163629055, 0.04752209782600403, 0.7026141881942749, 0.11499857902526855, -0.14029523730278015, -0.13398128747940063, -0.2580876052379608, -0.0005196412676014006, -0.329835444688797, 0.20005005598068237, -0.30358049273490906 ]
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).
[ -0.06431587040424347, -0.46592244505882263, -0.08466391265392303, -0.031062781810760498, -0.293763130903244, 0.014267781749367714, 0.27487194538116455, -0.0015316011849790812, 0.3273957669734955, -0.08819643408060074, -0.14839482307434082, 0.10962057113647461, -0.16663450002670288, 0.43982648849487305, 0.3444589376449585, -0.36647915840148926, 0.03876926377415657, -0.2969401776790619, -0.27085423469543457, 0.3853916823863983, 0.038084227591753006, 0.041931647807359695, -0.13354302942752838, 0.0882256031036377, -0.28080931305885315, -0.2636696994304657, -0.38970762491226196, 0.034434594213962555, 0.13997597992420197, 0.05134959518909454, 0.23408912122249603, 0.30755189061164856, 0.04861689358949661, 0.2791938781738281, -0.00011687174992403015, -0.08377621322870255, 0.23081006109714508, -0.019136475399136543, 0.11680866777896881, -0.0018153757555410266, 0.15380094945430756, -0.4783226251602173, 0.005464197602123022, -0.11083629727363586, 0.2739706337451935, 0.15680420398712158, 0.25918132066726685, 0.10646320879459381, 0.2886654734611511, -0.3447948098182678, 0.09772258996963501, 0.27468129992485046, -0.21083857119083405, 0.22634682059288025, -0.1184612587094307, -0.014192931354045868, 0.07735446095466614, 0.26061055064201355, 0.7055600881576538, -0.5410451889038086, -0.07276538759469986, 0.07944449037313461, 0.04223458468914032, 0.0119698541238904, 0.10907929390668869, -0.17149630188941956, -0.2628779709339142, -0.2975196838378906, -0.32956618070602417, 0.2208573967218399, -0.0858788937330246, -0.09882543981075287, -0.2601616084575653, -0.6264213919639587, -0.036425214260816574, 0.03155345469713211, -0.3546120226383209, 0.16949917376041412, -0.20453102886676788, -0.12913064658641815, -0.23460280895233154, -0.10024102032184601, -0.02642105519771576, 0.07539790868759155, 0.34746405482292175, 0.5395691990852356, 0.041969940066337585, 0.12722018361091614, 0.39268097281455994, -0.03826475143432617, -0.08897179365158081, -0.2737717032432556, 0.25671160221099854, 0.1343371868133545, -0.39920148253440857, -0.29389309883117676, 0.15462535619735718, -0.26450997591018677, 0.3152872920036316, 0.0821952298283577, 0.0817451924085617, 0.05161217600107193, -0.1418321281671524, 0.19767308235168457, 0.4014681279659271, 0.12532830238342285, -0.09977924078702927, -0.11220315098762512, 0.04722500592470169, -0.1615312248468399, 0.21224670112133026, 0.18241600692272186, 0.1426578313112259, -0.1386057734489441, -0.5396937131881714, -0.167190283536911, -0.2762545645236969, 0.14143991470336914, -0.18073825538158417, -0.5744891166687012, -0.020447004586458206, -0.1508355438709259, 0.020509544759988785, -0.09436818957328796, -0.27212557196617126, 0.0960921123623848, -0.11833245307207108, 0.2690684199333191, -0.1600545048713684, 0.07366286963224411, -0.05052119866013527, 0.04711301997303963, -0.470763236284256, -0.10468701273202896, 0.2876484990119934, 0.17828714847564697, 0.0942668467760086, 0.18709884583950043, -0.031958162784576416, 0.08593133091926575, 0.3025982677936554, -0.06191815808415413, -0.05103863775730133, 0.03935518488287926, 0.2460584044456482, -0.26802653074264526, -0.08793412148952484, 0.2486288696527481, -0.37916046380996704, -0.059033676981925964, -0.23170897364616394, -0.15932095050811768, 0.15850932896137238, -0.002280682558193803, -0.14439606666564941, -0.25043046474456787, -0.5782619118690491, 0.6736924648284912, 0.062367670238018036, -0.05668472498655319, -0.16031412780284882, 0.22042061388492584, -0.1436547040939331, -0.18698126077651978, 0.05995578318834305, 0.018210258334875107, -0.03283465653657913, -0.33299553394317627, 0.03664229437708855, -0.035241302102804184, 0.0035876769106835127, 0.4364962577819824, -0.18516071140766144, 0.2895748019218445, 0.08100868761539459, 0.12182892113924026, 0.26645246148109436, -0.4424838423728943, -0.011515277437865734, 0.4173332452774048, 0.05032644048333168, 0.33077776432037354, 0.12340619415044785, 0.2318916618824005, 0.04421757161617279, -0.20737791061401367, 0.2358897179365158, 0.9741354584693909, -0.37452685832977295, 0.09668085724115372, -0.05804974585771561, -0.465802401304245, 0.441190630197525, 0.22587624192237854, 0.08976144343614578, -0.2925751507282257, -0.07390519231557846, -0.054094232618808746, 0.05737683176994324, 0.08964520692825317, 0.09594698995351791, 0.04774607717990875, -0.4209631383419037, 0.09648653864860535, -0.18344923853874207, -0.0505203977227211, -0.44608911871910095, -0.010563768446445465, -0.08277756720781326, 0.19561372697353363, 0.37465617060661316, -0.09083868563175201, 0.3362385630607605, -0.32630735635757446, -0.09256407618522644, -0.34133216738700867, 0.06611587852239609, -0.1217472180724144, 0.0976746678352356, -0.30773481726646423, 0.0007403549971058965, 0.3354277014732361, 0.09459517896175385, -0.10449416190385818, -0.497672975063324, 0.1759321689605713, -0.036461103707551956, -0.008629939518868923, 0.05814779922366142, 0.5408447980880737, -0.19120100140571594, -0.08414333313703537, 0.10430626571178436, 0.28971898555755615, -0.2685343027114868, 0.23165223002433777, 0.31108567118644714, 0.26073792576789856, 0.21135930716991425, 0.11839224398136139, 0.20557259023189545, -0.10977377742528915, -0.09943708032369614, -0.22492994368076324, -0.10347151011228561, 0.4607519805431366, -0.13459333777427673, 0.5010107159614563, -0.029697410762310028, -0.07168646156787872, -0.18867164850234985, -0.017303381115198135, -0.16132676601409912, 0.4569726884365082, 0.3395250737667084, -0.1707572042942047, 0.3268982768058777, -0.09833278506994247, -0.35604262351989746, 0.05383467674255371, 0.022897645831108093, 0.05386388301849365, 0.08113547414541245, -0.008003332652151585, -0.03082805871963501, -0.29537585377693176, -0.12621480226516724, 0.3653465807437897, 0.6546182632446289, 0.2216179519891739, 0.21824203431606293, -0.06149447336792946, -0.22549642622470856, -0.1522151529788971, -0.02587454579770565, 0.20214490592479706, 0.0066399830393493176, 0.16049885749816895, 0.02629823237657547, -0.005736880004405975, 0.08471235632896423, -0.10533811897039413, 0.11840743571519852, -0.22522178292274475, -0.08519274741411209, 0.005405190866440535, -0.22219160199165344, -0.6311858892440796, -0.3055891990661621, 0.015079915523529053, -0.30497029423713684, -0.2313874214887619, 0.03909817337989807, 0.07374665886163712, -0.11424057185649872, 0.08188299834728241, 0.21158647537231445, 0.6373735666275024, -0.43466493487358093, -0.1099824607372284, -0.03147651255130768, -0.24741141498088837, -0.1409846991300583, 0.06759544461965561, 0.4684334695339203, 0.20970197021961212, 0.4605120122432709, 0.32193660736083984, -0.1092420220375061, 0.1142004132270813, -0.3225853145122528, 0.026337917894124985, -0.20052911341190338, 0.24238158762454987, 0.04370403289794922, -0.1586240530014038, 0.17650482058525085, -0.33668431639671326, -0.07314218580722809, -0.061084628105163574, -0.052852507680654526, -0.17358389496803284, -0.07025722414255142, 0.056488268077373505, -0.1488358974456787, -0.10039733350276947, -0.6727859377861023, -0.39646661281585693, 0.3388640880584717, -0.2569657564163208, 0.06550877541303635, 0.025421129539608955, -0.1942608654499054, 0.07539032399654388, -0.14416258037090302, -0.017736613750457764, -0.07496708631515503, -0.12288626283407211, -0.02914421819150448, -0.4028763473033905, -0.1595146805047989, -0.37320852279663086, -0.02167394943535328, 0.19834546744823456, 0.01705765537917614, -0.06768278777599335, -0.09868618845939636, 0.02911689691245556, -0.1789454072713852, 0.1095881536602974, 0.348138689994812, 0.08324074745178223, -0.3999960720539093, 0.1514010727405548, -0.05563667044043541, -0.24908708035945892, 0.14666150510311127, 0.2859279215335846, 0.49135735630989075, 0.21992047131061554, 0.008941972628235817, 0.15517908334732056, 0.7662395238876343, 0.2730219066143036, 0.023622997105121613, 0.09859321266412735, 0.2874361574649811, 0.02550465241074562, 0.1275710165500641, -0.11887825280427933, 0.33744630217552185, -0.17567263543605804, 0.43661990761756897, 0.06794393062591553, 0.20402318239212036, -0.16058138012886047, -0.23380239307880402, -0.03952769190073013, -0.32576459646224976, -0.17143838107585907, 0.2507261037826538, -0.25176531076431274, 0.07811984419822693, -0.14358218014240265, -0.4316743016242981, -0.29271435737609863, -0.01696101948618889, -0.06760821491479874, 0.3017238676548004, -0.08189839869737625, 0.20029890537261963, -0.3350658714771271, -0.16427896916866302, -0.3893592059612274, 0.17401213943958282, -0.002374152420088649, 0.29987451434135437, -0.03779637813568115, -0.014264467172324657, -0.07898098975419998, 0.17510314285755157, 0.5862995982170105, -0.4357357323169708, -0.14293219149112701, -0.014258159324526787, 0.11169269680976868, -0.14096778631210327, -0.19628870487213135, -0.2515057921409607, 0.05307263135910034, 0.507483959197998, 0.3795272409915924, -0.3611525297164917, -0.31687700748443604, 0.19735871255397797, -0.19273711740970612, -0.09322507679462433, -0.22613458335399628, -0.21406161785125732, -0.16825449466705322, -0.07910854369401932, 0.14117713272571564, 0.139517143368721, 0.1591918021440506, -0.23411113023757935, 0.14947561919689178, -0.0783836618065834, 0.2822117507457733, 0.1764059066772461, -0.01885443553328514, -0.0787944495677948, 0.2480190098285675, 0.1028505191206932, 0.01776975207030773, -0.0868244394659996, -0.16271831095218658, 0.03729325160384178, -0.13130171597003937, 0.15576718747615814, 0.29673823714256287, 0.013812568970024586, 0.6570737957954407, 0.10724861174821854, 0.4018983840942383, 0.2005971223115921, -0.11568356305360794, 0.05698535963892937, -0.19889219105243683, 0.1270313709974289, 0.23611541092395782, 0.24626265466213226, -0.22283613681793213, -0.2472972720861435, 0.5878699421882629, 0.38168710470199585, -0.05445447191596031, 0.40503573417663574, -0.6244819760322571, -0.2605520188808441, 0.23180074989795685, 0.024833189323544502, 0.5990921258926392, -0.013116811402142048, 0.2996212840080261, -0.0794404074549675, -0.35769349336624146, 0.45935380458831787, -0.3115876615047455, 0.16311439871788025, -0.07900001853704453, -0.08968082070350647, 0.07022051513195038, -0.13893496990203857, 0.01839802972972393, 0.2944928705692291, -0.23567348718643188, 0.3794519901275635, 0.41210439801216125, -0.25065794587135315, -0.15530171990394592, 0.47581747174263, -0.034751102328300476, -0.3831046521663666, 0.05561349540948868, 0.024736687541007996, -0.2695632576942444, 0.17046712338924408, -0.21481549739837646, -0.2006016969680786, 0.2319072037935257, 0.1601427048444748, -0.35234180092811584, 0.023304594680666924, -0.016125433146953583, 0.22780896723270416, 0.2794570028781891, -0.04916458576917648, 0.2548295557498932, -0.16728545725345612, -0.10953829437494278, -0.12556108832359314, -0.03871754929423332, 0.10054513067007065, -0.11218106001615524, 0.03601542487740517, 0.03847881406545639, -0.17122693359851837, 0.43994632363319397, -0.05404620245099068, -0.08606567233800888, -0.23325935006141663, 0.223176971077919, -0.2976759374141693, -0.296324759721756, 0.13378047943115234, 0.12846267223358154, 0.15110740065574646, -0.31660380959510803, 0.5542107820510864, 0.1613525152206421, 0.09315459430217743, 0.007156998850405216, 0.1634625941514969, -0.12557081878185272, 0.6486729979515076, -0.3296874463558197, -0.010702726431190968, 0.031420037150382996, -0.09077972918748856, 0.006326587870717049, -0.4751659333705902, -0.05922213941812515, 0.2901131808757782, 0.11815205961465836, 0.04533170908689499, 0.092624731361866, 0.09316766262054443, 0.02672472596168518, -0.01084193866699934, -0.10102423280477524, -0.2596002519130707, 0.1727825105190277, 0.16765479743480682, 0.1894177347421646, 0.1972038745880127, -0.05612848326563835, -0.2705429196357727, -0.10723843425512314, 0.4969041347503662, -0.019268669188022614, 0.21876296401023865, -0.11857294291257858, 0.33743011951446533, 0.49458253383636475, 0.599609911441803, -0.1953844279050827, 0.09189865738153458, 0.15111702680587769, 0.09730381518602371, 0.4130695164203644, -0.08511385321617126, 0.01588011160492897, 0.3530912399291992, -0.20409706234931946, 0.037824105471372604, -0.2977035641670227, -0.11826495081186295, -0.19182096421718597, 0.11325029283761978, 0.1212540715932846, 0.04695969074964523, 0.028699742630124092, -0.1480221152305603, 0.009298655204474926, -0.05196166783571243, 0.06106181442737579, -0.2568272054195404, -0.07784241437911987, 0.11179006844758987, 0.0003042785683646798, 0.08506160974502563, 0.14502768218517303, 0.37398847937583923, 0.04396071285009384, -0.06393072754144669, 0.34793418645858765, 0.10780230164527893, 0.1781652867794037, -0.1582489162683487, -0.2107108235359192, -0.05424998328089714, 0.29739946126937866, 0.142969012260437, -0.17916160821914673, 0.0863334983587265, -0.18172824382781982, -0.12257353216409683, 0.14351682364940643, 0.286325603723526, 0.0899973139166832, -0.11031533777713776, 0.2158232033252716, 0.06141496077179909, 0.006081878673285246, 0.002305209869518876, 0.17192623019218445, 0.10082147270441055, -0.004941380117088556, 0.19742067158222198, 0.1723507046699524, 0.4390968382358551, -0.27887046337127686, 0.029636671766638756, 0.09111112356185913, -0.34259936213493347, 0.09730216860771179, -0.13354699313640594, -0.0017446267884224653, -0.021511640399694443, 0.2189338505268097, 0.0036811528261750937, 0.14514662325382233, -0.31865283846855164, 0.3492978513240814, 0.2836127281188965, 0.041865188628435135, -0.03126326575875282, 0.16973207890987396, -0.1356528103351593, -0.23886209726333618, 0.28188613057136536, -0.0074556125327944756, 0.2914676070213318, 0.006257299333810806, 0.8371667265892029, -0.2760741412639618, -0.4076002240180969, 0.11187376081943512, 0.4058810770511627, -0.0084144938737154, -0.24195267260074615, 0.05891663581132889, 0.13566073775291443, -0.21899594366550446, 0.18768540024757385, 0.005946444347500801, -0.15572099387645721, 0.6309083104133606, -0.08685927838087082, 0.15201377868652344, -0.13946126401424408, -0.1847788244485855, 0.3204898238182068, 0.06872235983610153, -0.04573502019047737, 0.10238151997327805, -0.0006035816040821373, -0.08211366087198257, -0.1377708613872528, 0.5545239448547363, 0.27132344245910645, 0.050693996250629425, -0.3948327600955963, -0.05802099406719208, 0.12167682498693466, 0.18451707065105438, -0.09898970276117325, 0.12961961328983307, 0.4572368562221527, -0.2189447283744812, 0.21530210971832275, -0.018487663939595222, -0.033718645572662354, -0.06840480118989944, 0.03277238458395004, -0.4699128270149231, -0.3582218289375305, 0.44198623299598694, 0.03924171254038811, -0.15338797867298126, 0.09869464486837387, 0.35182297229766846, -0.6634209156036377, -0.22207516431808472, 0.3930754065513611, -0.15513265132904053, -0.1185300201177597, -0.1978643238544464, 0.057037923485040665, 0.12376957386732101, 0.650640606880188, 0.3044005334377289, 0.3155849277973175, -0.33722469210624695, -0.15077972412109375, -0.5069020986557007, -0.21034426987171173, -0.015344711020588875, -0.04891404137015343, -0.31534746289253235, 0.21218302845954895, -0.13832111656665802, 0.16271927952766418, -0.12016802281141281, 0.09354369342327118, 0.22755928337574005, 0.18092063069343567, -0.460758239030838, 0.12454532086849213, 0.09379174560308456, 0.10597728937864304, 0.27192190289497375, -0.08295982331037521, -0.15479180216789246, 0.0831175297498703, -0.13035976886749268, -0.06441977620124817, -0.0050103385001420975, 0.3701482117176056, 0.1331891268491745, 0.08900205045938492, 0.06753271073102951, 0.5516718626022339, 0.08134463429450989, -0.16196119785308838, -0.06577114760875702, 0.03485569357872009, -0.3513205945491791, -0.13600315153598785, -0.08042477816343307, 0.11832907050848007, -0.47275206446647644, 0.3081987798213959, 0.06809177249670029, -0.15013962984085083, 0.0518043115735054, 0.016798637807369232, -0.035492878407239914, -0.10962068289518356, 0.10881708562374115, 0.1547781229019165, 0.34218040108680725, 0.7173804640769958, 0.0134887071326375, 0.05112417787313461, -0.11207682639360428, 0.01637897454202175, -0.041447121649980545, -0.10039996355772018, -0.2099619358778, -0.24219323694705963, 0.09355347603559494, -0.04354920610785484, 0.12603607773780823, -0.5950585007667542, -0.3449786305427551, 0.0035272808745503426, 0.05697784200310707, 0.09147065132856369, 0.692209005355835, 0.08021365851163864, -0.19983097910881042, -0.12055398523807526, -0.36738693714141846, 0.0513145811855793, -0.2942933738231659, 0.09561268240213394, -0.31406649947166443 ]
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.
[ -0.03330754488706589, -0.5136857628822327, -0.027779709547758102, 0.06664903461933136, -0.25032734870910645, 0.04263020679354668, 0.17539633810520172, 0.060455191880464554, 0.20672662556171417, -0.1286959946155548, -0.2596334218978882, 0.1723332405090332, -0.16145706176757812, 0.5526629686355591, 0.36890390515327454, -0.4282805323600769, -0.03364429250359535, -0.22553154826164246, -0.3311319649219513, 0.35862815380096436, 0.1625538468360901, 0.12674084305763245, -0.13107387721538544, 0.011625549755990505, -0.3748590052127838, -0.05233386531472206, -0.32901275157928467, -0.014158841222524643, 0.07128332555294037, 0.15654471516609192, 0.1405659168958664, 0.34378743171691895, -0.04573135823011398, 0.30428168177604675, -0.00012068864452885464, -0.00085714046144858, 0.06160658970475197, -0.12618549168109894, 0.11212602257728577, -0.03864641487598419, 0.2874490022659302, -0.3399030566215515, 0.05313091352581978, -0.17077599465847015, 0.13239780068397522, 0.1373211145401001, 0.23634591698646545, 0.07320460677146912, 0.42596885561943054, -0.22620750963687897, 0.04279351234436035, 0.4667571485042572, -0.28901153802871704, 0.2969694137573242, 0.0031429913360625505, -0.03748799487948418, -0.033627189695835114, 0.21709585189819336, 0.7237202525138855, -0.40428298711776733, 0.06341966241598129, 0.040229760110378265, -0.013890853151679039, -0.011367504484951496, 0.21762849390506744, -0.007717309053987265, -0.3298623859882355, -0.29819416999816895, -0.3776232898235321, 0.24901816248893738, 0.003808068810030818, -0.18128229677677155, -0.22731953859329224, -0.5786260962486267, -0.052298106253147125, -0.008305580355226994, -0.3058779239654541, 0.028880739584565163, -0.2804422080516815, -0.07985898107290268, -0.1518014669418335, -0.1365441530942917, -0.11704957485198975, 0.1804959774017334, 0.3751618266105652, 0.4616806209087372, 0.0698355883359909, 0.17935670912265778, 0.2523778975009918, 0.13546830415725708, -0.05440978705883026, -0.3006082773208618, 0.2920211851596832, 0.18487870693206787, -0.3568213880062103, -0.3436088263988495, 0.07528000324964523, -0.38873907923698425, 0.3567281663417816, 0.12050748616456985, 0.16586092114448547, 0.010490522719919682, -0.14146852493286133, 0.11952178925275803, 0.3928675651550293, -0.09637751430273056, -0.16700053215026855, -0.11347848176956177, -0.025100113824009895, -0.2513059377670288, 0.07018406689167023, 0.09315341711044312, 0.16992315649986267, -0.10998518019914627, -0.3625374436378479, -0.12453193962574005, -0.1382128894329071, 0.09857959300279617, -0.032947491854429245, -0.522433876991272, -0.215951070189476, -0.1834096759557724, -0.0349707156419754, -0.08714616298675537, -0.2840886414051056, 0.04984014108777046, -0.0025753506924957037, 0.21522334218025208, -0.2706935703754425, 0.041582584381103516, -0.0405673012137413, 0.02727462351322174, -0.3746244013309479, -0.04783236235380173, 0.2949211597442627, 0.05317302048206329, 0.11462042480707169, 0.16285604238510132, 0.19290176033973694, 0.12154608219861984, 0.4489947259426117, -0.028123242780566216, -0.010045209899544716, -0.023936811834573746, 0.14568489789962769, -0.21127969026565552, -0.019119415432214737, 0.3398950397968292, -0.3035699427127838, -0.06994908303022385, -0.1554131805896759, -0.1905195713043213, 0.09611625969409943, -0.05094453692436218, -0.20126138627529144, -0.3853228688240051, -0.4683864414691925, 0.6835903525352478, 0.06396577507257462, -0.09593910723924637, -0.20830392837524414, 0.20528607070446014, -0.21099446713924408, -0.08647185564041138, 0.16142570972442627, 0.06823400408029556, -0.013201345689594746, -0.37244027853012085, 0.04198436811566353, -0.10053999722003937, 0.008564993739128113, 0.3029564619064331, -0.15408708155155182, 0.2219998985528946, 0.08476704359054565, 0.11213966459035873, 0.3399461507797241, -0.4791812598705292, -0.009698682464659214, 0.35266876220703125, -0.042694222182035446, 0.29657700657844543, 0.14705240726470947, 0.2737155258655548, -0.0726073756814003, -0.15075425803661346, 0.1296713799238205, 0.9934559464454651, -0.3805124759674072, -0.017214391380548477, -0.07959849387407303, -0.36450186371803284, 0.4801662564277649, 0.35858508944511414, 0.19760900735855103, -0.36574697494506836, -0.18771949410438538, -0.005163798574358225, 0.08525040000677109, 0.006785111036151648, 0.08330795913934708, 0.024008363485336304, -0.23442047834396362, 0.10306406021118164, -0.1965169459581375, -0.23357383906841278, -0.4084451198577881, 0.04143530875444412, -0.050924383103847504, 0.3780595362186432, 0.3517451286315918, -0.05343768745660782, 0.2586512565612793, -0.27821582555770874, -0.12434390187263489, -0.31017500162124634, 0.034565940499305725, -0.023269107565283775, 0.0006010318757034838, -0.2645055949687958, -0.09321192651987076, 0.26055610179901123, 0.03617451339960098, -0.09498154371976852, -0.4310053288936615, 0.42751801013946533, 0.022629091516137123, -0.034638550132513046, -0.03160133212804794, 0.5317654013633728, -0.2460523396730423, -0.1196654662489891, 0.10470791161060333, 0.21626342833042145, -0.2052464783191681, 0.15738530457019806, 0.22795146703720093, 0.39875829219818115, 0.2398860603570938, -0.0832396000623703, 0.05731631815433502, -0.051717858761548996, -0.06435872614383698, -0.21896956861019135, -0.004287458024919033, 0.40685945749282837, -0.27480626106262207, 0.40373408794403076, -0.007485665380954742, -0.02760309912264347, -0.22626402974128723, 0.01046435721218586, -0.12387342751026154, 0.4250917434692383, 0.34055134654045105, -0.24968084692955017, 0.31807154417037964, -0.13490435481071472, -0.3477175533771515, 0.233708918094635, 0.12683452665805817, 0.0192521121352911, 0.00021515105618163943, -0.040310077369213104, 0.01534811407327652, -0.29494258761405945, -0.13681773841381073, 0.31067588925361633, 0.6342357993125916, 0.19913354516029358, 0.13627761602401733, -0.06178688257932663, -0.18401046097278595, -0.06175120174884796, 0.07712710648775101, 0.17387591302394867, -0.07647083699703217, 0.21138343214988708, -0.013369879685342312, 0.1067885085940361, 0.041692137718200684, -0.16831190884113312, 0.2519674003124237, -0.17216473817825317, -0.16441869735717773, 0.0485871322453022, -0.17014385759830475, -0.6914836764335632, -0.11620097607374191, 0.08361636102199554, -0.29779016971588135, -0.25566723942756653, 0.1518940031528473, 0.13057979941368103, -0.09642692655324936, 0.17519427835941315, 0.16043385863304138, 0.4907010793685913, -0.30469486117362976, -0.33770090341567993, -0.003231876064091921, -0.2629941999912262, -0.0777212455868721, 0.0428459458053112, 0.43576282262802124, 0.41830974817276, 0.4902556240558624, 0.1984889805316925, -0.07153218239545822, -0.04202013835310936, -0.3898181915283203, 0.1562468260526657, -0.36184194684028625, 0.3624504804611206, 0.06319837272167206, -0.12011867761611938, 0.07015428692102432, -0.31865814328193665, -0.014651053585112095, -0.0883055329322815, -0.148350790143013, -0.06262724846601486, -0.0902751013636589, 0.07648733258247375, -0.13062770664691925, -0.16166286170482635, -0.607825517654419, -0.3920362591743469, 0.19814470410346985, -0.18852907419204712, 0.07596119493246078, 0.15446582436561584, -0.28402748703956604, 0.24272438883781433, -0.12732911109924316, 0.08162112534046173, -0.13222485780715942, -0.1482553631067276, -0.01692887395620346, -0.40536513924598694, -0.2201676070690155, -0.40197211503982544, -0.019565217196941376, 0.16767044365406036, 0.05245676636695862, -0.09963592886924744, -0.39100706577301025, 0.0016658894019201398, 0.027672257274389267, 0.05836043879389763, 0.25692451000213623, 0.11975843459367752, -0.281875342130661, 0.16816411912441254, -0.14121802151203156, -0.18737374246120453, 0.2724555432796478, 0.3120082914829254, 0.36048638820648193, 0.17845942080020905, -0.0021586560178548098, 0.1279565989971161, 0.9331133365631104, 0.294748455286026, 0.004239199683070183, 0.184007927775383, 0.13027030229568481, 0.019158830866217613, 0.08279602229595184, -0.19149279594421387, 0.20076076686382294, -0.11971075087785721, 0.4675973355770111, 0.10902047157287598, 0.08792636543512344, -0.016809310764074326, -0.2927389144897461, 0.09595129638910294, -0.3363725244998932, -0.25714975595474243, 0.20428664982318878, -0.34242284297943115, 0.10670763999223709, -0.1269156038761139, -0.4086247682571411, -0.37843674421310425, -0.07600954920053482, -0.09919127076864243, 0.2852097749710083, 0.1052638366818428, 0.17052415013313293, -0.38959255814552307, -0.050177209079265594, -0.4277201294898987, 0.14170695841312408, 0.12198332697153091, 0.29321420192718506, -0.08066632598638535, -0.0009058640571311116, -0.061951447278261185, 0.139470174908638, 0.576267421245575, -0.4894058108329773, -0.09264400601387024, 0.13218410313129425, -0.09517445415258408, -0.16487501561641693, -0.029359448701143265, -0.3096030056476593, 0.05311180278658867, 0.3403048515319824, 0.34198251366615295, -0.482268750667572, -0.44317707419395447, 0.285530686378479, -0.0707065761089325, -0.06597491353750229, -0.16693177819252014, -0.16285686194896698, -0.2332926094532013, -0.0486726351082325, -0.08596408367156982, 0.35172295570373535, 0.24646823108196259, -0.14004097878932953, 0.20227612555027008, -0.03201519697904587, 0.2508803606033325, 0.20811614394187927, -0.08595175296068192, -0.07969796657562256, 0.1945638805627823, 0.25474798679351807, 0.0349113754928112, -0.10541535913944244, -0.30441221594810486, 0.1429985910654068, -0.14487148821353912, -0.11757615208625793, 0.28176024556159973, 0.06203873082995415, 0.6452472805976868, 0.07200423628091812, 0.3485735058784485, 0.3418126404285431, -0.14397554099559784, 0.08821761608123779, -0.10354967415332794, 0.20663753151893616, 0.22724981606006622, 0.25939422845840454, -0.11130652576684952, -0.27035731077194214, 0.5457841157913208, 0.31183570623397827, 0.01879657618701458, 0.46680858731269836, -0.5664702653884888, -0.2669517397880554, 0.20873482525348663, 0.04355107620358467, 0.7242980599403381, -0.0649958997964859, 0.4172831177711487, 0.1482737958431244, -0.3961779475212097, 0.5780621767044067, -0.21578052639961243, 0.041985925287008286, -0.14468608796596527, -0.32053905725479126, 0.02411653846502304, -0.13038361072540283, -0.20336784422397614, 0.22062033414840698, -0.2274606078863144, 0.2735256850719452, 0.3709960877895355, -0.21724523603916168, -0.03466949984431267, 0.4872439503669739, -0.08320411294698715, -0.31027331948280334, -0.007831919938325882, -0.005010316614061594, -0.2720608711242676, 0.1534225344657898, -0.15633101761341095, -0.27250996232032776, 0.18027885258197784, 0.22555729746818542, -0.36308756470680237, -0.03686486557126045, -0.041071727871894836, 0.17539925873279572, 0.2798943519592285, -0.06461703777313232, 0.22811898589134216, -0.05899588391184807, -0.04431465268135071, -0.18439985811710358, -0.13139377534389496, 0.1105046421289444, -0.2541779577732086, 0.03881360962986946, 0.13085909187793732, -0.08861712366342545, 0.47457191348075867, -0.10375441610813141, -0.14080709218978882, -0.21365079283714294, 0.20078150928020477, -0.4532194435596466, -0.2897719442844391, -0.028219154104590416, 0.04603630304336548, 0.14461570978164673, -0.3120005428791046, 0.5833739042282104, 0.020649399608373642, 0.05548350140452385, -0.032347891479730606, 0.2065165787935257, -0.17367291450500488, 0.42533132433891296, -0.35417553782463074, -0.04309534654021263, 0.0554853193461895, 0.03814931958913803, -0.049222107976675034, -0.3953079879283905, 0.07578074187040329, 0.4156872630119324, 0.16333776712417603, 0.03143160790205002, 0.1543506532907486, 0.10619034618139267, -0.18556231260299683, 0.1107318252325058, -0.07481797784566879, -0.19344587624073029, 0.1271693855524063, 0.3633469343185425, 0.16826605796813965, 0.12084880471229553, -0.06369728595018387, -0.23129472136497498, -0.18166881799697876, 0.5214848518371582, -0.014888892881572247, 0.2288406938314438, -0.1573706716299057, 0.4722117781639099, 0.42730844020843506, 0.4741534888744354, -0.19213160872459412, 0.11956620216369629, 0.11680866032838821, 0.08993300050497055, 0.5097171068191528, -0.24610285460948944, 0.049917228519916534, 0.2659686207771301, -0.21029981970787048, -0.0020564410369843245, -0.31299203634262085, -0.08803526312112808, -0.26381608843803406, 0.15460611879825592, 0.13262192904949188, 0.017633749172091484, 0.08132210373878479, -0.10233201086521149, -0.20836415886878967, -0.1031864732503891, 0.1371445208787918, -0.22865700721740723, -0.07985033839941025, 0.21865065395832062, 0.004845073912292719, -0.015345177613198757, -0.068305104970932, 0.3252384662628174, 0.0670614242553711, -0.0039597731083631516, 0.34234175086021423, 0.15840297937393188, 0.16730651259422302, -0.1784553974866867, -0.16622062027454376, -0.027407735586166382, 0.3266688883304596, 0.09608421474695206, -0.06478874385356903, 0.05887962877750397, -0.11979376524686813, -0.18889731168746948, 0.1873275488615036, 0.2522062063217163, 0.030034556984901428, -0.02713894471526146, 0.21464209258556366, 0.03223833069205284, -0.03687312826514244, 0.00163279939442873, 0.13352438807487488, 0.005256152246147394, 0.08753187954425812, 0.12357042729854584, 0.009909730404615402, 0.48670947551727295, -0.1714266836643219, -0.04055018723011017, 0.00025952974101528525, -0.2595716714859009, 0.16239126026630402, 0.03387768194079399, -0.05490383878350258, -0.018305199220776558, 0.07210617512464523, -0.036743875592947006, 0.12226538360118866, -0.33030396699905396, 0.3012450039386749, 0.27697938680648804, -0.08719217777252197, -0.015070118941366673, 0.230384960770607, -0.08356542885303497, -0.1979319304227829, 0.31108635663986206, 0.07949624955654144, 0.41786521673202515, -0.009903552010655403, 0.918045163154602, -0.3006817400455475, -0.4647100269794464, 0.011498858220875263, 0.5212085843086243, 0.1481102854013443, -0.252360075712204, 0.06891943514347076, 0.08991701155900955, -0.2977738082408905, 0.09773916751146317, -0.009851265698671341, -0.1815321296453476, 0.39744827151298523, -0.010155856609344482, 0.19723689556121826, -0.13021734356880188, -0.14882870018482208, 0.3492199182510376, 0.17988654971122742, -0.006385840941220522, 0.0832379162311554, 0.039727941155433655, -0.10503031313419342, -0.1717856228351593, 0.5120560526847839, 0.08268147706985474, -0.07022836804389954, -0.5144467353820801, 0.02289748005568981, 0.1550082564353943, 0.1871427297592163, 0.12166742980480194, 0.20464295148849487, 0.4158379137516022, -0.1962643563747406, 0.36525964736938477, -0.08221448212862015, -0.03379262611269951, -0.18054719269275665, 0.060573212802410126, -0.4283541738986969, -0.4648219048976898, 0.468281090259552, 0.05242529883980751, -0.14958178997039795, 0.16356700658798218, 0.2457016557455063, -0.6194168329238892, -0.2717796862125397, 0.32530903816223145, -0.17062543332576752, -0.09467189013957977, -0.14008164405822754, 0.03522855043411255, 0.015107734128832817, 0.7085170149803162, 0.24064163863658905, 0.44719865918159485, -0.3157658278942108, -0.06770215183496475, -0.5137104392051697, -0.1357128620147705, 0.1859477013349533, -0.07559128105640411, -0.1854136735200882, 0.05579189583659172, 0.04734194278717041, 0.31720682978630066, -0.08298147469758987, 0.1642475575208664, 0.1825139969587326, 0.19075407087802887, -0.5117712616920471, -0.04429370164871216, -0.023752178996801376, 0.233748197555542, 0.2409575879573822, -0.13445545732975006, -0.21129363775253296, 0.11336632817983627, -0.11925768852233887, -0.10776656866073608, -0.12228390574455261, 0.34944605827331543, 0.057619329541921616, 0.10408392548561096, 0.0757715180516243, 0.5753723978996277, 0.06265409290790558, -0.19474005699157715, 0.04764502868056297, -0.051162317395210266, -0.2982564866542816, -0.09186168760061264, -0.07321720570325851, 0.11370617151260376, -0.35987424850463867, 0.35187625885009766, 0.013054802082479, -0.026847243309020996, 0.11974912136793137, -0.05373901501297951, -0.07982296496629715, -0.1371590793132782, 0.060858532786369324, 0.19985537230968475, 0.2587178349494934, 0.7949316501617432, -0.15307685732841492, 0.13903604447841644, -0.13647979497909546, 0.05447745695710182, 0.1046760156750679, -0.1295953243970871, -0.2167316973209381, -0.1166486144065857, 0.05686790123581886, 0.04489805921912193, 0.1433515101671219, -0.535740852355957, -0.2840757369995117, 0.1221645399928093, 0.0052573769353330135, 0.017481917515397072, 0.7596601843833923, 0.2824135720729828, -0.1751398891210556, -0.051766544580459595, -0.1894175112247467, 0.008700874634087086, -0.2264738380908966, 0.09785183519124985, -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])`
[ 0.08258108794689178, -0.44010573625564575, -0.09260918200016022, 0.38070473074913025, -0.05949300155043602, -0.13712067902088165, -0.3506771922111511, 0.09061781316995621, 0.02350255846977234, 0.23529517650604248, -0.2559319734573364, 0.16841153800487518, -0.02155134081840515, 0.002206427277997136, 0.09740174561738968, -0.3178662359714508, 0.037423208355903625, -0.03475465998053551, 0.16254763305187225, -0.2055862993001938, -0.15831181406974792, 0.35643163323402405, -0.06628065556287766, 0.2676071524620056, 0.054910048842430115, 0.22830070555210114, 0.18456697463989258, 0.20209889113903046, -0.4124771058559418, -0.17358016967773438, 0.14888514578342438, 0.005113185383379459, 0.23019911348819733, 0.2997414469718933, -0.0001134505364461802, -0.22685761749744415, 0.16347812116146088, -0.09964212775230408, 0.10824383050203323, -0.5389816761016846, 0.06842257082462311, -0.2803027927875519, -0.11710099130868912, -0.050361353904008865, -0.07701847702264786, -0.2652796804904938, -0.3086015582084656, -0.15958653390407562, 0.3000631034374237, 0.15584179759025574, 0.13660414516925812, -0.20554138720035553, -0.06221542879939079, -0.10612603276968002, 0.2618379592895508, -0.10174722224473953, 0.11459240317344666, 0.5414437055587769, 0.2648923695087433, 0.0035379035398364067, 0.12690585851669312, 0.14564721286296844, 0.25746795535087585, 0.06553088873624802, 0.3002890646457672, 0.21130594611167908, 0.31865444779396057, -0.11210065335035324, -0.19202956557273865, 0.2805584669113159, 0.12960253655910492, -0.134212926030159, -0.302421897649765, 0.10026529431343079, 0.07765430957078934, -0.5770217180252075, -0.2875574231147766, -0.10818591713905334, 0.1591964215040207, -0.21189624071121216, -0.1913762241601944, -0.04324186593294144, -0.15449917316436768, 0.10635083168745041, 0.0733344703912735, -0.011296848766505718, -0.05365898832678795, 0.19882231950759888, 0.184065580368042, -0.0033977406565099955, -0.5394580960273743, 0.2152118682861328, -0.31420913338661194, 0.2942137122154236, -0.3065519332885742, -0.09514333307743073, 0.18695026636123657, 0.40665125846862793, 0.10325281322002411, 0.348979651927948, 0.11953458189964294, -0.2580718696117401, -0.2762499749660492, 0.0872943177819252, -0.06995920091867447, 0.4045152962207794, 0.19604967534542084, 0.3381277322769165, -0.028442785143852234, 0.06927015632390976, 0.1954554319381714, -0.06287286430597305, 0.24693505465984344, -0.477272629737854, -0.06356951594352722, 0.4770830273628235, -0.007767843082547188, -0.5834072828292847, -0.4587993919849396, -0.0287332683801651, -0.3379417359828949, -0.20797038078308105, 0.15752945840358734, 0.10549367219209671, -0.2072853446006775, 0.3233102858066559, -0.034194011241197586, 0.2794032394886017, -0.32626044750213623, -0.113673135638237, -0.15017735958099365, 0.20628602802753448, -0.2022605687379837, 0.22359508275985718, 0.15393975377082825, 0.1776575744152069, 0.2562236487865448, 0.21195414662361145, 0.2691629230976105, 0.0598432794213295, 0.20626841485500336, -0.23722785711288452, -0.11752679944038391, -0.14814883470535278, -0.03447916731238365, 0.07741101086139679, 0.18840712308883667, -0.27248063683509827, -0.1226748675107956, -0.07505357265472412, -0.3015037178993225, -0.008742103353142738, 0.15920142829418182, 0.13326065242290497, 0.12586045265197754, 0.016422919929027557, -0.21359150111675262, 0.3548073470592499, -0.15039581060409546, -0.2694520354270935, 0.20135001838207245, -0.07212784141302109, -0.25833097100257874, -0.12636488676071167, 0.2293127030134201, -0.326433390378952, -0.17086613178253174, 0.044310312718153, 0.267841100692749, -0.08543562889099121, 0.05896719545125961, 0.36661991477012634, 0.0015421402640640736, 0.19801343977451324, -0.15210577845573425, -0.5923178791999817, 0.6651850342750549, -0.7773893475532532, -0.06579919904470444, -0.0005513983778655529, -0.11361727118492126, -0.15772943198680878, 0.13908828794956207, 0.02933243103325367, -0.2963535785675049, -0.20137320458889008, 0.3400661051273346, -0.031104421243071556, -0.1501091569662094, -0.3194095194339752, -0.0926348939538002, 0.33678919076919556, 0.45280513167381287, -0.16289211809635162, -0.042493198066949844, -0.11781184375286102, -0.08548703044652939, -0.04174085706472397, 0.40881630778312683, 0.06966614723205566, -0.12715521454811096, 0.021893568336963654, -0.18314962089061737, 0.16435641050338745, 0.2299865186214447, -0.2909388840198517, 0.2540517747402191, 0.10766568779945374, -0.9638234376907349, -0.22757463157176971, 0.4533107280731201, -0.25475794076919556, -0.6550407409667969, -0.20376750826835632, 0.3090624213218689, -0.26727256178855896, 0.09786073863506317, -0.16166381537914276, 0.36613020300865173, 0.061518024653196335, -0.12694677710533142, 0.09669619053602219, 0.060176167637109756, -0.00419210409745574, -0.3270716965198517, -0.13261856138706207, 0.1051754504442215, 0.07881274074316025, 0.2529539167881012, 0.4529830515384674, 0.4550013542175293, 0.26617828011512756, 0.14760729670524597, 0.2340746968984604, 0.22817136347293854, 0.11697326600551605, 0.3165964186191559, 0.28234386444091797, 0.16588670015335083, 0.06816942989826202, -0.06692624092102051, 0.45915114879608154, 0.3160407245159149, -0.1627158373594284, -0.19575335085391998, 0.4131931960582733, -0.48462289571762085, 0.12427667528390884, 0.027903972193598747, 0.012188353575766087, -0.20597583055496216, 0.1272018551826477, -0.4066193401813507, 0.004663374740630388, 0.23972083628177643, -0.10887447744607925, 0.1383102834224701, -0.2654801905155182, 0.07673676311969757, -0.06142575293779373, 0.19538305699825287, 0.02158014103770256, -0.036684345453977585, -0.1353597342967987, -0.0005482809501700103, -0.043356187641620636, -0.09061439335346222, -0.208299919962883, 0.1854833960533142, 0.1760423481464386, 0.2232331484556198, 0.12233243882656097, -0.368216872215271, -0.1841251105070114, -0.053566161543130875, -0.057328227907419205, 0.016051121056079865, 0.023728853091597557, 0.11102589964866638, -0.04540031775832176, 0.13835108280181885, -0.2840757369995117, -0.259244441986084, 0.10575356334447861, -0.3508681356906891, 0.10057410597801208, -0.09940065443515778, -0.11004474014043808, -0.17763330042362213, -0.5282848477363586, -0.1343933641910553, -0.34777483344078064, -0.026153311133384705, 0.07551545649766922, 0.2976187467575073, -0.20343030989170074, 0.21936553716659546, 0.18032394349575043, -0.1081443801522255, 0.3281557261943817, 0.10603395849466324, -0.21061021089553833, -0.2679210603237152, 0.17807485163211823, -0.22155314683914185, 0.007940308190882206, 0.24635569751262665, -0.12963169813156128, -0.007006484549492598, 0.15435968339443207, -0.2549612522125244, -0.09858131408691406, -0.0539374016225338, 0.22666507959365845, 0.06242278218269348, -0.28835681080818176, -0.20405732095241547, 0.30453068017959595, 0.2358664870262146, -0.024508671835064888, -0.2616881728172302, 0.034093260765075684, -0.19851471483707428, 0.20603370666503906, 0.052129946649074554, -0.1808527410030365, 0.06330021470785141, -0.08067402243614197, 0.4839155077934265, 0.24622252583503723, -0.07296391576528549, 0.27919918298721313, -0.3118188977241516, 0.25193139910697937, 0.3141075372695923, 0.37752383947372437, -0.3375563323497772, -0.2846830487251282, 0.31501662731170654, -0.4162308871746063, -0.3131349980831146, -0.0967823937535286, 0.06253836303949356, -0.013087349943816662, -0.08516395092010498, 0.06592933088541031, -0.3470349609851837, 0.03881412371993065, -0.041319023817777634, -0.12303409725427628, 0.18541383743286133, 0.24729037284851074, -0.26667720079421997, -0.059141118079423904, -0.10685432702302933, 0.08153656125068665, -0.013448418118059635, 0.10353488475084305, 0.5398585796356201, -0.2782626152038574, 0.1178239956498146, 0.13101090490818024, 0.34448790550231934, 0.37147215008735657, 0.43438324332237244, 0.05482892319560051, 0.16002950072288513, 0.02915610559284687, 0.08967442065477371, -0.17208319902420044, 0.3043067753314972, 0.07725416123867035, -0.15839801728725433, 0.07002726197242737, 0.1549822986125946, 0.42408961057662964, -0.2869345545768738, -0.03230841085314751, -0.10816027969121933, 0.10679012537002563, -0.08260093629360199, -0.2662000060081482, -0.03960119187831879, -0.08259453624486923, 0.04221712425351143, -0.12649759650230408, -0.10205943882465363, 0.30802661180496216, 0.036289993673563004, 0.20763370394706726, 0.19738133251667023, -0.37305402755737305, -0.08608078211545944, -0.30504900217056274, 0.4734289348125458, 0.09159303456544876, -0.003033858025446534, 0.12427829951047897, 0.09895651787519455, -0.017055075615644455, -0.194362610578537, 0.44468164443969727, 0.036086827516555786, 0.1932583898305893, 0.2945859134197235, -0.19206926226615906, -0.6826239228248596, -0.17620937526226044, -0.31174904108047485, 0.27777206897735596, 0.16995863616466522, 0.08621390908956528, -0.1744394302368164, 0.05788708105683327, 0.19328711926937103, 0.0027840635739266872, -0.1897600293159485, 0.027196373790502548, -0.31895676255226135, 0.015282286331057549, 0.32539814710617065, -0.21736924350261688, -0.1152779683470726, 0.5167666077613831, -0.08533590286970139, 0.3249632716178894, 0.09882189333438873, -0.25806882977485657, 0.4550524353981018, -0.0021248958073556423, 0.24309459328651428, 0.11786975711584091, -0.18184968829154968, 0.15938439965248108, -0.1389511525630951, -0.5745753049850464, 0.33481112122535706, 0.09035530686378479, -0.5968517065048218, -0.10880051553249359, 0.008119825273752213, 0.4826161563396454, 0.030374696478247643, -0.05051341652870178, -0.08396965265274048, -0.44186124205589294, -0.11237736791372299, -0.5094913244247437, -0.13879278302192688, 0.27017685770988464, -0.006576849147677422, -0.11409135907888412, -0.3621307909488678, 0.5230098366737366, 0.3907357156276703, -0.3316461443901062, 0.35048046708106995, -0.2329195737838745, -0.2078518569469452, 0.223556250333786, 0.07861416786909103, 0.7750366926193237, 0.23556752502918243, -0.03162870556116104, 0.12106242030858994, -0.25697919726371765, 0.5335091352462769, -0.6500672101974487, 0.07334878295660019, -0.21515604853630066, -0.2684263288974762, -0.03998471423983574, 0.06009352579712868, 0.04821157827973366, 0.18227219581604004, 0.08014305680990219, 0.5681716203689575, -0.06007957085967064, -0.15259939432144165, -0.030920621007680893, 0.253612220287323, 0.09966512769460678, -0.12943445146083832, -0.1757190078496933, 0.167453795671463, -0.132917121052742, 0.00017705981736071408, -0.2708413004875183, 0.11507507413625717, -0.5599819421768188, 0.06690455228090286, -0.14745117723941803, -0.13702526688575745, -0.18371480703353882, 0.1779044270515442, 0.1358851194381714, -0.4225405752658844, -0.3310343027114868, 0.5675361752510071, 0.7250475287437439, 0.06679403781890869, -0.2974839210510254, 0.2634715735912323, -0.17508479952812195, 0.11545813083648682, 0.03641780465841293, -0.09256527572870255, 0.05644349753856659, -0.028397085145115852, -0.02011876180768013, 0.32481053471565247, 0.3469849228858948, -0.0021527993958443403, -0.17017725110054016, 0.009425870142877102, -0.14369596540927887, 0.04415113106369972, -0.059258487075567245, 0.24112538993358612, 0.3410099744796753, 0.07543081045150757, 0.0871412605047226, 0.2268550544977188, -0.5656340718269348, -0.05264550819993019, 0.2004537433385849, -0.19524171948432922, 0.14268159866333008, 0.29056382179260254, 0.037577249109745026, -0.29353657364845276, -0.22153684496879578, 0.08756282180547714, 0.07894828170537949, -0.005834947805851698, -0.04641580581665039, 0.01647283509373665, -0.06334859132766724, 0.018273215740919113, 0.28729668259620667, 0.1504465490579605, 0.03996027633547783, 0.5065686106681824, 0.31823673844337463, -0.13108064234256744, 0.252267062664032, -0.45389890670776367, -0.22410722076892853, 0.19057050347328186, 0.08486615866422653, -0.08753068745136261, -0.08937274664640427, -0.0839211642742157, 0.13785184919834137, 0.5422527194023132, -0.28925877809524536, -0.23756732046604156, -0.04240002483129501, 0.14428775012493134, -0.16340014338493347, -0.1392481029033661, -0.10749548673629761, 0.14073367416858673, 0.06438962370157242, 0.044967614114284515, -0.18612565100193024, -0.1794562041759491, -0.24945789575576782, 0.1313876062631607, -0.14952464401721954, 0.1688951700925827, -0.06284710019826889, 0.07508017867803574, 0.18467943370342255, -0.08171840012073517, -0.05887404829263687, 0.3747631013393402, 0.3735155165195465, 0.3379555940628052, -0.03439231216907501, 0.10218087583780289, -0.056325044482946396, -0.01782965660095215, 0.21329346299171448, 0.20780132710933685, 0.07900930941104889, -0.008537634275853634, 0.14385944604873657, -0.01686854287981987, -0.28753602504730225, -0.0898604616522789, -0.14779497683048248, -0.22689500451087952, 0.14476263523101807, -0.018569713458418846, 0.20657122135162354, -0.21333353221416473, 0.37987709045410156, 0.1326860934495926, -0.0760842114686966, 0.18973122537136078, -0.18290534615516663, 0.14285613596439362, -0.11980106681585312, 0.40617966651916504, 0.17260000109672546, 0.25368472933769226, -0.21503499150276184, 0.25820741057395935, -0.0332847461104393, 0.15419074892997742, -0.011892584152519703, -0.008982140570878983, -0.027472559362649918, -0.3529619574546814, 0.21023716032505035, -0.45130008459091187, 0.13748310506343842, -0.07608523219823837, 0.34516972303390503, -0.14173294603824615, 0.08663562685251236, 0.09466512501239777, 0.11317455023527145, 0.20618678629398346, 0.05709505081176758, 0.38195016980171204, 0.05865364894270897, 0.16471202671527863, -0.335029274225235, 0.2503335475921631, 0.03729325905442238, -0.05603472515940666, 0.060577839612960815, 0.1348673403263092, 0.040382616221904755, -0.19460268318653107, -0.12172941118478775, 0.10459189116954803, 0.12302757054567337, -0.3937948942184448, -0.3598281741142273, -0.25772807002067566, -0.19753174483776093, -0.056885018944740295, -0.00442329840734601, 0.09338748455047607, -0.14682897925376892, 0.056325193494558334, -0.05960000678896904, -0.2923716604709625, -0.39882606267929077, 0.15627144277095795, 0.3441051244735718, -0.07433148473501205, 0.1079130619764328, 0.20615999400615692, 0.31274667382240295, 0.3157897889614105, 0.3679596483707428, 0.13274118304252625, 0.08966731280088425, -0.5088683366775513, 0.05510058254003525, -0.18205226957798004, -0.07493659108877182, 0.036594729870557785, 0.2758980393409729, 0.19292792677879333, -0.17557178437709808, 0.3166452944278717, 0.1541426181793213, -0.10532337427139282, 0.33918431401252747, -0.14225678145885468, 0.05937488004565239, -0.4119408130645752, 0.5525146722793579, 0.036496296525001526, 0.20788812637329102, -0.20443928241729736, 0.059173498302698135, -0.4736991822719574, -0.18732237815856934, -0.5115628242492676, 0.06559556722640991, 0.061522502452135086, -0.3025530278682709, 0.07796715199947357, 0.0898432582616806, 0.024084795266389847, 0.08686596155166626, 0.42110973596572876, -0.3376677930355072, -0.19173680245876312, -0.3841715157032013, -0.3622981011867523, -0.048488907516002655, 0.1175469160079956, -0.24215450882911682, 0.2872098684310913, -0.06774698942899704, 0.5412330627441406, 0.15837736427783966, 0.48342883586883545, 0.05399332195520401, 0.17747843265533447, -0.20079396665096283, -0.012257048860192299, -0.09126206487417221, -0.10455874353647232, -0.10259959101676941, 0.2286400943994522, -0.21448925137519836, -0.17608626186847687, 0.0368305929005146, -0.15790441632270813, -0.26630881428718567, 0.24678461253643036, 0.6915801167488098, 0.31457579135894775, 0.3558570444583893, 0.4649415910243988, -0.01913045160472393, 0.03927028924226761, -0.0585143119096756, -0.0018535570707172155, -0.05582133308053017, 0.15834499895572662, -0.3841409385204315, 0.31250423192977905, -0.3025667369365692, -0.2276661992073059, 0.10168594866991043, 0.6001898646354675, 0.009413305670022964, -0.20100083947181702, -0.3931237757205963, -0.07098322361707687, 0.018479008227586746, -0.149862602353096, -0.4123908281326294, 0.3378526270389557, -0.015125593170523643, 0.442717581987381, -0.00961952656507492, -0.13172684609889984, 0.4783709645271301, -0.365596741437912, 0.06721769273281097, 0.06162620708346367, 0.38429132103919983, 0.2618562579154968, -0.07773256301879883, -0.1968768686056137, -0.03381256014108658, 0.3542481064796448, 0.19263818860054016, -0.0012843633303418756, 0.16451609134674072, -0.18422232568264008, -0.10394728928804398, -0.15975959599018097, -0.019801955670118332, 0.06336132436990738, 0.04165356606245041, 0.10203631967306137, -0.16577692329883575 ]
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 :)
[ 0.12398974597454071, -0.49795177578926086, -0.08898458629846573, 0.34631484746932983, -0.08415486663579941, -0.2206081598997116, -0.3729410171508789, 0.08035808801651001, 0.036045778542757034, 0.22104908525943756, -0.21847876906394958, 0.23340606689453125, -0.011109297163784504, -0.0014881264651194215, 0.07686209678649902, -0.28436294198036194, -0.011588181369006634, -0.03263481706380844, 0.08533447980880737, -0.24822008609771729, -0.10775785148143768, 0.4054568111896515, -0.1544710099697113, 0.2524570822715759, 0.06542956829071045, 0.3166962265968323, 0.17007917165756226, 0.21432313323020935, -0.432865709066391, -0.18212704360485077, 0.1652715504169464, 0.05036298930644989, 0.14969606697559357, 0.23267799615859985, -0.000114266193122603, -0.29253748059272766, 0.18494386970996857, -0.1414392739534378, 0.17386220395565033, -0.6005355715751648, 0.058325670659542084, -0.26914966106414795, -0.08119751513004303, -0.06456456333398819, -0.1110231876373291, -0.19320018589496613, -0.2885154187679291, -0.12736785411834717, 0.3743639588356018, 0.23226206004619598, 0.12809443473815918, -0.20203980803489685, -0.06131405383348465, -0.05112752318382263, 0.24692314863204956, -0.08830821514129639, 0.08550667762756348, 0.5601397156715393, 0.2660447657108307, -0.03746751323342323, 0.08142395317554474, 0.15097463130950928, 0.23171237111091614, 0.009744388982653618, 0.39035800099372864, 0.1786264181137085, 0.24379940330982208, -0.09382632374763489, -0.22313937544822693, 0.3274887204170227, 0.1833609640598297, -0.15424370765686035, -0.2336723953485489, 0.15244022011756897, 0.027276137843728065, -0.5276973247528076, -0.25022363662719727, -0.09636007249355316, 0.2021206021308899, -0.15361136198043823, -0.24245630204677582, -0.030754096806049347, -0.21640455722808838, 0.08503448218107224, 0.00018079168512485921, -0.007077925372868776, -0.0669671967625618, 0.16879887878894806, 0.20319350063800812, 0.011866003274917603, -0.5337000489234924, 0.2507757842540741, -0.30721724033355713, 0.34942132234573364, -0.3709573745727539, -0.16411645710468292, 0.1327211707830429, 0.4145556390285492, 0.10024318844079971, 0.40116065740585327, 0.22469407320022583, -0.14048723876476288, -0.26722341775894165, 0.10991979390382767, -0.033082373440265656, 0.44693198800086975, 0.22990910708904266, 0.31486475467681885, -0.02780299447476864, 0.02518315613269806, 0.14823845028877258, -0.022644950076937675, 0.23133665323257446, -0.4569520354270935, -0.038008298724889755, 0.39786890149116516, -0.02740120142698288, -0.6806554794311523, -0.4113559126853943, 0.0514223575592041, -0.32168155908584595, -0.14020556211471558, 0.11257831007242203, 0.09805741161108017, -0.19216451048851013, 0.33024030923843384, -0.007844310253858566, 0.23737461864948273, -0.33151909708976746, -0.14027731120586395, -0.0784640684723854, 0.22878101468086243, -0.25133216381073, 0.21520061790943146, 0.18882836401462555, 0.2151869833469391, 0.24487315118312836, 0.20560823380947113, 0.3032243251800537, 0.029208682477474213, 0.28292566537857056, -0.2480926364660263, -0.027120990678668022, -0.14943161606788635, -0.07167106866836548, 0.060484252870082855, 0.11274951696395874, -0.2844977378845215, -0.09193714708089828, -0.05520515888929367, -0.29216551780700684, 0.009701506234705448, 0.14610488712787628, 0.13343557715415955, 0.07843906432390213, 0.012796790339052677, -0.11202716082334518, 0.37731578946113586, -0.20596793293952942, -0.2661988139152527, 0.21629920601844788, -0.14620402455329895, -0.28263208270072937, -0.12567786872386932, 0.21286290884017944, -0.3043462038040161, -0.15618211030960083, 0.0847708061337471, 0.25281721353530884, -0.05326251685619354, 0.014622438699007034, 0.4060993194580078, 0.0686083734035492, 0.1522849053144455, -0.09097420424222946, -0.5238792300224304, 0.6502181887626648, -0.750755786895752, -0.09616753458976746, 0.0344628281891346, -0.08055095374584198, -0.10891734808683395, 0.08223561942577362, 0.02192322351038456, -0.31465744972229004, -0.14471237361431122, 0.281575471162796, 0.010718680918216705, -0.08700933307409286, -0.3107842803001404, -0.15044131875038147, 0.39708665013313293, 0.40273284912109375, -0.07194825261831284, -0.08274509012699127, -0.1770651638507843, -0.1273576319217682, -0.01692323386669159, 0.4076850414276123, 0.0375572107732296, -0.12502345442771912, 0.11389756947755814, -0.19675946235656738, 0.14640359580516815, 0.2724830210208893, -0.3126002848148346, 0.3493419289588928, 0.07731390744447708, -0.8894879221916199, -0.17240509390830994, 0.46221601963043213, -0.22569233179092407, -0.66511070728302, -0.21515187621116638, 0.3109322190284729, -0.31003352999687195, 0.10217571258544922, -0.22574594616889954, 0.3516812026500702, 0.11397837847471237, -0.15364255011081696, 0.20693844556808472, 0.06197977438569069, 0.006930873729288578, -0.31909748911857605, -0.16207370162010193, 0.10615438967943192, 0.05850539728999138, 0.2559981048107147, 0.47424182295799255, 0.4499497711658478, 0.22688031196594238, 0.16882668435573578, 0.21439826488494873, 0.25388431549072266, 0.034971654415130615, 0.25829920172691345, 0.20957612991333008, 0.19598034024238586, -0.018469778820872307, -0.07744134962558746, 0.519900918006897, 0.3037612736225128, -0.15413594245910645, -0.194383904337883, 0.46627485752105713, -0.490393728017807, 0.05744209513068199, 0.030122855678200722, 0.0035354280844330788, -0.23440879583358765, 0.04894740879535675, -0.4269280731678009, 0.008210213854908943, 0.33786410093307495, -0.07221478223800659, 0.15080982446670532, -0.21438343822956085, 0.0776558518409729, -0.05872403457760811, 0.16696782410144806, -0.023158876225352287, -0.0700705274939537, -0.09916809946298599, 0.048283111304044724, -0.052478209137916565, -0.08713286370038986, -0.1464821845293045, 0.1367141157388687, 0.14246423542499542, 0.1714952290058136, 0.21921026706695557, -0.4117475748062134, -0.18401677906513214, -0.007912248373031616, -0.15325385332107544, 0.01218412071466446, 0.015409769490361214, 0.06150927394628525, -0.07383888959884644, 0.05183999985456467, -0.30490610003471375, -0.3022274076938629, 0.067902572453022, -0.32608234882354736, 0.08883699029684067, -0.10181156545877457, -0.14128339290618896, -0.1743478775024414, -0.45969873666763306, -0.15071174502372742, -0.30679911375045776, -0.037475671619176865, 0.07321751862764359, 0.29975634813308716, -0.2984480857849121, 0.21187351644039154, 0.2086879014968872, -0.09639378637075424, 0.35062721371650696, 0.09900624305009842, -0.29849109053611755, -0.21889391541481018, 0.14851537346839905, -0.22972337901592255, 0.010996043682098389, 0.20694641768932343, -0.1500386893749237, -0.0032071031164377928, 0.1473246067762375, -0.28855597972869873, -0.09995412826538086, -0.025258325040340424, 0.1648847907781601, 0.05572684109210968, -0.33034318685531616, -0.21322619915008545, 0.3580798804759979, 0.15705786645412445, -0.04325280711054802, -0.25011059641838074, 0.09181642532348633, -0.12921281158924103, 0.13602451980113983, -0.005322074983268976, -0.2049805223941803, 0.07565262913703918, -0.01680372841656208, 0.48908257484436035, 0.23215287923812866, -0.025978965684771538, 0.285144567489624, -0.3540247976779938, 0.21252380311489105, 0.25654512643814087, 0.3914877772331238, -0.32673758268356323, -0.2725764513015747, 0.3942360281944275, -0.4358575642108917, -0.3721911907196045, -0.10294616967439651, 0.10516532510519028, 0.0294391717761755, -0.04442441463470459, 0.05712832883000374, -0.32959747314453125, 0.05410843342542648, -0.09084338694810867, -0.17520205676555634, 0.1330193132162094, 0.2248002141714096, -0.26698803901672363, -0.04010266438126564, -0.12790317833423615, 0.09605062752962112, -0.002312297699972987, 0.13142825663089752, 0.5086545348167419, -0.24810613691806793, 0.07726975530385971, 0.1404632180929184, 0.275656133890152, 0.4029240012168884, 0.44602423906326294, 0.04914869740605354, 0.16979767382144928, 0.022712398320436478, 0.08366754651069641, -0.1309291571378708, 0.28202980756759644, 0.04214998334646225, -0.12101101130247116, 0.10032171756029129, 0.18470127880573273, 0.3316660225391388, -0.34148114919662476, -0.0804634615778923, -0.09187661856412888, 0.09235560894012451, -0.06381132453680038, -0.2443145513534546, -0.03437358886003494, -0.05484957620501518, 0.05709172412753105, -0.09766263514757156, -0.1349439173936844, 0.31029772758483887, 0.0227067694067955, 0.2167983502149582, 0.16918806731700897, -0.37488991022109985, -0.028955239802598953, -0.32199347019195557, 0.42253392934799194, 0.06536121666431427, -0.03763526305556297, 0.13242276012897491, 0.06528027355670929, 0.00621345778927207, -0.2297491729259491, 0.38630080223083496, 0.06365136057138443, 0.2125621885061264, 0.26525598764419556, -0.22254760563373566, -0.6708531379699707, -0.18083655834197998, -0.3159850835800171, 0.2895922362804413, 0.19786790013313293, 0.023055072873830795, -0.156122088432312, 0.08311517536640167, 0.18468418717384338, -0.02811334654688835, -0.18701855838298798, 0.04569626599550247, -0.258105993270874, 0.009999394416809082, 0.3646007776260376, -0.22924275696277618, -0.14277920126914978, 0.4782615900039673, -0.06073041632771492, 0.3421841859817505, 0.14983788132667542, -0.21545158326625824, 0.453617125749588, 0.04777456820011139, 0.2763141989707947, 0.026590339839458466, -0.20807862281799316, 0.16249078512191772, -0.09441923350095749, -0.6795374155044556, 0.3389071822166443, 0.1927538961172104, -0.590570867061615, -0.16465143859386444, -0.03763667866587639, 0.47643160820007324, 0.054586999118328094, -0.011679421178996563, -0.09295836091041565, -0.4811422824859619, -0.06214715167880058, -0.5296534299850464, -0.1239025741815567, 0.2654699683189392, -0.008956363424658775, -0.11165163666009903, -0.3016607463359833, 0.4860132038593292, 0.3428497314453125, -0.3132480978965759, 0.31255537271499634, -0.19474507868289948, -0.2029215395450592, 0.25724291801452637, 0.10273777693510056, 0.8113812804222107, 0.21375532448291779, 0.0021939361467957497, 0.07617845386266708, -0.35134586691856384, 0.47075146436691284, -0.7156766653060913, 0.044999174773693085, -0.26291927695274353, -0.25327327847480774, -0.04182609170675278, 0.04671785607933998, 0.04493233188986778, 0.21422645449638367, 0.06681875139474869, 0.5717447996139526, 0.06466982513666153, -0.1611304134130478, -0.003507071640342474, 0.2737407088279724, 0.13994696736335754, -0.05350326746702194, -0.14959971606731415, 0.1455356925725937, -0.12895630300045013, -0.027124721556901932, -0.2474437952041626, 0.0909351035952568, -0.5433633327484131, 0.0771607905626297, -0.16859158873558044, -0.12107563018798828, -0.14069585502147675, 0.11651507765054703, 0.13030724227428436, -0.39353442192077637, -0.2577757239341736, 0.6006634831428528, 0.7157096266746521, 0.12420652061700821, -0.37131813168525696, 0.2642662823200226, -0.2064560055732727, 0.15178179740905762, 0.10992161184549332, -0.054154668003320694, 0.045877669006586075, -0.057032693177461624, -0.06457602977752686, 0.27664175629615784, 0.3590576648712158, -0.010357392951846123, -0.22056956589221954, -0.0072340103797614574, -0.16538074612617493, 0.0703059658408165, -0.039750855416059494, 0.20711548626422882, 0.30184125900268555, 0.10604222118854523, 0.0886320024728775, 0.21996474266052246, -0.5469436049461365, -0.029552219435572624, 0.24314655363559723, -0.2311175912618637, 0.16570527851581573, 0.33130156993865967, 0.008900724351406097, -0.29333505034446716, -0.25438016653060913, 0.07475241273641586, 0.08233649283647537, -0.005110377445816994, -0.035461440682411194, -0.02048584818840027, -0.04117601364850998, 0.02023235522210598, 0.2510388195514679, 0.13301168382167816, 0.021783823147416115, 0.43461349606513977, 0.3064218759536743, -0.11115419119596481, 0.23509368300437927, -0.41880616545677185, -0.19548571109771729, 0.13808929920196533, 0.08050452172756195, -0.11472421139478683, -0.031742554157972336, -0.15798787772655487, 0.17435520887374878, 0.5180925726890564, -0.28377917408943176, -0.18762148916721344, -0.028949817642569542, 0.15590381622314453, -0.13453319668769836, -0.20898251235485077, -0.20269304513931274, 0.07883042842149734, 0.046424590051174164, 0.06440547853708267, -0.18087103962898254, -0.1644832342863083, -0.2384839802980423, 0.14039766788482666, -0.13585084676742554, 0.13904498517513275, -0.09388531744480133, 0.053113289177417755, 0.18270796537399292, -0.0889543890953064, -0.07194390892982483, 0.3491808772087097, 0.3699827492237091, 0.4574727714061737, -0.10201860964298248, 0.084199458360672, -0.11546287685632706, -0.03458700701594353, 0.21736612915992737, 0.22923868894577026, 0.080907441675663, -0.024483870714902878, 0.172731414437294, -0.036192528903484344, -0.2474125325679779, -0.0541042722761631, -0.20524117350578308, -0.27699846029281616, 0.20450451970100403, -0.0031602687668055296, 0.22931593656539917, -0.2594713270664215, 0.4298001825809479, 0.1612980216741562, -0.09349460899829865, 0.1403721421957016, -0.18729948997497559, 0.13596051931381226, -0.11948466300964355, 0.3579461872577667, 0.20658127963542938, 0.2541393041610718, -0.2403789907693863, 0.22333191335201263, -0.01607438549399376, 0.08376472443342209, 0.06476959586143494, -0.031833261251449585, -0.028623048216104507, -0.32644137740135193, 0.18215464055538177, -0.42984628677368164, 0.15969720482826233, -0.10262606292963028, 0.2641347348690033, -0.1583990752696991, 0.06268264353275299, 0.203857883810997, 0.11040244251489639, 0.22919917106628418, 0.02511708438396454, 0.4261513948440552, 0.04126840829849243, 0.2479865849018097, -0.30988121032714844, 0.26898637413978577, 0.07964642345905304, -0.03128261864185333, 0.10673385858535767, 0.14089955389499664, 0.11347521841526031, -0.2333458811044693, -0.1448434293270111, 0.06972701847553253, 0.12309956550598145, -0.397167831659317, -0.3614509403705597, -0.2509462237358093, -0.2693314850330353, -0.01795695722103119, -0.051786765456199646, 0.12138793617486954, -0.10855112224817276, 0.10018299520015717, -0.04484131932258606, -0.24225087463855743, -0.3757956624031067, 0.0737776905298233, 0.4461503028869629, -0.10802561789751053, 0.10562580823898315, 0.18961551785469055, 0.3804336190223694, 0.29856276512145996, 0.330020010471344, 0.12830062210559845, 0.0885034129023552, -0.5871761441230774, 0.10413949936628342, -0.18798497319221497, -0.08341574668884277, 0.07672002911567688, 0.32121095061302185, 0.1787658929824829, -0.16076792776584625, 0.3263953924179077, 0.12612111866474152, -0.06357874721288681, 0.34875622391700745, -0.11649837344884872, 0.05914182588458061, -0.3998605012893677, 0.4350467920303345, 0.11445902287960052, 0.21158495545387268, -0.25026413798332214, 0.018331041559576988, -0.5301826596260071, -0.18016090989112854, -0.5192384123802185, 0.1562502086162567, 0.057313911616802216, -0.33192959427833557, 0.07285402715206146, 0.1262475848197937, 0.0742608904838562, 0.055765300989151, 0.3771967589855194, -0.3097306787967682, -0.28629428148269653, -0.4125562608242035, -0.3241795003414154, 0.023713432252407074, 0.07984787970781326, -0.2076796293258667, 0.22943317890167236, -0.048436488956213, 0.5633387565612793, 0.10817000269889832, 0.46859872341156006, 0.05568800866603851, 0.17474563419818878, -0.2406177520751953, 0.055369287729263306, -0.09380529820919037, -0.13297009468078613, -0.14306601881980896, 0.2436823546886444, -0.17661787569522858, -0.18919774889945984, 0.030148208141326904, -0.12438838183879852, -0.2287803590297699, 0.21405881643295288, 0.7476032376289368, 0.26726147532463074, 0.3172944188117981, 0.4162449538707733, -0.0028229469899088144, 0.0363430455327034, -0.007560946047306061, 0.06603840738534927, 0.007561179809272289, 0.2057744860649109, -0.302584171295166, 0.25753313302993774, -0.21722078323364258, -0.2541283965110779, 0.1553342640399933, 0.6273364424705505, -0.040247052907943726, -0.20091111958026886, -0.3534561097621918, -0.15675175189971924, 0.05964139476418495, -0.16473792493343353, -0.5186472535133362, 0.39035049080848694, -0.04144960641860962, 0.4093334972858429, -0.0025365902110934258, -0.08588705956935883, 0.4886842370033264, -0.3368968367576599, 0.11393242329359055, 0.11974430084228516, 0.3727896809577942, 0.2416967749595642, -0.11075741052627563, -0.23718173801898956, -0.05154840648174286, 0.29273539781570435, 0.16268786787986755, 0.0006722693215124309, 0.20673109591007233, -0.15648125112056732, -0.10198014229536057, -0.171591117978096, 0.06903015077114105, 0.11228375881910324, 0.060577116906642914, 0.06237391009926796, -0.16444090008735657 ]
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
[ -0.13186056911945343, 0.1403096318244934, 0.05940936133265495, 0.421039879322052, -0.10276124626398087, 0.063910573720932, -0.09939055144786835, 0.24033157527446747, -0.16995196044445038, 0.16742537915706635, -0.0025767888873815536, 0.269363135099411, 0.10592728853225708, -0.029215428978204727, -0.06128936633467674, 0.11617539823055267, -0.0757436603307724, 0.2383858859539032, 0.06643012911081314, 0.009823443368077278, -0.02960888296365738, 0.4090341329574585, -0.290790855884552, -0.08376244455575943, -0.26350855827331543, -0.20447687804698944, -0.007695620413869619, 0.2664903402328491, -0.02803812175989151, -0.2925417125225067, 0.2720768451690674, 0.011068368330597878, 0.13318149745464325, 0.19191612303256989, -0.00012480426812544465, 0.022844407707452774, 0.5506138205528259, -0.25053170323371887, -0.21130093932151794, -0.24919673800468445, -0.41408583521842957, -0.2816128432750702, 0.03941405192017555, -0.30595383048057556, -0.07530731707811356, 0.13373973965644836, 0.1635826826095581, -0.05297669768333435, 0.2451406717300415, 0.3204348087310791, 0.11070699989795685, 0.07963921129703522, -0.268645316362381, 0.13458070158958435, 0.12662169337272644, 0.08180831372737885, -0.036990564316511154, -0.0744713619351387, -0.18289309740066528, -0.17931097745895386, -0.0033600919414311647, 0.46177735924720764, -0.27765512466430664, 0.13988223671913147, 0.19181622564792633, 0.1794872134923935, 0.3482629656791687, -0.13864772021770477, 0.15919561684131622, 0.31704673171043396, 0.4259490370750427, 0.15770450234413147, -0.20480965077877045, -0.42427900433540344, -0.021961674094200134, 0.1000383049249649, 0.4799434542655945, 0.32488197088241577, -0.23730996251106262, 0.03319874405860901, -0.4330621063709259, -0.07635178416967392, -0.036818236112594604, 0.009392685256898403, 0.1595870703458786, 0.32183754444122314, 0.08606140315532684, 0.07757250219583511, 0.09996235370635986, 0.013011477887630463, 0.34360387921333313, -0.27978405356407166, -0.21615900099277496, 0.042084161192178726, -0.10203374177217484, -0.1977805346250534, -0.18273743987083435, 0.062441401183605194, 0.15654750168323517, 0.0754769816994667, 0.3709210157394409, -0.08625367283821106, 0.17984481155872345, 0.11776161938905716, 0.37153828144073486, -0.013676989823579788, -0.028967399150133133, 0.5226830840110779, 0.10686898231506348, -0.009327389299869537, 0.11763446778059006, 0.0450279638171196, 0.29349982738494873, -0.14998497068881989, -0.19445747137069702, 0.12557068467140198, 0.1616523712873459, -0.6070356369018555, -0.44507625699043274, 0.016515955328941345, -0.5715518593788147, -0.20781360566616058, -0.05524887517094612, 0.24616685509681702, -0.3623117208480835, 0.2653515636920929, -0.18779286742210388, -0.037591300904750824, -0.3258187174797058, -0.017644107341766357, -0.18110668659210205, 0.16775602102279663, -0.13677240908145905, 0.25721171498298645, 0.14054319262504578, -0.18570256233215332, 0.4198707938194275, -0.17853720486164093, -0.030709201470017433, -0.1549042910337448, 0.26063063740730286, -0.1932947337627411, 0.07086912542581558, 0.22536779940128326, 0.0033085390459746122, 0.16629421710968018, -0.09200582653284073, -0.2952805161476135, -0.3580852746963501, 0.15888839960098267, -0.25167983770370483, -0.5216352939605713, 0.10345882177352905, 0.12058106809854507, -0.4729638397693634, -0.03648734465241432, -0.03690236061811447, 0.23291394114494324, 0.45762595534324646, -0.3340519368648529, -0.026959959417581558, -0.17522595822811127, -0.21760845184326172, -0.04367905855178833, -0.2373843491077423, 0.4392346441745758, 0.14354491233825684, -0.23715516924858093, 0.13986900448799133, 0.06780431419610977, 0.6044769287109375, 0.3637809455394745, -0.002590578980743885, 0.0724763497710228, -0.15798911452293396, 0.33445456624031067, 0.13933129608631134, -0.37921035289764404, -0.22019103169441223, 0.4759369194507599, 0.07704789936542511, 0.2783878445625305, 0.2617093324661255, 0.12012125551700592, 0.1928369402885437, -0.1457986831665039, 0.1319238841533661, 0.39678218960762024, -0.016017140820622444, 0.06469926983118057, -0.47572800517082214, -0.33400624990463257, 0.45108911395072937, 0.1048198863863945, 0.10609763115644455, 0.016658395528793335, -0.06624110788106918, 0.45338624715805054, 0.2774916887283325, -0.1179683655500412, -0.03857571259140968, 0.17891529202461243, -0.2975045144557953, -0.003019645344465971, -0.008994498290121555, 0.04593990743160248, -0.4310563802719116, -0.061897121369838715, -0.16155476868152618, 0.25738972425460815, 0.02154708281159401, -0.3396950960159302, -0.12602578103542328, -0.3853747248649597, -0.01261109672486782, -0.038151565939188004, -0.024999549612402916, 0.32791686058044434, 0.38916337490081787, 0.02011387050151825, 0.04731040075421333, 0.5165188312530518, -0.07561025768518448, 0.1980002373456955, -0.7434923648834229, 0.24971456825733185, -0.09700153023004532, -0.2209165096282959, 0.4876713454723358, 0.29407212138175964, 0.18833619356155396, -0.20826837420463562, -0.12820078432559967, 0.42031005024909973, -0.09913235157728195, 0.41020530462265015, -0.3028305172920227, 0.009452998638153076, 0.28660091757774353, -0.21590359508991241, -0.09509079903364182, 0.3127249479293823, -0.014362148940563202, -0.029679683968424797, -0.08529895544052124, 0.45606181025505066, -0.3552294671535492, 0.19494345784187317, 0.06018049642443657, -0.07444185763597488, 0.11264750361442566, -0.21182699501514435, -0.22562164068222046, -0.46470919251441956, 0.5550782680511475, -0.011282187886536121, -0.3389955461025238, 0.09569761902093887, -0.22145257890224457, 0.09830474108457565, 0.471059113740921, 0.09217573702335358, -0.13324379920959473, -0.10313479602336884, 0.10757987201213837, -0.1303173452615738, 0.24420149624347687, 0.21329867839813232, 0.4979694187641144, 0.12661859393119812, -0.025470906868577003, 0.05887512117624283, -0.14599910378456116, -0.38097327947616577, 0.13225600123405457, -0.017310818657279015, 0.14334973692893982, 0.288001149892807, 0.09366246312856674, -0.04037437215447426, -0.2727147042751312, -0.04973515868186951, 0.1679784208536148, 0.2461203783750534, -0.29215094447135925, -0.040914133191108704, -0.3865090608596802, -0.47475749254226685, -0.2583889663219452, -0.15843446552753448, -0.21669268608093262, -0.33608385920524597, 0.1613166332244873, 0.24778468906879425, -0.13261094689369202, 0.0692504346370697, -0.16835513710975647, 0.06610103696584702, -0.31220895051956177, -0.0732150748372078, -0.06265087425708771, -0.28624188899993896, -0.243129163980484, -0.04386449232697487, 0.1981617510318756, 0.37293609976768494, 0.20360082387924194, -0.425394207239151, -0.22502321004867554, 0.1412775218486786, -0.2914532423019409, 0.14501717686653137, -0.026525840163230896, -0.12445498257875443, 0.08017557114362717, -0.21738466620445251, 0.2646991014480591, -0.24174991250038147, 0.04491952061653137, -0.10440589487552643, -0.33106282353401184, -0.10031004995107651, 0.1255841851234436, 0.004528301768004894, -0.21038508415222168, -0.41956835985183716, -0.2590782046318054, -0.3447067141532898, -0.10632053762674332, 0.2822980284690857, 0.30478182435035706, 0.11865757405757904, -0.15986110270023346, -0.11309634149074554, -0.05772757530212402, 0.2196374535560608, -0.340501070022583, -0.09820781648159027, 0.23921044170856476, -0.08729135990142822, -0.11389192938804626, 0.011300330981612206, -0.05697092413902283, -0.033428240567445755, 0.1522318422794342, -0.5655203461647034, -0.2071274071931839, -0.16636061668395996, -0.10165060311555862, -0.2521023750305176, -0.2789224088191986, 0.5203964114189148, -0.05680593475699425, 0.03545086458325386, 0.065879687666893, -0.45105594396591187, 0.032256294041872025, 0.16768893599510193, 0.5530170202255249, -0.18503110110759735, 0.20177876949310303, -0.059397172182798386, 0.48935577273368835, 0.5270174145698547, 0.0005123536684550345, -0.15433135628700256, 0.15243998169898987, 0.14646750688552856, 0.07625634968280792, -0.03768443316221237, 0.3673475682735443, 0.05485771596431732, -0.222249373793602, 0.4481472671031952, -0.005777908954769373, -0.22578762471675873, -0.187069371342659, -0.10959038883447647, -0.26541051268577576, -0.28333085775375366, 0.1376711130142212, -0.16000208258628845, 0.1310962587594986, 0.2893316149711609, 0.32105252146720886, -0.07106977701187134, -0.5795767307281494, -0.07721071690320969, 0.3904203176498413, 0.1502409726381302, 0.12120439857244492, -0.39578673243522644, 0.06771257519721985, -0.5080149173736572, 0.34322646260261536, 0.24453188478946686, 0.282988041639328, -0.1522451937198639, -0.07149992138147354, 0.10070983320474625, -0.26286599040031433, 0.2946961522102356, -0.4508117735385895, -0.10308383405208588, -0.05667043849825859, -0.032913729548454285, -0.3269297182559967, -0.19002746045589447, -0.16598442196846008, 0.4601232409477234, 0.47317567467689514, 0.4658532440662384, -0.3707205653190613, -0.08754486590623856, 0.20380374789237976, -0.12906351685523987, -0.13388794660568237, -0.17332899570465088, -0.32800331711769104, -0.4676930904388428, -0.16358770430088043, -0.14629974961280823, -0.1255982369184494, 0.3344830274581909, -0.013892768882215023, -0.04778725281357765, 0.08378391712903976, 0.09194684773683548, 0.3580940067768097, 0.34259912371635437, 0.07846498489379883, 0.11075633764266968, 0.10656090080738068, 0.21385900676250458, 0.5311782956123352, 0.1292661428451538, 0.58072829246521, 0.1455209106206894, -0.0663343295454979, -0.019351042807102203, -0.24416673183441162, 0.1489509642124176, 0.49089285731315613, 0.13517732918262482, 0.05381626635789871, 0.037483759224414825, -0.08207698166370392, -0.27816081047058105, 0.07523506879806519, -0.0076915062963962555, -0.05447138100862503, -0.46967214345932007, -0.33327409625053406, 0.2758778929710388, 0.2949088215827942, -0.14097784459590912, -0.15513485670089722, -0.24002999067306519, -0.34685903787612915, 0.28352147340774536, 0.2437196969985962, 1.0874789953231812, 0.24174873530864716, -0.02210628055036068, -0.0018389469478279352, -0.15207736194133759, 0.9227550029754639, -0.2762623131275177, 0.4379432797431946, -0.39078086614608765, -0.20731131732463837, -0.04219413176178932, -0.16487115621566772, 0.12216304242610931, 0.02207280695438385, -0.38290154933929443, 0.309418261051178, 0.3011412024497986, 0.020166663452982903, -0.1154511570930481, 0.38712069392204285, -0.21533192694187164, -0.14622898399829865, -0.2855426073074341, -0.02860606648027897, -0.30701208114624023, 0.4314887523651123, -0.02508620172739029, -0.058468692004680634, -0.21924768388271332, -0.21921731531620026, -0.3474550247192383, 0.2333538681268692, 0.03732389584183693, -0.03378737345337868, 0.2643851935863495, -0.2253139764070511, -0.05423035845160484, 0.05196896567940712, 0.3515734076499939, -0.08515332639217377, -0.10732724517583847, -0.07103361934423447, -0.18470562994480133, -0.14279155433177948, 0.12849219143390656, 0.014371493831276894, 0.23452404141426086, -0.08470719307661057, 0.06245047226548195, 0.06788531690835953, -0.09753070771694183, -0.3518665134906769, -0.11455526947975159, -0.06113250553607941, -0.20195890963077545, -0.15386658906936646, -0.04333491995930672, 0.1310223639011383, -0.0852956548333168, -0.03876195475459099, 0.002993806963786483, -0.14414682984352112, -0.22105346620082855, 0.27000850439071655, -0.06912074238061905, -0.3818735182285309, -0.15728218853473663, 0.352618932723999, 0.44284096360206604, 0.03658917173743248, 0.36675116419792175, -0.15284796059131622, -0.07765895128250122, -0.15032604336738586, -0.15030474960803986, -0.3644890785217285, -0.20347844064235687, 0.2349226474761963, -0.23662813007831573, -0.12594081461429596, 0.14240814745426178, 0.4553373157978058, 0.18074458837509155, 0.3331873416900635, -0.0611843466758728, -0.510886549949646, -0.00819066446274519, 0.17526467144489288, 0.03207870200276375, 0.05828062444925308, -0.058064717799425125, 0.5259192585945129, -0.25158560276031494, 0.21031662821769714, -0.19090056419372559, -0.1488889902830124, -0.44389599561691284, 0.3342975378036499, 0.13203087449073792, -0.043771591037511826, -0.009572383016347885, 0.0000543913192814216, -0.16391301155090332, 0.31393057107925415, -0.06522489339113235, -0.005955784115940332, -0.06145972013473511, 0.2064998894929886, 0.17853792011737823, -0.1087699756026268, -0.174637109041214, 0.12480117380619049, 0.07677895575761795, 0.03751732036471367, 0.034636665135622025, 0.24733790755271912, 0.04405101016163826, 0.28184279799461365, -0.3356453478336334, 0.11022801697254181, -0.0783042162656784, 0.2276509404182434, -0.2985171675682068, 0.3604101240634918, -0.05913544073700905, 0.3064751625061035, -0.10010342299938202, 0.015594580210745335, 0.15056976675987244, -0.03606424853205681, 0.048481643199920654, 0.09732730686664581, 0.07677587866783142, -0.17980925738811493, 0.14732134342193604, 0.18416376411914825, -0.08916469663381577, 0.6363031268119812, -0.0754886269569397, -0.19107241928577423, 0.10431274771690369, 0.045415762811899185, -0.08597462624311447, 0.0015468659112229943, 0.222201868891716, 0.025047052651643753, -0.056129686534404755, 0.5138841867446899, 0.08978401124477386, -0.1323106288909912, -0.13743694126605988, 0.11166290193796158, 0.3375561237335205, -0.07570730894804001, 0.10394448786973953, 0.19666709005832672, -0.2053917795419693, 0.09556069225072861, 0.3046140670776367, -0.2812727093696594, 0.15195699036121368, -0.000051470309699652717, 0.027243077754974365, 0.4766053855419159, 0.1611965298652649, 0.3706956207752228, -0.024317413568496704, -0.11332197487354279, 0.029959410429000854, 0.6413366198539734, 0.28227749466896057, 0.3716781735420227, 0.28426894545555115, 0.38737183809280396, -0.27664798498153687, 0.028786512091755867, -0.36788806319236755, -0.17549045383930206, -0.06385763734579086, -0.14051619172096252, -0.4606216251850128, -0.048443522304296494, -0.17240498960018158, 0.20627911388874054, -0.1402350813150406, 0.12604166567325592, 0.45997321605682373, -0.0754479318857193, -0.20757713913917542, -0.3750883936882019, 0.12007821351289749, 0.028892816975712776, 0.16698631644248962, -0.23685862123966217, 0.3165041208267212, -0.07666324079036713, -0.08664529770612717, 0.3203832805156708, 0.5079479813575745, 0.4745161831378937, 0.30628955364227295, -0.057672660797834396, -0.20901250839233398, 0.08152627944946289, 0.11034237593412399, 0.22965645790100098, 0.7636836171150208, 0.29754242300987244, -0.25712811946868896, 0.38817867636680603, -0.06476613879203796, -0.07674138247966766, 0.4682406485080719, 0.07057199627161026, -0.08752609044313431, -0.022166138514876366, 0.40950265526771545, -0.07317901402711868, 0.11578691750764847, 0.021555408835411072, 0.13730399310588837, -0.2554469108581543, -0.27098944783210754, 0.06821254640817642, -0.27079305052757263, 0.3050960302352905, 0.013167486526072025, 0.022203581407666206, -0.3044309914112091, 0.6118449568748474, 0.38975849747657776, -0.12752358615398407, -0.2048281580209732, -0.07230476289987564, -0.4306270182132721, 0.34660279750823975, -0.4226308763027191, -0.239922434091568, -0.2996874153614044, 0.5030856728553772, -0.22747178375720978, 0.40453994274139404, -0.287725031375885, 0.3795645534992218, -0.06517502665519714, -0.03167552500963211, -0.22983898222446442, -0.0672011449933052, -0.1227160096168518, -0.2966387867927551, 0.020048856735229492, -0.2617558240890503, 0.22259119153022766, 0.07699605077505112, -0.13086310029029846, 0.0015137623995542526, -0.016435738652944565, 0.016017897054553032, 0.4078022539615631, -0.012907546013593674, 0.5090907216072083, 0.3904215395450592, -0.14212581515312195, -0.150312140583992, -0.0055238306522369385, -0.3197629153728485, -0.31113988161087036, 0.4149196743965149, 0.2139553278684616, 0.35975536704063416, -0.17412549257278442, 0.6665087342262268, -0.0674726590514183, 0.31923648715019226, 0.05682764947414398, 0.02263069897890091, -0.25127550959587097, -0.05701703578233719, -0.24597547948360443, 0.03582845628261566, 0.06754431128501892, 0.45341363549232483, 0.15364199876785278, -0.027757937088608742, -0.24259701371192932, -0.34912988543510437, 0.4797302782535553, -0.3962058126926422, 0.02929595485329628, -0.006679664831608534, 0.18370367586612701, 0.24500302970409393, -0.22634081542491913, -0.7466571927070618, -0.070057712495327, 0.25677376985549927, -0.21241332590579987, 0.050724096596241, -0.02079630270600319, -0.0501658171415329, -0.22436003386974335, -0.12880520522594452, -0.233914315700531, 0.21499398350715637, -0.41724690794944763, 0.2863387167453766, -0.12938715517520905 ]
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
[ -0.21565428376197815, 0.12347615510225296, 0.027687575668096542, 0.4813612401485443, -0.011550177820026875, 0.09883323311805725, -0.1319705694913864, 0.36858728528022766, -0.051696643233299255, 0.18515697121620178, 0.0752982422709465, 0.13688449561595917, 0.08542843908071518, -0.13447429239749908, -0.032238978892564774, 0.025298796594142914, -0.09251768887042999, 0.19295059144496918, 0.03854679688811302, -0.09255436062812805, -0.17508384585380554, 0.41649696230888367, -0.3340628445148468, 0.007659576367586851, -0.21158111095428467, -0.10277745127677917, 0.08266913145780563, 0.42487096786499023, 0.005414085928350687, -0.3878881633281708, 0.017097333446145058, -0.015491513535380363, 0.07717103511095047, 0.2035766839981079, -0.00012377271195873618, 0.052519019693136215, 0.6025202870368958, -0.2609117329120636, -0.15160629153251648, -0.24732503294944763, -0.3983467221260071, -0.3468353748321533, -0.1532551497220993, -0.30981290340423584, 0.043405696749687195, -0.13741986453533173, 0.27026912569999695, -0.21010436117649078, 0.30159035325050354, 0.3589797914028168, 0.10676814615726471, 0.17280328273773193, -0.06732206046581268, 0.13837482035160065, 0.27825385332107544, 0.06059809774160385, -0.0570702888071537, -0.07430313527584076, -0.030549094080924988, -0.21056655049324036, -0.09983690083026886, 0.3154500126838684, -0.29578322172164917, -0.004714201204478741, 0.17416711151599884, 0.09821542352437973, 0.3729248046875, -0.204134002327919, 0.1727689653635025, 0.2954416871070862, 0.5296699404716492, 0.13004104793071747, -0.21195340156555176, -0.4479368031024933, -0.15491196513175964, 0.117674320936203, 0.4065712094306946, 0.4266495406627655, -0.20510540902614594, -0.0010265090968459845, -0.26339825987815857, -0.03466137871146202, -0.19299177825450897, 0.06627204269170761, 0.24316449463367462, 0.24165336787700653, 0.08146437257528305, 0.06608112901449203, -0.02012070268392563, 0.0718880221247673, 0.09668935835361481, -0.2602766454219818, -0.1418871283531189, 0.09542859345674515, 0.029838792979717255, -0.07551881670951843, -0.11887253075838089, -0.07541254162788391, 0.20608893036842346, 0.20481669902801514, 0.3663531243801117, 0.019850632175803185, 0.20985360443592072, 0.13504061102867126, 0.33779868483543396, -0.07288730889558792, 0.04885515570640564, 0.42680636048316956, 0.20002351701259613, -0.035383667796850204, 0.08910273760557175, 0.013300261460244656, 0.2252694070339203, -0.27365803718566895, -0.15773573517799377, 0.17779532074928284, 0.24600405991077423, -0.49686169624328613, -0.45548269152641296, 0.1475425809621811, -0.4238388240337372, -0.27335843443870544, -0.25034913420677185, 0.2649309039115906, -0.3046392798423767, 0.36598506569862366, -0.13079877197742462, 0.08853164315223694, -0.3846737742424011, -0.10326589643955231, -0.13025547564029694, 0.16457194089889526, -0.1479644626379013, 0.20909665524959564, 0.17410166561603546, -0.1450197994709015, 0.36790353059768677, -0.17715902626514435, -0.06438592076301575, -0.21380110085010529, 0.3670308291912079, -0.17966464161872864, 0.0469617024064064, 0.25024017691612244, 0.10622447729110718, 0.17181405425071716, -0.11795474588871002, -0.29181575775146484, -0.28508204221725464, 0.09585822373628616, -0.2795941233634949, -0.39303961396217346, 0.11864612996578217, 0.14444108307361603, -0.3303494155406952, -0.04221527278423309, 0.0714510977268219, 0.04952285438776016, 0.41477566957473755, -0.48032835125923157, -0.09707242250442505, -0.22827786207199097, -0.15888935327529907, -0.12854725122451782, -0.2621921896934509, 0.6329420208930969, 0.13807933032512665, -0.10230572521686554, 0.2826194763183594, -0.029114635661244392, 0.5553821921348572, 0.35930150747299194, 0.06276644766330719, 0.128160297870636, -0.193980410695076, 0.4745140075683594, 0.11766660213470459, -0.41009339690208435, -0.1508030891418457, 0.4691442549228668, -0.041333843022584915, 0.3392701745033264, 0.43024930357933044, 0.10246618092060089, 0.2844761610031128, -0.17926861345767975, -0.042457375675439835, 0.31460902094841003, 0.06323031336069107, -0.028442680835723877, -0.4521333873271942, -0.14667314291000366, 0.43677619099617004, 0.1048913225531578, 0.1847425103187561, 0.010159561410546303, 0.14154885709285736, 0.4389922022819519, 0.31315934658050537, -0.10660909116268158, 0.10542189329862595, 0.2982521057128906, -0.44767990708351135, 0.1695099025964737, 0.1006329208612442, -0.10743016004562378, -0.3479069173336029, -0.07296053320169449, -0.3559930622577667, 0.14380571246147156, 0.0009279443183913827, -0.2238154113292694, -0.17331533133983612, -0.39921292662620544, -0.07715792953968048, -0.003457051934674382, -0.022262124344706535, 0.36639970541000366, 0.28739404678344727, 0.15451817214488983, -0.05905536189675331, 0.36677902936935425, -0.033696725964546204, 0.1865333467721939, -0.8968738317489624, 0.23865024745464325, -0.15186846256256104, -0.22171463072299957, 0.4185650050640106, 0.31582826375961304, 0.19127628207206726, -0.19332468509674072, -0.07729366421699524, 0.38104313611984253, -0.10586770623922348, 0.4759340286254883, -0.23418711125850677, 0.0746661052107811, 0.18912164866924286, -0.31206458806991577, -0.020262205973267555, 0.40568751096725464, 0.0988728255033493, -0.11943958699703217, -0.03732908517122269, 0.4227374792098999, -0.3420753479003906, 0.1942281723022461, 0.14112333953380585, -0.0657077357172966, 0.063754141330719, -0.267714262008667, -0.2454354614019394, -0.5215454697608948, 0.6221480965614319, -0.06730528175830841, -0.26908349990844727, 0.04679862782359123, -0.25621914863586426, 0.02593125030398369, 0.5729617476463318, 0.002075602300465107, -0.08300840854644775, -0.06289930641651154, 0.06719449162483215, -0.0630129799246788, -0.018116960301995277, 0.18459776043891907, 0.4571347236633301, 0.09094741195440292, -0.05557713657617569, 0.06711209565401077, -0.13181912899017334, -0.45842882990837097, 0.1281435191631317, -0.04832867905497551, -0.027972210198640823, 0.5724953413009644, -0.011887330561876297, 0.06637772172689438, -0.3477862775325775, -0.021248703822493553, 0.11456356942653656, 0.13852635025978088, -0.3325786888599396, -0.09293337166309357, -0.329971045255661, -0.40572041273117065, -0.3795689046382904, -0.07616916298866272, -0.1871165633201599, -0.38549700379371643, 0.1835038810968399, 0.21754328906536102, -0.2500668466091156, 0.02811366692185402, -0.19778531789779663, 0.0052064573392271996, -0.34196656942367554, 0.0488547757267952, -0.1312798708677292, -0.19375477731227875, -0.21058040857315063, -0.06514135748147964, 0.27047839760780334, 0.3898128271102905, 0.220295712351799, -0.5454668402671814, -0.21008794009685516, -0.04477061703801155, -0.2075357288122177, 0.020225610584020615, -0.05990811809897423, -0.17760729789733887, 0.09295949339866638, -0.1559802144765854, 0.5214743614196777, -0.1610417366027832, 0.14776457846164703, -0.12871915102005005, -0.270849347114563, 0.12490467727184296, 0.22166502475738525, -0.1444789469242096, -0.22062832117080688, -0.5866004228591919, -0.1790795922279358, -0.45689481496810913, -0.09897767007350922, 0.2818934917449951, 0.2948870062828064, 0.042354509234428406, -0.16215060651302338, -0.10634417831897736, -0.21439439058303833, 0.3561648726463318, -0.16313152015209198, 0.05585523322224617, 0.3234180212020874, -0.18394799530506134, -0.27022501826286316, -0.021707231178879738, 0.13760602474212646, -0.05147722363471985, 0.141585111618042, -0.4621351659297943, -0.059683240950107574, -0.18343858420848846, -0.15344783663749695, -0.2829677164554596, -0.11173205822706223, 0.4699302911758423, -0.03412957116961479, -0.004534973297268152, -0.05661594122648239, -0.26839593052864075, 0.017017748206853867, 0.054796528071165085, 0.30072399973869324, -0.128115713596344, 0.23388998210430145, -0.09957718104124069, 0.41609007120132446, 0.5284245610237122, 0.18462376296520233, -0.0677548199892044, 0.07574277371168137, 0.2727704644203186, 0.16633442044258118, -0.016635309904813766, 0.3361004889011383, -0.02377231977880001, -0.09101875126361847, 0.43315330147743225, -0.0023465310223400593, -0.09685899317264557, -0.14553619921207428, -0.24594701826572418, -0.34901535511016846, -0.1934383660554886, 0.1462906002998352, -0.16669544577598572, 0.14870807528495789, 0.22088977694511414, 0.30664387345314026, -0.03872033953666687, -0.5155574083328247, 0.024541711434721947, 0.45322540402412415, 0.32034775614738464, 0.18500791490077972, -0.4796754717826843, 0.09125629812479019, -0.42474207282066345, 0.2979332506656647, 0.14495420455932617, 0.3121427297592163, -0.1828320324420929, -0.1499021053314209, 0.1062329038977623, -0.2515309751033783, 0.23255720734596252, -0.41618290543556213, 0.01984391361474991, 0.059046629816293716, -0.006551284808665514, -0.3043076694011688, -0.17955151200294495, -0.26142266392707825, 0.40534546971321106, 0.386575311422348, 0.43123412132263184, -0.38310471177101135, -0.16495147347450256, 0.09735047817230225, 0.005135773215442896, -0.10864173620939255, -0.08253532648086548, -0.2994462549686432, -0.3675910532474518, -0.22960813343524933, -0.2945684790611267, -0.08849803358316422, 0.26921507716178894, -0.09233544021844864, -0.027086332440376282, 0.03468587249517441, 0.10009197890758514, 0.3609338402748108, 0.2708033323287964, 0.08658643811941147, -0.10598386824131012, 0.07422317564487457, 0.15665042400360107, 0.5734145045280457, 0.11814238131046295, 0.5895763039588928, 0.06899487227201462, -0.181111142039299, 0.024747006595134735, -0.43041327595710754, 0.19162803888320923, 0.4658256769180298, 0.08184243738651276, 0.044300489127635956, 0.12616175413131714, -0.024200286716222763, -0.4446396231651306, -0.05971948802471161, 0.16697156429290771, -0.08109409362077713, -0.40408650040626526, -0.42042428255081177, 0.2781201899051666, 0.20545512437820435, -0.1348200887441635, -0.16485343873500824, -0.13094420731067657, -0.25973016023635864, 0.3258019685745239, 0.13428537547588348, 1.0393661260604858, 0.26532912254333496, 0.07107430696487427, -0.06675097346305847, -0.1936645805835724, 0.7017490267753601, -0.23345685005187988, 0.46105635166168213, -0.4527300298213959, -0.1818924844264984, -0.0957932397723198, -0.18577446043491364, 0.28010982275009155, 0.1366838812828064, -0.21743790805339813, 0.3387283384799957, 0.20434336364269257, -0.02720686048269272, -0.10105747729539871, 0.4760473668575287, -0.2647891640663147, -0.03463730588555336, -0.10286465287208557, -0.052990637719631195, -0.18343371152877808, 0.456830233335495, -0.0012250016443431377, -0.13717572391033173, -0.16188181936740875, -0.2715013027191162, -0.35441941022872925, 0.286967009305954, 0.11578188091516495, -0.07829167693853378, 0.18359075486660004, -0.19995829463005066, 0.05459512397646904, 0.20607008039951324, 0.22772066295146942, -0.08420197665691376, -0.038650624454021454, -0.0967053771018982, -0.1803741157054901, -0.199497789144516, 0.18710355460643768, 0.1252504587173462, 0.347232848405838, -0.11429060250520706, 0.06230473145842552, 0.18676583468914032, -0.15038491785526276, -0.2784910798072815, -0.32652369141578674, 0.020361097529530525, -0.2225656509399414, -0.18651767075061798, -0.1451815962791443, 0.015972046181559563, -0.16698461771011353, -0.047240614891052246, -0.003171590156853199, -0.08467826247215271, -0.09634073078632355, 0.19067367911338806, -0.17764610052108765, -0.45024317502975464, -0.12551483511924744, 0.2783941924571991, 0.3640978932380676, -0.03016940876841545, 0.4310389459133148, -0.09358812123537064, -0.12448455393314362, -0.19571982324123383, -0.1521316021680832, -0.23456506431102753, -0.40158259868621826, 0.17225410044193268, -0.20613956451416016, -0.10457728058099747, 0.17576442658901215, 0.24396368861198425, 0.08783140778541565, 0.19602708518505096, -0.12787166237831116, -0.6653318405151367, 0.12647399306297302, 0.1522240787744522, 0.21884332597255707, -0.01905340515077114, -0.048901405185461044, 0.3229835629463196, -0.23107978701591492, 0.16286778450012207, -0.18060225248336792, -0.07511365413665771, -0.30882537364959717, 0.3002736270427704, 0.16600309312343597, 0.040770869702100754, -0.10946870595216751, 0.029191315174102783, -0.13556687533855438, 0.24872848391532898, -0.17746102809906006, 0.039268046617507935, -0.14119020104408264, 0.22275471687316895, 0.15719321370124817, -0.059652842581272125, -0.10897006839513779, 0.019748304039239883, 0.09330267459154129, -0.02179856039583683, 0.03392081335186958, 0.29959678649902344, -0.038626529276371, 0.3489004075527191, -0.3989141583442688, 0.14669689536094666, -0.09037598967552185, 0.2649860680103302, -0.3462004065513611, 0.3897472023963928, -0.20149537920951843, 0.2406027466058731, -0.03934541717171669, 0.02292019873857498, -0.09879719465970993, -0.03177813068032265, 0.12119586765766144, 0.06413544714450836, 0.31527402997016907, -0.15873879194259644, 0.21573196351528168, 0.2088233083486557, -0.03734006732702255, 0.7320029139518738, -0.12561549246311188, -0.11623475700616837, 0.06201779097318649, 0.05478941649198532, -0.11067062616348267, -0.017984328791499138, 0.1833477020263672, -0.08880683779716492, -0.08237114548683167, 0.3635832667350769, 0.19489268958568573, -0.25308409333229065, -0.18858081102371216, 0.10765570402145386, 0.4120692312717438, -0.08131398260593414, 0.09422257542610168, 0.2386835813522339, -0.011601055972278118, 0.017933562397956848, 0.19761031866073608, -0.1533752679824829, 0.07994953542947769, 0.11493546515703201, 0.09460413455963135, 0.5233208537101746, 0.1414514183998108, 0.4351024925708771, -0.23179787397384644, -0.12785860896110535, 0.018449854105710983, 0.6700650453567505, 0.19276079535484314, 0.4858522117137909, 0.2512127459049225, 0.3217865228652954, -0.2963058352470398, 0.1505880504846573, -0.26697838306427, -0.17418241500854492, -0.03776205703616142, -0.12803590297698975, -0.4697096049785614, -0.12692955136299133, -0.2004304826259613, 0.06337606906890869, -0.07095862925052643, 0.2730650007724762, 0.3169323801994324, -0.05591943860054016, -0.19864153861999512, -0.2947254776954651, 0.17275680601596832, -0.15478487312793732, 0.1828126609325409, -0.3227280080318451, 0.2761794626712799, 0.10261641442775726, -0.03287261724472046, 0.3759279251098633, 0.5440007448196411, 0.6268026232719421, 0.40504395961761475, -0.10989481955766678, -0.12854227423667908, -0.054338403046131134, 0.0478566437959671, 0.1882709115743637, 0.865935206413269, 0.21049334108829498, -0.1732793003320694, 0.3700045347213745, -0.039264701306819916, -0.08896409720182419, 0.33758389949798584, 0.16805610060691833, -0.0406322218477726, 0.047046635299921036, 0.3882299065589905, -0.19839532673358917, -0.00210387515835464, -0.022414233535528183, 0.03416965901851654, -0.41420596837997437, -0.26651161909103394, 0.1787838190793991, -0.1447102129459381, 0.2802106738090515, -0.15230464935302734, 0.019083596765995026, -0.17731137573719025, 0.5529253482818604, 0.47746339440345764, -0.17004431784152985, -0.09665437042713165, -0.022314919158816338, -0.5047860145568848, 0.5072606801986694, -0.4191436767578125, -0.31876417994499207, -0.2432268261909485, 0.3742891252040863, -0.19792024791240692, 0.2628782093524933, -0.09300830960273743, 0.4435424208641052, -0.13743454217910767, 0.10049083828926086, -0.21290765702724457, -0.019861897453665733, 0.014265201054513454, -0.15694935619831085, -0.09848302602767944, -0.2454884797334671, 0.2524297535419464, 0.040925342589616776, -0.07209344953298569, -0.010972476564347744, 0.15720804035663605, -0.0024537364952266216, 0.2743946313858032, -0.026171546429395676, 0.4778503477573395, 0.44526010751724243, -0.2909838855266571, -0.21393759548664093, -0.11031880974769592, -0.23036077618598938, -0.28157639503479004, 0.40109983086586, 0.22582021355628967, 0.36033618450164795, -0.20492464303970337, 0.5031102299690247, -0.092566579580307, 0.26268622279167175, -0.15710875391960144, 0.14505566656589508, -0.21884785592556, 0.03149876743555069, -0.23324857652187347, -0.016331445425748825, 0.15726901590824127, 0.4576330780982971, 0.18879978358745575, -0.12539458274841309, -0.23119832575321198, -0.3733322024345398, 0.4127679169178009, -0.39108720421791077, -0.07988396286964417, -0.017234787344932556, 0.19397448003292084, 0.12862266600131989, -0.041532065719366074, -0.6443721055984497, -0.05958766117691994, 0.22831961512565613, -0.11385010927915573, 0.02982274629175663, -0.06642250716686249, -0.15205010771751404, -0.3520394265651703, -0.034354615956544876, -0.2507988214492798, 0.27073603868484497, -0.41924959421157837, 0.40055137872695923, -0.10350929200649261 ]
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.
[ -0.15917639434337616, 0.0613841712474823, 0.06137017905712128, 0.3985083997249603, -0.09111960232257843, 0.06813949346542358, -0.15415671467781067, 0.37438714504241943, -0.05398888885974884, 0.1291232854127884, 0.09491007030010223, 0.1696779429912567, 0.05089188739657402, -0.07023661583662033, -0.029942329972982407, -0.03138881176710129, -0.04626782238483429, 0.20410291850566864, 0.04388732463121414, 0.0010295415995642543, -0.019636930897831917, 0.44040238857269287, -0.30174973607063293, -0.022966843098402023, -0.2469935566186905, -0.1683308333158493, -0.042130157351493835, 0.32819750905036926, 0.003955312073230743, -0.3751760423183441, 0.24395912885665894, -0.07514718919992447, 0.13529184460639954, 0.19081078469753265, -0.00012187100219307467, -0.013404469937086105, 0.5785961747169495, -0.2156340479850769, -0.20625454187393188, -0.25492191314697266, -0.40470558404922485, -0.36227986216545105, -0.0836014524102211, -0.23552004992961884, -0.021686792373657227, -0.055038027465343475, 0.1720554232597351, -0.08597995340824127, 0.27252909541130066, 0.28542202711105347, 0.11910402774810791, 0.15230627357959747, -0.11947862058877945, 0.03129195421934128, 0.13146868348121643, 0.11148048937320709, -0.03822549059987068, -0.09602130949497223, -0.08761485666036606, -0.2862390875816345, -0.10378145426511765, 0.43996769189834595, -0.1971522867679596, 0.15519708395004272, 0.22587357461452484, 0.11542966961860657, 0.40087375044822693, -0.114561066031456, 0.03758140653371811, 0.3904353082180023, 0.5094951391220093, 0.16033302247524261, -0.2086891531944275, -0.5066429376602173, -0.05045638233423233, 0.19382120668888092, 0.41565215587615967, 0.36488479375839233, -0.23909980058670044, 0.008471298962831497, -0.41110947728157043, -0.07393155246973038, -0.0905098021030426, 0.10173982381820679, 0.2668371796607971, 0.1985688954591751, 0.11837206035852432, 0.006740066688507795, 0.12967176735401154, -0.04109877720475197, 0.11463364958763123, -0.2442883402109146, -0.13999295234680176, 0.11282779276371002, -0.183878555893898, -0.19307436048984528, -0.14485175907611847, -0.017406532540917397, 0.21208812296390533, 0.21555636823177338, 0.35182857513427734, -0.05405342951416969, 0.22565507888793945, 0.11196926981210709, 0.4283163845539093, 0.047907084226608276, -0.01073585357517004, 0.5196449160575867, 0.0832175463438034, 0.003371703904122114, 0.11515562981367111, 0.05967355892062187, 0.265980988740921, -0.21909572184085846, -0.14518532156944275, 0.11991679668426514, 0.15946316719055176, -0.5667431354522705, -0.4398611783981323, 0.10815716534852982, -0.5615881085395813, -0.24343453347682953, -0.09248387068510056, 0.17566193640232086, -0.3614133596420288, 0.33200380206108093, -0.15387770533561707, 0.047798965126276016, -0.3616383969783783, -0.059445932507514954, -0.20667265355587006, 0.17067916691303253, -0.1974988877773285, 0.20532430708408356, 0.2014605849981308, -0.15544484555721283, 0.4026421308517456, -0.10782964527606964, -0.058672163635492325, -0.19689887762069702, 0.40113580226898193, -0.10814110934734344, 0.020010463893413544, 0.15724490582942963, -0.02918046899139881, 0.16998465359210968, -0.0008437033393420279, -0.21693295240402222, -0.38861724734306335, 0.07954568415880203, -0.223024383187294, -0.5368486642837524, 0.10468994081020355, 0.15302309393882751, -0.44328877329826355, -0.06129140406847, 0.09627190977334976, 0.21096163988113403, 0.36438462138175964, -0.33245113492012024, -0.09742452204227448, -0.22020068764686584, -0.12947265803813934, -0.054568421095609665, -0.20552793145179749, 0.495089590549469, 0.14020086824893951, -0.2660529911518097, 0.13045355677604675, -0.018282469362020493, 0.533545196056366, 0.34930914640426636, 0.03433435410261154, 0.09644102305173874, -0.19258973002433777, 0.4875858426094055, 0.13571123778820038, -0.35429832339286804, -0.16301178932189941, 0.5454863905906677, 0.02868376299738884, 0.2807048261165619, 0.25120458006858826, 0.11916995048522949, 0.1942964345216751, -0.0937625840306282, 0.010771148838102818, 0.4080868065357208, -0.021601947024464607, 0.03519253432750702, -0.45149222016334534, -0.3096882104873657, 0.4148995876312256, 0.20669463276863098, 0.11799579858779907, -0.04226293787360191, -0.0019412139663472772, 0.49649927020072937, 0.3046288788318634, -0.1359483152627945, 0.053966037929058075, 0.29197070002555847, -0.3540268540382385, 0.12886853516101837, 0.038535408675670624, -0.11504382640123367, -0.3697325885295868, -0.0230215135961771, -0.2869825065135956, 0.28181374073028564, 0.10698973387479782, -0.25840261578559875, -0.19961754977703094, -0.4155748188495636, -0.025226671248674393, -0.0299625713378191, 0.0107140326872468, 0.28152215480804443, 0.3974580466747284, 0.05119474232196808, 0.06851131469011307, 0.4284546375274658, -0.06218886747956276, 0.17390738427639008, -0.7967963218688965, 0.2320181131362915, -0.11876522749662399, -0.21117736399173737, 0.5127646327018738, 0.34133416414260864, 0.2236994355916977, -0.16931122541427612, -0.09573232382535934, 0.43656110763549805, -0.10373667627573013, 0.40195849537849426, -0.17406702041625977, 0.010552867315709591, 0.20788688957691193, -0.29545918107032776, 0.018217824399471283, 0.37473681569099426, -0.028403757140040398, -0.06868940591812134, -0.1335936039686203, 0.416767418384552, -0.33828744292259216, 0.16545313596725464, 0.06876230984926224, 0.01500897016376257, 0.07816552370786667, -0.23879611492156982, -0.19139669835567474, -0.4554643929004669, 0.6415621638298035, -0.09513678401708603, -0.32226401567459106, 0.04229302704334259, -0.201063334941864, 0.03303178399801254, 0.4396530091762543, 0.02186592109501362, -0.09082556515932083, -0.026410141959786415, 0.028389334678649902, -0.160865917801857, 0.09572149068117142, 0.275053471326828, 0.513839602470398, 0.12376830726861954, 0.02331407554447651, 0.015095884911715984, -0.16777153313159943, -0.3314583897590637, 0.12296842038631439, -0.06294000148773193, 0.027277695015072823, 0.3670690655708313, -0.0033676226157695055, -0.031909964978694916, -0.3108043968677521, -0.012369398027658463, 0.15956227481365204, 0.24881479144096375, -0.32820868492126465, -0.1341906636953354, -0.3583877980709076, -0.4044027626514435, -0.34532350301742554, -0.1350657343864441, -0.2826967239379883, -0.3694404363632202, 0.20206350088119507, 0.1999654918909073, -0.16825909912586212, 0.13886958360671997, -0.16680829226970673, 0.04333510622382164, -0.3156353533267975, -0.05282656103372574, -0.018515927717089653, -0.2104291021823883, -0.27342864871025085, -0.026242367923259735, 0.2240419089794159, 0.4267178475856781, 0.242521733045578, -0.39223501086235046, -0.15776918828487396, 0.04673954099416733, -0.25808802247047424, 0.06250391155481339, -0.05618460476398468, -0.15136493742465973, 0.1768350601196289, -0.13862520456314087, 0.3756524920463562, -0.22181439399719238, 0.10164055228233337, -0.18995557725429535, -0.3374049663543701, 0.027317050844430923, 0.1791120022535324, 0.035421404987573624, -0.19235365092754364, -0.5341887474060059, -0.2685714662075043, -0.36351487040519714, -0.062366288155317307, 0.28557682037353516, 0.24182604253292084, 0.03172805532813072, -0.12043987959623337, -0.13270403444766998, -0.17049212753772736, 0.2503780722618103, -0.30170774459838867, -0.05889030918478966, 0.24042052030563354, -0.09786730259656906, -0.16751530766487122, -0.031470347195863724, 0.0028388393111526966, -0.05066511780023575, 0.12287355959415436, -0.5468324422836304, -0.10958743095397949, -0.20560334622859955, -0.07945793867111206, -0.3269422650337219, -0.25753456354141235, 0.5263425707817078, -0.07347506284713745, -0.024189038202166557, 0.013537646271288395, -0.3449535071849823, 0.0683690682053566, 0.10190682858228683, 0.5138319730758667, -0.16110704839229584, 0.18403847515583038, -0.014320788905024529, 0.5651073455810547, 0.5774591565132141, 0.07665672153234482, -0.11549903452396393, 0.17679168283939362, 0.09244528412818909, 0.1740480363368988, 0.009828051552176476, 0.3770267069339752, 0.02252894639968872, -0.10745499283075333, 0.4157788157463074, 0.08360448479652405, -0.19980140030384064, -0.11438950896263123, -0.15785172581672668, -0.31042400002479553, -0.1871033012866974, 0.16156162321567535, -0.22997477650642395, 0.200632244348526, 0.20180676877498627, 0.23883280158042908, -0.1851041615009308, -0.5294701457023621, -0.012654189951717854, 0.35910794138908386, 0.22725659608840942, 0.17686016857624054, -0.47010335326194763, 0.1427343785762787, -0.4302375316619873, 0.29585516452789307, 0.21121416985988617, 0.2849108874797821, -0.21085679531097412, -0.15540389716625214, 0.0810379683971405, -0.19880688190460205, 0.295551598072052, -0.40564268827438354, -0.015019990503787994, -0.046211279928684235, -0.031895462423563004, -0.3422420620918274, -0.2017231434583664, -0.13815274834632874, 0.3731078505516052, 0.5086653828620911, 0.4213862121105194, -0.427228182554245, -0.12459463626146317, 0.16328127682209015, -0.05940474569797516, -0.13091090321540833, -0.16943760216236115, -0.2525332272052765, -0.3333899676799774, -0.17163117229938507, -0.14729198813438416, -0.12542997300624847, 0.3001165986061096, -0.02195604145526886, -0.0020847776904702187, 0.1447325348854065, 0.1252201944589615, 0.3355752229690552, 0.25718235969543457, 0.16852502524852753, -0.03306203708052635, 0.10766857862472534, 0.27886953949928284, 0.5404438376426697, 0.08358275890350342, 0.5131436586380005, 0.10595914721488953, -0.05891058221459389, -0.02509249746799469, -0.3047672510147095, 0.07861294597387314, 0.4301782250404358, 0.1747911423444748, 0.08524955064058304, 0.14247451722621918, -0.02743455581367016, -0.46311163902282715, -0.02574387565255165, 0.1107131689786911, -0.07656635344028473, -0.37168195843696594, -0.39093032479286194, 0.25883957743644714, 0.2693789601325989, -0.14291100203990936, -0.14606873691082, -0.19557955861091614, -0.25535714626312256, 0.2318812906742096, 0.15940804779529572, 0.9950311183929443, 0.2600904107093811, -0.026021402329206467, -0.02760963700711727, -0.22299730777740479, 0.8679137825965881, -0.2555355727672577, 0.4108973741531372, -0.4282158315181732, -0.2512688636779785, -0.062348220497369766, -0.20891477167606354, 0.15235136449337006, 0.11817735433578491, -0.29969659447669983, 0.3315716087818146, 0.2687565088272095, -0.011667536571621895, -0.10332244634628296, 0.44522392749786377, -0.2453589141368866, -0.07309813052415848, -0.17734669148921967, 0.014801668003201485, -0.28842222690582275, 0.5008421540260315, -0.017722375690937042, -0.1215856745839119, -0.21846354007720947, -0.21833476424217224, -0.4154762327671051, 0.2596202492713928, 0.1806759536266327, 0.0051103150472044945, 0.27800053358078003, -0.18391935527324677, -0.0008593745878897607, 0.09157321602106094, 0.32835930585861206, -0.18273340165615082, -0.1748187243938446, 0.006974640302360058, -0.20243743062019348, -0.13968566060066223, 0.19291824102401733, 0.07448733597993851, 0.3397161364555359, -0.1254936158657074, 0.0013904923107475042, 0.05642960965633392, -0.12228957563638687, -0.3300777077674866, -0.3035987317562103, -0.1133047565817833, -0.20215804874897003, -0.18769226968288422, -0.08816833049058914, 0.07827648520469666, -0.10347459465265274, -0.0614963099360466, 0.04047756269574165, -0.04023399576544762, -0.21182693541049957, 0.23607872426509857, -0.0904947817325592, -0.330422580242157, -0.11680590361356735, 0.39779970049858093, 0.3670256733894348, -0.034113213419914246, 0.44290444254875183, -0.10410875827074051, -0.09311021119356155, -0.20934076607227325, -0.12918995320796967, -0.3109356164932251, -0.3431035876274109, 0.1643272191286087, -0.26846158504486084, -0.08667001128196716, 0.194707453250885, 0.29301825165748596, 0.23840956389904022, 0.3419274091720581, -0.0016040417831391096, -0.5291139483451843, 0.016075143590569496, 0.1371026635169983, -0.024978315457701683, 0.07527963817119598, -0.1436309963464737, 0.42183157801628113, -0.20619754493236542, 0.24041813611984253, -0.2192307859659195, -0.08254683762788773, -0.40393707156181335, 0.32245495915412903, 0.1132710874080658, -0.1017618402838707, -0.03889911621809006, 0.015868941321969032, -0.1192547157406807, 0.2172648161649704, -0.14230109751224518, 0.0013298835838213563, -0.1154848113656044, 0.20137101411819458, 0.1473950892686844, -0.081973597407341, -0.10277977585792542, 0.11400680989027023, 0.07142944633960724, 0.029016893357038498, -0.05398690328001976, 0.2069932073354721, -0.02211366593837738, 0.3232147693634033, -0.3484603762626648, 0.042323265224695206, -0.06938488781452179, 0.23659782111644745, -0.3490240275859833, 0.29365915060043335, -0.09470124542713165, 0.21986159682273865, -0.08393590152263641, -0.00015827368770260364, -0.03363371640443802, 0.01332928892225027, 0.08372699469327927, 0.07243608683347702, 0.18849563598632812, -0.17304058372974396, 0.10398305207490921, 0.22139424085617065, 0.021713465452194214, 0.5436680316925049, -0.07953498512506485, -0.10975425690412521, 0.07915406674146652, 0.10070711374282837, -0.12908624112606049, -0.03161069378256798, 0.2547113597393036, -0.030462659895420074, -0.12914377450942993, 0.35050180554389954, 0.14029528200626373, -0.16056475043296814, -0.20051006972789764, 0.10796533524990082, 0.39615753293037415, -0.1329914629459381, 0.12379759550094604, 0.29219943284988403, -0.1617976278066635, -0.01571020670235157, 0.2393847405910492, -0.263828307390213, 0.1348326951265335, 0.010988039895892143, 0.04418562352657318, 0.4613370895385742, 0.0975368320941925, 0.39205455780029297, -0.14451506733894348, -0.1371356099843979, -0.05386106297373772, 0.7125877737998962, 0.3032740354537964, 0.4075475037097931, 0.23052358627319336, 0.47541192173957825, -0.3398899435997009, -0.0037838707212358713, -0.35419607162475586, -0.18464218080043793, -0.03735377639532089, -0.10190197825431824, -0.5155590176582336, -0.10836471617221832, -0.23224732279777527, 0.12218460440635681, -0.08818589150905609, 0.14326104521751404, 0.4093429744243622, -0.07159333676099777, -0.18918584287166595, -0.4787326157093048, 0.053078584372997284, 0.014809314161539078, 0.1754535287618637, -0.32879072427749634, 0.1579628586769104, 0.036844056099653244, -0.1626705527305603, 0.39498794078826904, 0.5192058682441711, 0.5169353485107422, 0.30216145515441895, -0.14195634424686432, -0.21230348944664001, 0.05457650125026703, 0.15161576867103577, 0.23794251680374146, 0.7681570053100586, 0.27525025606155396, -0.2857542932033539, 0.39371541142463684, -0.05130109190940857, -0.08288402110338211, 0.40311741828918457, 0.05604325607419014, -0.0689929872751236, -0.026875507086515427, 0.4426216185092926, -0.13705366849899292, 0.11488331109285355, -0.004703478887677193, 0.11616142094135284, -0.2987505793571472, -0.31024661660194397, 0.10357487946748734, -0.17181900143623352, 0.2889469563961029, -0.07533535361289978, 0.034251775592565536, -0.3127897083759308, 0.5959122776985168, 0.43539923429489136, -0.2124715894460678, -0.2104102522134781, -0.1160922721028328, -0.4998016357421875, 0.3454272150993347, -0.4612220823764801, -0.23284472525119781, -0.18967832624912262, 0.47722503542900085, -0.20990623533725739, 0.2936575412750244, -0.17123892903327942, 0.46486976742744446, -0.1285218745470047, 0.07755838334560394, -0.2698145806789398, -0.05188264325261116, -0.08174455165863037, -0.1910950243473053, 0.021983984857797623, -0.22614005208015442, 0.19919104874134064, 0.10465013980865479, -0.08411658555269241, 0.03213822469115257, 0.05748937651515007, 0.02699802629649639, 0.42762646079063416, -0.08215101808309555, 0.5023900866508484, 0.4290034770965576, -0.14516238868236542, -0.21732516586780548, -0.08768822252750397, -0.288800448179245, -0.29746636748313904, 0.37864717841148376, 0.13621743023395538, 0.421730101108551, -0.1751047968864441, 0.6471749544143677, -0.1287107914686203, 0.3200821280479431, 0.019590793177485466, 0.011846323497593403, -0.32712477445602417, -0.03769777715206146, -0.2145235240459442, 0.02605842985212803, 0.15193980932235718, 0.47100573778152466, 0.17633001506328583, -0.017447058111429214, -0.18155966699123383, -0.30553656816482544, 0.42345279455184937, -0.31965968012809753, 0.034292470663785934, -0.02669629640877247, 0.1810697317123413, 0.2032093107700348, -0.11101249605417252, -0.6931033134460449, -0.07680679857730865, 0.2931540906429291, -0.11016223579645157, 0.08574342727661133, -0.035791605710983276, -0.0713716521859169, -0.29009515047073364, -0.022308049723505974, -0.2207995057106018, 0.23416641354560852, -0.3639991581439972, 0.3844619691371918, -0.08561233431100845 ]
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.
[ -0.10609117150306702, 0.17192895710468292, 0.08614687621593475, 0.37152907252311707, -0.07061878591775894, 0.19752028584480286, -0.08851543813943863, 0.3589628338813782, -0.07059337943792343, 0.13236770033836365, 0.01096312329173088, 0.19261184334754944, 0.01973343826830387, -0.13543295860290527, -0.06135297939181328, 0.0010895621962845325, -0.049684617668390274, 0.1512274146080017, 0.08074230700731277, -0.0007667046738788486, -0.019083166494965553, 0.43668195605278015, -0.35079941153526306, -0.06492381542921066, -0.2797684669494629, -0.17144761979579926, 0.005060211289674044, 0.3068116009235382, 0.06779024749994278, -0.3383015990257263, 0.24676108360290527, -0.08211202174425125, 0.10468740016222, 0.25874507427215576, -0.0001232282374985516, 0.034662846475839615, 0.6713200807571411, -0.23092050850391388, -0.19163355231285095, -0.3499613106250763, -0.33645811676979065, -0.2255391925573349, -0.007184400223195553, -0.21592767536640167, 0.00239318935200572, -0.07125561684370041, 0.19345597922801971, -0.1592055708169937, 0.1716734915971756, 0.30346256494522095, 0.09692040085792542, 0.16826315224170685, -0.1791478842496872, 0.05318178981542587, 0.22526121139526367, 0.1842544972896576, -0.08278973400592804, -0.08363866806030273, -0.09633500128984451, -0.1391972452402115, -0.13139405846595764, 0.43920573592185974, -0.24314352869987488, 0.1521119773387909, 0.27531197667121887, 0.16114847362041473, 0.2986013889312744, -0.20217826962471008, 0.01869613490998745, 0.37583059072494507, 0.48768576979637146, 0.10426031053066254, -0.313458114862442, -0.47657039761543274, -0.022540271282196045, 0.18319016695022583, 0.3666224181652069, 0.34969672560691833, -0.24039402604103088, 0.03685646131634712, -0.4264366030693054, -0.029454180970788002, -0.13254496455192566, 0.08995290100574493, 0.21369363367557526, 0.10086234658956528, 0.15537020564079285, -0.0032710633240640163, 0.044661011546850204, -0.09545882791280746, 0.15903574228286743, -0.23730194568634033, -0.1132032498717308, 0.04265628755092621, -0.13779093325138092, -0.13386566936969757, -0.09080662578344345, 0.01538905594497919, 0.19063514471054077, 0.20482173562049866, 0.3135945200920105, 0.05322421342134476, 0.22500592470169067, 0.21918317675590515, 0.4718342125415802, -0.045839522033929825, -0.08414586633443832, 0.5593453645706177, 0.24461215734481812, 0.03002527728676796, 0.060235753655433655, -0.010037832893431187, 0.27793043851852417, -0.12376939505338669, -0.08744198083877563, 0.025136172771453857, 0.2521754205226898, -0.5562328696250916, -0.46968987584114075, 0.10254060477018356, -0.4754388928413391, -0.1108555793762207, -0.0499454066157341, 0.2599771022796631, -0.3506721556186676, 0.3180251121520996, -0.11212287098169327, 0.11180008202791214, -0.3332262635231018, -0.04516750946640968, -0.1908116340637207, 0.1568988412618637, -0.10932116210460663, 0.189616858959198, 0.3031001389026642, -0.09979411214590073, 0.46441447734832764, -0.14836932718753815, 0.05276639387011528, -0.16524428129196167, 0.42668625712394714, -0.17668236792087555, 0.1208886206150055, 0.22699223458766937, 0.07362133264541626, 0.2308327555656433, -0.02240685187280178, -0.14349481463432312, -0.3749016225337982, 0.14235320687294006, -0.18751944601535797, -0.5286054611206055, 0.11082935333251953, 0.12504611909389496, -0.39939752221107483, -0.05795292183756828, -0.010794282890856266, 0.19038978219032288, 0.34667593240737915, -0.385879248380661, -0.08683255314826965, -0.1637044996023178, -0.07959925383329391, 0.00499830162152648, -0.15899966657161713, 0.46130797266960144, 0.1852806806564331, -0.20224332809448242, 0.1531987339258194, 0.05393001064658165, 0.5437244176864624, 0.35257890820503235, 0.024558058008551598, 0.022253142669796944, -0.14071273803710938, 0.4579937756061554, 0.18906010687351227, -0.4663136601448059, -0.12786968052387238, 0.5851624011993408, 0.09912581741809845, 0.31708332896232605, 0.2227521687746048, 0.14682312309741974, 0.23375354707241058, -0.06082300469279289, 0.040691398084163666, 0.40092599391937256, -0.08982719480991364, 0.07596172392368317, -0.3674881160259247, -0.31407037377357483, 0.4080743193626404, 0.17905965447425842, 0.13126644492149353, 0.01079980656504631, -0.051287129521369934, 0.486693799495697, 0.2602527439594269, -0.08486111462116241, 0.09025520086288452, 0.2709386348724365, -0.40591466426849365, 0.1325220763683319, 0.07757306098937988, -0.09696792811155319, -0.36799952387809753, 0.0032128016464412212, -0.1876503825187683, 0.2800232172012329, 0.057301200926303864, -0.3265429437160492, -0.1517820805311203, -0.4228247404098511, -0.007933687418699265, 0.024887682870030403, -0.017824305221438408, 0.25395527482032776, 0.3247409164905548, 0.09333178400993347, -0.019458230584859848, 0.37591367959976196, -0.02651895582675934, 0.240707665681839, -0.8501719236373901, 0.13954223692417145, -0.1763751208782196, -0.28393998742103577, 0.4763646423816681, 0.2768564224243164, 0.18114015460014343, -0.14808835089206696, -0.07467491179704666, 0.44800135493278503, -0.07221224904060364, 0.46613064408302307, -0.3489920198917389, 0.022523513063788414, 0.3167482912540436, -0.2829040586948395, -0.019446173682808876, 0.3597462773323059, 0.01224728487432003, -0.06113005056977272, -0.17984221875667572, 0.3726075291633606, -0.30086421966552734, 0.2264280915260315, 0.10579586029052734, 0.00954275019466877, 0.12567439675331116, -0.2296978235244751, -0.24622677266597748, -0.5783402323722839, 0.5604590773582458, -0.10399378836154938, -0.339101105928421, -0.001096774940378964, -0.2969704270362854, -0.019728180021047592, 0.4288889169692993, 0.11643414199352264, -0.0657128393650055, -0.036480434238910675, -0.00525696249678731, -0.15044261515140533, 0.19341471791267395, 0.1804913580417633, 0.4251815974712372, 0.1282399594783783, 0.0847087949514389, 0.01277192123234272, -0.17027877271175385, -0.35771438479423523, 0.11679821461439133, 0.05171661451458931, 0.0034517154563218355, 0.38919010758399963, 0.0633620098233223, 0.032328780740499496, -0.3454793691635132, -0.027957946062088013, 0.20833463966846466, 0.10395725816488266, -0.38145801424980164, -0.12965857982635498, -0.4263536036014557, -0.3968108594417572, -0.23945912718772888, -0.17718368768692017, -0.12151685357093811, -0.376502126455307, 0.20473907887935638, 0.335532009601593, -0.20929203927516937, 0.11408489942550659, -0.2720203697681427, -0.012221505865454674, -0.3346617519855499, -0.0690799430012703, -0.07071845978498459, -0.2245493233203888, -0.2937887907028198, -0.03721293434500694, 0.19566018879413605, 0.45404258370399475, 0.22492265701293945, -0.3926711678504944, -0.16009245812892914, 0.05446169897913933, -0.262469619512558, 0.08240685611963272, -0.08520621806383133, -0.07174205780029297, 0.11912079900503159, -0.16228096187114716, 0.37041449546813965, -0.28497833013534546, 0.0295724980533123, -0.16127033531665802, -0.2947842478752136, 0.03648298606276512, 0.15339204668998718, 0.0339839868247509, -0.25572696328163147, -0.5162033438682556, -0.21861346065998077, -0.36561375856399536, -0.10801643133163452, 0.26958945393562317, 0.18670199811458588, 0.09474583715200424, -0.15174278616905212, -0.09986454993486404, -0.03366941958665848, 0.3037441670894623, -0.2530958950519562, -0.11181223392486572, 0.3357354700565338, -0.2215278148651123, -0.13774649798870087, 0.054698433727025986, -0.05541544407606125, -0.057856492698192596, 0.12819023430347443, -0.5356528759002686, -0.11282770335674286, -0.22534430027008057, -0.05317835509777069, -0.2187706083059311, -0.3081255853176117, 0.46694305539131165, -0.07963374257087708, -0.024783872067928314, -0.037710074335336685, -0.38316160440444946, 0.07807575166225433, 0.06121492013335228, 0.4086659550666809, -0.11687027662992477, 0.2762600779533386, -0.0010491084540262818, 0.4330560564994812, 0.5189310908317566, 0.0885111391544342, 0.0033367027062922716, 0.13722145557403564, 0.12602093815803528, 0.10288118571043015, 0.02816195785999298, 0.30397534370422363, -0.02097632922232151, -0.1285865157842636, 0.35867956280708313, 0.06537821888923645, -0.2829356789588928, -0.1473606526851654, -0.11749193072319031, -0.31570926308631897, -0.2660256326198578, 0.12340932339429855, -0.34433063864707947, 0.12941239774227142, 0.19345565140247345, 0.34079352021217346, -0.0818231925368309, -0.5395624041557312, -0.018391737714409828, 0.28625595569610596, 0.2730642557144165, 0.13522283732891083, -0.407380074262619, 0.1556648313999176, -0.4685329496860504, 0.2718777060508728, 0.18034234642982483, 0.34757328033447266, -0.1694793850183487, -0.15780410170555115, 0.05548808351159096, -0.2014869898557663, 0.3394179940223694, -0.4746237099170685, 0.013200893066823483, -0.0041796318255364895, -0.07114434987306595, -0.42376336455345154, -0.18968746066093445, -0.18537111580371857, 0.4030250310897827, 0.4932022988796234, 0.414958119392395, -0.3782067596912384, -0.07719452679157257, 0.2414441704750061, -0.08462473750114441, -0.05414070934057236, -0.1513761579990387, -0.21289479732513428, -0.32165291905403137, -0.2734091877937317, -0.1536313146352768, 0.005820155143737793, 0.25832733511924744, -0.10156765580177307, -0.137563556432724, 0.09534493833780289, 0.14404438436031342, 0.41291895508766174, 0.29458481073379517, 0.1835898905992508, 0.018144510686397552, 0.1495043784379959, 0.3504321873188019, 0.4505656957626343, 0.17933346331119537, 0.48814189434051514, 0.06032615527510643, -0.15531295537948608, 0.06298680603504181, -0.3202923536300659, 0.04378964379429817, 0.4678903818130493, 0.1995553821325302, 0.021901927888393402, 0.1291140615940094, -0.14375892281532288, -0.394074410200119, -0.005321691278368235, 0.06702926009893417, -0.17417603731155396, -0.33612459897994995, -0.3677898943424225, 0.26819851994514465, 0.34013640880584717, -0.11092543601989746, -0.16600114107131958, -0.15098725259304047, -0.3498978316783905, 0.33998095989227295, 0.17212849855422974, 0.945643961429596, 0.2882978618144989, -0.015686368569731712, -0.0824362114071846, -0.17538736760616302, 0.816389799118042, -0.2360057234764099, 0.4304831326007843, -0.3811832666397095, -0.2464487999677658, -0.08875583857297897, -0.17461080849170685, 0.1384231150150299, 0.11361932754516602, -0.3597879409790039, 0.32473060488700867, 0.13030126690864563, 0.03461398556828499, -0.19171525537967682, 0.4030137360095978, -0.28494516015052795, -0.19131146371364594, -0.15553031861782074, 0.02369290217757225, -0.16818122565746307, 0.384304016828537, 0.055374275892972946, -0.07974914461374283, -0.23064042627811432, -0.21798522770404816, -0.434041291475296, 0.2383929342031479, 0.15239717066287994, 0.04199264943599701, 0.2618044316768646, -0.21192942559719086, -0.021444804966449738, 0.12297068536281586, 0.3339596390724182, -0.17732936143875122, -0.17432674765586853, -0.005739609245210886, -0.10337155312299728, -0.18696127831935883, 0.14108417928218842, 0.05936390534043312, 0.38448792695999146, -0.09873640537261963, -0.01170400995761156, 0.08156457543373108, -0.1594618558883667, -0.2650093734264374, -0.17423102259635925, -0.13260604441165924, -0.1864423304796219, -0.1896415799856186, -0.16944332420825958, 0.01309659518301487, -0.06749573349952698, -0.08927705883979797, 0.013088792562484741, -0.1344013810157776, -0.15610624849796295, 0.21438787877559662, -0.0344940721988678, -0.33171695470809937, -0.10979076474905014, 0.4217984676361084, 0.33537793159484863, -0.0034991507418453693, 0.46919748187065125, -0.06988668441772461, -0.11542746424674988, -0.15430660545825958, -0.07997769117355347, -0.14900022745132446, -0.3325285017490387, 0.20993535220623016, -0.22150228917598724, -0.0826890617609024, 0.23825156688690186, 0.28522154688835144, 0.133651003241539, 0.2915157079696655, -0.12268853187561035, -0.4712011516094208, -0.009561492130160332, 0.25677797198295593, -0.01919396035373211, 0.114813432097435, -0.14997372031211853, 0.39572620391845703, -0.18629440665245056, 0.15357011556625366, -0.19016435742378235, -0.073128342628479, -0.4806310534477234, 0.3539181053638458, 0.09942641109228134, -0.0079760467633605, -0.08883413672447205, -0.02764899656176567, -0.1656598150730133, 0.28192415833473206, -0.019158799201250076, 0.006708940025418997, -0.13281415402889252, 0.19781982898712158, 0.09791278839111328, -0.10110468417406082, -0.1399448812007904, 0.11612596362829208, 0.02054673247039318, -0.004847961012274027, 0.03788449615240097, 0.2670261263847351, -0.008993883617222309, 0.38938817381858826, -0.3677164614200592, 0.09063168615102768, -0.0589553527534008, 0.18782047927379608, -0.32415103912353516, 0.28103572130203247, -0.12034691870212555, 0.21074874699115753, -0.04493287578225136, 0.04436357319355011, 0.0336267352104187, -0.028521357104182243, 0.03280775621533394, 0.04359261319041252, 0.14903274178504944, -0.16278477013111115, 0.14060956239700317, 0.3381780982017517, -0.023935535922646523, 0.5712107419967651, -0.06357821822166443, -0.161583811044693, 0.03010021336376667, 0.08797809481620789, -0.0846201553940773, -0.0033467020839452744, 0.20258072018623352, -0.06047194078564644, -0.06835407018661499, 0.35542935132980347, 0.18379707634449005, -0.10070468485355377, -0.18464207649230957, 0.1138509139418602, 0.4215512275695801, -0.17783133685588837, 0.19242459535598755, 0.1978379487991333, -0.21465827524662018, -0.01578359119594097, 0.16785676777362823, -0.20996689796447754, 0.19654665887355804, 0.045778024941682816, 0.020209504291415215, 0.4743588864803314, 0.15783260762691498, 0.3953704237937927, -0.07260807603597641, -0.2118464857339859, -0.06808539479970932, 0.6018929481506348, 0.30826687812805176, 0.43376424908638, 0.2270936369895935, 0.325705349445343, -0.27832555770874023, -0.00010910122364293784, -0.3504144847393036, -0.1837596744298935, -0.011417523957788944, -0.16466373205184937, -0.4718993902206421, -0.09388578683137894, -0.23944686353206635, 0.09697260707616806, -0.15479232370853424, 0.15204107761383057, 0.3951610326766968, -0.044418979436159134, -0.20598699152469635, -0.554853618144989, 0.13787615299224854, -0.13042554259300232, 0.19311967492103577, -0.267971396446228, 0.18461120128631592, 0.0576757937669754, -0.11533717066049576, 0.4310166835784912, 0.5478077530860901, 0.5472566485404968, 0.32145917415618896, -0.17351771891117096, -0.11064144223928452, 0.03344118595123291, 0.03277946263551712, 0.19032764434814453, 0.7335729002952576, 0.3221980631351471, -0.2711016833782196, 0.3745809495449066, -0.05907512456178665, -0.054816264659166336, 0.4139440655708313, 0.127464160323143, 0.015450801700353622, -0.06342169642448425, 0.5102360248565674, -0.17088837921619415, 0.10250909626483917, 0.00685124471783638, 0.025249730795621872, -0.274665892124176, -0.35111021995544434, 0.2094680666923523, -0.3275745213031769, 0.22225143015384674, -0.030705543234944344, 0.018956299871206284, -0.2546972334384918, 0.5723085403442383, 0.43846967816352844, -0.14093269407749176, -0.21884943544864655, -0.010652990080416203, -0.484534353017807, 0.3560207784175873, -0.3736162781715393, -0.16402579843997955, -0.13282248377799988, 0.553136944770813, -0.28624653816223145, 0.23929327726364136, -0.27799201011657715, 0.3856365978717804, -0.17468683421611786, 0.16580112278461456, -0.25717693567276, -0.013567949645221233, -0.10855495184659958, -0.22183676064014435, 0.006634545978158712, -0.24003097414970398, 0.1432495266199112, 0.042736902832984924, -0.12452109158039093, 0.03154304623603821, -0.02433457039296627, 0.06714532524347305, 0.4972288906574249, -0.003902207361534238, 0.4787132740020752, 0.46563294529914856, -0.15694773197174072, -0.2309427559375763, -0.16266821324825287, -0.37804359197616577, -0.34398844838142395, 0.36818498373031616, 0.1728876680135727, 0.4330158829689026, -0.17476195096969604, 0.5583015084266663, -0.046182192862033844, 0.2748160660266876, -0.043205443769693375, -0.004776459652930498, -0.32028597593307495, 0.012487725354731083, -0.27112919092178345, 0.025067660957574844, 0.12636417150497437, 0.37036895751953125, 0.1730210930109024, -0.020745497196912766, -0.1875908225774765, -0.41432133316993713, 0.43017494678497314, -0.3541783392429352, 0.053208932280540466, -0.13932479918003082, 0.15484224259853363, 0.21726094186306, -0.1422218680381775, -0.7538328766822815, -0.08064015954732895, 0.2726384401321411, -0.1609555184841156, 0.05061452090740204, -0.013103796169161797, -0.014781984500586987, -0.2709462344646454, -0.15545181930065155, -0.2552731931209564, 0.24762403964996338, -0.3734521269798279, 0.3072861433029175, -0.09805580973625183 ]
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.
[ -0.17600999772548676, 0.19232304394245148, 0.06580719351768494, 0.4156305491924286, 0.012082552537322044, 0.19369153678417206, -0.06068260595202446, 0.38789743185043335, -0.059173647314310074, 0.13248220086097717, -0.07451546937227249, 0.17166665196418762, 0.009341476485133171, -0.1365591287612915, -0.07436814159154892, 0.00014037142682354897, -0.06103062257170677, 0.14836759865283966, 0.07987785339355469, 0.035928815603256226, -0.06956411153078079, 0.4301077127456665, -0.3462487459182739, -0.08884061127901077, -0.23781456053256989, -0.11275550723075867, 0.02941953018307686, 0.3264773488044739, 0.04625590518116951, -0.3628298342227936, 0.25429630279541016, -0.06689684838056564, 0.12759649753570557, 0.3013107478618622, -0.00012377803795970976, 0.01873684488236904, 0.6639290452003479, -0.24313697218894958, -0.1930365413427353, -0.3823126554489136, -0.29838991165161133, -0.27950313687324524, -0.031189948320388794, -0.23511436581611633, 0.000046571207349188626, -0.04557729512453079, 0.1376456320285797, -0.1730501502752304, 0.2249150425195694, 0.3440399467945099, 0.0974617749452591, 0.18349337577819824, -0.09255971759557724, 0.11123856157064438, 0.21578530967235565, 0.12078319489955902, -0.04892555624246597, -0.014452249743044376, -0.01777915097773075, -0.15842963755130768, -0.038526955991983414, 0.3862316608428955, -0.3328869342803955, 0.11739921569824219, 0.2570723295211792, 0.14909522235393524, 0.3444841802120209, -0.19640597701072693, 0.03148142248392105, 0.3853250741958618, 0.4552387595176697, 0.020421508699655533, -0.34347280859947205, -0.43859270215034485, -0.00333142071031034, 0.09549713134765625, 0.3857690095901489, 0.365649551153183, -0.2501450181007385, 0.04120158776640892, -0.34488174319267273, 0.02025480568408966, -0.1232113391160965, 0.10563599318265915, 0.16576208174228668, 0.08337272703647614, 0.15274570882320404, -0.013326426967978477, 0.0407768152654171, -0.036118034273386, 0.14461684226989746, -0.2656342685222626, -0.10599914193153381, 0.09341870248317719, -0.16324201226234436, -0.12530267238616943, -0.18923765420913696, -0.011101354844868183, 0.16719719767570496, 0.199948251247406, 0.29252344369888306, 0.010139892809092999, 0.19105949997901917, 0.18825814127922058, 0.46921658515930176, -0.05503103509545326, -0.10778693854808807, 0.5959746241569519, 0.21419717371463776, 0.08869075775146484, -0.005387232173234224, -0.03111196681857109, 0.2485341876745224, -0.2011578232049942, 0.02705816552042961, 0.1297847479581833, 0.24780957400798798, -0.5492246150970459, -0.5105423331260681, 0.17850813269615173, -0.47248145937919617, -0.12943409383296967, -0.0333772748708725, 0.21454064548015594, -0.39336276054382324, 0.3246033191680908, -0.06497447937726974, 0.12263613939285278, -0.31911784410476685, -0.004285535309463739, -0.15956464409828186, 0.12740518152713776, -0.13459895551204681, 0.16218183934688568, 0.2891724109649658, -0.17284785211086273, 0.4029974937438965, -0.11030706763267517, 0.06361120939254761, -0.2159266173839569, 0.30709004402160645, -0.2068900167942047, 0.06568188965320587, 0.23887242376804352, 0.047185543924570084, 0.24789872765541077, 0.0007344565819948912, -0.20997630059719086, -0.3573243319988251, 0.13655652105808258, -0.20479100942611694, -0.4814678728580475, 0.07134072482585907, 0.12073981016874313, -0.39544281363487244, -0.047842781990766525, -0.07426504045724869, 0.036178186535835266, 0.36518311500549316, -0.42672961950302124, -0.09594763070344925, -0.22851383686065674, -0.07225894927978516, -0.03593417629599571, -0.1052926629781723, 0.5175380110740662, 0.15426747500896454, -0.16895270347595215, 0.13861241936683655, 0.059458617120981216, 0.5408572554588318, 0.39143240451812744, 0.0583207905292511, 0.025509856641292572, -0.19143226742744446, 0.3766968846321106, 0.20354320108890533, -0.5307167172431946, -0.23676137626171112, 0.535023033618927, 0.13700172305107117, 0.36929187178611755, 0.22536343336105347, 0.14809542894363403, 0.2658117711544037, -0.062012966722249985, 0.016567470505833626, 0.42391592264175415, -0.046327706426382065, 0.07779451459646225, -0.3728450834751129, -0.2930847704410553, 0.39219409227371216, 0.14195124804973602, 0.15030735731124878, 0.043526068329811096, -0.021334905177354813, 0.44595345854759216, 0.32378950715065, -0.04585505276918411, 0.09353774785995483, 0.26752784848213196, -0.28064799308776855, 0.25819674134254456, 0.04942989721894264, -0.22007198631763458, -0.3241133689880371, 0.02764584869146347, -0.1992626190185547, 0.2994343936443329, -0.09630535542964935, -0.2697671353816986, -0.1718752086162567, -0.4570291340351105, -0.04208105057477951, -0.007489645387977362, -0.006650620140135288, 0.20860178768634796, 0.28892529010772705, 0.13070359826087952, -0.06835632026195526, 0.4250880479812622, -0.09385346621274948, 0.2411915510892868, -0.8024853467941284, 0.24098916351795197, -0.19558897614479065, -0.3173595368862152, 0.4423464834690094, 0.2632880210876465, 0.19873642921447754, -0.14953139424324036, -0.14108479022979736, 0.44248977303504944, -0.031132962554693222, 0.420570969581604, -0.39021065831184387, 0.07099062204360962, 0.32512366771698, -0.3283196985721588, -0.01364715676754713, 0.37926769256591797, 0.06837360560894012, -0.09130765497684479, -0.09971344470977783, 0.373708039522171, -0.3763323426246643, 0.21936798095703125, 0.13083049654960632, 0.029671061784029007, 0.133592426776886, -0.2117363065481186, -0.2728498578071594, -0.5291529297828674, 0.5458309650421143, -0.06431109458208084, -0.3128312826156616, 0.004712661728262901, -0.23172351717948914, -0.06337649375200272, 0.4167565703392029, 0.039076000452041626, -0.07154951244592667, -0.016354016959667206, -0.005610610358417034, -0.09737548232078552, 0.07564843446016312, 0.20833896100521088, 0.4420287311077118, 0.09718986600637436, 0.013755002990365028, 0.07849980890750885, -0.20874273777008057, -0.31791338324546814, 0.16330759227275848, 0.01875641942024231, 0.0033223172649741173, 0.4489208459854126, 0.013933316804468632, 0.08493068814277649, -0.2789890766143799, -0.03021603636443615, 0.11369866877794266, 0.1889389455318451, -0.4271094501018524, -0.09499021619558334, -0.42307060956954956, -0.3812929391860962, -0.3252098560333252, -0.1917273998260498, -0.15214769542217255, -0.374156653881073, 0.10221859067678452, 0.3356717824935913, -0.09335115551948547, 0.15202771127223969, -0.27478060126304626, -0.07710935920476913, -0.28563815355300903, -0.16717529296875, -0.04007285088300705, -0.2553861439228058, -0.22221335768699646, -0.060300275683403015, 0.28512105345726013, 0.4162530303001404, 0.2191535234451294, -0.4314366281032562, -0.18763339519500732, -0.003962244838476181, -0.21282903850078583, 0.04713873565196991, -0.09212151169776917, -0.010913543403148651, 0.1246248334646225, -0.09668256342411041, 0.35306134819984436, -0.30845510959625244, 0.13665351271629333, -0.18659065663814545, -0.2865261137485504, 0.0870191901922226, 0.16549675166606903, 0.005900980904698372, -0.16749252378940582, -0.4755881428718567, -0.26043933629989624, -0.35199999809265137, -0.15064655244350433, 0.2697538435459137, 0.18269282579421997, 0.09125354140996933, -0.12194888293743134, -0.10851747542619705, -0.03932389244437218, 0.3502365052700043, -0.21496868133544922, -0.1261640340089798, 0.299368679523468, -0.1570669561624527, -0.17148824036121368, 0.006666142027825117, -0.019033681601285934, -0.028024304658174515, 0.12993314862251282, -0.5250760912895203, -0.12667794525623322, -0.25535327196121216, -0.019448013976216316, -0.2235792875289917, -0.21697387099266052, 0.4017181396484375, -0.01997002586722374, 0.007223943248391151, -0.007093958556652069, -0.35230350494384766, 0.02741340361535549, 0.053191326558589935, 0.38160035014152527, -0.13469237089157104, 0.30944904685020447, -0.03444178029894829, 0.42649295926094055, 0.5960378050804138, 0.03605317696928978, 0.027640048414468765, 0.07134273648262024, 0.14365197718143463, 0.0747460350394249, -0.10545036941766739, 0.26571550965309143, -0.07296650111675262, -0.07890303432941437, 0.2505124509334564, 0.00891765858978033, -0.223832368850708, -0.17834444344043732, -0.1296463906764984, -0.4514029920101166, -0.27320653200149536, 0.10059574991464615, -0.3199055790901184, 0.17668430507183075, 0.13752347230911255, 0.28149253129959106, -0.055178865790367126, -0.5646655559539795, 0.029398372396826744, 0.415056437253952, 0.21458126604557037, 0.11504068225622177, -0.3385315239429474, 0.14205677807331085, -0.34117352962493896, 0.2643872797489166, 0.2061053365468979, 0.3740829527378082, -0.11116916686296463, -0.11376369744539261, 0.03909456357359886, -0.1925828456878662, 0.37881171703338623, -0.47466418147087097, 0.055517103523015976, -0.004085284657776356, -0.07041024416685104, -0.4118960499763489, -0.21917130053043365, -0.18260164558887482, 0.36919209361076355, 0.4657863974571228, 0.4691725969314575, -0.35209357738494873, -0.06196654587984085, 0.25569477677345276, 0.02139100432395935, -0.10110528767108917, -0.174580916762352, -0.22858765721321106, -0.33551132678985596, -0.2618897259235382, -0.1732967644929886, -0.04387655481696129, 0.31121695041656494, -0.07534463703632355, -0.08241944760084152, 0.042250748723745346, 0.16651956737041473, 0.42532646656036377, 0.2312401682138443, 0.20768509805202484, -0.032401058822870255, 0.22409984469413757, 0.25176844000816345, 0.4094744324684143, 0.21285977959632874, 0.5298259258270264, -0.002762923017144203, -0.19281630218029022, -0.07655420154333115, -0.30114462971687317, 0.005492191296070814, 0.4374125897884369, 0.2044098824262619, -0.03044983744621277, 0.12637124955654144, -0.13738682866096497, -0.42967459559440613, -0.05294406786561012, 0.12660500407218933, -0.10131683945655823, -0.3854479193687439, -0.4973113238811493, 0.29982176423072815, 0.35049939155578613, -0.11287233978509903, -0.14339733123779297, -0.09955069422721863, -0.3347935080528259, 0.34836745262145996, 0.17213325202465057, 0.9992427825927734, 0.22708755731582642, -0.02504495158791542, -0.06836090981960297, -0.2009785920381546, 0.7844553589820862, -0.2129828780889511, 0.35989701747894287, -0.435026079416275, -0.2963213622570038, -0.1060200035572052, -0.18125876784324646, 0.19104528427124023, 0.1032356396317482, -0.33262673020362854, 0.38838034868240356, 0.10925755649805069, -0.017338989302515984, -0.2038635015487671, 0.405789315700531, -0.28847700357437134, -0.11453773826360703, -0.15640367567539215, 0.020807741209864616, -0.1877889335155487, 0.4317876994609833, 0.08245303481817245, -0.05726049095392227, -0.26244407892227173, -0.1683528572320938, -0.5494994521141052, 0.2216644287109375, 0.14272376894950867, 0.02637338824570179, 0.21179987490177155, -0.22485175728797913, -0.013590620830655098, 0.1342349648475647, 0.32788577675819397, -0.11514616757631302, -0.10681679099798203, -0.0016027115052565932, -0.02261585369706154, -0.1557306945323944, 0.20076411962509155, 0.05771078169345856, 0.43403688073158264, -0.10155397653579712, -0.06176362931728363, 0.10912110656499863, -0.07918648421764374, -0.2611496150493622, -0.13389351963996887, -0.13959360122680664, -0.18021482229232788, -0.21085703372955322, -0.20317159593105316, 0.014056196436285973, -0.02925691567361355, -0.09731964766979218, 0.017006656154990196, -0.05921439453959465, -0.19780591130256653, 0.2540116608142853, -0.0918474793434143, -0.35133200883865356, -0.09548141807317734, 0.4479711353778839, 0.2841583490371704, -0.04742572084069252, 0.46550804376602173, -0.04476882144808769, -0.11638546735048294, -0.14287014305591583, -0.07991968095302582, -0.22824028134346008, -0.3608662486076355, 0.17055602371692657, -0.25035586953163147, 0.003219874342903495, 0.21465753018856049, 0.2766205072402954, 0.23138055205345154, 0.2816077172756195, -0.09972801804542542, -0.5418651700019836, -0.005885561462491751, 0.16270044445991516, 0.09235355257987976, 0.1533018946647644, -0.19041426479816437, 0.3851989507675171, -0.20321133732795715, 0.08855850249528885, -0.1887628585100174, 0.0016791120870038867, -0.42164504528045654, 0.31907883286476135, 0.09868377447128296, -0.01033435482531786, -0.10648008435964584, -0.05362044274806976, -0.10954698175191879, 0.23286379873752594, -0.008623589761555195, -0.002102932659909129, -0.17353101074695587, 0.19913393259048462, 0.07818426191806793, -0.1462133824825287, -0.1277797371149063, 0.029297275468707085, 0.07194790989160538, -0.03990570455789566, -0.032883912324905396, 0.3182600736618042, 0.01694619655609131, 0.4303995668888092, -0.29679208993911743, 0.00705740274861455, -0.05222383141517639, 0.2537623643875122, -0.3740804195404053, 0.32040345668792725, -0.06483308970928192, 0.2339765727519989, -0.0029911145102232695, 0.07607699185609818, 0.005250284913927317, -0.010939661413431168, -0.03330593928694725, 0.0442507341504097, 0.26031821966171265, -0.18230952322483063, 0.15160922706127167, 0.32696476578712463, 0.06302893906831741, 0.6621745228767395, -0.09779175370931625, -0.14357520639896393, 0.10021929442882538, 0.08938563615083694, -0.12129492312669754, 0.01768462359905243, 0.2556926906108856, -0.06330011039972305, -0.08007574826478958, 0.37514686584472656, 0.1603878140449524, -0.11737290024757385, -0.24033223092556, 0.09514569491147995, 0.5035053491592407, -0.1352921724319458, 0.18721415102481842, 0.267254501581192, -0.05208104848861694, -0.0351175032556057, 0.11959386616945267, -0.21357035636901855, 0.11305532604455948, 0.08282125741243362, -0.016397496685385704, 0.4081760346889496, 0.05979743227362633, 0.38107171654701233, -0.11079796403646469, -0.19281327724456787, -0.01688435859978199, 0.5765299201011658, 0.1995476484298706, 0.44371116161346436, 0.23778961598873138, 0.34543341398239136, -0.33760300278663635, 0.025902410969138145, -0.3117019832134247, -0.13533563911914825, -0.013288400135934353, -0.17353428900241852, -0.4493544101715088, -0.13270969688892365, -0.22633272409439087, 0.11731532216072083, -0.13277378678321838, 0.16289079189300537, 0.3212730586528778, 0.033501919358968735, -0.2087709754705429, -0.512785017490387, 0.06974243372678757, -0.11926914006471634, 0.22779178619384766, -0.2851019501686096, 0.11935069411993027, 0.12183374166488647, -0.06800273060798645, 0.3805285096168518, 0.5955418348312378, 0.6343860030174255, 0.39273378252983093, -0.19025839865207672, -0.056027647107839584, 0.03639761731028557, 0.05950076878070831, 0.2065487653017044, 0.7163811922073364, 0.24613188207149506, -0.2202015519142151, 0.34021157026290894, -0.033732473850250244, -0.06545177102088928, 0.3719547390937805, 0.07000844925642014, 0.025595800951123238, -0.08627064526081085, 0.5236676335334778, -0.1916217803955078, 0.11693498492240906, -0.05106046795845032, 0.06795037537813187, -0.32060742378234863, -0.3236776292324066, 0.2613776922225952, -0.2641170024871826, 0.24652720987796783, -0.0030777130741626024, 0.015305610373616219, -0.2047778218984604, 0.5182005167007446, 0.5190637111663818, -0.0969170406460762, -0.24339567124843597, -0.011231021955609322, -0.5534392595291138, 0.3506773114204407, -0.3848564326763153, -0.21972522139549255, -0.1404455453157425, 0.5029476881027222, -0.29132479429244995, 0.2718918025493622, -0.16698041558265686, 0.42171674966812134, -0.19998954236507416, 0.2392648607492447, -0.2612939178943634, -0.048200931400060654, -0.0856674313545227, -0.24912534654140472, -0.041207849979400635, -0.23633699119091034, 0.1247376799583435, 0.0006039340514689684, -0.09424842894077301, 0.030080674216151237, -0.02639947645366192, 0.06136699020862579, 0.49497437477111816, 0.02486834116280079, 0.4822665750980377, 0.5218866467475891, -0.14398781955242157, -0.26332685351371765, -0.17555581033229828, -0.33674079179763794, -0.30829402804374695, 0.4396419823169708, 0.13124732673168182, 0.37442272901535034, -0.18796025216579437, 0.4798942804336548, -0.13751068711280823, 0.3361934721469879, -0.025599269196391106, -0.11676771938800812, -0.37553226947784424, -0.051263485103845596, -0.2222040593624115, -0.03113817237317562, 0.0988994762301445, 0.4211767911911011, 0.1588224619626999, 0.08846072852611542, -0.17987798154354095, -0.4090624153614044, 0.43700626492500305, -0.32739946246147156, 0.04309256002306938, -0.11276346445083618, 0.12388124316930771, 0.14992275834083557, -0.039151690900325775, -0.7592728137969971, -0.035973768681287766, 0.3005138039588928, -0.17703604698181152, -0.027316739782691002, -0.013707579113543034, -0.01945149153470993, -0.22175200283527374, -0.11064080893993378, -0.3095339834690094, 0.2275262475013733, -0.359622597694397, 0.3135587275028229, -0.08145656436681747 ]
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')
[ -0.05952109023928642, 0.15316857397556305, 0.07047887146472931, 0.3878919184207916, -0.015148007310926914, 0.14190790057182312, -0.09238557517528534, 0.32008156180381775, -0.04935650900006294, 0.008110440336167812, -0.08665408194065094, 0.15768754482269287, 0.03654864430427551, 0.04614751785993576, -0.019717158749699593, 0.08979135006666183, -0.04565105587244034, 0.20413713157176971, 0.04544680565595627, 0.029078947380185127, 0.0009836388053372502, 0.5112384557723999, -0.24448399245738983, -0.06453919410705566, -0.3506303131580353, -0.14984428882598877, 0.05441681295633316, 0.3629709482192993, 0.010663417167961597, -0.3910156190395355, 0.31971773505210876, -0.04728400334715843, 0.11418864130973816, 0.24019668996334076, -0.00012255208275746554, -0.002842358313500881, 0.4699680209159851, -0.21914030611515045, -0.2050270140171051, -0.2703422009944916, -0.21681587398052216, -0.2991621792316437, -0.04636301472783089, -0.21825683116912842, -0.06674200296401978, -0.06653834134340286, 0.07996346801519394, -0.051992058753967285, 0.37194857001304626, 0.2974585294723511, 0.1255473792552948, 0.09593866765499115, -0.12729498744010925, 0.10268136858940125, 0.08485591411590576, 0.06652484089136124, -0.011885037645697594, -0.0418551079928875, -0.17200250923633575, -0.3219939172267914, -0.12819784879684448, 0.5180578827857971, -0.1915794163942337, 0.12624691426753998, 0.1819387972354889, 0.08008380979299545, 0.2830647826194763, -0.11941685527563095, 0.0679110437631607, 0.3065376877784729, 0.40647900104522705, 0.10857941210269928, -0.2249048799276352, -0.5339202284812927, 0.007108790799975395, 0.051360856741666794, 0.45279985666275024, 0.34143391251564026, -0.19766207039356232, 0.10054163634777069, -0.41586530208587646, -0.22632424533367157, -0.054598368704319, 0.007278733886778355, 0.2139008790254593, 0.2067917287349701, 0.03881853446364403, -0.010091419331729412, 0.13749277591705322, -0.03788391128182411, 0.22440952062606812, -0.09224067628383636, -0.14465269446372986, 0.08490952849388123, -0.21261776983737946, -0.1594933271408081, -0.12389103323221207, 0.03781513124704361, 0.17408967018127441, 0.1535213589668274, 0.2372104376554489, -0.07823171466588974, 0.14463821053504944, 0.04192867875099182, 0.3693922162055969, 0.17495577037334442, -0.0018965292256325483, 0.428241491317749, 0.14054057002067566, 0.03328912705183029, 0.11029240489006042, 0.003799479454755783, 0.29398199915885925, -0.15723003447055817, -0.06706347316503525, 0.04519667476415634, 0.2099599987268448, -0.5436288714408875, -0.4165375232696533, 0.07986797392368317, -0.49174922704696655, -0.20235218107700348, -0.12295106798410416, 0.1320471465587616, -0.34991830587387085, 0.3523443341255188, -0.07884306460618973, 0.03780721127986908, -0.3732772469520569, -0.05599600449204445, -0.2014297991991043, 0.10900107771158218, -0.19364415109157562, 0.2363348752260208, 0.2714063823223114, -0.2821212112903595, 0.4370521903038025, -0.05147692933678627, -0.09336435794830322, -0.2561877965927124, 0.26467159390449524, -0.03078082576394081, -0.023292407393455505, 0.19040650129318237, -0.13137570023536682, 0.23993590474128723, 0.00858051236718893, -0.20084357261657715, -0.28731587529182434, 0.016373224556446075, -0.27687516808509827, -0.6388428211212158, 0.08123832941055298, 0.11715029925107956, -0.40461617708206177, -0.05964155122637749, 0.01749771647155285, 0.22032077610492706, 0.37056422233581543, -0.2604246735572815, -0.10591354966163635, -0.16875965893268585, -0.10133932530879974, -0.02985309064388275, -0.16532757878303528, 0.5042769312858582, 0.13272729516029358, -0.3224141299724579, 0.16070258617401123, -0.031384795904159546, 0.47419580817222595, 0.4351547956466675, -0.006501981522887945, 0.10734240710735321, -0.17257246375083923, 0.35535699129104614, 0.14280350506305695, -0.3949947953224182, -0.25507932901382446, 0.4000755846500397, -0.0010333555983379483, 0.2915745675563812, 0.18328069150447845, 0.13108763098716736, 0.09201546758413315, -0.06459511816501617, 0.05912270396947861, 0.47297513484954834, 0.07429323345422745, 0.09923028945922852, -0.45281118154525757, -0.4169975817203522, 0.43306517601013184, 0.16881917417049408, 0.10977055132389069, 0.07702751457691193, -0.019928576424717903, 0.488028883934021, 0.3150663375854492, -0.1443878710269928, 0.06338495761156082, 0.3153185546398163, -0.33134543895721436, 0.15865164995193481, 0.04939116910099983, -0.08946047723293304, -0.5167126059532166, 0.007718370296061039, -0.2730107009410858, 0.3195575475692749, 0.0504484549164772, -0.19811008870601654, -0.23507997393608093, -0.4627164304256439, -0.07891365885734558, -0.18066349625587463, 0.0007564787520095706, 0.23502181470394135, 0.4090336859226227, 0.0545513741672039, -0.018292918801307678, 0.5882973670959473, -0.09648583084344864, 0.2448498159646988, -0.8845171928405762, 0.20832650363445282, -0.12776292860507965, -0.1689957231283188, 0.4122883677482605, 0.28587228059768677, 0.27037739753723145, -0.14165320992469788, -0.13194648921489716, 0.46327099204063416, -0.22580678761005402, 0.4788297712802887, -0.22793681919574738, -0.0022638735827058554, 0.27312570810317993, -0.29561716318130493, 0.021120626479387283, 0.21801593899726868, -0.04656480625271797, -0.05527070164680481, -0.06619327515363693, 0.37105897068977356, -0.36957621574401855, 0.13853022456169128, 0.01757306233048439, -0.04477173835039139, 0.10005342215299606, -0.2872810661792755, -0.14379781484603882, -0.4895385205745697, 0.5787146091461182, -0.15024587512016296, -0.2525637447834015, 0.021484972909092903, -0.1338326781988144, 0.016381416469812393, 0.3793988823890686, 0.0649048238992691, -0.09646476805210114, -0.03197633475065231, -0.01605347916483879, -0.2002706229686737, 0.053591351956129074, 0.2453748732805252, 0.5202928185462952, 0.1364886313676834, 0.02637467347085476, 0.09233342111110687, -0.16415835916996002, -0.3306780457496643, 0.15244348347187042, -0.06436435878276825, 0.08235237002372742, 0.3692174255847931, 0.17514674365520477, -0.0021436824463307858, -0.5159104466438293, 0.10421571135520935, 0.21197280287742615, 0.24466362595558167, -0.30409905314445496, -0.13484099507331848, -0.36489418148994446, -0.38661524653434753, -0.2667770981788635, -0.19045612215995789, -0.3162619471549988, -0.33005276322364807, 0.2057834267616272, 0.285591721534729, -0.11775761842727661, 0.15927451848983765, -0.09243503212928772, 0.060767337679862976, -0.3277832269668579, 0.009110864251852036, -0.10953313857316971, -0.2758679986000061, -0.20552201569080353, -0.05036259442567825, 0.1038396954536438, 0.3221498429775238, 0.18946275115013123, -0.35606250166893005, -0.19915364682674408, -0.07048308849334717, -0.337956041097641, 0.04746350273489952, -0.08756683021783829, -0.07165676355361938, 0.08050284534692764, -0.06838096678256989, 0.39283299446105957, -0.15853343904018402, 0.14054343104362488, -0.2609800696372986, -0.3093511164188385, -0.041610606014728546, 0.17137835919857025, 0.19560803472995758, -0.16764119267463684, -0.5419102311134338, -0.24373039603233337, -0.3634907007217407, -0.015156951732933521, 0.21412551403045654, 0.18519513309001923, 0.17180371284484863, -0.13248224556446075, -0.20916211605072021, -0.11133086681365967, 0.2509463429450989, -0.3925293982028961, -0.1390988528728485, 0.19587847590446472, -0.07682519406080246, -0.1303209662437439, -0.07141534984111786, 0.10173920542001724, -0.0858042761683464, 0.07061208784580231, -0.6237039566040039, -0.22018077969551086, -0.24198280274868011, -0.09389495849609375, -0.30282995104789734, -0.17801034450531006, 0.5007230043411255, -0.04541129618883133, -0.0041328067891299725, 0.08095008879899979, -0.4596955478191376, 0.01794758252799511, 0.1331704705953598, 0.4873124063014984, -0.17907072603702545, 0.238106831908226, -0.019556008279323578, 0.5533110499382019, 0.5295947790145874, 0.0531759038567543, -0.009988955222070217, 0.16884203255176544, 0.22626478970050812, 0.04008089378476143, -0.13294954597949982, 0.333008348941803, -0.06891163438558578, 0.03726346045732498, 0.3456495404243469, 0.10086295008659363, -0.14929670095443726, -0.14536269009113312, -0.16622063517570496, -0.3622744679450989, -0.1801701933145523, 0.13737712800502777, -0.1392013430595398, 0.12042481452226639, 0.2959354817867279, 0.1435416340827942, -0.17569319903850555, -0.47165796160697937, -0.008708763867616653, 0.24278490245342255, 0.14365404844284058, 0.19786420464515686, -0.457914799451828, 0.17113885283470154, -0.5147839188575745, 0.2657477855682373, 0.15840059518814087, 0.264661580324173, -0.13183236122131348, -0.14703072607517242, 0.09939615428447723, -0.1728263944387436, 0.39402201771736145, -0.40494948625564575, -0.0772494450211525, -0.07967578619718552, -0.01671557128429413, -0.33635610342025757, -0.1784154325723648, -0.03545081615447998, 0.30832505226135254, 0.47513121366500854, 0.49083882570266724, -0.39141011238098145, -0.15796515345573425, 0.186876580119133, -0.009669514372944832, -0.13670526444911957, -0.13703151047229767, -0.22997455298900604, -0.38902580738067627, -0.2763266861438751, -0.11392204463481903, -0.13009390234947205, 0.3114844560623169, 0.025374671444296837, 0.01682475581765175, 0.12226442247629166, 0.0988527312874794, 0.25107961893081665, 0.18979117274284363, 0.13215506076812744, 0.025374509394168854, 0.07778174430131912, 0.38310813903808594, 0.5103030204772949, 0.03691963851451874, 0.629610002040863, -0.010878411121666431, -0.03791653737425804, -0.08799266070127487, -0.08143594115972519, 0.02740647830069065, 0.5505659580230713, 0.20019091665744781, 0.0538078248500824, 0.13748937845230103, 0.01043148897588253, -0.4142746925354004, -0.04083781689405441, 0.11529915779829025, -0.08835110813379288, -0.42787033319473267, -0.2941100597381592, 0.3169785737991333, 0.23538519442081451, -0.13394257426261902, -0.1313214898109436, -0.13563622534275055, -0.27198243141174316, 0.24533580243587494, 0.05523395910859108, 1.041927456855774, 0.20728473365306854, 0.04186413064599037, -0.05581316351890564, -0.14920112490653992, 0.9380416870117188, -0.30007874965667725, 0.41763725876808167, -0.3517111539840698, -0.1780039221048355, -0.05750351399183273, -0.14746712148189545, 0.036034245043992996, 0.09087656438350677, -0.15406090021133423, 0.37495267391204834, 0.2888360321521759, 0.05043919384479523, -0.12973637878894806, 0.42349618673324585, -0.24959714710712433, -0.15231406688690186, -0.3030504882335663, 0.01705758646130562, -0.3022838532924652, 0.5400670766830444, -0.03814375400543213, -0.07119669765233994, -0.08948308974504471, -0.20423376560211182, -0.37206488847732544, 0.25258561968803406, 0.02417220175266266, 0.12936842441558838, 0.2318013310432434, -0.19619135558605194, -0.07764197140932083, 0.10560866445302963, 0.40050604939460754, -0.1648746132850647, -0.12909553945064545, -0.006210960913449526, -0.1460067629814148, -0.12728795409202576, 0.11892815679311752, 0.0915602296590805, 0.28809672594070435, -0.09807279706001282, -0.03008093871176243, -0.01380379218608141, -0.14492282271385193, -0.33086657524108887, -0.24841293692588806, -0.05192225053906441, -0.16080878674983978, -0.22444875538349152, -0.014997085556387901, 0.17251567542552948, -0.028158875182271004, -0.13444553315639496, 0.02083178050816059, -0.0476456917822361, -0.1695941686630249, 0.1148095428943634, -0.09480088949203491, -0.35478663444519043, -0.1866084635257721, 0.46530088782310486, 0.41192322969436646, -0.1300048679113388, 0.4889170527458191, -0.10779226571321487, -0.04715241491794586, -0.161250501871109, -0.18598581850528717, -0.2622639536857605, -0.4254757761955261, 0.20580707490444183, -0.21055163443088531, -0.1763060837984085, 0.2077464461326599, 0.26054131984710693, 0.23746612668037415, 0.3127463757991791, -0.013969323597848415, -0.4817532002925873, -0.039254821836948395, 0.1360337883234024, -0.011189009062945843, 0.09119061380624771, -0.024539882317185402, 0.3797798454761505, -0.2409822642803192, 0.3181031346321106, -0.21519367396831512, 0.03928845748305321, -0.3406311869621277, 0.30292072892189026, 0.19331972301006317, -0.1392337679862976, 0.06780169159173965, 0.012921526096761227, -0.14105567336082458, 0.26413536071777344, -0.10220292955636978, -0.017113514244556427, -0.1405882090330124, 0.19693754613399506, 0.15148647129535675, -0.11185339093208313, -0.15570729970932007, 0.0798281729221344, 0.06451967358589172, 0.016015835106372833, 0.06329796463251114, 0.28377753496170044, 0.06752504408359528, 0.3050289452075958, -0.2870916724205017, -0.03953512758016586, 0.0005169676151126623, 0.19909076392650604, -0.24377386271953583, 0.22092698514461517, -0.0393502451479435, 0.21276777982711792, -0.16374030709266663, 0.05936998873949051, -0.09963328391313553, 0.000898944737855345, 0.08639172464609146, 0.08966086804866791, 0.08942973613739014, -0.25115662813186646, 0.034797344356775284, 0.2602269649505615, 0.0537942610681057, 0.5116792321205139, -0.083155557513237, -0.15430915355682373, 0.1248418316245079, 0.08711467683315277, -0.14101749658584595, -0.06204991787672043, 0.31024307012557983, -0.07737144827842712, -0.09637970477342606, 0.38103196024894714, 0.17304831743240356, -0.12971392273902893, -0.14074601233005524, 0.10686633735895157, 0.4532068371772766, -0.19162896275520325, 0.1509833186864853, 0.12681357562541962, -0.10557367652654648, -0.08674145489931107, 0.3743513524532318, -0.20107807219028473, 0.12214754521846771, 0.11047165095806122, 0.04505417123436928, 0.3720736503601074, 0.05472865328192711, 0.4531284272670746, -0.16516496241092682, -0.1589662730693817, -0.03961110860109329, 0.6585906744003296, 0.1879715770483017, 0.43030044436454773, 0.1421811878681183, 0.5052537322044373, -0.3357340097427368, -0.015215990133583546, -0.3209925591945648, 0.021360963582992554, -0.092747263610363, -0.1665078103542328, -0.3016425669193268, -0.11755891889333725, -0.10489368438720703, 0.1544775813817978, -0.15096943080425262, 0.05504562333226204, 0.4208000898361206, -0.06759672611951828, -0.29463106393814087, -0.5726761221885681, -0.003766485257074237, 0.05557888746261597, 0.27704957127571106, -0.2272968888282776, 0.2844381630420685, 0.07585090398788452, -0.20439967513084412, 0.3192933201789856, 0.6775825023651123, 0.5208287239074707, 0.23217332363128662, -0.032789431512355804, -0.0775182917714119, 0.07033335417509079, 0.12693887948989868, 0.21077430248260498, 0.7450385093688965, 0.2959689199924469, -0.31335198879241943, 0.27542445063591003, -0.03803357109427452, -0.11196869611740112, 0.3472534418106079, 0.02500358410179615, 0.07327773422002792, -0.036580272018909454, 0.3870857059955597, -0.16182906925678253, 0.08336400985717773, -0.06547269225120544, 0.09793063253164291, -0.3249248266220093, -0.21097224950790405, 0.16343876719474792, -0.17252622544765472, 0.3017396926879883, -0.08095758408308029, 0.0441569983959198, -0.2473849207162857, 0.5227412581443787, 0.5025138258934021, -0.31866219639778137, -0.15052737295627594, -0.026378240436315536, -0.5267685651779175, 0.29629507660865784, -0.3916666805744171, -0.08135951310396194, -0.20274895429611206, 0.44821152091026306, -0.19692084193229675, 0.2801699936389923, -0.17557309567928314, 0.41491788625717163, -0.07420129328966141, -0.01893754117190838, -0.23684784770011902, 0.07650018483400345, -0.03134005144238472, -0.1934029906988144, 0.0715756043791771, -0.24616371095180511, 0.21737603843212128, 0.1577955037355423, -0.10438890755176544, 0.03160786256194115, 0.10152055323123932, -0.016133032739162445, 0.39007246494293213, -0.07530254870653152, 0.5313584208488464, 0.29228490591049194, -0.03974778205156326, -0.28395146131515503, -0.14522872865200043, -0.402590811252594, -0.31096336245536804, 0.34221339225769043, 0.202791228890419, 0.48639532923698425, -0.22398023307323456, 0.47227513790130615, -0.16622622311115265, 0.3110293447971344, 0.027511898428201675, -0.06190222501754761, -0.31948888301849365, -0.026342282071709633, -0.174051433801651, 0.044958025217056274, 0.20531387627124786, 0.5501669645309448, 0.1474919468164444, 0.05283035710453987, -0.2644667327404022, -0.41413289308547974, 0.3889918327331543, -0.3854779005050659, 0.003921016585081816, 0.07362981140613556, 0.23610728979110718, 0.27451732754707336, -0.2053195685148239, -0.7636072635650635, 0.043448302894830704, 0.22158853709697723, -0.1319129467010498, -0.008146198466420174, 0.06734827905893326, -0.1885165572166443, -0.272627055644989, -0.014860718511044979, -0.1771647185087204, 0.3695697486400604, -0.40832576155662537, 0.417351633310318, -0.019796544685959816 ]
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)
[ -0.04786314070224762, 0.2781369984149933, 0.07765761762857437, 0.3915258049964905, -0.10355501621961594, 0.03640372306108475, -0.08268313854932785, 0.24596016108989716, -0.083533376455307, 0.04618573188781738, 0.08917640894651413, 0.17450423538684845, 0.04060833901166916, -0.1594097763299942, 0.01790868304669857, 0.05279991403222084, -0.06446697562932968, 0.12483137845993042, 0.10597913712263107, 0.0428970605134964, -0.01056397520005703, 0.39612412452697754, -0.1937781274318695, -0.04041885584592819, -0.3053344488143921, -0.23367752134799957, 0.029022276401519775, 0.23348157107830048, 0.04022820293903351, -0.4769928753376007, 0.24664802849292755, 0.05662958696484566, 0.1005319356918335, 0.1670154631137848, -0.00012176283780718222, -0.0632682666182518, 0.5389478802680969, -0.19484315812587738, -0.25188639760017395, -0.029789617285132408, -0.38360294699668884, -0.27975690364837646, 0.003872625296935439, -0.28427693247795105, 0.015178422443568707, 0.023537253960967064, 0.17921505868434906, -0.16523881256580353, 0.2291364222764969, 0.3027588725090027, 0.12136121839284897, 0.10980752110481262, -0.18810933828353882, 0.1678948849439621, 0.173204705119133, -0.02763039618730545, -0.06475888192653656, -0.15657655894756317, -0.11171024292707443, -0.28193017840385437, -0.24037230014801025, 0.45989373326301575, -0.23572422564029694, 0.15005967020988464, 0.26409244537353516, 0.08830887824296951, 0.32167714834213257, -0.08726483583450317, 0.013093874789774418, 0.39029112458229065, 0.5862076878547668, -0.046833451837301254, -0.28496620059013367, -0.49856555461883545, -0.12967394292354584, 0.028640318661928177, 0.5057664513587952, 0.2795506715774536, -0.15703769028186798, 0.053701501339673996, -0.4670746922492981, -0.17777825891971588, -0.05261130630970001, -0.0008108550682663918, 0.31687161326408386, 0.19807468354701996, 0.125627800822258, -0.051634710282087326, 0.03075571171939373, -0.010176900774240494, 0.2684498727321625, -0.18169057369232178, -0.12499357014894485, 0.08544931560754776, -0.23216646909713745, -0.20516401529312134, -0.026725610718131065, 0.09116508066654205, 0.25675061345100403, 0.21745604276657104, 0.40052512288093567, 0.005886257626116276, 0.21397140622138977, 0.12421656399965286, 0.335857093334198, 0.20841912925243378, -0.01487102173268795, 0.3932430148124695, 0.15307916700839996, -0.0024872615467756987, 0.13732926547527313, 0.057575568556785583, 0.35157957673072815, -0.20486624538898468, -0.011022225022315979, 0.20190492272377014, 0.15973494946956635, -0.4877757728099823, -0.35076817870140076, 0.10072839260101318, -0.39977991580963135, -0.26050668954849243, -0.05867193266749382, 0.24248526990413666, -0.3720569312572479, 0.32263028621673584, -0.2690640687942505, -0.033522795885801315, -0.4499552547931671, -0.12790533900260925, -0.23330014944076538, 0.17981591820716858, -0.14685574173927307, 0.3382386267185211, 0.2391362339258194, -0.07944651693105698, 0.3943147361278534, -0.1676049679517746, -0.09943617880344391, -0.13192246854305267, 0.35655441880226135, -0.10957295447587967, 0.08933659642934799, 0.20857872068881989, -0.09280368685722351, 0.1552496999502182, -0.048470690846443176, -0.09950074553489685, -0.4851606786251068, 0.04101800546050072, -0.32748943567276, -0.6717568635940552, 0.09786220639944077, 0.11474645882844925, -0.3720282018184662, -0.0490008145570755, 0.15988019108772278, 0.3045908510684967, 0.5192114114761353, -0.37980982661247253, 0.0015788148157298565, -0.08771120011806488, -0.19829161465168, -0.1586369127035141, -0.17053329944610596, 0.5231292247772217, 0.07261897623538971, -0.30685606598854065, 0.12426850944757462, 0.01171756163239479, 0.4826217293739319, 0.3697670102119446, -0.035665061324834824, 0.15336906909942627, -0.16418781876564026, 0.35587459802627563, 0.17119933664798737, -0.4163646101951599, -0.27680957317352295, 0.38038891553878784, 0.10637687146663666, 0.27490153908729553, 0.2496737241744995, 0.1237708330154419, 0.24420638382434845, -0.1929771900177002, 0.016356267035007477, 0.429861456155777, -0.06306522339582443, 0.09338270872831345, -0.5093786120414734, -0.4187878370285034, 0.4363305866718292, 0.11583667993545532, 0.06944459676742554, 0.03629539906978607, -0.0036747155245393515, 0.47427085041999817, 0.16346047818660736, -0.08560045808553696, 0.09851223975419998, 0.25446826219558716, -0.34776681661605835, 0.13968248665332794, 0.0839439108967781, -0.07341696321964264, -0.4933946430683136, -0.029929086565971375, -0.27951580286026, 0.1781543344259262, 0.14817465841770172, -0.3201976418495178, -0.09828801453113556, -0.4774523377418518, -0.09374042600393295, -0.13652074337005615, 0.02948085218667984, 0.20342418551445007, 0.5127836465835571, 0.08227691054344177, 0.019414393231272697, 0.5455952286720276, -0.006968056783080101, 0.16135865449905396, -0.8287150263786316, 0.06388166546821594, -0.1368926465511322, -0.18384499847888947, 0.36261171102523804, 0.3595055341720581, 0.21638263761997223, -0.05094464123249054, -0.0946415364742279, 0.5340720415115356, -0.14141790568828583, 0.4370478093624115, -0.22054404020309448, 0.15210264921188354, 0.2498190701007843, -0.2066888064146042, 0.11677218228578568, 0.21492525935173035, -0.13724827766418457, -0.0022524099331349134, -0.09489517658948898, 0.3331184685230255, -0.37468221783638, 0.08620353043079376, 0.06745458394289017, -0.10661545395851135, 0.05747632682323456, -0.25573617219924927, -0.2727254331111908, -0.45996788144111633, 0.5414716601371765, -0.053103189915418625, -0.23666399717330933, 0.01362602785229683, -0.1199130043387413, 0.10136976838111877, 0.46203017234802246, 0.10436481237411499, -0.015318724326789379, -0.06299618631601334, 0.09480459243059158, -0.25104615092277527, 0.20877355337142944, 0.2167145162820816, 0.4686731994152069, 0.1513730138540268, 0.07371979206800461, 0.019000455737113953, -0.18847370147705078, -0.3368954062461853, 0.1008831188082695, -0.01621178165078163, -0.024240266531705856, 0.2759980857372284, 0.011330670677125454, -0.0014324454823508859, -0.30334189534187317, 0.02210836671292782, 0.1912081241607666, 0.17614950239658356, -0.33251288533210754, -0.06196731701493263, -0.39287182688713074, -0.32350003719329834, -0.2495744377374649, -0.1301070600748062, -0.202510267496109, -0.35886088013648987, 0.3024478256702423, 0.16726498305797577, -0.25919556617736816, 0.13086920976638794, -0.2515992820262909, 0.185903400182724, -0.2763068974018097, -0.10526377707719803, -0.07306026667356491, -0.2778659462928772, -0.2761033773422241, -0.06377299875020981, 0.11679274588823318, 0.31539833545684814, 0.3178684115409851, -0.41521841287612915, -0.21306531131267548, 0.040672965347766876, -0.3145899772644043, 0.06718379259109497, 0.059937696903944016, -0.10611498355865479, 0.10012196749448776, -0.2141467183828354, 0.3740347921848297, -0.22101722657680511, -0.05585438385605812, -0.2211199849843979, -0.25211039185523987, -0.06747744232416153, 0.20162172615528107, 0.24550214409828186, -0.19950488209724426, -0.5588045120239258, -0.3258547782897949, -0.3663261830806732, -0.05702010542154312, 0.10900697857141495, 0.2995641231536865, 0.08219178020954132, -0.21076975762844086, -0.21016107499599457, -0.18786542117595673, 0.31254082918167114, -0.37954744696617126, -0.16327403485774994, 0.35842159390449524, -0.19167783856391907, -0.07630977779626846, 0.04411058872938156, -0.042466674000024796, -0.07117578387260437, 0.12719349563121796, -0.548882246017456, -0.17458398640155792, -0.11559586971998215, -0.041599515825510025, -0.2191823422908783, -0.22065357863903046, 0.6390054821968079, -0.04464628919959068, 0.04780702292919159, 0.03887385502457619, -0.36135271191596985, 0.0701281800866127, 0.03474855795502663, 0.5672550797462463, -0.04162020608782768, 0.23606474697589874, 0.06656751036643982, 0.5585213899612427, 0.4746895432472229, 0.14414533972740173, -0.024808110669255257, 0.3020041584968567, 0.1700400412082672, 0.0851999968290329, 0.09485101699829102, 0.3581473231315613, -0.07355272024869919, -0.20108456909656525, 0.35728219151496887, 0.14081469178199768, -0.14786876738071442, -0.08919966220855713, -0.04618338122963905, -0.284812867641449, -0.2670905292034149, 0.17356161773204803, -0.32631319761276245, 0.2059573382139206, 0.2735864222049713, 0.1817912757396698, -0.1691218614578247, -0.4797636568546295, 0.05117836222052574, 0.32504409551620483, 0.2547908425331116, 0.25988906621932983, -0.4101264774799347, 0.18531833589076996, -0.548197865486145, 0.1673712283372879, 0.12330935895442963, 0.2696002423763275, -0.18979103863239288, -0.17715856432914734, 0.04420550912618637, -0.15119192004203796, 0.29357293248176575, -0.5013009309768677, -0.010340968146920204, -0.11806613206863403, -0.018266137689352036, -0.19106526672840118, -0.17548471689224243, 0.01709083467721939, 0.3075025975704193, 0.43538376688957214, 0.3108360767364502, -0.477939248085022, -0.07290248572826385, 0.11803130060434341, -0.08828974515199661, -0.1425059735774994, -0.22670400142669678, -0.17830948531627655, -0.2467595487833023, -0.17376552522182465, -0.14634254574775696, -0.06844373047351837, 0.20476174354553223, -0.0036695136222988367, 0.05763374641537666, 0.0780763328075409, 0.1945192515850067, 0.26092129945755005, 0.23615522682666779, 0.24638833105564117, -0.07365662604570389, 0.10551038384437561, 0.344176709651947, 0.38107186555862427, 0.045800842344760895, 0.4809632897377014, -0.04737718775868416, -0.0333527997136116, 0.03335115313529968, -0.15820349752902985, -0.05585439130663872, 0.4643746316432953, 0.14981040358543396, 0.058227378875017166, 0.04612553492188454, -0.1169341653585434, -0.3775767982006073, 0.0007062580552883446, -0.055076103657484055, -0.14003051817417145, -0.29990699887275696, -0.40095624327659607, 0.261601984500885, 0.3482019007205963, -0.1439940333366394, -0.14029110968112946, -0.17629589140415192, -0.3403513729572296, 0.2590389549732208, 0.23190838098526, 1.04195237159729, 0.28712815046310425, 0.07867696136236191, -0.14152082800865173, -0.19598767161369324, 0.8950440883636475, -0.45846959948539734, 0.38842883706092834, -0.37975841760635376, -0.16995207965373993, -0.04971786215901375, -0.13921768963336945, 0.06719168275594711, 0.07828181982040405, -0.3491458296775818, 0.4424484968185425, 0.23112912476062775, 0.06170234456658363, -0.14359599351882935, 0.3393727242946625, -0.24981382489204407, -0.18752287328243256, -0.1950543075799942, -0.005369895603507757, -0.23499374091625214, 0.34888219833374023, -0.009586174972355366, -0.08522570133209229, -0.059861309826374054, -0.182704359292984, -0.3571864068508148, 0.11464280635118484, 0.0939713642001152, 0.012490744702517986, 0.23694084584712982, -0.20036745071411133, -0.11602853983640671, 0.09369588643312454, 0.36601394414901733, -0.18719875812530518, -0.16472142934799194, -0.050345368683338165, -0.13756588101387024, -0.07333612442016602, 0.11516939848661423, 0.06339471787214279, 0.3432280123233795, -0.08296255767345428, 0.061176225543022156, 0.11704758554697037, -0.1985531449317932, -0.4443513751029968, -0.21357157826423645, -0.13070213794708252, -0.14821800589561462, -0.11806493252515793, -0.023201489821076393, 0.08216460794210434, -0.013901160098612309, -0.05310792848467827, 0.042126256972551346, 0.0036880308762192726, -0.15956957638263702, 0.3063168525695801, -0.25279363989830017, -0.23129938542842865, -0.16731888055801392, 0.3863881230354309, 0.35395264625549316, -0.022715670987963676, 0.38353434205055237, -0.1494130790233612, -0.1633327305316925, -0.14232711493968964, -0.220254048705101, -0.24535730481147766, -0.3556593358516693, 0.11342097073793411, -0.32436898350715637, -0.2375129908323288, 0.0965530127286911, 0.19885846972465515, 0.1759381741285324, 0.3981854319572449, -0.059355054050683975, -0.4329327344894409, -0.08512745052576065, 0.16189219057559967, -0.08562620729207993, 0.10745801031589508, -0.034591272473335266, 0.2953035831451416, -0.23416543006896973, 0.3081858456134796, -0.22415941953659058, -0.2144247591495514, -0.3940699100494385, 0.2727060914039612, 0.08486264199018478, -0.03858989477157593, 0.014369008131325245, 0.045115288347005844, -0.21619157493114471, 0.2673085331916809, -0.1653190404176712, 0.005659074988216162, -0.051704321056604385, 0.20624282956123352, 0.13152091205120087, -0.03179598227143288, -0.18519632518291473, 0.12904204428195953, 0.09241293370723724, 0.04547376558184624, 0.028971536085009575, 0.2819371521472931, 0.040343862026929855, 0.403765469789505, -0.2569805681705475, 0.08450248092412949, -0.01948445662856102, 0.20341457426548004, -0.27178582549095154, 0.324747771024704, -0.08645588904619217, 0.30650508403778076, -0.20353427529335022, 0.029180996119976044, -0.02076478861272335, 0.07677080482244492, 0.12723584473133087, 0.011338811367750168, 0.1532256007194519, -0.142635315656662, 0.027951473370194435, 0.27671846747398376, -0.07626078277826309, 0.557901918888092, -0.11840227991342545, -0.27169784903526306, 0.13532356917858124, 0.0732983946800232, -0.02431321330368519, -0.06569481641054153, 0.09087392687797546, 0.02758839726448059, -0.07582658529281616, 0.42469170689582825, 0.2287529706954956, -0.13601508736610413, -0.012512724846601486, 0.23883222043514252, 0.48960036039352417, -0.2690233290195465, 0.07844089716672897, 0.1427542120218277, -0.1867327243089676, 0.057876162230968475, 0.3953358829021454, -0.23366136848926544, 0.14236798882484436, 0.06332115083932877, 0.113039031624794, 0.46257084608078003, 0.20257923007011414, 0.3750729560852051, -0.18873676657676697, -0.10802369564771652, -0.056961771100759506, 0.6232266426086426, 0.2760690152645111, 0.33051931858062744, 0.1989181488752365, 0.3161472678184509, -0.14343416690826416, 0.010835017077624798, -0.33690357208251953, -0.18920433521270752, -0.03781334310770035, -0.14686629176139832, -0.3757626414299011, -0.09626255929470062, -0.08604573458433151, 0.08358065783977509, -0.13657616078853607, 0.10390136390924454, 0.5570070147514343, -0.17096586525440216, -0.18252676725387573, -0.5047012567520142, 0.2223159372806549, 0.003294002264738083, 0.20471413433551788, -0.22498585283756256, 0.3006284534931183, -0.0441913977265358, -0.2016669064760208, 0.28789830207824707, 0.5239024758338928, 0.425033837556839, 0.29359671473503113, -0.04037963226437569, -0.1795707494020462, 0.07012078911066055, 0.1715540885925293, 0.19459746778011322, 0.7336777448654175, 0.2264082133769989, -0.36697906255722046, 0.27984780073165894, -0.07750450074672699, -0.04813347011804581, 0.4749353229999542, 0.010236356407403946, 0.03268752992153168, -0.0273013636469841, 0.42502889037132263, 0.010575998574495316, 0.11825019121170044, 0.11753829568624496, 0.14925137162208557, -0.22901615500450134, -0.285824716091156, 0.13254712522029877, -0.17832788825035095, 0.28705671429634094, -0.14238835871219635, 0.04563367739319801, -0.208994522690773, 0.6096136569976807, 0.4575616121292114, -0.23569664359092712, -0.22168295085430145, 0.0019249219913035631, -0.5511803030967712, 0.3374859690666199, -0.3613300025463104, -0.20019187033176422, -0.17516089975833893, 0.5507521629333496, -0.34127509593963623, 0.3287826478481293, -0.3048567473888397, 0.3189481496810913, -0.03687405213713646, 0.030297020450234413, -0.2732524871826172, -0.09151341766119003, -0.08998353034257889, -0.14913126826286316, 0.08855627477169037, -0.2684873044490814, 0.25103896856307983, 0.23709242045879364, -0.10101060569286346, -0.020666291937232018, -0.027044784277677536, 0.17944826185703278, 0.5358685851097107, -0.02898695319890976, 0.5525885224342346, 0.21866723895072937, -0.11766812950372696, -0.1761791855096817, -0.21538430452346802, -0.24386966228485107, -0.3098510205745697, 0.15333765745162964, 0.13062867522239685, 0.3093512952327728, -0.27565744519233704, 0.49836021661758423, -0.1463242471218109, 0.4165276885032654, -0.01683097332715988, 0.00833418220281601, -0.23042701184749603, -0.04797769710421562, -0.09354934096336365, 0.12309977412223816, 0.18964487314224243, 0.41267913579940796, 0.17846405506134033, 0.141026571393013, -0.19011090695858002, -0.3322608470916748, 0.3096320927143097, -0.35968977212905884, 0.08010376244783401, -0.03542778640985489, 0.2624865472316742, 0.2798443138599396, -0.2487337738275528, -0.7318595051765442, -0.06884338706731796, 0.2349819391965866, -0.12555165588855743, 0.0618446059525013, 0.045738335698843, -0.11327081173658371, -0.2084636688232422, -0.09276928752660751, -0.1714307814836502, 0.28795915842056274, -0.3046320974826813, 0.3468710780143738, -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 ```
[ 0.04782287776470184, -0.01499373558908701, 0.09996049851179123, 0.48276638984680176, 0.252623051404953, 0.11827041208744049, 0.12425930052995682, 0.4629976451396942, 0.20459404587745667, -0.11430198699235916, 0.1954910308122635, 0.5628966093063354, -0.11228950321674347, -0.2145102620124817, -0.06854420900344849, -0.18655364215373993, 0.11651553213596344, 0.24228167533874512, -0.003951522056013346, -0.019452199339866638, -0.12144707143306732, 0.12436161190271378, -0.26443982124328613, 0.16095416247844696, -0.4509822428226471, -0.1343632936477661, 0.10060327500104904, 0.07057586312294006, -0.16652822494506836, -0.45291265845298767, 0.13311618566513062, -0.3863157629966736, -0.01449435856193304, 0.26769042015075684, -0.00012025610340060666, -0.10239151865243912, 0.3526807427406311, 0.04843990132212639, -0.2610151171684265, -0.24791716039180756, -0.4819885790348053, 0.09865826368331909, 0.30856218934059143, -0.051061537116765976, 0.22519220411777496, 0.024379700422286987, 0.023879488930106163, -0.5974245667457581, 0.18228338658809662, 0.5890494585037231, 0.18119513988494873, 0.3117515444755554, -0.05861635506153107, 0.08240847289562225, 0.22701819241046906, 0.08643072098493576, -0.3474990129470825, 0.15860097110271454, -0.04186518117785454, -0.16374847292900085, -0.09725937992334366, 0.42489778995513916, -0.1737871617078781, 0.12876741588115692, 0.6042643189430237, 0.10137630254030228, 0.10312210768461227, -0.3009955883026123, 0.18946418166160583, 0.002262386027723551, 0.27389997243881226, -0.18591836094856262, -0.08067546784877777, 0.030915185809135437, -0.14413128793239594, -0.36947283148765564, 0.20961248874664307, 0.17653074860572815, -0.3138512670993805, 0.01766219735145569, 0.01131492666900158, 0.08406320214271545, -0.06323167681694031, -0.07367584854364395, -0.059721991419792175, 0.06742102652788162, 0.013082757592201233, 0.1142643466591835, -0.05431639030575752, -0.1860046535730362, -0.18972696363925934, -0.22003130614757538, -0.20488381385803223, 0.24149301648139954, -0.3378831446170807, 0.04840727522969246, -0.057925377041101456, -0.20891262590885162, -0.1648835837841034, -0.11893317103385925, 0.2784707248210907, -0.0002532057114876807, 0.14469996094703674, 0.12964124977588654, 0.33332404494285583, 0.28613728284835815, 0.32042112946510315, 0.24878503382205963, 0.09861896932125092, -0.3073052167892456, -0.18543606996536255, -0.013636868447065353, -0.011412159539759159, -0.009702508337795734, 0.514110803604126, 0.016387760639190674, 0.30025094747543335, -0.36010584235191345, -0.3549935817718506, -0.011048955842852592, -0.735805332660675, 0.15996381640434265, -0.16902349889278412, 0.10665591061115265, -0.008254085667431355, 0.4835464060306549, 0.24768288433551788, 0.2814024090766907, -0.33920010924339294, -0.1926283836364746, -0.11420819163322449, -0.0042968448251485825, -0.2985198199748993, 0.015622679144144058, 0.3052986264228821, -0.03636542707681656, 0.05392400175333023, 0.06570790708065033, -0.40770700573921204, -0.20890949666500092, 0.37390971183776855, -0.4226529002189636, 0.17886769771575928, -0.008707615546882153, -0.32873713970184326, 0.4152504801750183, 0.24119026958942413, -0.4089016616344452, -0.21291399002075195, 0.17159046232700348, -0.13049687445163727, 0.03689742833375931, -0.1928088665008545, 0.12478375434875488, 0.15921339392662048, -0.2186257392168045, -0.1702643483877182, -0.06532663106918335, 0.3925437927246094, -0.10832776129245758, 0.10371731221675873, -0.29295504093170166, -0.04032422602176666, -0.24107548594474792, -0.08629397302865982, 0.47886693477630615, -0.22314198315143585, 0.06661126762628555, -0.055577512830495834, 0.1835537850856781, 0.36414003372192383, 0.3781099319458008, -0.020374132320284843, 0.17711544036865234, 0.10718656331300735, 0.28203487396240234, 0.4739598333835602, -0.30076542496681213, -0.10005531460046768, 0.0992065966129303, -0.028822122141718864, -0.007086739409714937, -0.04180622473359108, 0.12140338867902756, 0.07706901431083679, -0.12026635557413101, 0.16744625568389893, 0.19477051496505737, -0.06231936439871788, 0.23532551527023315, -0.42068955302238464, -0.10153261572122574, 0.5093356966972351, -0.1742011457681656, -0.07566080242395401, 0.026094771921634674, -0.07687555998563766, 0.05020191892981529, -0.06762654334306717, -0.013059054501354694, 0.428176611661911, 0.2600767910480499, -0.09044042974710464, -0.08258427679538727, 0.0832560732960701, -0.18829387426376343, -0.38731589913368225, -0.052016448229551315, -0.10536517202854156, 0.13355879485607147, -0.17240865528583527, -0.04038713499903679, -0.34554558992385864, 0.07457040995359421, -0.02713192068040371, -0.19891318678855896, 0.12534509599208832, 0.10824761539697647, 0.15891990065574646, 0.05802815407514572, -0.22786828875541687, 0.1251097172498703, 0.0048242248594760895, 0.12739740312099457, -0.39792609214782715, -0.14816665649414062, -0.08023770153522491, -0.35545337200164795, -0.09916698187589645, 0.1799587458372116, 0.23762933909893036, 0.16755390167236328, -0.2698639929294586, 0.18731684982776642, 0.018680427223443985, 0.2171434611082077, -0.25498172640800476, 0.029901184141635895, 0.018767017871141434, -0.40214940905570984, 0.22020144760608673, 0.3269733190536499, 0.3616490066051483, -0.24359026551246643, -0.396494060754776, -0.02146218530833721, 0.014599384739995003, 0.5858549475669861, -0.13255612552165985, 0.10853222757577896, -0.0767752155661583, -0.03623275086283684, 0.09410489350557327, -0.07976377010345459, 0.09601124376058578, -0.1133064329624176, 0.1979532092809677, 0.29243695735931396, -0.30972328782081604, 0.2726483643054962, 0.42169561982154846, 0.30025026202201843, 0.13427652418613434, 0.159810870885849, -0.3345215916633606, -0.08566444367170334, 0.26441654562950134, -0.012384884990751743, 0.46109122037887573, -0.043286729604005814, 0.2145605832338333, -0.08492756634950638, -0.15070241689682007, -0.14859645068645477, 0.35784000158309937, -0.09661873430013657, 0.01726602390408516, -0.0092660216614604, 0.49071139097213745, 0.010626101866364479, -0.1787145733833313, 0.20089192688465118, 0.005513889715075493, 0.23830604553222656, -0.3552827537059784, 0.3290382921695709, -0.33911269903182983, -0.1678951382637024, -0.38104119896888733, -0.2544398605823517, -0.018507881090044975, -0.5028191208839417, -0.1405358463525772, 0.07923384010791779, 0.006112152710556984, 0.12038957327604294, -0.25411632657051086, -0.14185765385627747, 0.018085617572069168, -0.11071284860372543, -0.33473527431488037, -0.3540615737438202, -0.37050169706344604, -0.06328784674406052, 0.46462732553482056, 0.32319602370262146, 0.18738240003585815, 0.0530339814722538, -0.06295061856508255, -0.06901077926158905, -0.36038732528686523, 0.03056301921606064, -0.19892184436321259, 0.057185735553503036, 0.12039303034543991, 0.07614051550626755, -0.136030912399292, -0.12729710340499878, 0.42562100291252136, 0.003182713408023119, -0.2406022548675537, 0.329190731048584, -0.020771285519003868, 0.04902200773358345, -0.2927949130535126, -0.48070812225341797, 0.013537250459194183, -0.4682866632938385, 0.13702988624572754, -0.11211804300546646, 0.2093205600976944, 0.1776328682899475, 0.12377601116895676, 0.31068408489227295, -0.05574404075741768, 0.18512551486492157, 0.20702672004699707, 0.3653651773929596, 0.6162763833999634, -0.13313642144203186, -0.23794665932655334, 0.2084658145904541, -0.07696927338838577, 0.3138101100921631, 0.20748816430568695, -0.1788216084241867, 0.33276981115341187, -0.4218789041042328, 0.3207613229751587, -0.1633068323135376, -0.1305055171251297, 0.22173672914505005, 0.2210211306810379, -0.04805248603224754, -0.060437850654125214, -0.0075170015916228294, 0.2183128297328949, -0.2177659124135971, 0.15179331600666046, 0.12625369429588318, 0.21420960128307343, -0.05345010757446289, 0.38171979784965515, 0.14399801194667816, -0.22840730845928192, 0.28303128480911255, 0.016808966174721718, -0.047595687210559845, -0.3022211194038391, -0.27842625975608826, -0.0881352499127388, -0.21157050132751465, 0.12770923972129822, -0.17738941311836243, 0.03394405543804169, -0.4855812191963196, -0.022265536710619926, 0.2889741063117981, -0.1829327791929245, -0.13431286811828613, 0.3022044897079468, -0.21380233764648438, 0.29014071822166443, 0.1346106231212616, 0.1258193701505661, -0.19792313873767853, -0.41243889927864075, 0.032936569303274155, -0.024657029658555984, 0.1568613499403, 0.024755194783210754, -0.10193730145692825, -0.05289766937494278, -0.47397783398628235, 0.28840139508247375, 0.17515558004379272, 0.4326752722263336, -0.07643638551235199, -0.14325161278247833, 0.19170793890953064, -0.11352070420980453, 0.29864081740379333, -0.1461237519979477, -0.0275951586663723, 0.37580907344818115, 0.27237647771835327, -0.4777471423149109, 0.26930779218673706, -0.2311147153377533, 0.34930220246315, 0.26166728138923645, 0.283454954624176, -0.3896859586238861, -0.3131129741668701, 0.25790199637413025, 0.16102828085422516, 0.016651904210448265, 0.08997220546007156, -0.3169800937175751, -0.0035179981496185064, -0.15553458034992218, -0.08673817664384842, 0.27570462226867676, 0.310602068901062, 0.082622230052948, 0.054902177304029465, -0.33018678426742554, 0.1789494901895523, 0.18064174056053162, 0.026372577995061874, 0.42826664447784424, -0.3036718964576721, 0.13769154250621796, -0.1568317860364914, 0.20898035168647766, -0.08677227795124054, 0.3240395486354828, -0.06348997354507446, -0.19840136170387268, 0.26688021421432495, -0.05753711611032486, 0.36443641781806946, 0.3292664587497711, 0.03609272092580795, 0.3470538258552551, -0.12237446755170822, -0.0959215760231018, -0.20330095291137695, 0.02960517629981041, 0.16758930683135986, -0.052809569984674454, 0.04342726990580559, -0.2120436578989029, 0.4822666645050049, 0.298149436712265, -0.03357063606381416, 0.6272772550582886, 0.1370311826467514, -0.28337642550468445, 0.23740075528621674, 0.18653206527233124, 1.1821579933166504, 0.15122579038143158, 0.14966827630996704, 0.13378287851810455, -0.4301355481147766, 0.3339086174964905, 0.06193789839744568, 0.325460284948349, -0.46283650398254395, 0.32495638728141785, 0.024498585611581802, -0.16893252730369568, 0.0673145279288292, -0.033895980566740036, -0.3024047315120697, 0.08304499089717865, -0.19602973759174347, 0.3092464804649353, 0.11238525807857513, 0.05807110667228699, -0.3026752769947052, -0.3165785074234009, -0.3749186396598816, 0.08060049265623093, -0.25055015087127686, -0.07662521302700043, 0.11036092787981033, -0.2155655175447464, -0.49149274826049805, -0.04550549015402794, -0.0685393437743187, 0.11479715257883072, -0.17067162692546844, 0.04652312025427818, 0.2367296814918518, -0.15768668055534363, 0.18111729621887207, 0.07256678491830826, -0.10370739549398422, -0.06586918234825134, -0.1338339000940323, -0.08608777821063995, 0.1967160850763321, 0.10096609592437744, 0.1206190288066864, -0.0757206380367279, 0.011200969107449055, 0.07623343169689178, 0.024821698665618896, 0.22761788964271545, -0.3680090606212616, -0.013541732914745808, -0.06473128497600555, 0.16641736030578613, 0.11926659196615219, -0.09275045990943909, -0.08846552670001984, 0.09907550364732742, -0.06051855906844139, 0.003113249782472849, 0.09320326149463654, 0.02566823549568653, 0.15167862176895142, 0.08057168126106262, -0.15691056847572327, -0.26000767946243286, 0.2690161168575287, 0.341331422328949, 0.04775964096188545, 0.24060401320457458, 0.7169974446296692, 0.059508442878723145, -0.18554158508777618, -0.1779445856809616, 0.07761102169752121, -0.05239583179354668, -0.7310898303985596, 0.3293781578540802, -0.3314170241355896, -0.2210783064365387, 0.03403409197926521, 0.45446813106536865, -0.20293815433979034, 0.08033964782953262, -0.1594124138355255, -0.2330784946680069, -0.21894456446170807, 0.17904840409755707, -0.030600422993302345, 0.034671682864427567, -0.45244818925857544, -0.05752864107489586, 0.09664376080036163, 0.12149567902088165, -0.22719337046146393, -0.010041764006018639, -0.17429354786872864, 0.3713476061820984, 0.09213235229253769, 0.03296452388167381, -0.050732456147670746, -0.18050986528396606, -0.011443926952779293, 0.5151131749153137, -0.0971551463007927, -0.14438803493976593, -0.2621200382709503, 0.14715495705604553, 0.07113245129585266, -0.1661830097436905, 0.07055412232875824, 0.024171914905309677, 0.07843512296676636, -0.21286918222904205, -0.15107883512973785, -0.10288645327091217, -0.1862982213497162, 0.05603847652673721, -0.08769993484020233, 0.20297834277153015, -0.17871834337711334, 0.09925536066293716, -0.14713378250598907, 0.056820325553417206, -0.24345898628234863, 0.10793516784906387, 0.21070460975170135, 0.05712292715907097, -0.20819702744483948, 0.04877778887748718, -0.09463672339916229, -0.15886454284191132, 0.30473557114601135, -0.37726423144340515, -0.16618822515010834, 0.05655044689774513, -0.19067196547985077, 0.41663098335266113, -0.09205548465251923, 0.002051863120868802, -0.0814971998333931, 0.09775546193122864, 0.07033379375934601, 0.31662464141845703, -0.17869514226913452, -0.16311214864253998, -0.11346429586410522, 0.19117067754268646, -0.02398461475968361, -0.15496575832366943, -0.030290868133306503, 0.09489066898822784, 0.033047500997781754, -0.44211697578430176, -0.164072647690773, 0.1360536366701126, -0.025477739050984383, -0.07160664349794388, 0.35693663358688354, -0.10453932732343674, 0.021537678316235542, 0.020302725955843925, 0.36647799611091614, 0.33404630422592163, 0.4364643394947052, 0.20343396067619324, -0.06764330714941025, 0.11950594186782837, -0.02497223950922489, 0.38919928669929504, 0.2442471832036972, 0.15110088884830475, 0.17533552646636963, 0.5244447588920593, -0.29807859659194946, -0.3281242251396179, 0.06573376059532166, 0.10832871496677399, -0.27272337675094604, -0.0876118391752243, -0.5703155994415283, -0.11522655189037323, -0.4419066309928894, -0.0880984216928482, -0.19053766131401062, 0.2544894814491272, 0.692865252494812, 0.23379051685333252, 0.016697481274604797, -0.5845616459846497, -0.12988463044166565, 0.13295462727546692, 0.3920844793319702, -0.24886012077331543, 0.4035947024822235, 0.10868512094020844, -0.07600285857915878, -0.13055263459682465, 0.49973464012145996, 0.46834486722946167, 0.5678134560585022, -0.28837940096855164, -0.06765210628509521, 0.05808155983686447, -0.18258094787597656, 0.23777775466442108, 0.5766584277153015, 0.3382429778575897, 0.14017944037914276, 0.03504369407892227, -0.0379658006131649, -0.1073760986328125, -0.09008006006479263, -0.029721369966864586, 0.27744683623313904, -0.16376084089279175, 0.2949458062648773, -0.23841068148612976, -0.2226387858390808, -0.046252477914094925, -0.21092990040779114, -0.3650150001049042, -0.04621884599328041, 0.2055167406797409, 0.04957400634884834, 0.015669140964746475, -0.33418184518814087, 0.057421352714300156, 0.2021227777004242, 0.5171067714691162, 0.2711423337459564, 0.13861320912837982, -0.47538435459136963, -0.11695386469364166, -0.7315421104431152, 0.17639417946338654, -0.3149000108242035, -0.15671318769454956, -0.01451869122684002, 0.36327865719795227, 0.045231569558382034, 0.314698725938797, -0.034921642392873764, 0.2289496809244156, 0.25130149722099304, 0.30536088347435, -0.3416901230812073, -0.11979299783706665, 0.3530432879924774, -0.3037295937538147, 0.023086393252015114, -0.5888966917991638, 0.26860249042510986, 0.05124136433005333, 0.06511209160089493, -0.008692586794495583, -0.2844812572002411, 0.2985394597053528, 0.48924577236175537, 0.27550697326660156, 0.26379212737083435, 0.2856043875217438, 0.2475820928812027, -0.5557767152786255, -0.21760626137256622, -0.1734551042318344, -0.20772720873355865, -0.02030578814446926, 0.08986926078796387, 0.5805660486221313, -0.14044851064682007, 0.06243743747472763, -0.3035687804222107, 0.3277711272239685, -0.22651749849319458, 0.03437754511833191, -0.1755726933479309, 0.06700565665960312, -0.34905537962913513, 0.16302521526813507, 0.07218949496746063, 0.12040051072835922, 0.1978316307067871, 0.24254406988620758, -0.5347839593887329, -0.3874291777610779, 0.30438604950904846, -0.519407331943512, -0.4466893970966339, 0.293190598487854, 0.1429884284734726, 0.44685113430023193, -0.012415909208357334, -0.7071852684020996, -0.26888057589530945, 0.21321892738342285, -0.0964372307062149, 0.025330590084195137, 0.023005368188023567, 0.026595422998070717, 0.07402954995632172, -0.19160310924053192, 0.36198484897613525, 0.008301947265863419, 0.049280859529972076, 0.04832898825407028, -0.22841104865074158 ]
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.
[ 0.12082318216562271, 0.0813007578253746, 0.08892317861318588, 0.408298522233963, 0.28144094347953796, 0.054601505398750305, 0.11376463621854782, 0.336548775434494, 0.18600939214229584, -0.18695002794265747, 0.2484576553106308, 0.5634831786155701, 0.0009862850420176983, -0.07831835746765137, -0.023558616638183594, -0.11811479926109314, 0.07306355983018875, 0.2508918046951294, 0.05062994360923767, -0.12445284426212311, -0.0924607440829277, 0.16164350509643555, -0.18193919956684113, 0.22253046929836273, -0.5130985975265503, -0.15872825682163239, 0.14412766695022583, 0.08026690781116486, -0.1196368932723999, -0.49901169538497925, 0.015516167506575584, -0.3473424017429352, 0.0308180283755064, 0.20836132764816284, -0.00011913953494513407, -0.1659439355134964, 0.3734931945800781, -0.052215103060007095, -0.1960315704345703, -0.21079623699188232, -0.5531389117240906, 0.1664823293685913, 0.3032819330692291, -0.06724192947149277, 0.23992842435836792, -0.10213454067707062, 0.06580587476491928, -0.5268752574920654, 0.09618109464645386, 0.4953773319721222, 0.19496862590312958, 0.1500275582075119, 0.11893060058355331, 0.08554046601057053, 0.26706448197364807, 0.11917169392108917, -0.2874544560909271, 0.043262265622615814, -0.13040447235107422, -0.2241946905851364, -0.06938821822404861, 0.34750625491142273, -0.04919620603322983, -0.00454467348754406, 0.5662892460823059, 0.044567957520484924, 0.19759435951709747, -0.18642765283584595, 0.2564823627471924, 0.058506790548563004, 0.35495689511299133, -0.061708033084869385, -0.10864054411649704, 0.09885500371456146, -0.12332399189472198, -0.3633820414543152, 0.28847548365592957, 0.22252194583415985, -0.2659066915512085, -0.012291861698031425, 0.11637645959854126, 0.06312402337789536, 0.00971482414752245, -0.12463652342557907, -0.1565367728471756, 0.14945465326309204, 0.018611717969179153, 0.11266802996397018, -0.05011352151632309, -0.1976817101240158, -0.18047164380550385, -0.07938512414693832, -0.3031899631023407, 0.21402019262313843, -0.2830847501754761, 0.01163876336067915, -0.051603760570287704, -0.16387298703193665, -0.20187902450561523, -0.0879097729921341, 0.4137456715106964, -0.00014438063954003155, -0.0053522647358477116, 0.06754860281944275, 0.3144072890281677, 0.4288618564605713, 0.46972107887268066, 0.09827239066362381, 0.13660790026187897, -0.304889440536499, -0.18265296518802643, -0.034475892782211304, -0.025242239236831665, -0.009740060195326805, 0.43271729350090027, 0.024083580821752548, 0.07819661498069763, -0.304749995470047, -0.2590410113334656, 0.02684151381254196, -0.7115641236305237, 0.22196777164936066, -0.21643325686454773, 0.13746309280395508, 0.06488554179668427, 0.40737178921699524, 0.2952428460121155, 0.29609623551368713, -0.20428626239299774, -0.20120781660079956, -0.11234617978334427, -0.09873449802398682, -0.3487413823604584, -0.15234221518039703, 0.19270949065685272, -0.08550131320953369, 0.08561785519123077, 0.02165353298187256, -0.5190793871879578, -0.22347432374954224, 0.2967345714569092, -0.2855612337589264, 0.2961820662021637, 0.06044938415288925, -0.2745220959186554, 0.4591626226902008, 0.21227797865867615, -0.5164349675178528, -0.09735002368688583, 0.09279739856719971, -0.20460103452205658, 0.07067336142063141, -0.181947261095047, 0.13979122042655945, 0.037309423089027405, -0.2744906544685364, -0.20205354690551758, -0.06351959705352783, 0.38846561312675476, -0.1685948371887207, 0.09392953664064407, -0.20465312898159027, -0.01268838532269001, -0.2841031551361084, -0.14891183376312256, 0.3834085762500763, -0.18657825887203217, 0.10843689739704132, -0.0751800611615181, 0.16698159277439117, 0.30440768599510193, 0.3455915153026581, 0.02288292907178402, 0.1997976154088974, 0.10551238059997559, 0.35004371404647827, 0.3927367627620697, -0.22957955300807953, -0.00972948968410492, -0.0022535522002726793, -0.017602289095520973, -0.007658645045012236, -0.0582815483212471, -0.002801567083224654, 0.04202045872807503, -0.11749878525733948, 0.0826951190829277, 0.12329750508069992, 0.01209899690002203, 0.25914663076400757, -0.3936404585838318, -0.0621940977871418, 0.4800756871700287, -0.3127317726612091, -0.20604078471660614, -0.030394570901989937, 0.01649952121078968, 0.10142698138952255, -0.026429008692502975, -0.09374558925628662, 0.3080482482910156, 0.17479299008846283, -0.07066210359334946, -0.14800585806369781, 0.15718461573123932, -0.04129195958375931, -0.39961478114128113, -0.059613559395074844, -0.11000096052885056, 0.13414865732192993, -0.051153603941202164, 0.0029528990853577852, -0.2747977674007416, 0.05228067561984062, -0.007796375546604395, -0.25278130173683167, 0.14248062670230865, 0.06316812336444855, 0.16465024650096893, 0.12633119523525238, -0.21560586988925934, 0.06162581965327263, 0.02169225551187992, 0.12927883863449097, -0.3672282099723816, -0.26253029704093933, 0.010546207427978516, -0.33143290877342224, -0.1912831813097, 0.23127010464668274, 0.22903749346733093, 0.13432107865810394, -0.2543700933456421, 0.19253778457641602, -0.009425869211554527, 0.3920418918132782, -0.21428416669368744, 0.07450930029153824, -0.013634995557367802, -0.504682183265686, 0.19313791394233704, 0.41021183133125305, 0.29952168464660645, -0.2957395911216736, -0.40343642234802246, -0.053517669439315796, -0.031673673540353775, 0.5904878973960876, -0.09309858083724976, 0.2558080852031708, -0.020294230431318283, -0.06831350177526474, 0.07016131281852722, -0.1256333738565445, 0.1901371330022812, -0.16906927525997162, 0.17845727503299713, 0.23488160967826843, -0.23076637089252472, 0.31589871644973755, 0.38045716285705566, 0.3273272216320038, 0.21452654898166656, 0.03870788961648941, -0.3840384781360626, -0.06015659123659134, 0.27095603942871094, 0.06697291135787964, 0.36886170506477356, -0.059140849858522415, 0.22426290810108185, -0.13389995694160461, -0.15443645417690277, -0.17635594308376312, 0.32109057903289795, -0.20176970958709717, 0.07100459188222885, 0.06617750972509384, 0.5125982165336609, 0.017989687621593475, -0.18942204117774963, 0.3008752167224884, -0.016682151705026627, 0.23172098398208618, -0.41760361194610596, 0.23898693919181824, -0.3116365671157837, -0.05878394842147827, -0.3408748507499695, -0.21178224682807922, -0.08250030875205994, -0.5657788515090942, -0.03033333644270897, -0.03345084562897682, -0.0700712725520134, 0.03609154745936394, -0.11940205842256546, -0.14799292385578156, -0.10889875888824463, -0.07129647582769394, -0.3734091818332672, -0.4300118386745453, -0.2597951591014862, -0.04753977805376053, 0.3729805648326874, 0.23948872089385986, 0.10474346578121185, -0.012625285424292088, -0.07535816729068756, 0.0003269163426011801, -0.4714285433292389, -0.004124713130295277, -0.2229703664779663, 0.06794353574514389, 0.18282301723957062, 0.02107108011841774, -0.030382636934518814, -0.01443744357675314, 0.36036181449890137, -0.0667395368218422, -0.2764347195625305, 0.35011234879493713, 0.06635923683643341, 0.05440802499651909, -0.29347917437553406, -0.3115222156047821, 0.08216547966003418, -0.46678465604782104, 0.3006451725959778, -0.096478670835495, 0.24163851141929626, 0.21147549152374268, -0.017977043986320496, 0.22887583076953888, -0.06533204019069672, 0.16171419620513916, 0.18838953971862793, 0.4239363670349121, 0.5407384037971497, -0.2293756902217865, -0.3174419105052948, 0.29581254720687866, -0.03764568641781807, 0.3881961405277252, 0.14681091904640198, -0.04843804985284805, 0.40579092502593994, -0.3843775689601898, 0.13194186985492706, -0.14741834998130798, -0.07091407477855682, 0.18097704648971558, 0.2352914810180664, -0.05996188148856163, -0.1455630660057068, 0.03315688669681549, 0.21858535706996918, -0.03075638972222805, 0.24850617349147797, 0.09110963344573975, 0.004534481093287468, -0.11354284733533859, 0.3820226192474365, 0.3040759861469269, -0.15708638727664948, 0.2629912793636322, 0.006090134847909212, -0.06353586167097092, -0.23008902370929718, -0.2707284986972809, -0.1288837492465973, -0.17353962361812592, 0.15583157539367676, -0.15172068774700165, 0.03358360379934311, -0.38190972805023193, -0.031434688717126846, 0.3377803564071655, -0.185100719332695, -0.14558905363082886, 0.2932656705379486, -0.2784808278083801, 0.23703889548778534, 0.08548399806022644, 0.1900608092546463, -0.12471495568752289, -0.407335489988327, 0.12500064074993134, -0.0006933578988537192, 0.1178487241268158, 0.03632887080311775, -0.1986619532108307, -0.12969739735126495, -0.4102334976196289, 0.31972983479499817, 0.039934124797582626, 0.36018669605255127, -0.1612703800201416, -0.15637290477752686, 0.1840427964925766, -0.15219642221927643, 0.4350282549858093, -0.36648014187812805, -0.059063032269477844, 0.30513790249824524, 0.34210556745529175, -0.4739353060722351, 0.34191977977752686, -0.2217642068862915, 0.2468215525150299, 0.16475601494312286, 0.14953553676605225, -0.3234815001487732, -0.23489408195018768, 0.17474143207073212, 0.020166169852018356, 0.02373514696955681, 0.17043374478816986, -0.33508649468421936, 0.026662146672606468, -0.07697547972202301, -0.17390745878219604, 0.20773646235466003, 0.24202142655849457, 0.018690073862671852, 0.07589276134967804, -0.4179534316062927, 0.1937553584575653, 0.21744440495967865, 0.08719127625226974, 0.399777889251709, -0.24007323384284973, 0.01152730267494917, -0.020229453220963478, 0.19943824410438538, -0.07045430690050125, 0.45299145579338074, -0.04165373742580414, -0.1785087138414383, 0.2557936906814575, -0.083693727850914, 0.4723662734031677, 0.38348934054374695, 0.0856892392039299, 0.32462579011917114, -0.2723216712474823, -0.22830942273139954, -0.32098644971847534, 0.0069497860968112946, 0.04117932915687561, 0.00267801433801651, 0.14303745329380035, -0.10382363945245743, 0.5234212279319763, 0.2486495077610016, -0.07634899765253067, 0.5133988261222839, 0.16703125834465027, -0.33578312397003174, 0.32315948605537415, 0.22793768346309662, 1.157378911972046, 0.05849013850092888, 0.10378313809633255, -0.05825201794505119, -0.46720436215400696, 0.440611332654953, 0.06488723307847977, 0.2514752447605133, -0.47175151109695435, 0.35401394963264465, 0.004888071212917566, -0.20057345926761627, 0.044348716735839844, -0.05194288119673729, -0.30089572072029114, 0.11113572120666504, -0.2160286158323288, 0.2448735237121582, 0.09667215496301651, 0.09772039204835892, -0.31009602546691895, -0.3052617609500885, -0.26761043071746826, 0.09327095746994019, -0.18762566149234772, -0.007991146296262741, 0.11706535518169403, -0.2904031276702881, -0.5326500535011292, -0.05518355593085289, -0.08857406675815582, 0.19176504015922546, -0.19879864156246185, -0.07285836338996887, 0.25250232219696045, -0.2095436453819275, 0.2399551272392273, 0.2320893406867981, 0.08816647529602051, -0.06347933411598206, -0.17391374707221985, -0.04728465899825096, 0.29204005002975464, 0.15689204633235931, 0.12021429091691971, -0.033129602670669556, 0.01418529637157917, 0.1007954478263855, 0.04560534283518791, 0.2517973482608795, -0.4120064675807953, 0.040763117372989655, -0.08722411841154099, 0.20695364475250244, 0.0925796627998352, -0.056116439402103424, 0.058333393186330795, 0.05692782998085022, -0.05763885751366615, -0.024936389178037643, 0.10847320407629013, 0.013143626973032951, 0.16904374957084656, 0.03570617735385895, -0.11537571996450424, -0.2564198672771454, 0.3108493685722351, 0.3111073970794678, -0.011715500615537167, 0.2638024091720581, 0.734185516834259, 0.027994589880108833, -0.15276385843753815, -0.19830739498138428, 0.02862994372844696, -0.024161167442798615, -0.7314626574516296, 0.31250956654548645, -0.3160664439201355, -0.367892324924469, 0.020080074667930603, 0.5516511797904968, -0.2353508621454239, 0.1846444457769394, -0.18532192707061768, -0.26003196835517883, -0.19181157648563385, 0.29425421357154846, -0.06804847717285156, 0.006813224870711565, -0.4693589210510254, -0.10694610327482224, 0.029502544552087784, 0.169457346200943, -0.2477700412273407, 0.03532423451542854, -0.1177477240562439, 0.3626188337802887, -0.04668591544032097, -0.005117643158882856, -0.004681900609284639, -0.2574218511581421, 0.01005152240395546, 0.5666002035140991, -0.12076964229345322, -0.14190931618213654, -0.2833370566368103, 0.14587777853012085, 0.08442196995019913, -0.1702071875333786, 0.16537943482398987, -0.017817465588450432, 0.06224702671170235, -0.2151457816362381, -0.07266779243946075, -0.20808857679367065, -0.1844860166311264, 0.07484980672597885, -0.02109970524907112, 0.2446187287569046, -0.2342161387205124, 0.08050157874822617, -0.07008993625640869, 0.06816431134939194, -0.3697086274623871, 0.0071485042572021484, -0.014693220146000385, 0.002423658734187484, -0.035316742956638336, 0.15229389071464539, -0.1864626109600067, -0.11169939488172531, 0.282320499420166, -0.29534295201301575, -0.12911127507686615, -0.0050055463798344135, -0.1942414492368698, 0.3379954695701599, -0.15246908366680145, 0.048097800463438034, -0.042307477444410324, 0.11493930965662003, 0.1366557478904724, 0.3571566641330719, -0.2620997428894043, 0.02110840566456318, -0.0831819698214531, 0.27575430274009705, 0.1245066225528717, -0.22823944687843323, -0.10933306813240051, 0.06607531011104584, -0.06436751782894135, -0.3699742257595062, -0.13122698664665222, 0.1212930977344513, -0.1415523886680603, -0.0027166150975972414, 0.3157491981983185, -0.15956249833106995, -0.03928849473595619, -0.0020726691000163555, 0.39843761920928955, 0.3103444278240204, 0.508105456829071, 0.2068909853696823, 0.0034973504953086376, 0.09589090943336487, 0.008702204562723637, 0.24191157519817352, 0.3457297384738922, 0.3615860641002655, 0.29702284932136536, 0.6127427816390991, -0.38364267349243164, -0.2681463956832886, -0.026009151712059975, 0.102792888879776, -0.2622459828853607, -0.14170044660568237, -0.583560049533844, -0.12105032056570053, -0.4420403242111206, -0.08055086433887482, -0.21707719564437866, 0.19709858298301697, 0.5802703499794006, 0.2196570634841919, -0.0454748161137104, -0.6290245652198792, -0.2098432183265686, 0.05817922204732895, 0.5017371773719788, -0.22398382425308228, 0.45599648356437683, 0.0160693172365427, -0.05682424455881119, -0.018257014453411102, 0.5782517790794373, 0.43931010365486145, 0.5464798808097839, -0.32134583592414856, -0.013242494314908981, 0.059082746505737305, -0.1702311784029007, 0.28284281492233276, 0.625999391078949, 0.37193650007247925, 0.07206609845161438, 0.049773089587688446, -0.02863713540136814, -0.0999382957816124, -0.2609720826148987, 0.021830735728144646, 0.45613375306129456, -0.009454252198338509, 0.21890711784362793, -0.29675400257110596, -0.13070404529571533, -0.07994186133146286, -0.2456391155719757, -0.30123281478881836, -0.07740312069654465, 0.1358340084552765, 0.01643814519047737, 0.020511003211140633, -0.3425389230251312, 0.07126214355230331, 0.16037502884864807, 0.5518670678138733, 0.26156020164489746, 0.12667244672775269, -0.417110413312912, -0.11879725009202957, -0.5855779647827148, 0.20088745653629303, -0.23853260278701782, -0.1507560759782791, -0.09952849894762039, 0.3261135220527649, 0.1053706631064415, 0.2445041835308075, -0.027682140469551086, 0.2034977525472641, 0.3321019113063812, 0.317922979593277, -0.30158963799476624, -0.062346164137125015, 0.3174630105495453, -0.32736736536026, 0.00846310704946518, -0.575022280216217, 0.29662445187568665, 0.06786375492811203, 0.06840580701828003, 0.13131755590438843, -0.3014718294143677, 0.20186877250671387, 0.4438701868057251, 0.3714680075645447, 0.21601645648479462, 0.2539946436882019, 0.2564387619495392, -0.4523238241672516, -0.2691608667373657, -0.20937904715538025, -0.23863402009010315, 0.10153625160455704, 0.10521861165761948, 0.5344889760017395, -0.08086108416318893, 0.1265621930360794, -0.25590890645980835, 0.15339775383472443, -0.17736391723155975, 0.1092863455414772, -0.26375189423561096, 0.14666573703289032, -0.3174298107624054, 0.08098752796649933, 0.12022746354341507, 0.12272718548774719, 0.1513156294822693, 0.1441611349582672, -0.48501792550086975, -0.4537809193134308, 0.3317624628543854, -0.37990859150886536, -0.33462366461753845, 0.39078158140182495, 0.2151692509651184, 0.4396241307258606, -0.10427286475896835, -0.7881212830543518, -0.35706236958503723, 0.1472138911485672, -0.13919539749622345, -0.05090014263987541, -0.11611425876617432, -0.05664416775107384, -0.005235396325588226, -0.21112190186977386, 0.3439016044139862, -0.008731525391340256, 0.08373036980628967, 0.090286485850811, -0.1682761013507843 ]
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.
[ 0.09355701506137848, -0.052180711179971695, 0.09173016250133514, 0.49227476119995117, 0.2525181472301483, 0.14921066164970398, 0.10871345549821854, 0.40769997239112854, 0.1446879506111145, -0.0935792550444603, 0.34540581703186035, 0.5846510529518127, -0.12629714608192444, -0.08148059248924255, -0.03906705975532532, -0.18967248499393463, 0.15724419057369232, 0.2686879336833954, 0.08629141002893448, -0.06003163754940033, -0.14310383796691895, 0.15054140985012054, -0.2789251208305359, 0.13217128813266754, -0.4200153648853302, -0.10315176844596863, 0.0767614021897316, 0.057116229087114334, -0.08841767907142639, -0.4528537690639496, 0.07614746689796448, -0.3256978690624237, -0.006575233303010464, 0.25914642214775085, -0.00012003633310087025, -0.1207483634352684, 0.41301584243774414, -0.040331266820430756, -0.26557815074920654, -0.249553844332695, -0.435833215713501, 0.03188949078321457, 0.33162370324134827, -0.04249090328812599, 0.18057318031787872, -0.063745878636837, 0.02380245551466942, -0.5875163078308105, 0.20610766112804413, 0.5833624005317688, 0.17648328840732574, 0.21741533279418945, -0.06840863823890686, 0.1302625983953476, 0.2942132353782654, 0.05315949022769928, -0.3383096754550934, -0.08270303159952164, 0.02263210341334343, -0.12178799510002136, -0.13316552340984344, 0.32143136858940125, -0.061751604080200195, 0.05857038497924805, 0.6247360110282898, 0.09327898919582367, 0.131283238530159, -0.3116702139377594, 0.14749658107757568, 0.1445363461971283, 0.39628154039382935, -0.08866605907678604, -0.08990998566150665, -0.01776966080069542, -0.1631641536951065, -0.35285326838493347, 0.2785201668739319, 0.180648073554039, -0.2984367907047272, -0.0033445130102336407, 0.07484900206327438, 0.14837560057640076, -0.08723291009664536, -0.06426297128200531, -0.11870861798524857, 0.15310707688331604, 0.03931673243641853, 0.06686517596244812, -0.10409839451313019, -0.2040967345237732, -0.10837477445602417, -0.2334074229001999, -0.25244686007499695, 0.3161037862300873, -0.4382796883583069, 0.037972696125507355, -0.05144258588552475, -0.13796301186084747, -0.14761584997177124, 0.06295923888683319, 0.3841148316860199, 0.09981708973646164, 0.027438869699835777, 0.07166073471307755, 0.3587128221988678, 0.36590608954429626, 0.31820422410964966, 0.1350378841161728, 0.0632365345954895, -0.22235749661922455, -0.08160142600536346, 0.017592336982488632, 0.016700301319360733, -0.039309874176979065, 0.4090128540992737, 0.009138804860413074, 0.1497364491224289, -0.33181342482566833, -0.24876725673675537, 0.0587410144507885, -0.7114651799201965, 0.23032695055007935, -0.1983431875705719, 0.0958956703543663, -0.024717260152101517, 0.5226476192474365, 0.26963475346565247, 0.32407721877098083, -0.22754086554050446, -0.1015940010547638, -0.14699961245059967, -0.13756293058395386, -0.2435147911310196, -0.04535192996263504, 0.21126894652843475, -0.07352497428655624, 0.16572386026382446, 0.024609185755252838, -0.49142882227897644, -0.20295217633247375, 0.2668694257736206, -0.3210103213787079, 0.17625252902507782, 0.06481709331274033, -0.3296066224575043, 0.45327839255332947, 0.1861165463924408, -0.3901230990886688, -0.18190498650074005, 0.1033877357840538, -0.1929485946893692, 0.09501943737268448, -0.1574057936668396, 0.1287296563386917, 0.024879178032279015, -0.1485254466533661, -0.30236488580703735, -0.03228052332997322, 0.3422373831272125, -0.11202170699834824, 0.15281781554222107, -0.24676938354969025, -0.1065731942653656, -0.2501501739025116, -0.06317604333162308, 0.4090137183666229, -0.19291439652442932, 0.08677593618631363, -0.01841961033642292, 0.14051774144172668, 0.25783857703208923, 0.2855260670185089, 0.016143959015607834, 0.121570885181427, 0.06346695125102997, 0.39460429549217224, 0.4315240681171417, -0.2790813744068146, 0.007539497222751379, 0.13285386562347412, -0.035576678812503815, -0.011855599470436573, 0.0022214094642549753, 0.10645253211259842, 0.03309090435504913, -0.22361288964748383, 0.10748088359832764, 0.19908709824085236, -0.06461162865161896, 0.2510938048362732, -0.4027029275894165, -0.06630891561508179, 0.4951448440551758, -0.20197105407714844, -0.1811574548482895, -0.054760370403528214, -0.05374516546726227, 0.15148337185382843, -0.014162302017211914, -0.02881745994091034, 0.3827754855155945, 0.32884690165519714, -0.14390325546264648, -0.1607179194688797, 0.1006360873579979, -0.10738272219896317, -0.46785417199134827, 0.01179454568773508, -0.18902328610420227, 0.20261067152023315, -0.08662525564432144, 0.02806076407432556, -0.34109851717948914, 0.12534897029399872, -0.10165315866470337, -0.22731059789657593, 0.14942365884780884, 0.05417387932538986, 0.16102658212184906, 0.14653590321540833, -0.12235655635595322, 0.1121622696518898, 0.013422824442386627, 0.08212617039680481, -0.46395185589790344, -0.08359185606241226, 0.000003371918637640192, -0.2866653800010681, -0.08776199817657471, 0.26526719331741333, 0.2704834043979645, 0.11229988187551498, -0.2972472608089447, 0.21523027122020721, 0.11378950625658035, 0.2788778841495514, -0.4021197557449341, 0.0717957392334938, 0.06320380419492722, -0.4297589361667633, 0.14214801788330078, 0.4353227913379669, 0.3686193823814392, -0.34248340129852295, -0.3888028860092163, -0.04157614707946777, 0.022348815575242043, 0.680326521396637, -0.13089554011821747, 0.14410725235939026, -0.032994844019412994, -0.03734279051423073, 0.07075369358062744, -0.05491745471954346, 0.16543714702129364, 0.017986003309488297, 0.12332218140363693, 0.309825599193573, -0.2486722618341446, 0.29118895530700684, 0.3783721625804901, 0.2824705243110657, 0.11098349094390869, 0.12216775864362717, -0.385292649269104, -0.2181706577539444, 0.3125985562801361, 0.025050396099686623, 0.3757638931274414, -0.05269104242324829, 0.18641406297683716, -0.05416793376207352, -0.18275831639766693, -0.17576918005943298, 0.35379335284233093, -0.058424804359674454, 0.19384722411632538, 0.036412570625543594, 0.5011975765228271, -0.015893854200839996, -0.13138151168823242, 0.24980774521827698, -0.00886713806539774, 0.20772431790828705, -0.41092759370803833, 0.33749786019325256, -0.40770524740219116, -0.13143028318881989, -0.3705584704875946, -0.1527719497680664, -0.024396412074565887, -0.6165550947189331, -0.04489489644765854, -0.018378552049398422, -0.030627377331256866, 0.12242858111858368, -0.2843605875968933, -0.13166244328022003, 0.013293451629579067, -0.07765188813209534, -0.40242794156074524, -0.42999574542045593, -0.4094289243221283, -0.0479501374065876, 0.4301172196865082, 0.3920787274837494, 0.10363438725471497, 0.032185740768909454, -0.10713537782430649, -0.036248333752155304, -0.4125104546546936, 0.03437483683228493, -0.2577618956565857, 0.09031236171722412, 0.1510431170463562, 0.060903966426849365, 0.01968560181558132, -0.0868513360619545, 0.32464176416397095, 0.0006714711198583245, -0.3420940637588501, 0.39525097608566284, 0.053129892796278, 0.14244547486305237, -0.3051871657371521, -0.46169474720954895, 0.039699822664260864, -0.4655473232269287, 0.1820293515920639, 0.006348194554448128, 0.18158914148807526, 0.23612697422504425, 0.05442686378955841, 0.2801811099052429, -0.04302286356687546, 0.16916759312152863, 0.19449324905872345, 0.47960251569747925, 0.5730380415916443, -0.14828312397003174, -0.36124661564826965, 0.267640620470047, -0.133779838681221, 0.29561087489128113, 0.17932794988155365, -0.15094290673732758, 0.3191586136817932, -0.3194130063056946, 0.1534697413444519, -0.13758273422718048, -0.12629768252372742, 0.2398141771554947, 0.18344947695732117, -0.04229184612631798, -0.10745669156312943, 0.03156435117125511, 0.2597667872905731, -0.18653303384780884, 0.1925482153892517, 0.11958298832178116, 0.2033814638853073, -0.04648163169622421, 0.31539806723594666, 0.18806947767734528, -0.25868120789527893, 0.33788979053497314, -0.035150863230228424, -0.008244960568845272, -0.2894054651260376, -0.10925953835248947, -0.08718881756067276, -0.14598098397254944, 0.15237286686897278, -0.0830240473151207, 0.07133299857378006, -0.45441362261772156, -0.028707435354590416, 0.2991608679294586, -0.10596832633018494, -0.14508503675460815, 0.2503579556941986, -0.27159374952316284, 0.23262329399585724, 0.07304379343986511, 0.1534188985824585, -0.11898382753133774, -0.5184164643287659, 0.04418317228555679, -0.03199419006705284, 0.09283922612667084, -0.020922038704156876, -0.20190294086933136, 0.017810314893722534, -0.4196454882621765, 0.2854287326335907, 0.1683174967765808, 0.4798279404640198, -0.04797327145934105, -0.19165821373462677, 0.17190392315387726, -0.1533501148223877, 0.38575035333633423, -0.2743549048900604, -0.013486948795616627, 0.36245667934417725, 0.36000770330429077, -0.48109689354896545, 0.26782628893852234, -0.25803759694099426, 0.2621379494667053, 0.17987315356731415, 0.19627141952514648, -0.4151630699634552, -0.3015875816345215, 0.14041142165660858, 0.1914491057395935, -0.02124146558344364, 0.16336624324321747, -0.2792511284351349, -0.023450039327144623, -0.06628598272800446, -0.17773990333080292, 0.26599323749542236, 0.3581661880016327, 0.03981303796172142, -0.0013969469582661986, -0.43300962448120117, 0.11331099271774292, 0.12204823642969131, 0.020940668880939484, 0.4554501175880432, -0.2969662547111511, 0.12399277091026306, -0.20001749694347382, 0.16423356533050537, -0.06789238750934601, 0.34366926550865173, -0.06636157631874084, -0.09572048485279083, 0.30548200011253357, -0.12876878678798676, 0.39053404331207275, 0.33944013714790344, -0.026474500074982643, 0.36377978324890137, -0.17648407816886902, -0.15365372598171234, -0.24443647265434265, -0.025315361097455025, 0.14097055792808533, -0.08740092068910599, 0.10841043293476105, -0.19618071615695953, 0.5411815643310547, 0.3007134199142456, -0.0316544771194458, 0.4439758360385895, 0.18964068591594696, -0.32779383659362793, 0.35079658031463623, 0.21915355324745178, 1.1128394603729248, 0.11890729516744614, 0.17537812888622284, 0.08425099402666092, -0.3473070561885834, 0.45130112767219543, 0.02242620661854744, 0.25419682264328003, -0.4858030080795288, 0.25739240646362305, -0.035548802465200424, -0.18759123980998993, 0.04030558094382286, 0.02434391900897026, -0.28639012575149536, 0.19320957362651825, -0.17377713322639465, 0.2628732919692993, 0.12903310358524323, 0.01991177164018154, -0.2860623598098755, -0.33544132113456726, -0.4839944541454315, 0.07514116168022156, -0.15866298973560333, -0.06433776766061783, 0.10971183329820633, -0.2614782750606537, -0.49332883954048157, -0.09190814197063446, -0.12715253233909607, 0.1436116248369217, -0.24136842787265778, -0.062289733439683914, 0.2582909166812897, -0.1797671914100647, 0.22817058861255646, 0.16256238520145416, -0.04886306822299957, -0.057419586926698685, -0.18875813484191895, -0.11977638304233551, 0.20588138699531555, 0.09073033928871155, 0.15809358656406403, -0.019870560616254807, 0.14279145002365112, 0.058030519634485245, 0.0911150723695755, 0.2642858624458313, -0.4077542722225189, -0.0016342384042218328, 0.0684535801410675, 0.16407984495162964, 0.035155024379491806, -0.021138392388820648, -0.1293342560529709, 0.1025448739528656, -0.01636497676372528, -0.034186892211437225, 0.0898297056555748, 0.04698120057582855, 0.08257462084293365, 0.05820891261100769, -0.10640508681535721, -0.22666841745376587, 0.26345938444137573, 0.3531939685344696, 0.05719923600554466, 0.14272744953632355, 0.7486616969108582, 0.16375748813152313, -0.13262835144996643, -0.19221454858779907, 0.08599633723497391, 0.018571488559246063, -0.7875836491584778, 0.28427737951278687, -0.2714296281337738, -0.32221052050590515, 0.03584720939397812, 0.6465659737586975, -0.23184798657894135, 0.08952020108699799, -0.1663869023323059, -0.2611517310142517, -0.1747758686542511, 0.17243972420692444, -0.03713057190179825, 0.06116108596324921, -0.5098087787628174, -0.1845780909061432, 0.10252305865287781, 0.13193513453006744, -0.2177727073431015, 0.0001780782185960561, -0.29474806785583496, 0.3535666763782501, -0.014866175130009651, 0.1016855239868164, 0.005318770185112953, -0.1716873049736023, -0.004655676893889904, 0.4534411132335663, -0.16379649937152863, -0.14825893938541412, -0.23846130073070526, 0.15354640781879425, 0.11453447490930557, -0.11689336597919464, 0.06979896873235703, -0.02155907452106476, 0.030570227652788162, -0.21219195425510406, -0.10736989229917526, -0.2103566825389862, -0.17519165575504303, -0.01351938210427761, 0.07630755752325058, 0.2623259127140045, -0.2717636227607727, 0.12272589653730392, -0.1723669022321701, 0.07838163524866104, -0.3097918927669525, 0.020727727562189102, 0.1177961528301239, 0.02143009565770626, -0.14642249047756195, 0.10597565025091171, -0.16352437436580658, -0.2105933576822281, 0.3242959678173065, -0.3493189513683319, -0.20741967856884003, 0.04359644651412964, -0.15914970636367798, 0.39598068594932556, -0.10549257695674896, 0.0026339604519307613, -0.07950553297996521, 0.08346440643072128, 0.05427969619631767, 0.3221624493598938, -0.19346070289611816, -0.11521819233894348, -0.04124798625707626, 0.2851088345050812, 0.1726255565881729, -0.21538755297660828, -0.026877455413341522, 0.11490360647439957, 0.028292326256632805, -0.4184589087963104, -0.08133628964424133, 0.18888023495674133, -0.04898955672979355, 0.01914878934621811, 0.26837459206581116, -0.19589760899543762, 0.028500894084572792, -0.07040543109178543, 0.4393863379955292, 0.41986116766929626, 0.4042418301105499, 0.1670338213443756, -0.024891043081879616, 0.08554995805025101, 0.05491339787840843, 0.35679569840431213, 0.34315094351768494, 0.2155420333147049, 0.25422748923301697, 0.5395472049713135, -0.176178440451622, -0.26875850558280945, -0.05656968429684639, 0.12042126804590225, -0.2214539498090744, -0.13927097618579865, -0.4548972547054291, -0.15396654605865479, -0.48003995418548584, -0.2001102715730667, -0.16493985056877136, 0.138153538107872, 0.5681527853012085, 0.2202722281217575, 0.04264845326542854, -0.5250182151794434, -0.13786688446998596, 0.10644254088401794, 0.3856298327445984, -0.17447851598262787, 0.4604707956314087, 0.006539936177432537, -0.0706382542848587, -0.030537068843841553, 0.42286747694015503, 0.42817622423171997, 0.5327765345573425, -0.3272398114204407, -0.06885863840579987, 0.12607939541339874, -0.1641942709684372, 0.24071301519870758, 0.5911162495613098, 0.2477376013994217, 0.06603623181581497, 0.06582177430391312, -0.05595839023590088, -0.08274543285369873, -0.1022566705942154, -0.022089505568146706, 0.3411242961883545, -0.12443612515926361, 0.4214778244495392, -0.2777312397956848, -0.14602144062519073, -0.10044360905885696, -0.18616950511932373, -0.28644582629203796, 0.03204294294118881, 0.10367290675640106, 0.03099711425602436, 0.07790996134281158, -0.32750871777534485, 0.06784393638372421, 0.1837603896856308, 0.5826795101165771, 0.22417347133159637, 0.10328651964664459, -0.3924230933189392, -0.09169911593198776, -0.7618373036384583, 0.17532828450202942, -0.2416837066411972, -0.21087594330310822, 0.10276062041521072, 0.3020721971988678, 0.08311298489570618, 0.3048584461212158, -0.04291004687547684, 0.27844706177711487, 0.2520662546157837, 0.28670451045036316, -0.38265180587768555, -0.19707739353179932, 0.34452691674232483, -0.2810957431793213, -0.05288337543606758, -0.5603268146514893, 0.23074860870838165, 0.016348782926797867, 0.0587674155831337, 0.03245599567890167, -0.2147497981786728, 0.1856909543275833, 0.44472208619117737, 0.20759901404380798, 0.1835721731185913, 0.17585769295692444, 0.27285701036453247, -0.43202731013298035, -0.35022422671318054, -0.2179291844367981, -0.19331373274326324, 0.03420933708548546, 0.03560377657413483, 0.4834500551223755, -0.16845636069774628, 0.03575209528207779, -0.2757134437561035, 0.12094942480325699, -0.13554754853248596, 0.1336909681558609, -0.2686522305011749, 0.18321827054023743, -0.41397568583488464, 0.19967925548553467, 0.22928152978420258, 0.02240593172609806, 0.12845195829868317, 0.14924877882003784, -0.42977893352508545, -0.4989635944366455, 0.25693821907043457, -0.4704398810863495, -0.32642415165901184, 0.2818504869937897, 0.14563821256160736, 0.3511768877506256, -0.060560330748558044, -0.7402788400650024, -0.25568169355392456, 0.23546068370342255, -0.12984555959701538, 0.017962539568543434, -0.04342928156256676, -0.04658237844705582, 0.026109736412763596, -0.23466292023658752, 0.23958736658096313, 0.0006647274713031948, 0.14029885828495026, 0.03313111513853073, -0.2577144503593445 ]
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.
[ -0.06821651011705399, 0.060931235551834106, 0.10538241267204285, 0.4737013280391693, 0.25208011269569397, 0.10438024252653122, 0.09826845675706863, 0.38123026490211487, 0.22328907251358032, -0.06250706315040588, 0.32705163955688477, 0.47998321056365967, -0.08507649600505829, -0.022185614332556725, -0.045739706605672836, -0.13106970489025116, 0.16129451990127563, 0.3236936032772064, 0.06354113668203354, -0.12184242159128189, -0.14383503794670105, 0.08432061225175858, -0.23519456386566162, 0.07620399445295334, -0.5430988073348999, -0.16950692236423492, 0.05933595448732376, 0.10318558663129807, -0.1546952873468399, -0.4448225498199463, 0.014604197815060616, -0.25054529309272766, 0.1371709406375885, 0.31009212136268616, -0.00011774220911320299, -0.1311674267053604, 0.3823639452457428, -0.04525528848171234, -0.21709232032299042, -0.15531226992607117, -0.4626956582069397, 0.06081005558371544, 0.26330968737602234, -0.06844905018806458, 0.1900985985994339, -0.0022966235410422087, 0.06410683691501617, -0.4999205768108368, 0.10783106833696365, 0.5157181620597839, 0.20820841193199158, 0.2796901762485504, -0.01879955641925335, 0.0332891047000885, 0.3279874920845032, 0.021057480946183205, -0.3656476140022278, 0.0005636226269416511, -0.028315642848610878, -0.15549589693546295, -0.1619083136320114, 0.3987066447734833, -0.08816149830818176, 0.05234837904572487, 0.6084596514701843, 0.0969526618719101, 0.11834550648927689, -0.2912597060203552, 0.19596001505851746, 0.12539060413837433, 0.4451828896999359, -0.10498583316802979, -0.11374666541814804, -0.003936172462999821, -0.1345483362674713, -0.3904246687889099, 0.23909559845924377, 0.1906234174966812, -0.25099965929985046, -0.0031805154867470264, 0.059596627950668335, 0.10651859641075134, -0.03581863269209862, -0.19298145174980164, -0.06624875217676163, 0.20389717817306519, 0.06159088388085365, 0.04267077520489693, -0.10467105358839035, -0.15812553465366364, -0.01255879271775484, -0.23138529062271118, -0.3124718964099884, 0.25019732117652893, -0.3363651931285858, 0.04203135520219803, 0.009231333620846272, -0.10786846280097961, -0.1433403640985489, -0.07465703040361404, 0.3727550506591797, 0.023687461391091347, 0.13817396759986877, 0.11149316281080246, 0.3602554202079773, 0.38036850094795227, 0.3734179437160492, 0.13893958926200867, 0.172773540019989, -0.2554154396057129, -0.20996040105819702, -0.018515704199671745, -0.02588639222085476, -0.026921173557639122, 0.48337626457214355, -0.018195074051618576, 0.1269228309392929, -0.342743843793869, -0.28641536831855774, 0.03796976059675217, -0.6072552800178528, 0.1798171103000641, -0.21649327874183655, 0.18091219663619995, 0.014957117848098278, 0.4404563903808594, 0.14114995300769806, 0.3197542428970337, -0.28661805391311646, -0.21293583512306213, -0.1690678894519806, -0.06719332188367844, -0.35299694538116455, -0.0474066324532032, 0.23163427412509918, 0.02534494362771511, 0.17099426686763763, 0.045751433819532394, -0.5268830060958862, -0.19393067061901093, 0.2752549946308136, -0.33649247884750366, 0.10100140422582626, 0.09800000488758087, -0.18175570666790009, 0.3801155984401703, 0.15239644050598145, -0.3950260877609253, -0.20507796108722687, 0.07577163726091385, -0.14228419959545135, -0.014657932333648205, -0.14263609051704407, 0.16258682310581207, 0.019487814977765083, -0.22843672335147858, -0.28548717498779297, -0.04417162388563156, 0.376835435628891, -0.08247692883014679, 0.033103302121162415, -0.1942519098520279, -0.07683183252811432, -0.2546739876270294, -0.14971092343330383, 0.4914356470108032, -0.17050638794898987, 0.16371464729309082, 0.05206769332289696, 0.15883421897888184, 0.32009196281433105, 0.3464753329753876, -0.05128078907728195, 0.2360602617263794, 0.07445210218429565, 0.32859447598457336, 0.47833189368247986, -0.27670979499816895, 0.054584894329309464, 0.14386872947216034, -0.04974477365612984, -0.014119517058134079, 0.055780552327632904, 0.004746762104332447, 0.1131347268819809, -0.18608130514621735, 0.1133463904261589, 0.17492209374904633, -0.1087213009595871, 0.299288809299469, -0.44598719477653503, -0.04967838525772095, 0.3986052870750427, -0.20753437280654907, -0.08162318915128708, 0.05908750370144844, 0.01001072395592928, 0.055392373353242874, 0.017926089465618134, -0.08591856062412262, 0.34544801712036133, 0.29823583364486694, -0.12419266253709793, -0.10706944018602371, 0.08703386783599854, -0.09944121539592743, -0.4447137415409088, 0.03243204951286316, -0.15776477754116058, 0.12304917722940445, 0.05034507438540459, -0.09754763543605804, -0.2944745123386383, 0.09615052491426468, -0.03822717443108559, -0.2505197525024414, 0.15453094244003296, 0.02319597825407982, 0.17318513989448547, 0.05925073102116585, -0.20210517942905426, 0.06613495200872421, -0.015744449570775032, 0.10741601884365082, -0.3512533903121948, -0.17887359857559204, -0.044257745146751404, -0.35093259811401367, -0.03988926485180855, 0.23988282680511475, 0.23701824247837067, 0.10021083056926727, -0.2585672438144684, 0.23891966044902802, -0.0016572182066738605, 0.3618730306625366, -0.24295112490653992, 0.08757434040307999, -0.04195784777402878, -0.2605339586734772, 0.1956302970647812, 0.27519071102142334, 0.3589315712451935, -0.29758134484291077, -0.4221459925174713, -0.006539307534694672, 0.037945061922073364, 0.5811726450920105, -0.04470035433769226, 0.1357862800359726, 0.033681079745292664, -0.07959746569395065, 0.13727323710918427, -0.1051982119679451, 0.17800746858119965, -0.07357250899076462, 0.06387796252965927, 0.33713626861572266, -0.2674141228199005, 0.31074610352516174, 0.42854467034339905, 0.29901769757270813, 0.11164862662553787, 0.0571209155023098, -0.29339706897735596, -0.13686692714691162, 0.3519337475299835, 0.13247975707054138, 0.4555586874485016, -0.015705527737736702, 0.20980583131313324, -0.07270660996437073, -0.16493405401706696, -0.2009923756122589, 0.31755512952804565, -0.14931774139404297, 0.19674576818943024, 0.06621259450912476, 0.4594135284423828, 0.020023226737976074, -0.21954357624053955, 0.21167488396167755, 0.03651801869273186, 0.21878306567668915, -0.37831518054008484, 0.2838200032711029, -0.285977840423584, -0.07724177092313766, -0.43136993050575256, -0.04244397208094597, -0.046468090265989304, -0.5650829672813416, -0.03836643695831299, 0.007924437522888184, -0.09303826093673706, 0.03654523193836212, -0.3283858597278595, -0.1192564070224762, -0.12170125544071198, -0.0653657540678978, -0.2936853766441345, -0.4041464030742645, -0.3502984344959259, -0.05848417431116104, 0.42138680815696716, 0.3856320381164551, 0.19977761805057526, 0.01642804592847824, -0.0054526785388588905, -0.028151454403996468, -0.4237631559371948, 0.04701807722449303, -0.22188527882099152, 0.04329710081219673, 0.16059277951717377, -0.005856356117874384, 0.05179909989237785, -0.10299594700336456, 0.3697010576725006, -0.051131151616573334, -0.2518928349018097, 0.35217660665512085, 0.022314956411719322, -0.03763391450047493, -0.2629382014274597, -0.40583866834640503, 0.11667193472385406, -0.5667975544929504, 0.1007881909608841, -0.04693639650940895, 0.2017465978860855, 0.17167221009731293, 0.08207520842552185, 0.22550222277641296, -0.06525587290525436, 0.1951523870229721, 0.28039154410362244, 0.3783937394618988, 0.5403062105178833, -0.2259807139635086, -0.3230758011341095, 0.29508620500564575, -0.1108960434794426, 0.36081844568252563, 0.23071269690990448, -0.1520203948020935, 0.405286967754364, -0.410354346036911, 0.1700480878353119, -0.08374297618865967, -0.1172337457537651, 0.3123687505722046, 0.1904551386833191, -0.06837334483861923, -0.09682267904281616, 0.01799457147717476, 0.21361666917800903, -0.20148392021656036, 0.1650697886943817, 0.008490183390676975, 0.19980520009994507, -0.10333201289176941, 0.3400726318359375, 0.12449978291988373, -0.16565190255641937, 0.3614375591278076, 0.01160025130957365, 0.012079576030373573, -0.25102949142456055, -0.14403469860553741, 0.01997211202979088, -0.21486946940422058, 0.061621829867362976, -0.04792575538158417, 0.05654408410191536, -0.4713309407234192, -0.03464672714471817, 0.23681098222732544, -0.14764033257961273, -0.16390962898731232, 0.3225649893283844, -0.33875900506973267, 0.2803708612918854, 0.09475945681333542, 0.1365749090909958, -0.11946546286344528, -0.4064585864543915, 0.023601841181516647, -0.05851197615265846, 0.10876665264368057, -0.016414403915405273, -0.26465657353401184, 0.002610167022794485, -0.534706175327301, 0.2495766431093216, 0.05217333137989044, 0.3960875868797302, -0.0433574840426445, -0.18558372557163239, 0.18215233087539673, -0.22928613424301147, 0.4075532555580139, -0.31268739700317383, 0.018774937838315964, 0.28596094250679016, 0.2986103892326355, -0.49828481674194336, 0.1848800778388977, -0.22405634820461273, 0.3302094042301178, 0.2257489264011383, 0.18554361164569855, -0.3720528185367584, -0.3285734951496124, 0.22981752455234528, 0.04473905637860298, -0.000650879112072289, 0.10367853939533234, -0.35246753692626953, -0.05329153686761856, -0.06974153965711594, -0.14215591549873352, 0.2665809392929077, 0.3564814031124115, -0.07092609256505966, 0.007950111292302608, -0.43612492084503174, 0.15831901133060455, 0.15794895589351654, 0.13093756139278412, 0.38802915811538696, -0.33989474177360535, 0.12485639750957489, -0.16583503782749176, 0.19165845215320587, -0.045501548796892166, 0.42883315682411194, -0.2038114368915558, -0.19006197154521942, 0.3096027672290802, -0.17455171048641205, 0.48622918128967285, 0.37902355194091797, 0.04215037077665329, 0.3114558458328247, -0.2637414038181305, -0.1782965213060379, -0.2773718535900116, -0.04351379722356796, 0.13077564537525177, -0.0889105275273323, 0.09123115986585617, -0.23243598639965057, 0.5531398057937622, 0.30162325501441956, -0.055548295378685, 0.4027084708213806, 0.18881963193416595, -0.3564831018447876, 0.4658268094062805, 0.25320449471473694, 1.1186978816986084, 0.10575845837593079, 0.22237147390842438, 0.02540394477546215, -0.40893781185150146, 0.3530542850494385, -0.03306620568037033, 0.27251744270324707, -0.4955560266971588, 0.22722315788269043, 0.027789417654275894, -0.12293659895658493, 0.07964405417442322, 0.05283242091536522, -0.2791074216365814, 0.15774770081043243, -0.1973516196012497, 0.33613017201423645, 0.07588654011487961, 0.12650950253009796, -0.4026108384132385, -0.34524786472320557, -0.48598966002464294, 0.08268748223781586, -0.08391553908586502, -0.08008456975221634, 0.026806309819221497, -0.22409038245677948, -0.40694746375083923, -0.1369733214378357, -0.13860897719860077, 0.13444143533706665, -0.15800277888774872, -0.1193525642156601, 0.25688982009887695, -0.15448132157325745, 0.19842121005058289, 0.2281161993741989, 0.04491613060235977, -0.05061188340187073, -0.16760894656181335, -0.014997262507677078, 0.27013784646987915, 0.11929093301296234, 0.1429566740989685, -0.027162913233041763, 0.07144784927368164, 0.04230790212750435, 0.08588588237762451, 0.23091308772563934, -0.41005703806877136, 0.011083151213824749, -0.12113738805055618, 0.20614750683307648, -0.02342264913022518, -0.13388663530349731, -0.08782537281513214, 0.11561545729637146, -0.034164126962423325, -0.08314475417137146, 0.1168229803442955, 0.02967073954641819, 0.11227257549762726, 0.09994053095579147, -0.15056633949279785, -0.2468956559896469, 0.22061564028263092, 0.31885525584220886, 0.056490276008844376, 0.23681806027889252, 0.6745471358299255, 0.08035559952259064, -0.1407402604818344, -0.23257215321063995, 0.04807344824075699, 0.05120323970913887, -0.7669646739959717, 0.33888334035873413, -0.2679663896560669, -0.4024944603443146, 0.13678407669067383, 0.5917763710021973, -0.1855551302433014, 0.013741334900259972, -0.09446405619382858, -0.2665630280971527, -0.09995357692241669, 0.19291740655899048, 0.0318586640059948, 0.07356487214565277, -0.4767533838748932, -0.1263430267572403, 0.07701265811920166, 0.22448205947875977, -0.26808515191078186, -0.04891302064061165, -0.23274105787277222, 0.37962883710861206, 0.014470970258116722, 0.0802760198712349, -0.029821308329701424, -0.15083567798137665, -0.006759953685104847, 0.4313046336174011, -0.14375345408916473, -0.15116329491138458, -0.27181383967399597, 0.1370278298854828, 0.07310190051794052, -0.02484004758298397, 0.15791858732700348, 0.029813511297106743, -0.012405457906425, -0.25564178824424744, -0.0019757957197725773, -0.18603631854057312, -0.22291265428066254, -0.0284400787204504, 0.058521632105112076, 0.2825692296028137, -0.20867152512073517, 0.12727829813957214, -0.18776819109916687, 0.0798763632774353, -0.33941376209259033, 0.07165033370256424, 0.07958443462848663, 0.015731187537312508, 0.03195734694600105, 0.09080924093723297, -0.14252522587776184, -0.07515716552734375, 0.3140859603881836, -0.34542208909988403, -0.10453874617815018, 0.11357157677412033, -0.11710676550865173, 0.3987208306789398, -0.08429915457963943, 0.014218470081686974, -0.026213787496089935, 0.13289308547973633, 0.07714062929153442, 0.257689893245697, -0.27105841040611267, -0.12639236450195312, -0.05107763409614563, 0.30452221632003784, 0.17767742276191711, -0.21350857615470886, -0.040585167706012726, 0.12846463918685913, 0.054264314472675323, -0.3916771411895752, -0.14858053624629974, 0.08976926654577255, -0.09947679191827774, 0.007058710791170597, 0.33671271800994873, -0.11197768896818161, 0.02153904177248478, 0.08279450237751007, 0.4989662766456604, 0.45822039246559143, 0.39912158250808716, 0.16864845156669617, -0.11093097925186157, -0.0017288238741457462, 0.021545039489865303, 0.3171040117740631, 0.3153465688228607, 0.3045440912246704, 0.15488997101783752, 0.5104096531867981, -0.30581825971603394, -0.2672061622142792, -0.0920049175620079, 0.06214560195803642, -0.25436434149742126, -0.14574913680553436, -0.5064579844474792, -0.14190496504306793, -0.42316171526908875, -0.10091479867696762, -0.1865212470293045, 0.26608893275260925, 0.5050715208053589, 0.1783142238855362, 0.002986324718222022, -0.5828506946563721, -0.05613260343670845, 0.031109027564525604, 0.39658311009407043, -0.24684059619903564, 0.5015382170677185, -0.03485250100493431, -0.09850386530160904, -0.09822675585746765, 0.5166681408882141, 0.4602740705013275, 0.658098042011261, -0.2757399082183838, -0.070160411298275, 0.0755319818854332, -0.13385120034217834, 0.15214860439300537, 0.560676097869873, 0.3276127576828003, 0.15973079204559326, 0.11919432878494263, -0.034418776631355286, -0.09463390707969666, -0.11874833703041077, -0.034373458474874496, 0.2678051292896271, -0.09244740009307861, 0.3559325039386749, -0.1539553999900818, -0.24337144196033478, -0.08671656996011734, -0.15515653789043427, -0.2883928120136261, -0.05635567009449005, 0.13199473917484283, -0.07147876173257828, 0.05464475229382515, -0.42725205421447754, 0.07915198057889938, 0.20936833322048187, 0.6322187781333923, 0.2378581017255783, 0.16091477870941162, -0.45841163396835327, -0.07986689358949661, -0.7588371634483337, 0.14081904292106628, -0.39341649413108826, -0.16126547753810883, 0.026848318055272102, 0.35101792216300964, -0.014347471296787262, 0.22938159108161926, -0.05591113492846489, 0.3212624192237854, 0.1782602220773697, 0.29825863242149353, -0.34541454911231995, -0.17916624248027802, 0.22054901719093323, -0.3443707227706909, -0.049058493226766586, -0.6017603278160095, 0.3270030617713928, 0.05008517950773239, 0.08173490315675735, -0.007283016573637724, -0.2222328931093216, 0.3362942934036255, 0.2882283926010132, 0.32577580213546753, 0.16702581942081451, 0.28842389583587646, 0.1926189512014389, -0.45870649814605713, -0.3980747163295746, -0.19980326294898987, -0.3066632151603699, -0.09984301030635834, 0.1319093108177185, 0.4690837562084198, -0.1796381175518036, 0.11805947870016098, -0.2406577616930008, 0.25994157791137695, -0.10598054528236389, 0.11541399359703064, -0.1505337357521057, 0.13514044880867004, -0.323625385761261, 0.11830691248178482, 0.26126542687416077, 0.16057124733924866, 0.13314545154571533, 0.12775316834449768, -0.4737323522567749, -0.4435844123363495, 0.2808891534805298, -0.38236764073371887, -0.3833833634853363, 0.25261783599853516, 0.22172538936138153, 0.3978998064994812, -0.09128411114215851, -0.7310699820518494, -0.30311787128448486, 0.1360621601343155, -0.08944404870271683, 0.004416808020323515, 0.06323639303445816, 0.07703431695699692, -0.07732460647821426, -0.21661363542079926, 0.28396862745285034, -0.055230043828487396, 0.02648361586034298, 0.054151035845279694, -0.2534834146499634 ]
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.
[ -0.13736297190189362, 0.17048197984695435, -0.1900716871023178, -0.2922055125236511, -0.17911440134048462, 0.044915471225976944, 0.2525579035282135, 0.5567149519920349, 0.055268917232751846, 0.0512448288500309, 0.014377288520336151, 0.17538641393184662, -0.3614473044872284, 0.08514589816331863, 0.07604798674583435, 0.1644962579011917, -0.3026101589202881, 0.5070188641548157, -0.13109830021858215, 0.12752577662467957, -0.2230493724346161, -0.20796161890029907, 0.00013458497414831072, 0.2439568042755127, 0.18622957170009613, -0.3947817385196686, 0.18797403573989868, -0.16565586626529694, -0.019231600686907768, -0.5875657796859741, -0.10718852281570435, 0.2003774493932724, -0.17094147205352783, -0.11505145579576492, -0.00009701006638351828, 0.0429435595870018, 0.16814802587032318, 0.12773190438747406, -0.2890503704547882, 0.10696441680192947, -0.1203126460313797, -0.05883170664310455, 0.27865785360336304, -0.13759110867977142, -0.08296463638544083, -0.22412897646427155, 0.14215491712093353, -0.22430932521820068, 0.10305406898260117, 0.35247331857681274, 0.3387940227985382, 0.35248643159866333, -0.308523565530777, 0.43407300114631653, 0.13730274140834808, -0.011704571545124054, -0.2752702832221985, 0.045189306139945984, 0.03301142156124115, 0.015701469033956528, -0.3985559344291687, 0.5405020117759705, -0.0640978217124939, -0.007953471504151821, -0.03828674927353859, -0.17734363675117493, 0.3933626711368561, -0.2694503366947174, -0.30618658661842346, 0.08690249919891357, 0.19845010340213776, -0.19477102160453796, -0.06235066056251526, -0.4192016124725342, 0.2383095920085907, 0.04672978073358536, 0.09231428802013397, 0.2671627104282379, -0.0586763359606266, 0.23186305165290833, -0.5527653694152832, -0.4114423394203186, -0.1109345555305481, 0.08389560878276825, 0.14198285341262817, -0.10454478859901428, 0.08363649249076843, -0.05657774582505226, 0.35432320833206177, -0.0869220718741417, 0.17274431884288788, -0.05581096559762955, -0.25806018710136414, 0.014445565640926361, -0.005055077373981476, -0.5017728209495544, 0.09996507316827774, -0.019057385623455048, -0.03359735384583473, 0.20497840642929077, 0.0802922323346138, 0.10392984002828598, -0.20463944971561432, 0.06354304403066635, 0.10912398248910904, 0.12340106815099716, 0.3510042428970337, 0.06238657981157303, 0.25975513458251953, 0.24740131199359894, 0.24320939183235168, 0.30436405539512634, 0.33762887120246887, -0.25187283754348755, -0.2752167582511902, 0.1266617327928543, 0.3540289103984833, -0.23466165363788605, -0.3460923433303833, 0.0034459985326975584, -0.5296183228492737, 0.014235052280128002, 0.01448008045554161, 0.10928115248680115, -0.0006172169814817607, 0.12058677524328232, 0.03531215712428093, 0.031738199293613434, -0.36195656657218933, -0.3578192889690399, -0.05442376807332039, 0.10798398405313492, -0.09068229794502258, 0.11934176832437515, 0.4057036340236664, 0.04340824857354164, 0.16046161949634552, -0.034201961010694504, 0.09332343935966492, 0.04058917239308357, 0.36983054876327515, 0.024624431505799294, 0.3463374674320221, -0.05756774917244911, -0.11738023906946182, -0.21008704602718353, -0.21587081253528595, -0.14201417565345764, -0.18773338198661804, 0.15288546681404114, 0.09354312717914581, -0.20848335325717926, 0.0526687316596508, 0.3505091071128845, -0.4395667016506195, -0.014217481017112732, 0.09178129583597183, 0.23471510410308838, -0.3213297426700592, -0.04506177827715874, 0.14012733101844788, -0.17458783090114594, -0.13457602262496948, 0.07385961711406708, -0.17334522306919098, 0.29054296016693115, -0.09705126285552979, -0.08341518044471741, -0.039976418018341064, -0.09570503234863281, -0.20757608115673065, -0.024172095581889153, -0.40049317479133606, 0.17513073980808258, -0.20396797358989716, -0.06718975305557251, 0.36196744441986084, -0.21314741671085358, -0.10236690193414688, 0.20034706592559814, -0.23160777986049652, 0.0169601421803236, -0.14347738027572632, 0.07153689861297607, -0.16221855580806732, -0.4502266049385071, -0.4792475700378418, 0.07163723558187485, 0.0601608045399189, 0.14337167143821716, -0.4493595063686371, -0.42560940980911255, 0.4917641878128052, 0.04578477516770363, -0.053064074367284775, -0.1359761655330658, -0.1581975519657135, 0.2587839365005493, 0.437486857175827, 0.028907854110002518, -0.039323072880506516, -0.23834072053432465, 0.20831507444381714, -0.32658976316452026, -0.2815675437450409, 0.007924179546535015, -0.13267382979393005, -0.1672874540090561, -0.08347128331661224, 0.18698111176490784, 0.420723021030426, -0.2829241454601288, -0.09006677567958832, -0.09678632020950317, -0.1517232209444046, -0.09776152670383453, 0.290005087852478, -0.14136971533298492, -0.09831613302230835, 0.13745447993278503, 0.09653123468160629, -0.1579168736934662, -0.17216789722442627, 0.03815498948097229, 0.020719952881336212, -0.17654582858085632, -0.053054291754961014, 0.10111165046691895, 0.31183362007141113, 0.3026478588581085, 0.07121890783309937, 0.06072904169559479, 0.07649913430213928, 0.17240801453590393, 0.06634180247783661, 0.10490567237138748, 0.23246634006500244, 0.27824175357818604, 0.07447274029254913, -0.44166073203086853, -0.014351587742567062, 0.13553385436534882, -0.15129199624061584, 0.2366538941860199, -0.3167857825756073, 0.2563724219799042, 0.16144393384456635, -0.07265864312648773, 0.003107847413048148, -0.029504047706723213, 0.04893167316913605, -0.33980318903923035, -0.13817612826824188, -0.33787328004837036, -0.026473134756088257, 0.22689393162727356, -0.1862308531999588, 0.2369191199541092, -0.3378424644470215, 0.171169251203537, 0.09643464535474777, -0.015973368659615517, -0.026086365804076195, 0.10548366606235504, 0.20493516325950623, -0.3523615002632141, 0.23605811595916748, 0.07625589519739151, 0.10919330269098282, 0.34950166940689087, 0.22472357749938965, 0.17186981439590454, -0.2100955992937088, -0.27264267206192017, 0.3151496648788452, 0.3302735686302185, 0.22651365399360657, -0.02484247460961342, 0.24631305038928986, -0.16664527356624603, -0.242401123046875, 0.07095299661159515, -0.10728851705789566, -0.05060545727610588, -0.37377095222473145, 0.024834854528307915, -0.1493522822856903, -0.5540045499801636, -0.16929616034030914, 0.1952504962682724, -0.09234797209501266, -0.2506900131702423, 0.5380776524543762, -0.028283774852752686, -0.3674882650375366, 0.2202184945344925, -0.29027265310287476, 0.6584696769714355, -0.20973868668079376, 0.223077192902565, -0.06582286953926086, -0.020816506817936897, -0.36914491653442383, 0.30891114473342896, 0.1798364520072937, 0.06083424761891365, 0.5227046012878418, 0.10138306766748428, -0.17568445205688477, -0.1762026995420456, -0.27201029658317566, 0.086508609354496, -0.03296935185790062, 0.3253641724586487, -0.0012541713658720255, -0.1657099723815918, -0.047609779983758926, -0.1483747363090515, -0.418618768453598, -0.22404445707798004, -0.00045919971307739615, -0.15232470631599426, 0.14578013122081757, 0.3848955035209656, -0.24102433025836945, -0.46456456184387207, 0.0017173383384943008, -0.21992343664169312, 0.02016071230173111, -0.014971098862588406, 0.20743225514888763, 0.20418497920036316, -0.26563626527786255, 0.005003367550671101, -0.14018665254116058, -0.05355347320437431, -0.1193709522485733, 0.04450157284736633, 0.012334800325334072, -0.27020201086997986, -0.28076639771461487, 0.01924579218029976, -0.25710126757621765, 0.14636051654815674, -0.17370179295539856, -0.6786444187164307, -0.3268857002258301, 0.18293775618076324, -0.09719103574752808, -0.04581698402762413, -0.5455359220504761, -0.11407434195280075, 0.17513559758663177, -0.2621878683567047, 0.12555119395256042, 0.08653374016284943, 0.2562340199947357, 0.005403110757470131, 0.14536041021347046, -0.10717713087797165, -0.10614831000566483, 0.2679671347141266, 0.15631695091724396, 0.27477389574050903, 0.11752208322286606, -0.007231110241264105, 0.0892392173409462, 0.260608434677124, -0.008756261318922043, -0.03511358052492142, 0.046268850564956665, 0.12895244359970093, 0.015699267387390137, 0.33214297890663147, -0.11922284215688705, -0.313361257314682, 0.051886215806007385, 0.08164501190185547, -0.21345089375972748, -0.40376460552215576, 0.14978055655956268, -0.40173637866973877, 0.13901296257972717, 0.12182221561670303, 0.04311315342783928, 0.06124322488903999, -0.5748979449272156, 0.05878445878624916, -0.1971299797296524, 0.22688144445419312, -0.06648250669240952, -0.20710505545139313, -0.013696709647774696, -0.275026798248291, -0.011230573058128357, 0.146232008934021, 0.38458818197250366, 0.01064344309270382, -0.23364901542663574, 0.24429190158843994, 0.16834433376789093, 0.2181382030248642, 0.061730675399303436, 0.20189152657985687, 0.13025836646556854, -0.12362964451313019, -0.30144640803337097, 0.18793031573295593, -0.21928931772708893, -0.09623381495475769, 0.15762071311473846, -0.02711084671318531, 0.029719773679971695, -0.24706484377384186, 0.36989444494247437, 0.1215464323759079, -0.23160703480243683, 0.024665988981723785, -0.013284539803862572, 0.01062963716685772, -0.03718185052275658, 0.2463870495557785, -0.1701395958662033, -0.02812214381992817, -0.36340102553367615, -0.1015850380063057, -0.024530205875635147, -0.2028278410434723, 0.3764396011829376, 0.2882792055606842, -0.01591498777270317, 0.2726738750934601, 0.09733471274375916, 0.2122035026550293, 0.03915707394480705, -0.19112013280391693, 0.18984784185886383, 0.247857928276062, 0.3424505591392517, -0.044840678572654724, -0.20925544202327728, 0.34741050004959106, 0.4122864603996277, 0.13903407752513885, 0.0013978914357721806, -0.38751810789108276, 0.05723714828491211, -0.2808181941509247, 0.29006293416023254, 0.07467541843652725, -0.1746477633714676, 0.05383219197392464, -0.12565743923187256, 0.24363867938518524, -0.12188348919153214, -0.36395159363746643, 0.0006385644664987922, 0.11595039814710617, -0.18264448642730713, 0.20929931104183197, 0.4568057656288147, 0.7854485511779785, 0.045313067734241486, 0.19970177114009857, 0.472280353307724, -0.15780223906040192, 0.4363836944103241, -0.15434885025024414, 0.25087302923202515, -0.13412980735301971, 0.09534630924463272, 0.059532128274440765, -0.11783861368894577, 0.11199357360601425, 0.2078060358762741, 0.0020548503380268812, 0.0006366065936163068, 0.26226547360420227, -0.3583504259586334, -0.11938659846782684, 0.3215346932411194, 0.04634028673171997, -0.021235475316643715, -0.3037979006767273, 0.2726385295391083, -0.1611575335264206, -0.2050417959690094, 0.1078731119632721, 0.02808612398803234, 0.3095526397228241, 0.045405615121126175, 0.18522441387176514, 0.13581502437591553, 0.112190380692482, 0.13306348025798798, -0.13726162910461426, -0.18840207159519196, 0.3965897560119629, 0.23549318313598633, -0.17092253267765045, 0.1243637427687645, -0.0892934575676918, -0.012548009864985943, -0.1636732518672943, 0.2965010106563568, 0.21388816833496094, 0.2833039462566376, 0.3150736391544342, -0.19931215047836304, -0.1702398806810379, 0.2649152874946594, 0.1970161348581314, 0.059724144637584686, -0.08262801915407181, -0.005859323311597109, 0.214253231883049, -0.47217699885368347, 0.03237346559762955, 0.03920355066657066, 0.11395160853862762, -0.15051263570785522, 0.28724613785743713, 0.04401286318898201, -0.03829522803425789, 0.0432286411523819, 0.3496764302253723, -0.23045490682125092, 0.10065100342035294, 0.05798111855983734, 0.09893618524074554, 0.1047602966427803, 0.4397663176059723, -0.1681392341852188, -0.24870672821998596, -0.4134778380393982, 0.25973162055015564, -0.12139621376991272, -0.2759031355381012, 0.10330599546432495, 0.12412790954113007, -0.05992041528224945, -0.03658779710531235, 0.21929137408733368, -0.10416144132614136, -0.006942478008568287, -0.09856211394071579, 0.004728677216917276, 0.002826184267178178, 0.16214455664157867, -0.1353158801794052, 0.26489484310150146, 0.23671166598796844, 0.17586541175842285, -0.07974199950695038, 0.16821017861366272, -0.45865800976753235, -0.2357047200202942, -0.3629606068134308, 0.16979216039180756, -0.3387680947780609, 0.16333022713661194, 0.025976575911045074, 0.0013711937936022878, 0.13112430274486542, -0.2586055397987366, -0.2390025109052658, -0.3924419581890106, -0.11555560678243637, 0.05777055025100708, -0.014545410871505737, -0.17148569226264954, 0.02073679119348526, 0.011869387701153755, 0.10929159075021744, 0.056947898119688034, -0.1109563335776329, -0.36126941442489624, 0.18716450035572052, 0.6549921035766602, -0.22639799118041992, -0.3768439292907715, 0.04805300384759903, 0.002164491219446063, -0.33640214800834656, 0.1804552525281906, 0.19234508275985718, 0.25698769092559814, 0.0788685604929924, -0.0773615837097168, 0.09211746603250504, 0.14207783341407776, -0.0038993745110929012, 0.24301700294017792, -0.2207941859960556, 0.056132614612579346, 0.016851361840963364, -0.029933124780654907, -0.025893379002809525, -0.02797340229153633, -0.30157986283302307, -0.29457831382751465, 0.29048869013786316, 0.3381894528865814, -0.3103622794151306, 0.027729274705052376, 0.32519039511680603, 0.04844710975885391, 0.07958588749170303, 0.21352910995483398, 0.08040473610162735, 0.02968490496277809, 0.29514914751052856, 0.030976353213191032, -0.00004245973832439631, -0.3247299790382385, 0.08860589563846588, 0.16464708745479584, -0.41913899779319763, -0.275046169757843, 0.12339789420366287, 0.12206250429153442, 0.1778123527765274, -0.019305627793073654, 0.19298261404037476, 0.6014687418937683, -0.04290521889925003, 0.332793653011322, -0.02913454733788967, -0.04987672343850136, -0.06779877096414566, 0.00636169221252203, -0.1729622185230255, 0.0800589770078659, 0.372113436460495, 0.3768931031227112, 0.22836576402187347, -0.12036114186048508, -0.06467600166797638, 0.048560284078121185, 0.01717180199921131, 0.17235644161701202, -0.2735919654369354, 0.11902841925621033, -0.09038835018873215, 0.13026052713394165, 0.07079996913671494, -0.20963987708091736, 0.10960590094327927, 0.11861184984445572, -0.1082654669880867, -0.1014232188463211, 0.17040757834911346, -0.03653162717819214, 0.2522593140602112, -0.09298384189605713, -0.19044847786426544, -0.0768468976020813, 0.34443727135658264, 0.16699516773223877, 0.3699305057525635, -0.1490648090839386, -0.18440505862236023, -0.026089079678058624, -0.18871206045150757, -0.09965531527996063, -0.00915070902556181, -0.3066670298576355, 0.2665649950504303, 0.0757514163851738, 0.14168696105480194, 0.25459733605384827, 0.21736100316047668, -0.09045853465795517, 0.0599961094558239, 0.34681805968284607, -0.2202678769826889, -0.26668524742126465, 0.26403459906578064, 0.5363548398017883, -0.3322302997112274, -0.06163094937801361, -0.11595650762319565, 0.05984336510300636, 0.10731906443834305, 0.2752533257007599, -0.24124681949615479, 0.09731175750494003, -0.2812129855155945, 0.15977288782596588, 0.277683287858963, 0.43047964572906494, 0.005765583366155624, 0.17219854891300201, -0.09262100607156754, -0.0726955384016037, -0.5211905241012573, 0.18362893164157867, 0.22089828550815582, -0.14337533712387085, 0.25904494524002075, -0.026720646768808365, -0.18254797160625458, 0.26212772727012634, -0.3430914580821991, 0.19634003937244415, -0.03144605830311775, -0.22313150763511658, -0.2514569163322449, 0.03564029559493065, -0.03186897560954094, 0.25194746255874634, -0.1788475513458252, -0.10400603711605072, 0.20293304324150085, -0.1882825791835785, 0.1587805300951004, 0.06955938041210175, 0.0849522277712822, 0.1425607055425644, 0.2857636511325836, 0.28124237060546875, 0.16929639875888824, 0.12046930193901062, -0.03955979272723198, 0.09596648067235947, 0.0005992445512674749, -0.018906721845269203, -0.18427985906600952, -0.20551921427249908, 0.09908116608858109, 0.1971748322248459, -0.09581474214792252, 0.2707020938396454, 0.125995934009552, 0.16755031049251556, -0.02738119475543499, 0.22046251595020294, -0.11364443600177765, 0.16477476060390472, 0.032816410064697266, 0.2928229868412018, -0.1284545511007309, 0.22897560894489288, -0.0958426296710968, -0.14297913014888763, -0.07893211394548416, -0.19231940805912018, 0.25615236163139343, -0.2987460196018219, -0.04730783775448799, 0.028888504952192307, 0.3727441728115082, 0.009691968560218811, -0.31185048818588257, -0.4535421133041382, -0.284932017326355, 0.2458081841468811, -0.08283760398626328, 0.24263237416744232, 0.4404948353767395, -0.11160420626401901, -0.14189916849136353, -0.08976632356643677, 0.31003084778785706, 0.0736427754163742, 0.136420339345932, 0.07103100419044495, -0.3468833565711975 ]
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.
[ -0.24451442062854767, 0.06389829516410828, 0.024110952392220497, 0.3628231883049011, 0.10962121933698654, 0.4387523829936981, 0.2699873149394989, 0.36037400364875793, 0.044979169964790344, -0.11644787341356277, -0.38148143887519836, 0.12374304234981537, 0.003628471167758107, 0.27198493480682373, 0.271316796541214, -0.5642150044441223, 0.05804473161697388, 0.2187536060810089, -0.520010232925415, 0.18503697216510773, 0.0075752087868750095, 0.0051262639462947845, -0.16544689238071442, 0.012677289545536041, -0.20304378867149353, 0.23494401574134827, -0.34226277470588684, -0.22793914377689362, 0.04638424888253212, 0.1566457450389862, 0.3019718527793884, 0.16707980632781982, 0.21462081372737885, 0.31491565704345703, -0.00011992744111921638, -0.07949520647525787, -0.016489040106534958, -0.0905236303806305, 0.14189806580543518, -0.19872921705245972, 0.09682495892047882, -0.36194998025894165, -0.06335493922233582, -0.15658247470855713, 0.08227702230215073, -0.23953083157539368, -0.23318324983119965, 0.03752860426902771, 0.26808595657348633, -0.16602982580661774, 0.0541115365922451, -0.18832068145275116, -0.25114956498146057, 0.34095123410224915, -0.07127418369054794, -0.1605682522058487, 0.06146535277366638, -0.06943398714065552, 0.0818907618522644, -0.3195144534111023, 0.5779259204864502, 0.21333347260951996, -0.035110920667648315, -0.11987721174955368, 0.08233243227005005, 0.14248351752758026, 0.14339250326156616, -0.10950890928506851, 0.0553349070250988, 0.19000250101089478, 0.7741910219192505, -0.059649717062711716, 0.008376138284802437, -0.2915319502353668, 0.4038734436035156, -0.13154703378677368, -0.19321182370185852, 0.1430453509092331, 0.017043350264430046, 0.02549837715923786, -0.2548193633556366, -0.1563199907541275, -0.01640666462481022, 0.3023141622543335, -0.04316403716802597, 0.3315824270248413, 0.11024914681911469, 0.27707967162132263, -0.06372196227312088, -0.15143761038780212, 0.1660323292016983, -0.3364267945289612, 0.0009914570255205035, 0.15475763380527496, -0.3073638379573822, -0.40461912751197815, -0.3330456614494324, -0.18422916531562805, 0.2726256549358368, -0.15348823368549347, -0.1543896198272705, -0.006970979273319244, -0.10903885960578918, 0.13872398436069489, 0.30268219113349915, 0.1582721471786499, 0.13861030340194702, 0.21253813803195953, -0.18847203254699707, -0.31628328561782837, -0.3603154122829437, 0.16161029040813446, 0.07011686265468597, -0.07283031195402145, -0.12339430302381516, -0.2811368703842163, -0.058452315628528595, -0.34144559502601624, 0.14554041624069214, -0.1935318261384964, -0.2755388617515564, -0.23108302056789398, -0.10256723314523697, -0.15160973370075226, -0.01819605939090252, 0.34525585174560547, -0.14899347722530365, 0.3091168701648712, -0.10607459396123886, -0.013937526382505894, -0.025030450895428658, 0.07381553947925568, -0.17078657448291779, 0.12827013432979584, 0.09456326812505722, 0.17032937705516815, -0.13089480996131897, 0.4186180830001831, -0.18310484290122986, 0.10155940800905228, 0.3208995461463928, -0.19957460463047028, -0.07107540965080261, -0.2422417849302292, 0.31291234493255615, 0.16381865739822388, 0.1630842089653015, 0.10079876333475113, -0.3006672263145447, 0.1673177182674408, -0.17366595566272736, -0.08247633278369904, 0.015731122344732285, 0.05949963629245758, -0.08394327014684677, -0.30121949315071106, -0.12381064891815186, 0.6368820071220398, 0.2376345843076706, -0.31690341234207153, -0.14158758521080017, 0.09555433690547943, -0.46418431401252747, -0.04789251834154129, 0.20856665074825287, 0.025119595229625702, -0.35011258721351624, -0.13585299253463745, 0.0010479476768523455, -0.006365607958287001, -0.050108496099710464, 0.2889215648174286, -0.2664487361907959, 0.20656618475914001, 0.09581831097602844, 0.06720900535583496, 0.6602531671524048, 0.017082657665014267, -0.21449194848537445, 0.22153334319591522, -0.15071608126163483, 0.1325872242450714, -0.10475025326013565, 0.11741261929273605, -0.3766990005970001, -0.04700368270277977, 0.18447884917259216, 0.8385098576545715, -0.29177460074424744, -0.15610267221927643, -0.06377507746219635, -0.32391485571861267, 0.33796343207359314, 0.1553741693496704, 0.10215350985527039, -0.2498815953731537, -0.14259573817253113, 0.16948065161705017, 0.16629907488822937, -0.3560658097267151, 0.16821667551994324, 0.18711234629154205, -0.02912464737892151, 0.11733947694301605, -0.22763417661190033, -0.2857433557510376, -0.26597291231155396, -0.13448572158813477, 0.11240153759717941, 0.04761617258191109, 0.44605621695518494, -0.45091041922569275, 0.0663452073931694, -0.6603162288665771, -0.020744534209370613, 0.09107723832130432, -0.000595186953432858, 0.31140559911727905, 0.1058814525604248, -0.2892327308654785, -0.17570382356643677, 0.3350965976715088, -0.05201460048556328, -0.16662965714931488, -0.1528075635433197, 0.4420303404331207, 0.03853099048137665, 0.07375138252973557, 0.009351847693324089, 0.5092471241950989, 0.04848824441432953, 0.03464528173208237, 0.1494193971157074, -0.12538084387779236, -0.5496698021888733, 0.1898002177476883, 0.18356989324092865, 0.02857387438416481, -0.15538743138313293, -0.06806912273168564, 0.354319304227829, -0.33442002534866333, -0.005318326409906149, -0.16024217009544373, -0.07414641231298447, 0.41108188033103943, -0.24898123741149902, 0.36992478370666504, -0.020390622317790985, -0.11749589443206787, 0.016925225034356117, -0.1534186750650406, 0.021901758387684822, -0.10193333774805069, 0.21405459940433502, -0.007762130815535784, 0.013551181182265282, 0.3495674431324005, -0.38780584931373596, 0.2461586892604828, 0.2839241027832031, 0.10369336605072021, -0.04670143872499466, -0.014283232390880585, -0.14048579335212708, -0.047391388565301895, 0.034929390996694565, 0.17373251914978027, 0.5730594992637634, 0.19658587872982025, 0.201826810836792, 0.2332468330860138, -0.15427257120609283, 0.12579300999641418, 0.01481454074382782, 0.1872757077217102, 0.06608603149652481, 0.3079126179218292, 0.09058104455471039, 0.2213158756494522, 0.1897101253271103, 0.3987444043159485, 0.203317329287529, -0.09153176099061966, -0.1926695704460144, -0.005747835151851177, -0.28501638770103455, -0.6398730278015137, -0.562881350517273, -0.12821361422538757, -0.2211201786994934, -0.05306992679834366, -0.09376560896635056, -0.005284679122269154, 0.008729584515094757, 0.18447613716125488, -0.11454116553068161, 0.42126086354255676, -0.47502198815345764, -0.277840256690979, 0.2787611186504364, 0.04162327200174332, -0.18180076777935028, 0.0709051862359047, 0.4079384505748749, 0.21213701367378235, 0.25454801321029663, -0.0574188269674778, -0.23279757797718048, -0.06010572239756584, -0.5316917896270752, 0.05019499734044075, -0.3654108941555023, -0.1488938182592392, -0.2518879175186157, -0.10150037705898285, -0.21497702598571777, -0.3766782283782959, 0.05155947804450989, 0.31181761622428894, 0.05274160951375961, -0.1287006437778473, 0.05726388469338417, 0.28694093227386475, -0.10500960052013397, -0.2945322096347809, -0.42148059606552124, -0.0598759762942791, 0.0423828661441803, -0.16191211342811584, -0.02081853151321411, -0.36425963044166565, -0.05395602434873581, -0.036509424448013306, -0.26720693707466125, -0.011379368603229523, -0.2016235589981079, 0.0008073665667325258, 0.17877979576587677, -0.13474725186824799, -0.1283462941646576, -0.08461106568574905, -0.18979698419570923, 0.29974427819252014, -0.030816268175840378, -0.10249115526676178, -0.15170349180698395, -0.1399437040090561, -0.002392246387898922, -0.14289122819900513, 0.37742260098457336, 0.17890049517154694, 0.01767212525010109, 0.1485680341720581, 0.07189610600471497, -0.22946836054325104, 0.021115431562066078, 0.09855346381664276, 0.4574984312057495, -0.19675150513648987, 0.25292739272117615, -0.14573287963867188, 0.6077175140380859, 0.5213464498519897, 0.1669510304927826, 0.06605769693851471, 0.2555110454559326, 0.14240215718746185, 0.06331991404294968, -0.20977550745010376, 0.20084503293037415, 0.19463007152080536, 0.0576697513461113, 0.26992419362068176, 0.10378645360469818, -0.053351953625679016, -0.33397653698921204, 0.09006878733634949, -0.3042643368244171, -0.11738701909780502, 0.3777521848678589, -0.2725737988948822, 0.11911562085151672, 0.01102694496512413, -0.4641566276550293, -0.3556419312953949, -0.1309647113084793, -0.216474711894989, 0.1183912456035614, -0.25855621695518494, 0.1361861526966095, -0.5622233748435974, -0.10364524275064468, -0.14068900048732758, 0.32386577129364014, 0.15147583186626434, 0.30835074186325073, 0.04973035678267479, 0.055048100650310516, -0.24609512090682983, -0.04382041096687317, 0.06486909836530685, -0.4924119710922241, -0.1347343623638153, 0.06497912853956223, 0.014051605947315693, -0.23107777535915375, -0.17589326202869415, 0.017376063391566277, -0.09139563888311386, -0.18299783766269684, 0.33646029233932495, -0.31847772002220154, -0.15623900294303894, 0.6944684982299805, 0.4387205243110657, -0.07578423619270325, -0.3850887715816498, 0.04643447697162628, -0.12323157489299774, -0.14339596033096313, -0.26000720262527466, -0.047510482370853424, 0.1632034033536911, -0.3689453899860382, 0.15044574439525604, 0.01641179621219635, 0.0803886204957962, 0.4504483640193939, 0.10332701355218887, 0.07213606685400009, 0.35266342759132385, 0.38764581084251404, -0.24028679728507996, -0.08153365552425385, -0.252618670463562, 0.4975554943084717, -0.07558052986860275, -0.3227156102657318, -0.17115657031536102, -0.1438242346048355, 0.7155787348747253, -0.06854939460754395, 0.1511317938566208, 0.2316923439502716, -0.3995976448059082, -0.3615087866783142, 0.00195317342877388, 0.2028462439775467, 0.2826133370399475, 0.32211485505104065, -0.3971284031867981, -0.44975557923316956, 0.6233007907867432, 0.4106391966342926, -0.020718945190310478, 0.2541210651397705, -0.25738000869750977, -0.1741575300693512, 0.09945567697286606, -0.10148286074399948, 0.8141265511512756, 0.021813640370965004, 0.38159307837486267, -0.0004787243960890919, -0.16811983287334442, 0.7002722024917603, -0.09663330763578415, 0.019921788945794106, 0.020172040909528732, -0.1437642127275467, -0.03538588434457779, -0.030096080154180527, -0.12921607494354248, 0.3702211380004883, -0.20778626203536987, 0.3171716034412384, -0.0004518145869951695, -0.005242947489023209, 0.003598739393055439, 0.40919196605682373, 0.06669451296329498, -0.2587724030017853, -0.3931366503238678, -0.030338620766997337, -0.30272606015205383, -0.053879257291555405, -0.1444965898990631, 0.07147324830293655, -0.1313784122467041, 0.07047852128744125, -0.1482679396867752, 0.2999640107154846, 0.5066033005714417, 0.044584814459085464, -0.4117036461830139, -0.4415697753429413, 0.2938781678676605, 0.14138120412826538, 0.3556128740310669, 0.029089635238051414, 0.02995366044342518, 0.014914795756340027, -0.35656335949897766, -0.015754682943224907, -0.11483736336231232, -0.29389652609825134, 0.2594217360019684, 0.020187318325042725, -0.2533404529094696, 0.06901320815086365, 0.41276490688323975, -0.0823650062084198, -0.254402756690979, -0.17855213582515717, 0.33148983120918274, -0.059461094439029694, -0.5873780846595764, 0.6571645736694336, 0.15120765566825867, -0.07791920751333237, 0.01005462184548378, -0.11371768265962601, -0.2892293632030487, 0.3845168650150299, -0.27283963561058044, -0.2554665505886078, 0.06018335372209549, 0.18039125204086304, 0.30600082874298096, -0.3658260405063629, 0.470186322927475, 0.18391893804073334, 0.2744099795818329, 0.02297348901629448, -0.06155148148536682, 0.2622483968734741, -0.07377257198095322, -0.35813167691230774, 0.10253635048866272, 0.2045326828956604, 0.11256218701601028, 0.1527308225631714, 0.16331548988819122, 0.12227483838796616, 0.20673204958438873, -0.06777944415807724, -0.37824222445487976, 0.2728559672832489, 0.20024073123931885, 0.26178982853889465, 0.2570183277130127, 0.312802791595459, 0.06446346640586853, 0.6031779646873474, -0.1696951985359192, -0.01909763738512993, 0.13622349500656128, 0.2852420210838318, 0.3465158939361572, 0.08450741320848465, -0.4070693850517273, 0.014368534088134766, -0.044221241027116776, 0.0018222603248432279, 0.11410646140575409, -0.1442725509405136, -0.07686270773410797, 0.17597182095050812, 0.3449315130710602, 0.001123090391047299, 0.060233499854803085, -0.28074148297309875, -0.011823081411421299, -0.02250852808356285, 0.06651259958744049, -0.09257772564888, -0.00026302237529307604, 0.38064044713974, -0.07365303486585617, 0.03403794392943382, -0.2856486439704895, 0.26426953077316284, -0.13235028088092804, -0.03567943349480629, 0.2590028941631317, 0.10173726081848145, 0.3916665315628052, -0.044483110308647156, -0.25726914405822754, -0.26786479353904724, 0.21714018285274506, 0.06505512446165085, -0.34551647305488586, -0.034969545900821686, 0.04609065130352974, -0.06114375591278076, 0.15212306380271912, 0.3250304162502289, -0.10721460729837418, -0.02506890520453453, 0.2345830798149109, 0.0007480124477297068, -0.19700098037719727, 0.07210493832826614, 0.23050753772258759, -0.06402739137411118, -0.1415853053331375, 0.2195843607187271, 0.09061222523450851, 0.37726160883903503, -0.3111916780471802, -0.012798964977264404, -0.007480953354388475, -0.03604814037680626, 0.32435402274131775, 0.5464572906494141, -0.03242063522338867, -0.10312500596046448, -0.012002094648778439, 0.08090408146381378, 0.05886415019631386, -0.05017813667654991, -0.1465628743171692, -0.06746087968349457, -0.08968833088874817, -0.07963152974843979, 0.23124250769615173, -0.08268310129642487, 0.06398668140172958, 0.2914492189884186, 0.22425895929336548, 0.1321953386068344, 0.059876635670661926, 0.275981068611145, 0.005738543346524239, -0.46567675471305847, 0.027943192049860954, 0.4166928231716156, -0.07972187548875809, 0.0029158787801861763, -0.39661502838134766, -0.003886318998411298, -0.14031349122524261, 0.20574171841144562, -0.12469922006130219, -0.3338131606578827, 0.5012161135673523, 0.0929168239235878, -0.14959874749183655, -0.2275853455066681, 0.12669172883033752, 0.5161341428756714, -0.03509517014026642, -0.04846249520778656, 0.052023690193891525, 0.16546930372714996, -0.12800726294517517, 0.03769943118095398, 0.25175681710243225, 0.11086656898260117, 0.029335731640458107, 0.00114126643165946, 0.024302998557686806, 0.14044071733951569, 0.122123584151268, 0.1034310832619667, 0.019178001210093498, 0.08414564281702042, -0.016205167397856712, 0.2771754264831543, 0.10079758614301682, -0.06125640869140625, 0.07313276827335358, -0.12367552518844604, 0.04407656565308571, -0.5888586044311523, 0.26579606533050537, 0.31551533937454224, -0.023779813200235367, -0.05252767726778984, 0.17175760865211487, -0.3293569087982178, -0.1878260374069214, 0.5251398682594299, 0.29269322752952576, -0.15609648823738098, -0.15337060391902924, 0.035338111221790314, -0.029740890488028526, 0.3186293840408325, 0.3571053743362427, 0.08333561569452286, -0.4211038649082184, -0.18525885045528412, -0.7274481654167175, 0.11188367009162903, -0.19989605247974396, -0.1187603622674942, -0.26006925106048584, 0.040484871715307236, 0.2637613117694855, 0.23323555290699005, 0.0643841102719307, 0.4323757290840149, 0.14185287058353424, 0.1557859182357788, -0.434123694896698, -0.15743933618068695, 0.08547850698232651, 0.1868932545185089, 0.05970900505781174, -0.11151831597089767, 0.024231700226664543, 0.35462501645088196, -0.18669794499874115, -0.06368324160575867, 0.019787654280662537, 0.13367481529712677, -0.0402192547917366, 0.23977045714855194, 0.036836374551057816, 0.6354695558547974, -0.16606947779655457, 0.09476257860660553, 0.11244731396436691, -0.05096949636936188, -0.5395125150680542, 0.3559364676475525, -0.15957552194595337, 0.019293615594506264, -0.6011695861816406, 0.6696321368217468, 0.12462131679058075, -0.28975358605384827, -0.08932241797447205, 0.12198381125926971, -0.05398508161306381, -0.13608358800411224, 0.2872994840145111, 0.0695643275976181, 0.28958457708358765, 0.6231248378753662, 0.04480632022023201, 0.08362893015146255, -0.18011711537837982, -0.09812998026609421, 0.23890714347362518, 0.1017700806260109, -0.4384743273258209, -0.09930212050676346, -0.1474403738975525, 0.020527837797999382, -0.08745389431715012, -0.532736599445343, -0.036130454391241074, 0.32340356707572937, -0.1769234985113144, 0.38899150490760803, 0.501995861530304, 0.016691256314516068, 0.06571802496910095, -0.29338493943214417, -0.06997093558311462, -0.07178298383951187, -0.07720521092414856, -0.12362583726644516, -0.2613336443901062 ]
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 ?
[ 0.34102779626846313, -0.05078335106372833, 0.049837399274110794, 0.20578496158123016, -0.012494249269366264, -0.09158316999673843, 0.1314336061477661, 0.24781465530395508, 0.0751405730843544, -0.2193237841129303, -0.16563907265663147, 0.10058040916919708, 0.06382808089256287, -0.2900847792625427, 0.24017465114593506, 0.07839706540107727, 0.09766232222318649, 0.11659568548202515, 0.14824886620044708, -0.19698044657707214, -0.11032877117395401, 0.4662145674228668, -0.11777035892009735, -0.12418773025274277, 0.02648157998919487, 0.05782897770404816, -0.25497734546661377, 0.22920295596122742, -0.17581433057785034, -0.2693144679069519, 0.22969695925712585, 0.4607448875904083, 0.010647718794643879, 0.2442495971918106, -0.00012602117203641683, -0.026637118309736252, 0.40851646661758423, 0.03749185428023338, -0.07534395903348923, -0.1284836083650589, 0.15254181623458862, -0.3127950429916382, 0.39784446358680725, -0.1065424457192421, 0.042281586676836014, -0.3930911719799042, 0.17608484625816345, -0.046389833092689514, 0.1798725724220276, 0.0012875365791842341, 0.08718014508485794, 0.12423455715179443, 0.06380897760391235, 0.06727614998817444, 0.4933622479438782, 0.010115666314959526, 0.06612017750740051, 0.4198921322822571, 0.2697267532348633, -0.16339384019374847, 0.13654913008213043, 0.24459710717201233, -0.11495183408260345, 0.15639546513557434, 0.17969723045825958, 0.4194624423980713, 0.059644218534231186, -0.46717801690101624, -0.20219317078590393, 0.3455394208431244, 0.014854674227535725, 0.01301858201622963, -0.04609852656722069, -0.467778742313385, -0.012200791388750076, 0.0002347747067688033, 0.20772789418697357, 0.4563741683959961, -0.16081508994102478, -0.06489236652851105, 0.0811813548207283, -0.2500705420970917, -0.06046299636363983, 0.5617239475250244, 0.16432194411754608, 0.04834136739373207, -0.19733893871307373, 0.36373698711395264, 0.4003930389881134, 0.16857297718524933, -0.262498140335083, -0.2875666320323944, 0.1975400596857071, 0.1875079870223999, -0.30293357372283936, 0.008081452921032906, 0.08924084156751633, -0.06685435026884079, 0.240162193775177, 0.34288290143013, 0.08698280155658722, -0.04315385967493057, 0.1938953697681427, 0.12176168709993362, 0.5296923518180847, -0.06093317270278931, -0.02097371220588684, 0.18622180819511414, 0.3687584102153778, -0.01659138686954975, 0.08867666125297546, 0.09830335527658463, 0.09169550240039825, -0.025422725826501846, -0.23397570848464966, -0.5981558561325073, 0.28797727823257446, -0.2756761908531189, 0.24873140454292297, -0.023608040064573288, -0.011191051453351974, -0.05599597096443176, -0.2106602042913437, 0.004625738598406315, 0.018986813724040985, 0.22263045608997345, 0.11902975291013718, 0.28089162707328796, -0.20559212565422058, 0.20877200365066528, -0.07826906442642212, 0.005981262307614088, -0.365675151348114, 0.03970423340797424, 0.2711541950702667, -0.27817487716674805, 0.167244091629982, 0.09166508167982101, -0.11322484910488129, -0.18401117622852325, 0.17918333411216736, -0.1315322071313858, 0.06672147661447525, 0.3698318898677826, 0.277812659740448, 0.5080726742744446, -0.001968075055629015, -0.25453245639801025, -0.2754872739315033, -0.06114352121949196, -0.2797985076904297, -0.16831764578819275, -0.18681003153324127, 0.029858101159334183, -0.390236496925354, 0.16661086678504944, 0.00011750298290280625, 0.2882414162158966, 0.15523885190486908, -0.170510932803154, -0.11550763249397278, -0.004567788448184729, -0.08512488752603531, -0.3703431189060211, 0.12198787927627563, 0.34375467896461487, 0.19060105085372925, -0.10569141060113907, -0.1270904392004013, -0.10054047405719757, 0.08209976553916931, 0.31115853786468506, -0.2761954367160797, 0.14208747446537018, -0.24165400862693787, 0.2310817986726761, 0.5904169082641602, -0.2710139751434326, -0.4526691436767578, 0.31155797839164734, -0.024965088814496994, 0.07489549368619919, 0.15223027765750885, 0.2749258875846863, -0.1387360394001007, -0.02256184071302414, 0.1659156233072281, 0.3668280243873596, 0.09084151685237885, -0.14627477526664734, -0.30497610569000244, -0.19426612555980682, 0.3023199439048767, -0.09074103087186813, 0.11595363169908524, 0.005938899703323841, -0.008974879048764706, 0.06970670074224472, 0.3066055178642273, 0.22949601709842682, 0.18607576191425323, 0.24630209803581238, -0.0749327540397644, -0.009011047892272472, 0.07086270302534103, -0.20000094175338745, -0.3342156410217285, 0.05639549717307091, -0.3555223047733307, 0.24316728115081787, 0.2622112035751343, -0.04236580431461334, -0.23279793560504913, -0.4931893050670624, -0.17978306114673615, -0.2683068513870239, 0.004727217368781567, 0.21896950900554657, 0.12241210788488388, 0.1607268899679184, -0.2678830623626709, 0.17435497045516968, 0.009839766658842564, 0.08805976808071136, -0.8079341650009155, 0.31794631481170654, -0.017846200615167618, -0.07045122236013412, 0.06313435733318329, 0.14330022037029266, 0.0012619447661563754, -0.16912347078323364, 0.324447363615036, 0.43658119440078735, -0.13749289512634277, 0.2282072901725769, 0.22558575868606567, 0.057517461478710175, 0.2537793517112732, -0.11084946244955063, 0.1513350009918213, -0.12579162418842316, 0.11191107332706451, -0.23113112151622772, -0.30425307154655457, 0.1168283149600029, -0.07708783447742462, 0.3747081756591797, 0.1469537615776062, -0.011496412567794323, 0.010349145159125328, -0.23012566566467285, -0.47462472319602966, -0.21605578064918518, 0.20148734748363495, 0.22801314294338226, -0.11197298765182495, 0.14593084156513214, -0.299826443195343, 0.0780768021941185, 0.07013919949531555, 0.01723259873688221, -0.020467063412070274, 0.17367787659168243, -0.2714231610298157, -0.2440957874059677, -0.18221411108970642, 0.5162525177001953, 0.8718119263648987, 0.19785279035568237, 0.3229740262031555, 0.10880719125270844, -0.1763385683298111, -0.1969754844903946, 0.08697979152202606, 0.0337328277528286, -0.128058522939682, 0.21404008567333221, 0.1713268756866455, -0.008642139844596386, -0.423177033662796, 0.3805910348892212, 0.1604657620191574, 0.021289879456162453, -0.3763745129108429, -0.19031330943107605, -0.40123873949050903, -0.30128464102745056, -0.6075862646102905, -0.18438860774040222, -0.3624557852745056, -0.3466329872608185, -0.06475558876991272, 0.1399504542350769, -0.0504867359995842, 0.024639088660478592, -0.2065400779247284, 0.1555737853050232, -0.23960037529468536, 0.06841801106929779, 0.14115744829177856, 0.12482795864343643, -0.30138179659843445, -0.08074276149272919, 0.036516401916742325, 0.3887549042701721, 0.5186222195625305, 0.05468956008553505, -0.2073938399553299, -0.08984853327274323, -0.4579674303531647, 0.2533680498600006, -0.20390865206718445, 0.4136425256729126, 0.28525519371032715, 0.09833887219429016, 0.3634144365787506, -0.0833398625254631, 0.28173157572746277, -0.1454361230134964, 0.03042089007794857, 0.008215281181037426, 0.0834251120686531, -0.020247509703040123, -0.285076767206192, -0.5693516731262207, -0.3060919940471649, -0.2529955506324768, -0.1100715696811676, -0.05502317473292351, 0.12144868075847626, 0.24505053460597992, -0.08112357556819916, 0.12399709224700928, -0.46485644578933716, -0.045524656772613525, -0.364004522562027, -0.08933587372303009, 0.24169166386127472, -0.22708259522914886, -0.4088027775287628, -0.04336238279938698, -0.02477342262864113, -0.09826140850782394, 0.1594512164592743, -0.38081422448158264, -0.4189055562019348, -0.21503789722919464, -0.5069785118103027, 0.17532086372375488, 0.4170130491256714, 0.30164089798927307, -0.33650606870651245, -0.10329844057559967, -0.08495713025331497, -0.16373036801815033, 0.188657745718956, 0.21752291917800903, 0.3610684871673584, 0.20806385576725006, -0.09732368588447571, -0.015406065620481968, 0.5112147331237793, 0.45359474420547485, 0.1888439804315567, 0.25888192653656006, -0.030591053888201714, 0.629603385925293, 0.08015437424182892, -0.3526654839515686, 0.002588399453088641, -0.023501774296164513, 0.2995644807815552, 0.3215644955635071, 0.24641649425029755, -0.17453652620315552, -0.09905529022216797, 0.14814409613609314, -0.3503150939941406, -0.3976816236972809, 0.2206421196460724, -0.1181030422449112, 0.03348469361662865, 0.21164005994796753, -0.37359678745269775, 0.05672076344490051, -0.05050113797187805, -0.3561001121997833, 0.39024850726127625, -0.13153056800365448, 0.12844505906105042, -0.5994896292686462, 0.1178508996963501, -0.4363969564437866, 0.13917545974254608, -0.12238948792219162, 0.16424328088760376, 0.11637768149375916, -0.06740279495716095, 0.3922005295753479, 0.02802746370434761, 0.5079420804977417, -0.13392257690429688, -0.21793556213378906, 0.18318776786327362, -0.0962795540690422, -0.8417407870292664, -0.16880609095096588, -0.07770325988531113, -0.11526867747306824, 0.27617591619491577, 0.3028244078159332, -0.5176627039909363, -0.25081339478492737, 0.20353537797927856, 0.1481359899044037, -0.17218632996082306, -0.24935148656368256, -0.2638273537158966, -0.44855746626853943, -0.31027752161026, -0.39140889048576355, -0.12274371832609177, -0.05234165117144585, -0.11759010702371597, -0.12879632413387299, 0.1574462503194809, -0.17029491066932678, 0.14247268438339233, -0.16090027987957, 0.2610478699207306, -0.02096601203083992, 0.03182753175497055, 0.46520236134529114, 0.13367347419261932, -0.5750645995140076, 0.5631839036941528, -0.187859445810318, 0.28968027234077454, 0.13970714807510376, -0.17257241904735565, 0.36342278122901917, 0.5340467691421509, -0.06888541579246521, 0.0670914351940155, 0.0395318828523159, -0.07136163115501404, -0.11491316556930542, 0.2556438744068146, 0.5111373066902161, -0.34914931654930115, -0.13880009949207306, -0.33767691254615784, 0.559675931930542, 0.08166491985321045, -0.4084246754646301, 0.1992589682340622, 0.26665207743644714, -0.24415436387062073, 0.2821599543094635, -0.35387280583381653, 1.0832159519195557, 0.04487154260277748, 0.31874552369117737, 0.08139211684465408, -0.2079673558473587, 0.8912696242332458, -0.019846847280859947, 0.02509354054927826, -0.2386833131313324, -0.04003005474805832, 0.01605117693543434, -0.18522711098194122, 0.05906528979539871, 0.08315662294626236, -0.0264119915664196, 0.29316964745521545, 0.22072955965995789, 0.09865696728229523, -0.10229635238647461, 0.5956615209579468, 0.04097189009189606, -0.082302987575531, -0.32925793528556824, 0.04090885445475578, -0.013093531131744385, 0.315154492855072, 0.06848498433828354, -0.1459345817565918, -0.22294282913208008, -0.406307578086853, -0.26716163754463196, 0.04904678463935852, 0.058977171778678894, 0.13230879604816437, 0.10286783427000046, -0.11787692457437515, -0.2950005829334259, 0.029081668704748154, -0.04239268600940704, 0.019790980964899063, -0.25581488013267517, -0.08757343143224716, 0.003971846308559179, -0.307388037443161, -0.1308579444885254, -0.22023814916610718, 0.27061933279037476, -0.12889155745506287, -0.18283432722091675, -0.23630546033382416, 0.04053317755460739, -0.15919391810894012, -0.4621095657348633, 0.31663990020751953, 0.2644231617450714, -0.08291379362344742, -0.2963554859161377, 0.36175650358200073, 0.11170415580272675, -0.002364293672144413, 0.029360229149460793, 0.35258543491363525, 0.0783945843577385, 0.36199334263801575, 0.13534210622310638, -0.19166703522205353, -0.06823708117008209, 0.3341466784477234, 0.5289828181266785, -0.09188293665647507, 0.255486398935318, -0.11541899293661118, 0.05190796032547951, -0.050998587161302567, -0.16937889158725739, -0.08383142203092575, -0.3399844765663147, 0.12293007224798203, -0.012368030846118927, 0.06921043992042542, 0.13143208622932434, 0.04305218160152435, 0.09093338996171951, -0.21702660620212555, -0.10895178467035294, -0.4528346061706543, 0.06527455896139145, 0.29334092140197754, -0.17236968874931335, 0.40974047780036926, 0.04056304320693016, -0.12469840794801712, 0.2502572238445282, 0.3126417100429535, -0.16058333218097687, 0.25755932927131653, 0.06299512088298798, 0.24464008212089539, 0.6299229860305786, 0.01803307980298996, 0.08861178904771805, -0.11040426790714264, -0.0822521448135376, -0.05710297450423241, -0.17737196385860443, -0.09637579321861267, -0.2691380977630615, 0.22645147144794464, 0.14458200335502625, 0.05872960761189461, 0.09794029593467712, -0.23562267422676086, 0.03194238245487213, 0.008571350947022438, 0.2074664980173111, -0.06657173484563828, -0.03942098468542099, -0.025018123909831047, -0.16160450875759125, 0.03848215192556381, 0.11174434423446655, 0.34071168303489685, -0.36419472098350525, -0.04273552820086479, 0.3421163260936737, 0.008133063092827797, -0.21149621903896332, -0.09788176417350769, -0.2852182686328888, -0.030537595972418785, 0.10420700907707214, -0.04821816459298134, 0.11100775003433228, 0.006433622911572456, -0.02997579611837864, -0.18221504986286163, 0.3628309965133667, 0.5864512324333191, -0.26623326539993286, 0.08765041083097458, -0.052409376949071884, 0.024477757513523102, -0.1797003298997879, -0.03848240152001381, 0.2230921983718872, 0.26938506960868835, -0.08858183771371841, 0.26846587657928467, 0.3248119652271271, 0.025378376245498657, -0.12275755405426025, -0.019717741757631302, 0.6393312811851501, -0.19335432350635529, 0.08591996878385544, -0.24014243483543396, -0.04187876358628273, -0.11093230545520782, -0.017787884920835495, -0.033447422087192535, 0.00756439846009016, -0.09867125749588013, 0.2037726193666458, 0.20150573551654816, -0.03543462976813316, 0.052559882402420044, 0.23083479702472687, -0.18457423150539398, 0.016510067507624626, 0.7022261023521423, 0.2064039707183838, 0.30672261118888855, 0.14233773946762085, 0.7357051372528076, -0.11739078164100647, -0.05003533884882927, -0.1399213969707489, 0.5564059615135193, 0.26854610443115234, -0.23760467767715454, 0.06799247115850449, -0.11369524896144867, -0.14237555861473083, 0.04823797568678856, -0.044042203575372696, -0.3282136619091034, 0.2714610993862152, -0.027660688385367393, -0.19461435079574585, -0.735490083694458, 0.16255097091197968, 0.07861092686653137, 0.10736706852912903, -0.1520148664712906, 0.18626007437705994, -0.07005293667316437, -0.05007312074303627, 0.21916493773460388, 0.5182145833969116, 0.2908778190612793, 0.3593868315219879, -0.006481168791651726, -0.16204559803009033, 0.24868382513523102, 0.1770855337381363, -0.20950035750865936, 0.15459460020065308, 0.3400493860244751, 0.07801326364278793, 0.3201008141040802, 0.042092371731996536, -0.10348353534936905, -0.06722695380449295, 0.07740709185600281, 0.029380669817328453, -0.32071974873542786, 0.5537846684455872, -0.28373536467552185, 0.08009999990463257, -0.01782398298382759, 0.06254684925079346, -0.6169646382331848, -0.19064868986606598, 0.21398425102233887, 0.005209355615079403, 0.16936850547790527, -0.4606564939022064, 0.014075801707804203, 0.17706593871116638, 0.5300926566123962, 0.04728841036558151, 0.06634968519210815, -0.1545790433883667, -0.24412722885608673, -0.7898310422897339, 0.1378147453069687, -0.22904358804225922, 0.14415238797664642, -0.048181161284446716, 0.39347073435783386, 0.03287399560213089, 0.3112463653087616, -0.022220024839043617, 0.13064299523830414, -0.5399712324142456, -0.13637645542621613, -0.17598359286785126, -0.04879879951477051, -0.21684108674526215, -0.05826788768172264, 0.005118986591696739, -0.11321891099214554, 0.10097973793745041, -0.2996004819869995, -0.07709416002035141, 0.09741343557834625, 0.07090741395950317, 0.07483363151550293, 0.054494455456733704, 0.40695109963417053, 0.23451268672943115, 0.41267144680023193, -0.0002772534789983183, -0.28445568680763245, -0.5316895842552185, 0.045554738491773605, -0.4169575273990631, 0.09936404973268509, 0.02154901809990406, 0.20746104419231415, -0.44158896803855896, 0.15417622029781342, -0.11783290654420853, -0.07033373415470123, 0.13117671012878418, 0.05652361363172531, 0.03648356348276138, -0.008407408371567726, 0.14452466368675232, 0.1382405012845993, 0.233476921916008, 0.41835591197013855, -0.22531266510486603, 0.1311774104833603, -0.19558264315128326, -0.2739400565624237, 0.2897918224334717, -0.36235591769218445, -0.13488730788230896, 0.12069548666477203, 0.24682088196277618, 0.07022982090711594, 0.06666937470436096, -0.6980176568031311, -0.2639472484588623, 0.3107895851135254, -0.04248126223683357, -0.18357066810131073, 0.2500549554824829, -0.13164953887462616, -0.47837743163108826, -0.022704219445586205, 0.04410530626773834, 0.1565810590982437, -0.38131529092788696, 0.011634189635515213, -0.23016807436943054 ]
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')` ```
[ 0.33375197649002075, 0.1518261432647705, 0.040382154285907745, 0.12209121882915497, -0.011582503095269203, 0.07017665356397629, 0.10876744240522385, 0.38690340518951416, -0.11695444583892822, -0.2211580127477646, -0.12466738373041153, 0.3359685242176056, 0.04204702377319336, -0.3219139873981476, 0.3353402018547058, 0.05978969857096672, 0.17052800953388214, 0.11987818777561188, 0.2469612956047058, -0.3221700191497803, -0.15091955661773682, 0.3626810908317566, -0.20078301429748535, -0.07502932101488113, 0.1717316210269928, 0.1704220175743103, -0.13634684681892395, 0.23558419942855835, -0.28036171197891235, -0.2833782136440277, 0.3049052953720093, 0.4118289053440094, 0.12482468783855438, 0.32855623960494995, -0.0001224009320139885, -0.051074590533971786, 0.39966580271720886, -0.11453273147344589, -0.12407517433166504, -0.24390459060668945, 0.0256196279078722, -0.27727851271629333, 0.36111265420913696, -0.15762540698051453, 0.10323170572519302, -0.5018043518066406, 0.22668278217315674, -0.10192952305078506, 0.0027054925449192524, 0.2615557610988617, 0.12145350128412247, -0.08232896775007248, 0.1363518238067627, 0.0036471381317824125, 0.32640454173088074, 0.01863124780356884, 0.09062410145998001, 0.3079109787940979, 0.17882327735424042, 0.0836486741900444, 0.19623695313930511, 0.21780842542648315, -0.12764735519886017, 0.07411874830722809, 0.3337647020816803, 0.3773285746574402, 0.04918453097343445, -0.4533299505710602, -0.09537652134895325, 0.25129425525665283, 0.4695171117782593, -0.08259282261133194, 0.03704744577407837, -0.20686861872673035, -0.04010647535324097, -0.08870533853769302, 0.1081099659204483, 0.3900112509727478, -0.2233264148235321, -0.048257146030664444, 0.021009469404816628, -0.38639742136001587, -0.1963813155889511, 0.5864763855934143, 0.16923370957374573, 0.1530344933271408, -0.08709679543972015, 0.3638681471347809, 0.21774189174175262, 0.13750766217708588, 0.03393637761473656, -0.3225853145122528, 0.2554100751876831, 0.28198620676994324, -0.29220548272132874, -0.007908441126346588, -0.03504044562578201, -0.2091565579175949, 0.1889667809009552, 0.32612454891204834, 0.07502587139606476, -0.06704691797494888, 0.20057809352874756, 0.1209656298160553, 0.4643363356590271, -0.15208934247493744, 0.08678403496742249, 0.3539924919605255, 0.21561086177825928, -0.18660417199134827, 0.03401937708258629, 0.01824498362839222, -0.0772193968296051, 0.07174420356750488, -0.22967688739299774, -0.6142797470092773, 0.14092853665351868, -0.4055895209312439, 0.44782325625419617, -0.02523888647556305, 0.21384772658348083, -0.10629221051931381, -0.23445355892181396, 0.0817299485206604, 0.10517361760139465, 0.20921184122562408, 0.0853356346487999, 0.176457941532135, -0.11764825880527496, 0.09517589956521988, -0.08684976398944855, 0.016505030915141106, -0.3610490560531616, -0.05815691500902176, 0.2363770604133606, -0.01321476511657238, 0.2610936164855957, -0.09314679354429245, 0.07390692085027695, -0.09016013145446777, 0.24405549466609955, -0.0596819631755352, 0.144902765750885, 0.49642637372016907, 0.3115929365158081, 0.5195354223251343, -0.00015595913282595575, -0.1992071121931076, -0.2772218883037567, 0.15920612215995789, -0.3895840346813202, -0.12449294328689575, -0.11835958808660507, 0.07398617267608643, -0.31114864349365234, -0.06145261228084564, 0.1078469529747963, 0.21074190735816956, 0.18148663640022278, -0.3582732379436493, -0.10828767716884613, 0.025629768148064613, 0.04021826013922691, -0.37532007694244385, 0.1704997718334198, 0.1312365084886551, 0.06589086353778839, -0.03890412300825119, -0.2859936058521271, -0.06429071724414825, 0.2266976684331894, 0.07826383411884308, -0.30781376361846924, 0.15587744116783142, -0.1210067942738533, 0.0539570152759552, 0.8645718097686768, -0.2199849933385849, -0.4071541428565979, 0.27943721413612366, 0.054437194019556046, -0.14355960488319397, 0.09176154434680939, 0.3109307587146759, -0.284670889377594, 0.051498670130968094, 0.09130977094173431, 0.288500040769577, -0.012590395286679268, -0.2461199313402176, -0.1448046714067459, -0.11279936879873276, 0.3086381256580353, -0.07593356817960739, 0.294357031583786, 0.16808992624282837, -0.015603047795593739, 0.14979897439479828, 0.41001641750335693, 0.187556654214859, 0.21378687024116516, 0.22659936547279358, -0.11479006707668304, -0.1092505231499672, 0.20860177278518677, -0.1760011613368988, -0.16421034932136536, 0.06642291694879532, -0.4373757243156433, 0.26424771547317505, 0.32023921608924866, 0.0615609847009182, -0.29039329290390015, -0.5201038122177124, -0.06402744352817535, -0.3106151819229126, -0.012993893586099148, 0.2909797728061676, -0.0014553242363035679, 0.0786508247256279, -0.35568273067474365, 0.04744627699255943, -0.12707829475402832, 0.014778073877096176, -0.76802659034729, 0.15682993829250336, -0.002586372895166278, -0.014122436754405499, 0.08916139602661133, -0.030356569215655327, -0.041416238993406296, -0.018181784078478813, 0.22616015374660492, 0.41557711362838745, -0.03187999874353409, 0.3567318916320801, 0.35596171021461487, -0.10024559497833252, 0.1163010224699974, -0.23534353077411652, 0.15720002353191376, -0.10833905637264252, 0.26213565468788147, -0.20873038470745087, -0.34462565183639526, 0.10563202202320099, -0.153032124042511, 0.3154359757900238, 0.15342330932617188, -0.0027970680966973305, 0.062303561717271805, -0.1272721290588379, -0.406589150428772, -0.1481286734342575, 0.1709788590669632, 0.2773889899253845, -0.3150274157524109, 0.09033000469207764, -0.26422280073165894, 0.03734511509537697, 0.12562447786331177, 0.006544988602399826, -0.17453692853450775, 0.18298523128032684, -0.31146565079689026, -0.2189500480890274, -0.13686034083366394, 0.46691977977752686, 0.9032633900642395, 0.25674012303352356, 0.38866621255874634, 0.09151763468980789, -0.058752723038196564, -0.16813616454601288, 0.18907040357589722, 0.13167816400527954, 0.012559453025460243, 0.16735856235027313, -0.06914019584655762, -0.04874490946531296, -0.4017598628997803, 0.27937451004981995, 0.16855424642562866, 0.06223829463124275, -0.3027976155281067, -0.0377119742333889, -0.37835752964019775, -0.19744837284088135, -0.5116117000579834, -0.2571057379245758, -0.25512048602104187, -0.3293741047382355, -0.16789817810058594, 0.010062061250209808, 0.054954953491687775, 0.019709981977939606, -0.5319440364837646, 0.04521989822387695, -0.132942333817482, -0.1507660150527954, 0.12295158207416534, -0.022946922108530998, -0.2495671659708023, 0.0073659722693264484, -0.06586237996816635, 0.38983458280563354, 0.487861692905426, -0.07025717198848724, -0.042927082628011703, 0.05286649987101555, -0.4029851257801056, 0.23029516637325287, -0.24395576119422913, 0.23640793561935425, 0.24152077734470367, 0.12906870245933533, 0.29346179962158203, -0.002964173210784793, 0.15736213326454163, -0.18025986850261688, 0.024904798716306686, -0.04032787308096886, 0.10109657049179077, -0.10887996107339859, -0.2754931151866913, -0.8172608017921448, -0.3088201582431793, -0.1809062957763672, -0.2388209104537964, 0.1894153207540512, 0.16927355527877808, 0.35746341943740845, 0.02086687460541725, 0.1280636340379715, -0.32717782258987427, 0.05769907310605049, -0.36877596378326416, -0.15246862173080444, 0.4007764458656311, -0.2736347019672394, -0.37948542833328247, 0.10961063206195831, 0.14097368717193604, 0.03880582004785538, 0.14969384670257568, -0.2708304226398468, -0.05013031139969826, -0.007900197058916092, -0.5291980504989624, 0.1509532630443573, 0.33623045682907104, 0.29922425746917725, -0.33691951632499695, -0.11374465376138687, -0.13982728123664856, -0.06552579998970032, 0.11963070183992386, 0.24273639917373657, 0.5703086256980896, 0.13668492436408997, -0.06114596500992775, -0.1879299283027649, 0.47023576498031616, 0.33278053998947144, 0.21936587989330292, 0.13947369158267975, -0.08077383041381836, 0.5153935551643372, 0.07098837196826935, -0.25418931245803833, 0.17766234278678894, 0.03409856557846069, 0.11521975696086884, 0.4011799395084381, 0.14756064116954803, -0.3179175853729248, -0.2271198183298111, 0.02921314351260662, -0.3635047674179077, -0.29176679253578186, 0.226384237408638, -0.16411003470420837, 0.10192427784204483, 0.2606687843799591, -0.4112979769706726, -0.1092563346028328, -0.02179042063653469, -0.42039117217063904, 0.43069276213645935, -0.15378375351428986, 0.1531316339969635, -0.6548850536346436, 0.16735485196113586, -0.44770383834838867, 0.11611583828926086, -0.224205881357193, 0.17975763976573944, 0.013182733207941055, -0.10940742492675781, 0.2008356899023056, 0.07203809171915054, 0.37853220105171204, -0.09238993376493454, -0.2777591645717621, 0.16358676552772522, 0.1110323965549469, -0.6928815841674805, -0.054513685405254364, 0.035483770072460175, 0.06715433299541473, 0.07046375423669815, 0.1560639590024948, -0.3221653699874878, -0.21822842955589294, 0.28061267733573914, 0.12941709160804749, -0.20678310096263885, -0.23308981955051422, -0.3231203258037567, -0.5719006061553955, -0.2354198694229126, -0.4505119323730469, -0.14295253157615662, -0.04531512409448624, -0.1363145112991333, -0.09380354732275009, 0.21509858965873718, -0.1203148290514946, 0.2540307939052582, -0.06086577847599983, 0.26602494716644287, -0.037856295704841614, -0.033680010586977005, 0.4772064685821533, 0.09539500623941422, -0.2791562080383301, 0.42964208126068115, -0.14281490445137024, 0.20302043855190277, 0.13304810225963593, -0.37200695276260376, 0.4784090220928192, 0.3028554320335388, -0.10998047143220901, -0.0059056575410068035, 0.19600848853588104, -0.2528354227542877, -0.00895637460052967, 0.1406579464673996, 0.4021000862121582, -0.3468552827835083, -0.32030221819877625, -0.5828097462654114, 0.5492744445800781, 0.0031259539537131786, -0.3688296377658844, 0.2551991641521454, -0.010728285647928715, -0.3315703570842743, 0.2326078712940216, -0.22382216155529022, 1.152660608291626, 0.1274232715368271, 0.32659414410591125, -0.0009181576315313578, -0.10209809988737106, 0.8224447965621948, -0.05445834621787071, 0.08067619055509567, -0.10956183820962906, -0.04315541312098503, -0.13305160403251648, -0.20595534145832062, 0.1671735793352127, 0.12997110188007355, -0.09159845858812332, 0.30789637565612793, 0.1755504012107849, 0.0021263526286929846, 0.07282941788434982, 0.4751617908477783, 0.12683342397212982, 0.004292669706046581, -0.3501724898815155, 0.06220252066850662, 0.031312715262174606, 0.15649756789207458, 0.04731954261660576, -0.23766149580478668, -0.48491787910461426, -0.41060134768486023, -0.096663698554039, 0.09266920387744904, 0.3221283555030823, 0.07424508035182953, -0.07986201345920563, -0.19807782769203186, -0.2268858551979065, 0.20781375467777252, -0.008842568844556808, -0.05847968906164169, -0.3350391983985901, 0.1760788857936859, -0.14393191039562225, -0.2794754207134247, -0.20071382820606232, -0.16527825593948364, 0.2625402808189392, -0.25820502638816833, -0.2651193141937256, -0.01071022916585207, -0.07903069257736206, -0.280061811208725, -0.3210082948207855, 0.29311153292655945, 0.42605647444725037, 0.01068341638892889, -0.2837388515472412, 0.4091000258922577, -0.024119563400745392, -0.004243076778948307, 0.05624665692448616, 0.40378645062446594, 0.024780085310339928, 0.4783613085746765, 0.1997634619474411, -0.307781457901001, -0.11372822523117065, 0.3047422468662262, 0.7428837418556213, -0.0702206939458847, 0.3021068572998047, -0.09806791692972183, -0.011664065532386303, -0.10569637268781662, -0.10398522019386292, -0.209678053855896, -0.1769353151321411, 0.032819345593452454, 0.03127056732773781, 0.25124984979629517, 0.21842877566814423, 0.011708798818290234, -0.01157316379249096, -0.18024525046348572, -0.18839316070079803, -0.46140462160110474, -0.13243597745895386, 0.0887051448225975, -0.3724980652332306, 0.4005109369754791, 0.07215944677591324, -0.1330244094133377, 0.3243177831172943, 0.1654852032661438, -0.20654752850532532, 0.15494374930858612, -0.03597881644964218, 0.2704535722732544, 0.4610445499420166, 0.11127429455518723, 0.055707577615976334, -0.2033534049987793, -0.06352432072162628, -0.19280755519866943, -0.11600182205438614, -0.10259058326482773, -0.27869701385498047, 0.22386643290519714, 0.1265096217393875, 0.04898814111948013, 0.013911798596382141, -0.23256485164165497, -0.19723458588123322, 0.10693588852882385, 0.012789405882358551, -0.09102213382720947, 0.04635263979434967, 0.12139232456684113, -0.13883914053440094, 0.014663845300674438, 0.04517364501953125, 0.3172227144241333, -0.28538015484809875, 0.11752543598413467, 0.3289766013622284, 0.11331646144390106, -0.2091711312532425, -0.13677534461021423, -0.26757189631462097, -0.014869807288050652, 0.12046253681182861, -0.17886066436767578, 0.30952656269073486, 0.023659367114305496, 0.06650545448064804, -0.21033519506454468, 0.23484085500240326, 0.7377378344535828, -0.26888683438301086, 0.1399058848619461, -0.0722811222076416, 0.026590991765260696, -0.2600739002227783, -0.035654641687870026, 0.14888791739940643, 0.2221624255180359, -0.16084358096122742, 0.15752698481082916, 0.14933452010154724, -0.07433696836233139, -0.18297098577022552, 0.010846203193068504, 0.7443419098854065, -0.16752232611179352, 0.10861494392156601, -0.13933522999286652, -0.006556110456585884, 0.04117744043469429, -0.015504859387874603, 0.17375357449054718, 0.021766668185591698, 0.10402984917163849, 0.09909947216510773, 0.2958822250366211, 0.046929504722356796, -0.08133497089147568, 0.156587153673172, -0.196852907538414, -0.02389422059059143, 0.7082401514053345, 0.10019342601299286, 0.3292529881000519, 0.21454057097434998, 0.500723123550415, 0.046171341091394424, 0.08842477947473526, -0.1524607092142105, 0.45210152864456177, 0.3075077533721924, -0.12500490248203278, -0.1658221185207367, -0.27666109800338745, -0.06631572544574738, -0.10191535204648972, 0.004147587809711695, -0.27595049142837524, 0.1621083766222, 0.013707409612834454, -0.2040812373161316, -0.6190454363822937, 0.3473253548145294, 0.02899208478629589, -0.08817186206579208, -0.15126633644104004, 0.05593828856945038, -0.051575884222984314, -0.10623007267713547, 0.36736711859703064, 0.33350250124931335, 0.2425776869058609, 0.293914258480072, 0.06989080458879471, -0.1643189638853073, 0.35103243589401245, 0.030839629471302032, -0.18330958485603333, 0.19911235570907593, 0.3185026943683624, 0.09088269621133804, 0.3007444143295288, 0.10069771856069565, -0.0364370159804821, 0.03841133415699005, 0.2743447422981262, 0.0016559138894081116, -0.2547132968902588, 0.4916631579399109, -0.1473436802625656, 0.06509997695684433, -0.011430059559643269, 0.1163802444934845, -0.6328656077384949, -0.22419321537017822, 0.29300522804260254, 0.09583326429128647, 0.11507991701364517, -0.34181585907936096, 0.01429931540042162, 0.09833113104104996, 0.633484959602356, -0.07817039638757706, 0.12740159034729004, -0.1227286159992218, -0.3408302962779999, -0.6841957569122314, 0.2265513837337494, -0.11939598619937897, 0.138151615858078, 0.0011201014276593924, 0.2716940641403198, 0.28507840633392334, 0.244632288813591, 0.08168921619653702, 0.05564681068062782, -0.6260476112365723, 0.01900019869208336, -0.04369782656431198, -0.22733040153980255, -0.27663105726242065, 0.2152407020330429, -0.08741721510887146, -0.039485298097133636, 0.10128537565469742, -0.35122138261795044, -0.028956789523363113, -0.07868756353855133, -0.08384835720062256, 0.29225456714630127, 0.1682388037443161, 0.5404804944992065, 0.11159100383520126, 0.5250939130783081, -0.18571801483631134, -0.2845458388328552, -0.3882056772708893, 0.11773787438869476, -0.3753584921360016, 0.10256745666265488, -0.002077273325994611, 0.06304693967103958, -0.4657909870147705, 0.32165253162384033, -0.07286246865987778, 0.013991961255669594, -0.04858586564660072, -0.22922465205192566, -0.0022873482666909695, -0.051074832677841187, 0.07794680446386337, 0.10831771790981293, 0.18994639813899994, 0.29205790162086487, -0.20192839205265045, 0.07437717169523239, -0.18357278406620026, -0.13972824811935425, 0.3861722946166992, -0.20822566747665405, -0.19402636587619781, 0.0987410694360733, 0.018719885498285294, -0.17191502451896667, 0.1618819385766983, -0.7090769410133362, -0.06251835078001022, 0.3703852593898773, -0.08114278316497803, -0.189642071723938, 0.10166220366954803, -0.0853828564286232, -0.3409070074558258, -0.10702265053987503, 0.207116037607193, 0.011787292547523975, -0.27496466040611267, -0.19342859089374542, -0.18986108899116516 ]
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.
[ -0.10731988400220871, -0.08736063539981842, 0.05800749361515045, 0.1967514604330063, 0.21142952144145966, 0.09753541648387909, 0.2802221477031708, 0.3764084577560425, -0.15019190311431885, 0.0037060044705867767, -0.06849070638418198, 0.5348145961761475, -0.11443071812391281, 0.0036036840174347162, 0.32549768686294556, -0.17846287786960602, -0.03447540104389191, 0.2562464475631714, -0.06642461568117142, -0.06997606158256531, -0.04355191811919212, -0.02456844598054886, -0.09649717807769775, -0.026774737983942032, 0.0568176731467247, 0.01807527430355549, 0.08552531152963638, -0.14649748802185059, 0.4081133008003235, 0.04641716927289963, 0.3898698091506958, 0.023344850167632103, -0.033621158450841904, 0.1560439169406891, -0.000114742862933781, 0.40614011883735657, 0.03541814535856247, -0.19170303642749786, 0.07188375294208527, -0.1159353256225586, 0.16804367303848267, -0.16358761489391327, 0.16523264348506927, -0.21809548139572144, -0.05620704963803291, 0.02299475111067295, 0.21298527717590332, -0.06016510725021362, 0.2803553640842438, 0.13797207176685333, 0.09116262942552567, 0.5434459447860718, -0.17063172161579132, 0.2633916437625885, 0.14321106672286987, -0.15265358984470367, -0.23502607643604279, -0.25438013672828674, -0.19450947642326355, 0.020992117002606392, 0.11463578790426254, 0.08944632858037949, -0.2371271699666977, 0.12987150251865387, 0.4616265296936035, 0.16119210422039032, -0.06941099464893341, -0.36647161841392517, -0.20929747819900513, 0.42911550402641296, 0.26164674758911133, -0.06549941748380661, -0.17502018809318542, -0.44658032059669495, -0.23368871212005615, -0.3110027313232422, -0.10490407049655914, 0.12665331363677979, -0.28064703941345215, 0.22548305988311768, -0.09822195023298264, -0.08737321943044662, -0.25664618611335754, 0.17651842534542084, 0.19647064805030823, -0.04481394961476326, 0.12743723392486572, 0.11077618598937988, 0.20774655044078827, 0.1545112580060959, 0.4083368182182312, 0.021357299759984016, 0.3032936453819275, 0.40228071808815, -0.2894180715084076, 0.02911565825343132, 0.1779962033033371, -0.5676333904266357, -0.06889713555574417, -0.23024223744869232, 0.3916996419429779, 0.16744358837604523, -0.3058200478553772, 0.25170308351516724, -0.010725905187427998, -0.09433718025684357, -0.4143867790699005, 0.1737426370382309, -0.1531388908624649, -0.515454888343811, 0.05342930555343628, 0.04493878036737442, -0.2540418207645416, -0.010109368711709976, 0.022336699068546295, -0.5388681292533875, -0.17005738615989685, 0.22495561838150024, -0.31306639313697815, -0.42528626322746277, -0.2523905336856842, -0.0495489165186882, 0.0026699863374233246, 0.2106463611125946, -0.12489945441484451, 0.5658942461013794, 0.14857420325279236, -0.13732144236564636, -0.3045380413532257, -0.04863616079092026, -0.19958896934986115, 0.08693216741085052, -0.06759434193372726, -0.08612379431724548, 0.13496658205986023, -0.17564798891544342, 0.17783553898334503, -0.15483304858207703, -0.16657812893390656, 0.1368260383605957, 0.29989466071128845, -0.10216787457466125, 0.09434036910533905, 0.24431225657463074, 0.046297840774059296, -0.05605976656079292, 0.04743880406022072, 0.07257149368524551, -0.21072910726070404, 0.36221182346343994, -0.5235639214515686, -0.3000706434249878, -0.04146771505475044, 0.09681680053472519, 0.026772506535053253, -0.19203858077526093, -0.3231532871723175, 0.2882842719554901, 0.1580442190170288, -0.13978281617164612, -0.17260849475860596, -0.1775745004415512, -0.4339958131313324, -0.2991337478160858, 0.3465137183666229, 0.0750119611620903, -0.5027420520782471, -0.12278882414102554, -0.01291547529399395, -0.0579272136092186, 0.2622867226600647, 0.42779237031936646, -0.22205692529678345, 0.5380483269691467, -0.0876772329211235, -0.02747095562517643, 0.8010430932044983, -0.07477036863565445, -0.17104285955429077, 0.24721184372901917, -0.11224551498889923, 0.16722844541072845, 0.028713002800941467, 0.38759464025497437, -0.0019189761951565742, -0.030818725004792213, 0.38124775886535645, 0.6379417181015015, -0.15168501436710358, 0.025483405217528343, -0.10614652931690216, -0.3495664894580841, 0.2707488536834717, 0.24196384847164154, 0.10449320822954178, -0.030045511201024055, -0.23651078343391418, 0.3986964821815491, 0.13126195967197418, -0.02434096485376358, -0.19287654757499695, 0.1461459994316101, -0.1458994746208191, 0.025187602266669273, -0.18390017747879028, 0.10199640691280365, -0.6190454363822937, 0.23352797329425812, 0.1444821059703827, 0.08388828486204147, 0.23681889474391937, -0.0443686917424202, 0.355576753616333, 0.12261112779378891, 0.05143442377448082, 0.15910975635051727, -0.0466657318174839, 0.12642867863178253, -0.05838800221681595, -0.04362766817212105, -0.3507680296897888, 0.13079912960529327, -0.35051777958869934, -0.0940881073474884, -0.3933817148208618, 0.29612061381340027, 0.37826573848724365, -0.036910466849803925, -0.11612760275602341, 0.25752580165863037, -0.33689436316490173, -0.10080019384622574, 0.0596044659614563, 0.20393362641334534, 0.008492193184792995, 0.32520851492881775, -0.721428632736206, 0.47629207372665405, 0.05751020833849907, -0.01792527548968792, -0.08479506522417068, 0.13055172562599182, 0.009945613332092762, -0.2209625095129013, -0.018168216571211815, 0.2750522196292877, -0.23169927299022675, 0.19882583618164062, 0.28876686096191406, -0.2876957058906555, 0.1481616199016571, 0.07205011695623398, 0.1864931881427765, -0.11101437360048294, 0.47623971104621887, 0.17128826677799225, 0.17858880758285522, 0.32878074049949646, -0.5321707725524902, 0.20310622453689575, 0.6362370848655701, -0.015154444612562656, -0.14370980858802795, 0.17936615645885468, 0.0180216021835804, -0.2019813507795334, -0.036248672753572464, -0.07885139435529709, 0.248616561293602, 0.10557550191879272, 0.044687338173389435, -0.018213389441370964, -0.09769776463508606, -0.28492289781570435, 0.21820440888404846, 0.10028056800365448, 0.40321066975593567, -0.09976130723953247, 0.026069767773151398, 0.23116253316402435, -0.17332783341407776, 0.08780349045991898, 0.24988079071044922, 0.1755169928073883, -0.24388736486434937, 0.01786128617823124, -0.28857606649398804, -0.2792429029941559, -0.055799450725317, 0.08746697753667831, 0.13963881134986877, -0.16931763291358948, -0.09134818613529205, 0.3872908651828766, -0.009226835332810879, -0.08408626914024353, -0.14456328749656677, 0.2974510192871094, 0.14736495912075043, -0.4759063124656677, -0.10680648684501648, -0.06507289409637451, -0.21300117671489716, 0.03152339160442352, 0.37497028708457947, 0.19188150763511658, 0.22233065962791443, 0.008133403956890106, -0.13919630646705627, 0.18115463852882385, -0.08397325128316879, 0.24168133735656738, -0.03193846344947815, 0.13781821727752686, -0.046962738037109375, 0.11223586648702621, -0.0727236196398735, -0.2634597420692444, -0.06183741241693497, -0.4290964901447296, 0.039671700447797775, 0.1868332475423813, 0.017481550574302673, 0.16025768220424652, -0.2362191081047058, -0.1386372447013855, -0.2890470623970032, -0.39600276947021484, -0.018298454582691193, 0.1862499713897705, 0.3226667046546936, 0.40426385402679443, 0.003960215486586094, 0.25405967235565186, 0.39306309819221497, 0.06724479049444199, 0.08527905493974686, -0.155490443110466, 0.4933822751045227, -0.2766979932785034, -0.26682931184768677, -0.14344798028469086, -0.14479811489582062, 0.24469906091690063, 0.3684459924697876, -0.6950926780700684, 0.2113608419895172, -0.09673392027616501, 0.14602145552635193, -0.0984129011631012, -0.015865342691540718, 0.2280787080526352, -0.20837779343128204, 0.0687466561794281, -0.017936747521162033, 0.09904535114765167, 0.13033030927181244, 0.025880349799990654, -0.1436578333377838, 0.4220219552516937, 0.4722154140472412, 0.31678733229637146, 0.7827595472335815, -0.12301818281412125, -0.24631249904632568, 0.28384608030319214, -0.22802229225635529, 0.25997820496559143, -0.1918625831604004, -0.016254613175988197, 0.0662870779633522, 0.04867067188024521, -0.0657782033085823, 0.2300899475812912, -0.10858768224716187, -0.13981178402900696, -0.15558448433876038, 0.2244042307138443, 0.09555871039628983, -0.22741486132144928, 0.22539545595645905, -0.20803256332874298, 0.21618564426898956, 0.09425466507673264, 0.22961722314357758, -0.3376860022544861, -0.42883211374282837, -0.02579832822084427, -0.17720970511436462, 0.4478469789028168, -0.05506245419383049, -0.28194940090179443, 0.1320948600769043, -0.6937563419342041, 0.09473726153373718, 0.04768557474017143, 0.3188590407371521, 0.07766364514827728, -0.24230384826660156, 0.004208113066852093, 0.06520780920982361, 0.6028181314468384, 0.04434824734926224, -0.08510841429233551, 0.21417249739170074, -0.19752852618694305, -0.5001964569091797, 0.022523948922753334, -0.031512074172496796, -0.08884141594171524, 0.2648521661758423, 0.12026442587375641, -0.29280275106430054, -0.5042574405670166, -0.07943661510944366, -0.00020387825497891754, -0.359518826007843, -0.09357297420501709, -0.22388960421085358, -0.2616744041442871, -0.028193486854434013, -0.3826903998851776, 0.021888135001063347, 0.3010371923446655, -0.12246380001306534, 0.020172584801912308, -0.17321889102458954, 0.0016045033698901534, 0.2339184731245041, 0.011684582568705082, 0.14289595186710358, 0.40144088864326477, 0.040681835263967514, -0.061452701687812805, -0.03429970517754555, -0.098766028881073, 0.20590727031230927, -0.09806313365697861, -0.1612320989370346, 0.036570191383361816, 0.15343105792999268, 0.3691771924495697, 0.031787797808647156, 0.023263197392225266, 0.12791933119297028, -0.14821922779083252, 0.03830105811357498, -0.14445969462394714, 0.37864595651626587, 0.0908232033252716, -0.3453333377838135, -0.07255462557077408, -0.5526955127716064, 0.5314422249794006, 0.29880082607269287, 0.15192614495754242, 0.48851916193962097, -0.19536952674388885, -0.3557860851287842, 0.4916248023509979, 0.39687469601631165, 0.5634970664978027, -0.3789253234863281, 0.48827844858169556, 0.7028647661209106, 0.521899938583374, 0.9219176173210144, -0.3840237855911255, 0.2837420403957367, -0.1113036572933197, -0.3852222263813019, -0.07143665850162506, -0.2753489315509796, -0.32653072476387024, 0.1312573254108429, 0.13309474289417267, 0.2455316185951233, 0.1267225742340088, 0.2848149538040161, 0.2412848323583603, 0.35134461522102356, 0.35856330394744873, -0.4072932302951813, -0.21222913265228271, 0.013731845654547215, 0.0516495555639267, -0.14455975592136383, -0.11486591398715973, -0.20799042284488678, 0.21388472616672516, 0.05818144232034683, -0.014009397476911545, 0.1565965712070465, -0.2624652087688446, 0.21527305245399475, 0.1364258974790573, -0.13545139133930206, 0.06786574423313141, 0.259735643863678, 0.12310533970594406, -0.18345610797405243, -0.07790283858776093, -0.0785764530301094, 0.0610424242913723, -0.01166672445833683, 0.14572294056415558, -0.049384575337171555, 0.6279584765434265, -0.05613865703344345, 0.19453667104244232, 0.1661250740289688, -0.08871284872293472, -0.571548581123352, -0.13409848511219025, -0.14428319036960602, 0.07634435594081879, 0.18374982476234436, -0.5355895161628723, 0.28123271465301514, -0.28242072463035583, 0.20707984268665314, 0.051031243056058884, 0.1385137289762497, -0.22972042858600616, 0.039810288697481155, 0.0166217889636755, -0.19243475794792175, 0.2076532244682312, 0.035004425793886185, 0.0930202305316925, 0.031202545389533043, 0.37337324023246765, 0.22355760633945465, 0.02071712724864483, -0.017776887863874435, 0.08544126152992249, 0.28490740060806274, -0.07938235998153687, 0.08008474111557007, 0.026714501902461052, 0.20169778168201447, 0.07773033529520035, 0.37332209944725037, -0.18203893303871155, 0.06518011540174484, -0.2921450436115265, -0.16977253556251526, -0.2766077220439911, 0.3394336998462677, -0.08160261809825897, 0.12775245308876038, -0.049459218978881836, 0.21621191501617432, -0.10642249137163162, 0.24101096391677856, -0.18489305675029755, -0.053981754928827286, 0.002244934206828475, 0.043473780155181885, -0.14907652139663696, -0.017295945435762405, 0.0480046309530735, 0.11890894174575806, 0.027415649965405464, 0.12441633641719818, -0.32206061482429504, -0.1428016871213913, -0.02780252881348133, 0.2020147144794464, 0.051952071487903595, -0.25063198804855347, -0.023800672963261604, -0.37163615226745605, -0.331290066242218, -0.2035333812236786, 0.1122814267873764, -0.11092832684516907, -0.284444659948349, 0.02243898995220661, -0.07873168587684631, 0.3604108691215515, -0.2519090473651886, 0.061954595148563385, -0.08628309518098831, 0.34060949087142944, -0.09359622746706009, 0.25554773211479187, 0.150189608335495, -0.022436978295445442, -0.37363675236701965, -0.1217745840549469, 0.17380200326442719, -0.15421359241008759, 0.12140514701604843, -0.32857099175453186, -0.07203701883554459, -0.25759828090667725, 0.02463829144835472, 0.3140627443790436, -0.1965523213148117, 0.1291932761669159, 0.22686834633350372, 0.06306243687868118, -0.1053435355424881, -0.11044997721910477, 0.21441757678985596, 0.026152392849326134, -0.030349820852279663, 0.22932736575603485, 0.07449863106012344, 0.15259428322315216, -0.10686428844928741, -0.030716748908162117, 0.04046765714883804, -0.57243412733078, -0.18100449442863464, 0.40833017230033875, -0.07960908859968185, 0.06841672211885452, -0.17246674001216888, 0.1047242060303688, 0.13364660739898682, 0.13838419318199158, 0.3073664903640747, 0.17160239815711975, -0.10191676020622253, -0.08759192377328873, -0.04754035919904709, -0.37738287448883057, -0.45394644141197205, 0.48343291878700256, -0.04196381941437721, 0.5714561939239502, -0.26525604724884033, 0.001651304424740374, 0.11766768991947174, -0.16160377860069275, -0.17609842121601105, 0.284598171710968, -0.10150372982025146, -0.018113253638148308, -0.2557772397994995, 0.058280132710933685, -0.16195739805698395, 0.2735615074634552, -0.09264281392097473, -0.2304714173078537, -0.21142837405204773, 0.3268531858921051, 0.12374890595674515, 0.1698049157857895, 0.08803125470876694, 0.08871432393789291, 0.20860154926776886, -0.1412581503391266, 0.2405453324317932, -0.2891339659690857, -0.09723280370235443, -0.3500290811061859, 0.01534821093082428, 0.08870017528533936, 0.31457027792930603, -0.3028629422187805, 0.12886188924312592, 0.10812978446483612, -0.06667667627334595, -0.015008030459284782, 0.26009833812713623, 0.23029494285583496, 0.11595502495765686, 0.24547183513641357, -0.045211419463157654, -0.02587457001209259, 0.14597797393798828, 0.18008218705654144, 0.1694282442331314, -0.5887804627418518, 0.6027831435203552, 0.18034839630126953, -0.0968208909034729, -0.008341452106833458, 0.18996256589889526, -0.05924634635448456, -0.039797987788915634, 0.4897262156009674, -0.26842498779296875, 0.03991954028606415, 0.05689035356044769, 0.013276171870529652, -0.1862027496099472, 0.6324124336242676, 0.08844928443431854, 0.4917036294937134, -0.10871755331754684, 0.026488741859793663, -0.46973147988319397, 0.12125226855278015, -0.11352187395095825, 0.13867069780826569, 0.2599014639854431, 0.27580705285072327, 0.1513167768716812, 0.1794598400592804, -0.4360818564891815, -0.18453656136989594, 0.005466044880449772, 0.09761213511228561, -0.12733066082000732, -0.15381614863872528, -0.4934641420841217, 0.12027501314878464, -0.15698713064193726, -0.49851179122924805, 0.0545867383480072, -0.41390278935432434, -0.1042356863617897, -0.09382913261651993, 0.25318077206611633, 0.32836928963661194, 0.4639802873134613, 0.2814911901950836, -0.033511191606521606, -0.04955583065748215, 0.04295967519283295, -0.08725797384977341, -0.11083071678876877, -0.028888249769806862, -0.35535362362861633, -0.049638886004686356, -0.17677125334739685, 0.07457365840673447, -0.27969491481781006, -0.025637516751885414, -0.3028116226196289, -0.2440079301595688, 0.01799813099205494, -0.08510948717594147, -0.1251230239868164, 0.07936551421880722, -0.29182323813438416, 0.18006469309329987, 0.32479482889175415, 0.3962631821632385, -0.12905381619930267, -0.1300887167453766, -0.29265543818473816, 0.14886410534381866, 0.3681568205356598, -0.6422927975654602, -0.17142178118228912, 0.04930243641138077, -0.15187054872512817, 0.24562571942806244, 0.06413291394710541, -0.27930423617362976, -0.2004627287387848, 0.30683743953704834, -0.12597061693668365, -0.1965668499469757, 0.19951164722442627, 0.33595865964889526, -0.051681745797395706, -0.11604362726211548, 0.2768477499485016, -0.031304731965065, -0.461712509393692, -0.23161335289478302, -0.439833402633667 ]
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 ?
[ -0.16860957443714142, -0.09398823231458664, 0.011844201013445854, 0.16997693479061127, 0.2723703682422638, 0.0823572427034378, 0.22564619779586792, 0.3897305130958557, -0.05037659406661987, 0.0051495712250471115, -0.08119423687458038, 0.6080824732780457, -0.14292088150978088, -0.014412962831556797, 0.25082582235336304, -0.21652522683143616, -0.10149352252483368, 0.2596185505390167, -0.14962516725063324, -0.09359432756900787, -0.01441264059394598, -0.03481737896800041, -0.1413574516773224, -0.02752400003373623, -0.06012889742851257, -0.08392553776502609, 0.07913276553153992, -0.0956684798002243, 0.37118011713027954, 0.03904271125793457, 0.29161450266838074, 0.002584736095741391, -0.024420855566859245, 0.16511139273643494, -0.00010320142609998584, 0.3215746283531189, 0.03437270224094391, -0.14773303270339966, 0.04883978143334389, -0.07913138717412949, 0.16531749069690704, -0.2739705741405487, 0.1919618546962738, -0.30282336473464966, -0.0543791837990284, 0.008168940432369709, 0.19384202361106873, -0.02881588414311409, 0.31412672996520996, 0.15881864726543427, 0.216972216963768, 0.4294995963573456, -0.11771456897258759, 0.30170270800590515, 0.11722395569086075, -0.17708152532577515, -0.19004736840724945, -0.22906669974327087, -0.21755385398864746, -0.014429661445319653, 0.032906174659729004, 0.2663012444972992, -0.20410442352294922, 0.13051070272922516, 0.45549753308296204, 0.14307801425457, -0.1110716238617897, -0.37035757303237915, -0.2649401128292084, 0.3789658844470978, 0.22448164224624634, -0.08395049721002579, -0.15154582262039185, -0.4547223150730133, -0.1709524393081665, -0.30649852752685547, -0.15057675540447235, 0.11813399195671082, -0.26003098487854004, 0.12261297553777695, -0.10149877518415451, -0.08801303803920746, -0.23205602169036865, 0.144835963845253, 0.16639283299446106, -0.04291250929236412, 0.08226314932107925, 0.10815662890672684, 0.15611262619495392, 0.04901382327079773, 0.2461540699005127, 0.058890923857688904, 0.3580818176269531, 0.36975130438804626, -0.3404170870780945, 0.05369764566421509, 0.1779208779335022, -0.5166316628456116, -0.0021290637087076902, -0.2505265474319458, 0.3231509327888489, 0.17214269936084747, -0.2449929118156433, 0.18357093632221222, -0.07256817817687988, 0.012310515157878399, -0.4178171157836914, 0.11400648206472397, -0.11787008494138718, -0.48339197039604187, -0.023928068578243256, 0.021701842546463013, -0.23941364884376526, -0.04856526479125023, 0.1193634495139122, -0.4675556719303131, -0.033896468579769135, 0.23527880012989044, -0.37585684657096863, -0.3709524869918823, -0.19662554562091827, 0.056530874222517014, 0.013881944119930267, 0.2716267704963684, -0.1850641816854477, 0.4295888841152191, 0.1524079591035843, -0.10676465928554535, -0.33988791704177856, 0.004123414866626263, -0.2882518172264099, 0.12291796505451202, -0.128438800573349, -0.11231756955385208, 0.14382179081439972, -0.07857149094343185, 0.20242609083652496, -0.0808301791548729, -0.17189067602157593, 0.13403837382793427, 0.28055429458618164, -0.10108017921447754, 0.08383210748434067, 0.2150493860244751, 0.05136352404952049, -0.01314830593764782, 0.044644203037023544, 0.09676922112703323, -0.16510340571403503, 0.3000507652759552, -0.5144059062004089, -0.2434292435646057, -0.034086063504219055, 0.2013801783323288, 0.11621687561273575, -0.1493268460035324, -0.25425833463668823, 0.30588042736053467, 0.13064292073249817, -0.0581664964556694, -0.17962639033794403, -0.1691518872976303, -0.4182865023612976, -0.3279014825820923, 0.3185228109359741, 0.05843176320195198, -0.4493255615234375, -0.07593996077775955, -0.022894207388162613, -0.00007993614417500794, 0.28403398394584656, 0.41180551052093506, -0.20259392261505127, 0.47592025995254517, -0.06604327261447906, 0.06232507526874542, 0.7272696495056152, -0.07970738410949707, -0.2046176791191101, 0.24727140367031097, -0.18754437565803528, 0.1665133535861969, -0.05649469047784805, 0.38813701272010803, 0.014578812755644321, 0.006611598655581474, 0.38259807229042053, 0.6268606781959534, -0.13027746975421906, 0.12627564370632172, -0.16897965967655182, -0.2685701847076416, 0.30446282029151917, 0.29096198081970215, 0.05505363270640373, -0.09941761940717697, -0.2690315544605255, 0.3530549705028534, 0.14091584086418152, -0.04823535308241844, -0.17051982879638672, 0.1743096113204956, -0.10140743851661682, 0.03389355167746544, -0.1964699774980545, 0.023537157103419304, -0.6025986671447754, 0.2567143142223358, 0.21885152161121368, 0.11899276077747345, 0.14820291101932526, -0.09703923761844635, 0.2708936035633087, 0.11006155610084534, 0.03228209167718887, 0.15630649030208588, 0.11244998127222061, 0.11767581105232239, -0.01731007546186447, -0.010448210872709751, -0.2507621943950653, 0.2139793485403061, -0.37623223662376404, -0.1627490520477295, -0.3574135899543762, 0.20851126313209534, 0.3017750382423401, -0.03539026156067848, 0.004790205508470535, 0.2599482536315918, -0.2557673454284668, -0.11474455147981644, 0.05725487694144249, 0.2753027379512787, -0.028166890144348145, 0.23908764123916626, -0.626108705997467, 0.39822620153427124, -0.01640932448208332, -0.06816994398832321, -0.05593926087021828, 0.14532631635665894, 0.04535277187824249, -0.17168740928173065, 0.016372807323932648, 0.2203739434480667, -0.19826211035251617, 0.08164208382368088, 0.21227948367595673, -0.25049862265586853, 0.17031967639923096, 0.02143128775060177, 0.17781001329421997, -0.09024900197982788, 0.3996799886226654, 0.05171949788928032, 0.12820924818515778, 0.2929064929485321, -0.5027150511741638, 0.24715270102024078, 0.6455686092376709, 0.0017065361607819796, -0.060609910637140274, 0.21915046870708466, -0.004608092363923788, -0.1864921599626541, -0.11462308466434479, -0.07552865892648697, 0.2204630970954895, 0.2333492487668991, 0.004561266861855984, -0.04915609955787659, -0.1739027202129364, -0.23643603920936584, 0.22989507019519806, 0.10348115116357803, 0.39502039551734924, -0.09640945494174957, 0.04655268415808678, 0.15590441226959229, -0.21102255582809448, 0.0506058931350708, 0.22332820296287537, 0.28673675656318665, -0.20513223111629486, -0.03855021670460701, -0.2432170957326889, -0.2629227638244629, -0.07050161063671112, 0.01839360222220421, 0.1004713848233223, -0.17765048146247864, -0.06376805901527405, 0.32688209414482117, -0.005278074182569981, -0.002292812569066882, -0.1070261001586914, 0.26609373092651367, 0.2539463937282562, -0.3675926923751831, -0.04812689498066902, -0.09654650092124939, -0.2597522437572479, 0.12326271086931229, 0.34578055143356323, 0.26507648825645447, 0.3060997724533081, 0.042402178049087524, -0.029359707608819008, 0.15714679658412933, -0.02930849976837635, 0.16931909322738647, -0.18968899548053741, 0.1289825439453125, -0.07206515967845917, 0.20253276824951172, -0.10920311510562897, -0.21997807919979095, -0.040136486291885376, -0.4229848086833954, 0.04551754146814346, 0.1925015151500702, 0.016064509749412537, 0.053125862032175064, -0.22644731402397156, -0.22007007896900177, -0.2942711114883423, -0.42354458570480347, -0.011317764408886433, 0.11504679173231125, 0.28670817613601685, 0.4668727219104767, -0.03242725506424904, 0.2744038999080658, 0.3595525920391083, 0.13477635383605957, 0.021002212539315224, -0.13184291124343872, 0.5142493844032288, -0.2914407551288605, -0.29239025712013245, -0.1259390115737915, -0.1577879786491394, 0.21387840807437897, 0.26455336809158325, -0.6998733878135681, 0.16331246495246887, -0.1801672726869583, 0.09063276648521423, -0.17192858457565308, -0.023034099489450455, 0.2796553075313568, -0.1828378140926361, -0.037739481776952744, -0.06421389430761337, 0.1282205730676651, 0.15930896997451782, -0.0026169673074036837, -0.1744406372308731, 0.2911267578601837, 0.43631643056869507, 0.2932071089744568, 0.6823475360870361, -0.04952193424105644, -0.24725832045078278, 0.2451164871454239, -0.2034217119216919, 0.17989708483219147, -0.15880480408668518, -0.1297353357076645, 0.06918282806873322, 0.0468607172369957, -0.019307158887386322, 0.2627427279949188, -0.07913024723529816, -0.19225265085697174, -0.08546389639377594, 0.19333995878696442, 0.02322886511683464, -0.24505116045475006, 0.2765367925167084, -0.19784554839134216, 0.1782020479440689, 0.10248082131147385, 0.1724614053964615, -0.38271021842956543, -0.43289703130722046, -0.028190594166517258, -0.21834078431129456, 0.3675694763660431, -0.05496892333030701, -0.35252678394317627, 0.12331684678792953, -0.6885600686073303, 0.07143351435661316, 0.09516598284244537, 0.2950817942619324, 0.04400752857327461, -0.2644680142402649, 0.027435747906565666, 0.06361381709575653, 0.5770883560180664, 0.08534589409828186, -0.03744039684534073, 0.19421915709972382, -0.060621242970228195, -0.5440307259559631, 0.027683285996317863, -0.08388607203960419, -0.02656659670174122, 0.25266405940055847, 0.13720040023326874, -0.3163083493709564, -0.5172994136810303, -0.05490101873874664, 0.011884232051670551, -0.3238771855831146, -0.10303252190351486, -0.25590819120407104, -0.2436225712299347, -0.0124460868537426, -0.2928731143474579, 0.06926300376653671, 0.2637307643890381, -0.11182887852191925, -0.007645932491868734, -0.105202317237854, -0.0017555290833115578, 0.22647637128829956, 0.021024281159043312, 0.2179868072271347, 0.497418612241745, 0.0919596403837204, -0.006784555967897177, 0.044760141521692276, -0.17461439967155457, 0.21001878380775452, -0.10321377962827682, -0.1212225928902626, -0.002921516075730324, 0.11823973059654236, 0.327433705329895, 0.043706733733415604, 0.04576566815376282, 0.08590240776538849, -0.21129553020000458, 0.11721080541610718, -0.21579007804393768, 0.4107755422592163, 0.1617114394903183, -0.28921428322792053, -0.1019580140709877, -0.4516589641571045, 0.48958835005760193, 0.22257591784000397, 0.1089932769536972, 0.4912770986557007, -0.17228460311889648, -0.36280912160873413, 0.43914857506752014, 0.33841997385025024, 0.5404882431030273, -0.321880966424942, 0.4336530268192291, 0.7526196241378784, 0.4858792722225189, 0.8558775186538696, -0.35158655047416687, 0.2858251929283142, -0.15107744932174683, -0.33993232250213623, 0.006259836722165346, -0.22546714544296265, -0.40141764283180237, 0.06774861365556717, 0.06809143722057343, 0.18563054502010345, 0.086480051279068, 0.22489555180072784, 0.1757575422525406, 0.39436423778533936, 0.3600969910621643, -0.4310416281223297, -0.21452510356903076, 0.14980153739452362, -0.03654584288597107, -0.15512952208518982, -0.05978812649846077, -0.20871582627296448, 0.18185420334339142, 0.06399622559547424, 0.022014357149600983, 0.15056121349334717, -0.18615038692951202, 0.2835209369659424, 0.05520603433251381, -0.06650618463754654, -0.013074575923383236, 0.26324015855789185, 0.1220790445804596, -0.17733289301395416, -0.0984831377863884, -0.0576421283185482, 0.020041273906826973, 0.0036127734929323196, 0.08650193363428116, -0.05334319546818733, 0.6400722861289978, -0.049178410321474075, 0.13697375357151031, 0.10900628566741943, -0.07215595245361328, -0.4897186756134033, -0.19590575993061066, -0.11551910638809204, 0.08588857203722, 0.08319614827632904, -0.547953724861145, 0.23642680048942566, -0.20112012326717377, 0.18663036823272705, 0.1621534824371338, 0.12269821017980576, -0.19497360289096832, 0.09241051971912384, 0.09988823533058167, -0.21338358521461487, 0.22173069417476654, 0.035102181136608124, 0.10077410191297531, 0.09124848991632462, 0.3558182716369629, 0.22108884155750275, -0.025953661650419235, -0.07953108102083206, 0.08418352901935577, 0.3034348785877228, -0.06420759111642838, 0.03545660153031349, 0.047307055443525314, 0.28831589221954346, 0.07017546147108078, 0.3478032350540161, -0.22790847718715668, 0.014587398618459702, -0.19920125603675842, -0.18210968375205994, -0.30683794617652893, 0.33475157618522644, -0.067508764564991, 0.1276506930589676, -0.025996357202529907, 0.21559038758277893, -0.07443256676197052, 0.25938668847084045, -0.3228646516799927, -0.042934004217386246, -0.022495346143841743, 0.07561054825782776, -0.10007314383983612, -0.12442083656787872, 0.09797751158475876, 0.09842246025800705, 0.14724938571453094, 0.12966026365756989, -0.35898712277412415, -0.27380213141441345, -0.02206338942050934, 0.11938446015119553, 0.021814612671732903, -0.2305418848991394, -0.024569008499383926, -0.38802221417427063, -0.3245273530483246, -0.17464719712734222, 0.10038825124502182, -0.14644455909729004, -0.29161927103996277, 0.05987212061882019, -0.11217344552278519, 0.27732664346694946, -0.09360374510288239, 0.008463237434625626, -0.17180176079273224, 0.2840941250324249, -0.0625641867518425, 0.19119977951049805, 0.13741609454154968, -0.0796872228384018, -0.38825127482414246, -0.13485877215862274, 0.1281261444091797, -0.10049577802419662, -0.0008421611273661256, -0.3645165264606476, -0.08045906573534012, -0.20343458652496338, 0.09064098447561264, 0.24004030227661133, -0.25075528025627136, 0.03492062911391258, 0.21327032148838043, 0.201282799243927, -0.12772291898727417, -0.10396608710289001, 0.19983384013175964, 0.023196768015623093, 0.06183882802724838, 0.12107592076063156, -0.02108730562031269, 0.1645386517047882, -0.10702891647815704, 0.015114840120077133, 0.051866564899683, -0.5421465635299683, -0.2646966576576233, 0.3222053647041321, -0.15680630505084991, -0.034058015793561935, -0.08664074540138245, 0.135659322142601, 0.1247236430644989, 0.25816550850868225, 0.3180684447288513, 0.23370856046676636, -0.0799102857708931, -0.07542449235916138, -0.10472945868968964, -0.3642435371875763, -0.4094370901584625, 0.4181024134159088, -0.13467049598693848, 0.5021534562110901, -0.2514813244342804, 0.027214817702770233, 0.1200004443526268, -0.2684727609157562, -0.17018385231494904, 0.30529212951660156, -0.21323718130588531, -0.011489173397421837, -0.24748314917087555, 0.0629420280456543, -0.11182481050491333, 0.2272559106349945, -0.06825123727321625, -0.19298402965068817, -0.18089807033538818, 0.290274441242218, 0.049313902854919434, 0.19488771259784698, 0.07147769629955292, 0.18252314627170563, 0.23883698880672455, -0.12766575813293457, 0.23899763822555542, -0.22714929282665253, -0.09096371382474899, -0.4048546552658081, 0.02497956156730652, 0.10322316735982895, 0.3352166712284088, -0.33990994095802307, 0.12908807396888733, 0.09992950409650803, -0.10254407674074173, -0.06795092672109604, 0.23784373700618744, 0.207545667886734, 0.12218790501356125, 0.2946440279483795, 0.07240206748247147, -0.1298852115869522, 0.14953747391700745, 0.19643741846084595, 0.17680421471595764, -0.4668307602405548, 0.49007320404052734, 0.19497525691986084, -0.020242813974618912, -0.10694972425699234, 0.14637663960456848, -0.1831081211566925, -0.08909329771995544, 0.5219667553901672, -0.24887995421886444, 0.04972748085856438, -0.08820932358503342, 0.07930807024240494, -0.22433112561702728, 0.6939451694488525, 0.03303081542253494, 0.4624566435813904, -0.1568175107240677, 0.01007481012493372, -0.4859359562397003, 0.0970560684800148, -0.19249561429023743, 0.15192370116710663, 0.24726273119449615, 0.33283013105392456, 0.1067349836230278, 0.18085795640945435, -0.43699437379837036, -0.18489599227905273, -0.05393601953983307, 0.11275482922792435, -0.20269320905208588, -0.06505995243787766, -0.5096167325973511, 0.11405155807733536, -0.07577052712440491, -0.47097811102867126, 0.11253917962312698, -0.4334464371204376, 0.05631051957607269, -0.09103845059871674, 0.27836522459983826, 0.2807081341743469, 0.4136729836463928, 0.333699494600296, -0.04148993268609047, -0.04722534492611885, -0.019791943952441216, -0.10569287836551666, -0.12665529549121857, 0.023457415401935577, -0.34460684657096863, -0.07571852207183838, -0.050206709653139114, 0.21800008416175842, -0.23613223433494568, -0.03747405484318733, -0.23950152099132538, -0.2059682160615921, 0.09328202158212662, -0.11421600729227066, -0.0988612249493599, 0.11413241177797318, -0.17164140939712524, 0.18321608006954193, 0.18438182771205902, 0.4326106905937195, -0.09274686127901077, -0.05656084790825844, -0.2720027565956116, -0.03757445886731148, 0.4600411057472229, -0.5970886945724487, -0.23472289741039276, 0.09746073186397552, -0.060075659304857254, 0.22572438418865204, 0.04289362579584122, -0.2987770736217499, -0.15917867422103882, 0.2910027503967285, -0.13124872744083405, -0.2402859926223755, 0.2773435413837433, 0.3229193091392517, -0.08262724429368973, -0.08083812892436981, 0.290490061044693, -0.06068548187613487, -0.39346352219581604, -0.15428996086120605, -0.4548889100551605 ]
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']} ```
[ -0.12176644802093506, -0.025931626558303833, 0.04709145799279213, 0.18982039391994476, 0.19800393283367157, 0.15442752838134766, 0.24837562441825867, 0.45679691433906555, -0.037150491029024124, -0.05661475285887718, -0.1416860669851303, 0.6093761920928955, -0.1823762208223343, -0.031365081667900085, 0.24518415331840515, -0.16024309396743774, -0.05153336748480797, 0.21017979085445404, -0.0837811827659607, -0.05111927539110184, 0.041434094309806824, -0.10304634273052216, -0.11337645351886749, -0.08961852639913559, -0.08411609381437302, 0.03781071677803993, 0.026804119348526, -0.15933367609977722, 0.4063456952571869, -0.015848148614168167, 0.3480542004108429, 0.035480938851833344, -0.061999402940273285, 0.10191080719232559, -0.00011394677858334035, 0.38300806283950806, 0.03163425624370575, -0.1697196513414383, 0.005115835461765528, -0.048683248460292816, 0.18722331523895264, -0.22094638645648956, 0.2476184070110321, -0.2797372043132782, 0.02296026051044464, -0.043306559324264526, 0.22824789583683014, 0.02149231918156147, 0.27410852909088135, 0.0563444159924984, 0.11358039826154709, 0.4625847637653351, -0.22616207599639893, 0.36655718088150024, 0.15347078442573547, -0.10670574754476547, -0.1382305920124054, -0.27631285786628723, -0.21094046533107758, 0.04131370410323143, 0.04629546403884888, 0.21486051380634308, -0.29671844840049744, 0.10010875761508942, 0.4883711338043213, 0.1346103847026825, -0.11587359011173248, -0.2901354730129242, -0.20416897535324097, 0.3537537753582001, 0.19012607634067535, -0.07160366326570511, -0.17686016857624054, -0.5691705942153931, -0.23419807851314545, -0.21981750428676605, -0.21410530805587769, 0.04874380677938461, -0.18753688037395477, 0.13866224884986877, -0.21671625971794128, -0.09149640053510666, -0.19212809205055237, 0.1696152538061142, 0.17300382256507874, 0.08846990019083023, 0.15939010679721832, 0.15273542702198029, 0.16368256509304047, 0.04027045518159866, 0.27683907747268677, -0.01118026115000248, 0.29415351152420044, 0.41327211260795593, -0.3397207260131836, -0.004970071837306023, 0.14997588098049164, -0.5529507994651794, -0.0025995017495006323, -0.21910414099693298, 0.3147839903831482, 0.2147974818944931, -0.2501295506954193, 0.1788455694913864, -0.006411454640328884, -0.06401266157627106, -0.39652758836746216, 0.14390504360198975, -0.17764893174171448, -0.5946311354637146, 0.0709008127450943, 0.02544749528169632, -0.2908797562122345, 0.09226299822330475, 0.05414511263370514, -0.5405895709991455, -0.090340755879879, 0.22742824256420135, -0.30823180079460144, -0.4565540850162506, -0.24953985214233398, 0.032221585512161255, 0.02825876511633396, 0.2659035623073578, -0.17248526215553284, 0.4949047863483429, 0.04680109769105911, -0.1260072886943817, -0.23579660058021545, 0.07146371155977249, -0.20548029243946075, 0.08548390120267868, -0.05327272787690163, -0.09084229916334152, 0.18941821157932281, -0.044649217277765274, 0.23619996011257172, -0.18066543340682983, -0.15809713304042816, 0.16095015406608582, 0.25847169756889343, -0.09254220128059387, 0.07163988053798676, 0.21768243610858917, -0.014462441205978394, 0.06174687668681145, 0.04735370725393295, 0.1133972704410553, -0.23381265997886658, 0.35670921206474304, -0.5510233640670776, -0.22753334045410156, -0.004840715788304806, 0.0849091112613678, 0.10352537035942078, -0.23398049175739288, -0.36409828066825867, 0.35490870475769043, 0.102419413626194, -0.12468791007995605, -0.14371711015701294, -0.148555189371109, -0.4110414981842041, -0.26339462399482727, 0.28865453600883484, 0.11818345636129379, -0.3723016679286957, -0.19962595403194427, 0.013992655090987682, 0.03928852081298828, 0.21835900843143463, 0.44404229521751404, -0.2287718504667282, 0.49103274941444397, -0.0685441642999649, 0.17658209800720215, 0.7466119527816772, -0.1683417707681656, -0.21639251708984375, 0.2665832042694092, -0.10920848697423935, 0.20277336239814758, 0.007078690454363823, 0.4875006675720215, 0.05561188608407974, -0.028244659304618835, 0.32784372568130493, 0.6070597767829895, -0.11004222184419632, 0.09091208875179291, -0.06263841688632965, -0.3540171682834625, 0.3833683133125305, 0.22360163927078247, 0.04943360388278961, -0.08908837288618088, -0.24513205885887146, 0.48383066058158875, 0.15792757272720337, 0.006073647178709507, -0.18220330774784088, 0.13027895987033844, -0.18597283959388733, 0.005775631871074438, -0.2210053652524948, 0.04396351799368858, -0.6262136697769165, 0.29450786113739014, 0.28980663418769836, 0.01980094239115715, 0.22485071420669556, -0.132051020860672, 0.30463236570358276, 0.08697158098220825, 0.05847356468439102, 0.15581567585468292, -0.036375612020492554, 0.15324527025222778, -0.008286317810416222, -0.05614304542541504, -0.339827299118042, 0.21357761323451996, -0.35561302304267883, -0.08592253178358078, -0.28733816742897034, 0.2686125636100769, 0.446281373500824, 0.006337815895676613, -0.1396053284406662, 0.25583577156066895, -0.23948097229003906, -0.09090786427259445, 0.09431412816047668, 0.2214578241109848, -0.037428222596645355, 0.3102302551269531, -0.6426636576652527, 0.5175511837005615, -0.00729947118088603, -0.05198518931865692, -0.038033995777368546, 0.15125678479671478, 0.040839239954948425, -0.22987127304077148, -0.045041099190711975, 0.20381379127502441, -0.18118436634540558, 0.2521125078201294, 0.2275700569152832, -0.29739847779273987, 0.08644036948680878, 0.09497062116861343, 0.1474914401769638, -0.1060410663485527, 0.381844162940979, 0.08208712935447693, 0.258959025144577, 0.2943732440471649, -0.42012202739715576, 0.20113754272460938, 0.5873521566390991, -0.033839039504528046, -0.08823426067829132, 0.2429153323173523, -0.08980494737625122, -0.21951763331890106, -0.04968225583434105, -0.15792885422706604, 0.23874789476394653, 0.15569522976875305, 0.04756355285644531, -0.01836024969816208, -0.08513352274894714, -0.18957819044589996, 0.23301167786121368, 0.1381789594888687, 0.3442911207675934, -0.10489454120397568, 0.07945506274700165, 0.24394969642162323, -0.20069317519664764, 0.12063653022050858, 0.20413270592689514, 0.16148772835731506, -0.18216626346111298, 0.020875010639429092, -0.27405843138694763, -0.32334423065185547, -0.06881655752658844, -0.015634089708328247, 0.11983805894851685, -0.17669232189655304, -0.08395644277334213, 0.35642561316490173, 0.038988146930933, -0.026908524334430695, -0.12038220465183258, 0.3014891445636749, 0.18296079337596893, -0.48521196842193604, -0.1409328132867813, -0.0722896009683609, -0.18762914836406708, 0.012351101264357567, 0.34143102169036865, 0.24763774871826172, 0.2139730155467987, 0.1156187355518341, -0.10640116780996323, 0.0933758094906807, -0.13714507222175598, 0.22248099744319916, -0.1411895602941513, 0.07441699504852295, -0.09138819575309753, 0.08587713539600372, -0.131709024310112, -0.23741334676742554, -0.13842520117759705, -0.40012267231941223, 0.015283937565982342, 0.17810261249542236, 0.04398313909769058, 0.14078165590763092, -0.1730545312166214, -0.05411168932914734, -0.31138134002685547, -0.3680551052093506, 0.025136630982160568, 0.16438302397727966, 0.26305514574050903, 0.4266565144062042, -0.09771596640348434, 0.28752461075782776, 0.28419381380081177, 0.0566006675362587, 0.08437542617321014, -0.05910993367433548, 0.5510889291763306, -0.21195611357688904, -0.20756617188453674, -0.19724524021148682, -0.1682773232460022, 0.12386776506900787, 0.2879941761493683, -0.7350271344184875, 0.25820624828338623, -0.1476232260465622, 0.05228978022933006, -0.1360730677843094, -0.043233875185251236, 0.30103906989097595, -0.24622772634029388, 0.0579821951687336, -0.019213825464248657, 0.04298124089837074, 0.07409311830997467, 0.01504160463809967, -0.21246129274368286, 0.2835453450679779, 0.4253314733505249, 0.35637781023979187, 0.8198502659797668, -0.08152434974908829, -0.19571200013160706, 0.2159963697195053, -0.24318550527095795, 0.2042168825864792, -0.2124854475259781, -0.07977620512247086, 0.07321709394454956, 0.01492096483707428, -0.01920689083635807, 0.3055554926395416, -0.12085457146167755, -0.2351139783859253, -0.06570639461278915, 0.16245457530021667, 0.07112406194210052, -0.1895555704832077, 0.253777414560318, -0.23828253149986267, 0.12286163121461868, 0.14369386434555054, 0.2271646410226822, -0.39819374680519104, -0.5145756006240845, -0.05292676016688347, -0.19155606627464294, 0.34394755959510803, -0.03879287466406822, -0.40476492047309875, 0.18230612576007843, -0.7144542932510376, 0.1373092234134674, 0.056387316435575485, 0.3346385955810547, 0.1076648160815239, -0.29960939288139343, 0.023291023448109627, 0.07969065010547638, 0.6119206547737122, 0.1648486703634262, -0.0709831565618515, 0.17361630499362946, -0.15104717016220093, -0.5380709171295166, 0.01287076622247696, -0.05156515911221504, 0.03713851049542427, 0.2749986946582794, 0.15970683097839355, -0.30154383182525635, -0.5077014565467834, -0.04973540082573891, 0.006743061356246471, -0.35283535718917847, -0.0637001097202301, -0.22928383946418762, -0.1595723181962967, -0.03656793385744095, -0.31081506609916687, 0.0701918751001358, 0.23305214941501617, -0.05486319214105606, 0.03534473851323128, -0.23602718114852905, -0.07313316315412521, 0.2959986925125122, -0.007531931158155203, 0.17190702259540558, 0.5554134845733643, 0.06869801878929138, -0.05782357603311539, 0.028072159737348557, -0.22288216650485992, 0.22437584400177002, -0.046551600098609924, -0.17790302634239197, 0.020541608333587646, 0.139253169298172, 0.42129504680633545, 0.09153363853693008, 0.042673323303461075, 0.177964985370636, -0.18196241557598114, 0.026482246816158295, -0.2325231283903122, 0.4516891837120056, 0.10564117878675461, -0.28788360953330994, -0.11276036500930786, -0.4171381890773773, 0.5051949620246887, 0.26663684844970703, 0.0977829322218895, 0.4054184556007385, -0.18455468118190765, -0.37959787249565125, 0.474712610244751, 0.36818772554397583, 0.5930815935134888, -0.3568764328956604, 0.5228298902511597, 0.6306701898574829, 0.5135153532028198, 0.9229068160057068, -0.36990681290626526, 0.25785955786705017, -0.10761423408985138, -0.2891254425048828, -0.034149471670389175, -0.2701140344142914, -0.3677405118942261, 0.0976247489452362, 0.14890345931053162, 0.2554054856300354, 0.05960751697421074, 0.32816022634506226, 0.22872532904148102, 0.4213240146636963, 0.4004003703594208, -0.4190787374973297, -0.13934600353240967, 0.04883721470832825, -0.00046816322719678283, -0.19989648461341858, -0.0864967331290245, -0.13067470490932465, 0.22034911811351776, -0.037847623229026794, -0.06312491744756699, 0.09271523356437683, -0.1913192719221115, 0.27331066131591797, 0.08860529959201813, -0.022878006100654602, -0.030652686953544617, 0.23596099019050598, 0.05160655453801155, -0.16254614293575287, -0.07398036867380142, -0.0782148689031601, 0.020682398229837418, 0.0128319738432765, 0.13242073357105255, -0.07321611046791077, 0.5777003765106201, -0.0037469242233783007, 0.23270916938781738, 0.20195789635181427, -0.07018756121397018, -0.6475853323936462, -0.24203364551067352, -0.07746285200119019, 0.16460146009922028, 0.12083671987056732, -0.5740077495574951, 0.27751532196998596, -0.21644049882888794, 0.20163752138614655, 0.04024413600564003, 0.05791091546416283, -0.18905194103717804, 0.05720207467675209, 0.08815035969018936, -0.22422665357589722, 0.23804381489753723, 0.07987496256828308, 0.08021479845046997, 0.004342293832451105, 0.4003140330314636, 0.09414904564619064, 0.02644476294517517, 0.0004956665798090398, 0.14066673815250397, 0.31899040937423706, -0.053173404186964035, -0.0355195589363575, -0.03209904953837395, 0.22842828929424286, -0.0022146953269839287, 0.3170289695262909, -0.22109930217266083, 0.04519497603178024, -0.3271239995956421, -0.22691553831100464, -0.2797817289829254, 0.40973997116088867, -0.03227902203798294, 0.17035353183746338, -0.029661212116479874, 0.23653832077980042, -0.21207363903522491, 0.2599020004272461, -0.19452981650829315, -0.0003527061780914664, 0.05574321374297142, 0.06459634006023407, -0.2063012570142746, -0.03718244284391403, 0.08591103553771973, 0.14745528995990753, 0.02978677675127983, 0.1458420306444168, -0.2863136827945709, -0.16016757488250732, -0.05519573390483856, 0.21042372286319733, 0.06889991462230682, -0.18370799720287323, -0.07681185007095337, -0.3510492444038391, -0.3054199516773224, -0.21098484098911285, 0.07133451104164124, -0.11096103489398956, -0.25265079736709595, -0.017708035185933113, -0.016698163002729416, 0.30331191420555115, -0.18731704354286194, 0.05895693600177765, -0.0388108529150486, 0.3205207288265228, -0.06789528578519821, 0.2631840109825134, 0.17212136089801788, -0.016811033710837364, -0.36097627878189087, -0.11104592680931091, 0.26176339387893677, -0.1292840987443924, -0.01932099089026451, -0.35425931215286255, -0.0280043613165617, -0.24651040136814117, 0.01943294331431389, 0.3817175030708313, -0.2598595917224884, 0.011901489458978176, 0.25529199838638306, 0.07606113702058792, -0.06930935382843018, -0.13709066808223724, 0.2898457646369934, 0.01564125530421734, -0.01702941581606865, 0.21574866771697998, 0.019055970013141632, 0.23018857836723328, -0.07639779150485992, -0.05945046991109848, 0.05092655122280121, -0.458130419254303, -0.24831117689609528, 0.34422624111175537, -0.14797507226467133, -0.04453127086162567, -0.15659978985786438, 0.15954947471618652, 0.15881989896297455, 0.20096929371356964, 0.2688859701156616, 0.272053599357605, -0.1309780776500702, -0.08041992038488388, -0.10907385498285294, -0.2988499402999878, -0.4159931540489197, 0.4332541823387146, -0.03095054440200329, 0.5543475151062012, -0.30145975947380066, -0.044290341436862946, 0.21874773502349854, -0.2889837622642517, -0.1426353007555008, 0.28796014189720154, -0.12157303094863892, -0.09826510399580002, -0.18458382785320282, 0.06264106929302216, -0.21620093286037445, 0.20930610597133636, -0.12552712857723236, -0.24214597046375275, -0.1884736865758896, 0.2828623056411743, 0.12130411714315414, 0.14187613129615784, 0.03304439038038254, 0.13925662636756897, 0.2624237537384033, -0.16844035685062408, 0.28532475233078003, -0.30386513471603394, -0.023454096168279648, -0.3080236613750458, 0.05243345722556114, 0.030013494193553925, 0.32330071926116943, -0.3477117121219635, 0.13549593091011047, 0.07323808968067169, -0.05230724811553955, -0.021849291399121284, 0.2548741102218628, 0.21410423517227173, 0.11187846213579178, 0.2443319857120514, -0.047419626265764236, -0.03806295990943909, 0.15545955300331116, 0.13566069304943085, 0.10434861481189728, -0.6335429549217224, 0.5223556756973267, 0.2589547038078308, -0.03485381230711937, 0.04822485148906708, 0.23610129952430725, -0.17642270028591156, -0.11756720393896103, 0.49291667342185974, -0.2662459909915924, 0.07922812551259995, 0.08962001651525497, 0.023243864998221397, -0.236037477850914, 0.6891622543334961, 0.015659349039196968, 0.5115364193916321, -0.12947972118854523, 0.08085674792528152, -0.48336195945739746, 0.046040140092372894, -0.13137909770011902, 0.1344796121120453, 0.24390119314193726, 0.2871461808681488, 0.15486115217208862, 0.21748219430446625, -0.37715384364128113, -0.29307132959365845, -0.07903537154197693, 0.005899378098547459, -0.12049271911382675, -0.09162639826536179, -0.4695657789707184, 0.20530226826667786, -0.06458189338445663, -0.48753708600997925, 0.053547684103250504, -0.31423696875572205, -0.10486916452646255, -0.06824928522109985, 0.20544999837875366, 0.27082765102386475, 0.3716529607772827, 0.32452279329299927, -0.059266578406095505, 0.0022979548666626215, 0.06226672977209091, -0.10009274631738663, -0.13290446996688843, -0.028408227488398552, -0.3497498035430908, -0.08344780653715134, -0.16258291900157928, 0.16817998886108398, -0.33418506383895874, 0.01335939485579729, -0.21613335609436035, -0.3234877288341522, 0.05041765421628952, -0.2010859102010727, -0.14042311906814575, 0.090306356549263, -0.3096466064453125, 0.31978508830070496, 0.13747629523277283, 0.4710559844970703, -0.1315186470746994, -0.038961563259363174, -0.19754301011562347, 0.12544260919094086, 0.47626370191574097, -0.6969991326332092, -0.1845293641090393, 0.08059531450271606, -0.13366729021072388, 0.2864728569984436, 0.08193147927522659, -0.34681081771850586, -0.2159189134836197, 0.36738553643226624, -0.06303628534078598, -0.2189219743013382, 0.23914101719856262, 0.3046948313713074, -0.1087450459599495, -0.1180870309472084, 0.30738383531570435, -0.08267620950937271, -0.36576804518699646, -0.1998542994260788, -0.5045342445373535 ]
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
[ -0.13054968416690826, -0.030859442427754402, 0.03630466386675835, 0.12274204939603806, 0.22812078893184662, 0.09177573025226593, 0.2730557322502136, 0.4065893888473511, -0.11111743003129959, 0.0027907316107302904, -0.08880873024463654, 0.5513769388198853, -0.15211926400661469, 0.047211214900016785, 0.3196074664592743, -0.2627655863761902, -0.06650406867265701, 0.2488582879304886, -0.1298067718744278, -0.014312619343400002, -0.04055102542042732, -0.05323069542646408, -0.09556213021278381, -0.016439583152532578, 0.015059716999530792, -0.03353499621152878, 0.05114983767271042, -0.15202964842319489, 0.41692981123924255, 0.021076951175928116, 0.3553370237350464, 0.0021882401779294014, -0.03409392759203911, 0.11227048188447952, -0.00010922730143647641, 0.38452470302581787, 0.03214536979794502, -0.14960572123527527, 0.06209931895136833, -0.07406581938266754, 0.20812633633613586, -0.22736458480358124, 0.1988503485918045, -0.26616284251213074, -0.04680865630507469, -0.005252201110124588, 0.23685550689697266, -0.03428574651479721, 0.2574250400066376, 0.11936886608600616, 0.15609537065029144, 0.5252829194068909, -0.17337894439697266, 0.2786279022693634, 0.13253091275691986, -0.15955960750579834, -0.21689729392528534, -0.27933043241500854, -0.13236485421657562, -0.010954257100820541, 0.0657459944486618, 0.14796705543994904, -0.22829413414001465, 0.09629999846220016, 0.43259772658348083, 0.1546454280614853, -0.11124934256076813, -0.3790855407714844, -0.2610156834125519, 0.4085840880870819, 0.20153634250164032, -0.07298098504543304, -0.17706932127475739, -0.4851958751678467, -0.17577533423900604, -0.29344362020492554, -0.17836521565914154, 0.08686396479606628, -0.2625953257083893, 0.18190978467464447, -0.10118099302053452, -0.12041549384593964, -0.21891945600509644, 0.12010437250137329, 0.21201230585575104, -0.03557480871677399, 0.16041907668113708, 0.1168626919388771, 0.22441312670707703, 0.10827701538801193, 0.349454402923584, 0.08066027611494064, 0.35486942529678345, 0.3675655126571655, -0.28405025601387024, -0.011491364799439907, 0.1692715883255005, -0.5806730389595032, -0.04443418234586716, -0.25259673595428467, 0.357106477022171, 0.1853385865688324, -0.29192596673965454, 0.2165098339319229, -0.013307726942002773, -0.07356644421815872, -0.42249467968940735, 0.13194596767425537, -0.13990892469882965, -0.5050945281982422, 0.04702818766236305, 0.046306267380714417, -0.22082184255123138, -0.019927917048335075, 0.06059923768043518, -0.5112511515617371, -0.14770333468914032, 0.2690739333629608, -0.2849746346473694, -0.4418404698371887, -0.23016899824142456, 0.02839222177863121, -0.014332803897559643, 0.17374639213085175, -0.1279168426990509, 0.5364342331886292, 0.13731837272644043, -0.12483685463666916, -0.2813413739204407, -0.029577868059277534, -0.23230735957622528, 0.11308996379375458, -0.06146422401070595, -0.09602812677621841, 0.15568773448467255, -0.1606728881597519, 0.17979362607002258, -0.12268879264593124, -0.12777180969715118, 0.1816205531358719, 0.2595547139644623, -0.08195066452026367, 0.10017409920692444, 0.25791484117507935, 0.01488631870597601, -0.06456943601369858, 0.038228463381528854, 0.0749143734574318, -0.21613091230392456, 0.35391825437545776, -0.5225857496261597, -0.2890230119228363, -0.04879701882600784, 0.13496695458889008, 0.08778105676174164, -0.21369022130966187, -0.3331678807735443, 0.326195627450943, 0.12096334993839264, -0.11928080022335052, -0.1852189153432846, -0.20750412344932556, -0.4238695800304413, -0.30116695165634155, 0.30410709977149963, 0.04064252972602844, -0.4613338112831116, -0.11523313075304031, 0.006508489139378071, -0.016593577340245247, 0.24931301176548004, 0.4456065595149994, -0.23294797539710999, 0.48019281029701233, -0.053436774760484695, 0.0012748086592182517, 0.7598669528961182, -0.08542151749134064, -0.16154102981090546, 0.22563156485557556, -0.12333734333515167, 0.20281203091144562, -0.004924946464598179, 0.405376136302948, 0.0003437797713559121, -0.029118018224835396, 0.3868637979030609, 0.6318567395210266, -0.15841254591941833, 0.07676416635513306, -0.10929054766893387, -0.34593266248703003, 0.3001142144203186, 0.2797212302684784, 0.09528956562280655, -0.0565822571516037, -0.2650609016418457, 0.39543265104293823, 0.13062497973442078, -0.034321967512369156, -0.18207554519176483, 0.06700465083122253, -0.12869006395339966, 0.00043369477498345077, -0.1707858294248581, 0.06576725095510483, -0.6022961139678955, 0.25331681966781616, 0.1912776231765747, 0.09608091413974762, 0.18939854204654694, -0.11190378665924072, 0.36263155937194824, 0.10908039659261703, 0.11826259642839432, 0.15904933214187622, 0.019094688817858696, 0.07548084855079651, -0.028345102444291115, -0.011725276708602905, -0.34432199597358704, 0.1589708775281906, -0.375441312789917, -0.12695889174938202, -0.34896770119667053, 0.26449695229530334, 0.3732421398162842, 0.017966771498322487, -0.08456075936555862, 0.3013322949409485, -0.36645761132240295, -0.12099935859441757, 0.11764267832040787, 0.21522918343544006, -0.030569283291697502, 0.2932139039039612, -0.667224109172821, 0.4424225389957428, 0.05159892141819, -0.022755660116672516, -0.10921429097652435, 0.12416110187768936, 0.0025713935028761625, -0.19744907319545746, 0.010840428061783314, 0.2805429697036743, -0.22586193680763245, 0.14711354672908783, 0.2610127627849579, -0.2442210465669632, 0.15271611511707306, 0.06381229311227798, 0.13426342606544495, -0.11270657181739807, 0.3706357777118683, 0.09925209730863571, 0.13777029514312744, 0.29518190026283264, -0.5292934775352478, 0.2312229424715042, 0.6706405878067017, -0.022198030725121498, -0.11366263031959534, 0.21687623858451843, 0.03733406215906143, -0.21183151006698608, -0.08162238448858261, -0.08116970211267471, 0.20193089544773102, 0.1457822173833847, 0.041635867208242416, 0.0004896327736787498, -0.164139524102211, -0.2583945095539093, 0.24762281775474548, 0.1311679482460022, 0.3808507025241852, -0.1395997703075409, 0.03004489466547966, 0.2035275399684906, -0.1915540099143982, 0.08323968201875687, 0.2106478363275528, 0.17547309398651123, -0.20561537146568298, 0.030216893181204796, -0.2745700478553772, -0.3066864013671875, -0.08983341604471207, 0.03501958027482033, 0.14405669271945953, -0.18368205428123474, -0.049924738705158234, 0.3294703960418701, -0.04942652955651283, -0.05361633375287056, -0.11888284236192703, 0.3149208724498749, 0.16494852304458618, -0.4449193775653839, -0.06582386791706085, -0.10550209879875183, -0.18507151305675507, 0.09022968262434006, 0.3607679307460785, 0.17744210362434387, 0.2429792433977127, 0.026393525302410126, -0.08753308653831482, 0.16812993586063385, -0.07798134535551071, 0.22266918420791626, -0.11164135485887527, 0.11009324342012405, -0.04960903897881508, 0.07181675732135773, -0.1287715882062912, -0.258893221616745, -0.11553193628787994, -0.4312824308872223, 0.05385315790772438, 0.15509124100208282, -0.002424989826977253, 0.14857859909534454, -0.24266725778579712, -0.1758737862110138, -0.30553820729255676, -0.3943731486797333, 0.018097609281539917, 0.16430498659610748, 0.3020918071269989, 0.42609959840774536, -0.029928933829069138, 0.28222182393074036, 0.38556817173957825, 0.05240766704082489, 0.08026599138975143, -0.1383151412010193, 0.4784577786922455, -0.27619487047195435, -0.2551717460155487, -0.15608848631381989, -0.17423506081104279, 0.1930471807718277, 0.31865304708480835, -0.7124727368354797, 0.14556312561035156, -0.08327455073595047, 0.08658715337514877, -0.0899091586470604, -0.04109234735369682, 0.23172472417354584, -0.20331773161888123, 0.020511100068688393, 0.004608738701790571, 0.16507463157176971, 0.14517609775066376, 0.02480635792016983, -0.20415599644184113, 0.3703915476799011, 0.4237150251865387, 0.33127135038375854, 0.7575737237930298, -0.08429167419672012, -0.21429255604743958, 0.23427152633666992, -0.18224991858005524, 0.19702783226966858, -0.23281574249267578, -0.06427574902772903, 0.042938992381095886, -0.013692792505025864, -0.005353040061891079, 0.22368399798870087, -0.10000862181186676, -0.19537688791751862, -0.11734012514352798, 0.22071106731891632, 0.11836304515600204, -0.23836751282215118, 0.25822973251342773, -0.21444033086299896, 0.18313203752040863, 0.11112025380134583, 0.21261003613471985, -0.32998502254486084, -0.3997202217578888, -0.01931042969226837, -0.23272429406642914, 0.41783589124679565, -0.04035947471857071, -0.3286675214767456, 0.10396495461463928, -0.7131767868995667, 0.06532417237758636, 0.11960608512163162, 0.3085826635360718, 0.08994664996862411, -0.29901474714279175, -0.006996363867074251, 0.06775638461112976, 0.5483503341674805, 0.09833700209856033, -0.08745644986629486, 0.2097495049238205, -0.17732758820056915, -0.48929446935653687, 0.03407828137278557, -0.05583784356713295, -0.07893278449773788, 0.24943765997886658, 0.14017464220523834, -0.2619260549545288, -0.4913126230239868, -0.024556659162044525, -0.01641920953989029, -0.3395271897315979, -0.04882008954882622, -0.1868024468421936, -0.24344764649868011, 0.0010783100733533502, -0.31565260887145996, 0.05761398375034332, 0.27913469076156616, -0.17108549177646637, 0.020696332678198814, -0.1508035957813263, -0.0014762248611077666, 0.2122334986925125, 0.0035505718551576138, 0.16318699717521667, 0.46328672766685486, 0.0473945215344429, -0.048913732171058655, -0.008087996393442154, -0.1308063119649887, 0.22405454516410828, -0.09222041815519333, -0.15226952731609344, 0.01875901408493519, 0.15853556990623474, 0.37199652194976807, 0.04845355823636055, 0.036094918847084045, 0.12852442264556885, -0.1741504669189453, 0.03312740474939346, -0.19866789877414703, 0.3670535683631897, 0.08676330000162125, -0.34173792600631714, -0.08246482908725739, -0.5109407901763916, 0.5615848302841187, 0.29596763849258423, 0.12883761525154114, 0.4830285906791687, -0.19704562425613403, -0.361695259809494, 0.5212910175323486, 0.3826305866241455, 0.5098997950553894, -0.37036898732185364, 0.5261973738670349, 0.7229401469230652, 0.47835925221443176, 0.923994243144989, -0.35906746983528137, 0.27109092473983765, -0.1238887831568718, -0.3586644232273102, -0.02879764884710312, -0.25565671920776367, -0.3750624358654022, 0.07888747751712799, 0.10311448574066162, 0.21673333644866943, 0.1310858279466629, 0.27752789855003357, 0.2179109901189804, 0.3121586740016937, 0.3466191291809082, -0.42222878336906433, -0.1941182017326355, 0.06759938597679138, 0.02967032976448536, -0.2340950220823288, -0.08185189962387085, -0.1964038610458374, 0.23074989020824432, 0.06485650688409805, 0.020356403663754463, 0.12005959451198578, -0.18482451140880585, 0.281927227973938, 0.1395895779132843, -0.09903062134981155, 0.04047880321741104, 0.24172845482826233, 0.12007835507392883, -0.2049071043729782, -0.1091783344745636, -0.08963549137115479, 0.06531930714845657, 0.01017042901366949, 0.14681874215602875, -0.044765107333660126, 0.6814283132553101, -0.021972212940454483, 0.19641365110874176, 0.1639900654554367, -0.07861166447401047, -0.5933406352996826, -0.18238712847232819, -0.11446233093738556, 0.08533694595098495, 0.1685638427734375, -0.5447580814361572, 0.2875628173351288, -0.1981663852930069, 0.1810290366411209, 0.09797929972410202, 0.11959954351186752, -0.20461012423038483, 0.05496637150645256, 0.055238112807273865, -0.18702736496925354, 0.22112692892551422, 0.029215414077043533, 0.1272468864917755, 0.034634072333574295, 0.35292571783065796, 0.2273520827293396, 0.004736424423754215, -0.025911390781402588, 0.10443616658449173, 0.2849634885787964, -0.04618540778756142, 0.06443586945533752, 0.03664526715874672, 0.22891297936439514, 0.06418907642364502, 0.35984182357788086, -0.22694505751132965, 0.031078645959496498, -0.2919079661369324, -0.17317940294742584, -0.30188632011413574, 0.36624348163604736, -0.059796594083309174, 0.12840865552425385, -0.0037668286822736263, 0.21896669268608093, -0.12371939420700073, 0.22889865934848785, -0.24168674647808075, -0.022742940112948418, 0.005892714485526085, 0.06760426610708237, -0.12601536512374878, -0.05863913521170616, 0.06964516639709473, 0.1309002786874771, 0.06438508629798889, 0.14221784472465515, -0.374716579914093, -0.20188695192337036, -0.04005305841565132, 0.16559568047523499, 0.032665740698575974, -0.2386941909790039, -0.00442203413695097, -0.3472810983657837, -0.3325616717338562, -0.16482393443584442, 0.1165752112865448, -0.125074103474617, -0.2715752124786377, 0.06031190603971481, -0.05946097522974014, 0.2978953421115875, -0.13933761417865753, 0.034537576138973236, -0.10700347274541855, 0.3132711350917816, -0.08346769958734512, 0.24981431663036346, 0.099190354347229, -0.05009348690509796, -0.3136666715145111, -0.11679372936487198, 0.13764715194702148, -0.11600009351968765, 0.0682290643453598, -0.3133593499660492, -0.0745454654097557, -0.24887824058532715, 0.04287295415997505, 0.2715887129306793, -0.23715412616729736, 0.09041192382574081, 0.24579183757305145, 0.1149819940328598, -0.09760556370019913, -0.15072190761566162, 0.20687198638916016, 0.08234644681215286, -0.024317177012562752, 0.20025289058685303, 0.08809415251016617, 0.16034311056137085, -0.06396133452653885, -0.040288325399160385, 0.01018658559769392, -0.5695467591285706, -0.1938687413930893, 0.4081522524356842, -0.13632282614707947, -0.014250031672418118, -0.16127094626426697, 0.09546259045600891, 0.1648010015487671, 0.21101625263690948, 0.27941858768463135, 0.22885385155677795, -0.07179965078830719, -0.09621987491846085, -0.06139996275305748, -0.3514270484447479, -0.4288907051086426, 0.4405209720134735, -0.061253923922777176, 0.5847461819648743, -0.2274625599384308, 0.035516053438186646, 0.14829818904399872, -0.21973007917404175, -0.16932722926139832, 0.29570356011390686, -0.14875158667564392, -0.06481295824050903, -0.23912525177001953, 0.06481734663248062, -0.15066871047019958, 0.29914548993110657, -0.10230916738510132, -0.21432219445705414, -0.1676589548587799, 0.2694036066532135, 0.09658154845237732, 0.22632619738578796, 0.1180020272731781, 0.07471587508916855, 0.24854056537151337, -0.1172582134604454, 0.19789014756679535, -0.26345086097717285, -0.1078857034444809, -0.3696136474609375, 0.011613118462264538, 0.09121473133563995, 0.2826962172985077, -0.33375421166419983, 0.12661008536815643, 0.07012615352869034, -0.045848995447158813, -0.020806381478905678, 0.24407020211219788, 0.23685352504253387, 0.09861885011196136, 0.30468037724494934, -0.0035824505612254143, -0.05681198835372925, 0.15154990553855896, 0.15705421566963196, 0.17538775503635406, -0.5788912773132324, 0.5717279314994812, 0.22617913782596588, -0.05557718500494957, -0.005985352676361799, 0.2032669186592102, -0.08927392214536667, -0.12076373398303986, 0.5214149355888367, -0.2620497941970825, 0.036717746406793594, -0.002508120145648718, 0.04671444743871689, -0.1930745244026184, 0.6902193427085876, 0.05742127075791359, 0.4553857147693634, -0.08694535493850708, 0.04564308747649193, -0.4419494569301605, 0.09158525615930557, -0.10719525814056396, 0.21197383105754852, 0.26062825322151184, 0.29831331968307495, 0.14649106562137604, 0.20516304671764374, -0.45577114820480347, -0.21373891830444336, -0.016207629814743996, 0.0483994297683239, -0.1488635241985321, -0.10379477590322495, -0.5044937133789062, 0.12454158067703247, -0.13039062917232513, -0.44239968061447144, 0.0877697765827179, -0.421377956867218, -0.03412601724267006, -0.08691316843032837, 0.22290672361850739, 0.3233673870563507, 0.44661396741867065, 0.33364853262901306, -0.03963198885321617, -0.10491995513439178, 0.049597542732954025, -0.07954303920269012, -0.11297344416379929, -0.027450688183307648, -0.3176705837249756, -0.0786447748541832, -0.1492995321750641, 0.12864680588245392, -0.29040277004241943, -0.04747023805975914, -0.22104139626026154, -0.262742817401886, 0.025212714448571205, -0.10299623012542725, -0.13584555685520172, 0.09924497455358505, -0.2605093717575073, 0.17538054287433624, 0.23030823469161987, 0.4477681815624237, -0.14069733023643494, -0.1299130767583847, -0.2153412252664566, 0.08180071413516998, 0.38260048627853394, -0.6043995022773743, -0.20278356969356537, 0.06831437349319458, -0.1658097803592682, 0.25549790263175964, 0.06379912793636322, -0.2777388095855713, -0.21997900307178497, 0.27987363934516907, -0.12841497361660004, -0.2145450860261917, 0.2562739849090576, 0.32339829206466675, -0.10076111555099487, -0.12213340401649475, 0.27809542417526245, -0.06114121899008751, -0.433061420917511, -0.21358846127986908, -0.486518532037735 ]
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)
[ -0.17388387024402618, -0.06410499662160873, -0.045787371695041656, 0.06958933919668198, -0.3965015113353729, 0.13637568056583405, 0.06761050224304199, 0.40038397908210754, -0.068243607878685, -0.32224205136299133, 0.13134290277957916, 0.18338929116725922, -0.08250488340854645, 0.07896018773317337, 0.39143097400665283, -0.04079137369990349, -0.05888555571436882, 0.2828741669654846, -0.03286243602633476, -0.03655635938048363, 0.16092759370803833, -0.021287214010953903, -0.10444439947605133, 0.007256662007421255, -0.19171863794326782, -0.3496658504009247, -0.16615340113639832, 0.15690942108631134, -0.14719173312187195, -0.4131219685077667, 0.11186317354440689, 0.08591295033693314, 0.5545093417167664, 0.15228885412216187, -0.00012087303184671327, -0.4648410975933075, 0.1461738497018814, -0.07211581617593765, -0.15252895653247833, -0.20889610052108765, -0.2326003760099411, -0.09182385355234146, 0.172158345580101, -0.33904990553855896, 0.07731151580810547, -0.44778406620025635, 0.17121952772140503, -0.2712027132511139, 0.6657806038856506, 0.34937092661857605, 0.07335954904556274, -0.10189507156610489, 0.18934044241905212, 0.22857528924942017, 0.4799536168575287, 0.5429452061653137, -0.3052915334701538, 0.1529545783996582, 0.2607788145542145, -0.20762114226818085, -0.2825186848640442, 0.15879979729652405, -0.09765517711639404, -0.2390972077846527, 0.3305780589580536, 0.1180301085114479, -0.27039965987205505, -0.3330766558647156, 0.04893827810883522, 0.3964375853538513, 0.3696252405643463, -0.4938056468963623, 0.048539161682128906, -0.0501629039645195, 0.18940632045269012, -0.4890809953212738, -0.1487528681755066, 0.6180802583694458, -0.5020692348480225, 0.35973885655403137, 0.33444279432296753, -0.38164615631103516, -0.4339837431907654, 0.12841801345348358, 0.4575885832309723, -0.0715697705745697, -0.0471397265791893, -0.07582766562700272, -0.05029013380408287, 0.1744946986436844, 0.21123643219470978, -0.24556134641170502, -0.14875836670398712, 0.34700560569763184, -0.14370547235012054, -0.16275890171527863, -0.4877949655056, -0.20781898498535156, 0.27582237124443054, -0.029754450544714928, 0.37007030844688416, -0.13018043339252472, -0.039095163345336914, -0.09167711436748505, 0.03946318104863167, -0.0950193777680397, 0.4510572552680969, 0.20620958507061005, -0.009002195671200752, -0.2627828121185303, 0.1622351109981537, -0.08094630390405655, -0.20558732748031616, 0.09666646271944046, -0.15640898048877716, 0.033218465745449066, 0.04962414130568504, -0.19279040396213531, 0.011112671345472336, 0.006019113585352898, -0.43664777278900146, -0.02457011491060257, -0.2640625536441803, 0.23352773487567902, 0.1825588196516037, -0.09135633707046509, 0.22798117995262146, 0.39805829524993896, 0.0025834888219833374, -0.19100555777549744, 0.12297520786523819, 0.07351357489824295, -0.2823682129383087, 0.28089261054992676, 0.24723415076732635, 0.04876604676246643, -0.0020065074786543846, -0.09525178372859955, 0.0477818064391613, -0.3019627034664154, 0.19040685892105103, -0.011526078917086124, 0.16694137454032898, -0.11928211897611618, 0.2679125666618347, 0.04077582061290741, -0.14994068443775177, -0.12755177915096283, -0.10135649889707565, 0.3692103326320648, -0.36875373125076294, -0.12114396691322327, -0.1366642713546753, -0.0014927085721865296, 0.05503595247864723, -0.14150629937648773, -0.277797132730484, 0.2512914836406708, 0.13343468308448792, 0.19284117221832275, 0.055643387138843536, 0.2801029086112976, -0.013974376954138279, -0.2512495219707489, 0.034251902252435684, 0.4760321378707886, -0.2327185422182083, 0.1828201413154602, -0.13035494089126587, 0.42202672362327576, 0.04189292713999748, 0.19998131692409515, -0.25090551376342773, 0.3373117446899414, 0.028533026576042175, 0.3784562945365906, 0.730847179889679, -0.07829371839761734, 0.22644934058189392, 0.40336307883262634, 0.18155477941036224, -0.3522214889526367, 0.1490733027458191, -0.04322519898414612, 0.10294289141893387, -0.09913064539432526, -0.3000408709049225, 0.4925093650817871, 0.11881759762763977, 0.03809338063001633, -0.07898520678281784, -0.20245958864688873, -0.18061652779579163, 0.01382903940975666, -0.14248500764369965, -0.15212266147136688, 0.09050925821065903, -0.4588903486728668, 0.4787382483482361, -0.2984761595726013, 0.4021229147911072, 0.2903977036476135, 0.04152550548315048, -0.014199175871908665, 0.24639372527599335, -0.11176910996437073, -0.18240371346473694, -0.3121176064014435, -0.40653079748153687, 0.29062819480895996, -0.2542754113674164, -0.27431434392929077, -0.25395360589027405, -0.1194472387433052, 0.07688555121421814, -0.03059476427733898, -0.03145666792988777, -0.24361342191696167, -0.21933604776859283, 0.01776423305273056, -0.2073882669210434, 0.25948426127433777, -0.279771625995636, 0.1349845826625824, -0.23029454052448273, 0.21568067371845245, 0.08199438452720642, -0.19584093987941742, 0.24660909175872803, 0.238074392080307, -0.2799946367740631, 0.14895698428153992, 0.06565982103347778, 0.1418318748474121, -0.05355468764901161, 0.20256243646144867, 0.4277292788028717, -0.28081637620925903, 0.08477290719747543, -0.6659769415855408, 0.303709477186203, -0.041348572820425034, 0.22573725879192352, -0.1121065765619278, -0.30820420384407043, 0.44492775201797485, -0.062423475086688995, 0.19242997467517853, -0.1610330194234848, 0.018283260986208916, -0.07411964237689972, -0.030854571610689163, -0.08382116258144379, 0.1798776537179947, 0.0551714189350605, 0.4346470236778259, 0.15664000809192657, -0.151343435049057, 0.027362657710909843, 0.08514789491891861, -0.12751780450344086, -0.17111225426197052, 0.23465584218502045, 0.23757700622081757, -0.3029536306858063, -0.1035986989736557, -0.16728395223617554, 0.048393137753009796, 0.09259168803691864, 0.18483610451221466, -0.15063852071762085, -0.040358055382966995, 0.08050721883773804, -0.09687429666519165, 0.37088558077812195, -0.03018070198595524, 0.29314425587654114, -0.16108772158622742, -0.09078805893659592, -0.023355230689048767, -0.17951402068138123, -0.2806975841522217, -0.04755215346813202, 0.1384667158126831, -0.21954290568828583, 0.028408918529748917, -0.0660051703453064, -0.37388548254966736, -0.33117762207984924, 0.12960989773273468, -0.04142502695322037, 0.25273463129997253, 0.012834284454584122, -0.24280166625976562, 0.16066338121891022, 0.0032888809219002724, -0.01538765151053667, 0.5403857827186584, 0.0907275527715683, 0.07187333703041077, -0.1886596828699112, -0.5223939418792725, -0.18206371366977692, 0.05640025809407234, 0.24623337388038635, 0.4204184114933014, 0.3654951751232147, -0.1879570335149765, 0.12019797414541245, -0.03721219301223755, -0.015586593188345432, 0.03923342004418373, -0.17580163478851318, -0.2756323218345642, -0.1856740415096283, 0.12907792627811432, -0.5696604251861572, -0.10519042611122131, -0.21616104245185852, 0.0416451096534729, -0.1757650226354599, -0.2118130773305893, -0.07265753298997879, -0.04017423465847969, 0.19620919227600098, -0.524660050868988, -0.15463657677173615, -0.12683823704719543, 0.35743263363838196, 0.30721619725227356, 0.2926774322986603, -0.024484500288963318, 0.23975226283073425, 0.11661981046199799, -0.30759090185165405, -0.05353693291544914, 0.19105195999145508, -0.06540253758430481, 0.3787030279636383, -0.3172714412212372, -0.53702312707901, 0.011981855146586895, -0.053312044590711594, -0.07275426387786865, 0.31094780564308167, -0.3548155725002289, 0.4196316599845886, 0.0823073610663414, 0.029532834887504578, -0.10327095538377762, -0.012858390808105469, 0.24516254663467407, -0.00042274361476302147, 0.06135663390159607, -0.05839310958981514, 0.09048913419246674, 0.3568626940250397, -0.12842382490634918, 0.3276352882385254, 0.34489211440086365, 0.14565619826316833, -0.04661427065730095, 0.5362755656242371, 0.2517479956150055, 0.06070588901638985, 0.17636649310588837, -0.11871366202831268, 0.07755324989557266, 0.10546045005321503, 0.031869396567344666, -0.0742826908826828, 0.12698684632778168, 0.037481825798749924, 0.18022377789020538, 0.31824931502342224, -0.32272589206695557, -0.20063075423240662, -0.2434098720550537, -0.372249960899353, -0.19315630197525024, 0.3950958251953125, -0.29584529995918274, 0.2528686225414276, -0.11756269633769989, -0.21010123193264008, -0.17794986069202423, -0.47201061248779297, -0.10274522751569748, 0.23868024349212646, 0.2641414999961853, 0.17290489375591278, -0.04200548678636551, 0.36136138439178467, -0.6206265091896057, 0.2994307279586792, 0.18296539783477783, -0.11638819426298141, 0.11488426476716995, 0.053787846118211746, 0.213324174284935, 0.08678209781646729, -0.11692533642053604, 0.16837002336978912, -0.05905267596244812, -0.014313637278974056, 0.33650946617126465, -0.3689158260822296, 0.32537195086479187, -0.2826526463031769, 0.03724048659205437, 0.22013774514198303, -0.11274105310440063, -0.43968942761421204, -0.2201676219701767, 0.17403534054756165, 0.14907991886138916, -0.110886350274086, -0.402035117149353, 0.12999650835990906, 0.005802448373287916, -0.3960942327976227, -0.14598964154720306, 0.3968571126461029, -0.0160329919308424, -0.12497652322053909, -0.15539225935935974, -0.01966412551701069, 0.10079915821552277, 0.22278159856796265, -0.23059988021850586, 0.2011769860982895, -0.2508758008480072, 0.021771615371108055, -0.08185719698667526, 0.3278181254863739, 0.06736352294683456, -0.0015114372363314033, 0.18697333335876465, -0.22211620211601257, -0.06431413441896439, -0.3994569778442383, 0.5800669193267822, 0.1728687584400177, 0.11208879947662354, 0.15619605779647827, 0.18525199592113495, -0.3044201135635376, -0.11992882192134857, -0.011086774058640003, 0.13509562611579895, -0.11089463531970978, -0.41002482175827026, -0.3892831802368164, 0.5749382972717285, 0.04820924624800682, -0.03414592891931534, 0.2953093647956848, 0.05651557445526123, -0.2902968227863312, 0.6706214547157288, 0.20020262897014618, 1.2255789041519165, -0.1759890466928482, 0.2871074974536896, 0.05487578362226486, -0.10766845941543579, 0.5923519730567932, -0.4202047288417816, -0.21309903264045715, -0.5059723854064941, 0.30423203110694885, -0.11025378853082657, -0.14544333517551422, 0.15465308725833893, 0.33180296421051025, -0.06618240475654602, 0.10463598370552063, 0.2753223776817322, -0.04574216529726982, 0.13521774113178253, 0.3248899281024933, -0.21979475021362305, -0.4144163429737091, -0.1502167284488678, -0.09407124668359756, -0.12854498624801636, 0.04692883789539337, 0.11305601894855499, -0.4174148738384247, -0.32015806436538696, 0.07482128590345383, -0.4243672490119934, 0.10035193711519241, -0.008133471012115479, -0.22221879661083221, -0.26612770557403564, -0.6077929735183716, 0.3085108995437622, 0.3429757058620453, 0.09421681612730026, -0.007780071347951889, -0.285155713558197, 0.09172757714986801, -0.2546539604663849, -0.3908627927303314, 0.2797311544418335, -0.12988482415676117, 0.07501404732465744, -0.1504439115524292, -0.2056218683719635, 0.15109021961688995, 0.053390566259622574, -0.2871480882167816, -0.006581729277968407, -0.22503302991390228, 0.49903053045272827, -0.12148301303386688, -0.06034611538052559, 0.3270648121833801, -0.41001591086387634, 0.005483092274516821, 0.07348689436912537, 0.38404181599617004, 0.10894716531038284, 0.05277856066823006, 0.26296374201774597, -0.2531401813030243, -0.04520682990550995, 0.17500491440296173, 0.3665909171104431, -0.22578535974025726, 0.500103235244751, 0.41960781812667847, -0.1537499725818634, -0.29532957077026367, 0.2761525809764862, 0.3888963460922241, -0.8429333567619324, -0.10552596300840378, -0.008339760825037956, 0.09867822378873825, 0.016229400411248207, 0.10067291557788849, -0.0775008499622345, -0.046553898602724075, 0.010311639867722988, -0.5420810580253601, -0.2696770429611206, 0.14990979433059692, -0.09666278213262558, 0.07544498890638351, 0.39812320470809937, 0.09652497619390488, 0.04153716564178467, 0.13556745648384094, -0.20735105872154236, 0.08010707795619965, 0.5916275382041931, 0.17521585524082184, 0.10622335970401764, 0.09708918631076813, 0.01988169550895691, 0.025556664913892746, -0.02478550747036934, 0.11710426956415176, 0.11115169525146484, -0.1314205378293991, -0.01541933510452509, 0.20284494757652283, -0.060861390084028244, -0.2965809106826782, -0.3098231256008148, -0.45137447118759155, -0.2555307149887085, -0.1911255121231079, -0.14729654788970947, -0.15576715767383575, -0.13286036252975464, 0.13090816140174866, 0.07797190546989441, 0.014390951953828335, -0.4827424883842468, 0.22568190097808838, -0.21510645747184753, 0.2515772581100464, -0.09725835919380188, 0.14951981604099274, -0.0002763932861853391, -0.05023441091179848, 0.04456472024321556, 0.12908244132995605, 0.2469748705625534, 0.18958783149719238, 0.5016096234321594, -0.11936333030462265, -0.04702065885066986, -0.37288737297058105, 0.20446960628032684, 0.2573812007904053, 0.0903845727443695, 0.06380900740623474, 0.29504814743995667, 0.09182357788085938, -0.22265169024467468, 0.2888721525669098, -0.23050346970558167, -0.20139411091804504, -0.18630313873291016, 0.12676368653774261, -0.02106987126171589, -0.19755548238754272, -0.3109135627746582, -0.16034027934074402, 0.18026578426361084, -0.015163795091211796, -0.3553696572780609, 0.1900033950805664, 0.225592240691185, -0.04182201623916626, -0.13116061687469482, 0.22774840891361237, 0.25549525022506714, -0.06322751194238663, -0.05395853519439697, 0.15032479166984558, 0.7298561930656433, 0.26674479246139526, 0.14226189255714417, -0.16784672439098358, 0.041503969579935074, 0.4384758770465851, 0.24947476387023926, 0.014321968890726566, 0.2309006005525589, 0.5041627287864685, 0.6058549880981445, 0.10812123119831085, -0.048714034259319305, 0.29230356216430664, 0.12374929338693619, 0.03654130548238754, -0.45995837450027466, -0.0659705102443695, -0.048481978476047516, -0.27365195751190186, -0.1924750953912735, -0.07600390166044235, -0.14734040200710297, 0.3707128167152405, 0.15805162489414215, -0.055774375796318054, 0.1649741381406784, 0.07142599672079086, 0.32485803961753845, -0.249286949634552, 0.0821080133318901, -0.12004061788320541, 0.015363839454948902, 0.037473760545253754, 0.43583163619041443, 0.33153727650642395, 0.17218957841396332, -0.13928700983524323, 0.14918874204158783, -0.00022749700292479247, -0.10717977583408356, 0.17634797096252441, 0.19603362679481506, 0.4006016254425049, 0.03843877464532852, 0.07309972494840622, 0.04524249583482742, 0.027988078072667122, -0.12604168057441711, 0.3155635595321655, -0.2134237438440323, 0.09988724440336227, -0.15060219168663025, 0.07479551434516907, -0.37400513887405396, -0.16975465416908264, 0.0998465046286583, -0.16131077706813812, 0.36335042119026184, 0.6061455607414246, -0.23281385004520416, -0.04642930254340172, -0.11821356415748596, 0.011206601746380329, -0.10335055738687515, 0.4182172119617462, 0.343479722738266, 0.1270029991865158, -0.2906665503978729, -0.2813268303871155, -0.047038234770298004, 0.10857336968183517, -0.03955458477139473, -0.23724061250686646, -0.05329858884215355, 0.03458246961236, 0.2613575756549835, 0.4125305116176605, 0.048847638070583344, 0.1342530995607376, -0.11785801500082016, 0.471804678440094, -0.2880139648914337, -0.04197230935096741, 0.27318766713142395, 0.5164782404899597, -0.16281402111053467, -0.10243474692106247, 0.15229009091854095, 0.35864946246147156, -0.06448976695537567, -0.23673920333385468, -0.08922381699085236, 0.3275838792324066, -0.17799565196037292, 0.22377046942710876, -0.09077418595552444, 0.15293285250663757, -0.05167486146092415, -0.12333337962627411, 0.0046212635934352875, -0.08744798600673676, 0.10571639239788055, 0.4439932703971863, 0.32237693667411804, -0.19108951091766357, -0.26522547006607056, -0.0015396400121971965, 0.04711379483342171, 0.27495086193084717, -0.28760039806365967, -0.050848137587308884, -0.27267852425575256, -0.023377306759357452, -0.2617731988430023, 0.14743784070014954, -0.034875672310590744, 0.28765201568603516, -0.21054615080356598, 0.1158321350812912, -0.12690186500549316, -0.13028740882873535, 0.09160185605287552, -0.25090867280960083, 0.03996600955724716, 0.26852041482925415, 0.21695385873317719, 0.5681495070457458, 0.36740532517433167, -0.5660693645477295, 0.2106538712978363, 0.234237402677536, 0.07025491446256638, 0.10360503941774368, 0.06293857842683792, 0.002586635062471032, 0.10403311252593994, -0.01696956157684326, 0.39650651812553406, -0.05848297104239464, -0.03820473328232765, -0.46145686507225037, -0.49836650490760803 ]
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 :)
[ -0.10410337150096893, -0.08714869618415833, -0.09965094923973083, 0.15440328419208527, -0.3756243586540222, 0.15951842069625854, 0.1173253059387207, 0.4278324842453003, -0.04359566420316696, -0.2937576174736023, 0.07305051386356354, 0.11510442197322845, -0.0833790972828865, 0.14750348031520844, 0.2515306770801544, -0.15385332703590393, -0.0880170464515686, 0.21654416620731354, -0.29338327050209045, 0.015712060034275055, 0.27406179904937744, -0.02474914863705635, -0.0877852588891983, 0.012240911833941936, -0.17144307494163513, -0.2362554371356964, -0.12816713750362396, -0.003313877386972308, -0.1450771689414978, -0.3376830518245697, -0.02538461424410343, -0.02935338392853737, 0.5116753578186035, 0.20511552691459656, -0.00012170406989753246, -0.346560001373291, 0.12741811573505402, -0.004115762654691935, -0.24111774563789368, -0.20824989676475525, -0.15198774635791779, -0.009040852077305317, 0.19447411596775055, -0.33560213446617126, -0.06002168357372284, -0.3797878324985504, 0.03028702735900879, -0.4688572585582733, 0.8392435312271118, 0.3928324282169342, 0.07577164471149445, -0.160225972533226, 0.18871460855007172, 0.2179347425699234, 0.3413509726524353, 0.5511065721511841, -0.3402370512485504, -0.0774216428399086, 0.14995823800563812, 0.03337555006146431, -0.1650860756635666, 0.19850070774555206, 0.04398881271481514, -0.15324635803699493, 0.3540644645690918, -0.03162284195423126, -0.2806328535079956, -0.2210429310798645, 0.12315770983695984, 0.32636207342147827, 0.4954078495502472, -0.4750733971595764, -0.04503718763589859, -0.01197209395468235, 0.2050464004278183, -0.42492932081222534, -0.2257211059331894, 0.575936496257782, -0.45890533924102783, 0.48547738790512085, 0.14960379898548126, -0.40089499950408936, -0.4561852514743805, -0.07455511391162872, 0.39253920316696167, -0.11834359914064407, -0.10440409183502197, -0.06381220370531082, -0.14212645590305328, 0.162960484623909, 0.2270011305809021, -0.3375423550605774, -0.1277940422296524, 0.2690400183200836, -0.042774997651576996, -0.21517284214496613, -0.5973444581031799, -0.1973356455564499, 0.18537627160549164, 0.07732852548360825, 0.38570842146873474, -0.11290761828422546, -0.04760909825563431, -0.2783754765987396, 0.07719892263412476, -0.07277941703796387, 0.4339834153652191, 0.1329328566789627, -0.007071024738252163, -0.26325637102127075, 0.17075693607330322, -0.11871547251939774, -0.09468071162700653, 0.01739884726703167, -0.13679398596286774, 0.20157107710838318, -0.010499742813408375, -0.27809855341911316, -0.03335605934262276, -0.019908525049686432, -0.38649529218673706, -0.07700017839670181, -0.21380306780338287, 0.026128947734832764, 0.20136485993862152, -0.14675681293010712, 0.2247103750705719, 0.3776317834854126, 0.05205437168478966, -0.2802138030529022, 0.17620940506458282, -0.018021680414676666, -0.23208190500736237, 0.2816554307937622, 0.08700833469629288, 0.16538026928901672, -0.13530714809894562, -0.0685046911239624, -0.058296702802181244, -0.14523963630199432, 0.2524913251399994, -0.03617075830698013, 0.21758884191513062, -0.06669122725725174, 0.2069646716117859, 0.016298513859510422, -0.12016601860523224, -0.007247450295835733, -0.07581135630607605, 0.3549630641937256, -0.18190164864063263, 0.03534028306603432, -0.04450514167547226, 0.015734365209937096, 0.07062416523694992, -0.08456771820783615, -0.2357398420572281, 0.1604422777891159, -0.0032508010044693947, 0.3336625397205353, 0.07029975205659866, 0.28684455156326294, 0.06052299961447716, -0.14232558012008667, -0.06134994700551033, 0.4451693892478943, -0.3385562300682068, 0.1481718271970749, -0.09775713086128235, 0.20304511487483978, -0.11506333947181702, 0.30969688296318054, -0.22147974371910095, 0.2855270206928253, -0.021732628345489502, 0.24805942177772522, 0.7609074711799622, -0.0031956490129232407, 0.19536159932613373, 0.4988290071487427, 0.03261345624923706, -0.37933841347694397, 0.09234792739152908, -0.08416333049535751, -0.0858478844165802, -0.029711076989769936, -0.28686606884002686, 0.5281494855880737, 0.11185731738805771, -0.0169502105563879, -0.11940541863441467, -0.1720544546842575, -0.11659663170576096, -0.053938161581754684, -0.23231510818004608, -0.05482996255159378, 0.09218282997608185, -0.4823830723762512, 0.48952716588974, -0.3356693387031555, 0.453250527381897, 0.237852543592453, 0.02512144297361374, -0.03932974860072136, 0.17496205866336823, -0.031960777938365936, -0.0986606553196907, -0.3560510277748108, -0.3879696726799011, 0.21412913501262665, -0.32696712017059326, -0.23821283876895905, -0.2696840167045593, -0.12529031932353973, -0.06862255930900574, 0.08229725807905197, -0.019796207547187805, -0.23528046905994415, -0.15579436719417572, 0.014363802038133144, -0.21273359656333923, 0.31802883744239807, -0.11018786579370499, 0.14913539588451385, -0.22364486753940582, 0.1163676455616951, 0.10570722818374634, -0.004393710754811764, 0.23759892582893372, 0.13671790063381195, -0.23089107871055603, 0.05242152884602547, -0.05458371713757515, 0.0032956537324935198, -0.07111270725727081, 0.17900705337524414, 0.5197830200195312, -0.2920800745487213, 0.07352745532989502, -0.6826451420783997, 0.3104526996612549, 0.017530804499983788, 0.21841387450695038, -0.10102801769971848, -0.46540236473083496, 0.48077988624572754, -0.08094535768032074, 0.19033676385879517, -0.14410781860351562, 0.04329710826277733, -0.13720950484275818, -0.11262884736061096, -0.1677834689617157, 0.1439184695482254, -0.06783585250377655, 0.3569745123386383, 0.21481411159038544, -0.09206952899694443, 0.09972027689218521, 0.031151819974184036, -0.19164346158504486, -0.19726820290088654, 0.23297223448753357, 0.2144039273262024, -0.23323842883110046, -0.07656899839639664, -0.061181314289569855, 0.04091774672269821, 0.11759545654058456, 0.14910195767879486, 0.021691594272851944, 0.07660017907619476, 0.22029931843280792, -0.13715267181396484, 0.3341182470321655, -0.13219459354877472, 0.3828600347042084, -0.15953008830547333, -0.16526345908641815, 0.02119910530745983, -0.09979046881198883, -0.1932457983493805, -0.10108159482479095, 0.14014822244644165, -0.0816582590341568, -0.038221001625061035, -0.13772614300251007, -0.4040078818798065, -0.3725436329841614, 0.2617993950843811, -0.03137795254588127, 0.2539512515068054, -0.07426423579454422, -0.4246176779270172, 0.14733339846134186, -0.15592803061008453, -0.07536529004573822, 0.5126532912254333, 0.07902297377586365, 0.060460690408945084, -0.283617228269577, -0.39828288555145264, -0.16304081678390503, 0.07741985470056534, 0.30698490142822266, 0.4057590365409851, 0.3491530418395996, -0.08209389448165894, 0.19486457109451294, -0.19154047966003418, 0.09424605220556259, 0.027147069573402405, -0.034980468451976776, -0.30891913175582886, -0.1563505381345749, 0.16644497215747833, -0.5549852848052979, 0.1195530965924263, -0.1840694397687912, 0.11688744276762009, -0.2389352023601532, -0.2792016863822937, -0.07527553290128708, -0.02139127440750599, 0.2210381180047989, -0.4112201929092407, -0.13717831671237946, -0.08414337784051895, 0.4223138689994812, 0.23396283388137817, 0.2500329911708832, -0.041658323258161545, 0.13203297555446625, 0.17086562514305115, -0.23383066058158875, -0.07233085483312607, 0.19056572020053864, -0.1538463830947876, 0.2510853707790375, -0.3682454228401184, -0.5960060954093933, -0.06363970041275024, -0.01584281027317047, 0.02100483886897564, 0.31264835596084595, -0.4215536415576935, 0.5755351781845093, -0.035364627838134766, 0.05393088236451149, -0.14537565410137177, -0.005995914340019226, 0.20333778858184814, -0.03361855819821358, 0.044098347425460815, -0.03009035252034664, 0.08547823131084442, 0.31540369987487793, -0.1693236529827118, 0.35248586535453796, 0.4192034602165222, 0.13491712510585785, 0.08188989758491516, 0.6222643256187439, 0.2455945611000061, 0.10145345330238342, 0.2365695983171463, -0.024821165949106216, 0.006416776683181524, 0.02607899159193039, 0.05190137028694153, -0.14044998586177826, 0.19511985778808594, 0.12384898215532303, 0.07361665368080139, 0.28938066959381104, -0.0958738699555397, -0.29619133472442627, -0.26517948508262634, -0.3516790568828583, -0.20116376876831055, 0.29699966311454773, -0.3599570393562317, 0.27954310178756714, -0.15597039461135864, -0.21718445420265198, -0.0630711168050766, -0.31133952736854553, 0.05996674671769142, 0.1286882609128952, 0.3190664052963257, 0.14906051754951477, -0.270694375038147, 0.34821560978889465, -0.6118109226226807, 0.34974852204322815, 0.051624149084091187, -0.10029671341180801, 0.05672121420502663, 0.03056403435766697, 0.14685799181461334, 0.08642436563968658, -0.17882384359836578, 0.11098542809486389, -0.025316115468740463, 0.009351910091936588, 0.28992727398872375, -0.42268019914627075, 0.3903346657752991, -0.228807270526886, 0.11045831441879272, 0.012456626631319523, 0.12471523135900497, -0.3796062469482422, -0.2781963348388672, 0.134401336312294, 0.24216265976428986, -0.1416957974433899, -0.3648134469985962, 0.25685226917266846, 0.008763075806200504, -0.1879395991563797, -0.11055942624807358, 0.3730453550815582, 0.02087300643324852, -0.07553970068693161, -0.19689378142356873, -0.059822022914886475, 0.15742227435112, 0.1272709220647812, -0.3288094997406006, 0.10629282146692276, -0.2729625105857849, 0.026045652106404305, 0.05754528567194939, 0.38840237259864807, 0.008848387748003006, 0.08767355978488922, 0.176996648311615, -0.2161041498184204, -0.06899284571409225, -0.30543607473373413, 0.6498404145240784, 0.2223314493894577, 0.12618370354175568, 0.1377881020307541, 0.21072694659233093, -0.2504119277000427, -0.07103022933006287, -0.1461544930934906, 0.041151825338602066, -0.06889675557613373, -0.42613595724105835, -0.3491380512714386, 0.5717189311981201, 0.05395360663533211, 0.0027202737983316183, 0.19814066588878632, 0.08373115211725235, -0.27170059084892273, 0.5873886346817017, 0.27925050258636475, 1.185522437095642, -0.27428171038627625, 0.27304884791374207, -0.13550782203674316, -0.3160178065299988, 0.4253911077976227, -0.27867892384529114, -0.2547184228897095, -0.5003958344459534, 0.4487968385219574, -0.07695702463388443, -0.1531475931406021, 0.4294324517250061, 0.41808271408081055, -0.04471583664417267, 0.023450562730431557, 0.21344734728336334, -0.19627445936203003, 0.11166632175445557, 0.4447547197341919, -0.33992692828178406, -0.30095475912094116, -0.09718886017799377, -0.054392918944358826, -0.12668253481388092, 0.14949753880500793, 0.12932433187961578, -0.4405776858329773, -0.37254300713539124, 0.0856233686208725, -0.2790434658527374, 0.06759664416313171, -0.16045573353767395, -0.2300845980644226, -0.30945348739624023, -0.6526291370391846, 0.3845920264720917, 0.3031029999256134, 0.16113770008087158, 0.018032778054475784, -0.14568151533603668, 0.06527067720890045, -0.30593782663345337, -0.21463368833065033, 0.14000776410102844, -0.09702998399734497, 0.014985824935138226, -0.1734144240617752, -0.1473105549812317, 0.13271589577198029, 0.07229083776473999, -0.22374694049358368, 0.03694796562194824, -0.2066386491060257, 0.5542380213737488, -0.22024871408939362, -0.05737234279513359, 0.20829914510250092, -0.38716986775398254, 0.09555791318416595, 0.07478734850883484, 0.3437504470348358, 0.010053051635622978, -0.019843365997076035, 0.29231759905815125, -0.19841405749320984, 0.062226392328739166, 0.19615542888641357, 0.20381462574005127, -0.19679823517799377, 0.43999478220939636, 0.38451290130615234, -0.09716875106096268, -0.3226234018802643, 0.5589429140090942, 0.48946574330329895, -0.7103239893913269, -0.05827125534415245, -0.08733200281858444, 0.17481109499931335, 0.007512325420975685, 0.07378509640693665, -0.047736432403326035, -0.02427542954683304, -0.041590314358472824, -0.4507732689380646, -0.30290931463241577, 0.021304849535226822, -0.05329320952296257, 0.09318288415670395, 0.3139236271381378, 0.018152732402086258, 0.075315922498703, 0.04191671311855316, -0.20519819855690002, 0.15909443795681, 0.6357864737510681, 0.17778655886650085, 0.0277660321444273, 0.04229682311415672, -0.05671681836247444, -0.00015122817421797663, -0.004827910102903843, 0.13855095207691193, 0.07041764259338379, -0.1743817925453186, -0.03296957537531853, 0.21718795597553253, -0.09015797823667526, -0.32971838116645813, -0.33929261565208435, -0.44494301080703735, -0.15494002401828766, -0.03558608889579773, -0.13536977767944336, -0.08628891408443451, -0.06770704686641693, 0.1622815728187561, 0.1766914576292038, -0.09132018685340881, -0.5623161792755127, 0.12945884466171265, -0.10235704481601715, 0.2173244059085846, -0.13209497928619385, 0.16996753215789795, 0.09857212007045746, -0.05798984318971634, 0.04792133346199989, 0.00912930816411972, 0.3277367055416107, 0.1631397157907486, 0.35374242067337036, -0.1717623770236969, 0.04795071855187416, -0.20626893639564514, 0.3029593527317047, 0.1416802555322647, -0.008553608320653439, 0.07447577267885208, 0.29503998160362244, 0.0773182138800621, -0.2086273580789566, 0.23932312428951263, -0.2136075645685196, -0.2218509316444397, -0.2649485170841217, 0.059707608073949814, 0.004555591847747564, 0.02655915543437004, -0.413725346326828, -0.24392764270305634, 0.040048930794000626, 0.00961831584572792, -0.2730497419834137, 0.3426245450973511, 0.14183808863162994, -0.03672187775373459, -0.01231218222528696, 0.22880734503269196, 0.16126644611358643, -0.05761438608169556, -0.10032612830400467, 0.14080865681171417, 0.7164051532745361, 0.25883299112319946, 0.2297167330980301, -0.14957694709300995, 0.10609422624111176, 0.340749591588974, 0.17358367145061493, 0.06045190989971161, 0.3084302842617035, 0.5016961097717285, 0.4905334711074829, -0.06533703207969666, 0.10396668314933777, 0.2332955002784729, 0.02619994804263115, 0.09245084226131439, -0.29751911759376526, -0.09762120991945267, -0.07974143326282501, -0.2705273926258087, -0.267034649848938, -0.06595662981271744, -0.12805074453353882, 0.3783091604709625, 0.22421585023403168, 0.006519712973386049, 0.09486527740955353, 0.07628809660673141, 0.3090254068374634, -0.19678981602191925, -0.08400526642799377, -0.04792492836713791, 0.06425434350967407, 0.01481980923563242, 0.44661569595336914, 0.14851853251457214, 0.11663661152124405, -0.18219131231307983, 0.21957488358020782, -0.03238433226943016, -0.13474982976913452, 0.2533149719238281, 0.24931137263774872, 0.4374347925186157, 0.02465912513434887, 0.16236263513565063, 0.03777836635708809, 0.06002860143780708, 0.010893750004470348, 0.1967502236366272, -0.16068576276302338, -0.10829715430736542, -0.04838685318827629, -0.07437445223331451, -0.4557873010635376, -0.21377208828926086, 0.07236259430646896, -0.13846880197525024, 0.20172831416130066, 0.6342796087265015, -0.30233147740364075, -0.05422424152493477, -0.22437326610088348, 0.012459770776331425, -0.11517810076475143, 0.4238155484199524, 0.3466523289680481, 0.14342570304870605, -0.29574817419052124, -0.28634220361709595, -0.06824126094579697, 0.11662798374891281, -0.023076074197888374, -0.0425318144261837, -0.09624223411083221, -0.03116641566157341, 0.27951696515083313, 0.5127826929092407, 0.13443037867546082, 0.1204080805182457, -0.17573867738246918, 0.4893505871295929, -0.23794105648994446, -0.01999320462346077, 0.45011642575263977, 0.5878044366836548, -0.15019407868385315, -0.18826636672019958, 0.3224354684352875, 0.6150745749473572, -0.060428470373153687, -0.19979703426361084, 0.007830471731722355, 0.33421868085861206, -0.17057214677333832, 0.2096032053232193, -0.013460854068398476, 0.09708040952682495, 0.02499501407146454, -0.0999695211648941, 0.1413358449935913, 0.03287169337272644, 0.12586309015750885, 0.5581203103065491, 0.250945121049881, -0.17178648710250854, -0.2090209573507309, -0.06832195073366165, 0.015021970495581627, 0.2349604219198227, -0.3354088068008423, -0.08596210926771164, -0.31260305643081665, -0.09726453572511673, -0.2712607681751251, 0.04189329594373703, 0.016883840784430504, 0.30374675989151, -0.17290377616882324, 0.01221131719648838, -0.20690612494945526, 0.02385246753692627, 0.11828352510929108, -0.2757699191570282, 0.03076813369989395, 0.36452439427375793, 0.10293444991111755, 0.622542142868042, 0.15047983825206757, -0.48482584953308105, 0.1481480747461319, 0.31665995717048645, -0.022683413699269295, 0.2637162506580353, 0.0905396044254303, 0.057136449962854385, 0.130814790725708, 0.010113371536135674, 0.5547729134559631, -0.05681564286351204, 0.0049941944889724255, -0.5158872008323669, -0.4087068736553192 ]
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)
[ -0.11623521149158478, -0.18411757051944733, -0.0716126412153244, 0.17512021958827972, -0.32802635431289673, 0.13923627138137817, 0.11795948445796967, 0.3817433714866638, 0.002252156613394618, -0.27272239327430725, 0.05691436678171158, 0.09839930385351181, -0.07723762094974518, 0.21921499073505402, 0.3109801709651947, -0.05802787095308304, -0.10293136537075043, 0.1872921735048294, -0.28998053073883057, 0.03419720381498337, 0.2486112415790558, 0.02355021983385086, -0.06273804605007172, 0.01480869296938181, -0.18604955077171326, -0.21685832738876343, -0.123759925365448, 0.10292933136224747, -0.11955875903367996, -0.40562543272972107, 0.04538268223404884, -0.0250192079693079, 0.519914448261261, 0.1765509992837906, -0.00012255784531589597, -0.3012935221195221, 0.14918231964111328, -0.008922329172492027, -0.2559577524662018, -0.35214313864707947, -0.09115905314683914, -0.05141798034310341, 0.3022273778915405, -0.2763158679008484, -0.1366528868675232, -0.3666139841079712, 0.044753946363925934, -0.46327656507492065, 0.8944177031517029, 0.3823668658733368, 0.06688162684440613, -0.06973019987344742, 0.12017393857240677, 0.17469850182533264, 0.2935894727706909, 0.6498957872390747, -0.2985197901725769, 0.017334166914224625, 0.22698600590229034, -0.02760334499180317, -0.18706849217414856, 0.2103872299194336, 0.03350253403186798, -0.1601579487323761, 0.41826868057250977, 0.032942995429039, -0.31972944736480713, -0.2667314410209656, 0.10459908097982407, 0.2918635308742523, 0.4435719847679138, -0.5081268548965454, -0.06863576173782349, -0.15882885456085205, 0.19593681395053864, -0.4047841727733612, -0.20237405598163605, 0.5912532806396484, -0.4560740888118744, 0.4850004315376282, 0.11115676164627075, -0.33012697100639343, -0.45467624068260193, -0.023367075249552727, 0.4130203425884247, -0.11923040449619293, -0.1424250304698944, -0.034220967441797256, -0.08781947940587997, 0.11395946890115738, 0.14466150104999542, -0.3402825593948364, -0.08196957409381866, 0.2702951431274414, -0.03041764535009861, -0.14818446338176727, -0.5688724517822266, -0.11655567586421967, 0.29623928666114807, 0.1948106735944748, 0.35812410712242126, -0.14922839403152466, -0.08302668482065201, -0.24126337468624115, 0.11707842350006104, -0.035984594374895096, 0.4621579051017761, 0.12294959276914597, -0.02287806011736393, -0.10338737815618515, 0.21610508859157562, -0.11804772168397903, -0.07146728783845901, -0.0071605113334953785, -0.1958075910806656, 0.1252121925354004, 0.06821301579475403, -0.3583875894546509, -0.06369628757238388, -0.06358204782009125, -0.31606045365333557, -0.06485763192176819, -0.1641440987586975, 0.12130096554756165, 0.15002985298633575, -0.148542582988739, 0.2026013880968094, 0.42098936438560486, 0.002455431269481778, -0.23466061055660248, 0.13133789598941803, -0.010917400009930134, -0.2892899811267853, 0.34125468134880066, 0.1514548659324646, 0.0410255566239357, -0.03367704525589943, -0.0651337280869484, 0.041876308619976044, -0.18663989007472992, 0.21144604682922363, 0.0035024217795580626, 0.25545671582221985, -0.027918081730604172, 0.22552098333835602, 0.05125497281551361, -0.08152381330728531, -0.015555879101157188, -0.1415529102087021, 0.3194228410720825, -0.19716249406337738, -0.0013498141197487712, -0.019552184268832207, -0.01256622839719057, -0.07126845419406891, -0.1170920804142952, -0.2966444492340088, 0.16671666502952576, -0.055329710245132446, 0.34108299016952515, 0.051545530557632446, 0.3041135370731354, -0.004431361332535744, -0.15689964592456818, 0.060645878314971924, 0.5751582980155945, -0.27983880043029785, 0.07748086005449295, -0.06367768347263336, 0.1921239197254181, -0.13204005360603333, 0.32104742527008057, -0.24032819271087646, 0.20451746881008148, -0.10772119462490082, 0.21508720517158508, 0.6738171577453613, -0.11156903207302094, 0.16387337446212769, 0.5188358426094055, 0.056450359523296356, -0.2370750457048416, 0.11921440809965134, -0.11641411483287811, -0.07434394210577011, -0.0028064704965800047, -0.3140355348587036, 0.5571512579917908, 0.08260925114154816, 0.018255937844514847, -0.1299574375152588, -0.21523691713809967, -0.09700988233089447, 0.004683936946094036, -0.23405605554580688, -0.014258374460041523, 0.060363538563251495, -0.5500680208206177, 0.4852309226989746, -0.3260747492313385, 0.4003874361515045, 0.2320333570241928, 0.0025298113469034433, 0.02298738621175289, 0.13997365534305573, -0.03116629831492901, -0.26830247044563293, -0.25660669803619385, -0.3145011067390442, 0.22397644817829132, -0.45106619596481323, -0.24881409108638763, -0.32507526874542236, -0.1321396827697754, -0.10724630951881409, -0.021122047677636147, -0.0625174343585968, -0.2079956829547882, -0.10984879732131958, 0.053021274507045746, -0.21073774993419647, 0.36422210931777954, -0.054530225694179535, 0.24731071293354034, -0.36918309330940247, 0.12507282197475433, 0.13142229616641998, -0.0102913286536932, 0.23380368947982788, 0.17634469270706177, -0.1896403580904007, 0.03401082754135132, 0.007023175712674856, 0.084144227206707, 0.010062295012176037, 0.2041340470314026, 0.4599419832229614, -0.2706069052219391, 0.1242067962884903, -0.713015079498291, 0.27949148416519165, 0.007425835356116295, 0.1846461147069931, -0.09204734861850739, -0.48254626989364624, 0.5747973322868347, -0.14493010938167572, 0.23111040890216827, -0.14596465229988098, 0.027441678568720818, -0.1334022730588913, -0.07772515714168549, -0.22561103105545044, 0.06281200796365738, -0.03213820233941078, 0.32594138383865356, 0.35702478885650635, -0.12372733652591705, 0.017253387719392776, 0.03389057517051697, -0.22877824306488037, -0.22015008330345154, 0.15034902095794678, 0.26336804032325745, -0.2448005974292755, -0.07236367464065552, -0.03271269425749779, -0.008668411523103714, 0.11415647715330124, 0.17046137154102325, -0.025074487552046776, 0.12046822160482407, 0.20249882340431213, -0.1406170278787613, 0.30452239513397217, -0.10037089884281158, 0.41077476739883423, -0.17753517627716064, -0.1377762109041214, -0.012871000915765762, -0.12090447545051575, -0.23075655102729797, -0.16143402457237244, 0.10699723660945892, -0.17705482244491577, -0.008561881259083748, -0.16584183275699615, -0.4955018162727356, -0.4169103801250458, 0.23199540376663208, -0.06793005019426346, 0.25156453251838684, -0.16291959583759308, -0.3335671126842499, 0.12842416763305664, -0.1341254860162735, -0.10172931849956512, 0.4734610915184021, -0.00704944459721446, 0.016294963657855988, -0.2771506905555725, -0.38154301047325134, -0.16488195955753326, 0.06597515195608139, 0.3686506450176239, 0.41833606362342834, 0.3976905047893524, -0.12005116790533066, 0.13771800696849823, -0.19119061529636383, 0.014920555055141449, 0.04858419671654701, -0.09082522243261337, -0.213777095079422, -0.12403147667646408, 0.13249340653419495, -0.524358868598938, 0.07383081316947937, -0.1173049807548523, 0.08368141204118729, -0.23695898056030273, -0.2871926426887512, -0.09819866716861725, 0.013112296350300312, 0.14931614696979523, -0.39084935188293457, -0.16516515612602234, -0.10185346007347107, 0.5026170611381531, 0.2510916292667389, 0.19040635228157043, 0.043757904320955276, 0.10605823248624802, 0.1319567710161209, -0.2570900619029999, -0.10649903863668442, 0.18408694863319397, -0.20155037939548492, 0.2977464199066162, -0.3921269178390503, -0.5845624804496765, -0.11088483035564423, -0.006884878501296043, 0.034188393503427505, 0.2856810986995697, -0.4618055820465088, 0.41268235445022583, -0.07110357284545898, 0.07117266207933426, -0.11788604408502579, -0.010395952500402927, 0.2041831761598587, -0.025251785293221474, 0.08497034013271332, -0.020421473309397697, -0.017280982807278633, 0.29869183897972107, -0.20029915869235992, 0.28786909580230713, 0.425469309091568, 0.17232239246368408, 0.05806047096848488, 0.5829076170921326, 0.30137336254119873, 0.10892760008573532, 0.28402993083000183, -0.08005820959806442, 0.05360938981175423, -0.026313748210668564, 0.013504449278116226, -0.20941540598869324, 0.12492363899946213, 0.14099840819835663, 0.1072889119386673, 0.3072505295276642, -0.03692935034632683, -0.33212608098983765, -0.2866545617580414, -0.41476649045944214, -0.21308189630508423, 0.30389177799224854, -0.38708099722862244, 0.23486047983169556, -0.1826099306344986, -0.1835949420928955, -0.0711633488535881, -0.31488901376724243, 0.05090329051017761, 0.22340433299541473, 0.35175609588623047, 0.2050117701292038, -0.26657724380493164, 0.4281851649284363, -0.6509033441543579, 0.368991494178772, 0.05844467133283615, -0.14979803562164307, 0.050097499042749405, -0.0044016228057444096, 0.17925050854682922, 0.0758807510137558, -0.10500716418027878, 0.1694646179676056, -0.01878582127392292, -0.022987667471170425, 0.14634031057357788, -0.5188398957252502, 0.39623937010765076, -0.17134422063827515, 0.027792606502771378, 0.012690299190580845, 0.19453561305999756, -0.4316796362400055, -0.3492642343044281, 0.11258299648761749, 0.22321465611457825, -0.12154306471347809, -0.33115944266319275, 0.21770139038562775, -0.060272831469774246, -0.2420458346605301, -0.0690440908074379, 0.3726751208305359, 0.033900585025548935, -0.04761365428566933, -0.13076604902744293, -0.04701988026499748, 0.12143313139677048, 0.11338670551776886, -0.2611369788646698, 0.13783736526966095, -0.28408941626548767, 0.09319933503866196, 0.1468048095703125, 0.4258978068828583, 0.07082612812519073, 0.21173226833343506, 0.1341327726840973, -0.3408617377281189, -0.12266618758440018, -0.2233576625585556, 0.6217302680015564, 0.27249765396118164, 0.1529465615749359, 0.09791581332683563, 0.28170090913772583, -0.21737393736839294, -0.17630137503147125, -0.09220089018344879, 0.029144108295440674, -0.06997984647750854, -0.42138606309890747, -0.37580591440200806, 0.572821855545044, 0.036477282643318176, -0.02153879962861538, 0.15765058994293213, 0.1421733796596527, -0.2886929214000702, 0.5798212885856628, 0.24263054132461548, 1.1859325170516968, -0.2814682126045227, 0.3137073814868927, -0.035962484776973724, -0.36061760783195496, 0.5188566446304321, -0.20172779262065887, -0.277855783700943, -0.49516627192497253, 0.3837963044643402, -0.04490651935338974, -0.18553084135055542, 0.3748290240764618, 0.3764258027076721, -0.03015538863837719, 0.12418948113918304, 0.1666647493839264, -0.12160614877939224, 0.05953516438603401, 0.47545504570007324, -0.31231001019477844, -0.346975713968277, -0.15084673464298248, -0.07249194383621216, -0.14247480034828186, 0.23283201456069946, 0.1278478056192398, -0.45531776547431946, -0.36892130970954895, 0.06390155106782913, -0.25991764664649963, 0.10600852966308594, -0.2166532427072525, -0.2464059293270111, -0.2637866735458374, -0.63893723487854, 0.4127871096134186, 0.36399340629577637, 0.22060680389404297, -0.047841064631938934, -0.210903137922287, 0.0887877345085144, -0.36310532689094543, -0.18496036529541016, 0.12916572391986847, -0.1481269896030426, 0.030227115377783775, -0.14576275646686554, -0.11701998859643936, 0.03543075546622276, 0.11575961858034134, -0.17766128480434418, -0.019819434732198715, -0.19169588387012482, 0.4527118504047394, -0.2664206326007843, -0.10804031044244766, 0.19887083768844604, -0.2783835530281067, 0.08723215758800507, 0.03405510634183884, 0.31136298179626465, 0.049848586320877075, -0.07525105029344559, 0.27844658493995667, -0.2262180894613266, 0.027817536145448685, 0.2285946160554886, 0.16525062918663025, -0.26487216353416443, 0.4430118799209595, 0.4258747100830078, -0.09706965833902359, -0.2546086013317108, 0.4642239809036255, 0.5814704298973083, -0.8045467734336853, -0.02388329990208149, -0.05021874979138374, 0.16765931248664856, 0.0249482411891222, 0.11201559007167816, -0.013125666417181492, -0.099068783223629, 0.03518941253423691, -0.4654947817325592, -0.3229875862598419, 0.08474715054035187, -0.05763579532504082, 0.10016078501939774, 0.335186630487442, 0.10849271714687347, 0.05705282464623451, 0.06976959854364395, -0.19205889105796814, 0.1754729151725769, 0.6514855623245239, 0.16727949678897858, 0.12899935245513916, 0.04435737431049347, 0.058668963611125946, -0.010351662524044514, -0.029424183070659637, 0.15425308048725128, 0.033484093844890594, -0.14044460654258728, -0.05531332269310951, 0.22434094548225403, -0.11151164025068283, -0.3583466410636902, -0.30706512928009033, -0.3529406189918518, -0.050478603690862656, -0.14498578011989594, -0.04875093698501587, -0.062427859753370285, -0.04851466417312622, 0.11967772990465164, 0.0911669135093689, -0.06606385856866837, -0.5188185572624207, 0.16587643325328827, -0.12332671135663986, 0.2130279839038849, -0.027788661420345306, 0.1678665429353714, -0.014582034200429916, -0.09240489453077316, 0.06336076557636261, 0.01959112100303173, 0.347075879573822, 0.20046520233154297, 0.33112630248069763, -0.11110370606184006, -0.01902996562421322, -0.20461267232894897, 0.3492090404033661, 0.14204396307468414, 0.04396390542387962, 0.04862743616104126, 0.323445588350296, 0.06121768057346344, -0.19304832816123962, 0.24553272128105164, -0.17272023856639862, -0.18861545622348785, -0.2519421875476837, 0.0988527461886406, 0.03157607093453407, 0.10778257250785828, -0.4496344029903412, -0.23619262874126434, 0.17203691601753235, 0.06814947724342346, -0.25897571444511414, 0.27036505937576294, 0.14384585618972778, -0.039956629276275635, -0.00792310293763876, 0.2794288992881775, 0.229250967502594, -0.08278076350688934, -0.14109373092651367, 0.17375165224075317, 0.6221064329147339, 0.3592700660228729, 0.29032203555107117, -0.23674069344997406, 0.06427210569381714, 0.3366849422454834, 0.1842469573020935, 0.03932708874344826, 0.204832524061203, 0.6484084129333496, 0.5068119764328003, -0.07449222356081009, 0.11960969120264053, 0.29356834292411804, 0.04900948703289032, 0.030896782875061035, -0.25908705592155457, -0.08743603527545929, -0.12445302307605743, -0.19977816939353943, -0.26839762926101685, -0.15598897635936737, -0.07701225578784943, 0.3239074647426605, 0.19428832828998566, -0.0283339936286211, 0.02810056507587433, 0.09170117974281311, 0.3089795708656311, -0.17272549867630005, 0.022006742656230927, 0.0015089386142790318, 0.010933786630630493, 0.04089725762605667, 0.43185797333717346, 0.2473565936088562, 0.09488633275032043, -0.12778674066066742, 0.22882217168807983, 0.026228975504636765, -0.09294770658016205, 0.21523043513298035, 0.1996612250804901, 0.3968442380428314, 0.035862311720848083, 0.1872345358133316, 0.03754758462309837, 0.06753217428922653, -0.055728621780872345, 0.18198072910308838, -0.1553834229707718, -0.07866616547107697, -0.15033338963985443, -0.05152933672070503, -0.39149346947669983, -0.19860796630382538, 0.06892616301774979, -0.17805670201778412, 0.2038513869047165, 0.5880889296531677, -0.2470197230577469, -0.049936551600694656, -0.16663649678230286, 0.00043435677071101964, -0.03191451355814934, 0.45689624547958374, 0.3662307858467102, 0.14163070917129517, -0.30244672298431396, -0.32491809129714966, -0.13995331525802612, 0.09046222269535065, -0.014410022646188736, -0.09643380343914032, -0.05399129167199135, -0.05557712912559509, 0.22304925322532654, 0.45404690504074097, 0.023301154375076294, 0.17148005962371826, -0.12571409344673157, 0.4537196159362793, -0.2217678725719452, -0.010113036260008812, 0.37255820631980896, 0.5218808650970459, -0.10128213465213776, -0.1328435242176056, 0.25423502922058105, 0.5032427906990051, -0.09317538142204285, -0.1573285609483719, 0.05690676346421242, 0.2313569337129593, -0.2506233751773834, 0.20016835629940033, 0.03537242114543915, 0.13309994339942932, -0.0070617590099573135, -0.15031686425209045, 0.09967019408941269, -0.02799469605088234, 0.059136297553777695, 0.6108745336532593, 0.2494727522134781, -0.10200589150190353, -0.21705736219882965, -0.11341239511966705, -0.055787570774555206, 0.2859750986099243, -0.2583238482475281, -0.12872764468193054, -0.3091915547847748, -0.06209133565425873, -0.2419230043888092, 0.02537052147090435, 0.0015791910700500011, 0.40263405442237854, -0.15525564551353455, 0.06434890627861023, -0.16876159608364105, -0.047382861375808716, 0.22028057277202606, -0.310128390789032, 0.08213496208190918, 0.3282664716243744, 0.1244424432516098, 0.6768011450767517, 0.10875971615314484, -0.5213271379470825, 0.19877088069915771, 0.27414005994796753, 0.004108988214284182, 0.21930505335330963, 0.09368600696325302, 0.0532282255589962, 0.07330287992954254, -0.03502950444817543, 0.4586234390735626, -0.029533665627241135, -0.029490873217582703, -0.4290834963321686, -0.43094947934150696 ]
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"
[ 0.1898328959941864, -0.40817776322364807, -0.02667149342596531, -0.07893203943967819, 0.1755460500717163, -0.1144011914730072, 0.2933265268802643, 0.23464733362197876, -0.11313798278570175, -0.006701200734823942, -0.07507722824811935, 0.5407453179359436, -0.25638601183891296, 0.21988412737846375, 0.0012711382005363703, -0.44989344477653503, 0.16272671520709991, 0.13243570923805237, -0.12136063724756241, -0.11142580211162567, -0.2248261570930481, -0.1842452436685562, -0.25302553176879883, -0.10311716049909592, -0.15764793753623962, 0.06863249838352203, -0.36664918065071106, -0.216035857796669, -0.37540432810783386, -0.3336399793624878, -0.19366517663002014, 0.2823522388935089, -0.2183115929365158, -0.036677248775959015, -0.00012514082482084632, -0.40157490968704224, -0.17735908925533295, -0.1747858077287674, -0.1102391928434372, -0.11791368573904037, -0.8842167854309082, -0.007036612834781408, 0.3844006657600403, 0.0185866579413414, 0.007791238371282816, 0.026044586673378944, 0.02693786658346653, -0.30778035521507263, 0.4810141921043396, -0.008772265166044235, 0.07200630009174347, -0.0372217632830143, 0.03685816377401352, 0.0641988143324852, 0.13089856505393982, 0.5086036324501038, 0.018672777339816093, -0.04318325221538544, 0.6300453543663025, 0.003554986324161291, -0.2057005763053894, 0.36134153604507446, 0.09690237045288086, 0.10612979531288147, 0.34404969215393066, -0.026201492175459862, -0.05275880545377731, -0.08939729630947113, -0.23088888823986053, 0.35122114419937134, 0.7839897274971008, -0.09999565780162811, -0.2693171501159668, -0.03707214444875717, 0.021386433392763138, 0.06871107220649719, -0.18641166388988495, 0.014609151519834995, -0.14011649787425995, -0.06115402653813362, -0.48590773344039917, -0.3253765106201172, -0.35618308186531067, 0.06938890367746353, -0.276447057723999, 0.42656177282333374, 0.008110176771879196, -0.03517892211675644, -0.05221103876829147, -0.27256423234939575, -0.021641450002789497, -0.4392715394496918, 0.39041706919670105, 0.14208759367465973, 0.10635891556739807, -0.3791533410549164, -0.18745829164981842, 0.07733555138111115, 0.3710348904132843, -0.2946627140045166, -0.09317933768033981, -0.0026955606881529093, -0.5204652547836304, -0.01247218158096075, 0.5711629986763, 0.002249016659334302, -0.04572524130344391, 0.07216694951057434, -0.34575361013412476, -0.29139578342437744, -0.2567962408065796, 0.041647475212812424, 0.1252155900001526, 0.22841838002204895, 0.07164231687784195, 0.12182885408401489, 0.05681445077061653, -0.3215623199939728, 0.06382258236408234, 0.07790414988994598, -0.3623918294906616, -0.2042142003774643, 0.03367449343204498, 0.15451613068580627, -0.08497214317321777, 0.5170228481292725, -0.19649603962898254, 0.20776647329330444, 0.08788136392831802, -0.3568958044052124, 0.13895876705646515, -0.10198044776916504, -0.5272908210754395, -0.2540569603443146, 0.19491809606552124, 0.2871019244194031, -0.01082152221351862, 0.3692452311515808, 0.37839576601982117, -0.2300276756286621, 0.07069496065378189, 0.02496250718832016, 0.5642063617706299, -0.183881476521492, -0.35279956459999084, 0.32675501704216003, 0.2906138300895691, 0.3297918736934662, -0.08868283778429031, 0.5011613965034485, -0.21087953448295593, -0.014086783863604069, -0.1533489227294922, 0.017584243789315224, 0.21666359901428223, -0.2115970402956009, 0.041118454188108444, 0.35120129585266113, 0.04503434896469116, -0.33556774258613586, 0.13014274835586548, -0.534650981426239, 0.04989206790924072, -0.29566827416419983, 0.0025753406807780266, 0.4049500524997711, -0.46270573139190674, 0.09012211859226227, -0.19625216722488403, 0.2359759658575058, 0.22379375994205475, 0.26253271102905273, -0.0753692090511322, -0.059609249234199524, -0.011268201284110546, 0.3569812476634979, 0.3155229985713959, -0.16380919516086578, -0.31837141513824463, 0.22978603839874268, 0.16081295907497406, 0.11740323901176453, -0.03637674078345299, 0.2344701588153839, 0.15473651885986328, 0.2690248489379883, 0.05115582421422005, -0.1387479305267334, -0.3022443652153015, 0.36625227332115173, -0.01800118014216423, -0.04290350154042244, 0.12097160518169403, -0.04484026879072189, -0.07994233071804047, -0.14226503670215607, -0.08018450438976288, 0.11475399136543274, 0.2186724692583084, -0.20548707246780396, -0.05747031047940254, -0.42732083797454834, 0.1527351438999176, 0.02900863066315651, -0.12003213912248611, -0.17264291644096375, -0.6358428597450256, -0.07989703118801117, -0.02884705364704132, -0.18700368702411652, 0.22696974873542786, -0.2735063135623932, -0.25807976722717285, 0.05828816071152687, -0.3363613188266754, 0.104703888297081, -0.01362052746117115, 0.020014598965644836, -0.24578604102134705, -0.2998508810997009, -0.2771869897842407, -0.06258732825517654, 0.5232808589935303, 0.018669789656996727, 0.05305277928709984, 0.11633952707052231, 0.09669694304466248, 0.002988656284287572, 0.13223695755004883, 0.35577648878097534, 0.19228725135326385, 0.25239166617393494, -0.03136174753308296, 0.23702430725097656, 0.4595884382724762, -0.04832523688673973, -0.03865768387913704, 0.11578782647848129, 0.40569859743118286, -0.23332004249095917, 0.32311803102493286, 0.17757369577884674, 0.0954631045460701, -0.013860391452908516, -0.8099333047866821, 0.18935805559158325, -0.0785285159945488, 0.11697987467050552, -0.14940430223941803, 0.2087828814983368, -0.19297218322753906, 0.2378270924091339, -0.06035572662949562, -0.47876113653182983, -0.12163158506155014, 0.2656581401824951, 0.150381937623024, 0.07501398772001266, -0.4771556258201599, 0.27814120054244995, 0.37942156195640564, -0.017873719334602356, 0.22816409170627594, -0.060224857181310654, -0.017266856506466866, -0.09874337911605835, 0.419706255197525, -0.21944653987884521, 0.23195888102054596, 0.1562119722366333, -0.06431928277015686, -0.14587189257144928, 0.008576720952987671, -0.12822511792182922, -0.0028877095319330692, -0.03820943087339401, -0.10356057435274124, 0.10288222134113312, 0.19065603613853455, -0.10545210540294647, 0.002276570303365588, 0.07329034805297852, -0.056763045489788055, -0.0949905514717102, -0.12194759398698807, 0.005623327102512121, -0.5228878259658813, -0.016359243541955948, -0.15502750873565674, -0.5842820405960083, -0.19491054117679596, -0.18189506232738495, 0.20081689953804016, -0.19690635800361633, -0.08773297816514969, 0.00636077718809247, 0.08728343993425369, 0.35097020864486694, 0.13916312158107758, 0.09050795435905457, 0.07922901958227158, -0.01744181104004383, 0.06981422007083893, 0.04823223873972893, 0.004030668176710606, 0.03277864307165146, 0.19473369419574738, 0.10559020191431046, 0.011595703661441803, -0.35372406244277954, 0.015375043265521526, 0.19428376853466034, -0.04946032166481018, 0.01033822726458311, 0.3358047902584076, 0.08736928552389145, -0.49544548988342285, 0.1399180293083191, 0.14058950543403625, 0.3530597984790802, -0.052326515316963196, 0.03083498775959015, 0.34738466143608093, 0.01778026856482029, 0.09395428001880646, -0.34880098700523376, -0.1854475885629654, 0.04607262462377548, 0.09167956560850143, -0.024083824828267097, 0.10844965279102325, -0.2396625578403473, 0.17827793955802917, -0.127211794257164, -0.34878525137901306, -0.24792645871639252, -0.2057136595249176, 0.11752516031265259, 0.04890201985836029, -0.14117518067359924, -0.01738825812935829, -0.21726755797863007, 0.14096352458000183, -0.13892069458961487, 0.18003645539283752, 0.27049022912979126, -0.2778679430484772, -0.045474905520677567, 0.12244544178247452, 0.0982368066906929, 0.4216066598892212, 0.30288514494895935, 0.39184486865997314, 0.08133646100759506, -0.07759655267000198, 0.10318748652935028, 0.110087089240551, 0.3211500644683838, 0.009788088500499725, 0.14529219269752502, 0.04953550174832344, -0.13081207871437073, 0.785081148147583, 0.4763053357601166, 0.144790381193161, 0.11252434551715851, 0.002753816545009613, -0.162567600607872, -0.13791371881961823, -0.10453391820192337, 0.2698311507701874, -0.07678443938493729, 0.19053776562213898, 0.4577789902687073, 0.08809155225753784, -0.06342189759016037, 0.035953644663095474, 0.29231885075569153, -0.23634709417819977, -0.2863631546497345, 0.42527368664741516, -0.17180883884429932, 0.25057095289230347, 0.014177893288433552, -0.11715614795684814, -0.25023069977760315, -0.12013711035251617, -0.06141102313995361, -0.11164262145757675, 0.34356552362442017, -0.3986707627773285, -0.09759148210287094, -0.2561989426612854, -0.15704847872257233, 0.29891884326934814, 0.4773337244987488, 0.10383447259664536, -0.013508148491382599, -0.07692836970090866, 0.16502925753593445, 0.3424926996231079, 0.5172504782676697, -0.03934943675994873, -0.11622600257396698, -0.09208124876022339, 0.05629943311214447, 0.17325063049793243, -0.0368112288415432, -0.6540710926055908, 0.32207489013671875, -0.29261514544487, 0.1334904581308365, -0.22548885643482208, -0.35192808508872986, 0.1718117594718933, 0.14538055658340454, -0.10473306477069855, -0.22611115872859955, 0.1381574422121048, 0.5807725787162781, 0.0008972196956165135, 0.2521328330039978, 0.2287890762090683, -0.1592705398797989, 0.19195649027824402, -0.09325671941041946, 0.18506544828414917, -0.006664315238595009, 0.3489796221256256, 0.21433040499687195, 0.06443074345588684, 0.16485701501369476, -0.3918600082397461, 0.22942380607128143, 0.15006227791309357, 0.1957373470067978, 0.20301273465156555, -0.6772516965866089, -0.060861025005578995, 0.08496491611003876, -0.11696657538414001, 0.6384945511817932, 0.31108519434928894, 0.2727362811565399, -0.08074880391359329, 0.09046327322721481, -0.01485990546643734, -0.18671004474163055, -0.07817531377077103, 0.30363425612449646, 0.08140745759010315, -0.2078102082014084, -0.4161689281463623, 0.3839748799800873, -0.11780275404453278, -0.2680470645427704, -0.13686561584472656, 0.5329189300537109, 0.0063745505176484585, 0.03252529352903366, 0.4108950197696686, 0.8881411552429199, -0.18266122043132782, 0.3090890944004059, 0.1333586573600769, 0.07515893876552582, 0.37113064527511597, -0.13379496335983276, 0.22359661757946014, -0.22098952531814575, 0.105465367436409, -0.184381902217865, -0.19609881937503815, -0.0003089199017267674, 0.5590389370918274, -0.6933908462524414, 0.27384448051452637, -0.15490415692329407, 0.1590493619441986, 0.14156140387058258, -0.062247004359960556, 0.31148838996887207, -0.13656528294086456, -0.1344509869813919, -0.012796035036444664, -0.2086646407842636, -0.13177864253520966, -0.15400856733322144, -0.04845895618200302, 0.03196616470813751, 0.0056092399172484875, -0.314815878868103, -0.06546628475189209, -0.5026322603225708, -0.2293134182691574, 0.0704682245850563, -0.1327109932899475, 0.49281883239746094, 0.47302383184432983, 0.11012998968362808, 0.1263779103755951, -0.08461256325244904, -0.1857496201992035, 0.13388442993164062, 0.2864261269569397, -0.2314491719007492, 0.1073823869228363, 0.5054354071617126, -0.02876114659011364, -0.3097285032272339, 0.236153706908226, -0.12689204514026642, -0.08921203017234802, -0.48502448201179504, 0.11372116208076477, 0.7107095718383789, -0.5381377339363098, -0.42165669798851013, -0.03158758953213692, 0.12211205065250397, 0.028233323246240616, -0.08895735442638397, 0.2638346254825592, 0.1101144477725029, 0.5779939889907837, 0.2360902726650238, 0.15750199556350708, 0.24454274773597717, -0.13318347930908203, 0.14268732070922852, -0.1016160175204277, 0.31638839840888977, 0.20700940489768982, -0.14301562309265137, -0.13768133521080017, 0.08183923363685608, 0.2291908711194992, -0.10338243842124939, 0.1376062035560608, -0.25277179479599, -0.08193712681531906, -0.03888688236474991, 0.33960390090942383, 0.3437860310077667, 0.3037855923175812, -0.00930476188659668, 0.04974672943353653, -0.11958934366703033, 0.2682492733001709, -0.22673341631889343, 0.17152810096740723, 0.19719453155994415, -0.15251611173152924, 0.06678895652294159, 0.2607317864894867, -0.176861971616745, -0.09415063261985779, 0.11468587070703506, 0.39735743403434753, -0.0025258820969611406, -0.20133469998836517, -0.225072979927063, -0.051654648035764694, -0.04072563722729683, -0.22952792048454285, -0.22013932466506958, -0.12144701927900314, -0.1096401959657669, 0.13896623253822327, 0.09552156180143356, -0.07645709067583084, -0.1243968978524208, -0.12791970372200012, -0.008751094341278076, -0.1249360516667366, -0.002041965490207076, -0.06993529945611954, -0.1673055738210678, 0.33069732785224915, 0.13859733939170837, -0.3417626917362213, -0.19411855936050415, 0.12415013462305069, 0.32734936475753784, -0.04631219059228897, 0.17618317902088165, 0.1268652081489563, 0.06882394105195999, -0.219927579164505, -0.0735199972987175, 0.43982812762260437, 0.27946895360946655, 0.0344991572201252, 0.02815595455467701, 0.0986635610461235, 0.16357477009296417, 0.0910554900765419, 0.2928510308265686, 0.058994486927986145, 0.048106785863637924, 0.09166964143514633, 0.04486198350787163, 0.08739276975393295, -0.10868413746356964, -0.21807754039764404, 0.5362935662269592, 0.1088518500328064, 0.16727374494075775, 0.3539818227291107, 0.05867838114500046, 0.01752348430454731, 0.3759634494781494, -0.028894320130348206, -0.35815808176994324, 0.06373865157365799, 0.014865760691463947, 0.616476833820343, -0.09675900638103485, 0.06004120782017708, 0.15797975659370422, 0.2112785428762436, 0.0631435215473175, -0.003104482078924775, 0.07378467917442322, 0.2152068167924881, 0.11956275254487991, -0.13529030978679657, 0.498638778924942, -0.016309067606925964, 0.2160298377275467, 0.22185900807380676, 0.10694853961467743, 0.4482228755950928, 0.351382315158844, 0.13832245767116547, -0.2611645758152008, -0.5070873498916626, -0.005821362137794495, 0.0829538032412529, -0.14927980303764343, -0.18763846158981323, -0.023021288216114044, -0.22036422789096832, -0.340963751077652, -0.030820438638329506, -0.6939731240272522, -0.13618916273117065, 0.4685724973678589, 0.32330089807510376, 0.5033947229385376, -0.12030553072690964, -0.049626607447862625, 0.10586351901292801, 0.29202643036842346, -0.12352419644594193, 0.185537651181221, 0.033744946122169495, 0.027988547459244728, 0.10734589397907257, 0.3419780433177948, -0.0009042052552103996, 0.2166770100593567, -0.18185116350650787, -0.03196090832352638, -0.33618834614753723, -0.08698032051324844, 0.2088312804698944, -0.2359447181224823, 0.22233763337135315, -0.07854956388473511, 0.15231919288635254, 0.01135796308517456, 0.05159815400838852, 0.15603023767471313, 0.06995556503534317, 0.006510213017463684, -0.23569224774837494, -0.5836463570594788, 0.21649818122386932, -0.2501598000526428, 0.08592181652784348, -0.031399164348840714, -0.6951836943626404, -0.0269295834004879, 0.05683054402470589, -0.10170410573482513, -0.0800093337893486, 0.14819765090942383, -0.012619532644748688, 0.2639652192592621, -0.029806356877088547, 0.36169660091400146, 0.014449702575802803, -0.32095715403556824, -0.17841561138629913, -0.2791803181171417, -0.11175759136676788, 0.2492118775844574, -0.12218441814184189, 0.07240936160087585, -0.021377718076109886, 0.05432356894016266, 0.3683212399482727, 0.055671583861112595, -0.579714298248291, -0.25978296995162964, 0.4736153781414032, -0.1560896337032318, 0.16441188752651215, 0.28888213634490967, 0.2590022385120392, -0.028212418779730797, -0.06761832535266876, 0.27827513217926025, 0.31076955795288086, -0.151582270860672, -0.2304987758398056, 0.01073895301669836, -0.2494553029537201, -0.0599178783595562, -0.22761648893356323, 0.36596617102622986, 0.2763817012310028, 0.10707353800535202, -0.20322714745998383, 0.2864724397659302, 0.4057307839393616, 0.11927694082260132, 0.09432689845561981, 0.11116787046194077, 0.602498471736908, -0.149407297372818, 0.2883364260196686, -0.054548025131225586, 0.0235117357224226, 0.0980963408946991, -0.1638515591621399, -0.6133045554161072, -0.02244398556649685, -0.023944303393363953, -0.32011786103248596, 0.13848400115966797, 0.41740289330482483, -0.12012782692909241, 0.18614065647125244, -0.24419823288917542, 0.2617799639701843, 0.13630349934101105, -0.22383342683315277, -0.34966400265693665, -0.30819395184516907, -0.04592224955558777, -0.24816520512104034, -0.014777283184230328, -0.7189198136329651, -0.2410823255777359, 0.23456346988677979, -0.115532785654068, -0.14388933777809143, 0.04211654141545296, -0.039306640625, -0.16625680029392242, -0.3629513382911682, 0.6250200271606445, 0.036941640079021454, 0.10133054107427597, -0.31193244457244873, -0.4519558846950531 ]
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) ```
[ 0.19923114776611328, -0.5758008360862732, 0.010819713585078716, -0.06993614882230759, 0.2859189808368683, -0.15366771817207336, 0.24460150301456451, 0.24167515337467194, -0.009871386922895908, 0.018018784001469612, -0.15490837395191193, 0.4558006525039673, -0.2713589370250702, 0.0753546729683876, 0.1255870908498764, -0.37995222210884094, 0.15262757241725922, 0.06899327039718628, -0.20162361860275269, -0.06752035021781921, -0.11300849169492722, 0.02811501920223236, -0.1665603518486023, -0.020647400990128517, -0.35132768750190735, 0.046458084136247635, -0.3601537346839905, -0.06562943011522293, -0.3025084137916565, -0.3981378674507141, -0.13272470235824585, 0.19807398319244385, -0.25676804780960083, 0.1135844886302948, -0.00013058973127044737, -0.31119444966316223, -0.21208475530147552, -0.21448419988155365, -0.09534190595149994, -0.03154377266764641, -0.7208172082901001, -0.011710870079696178, 0.3938955068588257, 0.10779350996017456, -0.02506088837981224, -0.02440764755010605, 0.0035536314826458693, -0.1321895867586136, 0.6255211234092712, -0.020353540778160095, 0.009229231625795364, 0.04004723206162453, 0.08175057172775269, 0.09900174289941788, 0.16138464212417603, 0.5654873847961426, -0.07321750372648239, 0.080771803855896, 0.5695477724075317, -0.08953171968460083, -0.2318437695503235, 0.5064967274665833, 0.03164707496762276, 0.0750545784831047, 0.2941821217536926, 0.11394388973712921, -0.015605799853801727, -0.08051545917987823, -0.22278165817260742, 0.4199117124080658, 0.571357011795044, -0.30700618028640747, -0.4449504613876343, -0.281170517206192, 0.07755253463983536, -0.053309254348278046, -0.18529577553272247, 0.04425617679953575, -0.2463555634021759, 0.0007777917198836803, -0.38704797625541687, -0.14347678422927856, -0.2480076104402542, 0.03583585470914841, -0.26923486590385437, 0.5118802189826965, -0.05074482038617134, 0.021875951439142227, -0.006453394424170256, -0.2313464879989624, -0.15191003680229187, -0.3797208070755005, 0.30629488825798035, 0.20800712704658508, -0.04424024373292923, -0.39680713415145874, -0.13806238770484924, -0.0766853615641594, 0.4508572816848755, -0.2433418333530426, -0.008573461323976517, 0.023109832778573036, -0.4934563934803009, -0.02274192124605179, 0.6486937403678894, 0.041839729994535446, -0.01199435442686081, -0.002554558450356126, -0.23035407066345215, -0.08692923933267593, -0.10430467128753662, 0.06293795257806778, 0.14950495958328247, 0.14896292984485626, 0.12328296154737473, 0.1073073223233223, 0.02461306005716324, -0.26230359077453613, -0.03937047719955444, 0.012350407429039478, -0.3285461664199829, -0.13844552636146545, 0.13970732688903809, 0.24745838344097137, -0.12764668464660645, 0.5036807060241699, -0.12741461396217346, 0.30862611532211304, -0.009251818060874939, -0.09967133402824402, 0.09143590182065964, -0.1582871675491333, -0.4034562110900879, -0.10639651119709015, 0.21177910268306732, 0.3301984667778015, 0.09043759852647781, 0.4027233421802521, 0.5290734767913818, -0.26109418272972107, 0.021614015102386475, 0.014538628980517387, 0.48413676023483276, 0.02228274755179882, -0.33286741375923157, 0.4092327654361725, 0.35748258233070374, 0.27040785551071167, -0.11678227037191391, 0.37184447050094604, -0.17706039547920227, -0.0374804362654686, -0.016465725377202034, -0.05915630981326103, 0.12640196084976196, -0.168894425034523, -0.1566007137298584, 0.36347872018814087, 0.12413905560970306, -0.1976219117641449, 0.0890340581536293, -0.5058836340904236, -0.01267828419804573, -0.1831262856721878, 0.1596052050590515, 0.5324587225914001, -0.28719890117645264, -0.14317451417446136, -0.02312348037958145, 0.14544817805290222, 0.07706060260534286, 0.36676445603370667, -0.08329937607049942, -0.09566771239042282, -0.028542716056108475, 0.40206536650657654, 0.20190654695034027, -0.2790754735469818, -0.2886228859424591, 0.18329240381717682, 0.320334255695343, 0.30419206619262695, -0.12321638315916061, 0.1302313506603241, 0.02096795290708542, 0.22887694835662842, 0.10841420292854309, -0.1621449887752533, -0.17627917230129242, 0.4218747913837433, -0.05920540168881416, -0.13189485669136047, 0.0792522206902504, -0.0365055613219738, -0.10871544480323792, -0.22112661600112915, -0.21198482811450958, 0.15718501806259155, 0.3730962574481964, -0.13576793670654297, -0.15733003616333008, -0.1915746033191681, 0.14541830122470856, -0.08411413431167603, -0.03216908872127533, -0.1841539740562439, -0.6730771660804749, 0.03187340870499611, -0.05882015824317932, -0.013976769521832466, -0.02356032468378544, -0.2371213138103485, -0.33111903071403503, 0.151394784450531, -0.3594209551811218, -0.06123194471001625, -0.06678392738103867, -0.011039607226848602, -0.09729558974504471, -0.2155425101518631, -0.23209547996520996, 0.05262404680252075, 0.48802900314331055, 0.1118907481431961, -0.011933110654354095, 0.19364570081233978, 0.19373191893100739, -0.12392769753932953, 0.06306754797697067, 0.3638550937175751, 0.238555446267128, 0.18986482918262482, 0.015624402090907097, 0.2106880247592926, 0.5333712697029114, -0.07160799205303192, -0.12715278565883636, 0.23659703135490417, 0.45395714044570923, -0.1962524801492691, 0.15113243460655212, 0.20681224763393402, 0.19266007840633392, -0.04783991351723671, -0.7790425419807434, 0.2532618045806885, 0.013713807798922062, 0.2599428594112396, -0.09325232356786728, 0.19287718832492828, -0.07523894309997559, 0.33153554797172546, -0.05952591076493263, -0.39366692304611206, -0.20147693157196045, 0.12461234629154205, 0.14804990589618683, -0.02956032194197178, -0.49889591336250305, 0.3210470676422119, 0.3738496005535126, -0.05215471237897873, 0.25738489627838135, -0.057632848620414734, -0.10829813778400421, -0.15855784714221954, 0.3733246624469757, -0.35355332493782043, 0.2838003933429718, 0.06916701793670654, -0.08368025720119476, -0.12158364802598953, -0.04242314398288727, -0.03475858271121979, 0.027157481759786606, 0.06743299961090088, -0.07191848009824753, 0.13305777311325073, 0.18482965230941772, -0.22043867409229279, -0.08031687140464783, -0.005193399731069803, -0.0877743810415268, -0.08955003321170807, -0.2064070701599121, 0.09672454744577408, -0.5124543905258179, 0.10381168127059937, -0.2801769971847534, -0.5547011494636536, -0.36573538184165955, -0.22905413806438446, 0.26497069001197815, -0.08991889655590057, -0.1655486524105072, 0.058604978024959564, 0.11672484874725342, 0.09931078553199768, 0.1955944448709488, -0.006008118856698275, -0.08944988995790482, 0.046014029532670975, 0.10184039175510406, -0.051307082176208496, -0.047684092074632645, -0.04785193130373955, 0.037648044526576996, 0.12882821261882782, -0.11320015043020248, -0.45335352420806885, -0.008424118161201477, 0.3111926317214966, -0.11082707345485687, 0.1549980491399765, 0.3773554265499115, 0.102915920317173, -0.4873550534248352, 0.04879263415932655, 0.21227015554904938, 0.2985992729663849, -0.0036081750877201557, 0.1128145232796669, 0.3003280460834503, -0.022947564721107483, -0.012343835085630417, -0.19157715141773224, -0.13206136226654053, -0.003897819435223937, 0.10018601268529892, 0.026955801993608475, 0.15248648822307587, -0.1670544147491455, 0.26794716715812683, -0.035998910665512085, -0.3355773389339447, -0.24153465032577515, -0.14004848897457123, 0.08696012943983078, 0.14672957360744476, -0.08437203615903854, -0.0522509329020977, -0.28386327624320984, 0.0482754111289978, -0.018866945058107376, 0.11450191587209702, 0.1831580400466919, -0.35383522510528564, -0.08702854067087173, -0.06288986653089523, 0.1746329814195633, 0.4848908483982086, 0.2857911288738251, 0.2647238075733185, 0.11941228806972504, -0.05617332085967064, -0.006127574481070042, 0.16391104459762573, 0.3485114872455597, 0.04665444791316986, 0.004912895616143942, -0.09494318813085556, 0.0010067186085507274, 1.0838125944137573, 0.4843839108943939, 0.24819226562976837, -0.0036785511765629053, -0.11817046254873276, -0.12630318105220795, -0.1567162573337555, -0.09003368765115738, 0.11810842901468277, -0.07367584109306335, 0.3027012348175049, 0.46514350175857544, 0.24424666166305542, 0.012060770764946938, 0.05346541851758957, 0.3858351409435272, -0.1621825098991394, -0.45231810212135315, 0.39658766984939575, -0.07453775405883789, 0.145846888422966, 0.018084444105625153, 0.06413005292415619, -0.18320025503635406, -0.12235718220472336, -0.07621140033006668, 0.047607164829969406, 0.3101416528224945, -0.19234783947467804, -0.040282778441905975, -0.2107449620962143, -0.25974375009536743, 0.28145676851272583, 0.5223879814147949, 0.1269291341304779, -0.0026250500231981277, -0.14970137178897858, 0.2083989530801773, 0.23716425895690918, 0.6356262564659119, 0.17853084206581116, -0.20981650054454803, -0.009059385396540165, -0.12181373685598373, -0.010507624596357346, -0.1275862753391266, -0.6342929005622864, 0.1838945895433426, -0.2725785970687866, 0.22634407877922058, -0.39750543236732483, -0.46626895666122437, -0.07813160866498947, 0.11758798360824585, -0.13534095883369446, -0.1077001765370369, -0.05650637298822403, 0.41584932804107666, -0.15804965794086456, 0.32520970702171326, 0.3609255254268646, -0.14264827966690063, 0.1712752878665924, -0.064505934715271, 0.22820799052715302, -0.05428944155573845, 0.3435716927051544, 0.33272430300712585, 0.13303756713867188, 0.09885752946138382, -0.3660648465156555, 0.22141103446483612, 0.08414465188980103, -0.009226267226040363, 0.32175785303115845, -0.5721667408943176, -0.2904278039932251, 0.14899942278862, -0.04958147928118706, 0.4967501163482666, 0.44231754541397095, 0.21683454513549805, -0.05097963660955429, 0.26420319080352783, 0.07970666885375977, -0.13998578488826752, 0.1115492731332779, 0.2770271599292755, 0.0872880071401596, -0.17061841487884521, -0.3983021676540375, 0.4744041860103607, -0.06491927802562714, -0.2802017331123352, 0.013689628802239895, 0.5781556367874146, -0.08784233033657074, -0.07902675122022629, 0.29351526498794556, 0.9808763861656189, -0.2313356250524521, 0.3998238742351532, 0.40193748474121094, 0.026256218552589417, 0.31943953037261963, -0.282145231962204, 0.26689913868904114, -0.22960396111011505, 0.11477553844451904, -0.12683045864105225, -0.189653217792511, 0.09107118844985962, 0.48549672961235046, -0.6954212188720703, 0.4039892554283142, -0.06320936232805252, 0.07957549393177032, 0.06719353049993515, -0.020808031782507896, 0.16373136639595032, -0.25643888115882874, -0.2649291455745697, -0.06672532856464386, -0.258674293756485, 0.0004027843242511153, -0.10235666483640671, -0.061161234974861145, -0.0066708954982459545, -0.12094172090291977, -0.3932930827140808, -0.014797128736972809, -0.45530563592910767, -0.17547591030597687, 0.29111671447753906, -0.15258732438087463, 0.39535677433013916, 0.4345223903656006, 0.0012732399627566338, 0.14556868374347687, -0.1255715787410736, -0.028710506856441498, -0.08400897681713104, 0.40583309531211853, -0.26161661744117737, 0.09170591086149216, 0.5665051341056824, -0.006685835774987936, -0.34218117594718933, 0.15743710100650787, -0.0809239074587822, -0.10101282596588135, -0.43462657928466797, 0.1806265413761139, 0.6312639713287354, -0.5566818714141846, -0.2709469199180603, 0.10570531338453293, 0.11234081536531448, -0.04159417748451233, -0.14434164762496948, 0.3904123604297638, -0.03776434808969498, 0.5050051808357239, 0.15269356966018677, 0.0753721222281456, 0.1519210934638977, -0.024606190621852875, 0.044686004519462585, -0.22016768157482147, 0.3574770987033844, 0.2970626652240753, -0.11576618999242783, -0.1050681620836258, -0.05706581473350525, 0.2073892056941986, -0.19489721953868866, 0.3421350419521332, -0.21411161124706268, -0.2222314178943634, -0.09522437304258347, 0.3001447319984436, 0.18653321266174316, 0.15205810964107513, 0.04173426330089569, -0.032338183373212814, -0.11665632575750351, 0.3286190330982208, -0.14338859915733337, 0.25442108511924744, 0.19086496531963348, -0.031950850039720535, 0.028394607827067375, 0.19577889144420624, -0.1289578527212143, -0.08749984204769135, 0.008275370113551617, 0.34924250841140747, -0.11323001235723495, -0.26551565527915955, -0.029821019619703293, 0.0035737380385398865, -0.03897583484649658, -0.09551606327295303, -0.2530173063278198, -0.06898950040340424, -0.25027474761009216, 0.16109591722488403, 0.047943372279405594, -0.12595921754837036, -0.026538383215665817, 0.03126094862818718, -0.02318301424384117, -0.2388312965631485, 0.19180476665496826, -0.08689827471971512, -0.14084838330745697, 0.3501247763633728, 0.007580718956887722, -0.21584482491016388, -0.13044041395187378, 0.02710779942572117, 0.1823073923587799, -0.035978205502033234, 0.33490628004074097, 0.15295854210853577, -0.035476136952638626, -0.23685063421726227, 0.03910638391971588, 0.4817042350769043, 0.15812788903713226, -0.01392139121890068, 0.07354774326086044, 0.08150297403335571, 0.07890521734952927, 0.017839275300502777, 0.2860039174556732, 0.008110882714390755, 0.07826101034879684, 0.0466495156288147, 0.14516781270503998, -0.006222786381840706, -0.10724987834692001, -0.29134559631347656, 0.45848000049591064, 0.10486217588186264, 0.20024657249450684, 0.40678220987319946, 0.017292706295847893, -0.003219063626602292, 0.3959061801433563, -0.053129151463508606, -0.27697765827178955, 0.06952836364507675, 0.011238818988204002, 0.6116731762886047, -0.1678517758846283, 0.12471406906843185, 0.07526926696300507, 0.2117461860179901, 0.17982648313045502, 0.03575935214757919, -0.18006429076194763, 0.20812657475471497, -0.1359260380268097, -0.15321247279644012, 0.5411540865898132, -0.13623884320259094, 0.014910691417753696, -0.00678713945671916, 0.021940533071756363, 0.459004670381546, 0.2312006950378418, 0.2578001320362091, -0.31634804606437683, -0.35048821568489075, 0.0007418016903102398, 0.10561441630125046, -0.11325452476739883, -0.19692261517047882, 0.20105038583278656, -0.11095329374074936, -0.30959588289260864, -0.06364303082227707, -0.5783079862594604, -0.3299182951450348, 0.48669689893722534, 0.22485007345676422, 0.5121278166770935, -0.16760654747486115, -0.11800265312194824, 0.1449575424194336, 0.36797526478767395, -0.15258802473545074, 0.15408699214458466, 0.30220746994018555, -0.03908358886837959, 0.018621627241373062, 0.20260365307331085, 0.08953496813774109, 0.13583548367023468, -0.20486095547676086, -0.014348790980875492, -0.3012125790119171, -0.02742961421608925, 0.18805719912052155, -0.15484298765659332, 0.18413937091827393, 0.026713434606790543, 0.2231922149658203, -0.0609477236866951, 0.06802735477685928, -0.11772585660219193, 0.03170761093497276, 0.1584734171628952, -0.18284845352172852, -0.3853932321071625, -0.013146483339369297, -0.17111201584339142, 0.04769596830010414, -0.0006279439548961818, -0.7032938003540039, -0.0632740929722786, 0.07264992594718933, -0.06937975436449051, -0.03406532108783722, 0.09923781454563141, -0.04312067851424217, 0.20425187051296234, 0.15444716811180115, 0.24497446417808533, 0.14178645610809326, -0.27186915278434753, 0.12096316367387772, -0.27578359842300415, -0.18676045536994934, 0.2845398187637329, -0.23186272382736206, 0.03978751227259636, 0.02078132890164852, 0.10187831521034241, 0.34108859300613403, 0.06822282820940018, -0.5816967487335205, -0.1138572245836258, 0.46572333574295044, -0.20126880705356598, 0.1362157016992569, 0.18447719514369965, 0.266882061958313, 0.10629048198461533, -0.1284310519695282, 0.21247434616088867, 0.3538288176059723, -0.21196262538433075, -0.17079640924930573, -0.0713018923997879, -0.3326978087425232, -0.058846890926361084, -0.24034546315670013, 0.5293840765953064, 0.20186176896095276, 0.12241372466087341, -0.2680336534976959, 0.14065146446228027, 0.28605449199676514, 0.056703679263591766, -0.02897876873612404, 0.027317937463521957, 0.7622185945510864, -0.09328214824199677, 0.10506929457187653, -0.11119569092988968, 0.21598084270954132, 0.12799905240535736, -0.3299758732318878, -0.4948563873767853, 0.08940299600362778, -0.025288475677371025, -0.23564685881137848, 0.10793591290712357, 0.48345842957496643, -0.12256097793579102, 0.1367386281490326, -0.2670828700065613, 0.1285669058561325, 0.3199920952320099, -0.2631490230560303, -0.3871842324733734, -0.39997541904449463, -0.07116082310676575, -0.035583384335041046, -0.032920677214860916, -0.8910265564918518, -0.17213229835033417, 0.11249084770679474, -0.03815918415784836, -0.253605455160141, 0.08522757142782211, -0.02235635183751583, -0.27161070704460144, -0.3777168095111847, 0.6619982719421387, 0.056897811591625214, 0.033962182700634, -0.22671565413475037, -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.
[ -0.2745409607887268, -0.07471389323472977, -0.2828442454338074, 0.08163276314735413, 0.10131540894508362, 0.08511558175086975, 0.1622699648141861, 0.1815718561410904, 0.15496589243412018, 0.028999315574765205, 0.09069543331861496, 0.2542612552642822, -0.24301405251026154, 0.09133557975292206, 0.25564438104629517, -0.3420153558254242, 0.20134352147579193, -0.07195044308900833, -0.24810552597045898, -0.3592264950275421, -0.10869905352592468, 0.09392467141151428, -0.26665619015693665, -0.21100138127803802, -0.13362126052379608, -0.06005677208304405, -0.24860863387584686, -0.04428120329976082, -0.42535167932510376, -0.23080092668533325, 0.06616629660129547, 0.2980559170246124, 0.23175840079784393, 0.35712164640426636, -0.00010253637447021902, -0.179146870970726, 0.19525428116321564, 0.0659598782658577, -0.09675730764865875, -0.3587574064731598, -0.3629663288593292, -0.266895592212677, 0.17846566438674927, 0.10846666991710663, 0.09927207231521606, 0.0903831422328949, -0.0391337089240551, -0.4096408486366272, 0.027926350012421608, 0.35471290349960327, 0.2387000173330307, 0.35676467418670654, 0.41479402780532837, -0.19605229794979095, 0.23967799544334412, 0.06696399301290512, -0.1821238398551941, 0.08796855062246323, 0.42116808891296387, 0.11522839218378067, 0.30466559529304504, 0.1468876153230667, 0.1961059421300888, 0.10997598618268967, 0.38002046942710876, 0.018322544172406197, 0.013787881471216679, -0.5149790644645691, -0.3262498676776886, 0.2852603495121002, 0.7866035103797913, -0.234496608376503, -0.23747438192367554, -0.10458064079284668, -0.03371288999915123, -0.2809472978115082, 0.03654810041189194, 0.031457655131816864, 0.2146727442741394, -0.16369494795799255, 0.06738591194152832, -0.40398886799812317, -0.3353690207004547, 0.40255019068717957, 0.13977502286434174, 0.4864601492881775, 0.12364175915718079, 0.12610043585300446, 0.09881314635276794, -0.14917509257793427, 0.39519381523132324, -0.0036636930890381336, 0.21630731225013733, 0.2663559317588806, -0.21953926980495453, -0.21650676429271698, -0.09907504916191101, -0.08511707931756973, 0.018533170223236084, -0.18484319746494293, -0.14839710295200348, 0.3597199320793152, -0.30942851305007935, 0.2582106590270996, 0.21343904733657837, -0.1501583307981491, -0.10434868931770325, -0.004755295813083649, 0.4263310730457306, -0.03369499370455742, 0.07132307440042496, 0.19717858731746674, -0.19616824388504028, 0.007611086592078209, -0.27579057216644287, 0.027275191619992256, 0.02952413819730282, -0.012821643613278866, 0.12252495437860489, -0.38320666551589966, 0.09562382102012634, -0.36199116706848145, -0.04778091982007027, -0.08453328907489777, -0.3333871066570282, 0.17141321301460266, -0.009538671933114529, 0.027227481827139854, -0.1498740166425705, -0.02315664477646351, -0.0725000873208046, 0.15157677233219147, -0.21522057056427002, 0.019892772659659386, 0.23342451453208923, -0.101899154484272, 0.08815053850412369, 0.012576698325574398, 0.20164772868156433, 0.155318483710289, 0.18760861456394196, 0.10861700773239136, -0.1332702785730362, 0.2291131615638733, -0.16917985677719116, 0.03806184604763985, -0.26937001943588257, 0.18624162673950195, -0.09615349769592285, -0.05022143945097923, -0.14967170357704163, -0.21687643229961395, -0.1686045080423355, 0.2626045346260071, -0.1451125293970108, -0.3540807068347931, -0.22397251427173615, 0.3776283860206604, -0.280754029750824, 0.2039029598236084, -0.17724822461605072, -0.1275351345539093, -0.09541001170873642, -0.3391963541507721, 0.005437586922198534, 0.07479571551084518, -0.41917309165000916, 0.15557457506656647, -0.13233238458633423, -0.0583365336060524, -0.23669745028018951, -0.054188381880521774, -0.19494986534118652, 0.015208465978503227, -0.09529165923595428, -0.12077133357524872, 0.45636507868766785, -0.5523671507835388, -0.06465821713209152, -0.1699967086315155, 0.22772063314914703, -0.22985945641994476, -0.1641162931919098, 0.19302169978618622, 0.03970059007406235, -0.14828254282474518, -0.20391404628753662, 0.3211025893688202, -0.1425626277923584, -0.2917487323284149, -0.20709408819675446, -0.11056310683488846, 0.05193530395627022, 0.11169508844614029, 0.3645804822444916, 0.017507199198007584, 0.16154718399047852, 0.21329596638679504, 0.01444335374981165, 0.1034305989742279, 0.03156667202711105, 0.20741118490695953, 0.41448691487312317, 0.21465051174163818, 0.46539106965065, -0.42616912722587585, -0.43927866220474243, 0.13807733356952667, 0.1953694224357605, 0.18815842270851135, 0.10498828440904617, 0.08624681830406189, -0.004376194439828396, 0.10294637829065323, -0.12055515497922897, 0.3178800940513611, 0.14472608268260956, -0.04055710509419441, 0.2105715125799179, -0.14622405171394348, -0.5968182682991028, 0.0926128476858139, -0.1041233241558075, 0.41401207447052, -0.4701022505760193, 0.3047417104244232, 0.015341708436608315, -0.15350067615509033, 0.1693393439054489, 0.244172602891922, 0.005732478573918343, -0.4595032334327698, 0.08032973855733871, 0.29646921157836914, 0.16137129068374634, 0.23572604358196259, -0.03380643576383591, 0.44494014978408813, 0.42598727345466614, -0.06091931089758873, 0.31197917461395264, -0.31688186526298523, -0.026507623493671417, -0.05693737417459488, 0.010205725207924843, 0.21473123133182526, -0.03721553459763527, -0.09473762661218643, 0.2684362232685089, 0.01656363345682621, 0.1671217679977417, 0.2664136290550232, -0.042163871228694916, -0.21579453349113464, 0.03004392236471176, -0.008427668362855911, -0.021350333467125893, -0.13367809355258942, -0.2969608008861542, 0.22897592186927795, 0.603425920009613, -0.12870314717292786, 0.03560444340109825, 0.06194150075316429, -0.24993275105953217, -0.008935476653277874, 0.09163692593574524, 0.2701999545097351, 0.19163952767848969, 0.14972522854804993, 0.21871419250965118, -0.018969422206282616, -0.2912890315055847, -0.12581434845924377, 0.03649473562836647, 0.1996532380580902, 0.14305704832077026, -0.06021953374147415, -0.013700060546398163, 0.0058994777500629425, -0.09948033094406128, -0.01922411471605301, -0.04581400007009506, 0.4715064465999603, 0.07942872494459152, -0.18269288539886475, -0.40998050570487976, -0.18739190697669983, -0.07155625522136688, -0.1772996485233307, -0.15517418086528778, -0.3065159320831299, 0.03282037377357483, -0.014600218273699284, 0.19896556437015533, 0.01642397604882717, 0.19826914370059967, 0.0992313027381897, -0.2671225368976593, 0.3675662875175476, -0.2708779573440552, 0.0022391665261238813, -0.2175019234418869, 0.19029958546161652, 0.25650930404663086, -0.08611151576042175, 0.41823428869247437, -0.03400939330458641, 0.23329392075538635, -0.1385355293750763, -0.6403663754463196, 0.17638801038265228, -0.14959028363227844, -0.02689189836382866, -0.020250003784894943, -0.0834307074546814, 0.17992787063121796, 0.11929910629987717, 0.07279115915298462, -0.2918449640274048, -0.051495641469955444, -0.04408704489469528, -0.197087362408638, -0.1112024337053299, -0.13188089430332184, -0.5581415295600891, -0.4593670666217804, -0.2616688013076782, 0.17642086744308472, 0.32629626989364624, 0.08081500977277756, 0.16245827078819275, 0.0841362476348877, 0.16439984738826752, -0.06314369291067123, 0.03947480767965317, 0.1127869188785553, -0.25843703746795654, 0.4067581593990326, -0.3071107864379883, -0.38439691066741943, 0.4961082339286804, -0.10421452671289444, 0.3492160439491272, 0.2127215713262558, -0.30541157722473145, -0.19868506491184235, -0.025309860706329346, -0.01599079556763172, 0.1663932502269745, -0.019606344401836395, 0.23422181606292725, -0.009280458092689514, 0.052656110376119614, -0.25241127610206604, -0.14401793479919434, -0.0861579030752182, 0.3769686222076416, 0.02944543957710266, 0.14342646300792694, 0.33475857973098755, -0.033224061131477356, 0.43053320050239563, 0.03644866496324539, -0.24522386491298676, 0.05040546506643295, 0.16883529722690582, 0.10867689549922943, 0.011381952092051506, -0.050217725336551666, 0.07712599635124207, 0.03132672980427742, 0.2990892231464386, 0.0489153228700161, 0.12088098376989365, -0.10746315121650696, -0.3140343129634857, 0.33713552355766296, -0.17844337224960327, -0.0018071294762194157, -0.0018807448213919997, -0.251693457365036, 0.3295738101005554, 0.08878365904092789, -0.19609875977039337, -0.28800493478775024, -0.14876814186573029, 0.15746676921844482, 0.35457471013069153, 0.24198202788829803, -0.12902610003948212, -0.537856936454773, -0.18528202176094055, -0.1731892079114914, 0.417476087808609, -0.10197830200195312, 0.17927807569503784, -0.19229264557361603, 0.11066799610853195, 0.11774320155382156, 0.0859801396727562, 0.1878301352262497, -0.47661346197128296, -0.5271234512329102, 0.2791646420955658, -0.13627272844314575, -0.2341882884502411, -0.024257078766822815, -0.3106382489204407, 0.11367221921682358, -0.029083291068673134, 0.44889023900032043, 0.015480734407901764, -0.3676788806915283, 0.2430873066186905, 0.13085678219795227, 0.08591002970933914, -0.14056414365768433, 0.26299378275871277, -0.010136370547115803, -0.11732549965381622, -0.26204460859298706, -0.105214424431324, -0.11233876645565033, -0.1167275458574295, -0.5168623328208923, 0.07123591005802155, -0.33272644877433777, 0.18476121127605438, -0.047048311680555344, -0.056590039283037186, 0.27492931485176086, 0.03683142364025116, 0.31681889295578003, -0.14537103474140167, 0.46666738390922546, -0.01151352096349001, -0.38040339946746826, -0.16965319216251373, 0.174065962433815, -0.09162192791700363, 0.19564731419086456, 0.15047921240329742, -0.05298185348510742, 0.1828361302614212, -0.0608179084956646, -0.031034231185913086, -0.12551723420619965, 0.16413475573062897, 0.4193401634693146, 0.22790886461734772, -0.28524813055992126, -0.6071879267692566, 0.11472314596176147, 0.016152488067746162, -0.048595864325761795, 0.42071545124053955, -0.13392095267772675, -0.174508199095726, 0.11098364740610123, 0.04105661064386368, 0.5863261818885803, -0.2631884515285492, -0.29278144240379333, 0.03474978357553482, -0.08629991114139557, 0.47325971722602844, -0.6111389994621277, 0.17423954606056213, -0.03426744416356087, 0.10870572924613953, -0.08511229604482651, -0.25487107038497925, 0.2109452188014984, 0.07227623462677002, -0.20136044919490814, 0.0995425283908844, 0.17709054052829742, 0.1380796581506729, -0.19421091675758362, 0.6597747802734375, -0.06312783062458038, -0.14037317037582397, -0.09899744391441345, 0.2516936957836151, -0.028308534994721413, 0.052182018756866455, -0.06026054918766022, -0.19137780368328094, -0.0318773128092289, 0.1811332255601883, -0.28495216369628906, -0.1687602549791336, -0.22131700813770294, 0.008319917134940624, 0.09131796658039093, -0.04329495504498482, 0.3468940258026123, 0.0962311401963234, 0.5915398597717285, -0.032024141401052475, -0.23760193586349487, 0.27973929047584534, -0.2682269215583801, -0.11788790673017502, -0.12288876622915268, 0.01720753312110901, 0.37216293811798096, -0.29674237966537476, -0.012082860805094242, -0.07735519856214523, -0.00027863544528372586, -0.24259774386882782, -0.36513465642929077, -0.07142263650894165, 0.32986247539520264, 0.09708577394485474, 0.19230249524116516, 0.3033458888530731, -0.09278284758329391, 0.007559990510344505, 0.24786654114723206, 0.4511789083480835, 0.014724146574735641, 0.4076642096042633, 0.04798503592610359, 0.0005297396564856172, -0.1513431817293167, -0.1505492925643921, 0.270860493183136, -0.2877451777458191, 0.07257485389709473, 0.4370303153991699, -0.19491270184516907, -0.20974591374397278, -0.19884449243545532, 0.18380042910575867, 0.05712978541851044, 0.07818335294723511, 0.15595902502536774, 0.09627259522676468, 0.1928562968969345, 0.09991782903671265, -0.023509211838245392, -0.41114306449890137, -0.4108960032463074, -0.293220192193985, -0.2243691235780716, 0.25665104389190674, 0.04946174472570419, 0.27945002913475037, -0.19637493789196014, -0.07223783433437347, 0.17411182820796967, 0.1493716686964035, -0.37918296456336975, 0.05701436847448349, 0.15133550763130188, 0.03618864715099335, 0.22612474858760834, -0.07250063121318817, -0.03612138330936432, -0.07173990458250046, 0.09386394172906876, -0.02794773317873478, -0.46794503927230835, -0.18472224473953247, 0.042026251554489136, 0.10021437704563141, -0.04263772442936897, -0.22871801257133484, 0.12667149305343628, -0.3141506016254425, -0.282792866230011, -0.024908829480409622, -0.03976767137646675, -0.30507704615592957, -0.1421205699443817, 0.33117493987083435, 0.1819978803396225, 0.17024774849414825, 0.558208167552948, 0.09513983130455017, 0.2124508023262024, 0.2048579603433609, 0.40218737721443176, 0.18952007591724396, 0.26010704040527344, -0.08211475610733032, -0.14819803833961487, 0.07816363126039505, -0.13820557296276093, -0.1794574111700058, 0.39828863739967346, -0.16925761103630066, 0.1535908579826355, 0.03235528990626335, 0.10143991559743881, 0.2642422318458557, 0.021073725074529648, 0.036964207887649536, 0.14142948389053345, 0.21311019361019135, -0.1344301700592041, -0.055899713188409805, 0.40270641446113586, 0.4979761242866516, 0.10187990963459015, 0.07686411589384079, 0.5620502829551697, 0.15833283960819244, 0.28498026728630066, 0.1785445213317871, 0.23345711827278137, -0.14217452704906464, 0.40867576003074646, -0.10305377095937729, -0.053733181208372116, 0.07344017922878265, 0.08991999179124832, 0.38149258494377136, 0.09467040747404099, 0.3289196193218231, 0.18405799567699432, 0.43973004817962646, 0.1460437774658203, -0.08007973432540894, 0.5982282161712646, -0.319163054227829, 0.0660490021109581, 0.1627112776041031, 0.060785915702581406, 0.09383033961057663, 0.07741843163967133, 0.4690661132335663, -0.009389854967594147, -0.13898444175720215, -0.05403550714254379, -0.11669953912496567, -0.1087820902466774, -0.005931607913225889, -0.041407510638237, -0.17655430734157562, -0.20635227859020233, -0.09554553031921387, 0.04245602712035179, -0.2446388155221939, 0.10841155797243118, -0.006783934775739908, 0.025074830278754234, -0.18302904069423676, 0.3499511182308197, -0.08808057010173798, -0.06958431750535965, 0.2082018405199051, 0.44161126017570496, -0.2254699170589447, -0.20558926463127136, -0.09230750799179077, -0.07757952809333801, -0.06999090313911438, 0.020393043756484985, -0.35975441336631775, 0.3094954490661621, 0.051701851189136505, -0.192958265542984, 0.20771794021129608, 0.4019233286380768, 0.07114540040493011, 0.3664080500602722, -0.11097875237464905, 0.15951359272003174, -0.0535556823015213, -0.022820642217993736, -0.2833409011363983, -0.021147089079022408, -0.41630250215530396, -0.08433759957551956, 0.1102449893951416, -0.15187132358551025, -0.04534957557916641, -0.04589666426181793, -0.4722164571285248, -0.5669897198677063, 0.5651453137397766, -0.1895698755979538, -0.15351346135139465, -0.25356122851371765, 0.1081458255648613, -0.09533147513866425, 0.2511746287345886, 0.12147358804941177, 0.4550127685070038, 0.05822407826781273, -0.3774459660053253, -0.2873684763908386, 0.005675969645380974, 0.14915886521339417, 0.28153520822525024, -0.11302042007446289, 0.19985251128673553, 0.24011631309986115, -0.03342708200216293, 0.43038812279701233, -0.2973405122756958, 0.09230455011129379, -0.004097955767065287, -0.32758742570877075, 0.009855608455836773, -0.2028946429491043, 0.04155029356479645, 0.11045961081981659, 0.2590784728527069, 0.07082255184650421, -0.26883020997047424, -0.025742627680301666, -0.4121481776237488, -0.023359287530183792, 0.33951470255851746, 0.08142480999231339, 0.42126867175102234, 0.14328190684318542, 0.2954126298427582, 0.18000535666942596, -0.04555296152830124, -0.416683167219162, -0.1639951765537262, -0.019758082926273346, 0.23976010084152222, -0.12995535135269165, 0.08060075342655182, -0.19504307210445404, 0.0144461989402771, -0.1657218188047409, 0.142140731215477, -0.16518989205360413, -0.35498932003974915, 0.03663388639688492, 0.06656569242477417, -0.08965345472097397, 0.032776378095149994, 0.4078938066959381, 0.13496799767017365, 0.233576238155365, -0.24015268683433533, -0.21379773318767548, 0.12585057318210602, 0.05555253103375435, -0.07777596265077591, -0.24822509288787842, -0.1938064843416214, -0.18507099151611328, 0.26290974020957947, 0.11359784752130508, -0.875744104385376, -0.21726343035697937, 0.29788175225257874, -0.09050770848989487, 0.06773224472999573, 0.26117274165153503, 0.46081215143203735, -0.011698182672262192, -0.1992035210132599, 0.08760407567024231, 0.0045105330646038055, -0.05850193649530411, -0.13027124106884003, -0.15369945764541626 ]
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 ?
[ -0.15169507265090942, 0.0031184398103505373, -0.017220066860318184, 0.3954118490219116, -0.06618139147758484, 0.05573353171348572, 0.043347932398319244, 0.3345067501068115, 0.3488543629646301, 0.11682785302400589, 0.3199305534362793, 0.1732225865125656, -0.1478523164987564, -0.08189837634563446, -0.01621335744857788, -0.4596622586250305, -0.05198794603347778, 0.24634332954883575, 0.03658349812030792, -0.05598783120512962, -0.2965039312839508, 0.2679034173488617, -0.2995699346065521, 0.10403323173522949, -0.06252704560756683, -0.14766892790794373, 0.15097148716449738, 0.02395017258822918, -0.2917690873146057, -0.37569132447242737, 0.041131407022476196, -0.22728337347507477, 0.2764594554901123, 0.1493854522705078, -0.00011269281822023913, -0.03307889401912689, 0.6667639017105103, 0.01335715688765049, -0.4484105110168457, -0.3978950083255768, -0.1365268975496292, -0.4073241651058197, 0.45148593187332153, -0.27835801243782043, 0.01081099919974804, -0.18324364721775055, 0.3193395733833313, 0.17982681095600128, 0.4127219021320343, 0.29741448163986206, 0.21510019898414612, 0.01329302228987217, 0.31618157029151917, 0.02897241897881031, 0.43313145637512207, 0.02345583587884903, 0.03768182173371315, -0.082879438996315, -0.1319962590932846, 0.034157413989305496, -0.12720957398414612, 0.151204913854599, -0.2554621398448944, 0.0856369137763977, 0.34953808784484863, -0.19634757936000824, -0.07118012011051178, -0.5439116954803467, 0.13230380415916443, 0.14068451523780823, 0.6836259365081787, -0.04624008387327194, 0.1660723090171814, 0.0973648726940155, 0.11649169027805328, 0.05789173021912575, 0.2387092411518097, 0.32420793175697327, -0.42637354135513306, -0.30791038274765015, -0.21799728274345398, -0.2496751993894577, -0.2927876114845276, 0.46477800607681274, 0.04294971376657486, 0.5524111390113831, -0.06267837435007095, 0.22764937579631805, -0.04183021932840347, -0.023902783170342445, -0.06886431574821472, -0.12929517030715942, 0.02879023365676403, 0.30135443806648254, 0.060495153069496155, 0.09357383102178574, 0.021500587463378906, 0.2295122593641281, 0.1968686431646347, -0.04341708868741989, -0.1734367311000824, -0.14541444182395935, 0.1463516503572464, 0.14654342830181122, 0.033377792686223984, 0.25672265887260437, 0.1367919147014618, -0.1662372201681137, 0.1898493468761444, 0.21545077860355377, -0.06594295799732208, 0.08795038610696793, -0.031795334070920944, -0.17879848182201385, -0.46619147062301636, 0.023531129583716393, 0.2609487473964691, -0.17632591724395752, -0.03625958412885666, 0.04170362278819084, -0.4259183704853058, -0.2416989654302597, -0.02936340868473053, 0.32280638813972473, -0.21127142012119293, 0.35249972343444824, 0.3740540146827698, 0.07673805207014084, -0.31700387597084045, -0.32719358801841736, 0.020153574645519257, 0.29330238699913025, -0.2867259979248047, 0.08495920896530151, 0.23058679699897766, 0.3047323226928711, 0.3511448800563812, -0.017778655514121056, -0.07731492817401886, 0.008208611980080605, 0.16384054720401764, -0.08172613382339478, -0.21706688404083252, 0.23867331445217133, 0.3142976760864258, 0.2701239585876465, 0.058487072587013245, -0.3325561583042145, -0.07982207834720612, 0.25485411286354065, -0.33965885639190674, -0.07506005465984344, -0.03394251689314842, 0.16418901085853577, -0.10578485578298569, 0.013759737834334373, -0.34825360774993896, 0.2475653737783432, 0.030831359326839447, -0.28464287519454956, -0.17346861958503723, -0.16041035950183868, -0.08583083003759384, -0.35577014088630676, 0.2626809775829315, 0.3350421190261841, 0.13158616423606873, -0.10666561126708984, -0.08848150074481964, 0.16788440942764282, 0.22503484785556793, 0.05638487637042999, -0.12499357014894485, 0.47265011072158813, -0.04442134499549866, -0.07877206057310104, 0.7422000169754028, -0.13415735960006714, -0.1724846363067627, 0.23400957882404327, 0.1318555325269699, 0.01906771957874298, 0.061424050480127335, -0.06255591660737991, 0.12045735865831375, 0.0860343873500824, -0.0515306256711483, 0.5668150186538696, -0.011338246054947376, 0.03431939333677292, -0.4098309576511383, -0.2514776885509491, 0.17665734887123108, 0.06843874603509903, 0.3860548734664917, 0.15072940289974213, 0.010939395055174828, 0.5279036164283752, 0.08195854723453522, -0.03969384357333183, 0.026423344388604164, 0.31999701261520386, -0.29706090688705444, -0.09037389606237411, -0.021785099059343338, 0.06996974349021912, -0.1751365214586258, -0.02547682262957096, -0.3948403596878052, 0.15085041522979736, 0.15849992632865906, 0.10617610812187195, -0.3540579378604889, -0.14297084510326385, -0.0524572916328907, -0.15488967299461365, 0.16197271645069122, 0.18550829589366913, 0.09486673027276993, 0.1506929099559784, 0.05413755029439926, 0.06682395935058594, -0.12752525508403778, -0.04091862589120865, -0.6120948791503906, 0.18550831079483032, -0.27361777424812317, -0.07680456340312958, 0.11377500742673874, 0.21021638810634613, 0.24915525317192078, 0.011264911852777004, -0.12344341725111008, 0.006876274012029171, -0.2011503428220749, -0.06263663619756699, -0.02723267860710621, -0.07050856947898865, 0.2022545039653778, -0.3002309501171112, 0.31451156735420227, 0.29330915212631226, 0.15260079503059387, -0.1323699802160263, 0.08773379772901535, -0.0768917053937912, 0.1184145137667656, 0.3404712975025177, 0.07403974235057831, 0.10327678918838501, 0.004244744312018156, -0.06270992010831833, -0.0017370390705764294, 0.2277480810880661, 0.3068397641181946, 0.4291990101337433, 0.0985184833407402, -0.1077880859375, 0.0011935855727642775, -0.20937363803386688, 0.36993661522865295, 0.044733043760061264, 0.04405386000871658, 0.2671770751476288, -0.4356001019477844, -0.015432283282279968, 0.11218671500682831, -0.16287559270858765, 0.16692359745502472, 0.22491173446178436, -0.15550872683525085, -0.0398738868534565, -0.08139237016439438, -0.12739381194114685, 0.20037710666656494, 0.17848215997219086, 0.3696184754371643, 0.11934082955121994, 0.08972058445215225, 0.07794467359781265, -0.0783376470208168, -0.13671420514583588, -0.1466950625181198, 0.31704777479171753, -0.3040970265865326, 0.07824001461267471, -0.023409830406308174, -0.4767172038555145, -0.24303022027015686, 0.044123731553554535, -0.45233333110809326, -0.38936376571655273, -0.04385971277952194, 0.26631513237953186, 0.10461790859699249, 0.0952170193195343, 0.3447333574295044, 0.14382542669773102, -0.01657198928296566, -0.09474831074476242, -0.1618870049715042, -0.238800048828125, -0.46269991993904114, 0.060326773673295975, 0.36436522006988525, 0.2716529071331024, 0.26252681016921997, -0.03581196814775467, 0.0011126883327960968, -0.01671990007162094, -0.2616986036300659, 0.03566299378871918, -0.08692256361246109, 0.19745075702667236, -0.1058528944849968, 0.4281996190547943, -0.06742069870233536, -0.19357015192508698, 0.26940402388572693, -0.03542931377887726, -0.09465449303388596, 0.0787811204791069, -0.13136345148086548, 0.019894620403647423, -0.018719948828220367, -0.3938399851322174, -0.19605620205402374, -0.572257399559021, 0.03628217428922653, 0.501702606678009, 0.15746331214904785, 0.15771831572055817, 0.22311857342720032, 0.07921316474676132, 0.3722139894962311, 0.3003672659397125, -0.07484935969114304, 0.08280735462903976, 0.35699647665023804, -0.24258501827716827, -0.45372623205184937, 0.006873347796499729, -0.2535282373428345, 0.20978476107120514, 0.11023588478565216, -0.5407999157905579, -0.06307707726955414, -0.03617767617106438, -0.09408997744321823, -0.07280648499727249, 0.09220920503139496, 0.402616947889328, -0.025193287059664726, 0.06426438689231873, 0.0335509330034256, -0.004460041411221027, -0.1506369262933731, -0.2979055345058441, 0.5397560596466064, 0.34649744629859924, 0.38570576906204224, -0.12387118488550186, 0.9763141870498657, 0.11937181651592255, -0.01083979383111, 0.2124878466129303, 0.1578993797302246, 0.04472494125366211, -0.2002597153186798, -0.21275119483470917, 0.14212368428707123, -0.015238681808114052, -0.051308486610651016, 0.3709128201007843, -0.09053636342287064, -0.2564307153224945, -0.289777547121048, -0.05816572532057762, -0.2621937394142151, -0.12282591313123703, 0.23912188410758972, 0.04297173023223877, 0.19683194160461426, -0.11754605919122696, 0.0472726970911026, -0.06097222864627838, -0.3704538345336914, 0.08413922786712646, 0.263555645942688, -0.05967564508318901, 0.08931136131286621, 0.2353031188249588, -0.17476285994052887, -0.5001494884490967, 0.40322357416152954, 0.02955080009996891, 0.09460233151912689, -0.23251745104789734, -0.025917787104845047, 0.17269614338874817, -0.04392090439796448, 0.22943970561027527, -0.17277543246746063, -0.18007849156856537, 0.3131050765514374, 0.27893170714378357, -0.493118554353714, -0.00716136721894145, -0.22605521976947784, 0.21649277210235596, 0.3188702166080475, -0.06501583755016327, -0.44683411717414856, 0.0118402736261487, 0.3586074113845825, 0.26255860924720764, -0.1611928939819336, -0.06786530464887619, -0.4241854250431061, -0.10542414337396622, -0.3580254912376404, -0.1601400077342987, -0.14326463639736176, 0.3781591057777405, 0.014348993077874184, 0.08232955634593964, -0.051163215190172195, -0.07533719390630722, 0.03009631484746933, 0.05655314773321152, 0.13161449134349823, 0.15762051939964294, -0.3129262328147888, -0.17733997106552124, 0.2570784091949463, -0.36674103140830994, 0.20740576088428497, 0.33988094329833984, 0.022439608350396156, 0.009034275077283382, 0.16051530838012695, 0.10108709335327148, 0.12116938829421997, 0.03866142779588699, -0.09372289478778839, -0.09350207448005676, -0.02033759281039238, -0.10446907579898834, 0.04888497292995453, 0.20298324525356293, -0.18986716866493225, -0.19062075018882751, -0.18611188232898712, 0.9091107249259949, 0.11542591452598572, -0.08426395058631897, 0.39913347363471985, -0.10438933223485947, -0.3004436194896698, 0.4643958806991577, 0.29235056042671204, 1.0337493419647217, -0.05213090404868126, 0.07410755008459091, 0.35504287481307983, 0.14539161324501038, 0.5975177884101868, -0.5634700059890747, 0.2468145191669464, -0.46771374344825745, 0.19664140045642853, -0.042075157165527344, -0.08887286484241486, 0.1818549782037735, -0.025462539866566658, -0.422960102558136, 0.30799657106399536, 0.2147320657968521, 0.053165536373853683, -0.004552327562123537, 0.49724313616752625, -0.0075475373305380344, -0.20341500639915466, -0.20266886055469513, 0.11879856139421463, -0.4527354836463928, 0.4196888506412506, -0.3137229084968567, -0.07739035785198212, -0.03726916387677193, -0.26645177602767944, -0.32122600078582764, 0.23819081485271454, -0.28426414728164673, 0.2605063021183014, -0.07888036966323853, -0.40404319763183594, 0.1364523023366928, 0.2001180797815323, -0.07382691651582718, 0.22187989950180054, -0.36752015352249146, 0.3076857328414917, -0.3412057161331177, -0.39709144830703735, -0.16703777015209198, 0.3038885295391083, 0.3126794397830963, -0.20412340760231018, -0.09111818671226501, 0.26975294947624207, -0.15746240317821503, -0.19343268871307373, -0.11627307534217834, -0.04585830122232437, 0.4036431908607483, -0.08228041976690292, -0.2840815484523773, 0.08862943947315216, -0.11740083992481232, 0.10527440160512924, 0.12881624698638916, -0.15789081156253815, -0.02384294755756855, -0.1117861196398735, 0.4168771207332611, -0.12245696783065796, 0.15307922661304474, 0.21601258218288422, 0.351275771856308, -0.04350973293185234, 0.624259889125824, 0.2489510178565979, -0.12671299278736115, -0.19333381950855255, -0.08216250687837601, -0.363400936126709, -0.22032548487186432, -0.13269220292568207, 0.06104382127523422, 0.026035677641630173, -0.1617572009563446, 0.3119337856769562, 0.045710694044828415, 0.14814749360084534, 0.085420161485672, -0.47049373388290405, -0.085733562707901, 0.36615777015686035, -0.21597404778003693, -0.026713842526078224, 0.1270246058702469, -0.1144794374704361, 0.4633326828479767, 0.13523617386817932, -0.2516553997993469, -0.19684094190597534, -0.2700386643409729, 0.20194362103939056, 0.33353689312934875, -0.03152091056108475, -0.09638075530529022, 0.18977481126785278, 0.03872325271368027, -0.18776564300060272, -0.19673050940036774, -0.19243742525577545, 0.07814828306436539, 0.16214874386787415, 0.20964819192886353, -0.05808482691645622, -0.10040751844644547, -0.25374794006347656, 0.04045737907290459, -0.17626835405826569, -0.17049220204353333, -0.2846742272377014, -0.16244208812713623, 0.07903779298067093, -0.06264528632164001, 0.2831449508666992, -0.3918219804763794, 0.21595905721187592, -0.15372972190380096, 0.4537636637687683, 0.11764970421791077, 0.042025499045848846, 0.5127453804016113, -0.05549104884266853, -0.7429887652397156, 0.03110474720597267, 0.013880796730518341, 0.11196247488260269, 0.25134676694869995, -0.1018093153834343, 0.0759131908416748, -0.05008993297815323, -0.1866135150194168, 0.017426058650016785, -0.12645620107650757, 0.07031653076410294, 0.11335652321577072, 0.15643493831157684, -0.16165702044963837, 0.3631627857685089, 0.39460477232933044, 0.17278079688549042, -0.1709178239107132, 0.07866593450307846, 0.034252580255270004, -0.09170495718717575, -0.003751617856323719, 0.2260856181383133, -0.0059778946451842785, -0.05571671575307846, 0.02320871129631996, -0.12723298370838165, -0.02531929314136505, -0.013456890359520912, 0.01118899043649435, -0.020251525565981865, 0.12814365327358246, 0.11533644050359726, -0.14658194780349731, 0.1840616762638092, 0.05437842011451721, 0.006913356948643923, 0.015742527320981026, -0.18640585243701935, 0.23530073463916779, 0.1520325392484665, 0.371134877204895, 0.047062866389751434, 0.049872174859046936, 0.04481039568781853, 0.11762675642967224, -0.14441263675689697, -0.40655258297920227, 0.1806311309337616, -0.12935298681259155, 0.1262434870004654, -0.4547826945781708, -0.07528579980134964, -0.289143830537796, 0.1415073722600937, 0.10705284774303436, -0.2766270637512207, 0.05717259645462036, 0.2894275486469269, -0.3486199676990509, -0.40836596488952637, 0.4397537410259247, 0.14451643824577332, 0.02715003862977028, -0.15341514348983765, 0.3643992245197296, -0.15128126740455627, 0.018108749762177467, -0.08264738321304321, 0.38363853096961975, 0.34362539649009705, 0.5497421026229858, -0.47595930099487305, -0.05491866171360016, 0.055613212287425995, -0.15051952004432678, -0.07915544509887695, 0.11907310038805008, 0.4115953743457794, 0.23406921327114105, 0.18227358162403107, 0.13737614452838898, -0.23148676753044128, -0.031673628836870193, 0.18226847052574158, -0.10571825504302979, -0.08261988312005997, 0.16354724764823914, 0.0025952039286494255, -0.17681822180747986, -0.249995619058609, -0.029853418469429016, -0.6789039373397827, 0.08470205217599869, -0.017861787229776382, 0.1937493085861206, -0.09923867136240005, -0.31724509596824646, 0.028229273855686188, -0.19291365146636963, 0.4531593322753906, 0.33714917302131653, 0.3244171738624573, -0.3690619170665741, -0.31667011976242065, -0.3890501856803894, 0.1598980724811554, -0.3162230849266052, -0.32014745473861694, -0.09974761307239532, 0.23386362195014954, -0.15514899790287018, 0.3409181833267212, -0.23970487713813782, -0.0007637504604645073, 0.07648152858018875, 0.029246345162391663, -0.2659043073654175, -0.16083358228206635, 0.04977165907621384, -0.11939354240894318, -0.3061857223510742, -0.1297396570444107, -0.004993685055524111, -0.27575036883354187, 0.06248512491583824, -0.026224583387374878, -0.2599043846130371, 0.27881893515586853, 0.053317077457904816, 0.2739654779434204, 0.07670801132917404, 0.6266771554946899, -0.1028347760438919, -0.0684027448296547, -0.004855201579630375, -0.06971421092748642, -0.063550665974617, 0.36973804235458374, 0.03681906685233116, 0.5085331201553345, -0.08119194954633713, 0.17832864820957184, -0.35964706540107727, 0.49303293228149414, -0.2634870111942291, 0.14966344833374023, -0.11382630467414856, -0.13014189898967743, -0.23576205968856812, 0.3155692219734192, -0.03986368700861931, 0.08727340400218964, -0.11603302508592606, 0.07417356967926025, -0.4722553789615631, -0.21190245449543, 0.3164288401603699, -0.7915247678756714, -0.32634255290031433, 0.11869125813245773, 0.09361738711595535, -0.08433335274457932, 0.023440012708306313, -0.5125057101249695, -0.06366837024688721, 0.2648855149745941, -0.07161737978458405, -0.10451541095972061, 0.03642409294843674, -0.11945183575153351, -0.05821073055267334, -0.11786194890737534, -0.22919000685214996, -0.12916304171085358, -0.22626133263111115, -0.1860431432723999, -0.3009594678878784 ]
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
[ 0.31512677669525146, -0.13530045747756958, -0.005951759871095419, 0.20902252197265625, 0.09835085272789001, -0.18703697621822357, 0.26089367270469666, 0.3269939422607422, -0.112034372985363, -0.12795452773571014, -0.10739795863628387, 0.4277700185775757, 0.08190377056598663, -0.2038930356502533, 0.0939093679189682, 0.0684095174074173, 0.16564159095287323, 0.11475612223148346, -0.11452048271894455, -0.04316352680325508, -0.17623738944530487, 0.10245221108198166, -0.08972562849521637, -0.11245464533567429, -0.46167999505996704, 0.1460483819246292, -0.2929147183895111, -0.15815584361553192, 0.21726329624652863, -0.39209821820259094, 0.3604980707168579, 0.07840597629547119, -0.2556278705596924, 0.2644166052341461, -0.00011643479956546798, -0.039437782019376755, 0.18988563120365143, -0.06047004833817482, -0.09947287291288376, 0.15082596242427826, 0.020187856629490852, -0.1497223824262619, 0.16033543646335602, -0.3118852376937866, -0.01663917861878872, 0.114776112139225, 0.0591144822537899, 0.13279883563518524, 0.5603551268577576, 0.06728857010602951, 0.19152888655662537, 0.2768821716308594, -0.3045635223388672, 0.1994515061378479, 0.158295139670372, -0.09230350703001022, -0.19662748277187347, -0.07651316374540329, 0.018527325242757797, -0.27605941891670227, -0.03462224826216698, 0.3285541236400604, -0.09903673082590103, 0.09711761772632599, -0.01888168789446354, 0.09337887167930603, 0.1654081791639328, -0.19152864813804626, 0.08571930974721909, 0.06397505104541779, -0.12460602819919586, -0.32999736070632935, -0.22829605638980865, -0.3659471869468689, -0.22737933695316315, -0.3725855052471161, 0.13380692899227142, 0.12148936837911606, -0.1510028839111328, -0.11508476734161377, -0.288733571767807, 0.2583758234977722, 0.18000313639640808, 0.06546580046415329, 0.1485079526901245, 0.34907928109169006, 0.170522540807724, 0.003995595034211874, 0.20543447136878967, 0.16464853286743164, -0.022904440760612488, -0.1676713228225708, 0.03933120146393776, 0.38251468539237976, -0.32864588499069214, -0.1350691318511963, 0.2580420672893524, -0.03761504963040352, -0.020884273573756218, -0.035171665251255035, 0.2935587167739868, -0.05806353688240051, 0.030032506212592125, 0.037809304893016815, 0.14594818651676178, 0.4408579468727112, -0.05845865607261658, 0.2622620761394501, 0.03631449490785599, -0.11190135031938553, -0.47820383310317993, 0.07153692841529846, 0.30434778332710266, 0.08766849339008331, 0.3045765161514282, 0.055590830743312836, -0.07608970254659653, -0.09867477416992188, 0.042736273258924484, 0.09191593527793884, -0.3604728579521179, -0.1665523499250412, 0.04151638224720955, 0.12388288229703903, 0.01051317434757948, 0.1854724884033203, -0.2612513601779938, 0.04123720899224281, -0.5030216574668884, 0.1690860092639923, -0.3158501386642456, 0.011568814516067505, -0.4746893346309662, 0.30212482810020447, 0.24822378158569336, -0.20883449912071228, 0.274618536233902, 0.3141806721687317, -0.24361920356750488, -0.27176880836486816, 0.29848533868789673, -0.20965223014354706, 0.3557649552822113, 0.09198484569787979, -0.11236383765935898, 0.20934996008872986, -0.05222262069582939, 0.1795109063386917, -0.20357470214366913, 0.10834092646837234, -0.10235641896724701, -0.18052856624126434, 0.34438642859458923, 0.16676533222198486, -0.26253756880760193, 0.1560012400150299, 0.32520684599876404, 0.2136511504650116, 0.7869170904159546, -0.3852565288543701, 0.16961777210235596, -0.07720302790403366, -0.2634733021259308, -0.3530150055885315, -0.03340952843427658, 0.2723202109336853, 0.12401345372200012, -0.3135710060596466, 0.1695547103881836, 0.23233522474765778, 0.0021243852097541094, 0.5360866785049438, -0.16725265979766846, 0.1547781378030777, -0.32080090045928955, 0.030222048982977867, 0.42413294315338135, -0.2938268184661865, -0.4632187783718109, 0.12359783053398132, -0.0104856938123703, 0.1751442402601242, 0.11865322291851044, 0.19418881833553314, 0.11643936485052109, -0.2828817665576935, 0.20412304997444153, 0.04053054749965668, -0.10490638762712479, 0.09167343378067017, -0.378133088350296, 0.014632136560976505, 0.3601005971431732, -0.24554415047168732, 0.000286048132693395, 0.061888113617897034, -0.1713062971830368, -0.00008954873192124069, 0.26358163356781006, 0.0021530180238187313, 0.13179706037044525, 0.02144596353173256, -0.05507618933916092, -0.1804734319448471, 0.07052119076251984, -0.18271571397781372, 0.010329277254641056, 0.20688186585903168, -0.2916732132434845, -0.030908700078725815, 0.24033476412296295, -0.11940094083547592, 0.10904709249734879, -0.2765055000782013, -0.26970383524894714, -0.49192118644714355, 0.17864225804805756, 0.05129965767264366, 0.48496031761169434, -0.00875514093786478, -0.08701670914888382, 0.29978764057159424, 0.3912486135959625, -0.1188090518116951, -0.32058095932006836, 0.11166096478700638, -0.0021976495627313852, -0.16492022573947906, -0.35923904180526733, 0.5168808698654175, 0.21017178893089294, -0.00870154146105051, -0.12740281224250793, -0.04566940292716026, 0.0050727506168186665, 0.08217795193195343, -0.2594386339187622, 0.08567077666521072, 0.10755643993616104, 0.1489415466785431, 0.03267945349216461, 0.11418824642896652, -0.03314933553338051, -0.2737234830856323, 0.1267450600862503, 0.4829828441143036, 0.33497604727745056, 0.10381138324737549, -0.00017383715021423995, -0.08665533363819122, -0.06366574019193649, -0.15970759093761444, -0.22713592648506165, -0.3474416732788086, 0.25259479880332947, -0.03174925968050957, 0.42569881677627563, 0.27518200874328613, -0.04971017688512802, 0.27717670798301697, 0.5075844526290894, 0.18606312572956085, -0.022791502997279167, -0.0917823314666748, -0.015192117542028427, -0.26299020648002625, -0.09401440620422363, 0.21855488419532776, 0.5466564297676086, 0.0993313118815422, 0.1461595743894577, 0.18917633593082428, -0.32118359208106995, -0.20500102639198303, -0.029941074550151825, -0.12286648899316788, 0.06329668313264847, -0.042650338262319565, 0.37207162380218506, -0.08435285091400146, -0.20664794743061066, 0.34665077924728394, 0.187712162733078, -0.12854506075382233, 0.03776826709508896, 0.21131959557533264, -0.2382613867521286, -0.4322883188724518, -0.3985506296157837, -0.20139996707439423, -0.12059184908866882, -0.30527961254119873, 0.12759974598884583, 0.30709290504455566, 0.09445147216320038, 0.2894100546836853, -0.20783600211143494, 0.08653520792722702, -0.17033307254314423, -0.15077237784862518, 0.002173089189454913, -0.3452705442905426, -0.3134087026119232, 0.032280076295137405, -0.016492556780576706, -0.27436041831970215, 0.2854226529598236, 0.01840953342616558, -0.1734154224395752, -0.14144903421401978, -0.39166516065597534, 0.20791563391685486, -0.008682796731591225, -0.2005302459001541, -0.19862216711044312, -0.3684796392917633, 0.09001240134239197, 0.01222099456936121, 0.33747461438179016, -0.36552637815475464, -0.24376428127288818, 0.05502287670969963, -0.08614614605903625, 0.10925310105085373, -0.2796725928783417, -0.10740385949611664, -0.14487726986408234, -0.04877816513180733, 0.20083212852478027, -0.2825668156147003, 0.2501552700996399, 0.34708723425865173, -0.2549814283847809, -0.09772169589996338, -0.15965329110622406, -0.16263166069984436, -0.462517112493515, -0.036857981234788895, 0.21641826629638672, -0.3161963224411011, -0.1890566647052765, -0.3839230537414551, -0.3472078740596771, 0.2669111490249634, 0.1701955497264862, -0.435641348361969, -0.4333494305610657, -0.06118948385119438, 0.07350443303585052, 0.17087772488594055, 0.04932286962866783, 0.3877250850200653, -0.17816820740699768, -0.01823333464562893, -0.153355211019516, -0.3007078170776367, 0.2078816294670105, 0.39638301730155945, 0.35678890347480774, -0.1756996214389801, 0.22636638581752777, 0.1310703307390213, 0.5557450652122498, 0.6314631104469299, 0.06918303668498993, 0.37321722507476807, -0.013673871755599976, 0.21526825428009033, -0.26185548305511475, -0.19998669624328613, 0.18395788967609406, -0.15264518558979034, -0.2572273313999176, 0.18328756093978882, 0.17619867622852325, -0.5797117948532104, -0.06663412600755692, 0.1652226448059082, -0.27863404154777527, -0.3541928827762604, 0.09116513282060623, -0.3137052059173584, 0.22862015664577484, 0.3108302056789398, 0.23395322263240814, -0.4184879958629608, -0.1160447970032692, -0.12488031387329102, -0.3066103756427765, 0.1455271989107132, 0.09425301849842072, -0.0962531790137291, -0.24743661284446716, -0.4561164677143097, 0.2061755508184433, 0.24228781461715698, 0.12783098220825195, 0.022665124386548996, -0.23083920776844025, 0.06999609619379044, 0.019404364749789238, 0.4736870527267456, -0.03395901992917061, -0.003952931147068739, 0.10121847689151764, -0.07110372930765152, -0.40342965722084045, -0.2792934477329254, 0.23190376162528992, 0.4614938199520111, 0.4238910675048828, 0.2808866500854492, -0.21179410815238953, -0.10709010809659958, -0.07216489315032959, 0.13284042477607727, -0.23876836895942688, -0.3905238211154938, -0.23795630037784576, -0.06291770190000534, 0.06605035811662674, 0.0712948814034462, -0.12635420262813568, -0.18954627215862274, -0.1981026530265808, -0.07180885970592499, 0.19534066319465637, -0.03914881870150566, 0.03154463320970535, 0.24644120037555695, 0.26728639006614685, -0.12621812522411346, 0.1792062520980835, 0.17825356125831604, -0.10764848440885544, -0.4886569380760193, 0.41116783022880554, -0.1816367208957672, 0.3603588938713074, 0.18541859090328217, -0.05674796923995018, -0.02005300670862198, 0.406605064868927, 0.19946905970573425, 0.1454412043094635, -0.07756073772907257, -0.07732515782117844, -0.026072150096297264, -0.2263958901166916, 0.32063496112823486, 0.0029674137476831675, -0.23820824921131134, -0.45535334944725037, 0.23632289469242096, 0.15792137384414673, -0.14968982338905334, 0.22838015854358673, -0.6113733649253845, -0.11635476350784302, 0.39091718196868896, -0.197150319814682, 0.7841209769248962, 0.024202222004532814, 0.4790370762348175, 0.2232556939125061, -0.05244653299450874, 0.380192369222641, -0.0466325581073761, 0.363625168800354, -0.39555591344833374, 0.18653647601604462, 0.12171313166618347, -0.17433656752109528, -0.3330659866333008, 0.021302781999111176, 0.06636732071638107, 0.4315520226955414, 0.2073076218366623, 0.48412689566612244, 0.02028004638850689, -0.21137624979019165, 0.20552673935890198, 0.02013992704451084, -0.3214283287525177, 0.17260447144508362, 0.05149531736969948, 0.24724842607975006, -0.0631638914346695, -0.22587206959724426, -0.0719214454293251, -0.1654234230518341, -0.03188123181462288, 0.2956011891365051, -0.13948607444763184, 0.4080139696598053, 0.35625553131103516, -0.20116782188415527, -0.6638271808624268, 0.09609922766685486, 0.13754169642925262, -0.3959575295448303, -0.01546770054847002, -0.111834816634655, 0.4741964340209961, 0.4154972434043884, -0.05152205005288124, -0.2530067265033722, 0.366411417722702, -0.0708453431725502, -0.048300851136446, -0.055002763867378235, -0.008599500171840191, -0.4087693989276886, -0.23523584008216858, 0.3175221085548401, -0.3334672152996063, 0.05075538903474808, 0.13472504913806915, 0.183037668466568, -0.07473669946193695, -0.11483439803123474, 0.11003778129816055, 0.0739767774939537, -0.1293463408946991, 0.35186001658439636, -0.35450479388237, -0.10786087810993195, 0.034888774156570435, 0.28572115302085876, 0.2867715656757355, 0.08036516606807709, 0.36261317133903503, -0.27099648118019104, 0.061727963387966156, -0.036722831428050995, -0.33266276121139526, -0.31491512060165405, -0.3201562166213989, 0.10790354758501053, -0.2535043954849243, -0.18748368322849274, 0.10250464081764221, 0.007121487054973841, 0.12069115042686462, 0.37156036496162415, -0.0805354192852974, -0.18930520117282867, -0.5517705678939819, -0.0539892353117466, -0.1033543050289154, 0.16097675263881683, -0.017603116109967232, 0.6225499510765076, 0.07026726752519608, 0.39354243874549866, -0.28945857286453247, -0.020572088658809662, -0.015657369047403336, 0.22764627635478973, -0.13291743397712708, -0.40480244159698486, -0.054158639162778854, 0.14537961781024933, 0.043657638132572174, 0.33607953786849976, -0.4868718981742859, -0.20181836187839508, -0.16914977133274078, 0.10196724534034729, 0.14670825004577637, 0.14324107766151428, -0.2514537274837494, 0.13051296770572662, 0.3261985778808594, -0.34623968601226807, 0.11617041379213333, 0.1698741614818573, 0.022485001012682915, 0.19154712557792664, 0.16412410140037537, 0.0671846941113472, 0.1441856026649475, 0.11707303673028946, -0.14140525460243225, -0.059945300221443176, 0.16280721127986908, 0.18861618638038635, -0.13985148072242737, 0.15012970566749573, -0.13678883016109467, -0.1625829041004181, -0.11337693780660629, -0.07932811975479126, -0.10075531154870987, -0.27143827080726624, -0.129256010055542, 0.07303952425718307, 0.21172848343849182, 0.22056743502616882, -0.08022292703390121, -0.17486289143562317, -0.10348451137542725, 0.16781766712665558, 0.0692361369729042, 0.10698186606168747, 0.3573373556137085, 0.20533914864063263, 0.07064763456583023, 0.35987505316734314, 0.4193614721298218, -0.00026027296553365886, 0.4682634770870209, 0.034651048481464386, -0.08014576137065887, -0.47159332036972046, 0.26750683784484863, 0.010570496320724487, 0.19505774974822998, 0.08407630026340485, 0.34856459498405457, -0.2840741276741028, -0.07184199243783951, 0.0636942982673645, 0.2873608469963074, 0.16066117584705353, -0.029453137889504433, -0.07150301337242126, 0.08202336728572845, 0.019564643502235413, -0.2752048075199127, 0.029206838458776474, -0.0937788337469101, 0.25051388144493103, 0.043637555092573166, 0.3715870678424835, -0.15463954210281372, -0.18234413862228394, -0.42075115442276, 0.3165491819381714, -0.28920915722846985, -0.28602907061576843, 0.2699078321456909, -0.30062466859817505, 0.2749161124229431, 0.1448916494846344, -0.04462623968720436, 0.19261574745178223, 0.7117821574211121, 0.06008920073509216, -0.027329545468091965, -0.15319839119911194, -0.3156096041202545, 0.29996347427368164, 0.39620769023895264, -0.10624177753925323, 0.3474767208099365, 0.0668395608663559, -0.3460542559623718, -0.15965624153614044, 0.18510864675045013, 0.39520499110221863, 0.27467307448387146, -0.2849203646183014, 0.16531150043010712, 0.31191617250442505, 0.043719884008169174, -0.14279258251190186, 0.28440621495246887, 0.46444469690322876, -0.2753710448741913, 0.10013259202241898, 0.07736315578222275, -0.16198551654815674, 0.3374539315700531, 0.09224109351634979, 0.26263776421546936, -0.2047155201435089, 0.5761633515357971, -0.05506069213151932, -0.07937562465667725, 0.008968982845544815, 0.20302973687648773, -0.5066409111022949, -0.12212088704109192, 0.06195890158414841, 0.05762312933802605, 0.1631636768579483, -0.22196762263774872, 0.09016381204128265, 0.10871439427137375, 0.3901247978210449, 0.12524515390396118, 0.13043217360973358, -0.2856215238571167, -0.016744056716561317, -0.5698695182800293, 0.33293047547340393, -0.3890037536621094, 0.24580851197242737, -0.1836739331483841, 0.27481549978256226, -0.26683568954467773, 0.039425186812877655, 0.14697863161563873, -0.21078899502754211, 0.18485164642333984, 0.10188571363687515, -0.4532124698162079, 0.025243164971470833, 0.0722878947854042, -0.19550499320030212, 0.25490912795066833, -0.23168110847473145, 0.02173997275531292, -0.019082728773355484, 0.13105084002017975, 0.0037310351617634296, -0.003355564083904028, 0.2672177851200104, 0.506138026714325, 0.2889290153980255, 0.42512595653533936, 0.20019449293613434, 0.04050192981958389, -0.18851466476917267, -0.39051181077957153, 0.03260098770260811, -0.031581778079271317, -0.24561244249343872, -0.0693579912185669, 0.3341865837574005, -0.31259840726852417, -0.11305195093154907, -0.28429922461509705, 0.22645717859268188, 0.08093835413455963, -0.39334404468536377, 0.17809896171092987, -0.036445870995521545, -0.12942686676979065, 0.060525987297296524, 0.03079359419643879, 0.725405216217041, 0.016145329922437668, 0.4419615864753723, -0.18536333739757538, -0.2571270763874054, 0.3819313049316406, -0.6724093556404114, -0.2614431381225586, 0.0037962165661156178, 0.3030821979045868, 0.09342256933450699, -0.17738237977027893, -0.530910849571228, -0.1243797242641449, 0.17773979902267456, -0.16756929457187653, -0.12121245265007019, 0.1805967092514038, -0.1937052309513092, -0.035032667219638824, -0.010602809488773346, 0.16422681510448456, 0.09443014115095139, -0.2684940695762634, 0.27279284596443176, -0.2511071264743805 ]
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.
[ 0.2502775192260742, -0.3640768527984619, -0.061697788536548615, 0.1393110603094101, -0.0684489980340004, -0.07843432575464249, 0.3446008861064911, 0.2630481719970703, 0.37012922763824463, 0.19107212126255035, -0.2715184688568115, 0.6841734051704407, -0.0655926838517189, 0.1188960000872612, 0.18207651376724243, -0.11214069277048111, -0.10166774690151215, 0.34712934494018555, -0.15045098960399628, -0.12176010757684708, -0.26120540499687195, 0.37858521938323975, -0.2576519548892975, 0.14147375524044037, -0.17962299287319183, -0.10173958539962769, -0.1501425802707672, 0.07673334330320358, -0.03307381272315979, -0.401468962430954, 0.10599258542060852, 0.14746442437171936, -0.1076185330748558, -0.12033388763666153, -0.0001057189583661966, -0.28695619106292725, 0.2894555926322937, 0.07167445868253708, 0.032457150518894196, 0.12928740680217743, -0.14674843847751617, -0.2972307801246643, -0.03970760107040405, -0.3864685297012329, -0.05714058503508568, -0.0009066477068699896, 0.04592796415090561, -0.12836851179599762, 0.16652292013168335, 0.16479052603244781, 0.24084724485874176, 0.1582944542169571, 0.08807902783155441, 0.13617277145385742, 0.45657655596733093, -0.4768168032169342, 0.044315386563539505, 0.24602989852428436, 0.09880666434764862, -0.34286805987358093, -0.016030285507440567, 0.45544934272766113, 0.12964677810668945, 0.1253584921360016, 0.039680927991867065, 0.1442376971244812, 0.03318913280963898, -0.35992300510406494, -0.0838700607419014, 0.42646491527557373, 0.1315079629421234, -0.3524518311023712, -0.31000685691833496, -0.2016560584306717, 0.21391206979751587, -0.3125111758708954, -0.06473948061466217, 0.14203603565692902, -0.01629730314016342, 0.00970698706805706, 0.03766300901770592, 0.10280205309391022, -0.12196812033653259, 0.1391846239566803, 0.13890483975410461, 0.5842684507369995, -0.02031157910823822, 0.06995432078838348, 0.36926349997520447, -0.029156845062971115, -0.1850247085094452, 0.09218516200780869, -0.020080531015992165, 0.17599335312843323, -0.28078654408454895, -0.1972288191318512, 0.10954350978136063, 0.1007743552327156, 0.021075034514069557, 0.044597078114748, 0.22502386569976807, 0.03807021304965019, 0.20733891427516937, 0.16420333087444305, 0.08044931292533875, 0.28569498658180237, 0.384816437959671, 0.28377097845077515, 0.028996869921684265, 0.06244770810008049, -0.12825199961662292, 0.20890246331691742, -0.014154298231005669, -0.10459049046039581, 0.27334558963775635, -0.09539487957954407, -0.07485385984182358, -0.12386628240346909, -0.3194418251514435, -0.0257934108376503, -0.28512465953826904, 0.002068599918857217, -0.1796579211950302, 0.244452565908432, 0.027862949296832085, 0.015347042120993137, -0.03464847430586815, 0.09597474336624146, -0.15768180787563324, -0.16153265535831451, -0.2969713509082794, 0.37302109599113464, -0.23291641473770142, -0.25091108679771423, 0.14201577007770538, 0.2079273909330368, 0.28334152698516846, 0.028486929833889008, -0.13069751858711243, 0.037279363721609116, 0.14972728490829468, -0.020442312583327293, 0.27757930755615234, 0.11669623851776123, -0.06863974034786224, 0.10494772344827652, -0.01046609878540039, -0.37980329990386963, -0.2766099274158478, -0.0801536962389946, 0.09214677661657333, 0.07429827749729156, 0.02317366935312748, 0.2386457324028015, -0.17580583691596985, 0.027032289654016495, 0.1805448979139328, 0.09859310835599899, -0.09635739773511887, 0.2063305824995041, 0.14499232172966003, -0.14029505848884583, -0.11530141532421112, -0.18978017568588257, 0.2916015386581421, 0.2018369436264038, -0.30105817317962646, -0.11395047605037689, 0.08180490881204605, -0.03524652495980263, -0.12665049731731415, 0.05246226117014885, 0.2215992957353592, 0.00540855573490262, -0.2682156562805176, 0.2221400886774063, 0.2920091152191162, -0.5699790120124817, -0.4601469039916992, 0.026117447763681412, -0.09821590781211853, -0.11541426926851273, 0.08403830230236053, 0.07993180304765701, -0.02733764611184597, 0.20734506845474243, 0.30574339628219604, 0.2759207785129547, 0.1464020162820816, -0.014055224135518074, -0.33543699979782104, 0.14553847908973694, 0.3818054497241974, 0.1446690559387207, -0.17248345911502838, -0.16292200982570648, -0.20003026723861694, -0.16491572558879852, 0.26382994651794434, 0.11475232988595963, -0.08811510354280472, 0.2681298553943634, 0.01201168354600668, -0.3535585105419159, 0.12525126338005066, -0.14601200819015503, -0.2344864457845688, 0.24857276678085327, -0.2334451675415039, 0.3903796076774597, 0.2793835699558258, -0.13543885946273804, -0.3910471200942993, -0.009539970196783543, -0.021591175347566605, -0.2226114273071289, 0.2764018774032593, 0.022155849263072014, 0.3353920578956604, 0.061246130615472794, -0.09688164293766022, 0.3609965145587921, 0.09802190214395523, 0.04160773381590843, -0.3697095513343811, 0.0968739241361618, -0.059088315814733505, -0.04003389552235603, 0.2472638338804245, 0.5104692578315735, 0.24413594603538513, -0.05340427905321121, 0.13028821349143982, 0.3582196831703186, -0.023660259321331978, 0.01289600320160389, 0.08358979970216751, -0.01738232746720314, -0.04296492785215378, -0.29577168822288513, 0.09626902639865875, 0.167236328125, 0.08046383410692215, -0.040201008319854736, -0.15581382811069489, 0.3673763573169708, 0.09383894503116608, 0.05186353623867035, 0.11298207938671112, 0.0980575904250145, -0.09670717269182205, -0.29650336503982544, -0.31454530358314514, -0.13945326209068298, 0.33484089374542236, -0.22544223070144653, 0.35534870624542236, 0.04052114486694336, -0.2502378821372986, 0.2089705765247345, 0.4467063844203949, 0.024678612127900124, 0.2037922590970993, -0.11887653172016144, -0.2853444814682007, -0.10063879936933517, 0.054819680750370026, 0.5218159556388855, 0.1674070507287979, 0.4208611249923706, 0.017650077119469643, -0.059704333543777466, -0.3939404785633087, -0.07345971465110779, 0.16774001717567444, -0.08496882766485214, 0.07969304919242859, -0.08924093097448349, 0.20133495330810547, -0.1482601761817932, -0.1896371692419052, 0.055628541857004166, -0.24803268909454346, 0.23431766033172607, -0.262809693813324, -0.19544807076454163, -0.08484268188476562, -0.497170090675354, -0.27955806255340576, -0.2954831123352051, -0.14680342376232147, -0.52729332447052, 0.13580846786499023, -0.24137598276138306, -0.020684940740466118, 0.22529463469982147, 0.054361067712306976, 0.1611199826002121, -0.18304020166397095, 0.07843846082687378, -0.2059391885995865, -0.1355559527873993, -0.21072614192962646, 0.15992183983325958, -0.2740512788295746, 0.5203292369842529, 0.39891332387924194, -0.09670918434858322, 0.07538852840662003, 0.31900420784950256, -0.21436801552772522, -0.0005631212261505425, -0.173783078789711, 0.24654273688793182, 0.4136919379234314, 0.0018742462852969766, -0.1277114450931549, -0.01299276016652584, 0.2429472953081131, -0.4388088881969452, -0.005683107301592827, -0.06721719354391098, 0.20725546777248383, -0.03914044424891472, -0.14822222292423248, -0.4048270583152771, -0.36483630537986755, -0.221210315823555, 0.23954303562641144, 0.2358979731798172, 0.25962066650390625, 0.11519578099250793, -0.33117029070854187, 0.32700517773628235, -0.1680487096309662, -0.03360208496451378, -0.1516864150762558, -0.05473510921001434, 0.2490205615758896, -0.1052173525094986, -0.2982955276966095, -0.24963775277137756, -0.1970975250005722, 0.22305850684642792, -0.2926654815673828, -0.21160611510276794, -0.17701639235019684, -0.07415018230676651, -0.1401941329240799, 0.027845030650496483, 0.22281566262245178, 0.11768671125173569, 0.032634519040584564, -0.15178057551383972, -0.1452549248933792, -0.24856722354888916, 0.3365793526172638, 0.05205540731549263, -0.14087170362472534, 0.038537535816431046, 0.03523281216621399, 0.014000759460031986, 0.17996066808700562, 0.11784464120864868, -0.12312301993370056, 0.19788867235183716, -0.2788589596748352, 0.29525014758110046, 0.23055239021778107, -0.2540299892425537, 0.010372533462941647, 0.3167248070240021, 0.2669711709022522, 0.13920293748378754, 0.11716976016759872, -0.3290160298347473, -0.09989453107118607, 0.11832111328840256, -0.26283538341522217, -0.059080570936203, 0.03934846818447113, -0.10333201289176941, -0.21532246470451355, 0.020680053159594536, -0.03607334569096565, -0.1080712303519249, -0.17250432074069977, -0.04275599867105484, -0.3998681902885437, -0.08388228714466095, -0.038334932178258896, -0.6121565699577332, -0.12144092470407486, -0.5194757580757141, 0.154316708445549, 0.048394717276096344, 0.4734872281551361, -0.010718360543251038, -0.09485800564289093, -0.031377073377370834, 0.03042140230536461, 0.5715877413749695, -0.4146971106529236, 0.2131909281015396, 0.4603511095046997, 0.18213942646980286, -0.69499671459198, 0.00564079312607646, -0.5489994287490845, -0.11691147089004517, 0.628441572189331, 0.2622632086277008, -0.24899600446224213, -0.15028272569179535, 0.2737717628479004, 0.016338087618350983, -0.09192773699760437, -0.1665167510509491, -0.2935802936553955, -0.22302329540252686, -0.036796052008867264, 0.14788106083869934, -0.009105554781854153, 0.26366716623306274, 0.19725726544857025, -0.07814469188451767, 0.05830588564276695, -0.21258725225925446, 0.14469681680202484, 0.22399559617042542, 0.2999829649925232, 0.06814359128475189, -0.03851281851530075, 0.11713021248579025, 0.2122262716293335, -0.678682804107666, 0.12109199911355972, -0.5201424360275269, 0.3546108603477478, -0.21757523715496063, -0.25405576825141907, 0.4651707112789154, 0.1619272232055664, 0.0466170459985733, -0.161482572555542, -0.5788978934288025, 0.21016088128089905, -0.047841817140579224, 0.020869985222816467, 0.20848549902439117, 0.1639750897884369, -0.23401431739330292, -0.07566925138235092, 0.5497065782546997, 0.06847739219665527, -0.22469212114810944, 0.33281779289245605, 0.19439487159252167, -0.35684239864349365, 0.41191190481185913, -0.08889129012823105, 0.9092603921890259, -0.09022390097379684, -0.007592327892780304, 0.25458452105522156, -0.36367470026016235, 0.6806228756904602, 0.14083077013492584, 0.12344548851251602, -0.5268475413322449, -0.3948061466217041, 0.06874209642410278, -0.08320681005716324, -0.12789751589298248, 0.2755148112773895, -0.2226203978061676, 0.41558218002319336, 0.07301351428031921, -0.19334717094898224, 0.022332554683089256, 0.3862769603729248, 0.0031905313953757286, -0.10484953969717026, -0.34154418110847473, 0.15480968356132507, -0.032970454543828964, 0.1674894541501999, -0.06146233528852463, -0.019678344950079918, -0.13237006962299347, -0.21145232021808624, -0.048341844230890274, 0.19770124554634094, -0.21826794743537903, -0.1657027304172516, 0.3732074201107025, -0.0305124893784523, -0.4524392783641815, 0.13093191385269165, 0.18288522958755493, 0.12453877925872803, 0.006406739819794893, 0.05262477695941925, 0.17659853398799896, 0.0018503527389839292, 0.5183125138282776, 0.2572228014469147, 0.29605716466903687, -0.10177110135555267, -0.057916902005672455, -0.10081509500741959, 0.0859796404838562, -0.10533280670642853, -0.06399115175008774, 0.058669012039899826, -0.05259701609611511, -0.1444300264120102, -0.34001392126083374, 0.06040729209780693, -0.013500235974788666, -0.197665274143219, 0.19822055101394653, 0.0939963310956955, -0.41108354926109314, 0.27004364132881165, 0.0935869887471199, -0.3252714276313782, -0.036869801580905914, 0.3700680136680603, -0.05986518785357475, 0.4367310702800751, 0.17269837856292725, 0.029365796595811844, -0.016895318403840065, -0.28269121050834656, -0.0653926432132721, 0.32145747542381287, -0.2296520471572876, 0.06664927303791046, -0.12342493236064911, -0.10889672487974167, -0.02172110043466091, 0.03522311896085739, 0.1964280605316162, 0.1793641448020935, 0.09442636370658875, -0.46983930468559265, -0.1270841658115387, 0.16168777644634247, 0.06652913242578506, 0.31714022159576416, 0.0377875491976738, -0.094145767390728, 0.2178998738527298, 0.3500141501426697, -0.4358420968055725, -0.023670386523008347, -0.2066919058561325, -0.07812854647636414, 0.4145659804344177, -0.023621035739779472, 0.07423564046621323, 0.0542699471116066, 0.3157196044921875, 0.07997900247573853, -0.09340909123420715, -0.30972805619239807, -0.18688155710697174, 0.019749455153942108, -0.07820963114500046, -0.011730517260730267, 0.025764578953385353, -0.12001346051692963, -0.076725073158741, -0.06595677882432938, -0.09149370342493057, -0.1465873420238495, 0.007814955897629261, 0.4473211169242859, 0.0004198035167064518, 0.09699166566133499, 0.06402931362390518, 0.11668186634778976, -0.021098578348755836, -0.2281688153743744, 0.032190222293138504, 0.16215144097805023, -0.051699597388505936, -0.098664291203022, -0.1505555510520935, -0.047165364027023315, -0.38828185200691223, -0.06876561045646667, 0.006775987800210714, -0.16050168871879578, -0.09582388401031494, 0.07739394903182983, 0.6340236067771912, 0.010784421116113663, -0.29655298590660095, -0.19216446578502655, 0.027980701997876167, 0.3456474542617798, -0.39493411779403687, -0.12690265476703644, 0.4542475640773773, 0.24674110114574432, 0.26460975408554077, 0.35296547412872314, 0.1391116827726364, -0.0840521827340126, -0.19155500829219818, 0.004000295419245958, 0.291332483291626, -0.40785154700279236, -0.16344194114208221, 0.17371097207069397, -0.1426512449979782, -0.1698189377784729, 0.32067838311195374, 0.05786518380045891, 0.16803520917892456, 0.08389309793710709, 0.3906584680080414, 0.5761870741844177, 0.14169670641422272, 0.030426444485783577, -0.05772709101438522, 0.11515689641237259, -0.0707182064652443, 0.5745169520378113, -0.02764991857111454, 0.24956761300563812, 0.15677978098392487, 0.589764416217804, -0.2613562047481537, -0.21765770018100739, -0.040920983999967575, 0.21259526908397675, -0.034246280789375305, -0.2194952666759491, 0.13715550303459167, -0.07145749777555466, -0.3883407711982727, -0.004371357150375843, 0.014580773189663887, -0.231360524892807, 0.3803722858428955, -0.04845964536070824, 0.1276191771030426, -0.2602132260799408, -0.11909821629524231, 0.024867817759513855, 0.4030859172344208, -0.39799314737319946, 0.06432215869426727, -0.2506292164325714, 0.09554290026426315, -0.01759813353419304, 0.3547400236129761, 0.17168110609054565, 0.017751198261976242, -0.251244455575943, -0.08426295220851898, -0.19496062397956848, -0.11999725550413132, -0.20278149843215942, 0.2279924750328064, 0.24091894924640656, 0.018384071066975594, 0.4225250482559204, 0.18160417675971985, -0.22556334733963013, 0.049137357622385025, 0.16723479330539703, 0.0984485000371933, -0.0070397937670350075, 0.4931472837924957, -0.10185009241104126, -0.2557196617126465, -0.2666809558868408, -0.0513104647397995, -0.40268993377685547, -0.06665530800819397, -0.11750078201293945, -0.03085358999669552, 0.17395590245723724, -0.383569598197937, 0.14161600172519684, -0.049735743552446365, 0.44530877470970154, 0.012515826150774956, 0.2989167273044586, -0.21124088764190674, -0.21040122210979462, -0.6262403130531311, -0.022991539910435677, -0.09374160319566727, -0.5266144871711731, 0.2152058333158493, 0.029598230496048927, -0.06426148861646652, 0.020876573398709297, 0.04277624562382698, 0.5819960832595825, -0.16318963468074799, -0.10652969032526016, -0.4230671226978302, 0.11542091518640518, -0.18298101425170898, -0.05020354688167572, 0.14122647047042847, -0.2908751964569092, 0.011113027110695839, -0.2384883612394333, 0.19092661142349243, 0.07287680357694626, 0.025542475283145905, 0.06769717484712601, 0.27018463611602783, 0.13873103260993958, 0.04367153346538544, 0.26167652010917664, -0.14459022879600525, 0.0857325866818428, -0.10288693755865097, -0.01684618927538395, -0.29823487997055054, 0.039442092180252075, 0.10759301483631134, 0.33385229110717773, 0.041206084191799164, 0.02199237234890461, -0.07287934422492981, 0.0933414176106453, 0.1682974249124527, 0.2568714916706085, 0.03382614627480507, 0.1943795084953308, 0.14239351451396942, -0.04101584106683731, 0.09539638459682465, 0.34074845910072327, 0.0053742676973342896, 0.05500371754169464, -0.012975335121154785, -0.34891965985298157, 0.4604705572128296, -0.3287239670753479, -0.13288937509059906, 0.09408177435398102, 0.34773772954940796, 0.26250341534614563, -0.24081836640834808, -0.5395860075950623, -0.13690701127052307, 0.3368120789527893, 0.042628053575754166, -0.2955171465873718, 0.3126371204853058, -0.04802176356315613, 0.12107111513614655, 0.054204415529966354, -0.015983974561095238, 0.2987123429775238, -0.2579360008239746, -0.10434567928314209, -0.2538382411003113 ]
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.
[ 0.13189564645290375, -0.12678305804729462, 0.05668973550200462, -0.028536442667245865, 0.11699824035167694, -0.1541464924812317, 0.3969920873641968, -0.06210575997829437, -0.21482831239700317, 0.25181275606155396, -0.08670715242624283, 0.15889912843704224, -0.17179937660694122, 0.2942129075527191, 0.3824806809425354, 0.10448379069566727, 0.0907263532280922, 0.31321364641189575, 0.03962354734539986, -0.12798276543617249, -0.26271045207977295, 0.05864562466740608, -0.01620396599173546, -0.07316631823778152, -0.07528543472290039, -0.18676677346229553, -0.13502037525177002, 0.2678753733634949, -0.22584213316440582, -0.3100968897342682, 0.4256783425807953, 0.20961815118789673, 0.34658509492874146, 0.5284522175788879, -0.00012068286014255136, -0.0640510693192482, 0.1734645515680313, -0.24845097959041595, -0.5683867931365967, -0.35898637771606445, 0.0835864320397377, -0.1341312825679779, 0.08103296905755997, -0.2865528464317322, 0.22721746563911438, 0.052917689085006714, 0.050873056054115295, 0.13236530125141144, 0.5366085171699524, 0.40950721502304077, 0.12447773665189743, 0.06691014766693115, -0.21695265173912048, -0.10151121765375137, 0.008197019807994366, 0.16872473061084747, -0.03125619515776634, 0.2934640645980835, 0.31920763850212097, -0.40811315178871155, 0.29175907373428345, -0.242682084441185, -0.039042599499225616, -0.007286003790795803, 0.5370671153068542, 0.05455397069454193, -0.053798187524080276, -0.4257263243198395, 0.004694019909948111, 0.15714412927627563, 0.28184235095977783, 0.033649034798145294, -0.32898202538490295, -0.11294952034950256, 0.11158490180969238, -0.5599850416183472, 0.209152489900589, 0.30534741282463074, -0.42565545439720154, -0.03852704167366028, -0.20674756169319153, 0.3366679847240448, -0.39196646213531494, 0.475935697555542, 0.11568892747163773, 0.3350134491920471, -0.03723018243908882, 0.13018055260181427, 0.2861829400062561, -0.09340392798185349, -0.00962584838271141, 0.07325876504182816, 0.02519134245812893, 0.2944716215133667, -0.16644224524497986, -0.18143440783023834, -0.030567606911063194, -0.2799895405769348, -0.11502803862094879, 0.09093757718801498, 0.46870723366737366, -0.18926194310188293, -0.2184365838766098, 0.16301031410694122, 0.11493553966283798, 0.3294888734817505, -0.02343812584877014, 0.05227096006274223, 0.1015276312828064, -0.025000033900141716, 0.3100617825984955, -0.08968440443277359, -0.25382405519485474, -0.10262658447027206, -0.1475948989391327, -0.023878805339336395, 0.27968502044677734, 0.12876713275909424, 0.012593789957463741, -0.13629192113876343, 0.05509081855416298, 0.06330803781747818, -0.16161301732063293, 0.3230098783969879, -0.08835263550281525, -0.26840290427207947, 0.0020212058443576097, 0.2783096134662628, 0.1604536920785904, -0.29687368869781494, 0.10166541486978531, 0.3348393440246582, -0.10520195960998535, -0.15593476593494415, 0.3859611451625824, 0.02612285315990448, 0.1852373629808426, -0.059755388647317886, 0.2265554517507553, -0.14647166430950165, 0.022025972604751587, -0.034872230142354965, -0.1694050431251526, 0.014915534295141697, 0.23729732632637024, 0.3928617537021637, -0.039098866283893585, -0.5896610617637634, -0.11377349495887756, 0.09599348902702332, -0.4078333377838135, -0.3410401940345764, -0.45826661586761475, 0.05670666694641113, 0.04793199151754379, -0.18805886805057526, -0.35080042481422424, -0.13403935730457306, 0.1007576659321785, -0.27891290187835693, -0.09075620025396347, 0.01032454427331686, -0.08747880905866623, -0.132045716047287, 0.06137003377079964, 0.04111156612634659, -0.5356751084327698, 0.2628818452358246, -0.08810143917798996, 0.5520691275596619, 0.358929842710495, 0.29697829484939575, -0.3132716715335846, 0.30237382650375366, -0.043867431581020355, 0.17262272536754608, 0.5356281995773315, -0.4726382791996002, -0.2246147096157074, 0.007902959361672401, -0.036810580641031265, -0.17611023783683777, 0.11352554708719254, 0.19710126519203186, 0.12219531834125519, -0.06313980370759964, 0.0917976051568985, 0.38595688343048096, 0.05037359148263931, 0.01872454769909382, -0.19749271869659424, -0.16739949584007263, 0.2501934766769409, -0.02691328339278698, -0.16875770688056946, 0.11825582385063171, 0.2636357843875885, 0.02896646223962307, 0.030211251229047775, -0.31333810091018677, 0.11959073692560196, 0.21816331148147583, 0.3369303047657013, -0.18201296031475067, -0.1540910005569458, 0.28230616450309753, -0.40281209349632263, 0.16599220037460327, 0.17265655100345612, 0.3474859595298767, -0.055268265306949615, -0.044146131724119186, -0.4338451325893402, -0.11113903671503067, -0.020370464771986008, -0.025351589545607567, 0.04577556252479553, 0.11559565365314484, 0.02293495461344719, -0.23246541619300842, -0.2718803584575653, 0.10688909888267517, -0.260445237159729, 0.049384523183107376, -0.2114427387714386, -0.03956311196088791, -0.32546699047088623, -0.06348545104265213, 0.25599968433380127, 0.11972057074308395, 0.203800767660141, -0.012565544806420803, -0.16738729178905487, 0.10784420371055603, -0.424612432718277, 0.26699554920196533, -0.03085540421307087, 0.1198558583855629, 0.007746321149170399, -0.006525013595819473, -0.2597024738788605, -0.07866542786359787, 0.22222734987735748, -0.20514318346977234, -0.38404110074043274, 0.16649070382118225, -0.009620128199458122, 0.0653088390827179, 0.12813542783260345, 0.008446226827800274, 0.29795268177986145, -0.1877760887145996, 0.06347144395112991, 0.14335060119628906, 0.3599367141723633, 0.11930446326732635, 0.380616158246994, -0.021553587168455124, -0.12040462344884872, -0.15498526394367218, 0.168571338057518, 0.010845196433365345, 0.13222698867321014, 0.058391790837049484, 0.06193958595395088, -0.07862592488527298, 0.3246951699256897, 0.5969868302345276, 0.38469329476356506, 0.03554309532046318, -0.06499948352575302, -0.00805889442563057, -0.10091211646795273, -0.23450498282909393, -0.025496019050478935, -0.04307086393237114, 0.04232781380414963, 0.1619136929512024, 0.3972325325012207, 0.2295970916748047, -0.0900120809674263, 0.027444973587989807, 0.1321055144071579, 0.42635658383369446, -0.21294178068637848, 0.26301318407058716, -0.10497554391622543, -0.41205841302871704, 0.2812418043613434, -0.24620576202869415, -0.06726260483264923, -0.15471290051937103, -0.18189290165901184, 0.1328592747449875, 0.10234759002923965, -0.1452893763780594, 0.156103253364563, 0.37681496143341064, -0.11184222996234894, -0.2587057948112488, -0.23002545535564423, -0.09443657845258713, -0.2048044204711914, 0.022891538217663765, 0.28228190541267395, -0.07962038367986679, 0.39604300260543823, 0.2977892756462097, 0.059124983847141266, 0.0036392686888575554, -0.30800044536590576, -0.018483592197299004, -0.0448782816529274, 0.10120942443609238, 0.09834782779216766, 0.38181567192077637, 0.011832541786134243, -0.5280577540397644, 0.1464276909828186, 0.07743372768163681, -0.2765723764896393, 0.24847868084907532, -0.02130456455051899, -0.3578559458255768, -0.48331713676452637, -0.02597077190876007, -0.23736800253391266, -0.37235867977142334, 0.23394285142421722, 0.34234634041786194, 0.20104339718818665, 0.6263755559921265, 0.033080730587244034, 0.10552367568016052, 0.3673250079154968, 0.15832488238811493, 0.03778167814016342, -0.14254778623580933, 0.4890120029449463, -0.21970705687999725, -0.4154713451862335, 0.14479434490203857, -0.15794134140014648, 0.5447534322738647, -0.07129577547311783, -0.3519459664821625, -0.18232230842113495, 0.0811840146780014, 0.14644137024879456, -0.0377870537340641, -0.0391998253762722, 0.31707510352134705, -0.24508918821811676, 0.06349536031484604, -0.07407766580581665, -0.3456256091594696, 0.1111992672085762, 0.6317629218101501, 0.2669854760169983, 0.0019010058604180813, 0.46685710549354553, 0.0014614997198805213, 0.7395949959754944, 0.07472385466098785, -0.29264482855796814, 0.4109627306461334, -0.10046782344579697, 0.2557585537433624, -0.17619848251342773, -0.32717758417129517, 0.027622871100902557, 0.18094271421432495, -0.12938292324543, 0.12235841900110245, -0.056772105395793915, -0.07398997247219086, -0.4756855070590973, 0.30633047223091125, -0.29413044452667236, 0.04968637973070145, 0.010090029798448086, 0.10700753331184387, -0.009691133163869381, -0.0785011425614357, -0.22861485183238983, -0.33661964535713196, -0.4889744222164154, -0.18894575536251068, 0.2726055085659027, -0.2198995053768158, 0.1720753163099289, -0.6565858125686646, -0.13269881904125214, -0.6450006365776062, 0.3082122206687927, -0.16790395975112915, 0.30234792828559875, -0.18335291743278503, 0.1802239567041397, 0.17021037638187408, -0.06942994147539139, 0.1479044407606125, -0.38897278904914856, -0.20179705321788788, -0.10400496423244476, 0.3181890547275543, -0.42693406343460083, -0.06536216288805008, -0.07280009239912033, 0.08206387609243393, 0.48182621598243713, 0.20355448126792908, -0.014485826715826988, 0.0970020517706871, -0.1921156346797943, -0.18257485330104828, -0.20688681304454803, -0.21783840656280518, -0.08103422075510025, -0.6945074200630188, -0.3678606152534485, -0.501621663570404, 0.12273713201284409, 0.13900983333587646, 0.14689333736896515, -0.030242672190070152, 0.21095696091651917, 0.08789132535457611, 0.48534300923347473, 0.004931125324219465, -0.05993637442588806, 0.6008954048156738, -0.16515691578388214, -0.4717911183834076, 0.4963057041168213, -0.1177106499671936, 0.37814241647720337, 0.09147702902555466, -0.15693695843219757, -0.13820070028305054, -0.3253120183944702, 0.47440239787101746, 0.23392291367053986, 0.10316669940948486, 0.28175362944602966, -0.06286624073982239, 0.026009436696767807, -0.15068212151527405, 0.47322386503219604, 0.3427961468696594, -0.1364431381225586, -0.3455304503440857, -0.1474507600069046, 0.4850681722164154, -0.20866632461547852, -0.061527978628873825, 0.2201673984527588, -0.20480971038341522, -0.23179969191551208, 0.21763405203819275, 0.16481976211071014, 1.2438884973526, -0.059596143662929535, 0.4256587624549866, 0.07456737011671066, -0.03275998681783676, 0.7743026614189148, -0.16633479297161102, 0.06815861165523529, -0.23189310729503632, 0.13499392569065094, -0.14140065014362335, -0.06125318259000778, -0.1139100044965744, 0.11026684194803238, -0.2297612577676773, 0.10841439664363861, -0.11354093253612518, -0.04289832338690758, 0.0006191310239955783, 0.45125293731689453, 0.10172713547945023, -0.3414742946624756, -0.1738097369670868, 0.06134425848722458, -0.25388655066490173, 0.31335514783859253, -0.17983432114124298, -0.21803918480873108, -0.3014184534549713, -0.23261965811252594, -0.18177835643291473, 0.24452091753482819, -0.26989027857780457, 0.07795186340808868, 0.0035697263665497303, -0.322989284992218, 0.17310549318790436, 0.4630567133426666, 0.5105308890342712, -0.06357796490192413, -0.06871581822633743, -0.009666213765740395, -0.08557648956775665, -0.042368337512016296, -0.09295255690813065, 0.06018519401550293, 0.13630172610282898, -0.29558876156806946, 0.07367662340402603, 0.32115501165390015, -0.08515814691781998, -0.3732733130455017, -0.210207000374794, 0.19486309587955475, 0.05941299721598625, -0.1340412050485611, -0.3536510765552521, 0.06596330553293228, -0.2124883234500885, -0.015740765258669853, 0.004534284118562937, 0.0331510528922081, 0.08300954103469849, -0.09015650302171707, 0.08961211144924164, -0.1979520171880722, 0.09447413682937622, 0.06312543898820877, -0.041648369282484055, -0.048991959542036057, 0.7007601261138916, -0.12110807001590729, 0.08024631440639496, -0.04312048479914665, -0.20529378950595856, -0.33442625403404236, -0.14639431238174438, -0.07787784188985825, 0.31912147998809814, -0.030446387827396393, 0.03915867581963539, 0.6615256667137146, 0.3981337547302246, 0.18721790611743927, 0.06805620342493057, -0.1111069768667221, -0.32315701246261597, -0.3133421838283539, -0.03441201522946358, 0.2028842717409134, -0.025104112923145294, 0.21315990388393402, 0.2098328173160553, 0.2966155409812927, -0.21475595235824585, -0.18158240616321564, -0.20306020975112915, 0.17764778435230255, 0.7618895173072815, -0.013123772107064724, -0.11110194027423859, -0.010394692420959473, -0.033159688115119934, 0.009593193419277668, -0.1480642557144165, -0.1033521443605423, -0.3793177902698517, 0.18479830026626587, 0.5306532979011536, -0.11960560828447342, -0.014834362082183361, -0.23938077688217163, -0.11199558526277542, -0.12136750668287277, -0.32977527379989624, 0.04165039211511612, -0.014645254239439964, 0.20553632080554962, -0.310273677110672, 0.3925666809082031, -0.15368615090847015, 0.1730085164308548, 0.02032838948071003, 0.2511637508869171, 0.06056467443704605, -0.05216382443904877, 0.15323281288146973, 0.07449052482843399, -0.39298903942108154, 0.006554153747856617, -0.177519753575325, -0.0990985855460167, 0.025729764252901077, -0.41810184717178345, 0.1579311639070511, -0.2215924859046936, 0.022271106019616127, 0.04608545079827309, -0.502012312412262, 0.21895769238471985, 0.4416904151439667, 0.08857271820306778, 0.04070447012782097, 0.1663506180047989, 0.07060454785823822, 0.21389886736869812, 0.024117914959788322, 0.2750561535358429, -0.0007001938647590578, -0.30314746499061584, -0.04899748042225838, -0.22703438997268677, -0.11426705121994019, -0.27213290333747864, 0.0433906689286232, -0.06087986379861832, -0.4242880642414093, 0.004609233699738979, 0.1704261749982834, 0.1590813547372818, -0.10351168364286423, 0.15583395957946777, -0.19883903861045837, 0.09525351226329803, 0.05639862269163132, -0.115884929895401, 0.18386009335517883, -0.23433223366737366, 0.12833593785762787, -0.04518166184425354, -0.007981596514582634, 0.33807313442230225, 0.1622440069913864, 0.6163567900657654, -0.3498658835887909, -0.011332525871694088, -0.2669870853424072, 0.058032989501953125, 0.02386714704334736, 0.17970941960811615, -0.66041499376297, -0.005375504028052092, 0.06842418760061264, -0.0865071713924408, 0.03860245645046234, -0.3267976641654968, -0.050380077213048935, 0.2629709243774414, -0.1634490191936493, 0.10786846280097961, -0.13042491674423218, 0.19259412586688995, 0.13164886832237244, -0.11263610422611237, 0.36844468116760254, 0.003191038267686963, -0.03597177192568779, -0.06830361485481262, 0.058665625751018524, 0.5559295415878296, 0.6175878643989563, 0.1681140810251236, 0.27055469155311584, 0.23017160594463348, -0.006906419992446899, 0.07748192548751831, 0.1667325645685196, 0.4616715908050537, 0.3019821345806122, 0.12015044689178467, 0.030812840908765793, -0.12133526802062988, 0.14843040704727173, 0.053318582475185394, 0.09564625471830368, -0.3158126175403595, 0.2280459851026535, -0.02910655550658703, -0.3997562825679779, -0.15640881657600403, 0.4017168879508972, -0.4579068720340729, -0.014089370146393776, 0.1900971680879593, 0.18232694268226624, 0.36742469668388367, -0.23387138545513153, 0.022305456921458244, -0.18495416641235352, 0.39756542444229126, 0.465614378452301, -0.0361272469162941, -0.1464562863111496, -0.21958988904953003, -0.4871543347835541, 0.17760154604911804, -0.22294239699840546, -0.20073741674423218, -0.3758894205093384, 0.1740434616804123, 0.06222346052527428, -0.018662741407752037, 0.07831194996833801, 0.08619540929794312, -0.27417758107185364, 0.023238951340317726, -0.3622927963733673, -0.02059962972998619, -0.12250679731369019, -0.0025124582462012768, -0.07683001458644867, -0.16656650602817535, 0.0395691841840744, 0.11009779572486877, 0.012876717373728752, 0.024147814139723778, 0.24240954220294952, 0.13704968988895416, 0.36730995774269104, 0.4786677360534668, -0.07187210768461227, 0.7685264348983765, 0.06704545766115189, 0.02057424746453762, 0.04930095747113228, -0.04225842282176018, -0.17712591588497162, -0.0018192871939390898, 0.17840659618377686, 0.37813419103622437, -0.05883794277906418, 0.34986403584480286, -0.19776244461536407, 0.06250468641519547, -0.022588305175304413, -0.5348162651062012, 0.05529927462339401, -0.006339341402053833, -0.08719674497842789, 0.2302379608154297, 0.423831045627594, 0.08845052123069763, 0.07060215622186661, -0.1437550187110901, -0.43358689546585083, -0.285406231880188, 0.41403675079345703, -0.5086386799812317, -0.300311416387558, 0.22360950708389282, 0.23953382670879364, -0.02934344671666622, 0.05272889882326126, -0.2230435311794281, -0.0850505381822586, 0.1356162428855896, -0.25086018443107605, -0.05188969522714615, 0.32419681549072266, -0.5186461210250854, 0.01627280004322529, -0.055287498980760574, 0.3049899637699127, 0.2024187445640564, -0.37965095043182373, -0.0708051323890686, -0.21507969498634338 ]
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' ```
[ 0.052604734897613525, -0.02810976654291153, 0.055330462753772736, -0.0020758789032697678, 0.14557722210884094, -0.18475818634033203, 0.33769690990448, -0.047722745686769485, -0.1926681399345398, 0.1797221153974533, -0.1303306668996811, 0.24369049072265625, -0.21156135201454163, 0.27915024757385254, 0.36935654282569885, 0.06986390799283981, 0.013265492394566536, 0.2872205972671509, 0.012621807865798473, -0.10832816362380981, -0.2543199360370636, 0.09193605929613113, -0.01930956169962883, -0.08095770329236984, -0.12342623621225357, -0.19380053877830505, -0.08119558542966843, 0.26023659110069275, -0.22382193803787231, -0.3616124093532562, 0.491118848323822, 0.23479163646697998, 0.36471033096313477, 0.5793420672416687, -0.00011945605365326628, -0.048161305487155914, 0.16207435727119446, -0.22125212848186493, -0.6242918968200684, -0.40795353055000305, 0.020531922578811646, -0.15109440684318542, 0.07041089981794357, -0.25698891282081604, 0.18926115334033966, 0.0231859739869833, 0.021706894040107727, 0.04705110937356949, 0.48964205384254456, 0.406698077917099, 0.1377582997083664, 0.1278693825006485, -0.2263694405555725, -0.13977232575416565, 0.007812514901161194, 0.17320719361305237, -0.05875520408153534, 0.25951093435287476, 0.334825336933136, -0.40046247839927673, 0.27344295382499695, -0.19609785079956055, 0.012751743197441101, -0.03790641203522682, 0.48874038457870483, -0.021753940731287003, -0.060725755989551544, -0.3897159993648529, -0.01117783784866333, 0.15420478582382202, 0.3593457043170929, -0.017493102699518204, -0.3905928134918213, -0.12496677041053772, 0.13513141870498657, -0.4883084297180176, 0.2485235184431076, 0.27409619092941284, -0.4021522104740143, -0.02720935456454754, -0.2619931697845459, 0.3479142487049103, -0.40307870507240295, 0.4445771872997284, 0.10771405696868896, 0.370847225189209, -0.03705955669283867, 0.09139969199895859, 0.2884839177131653, -0.1540011763572693, 0.022260889410972595, 0.03814223036170006, 0.028013020753860474, 0.2681345045566559, -0.21565786004066467, -0.15431611239910126, 0.011639843694865704, -0.23738563060760498, -0.0875120684504509, 0.11485078930854797, 0.4278506934642792, -0.12593182921409607, -0.24697120487689972, 0.14052724838256836, 0.10184258222579956, 0.33750784397125244, 0.023061582818627357, 0.12450449913740158, 0.1315499097108841, 0.0036603352054953575, 0.24482892453670502, -0.0774664357304573, -0.29436540603637695, -0.08862260729074478, -0.11200881749391556, -0.052120137959718704, 0.26885518431663513, 0.07469945400953293, -0.007578371092677116, -0.10847905278205872, 0.07862365990877151, 0.04231103137135506, -0.08323118835687637, 0.3526008427143097, -0.05940303951501846, -0.23659157752990723, -0.02016669698059559, 0.2576301097869873, 0.14143432676792145, -0.26095637679100037, 0.0658906102180481, 0.2955015301704407, -0.12431101500988007, -0.14895279705524445, 0.4141160845756531, 0.023555513471364975, 0.19150902330875397, -0.06857930123806, 0.2129645198583603, -0.11405239254236221, 0.056869808584451675, 0.029537547379732132, -0.16482406854629517, 0.01724584773182869, 0.2235615998506546, 0.3928874731063843, -0.05233896151185036, -0.5404903292655945, -0.14034460484981537, 0.023450428619980812, -0.34771981835365295, -0.3577896058559418, -0.42381617426872253, 0.07509469240903854, 0.020578566938638687, -0.24025282263755798, -0.3375001549720764, -0.09723293781280518, 0.0914815217256546, -0.28204506635665894, -0.09507939964532852, -0.0451882965862751, -0.15627166628837585, -0.116360142827034, 0.11207814514636993, 0.06296421587467194, -0.4992424547672272, 0.2119518369436264, -0.07937256991863251, 0.5196447968482971, 0.27601704001426697, 0.3038299083709717, -0.34270745515823364, 0.23301410675048828, -0.06637686491012573, 0.18478037416934967, 0.5762104392051697, -0.4968986511230469, -0.3117561638355255, 0.04671184718608856, -0.030563941225409508, -0.11799074709415436, 0.07224192470312119, 0.17157958447933197, 0.11363720893859863, -0.08424389362335205, 0.10218102484941483, 0.4385213255882263, 0.06725272536277771, 0.04018286615610123, -0.2208322286605835, -0.15429775416851044, 0.25438058376312256, 0.036983996629714966, -0.1506790816783905, 0.08312273025512695, 0.26047438383102417, 0.06287042796611786, 0.010456550866365433, -0.32887977361679077, 0.14159712195396423, 0.21016982197761536, 0.35805782675743103, -0.16867801547050476, -0.12786753475666046, 0.21730636060237885, -0.43105441331863403, 0.21843907237052917, 0.18919073045253754, 0.3485828936100006, -0.1336555927991867, -0.036464203149080276, -0.43643492460250854, -0.10224472731351852, -0.10477100312709808, -0.048560816794633865, 0.04172855615615845, 0.1295848935842514, 0.0738750621676445, -0.1800091564655304, -0.22281429171562195, 0.15982909500598907, -0.2588523328304291, 0.06837025284767151, -0.2100192755460739, 0.03586219996213913, -0.27388012409210205, -0.09568636119365692, 0.26330748200416565, 0.16703712940216064, 0.21565207839012146, -0.01599401980638504, -0.20171095430850983, 0.09329312294721603, -0.4534428119659424, 0.28580009937286377, 0.017559092491865158, 0.10015597194433212, 0.0362418070435524, -0.05520618334412575, -0.21382001042366028, -0.10987984389066696, 0.17156246304512024, -0.15249338746070862, -0.3677259087562561, 0.17456941306591034, -0.004488814156502485, 0.03608423471450806, 0.16663658618927002, 0.010147941298782825, 0.29598468542099, -0.17794927954673767, 0.09211887419223785, 0.09837894886732101, 0.41748321056365967, 0.09317213296890259, 0.4033571779727936, -0.011018456891179085, -0.14266125857830048, -0.08572538942098618, 0.16825567185878754, -0.0006486927741207182, 0.15058524906635284, 0.07491092383861542, 0.024227704852819443, -0.05125832557678223, 0.3066673278808594, 0.6624006032943726, 0.3770545423030853, 0.08588311076164246, -0.026365306228399277, -0.051960110664367676, -0.12508495151996613, -0.21065685153007507, -0.04003207013010979, -0.07790292799472809, 0.05889212340116501, 0.18291549384593964, 0.4417642056941986, 0.23638854920864105, -0.13617059588432312, -0.0484122596681118, 0.1532735526561737, 0.4534132480621338, -0.19562551379203796, 0.2583774924278259, -0.16094240546226501, -0.42994070053100586, 0.24248899519443512, -0.2555161118507385, -0.06268535554409027, -0.18169526755809784, -0.21708352863788605, 0.14613744616508484, 0.1071714535355568, -0.08416219055652618, 0.10070189088582993, 0.33156877756118774, -0.10592742264270782, -0.28663918375968933, -0.2544068396091461, -0.11374922096729279, -0.1918031871318817, 0.027764160186052322, 0.3250078558921814, -0.08840090036392212, 0.4124763011932373, 0.32488036155700684, 0.08864571154117584, -0.04613983631134033, -0.3268384635448456, -0.004596880171447992, -0.0374094694852829, 0.1208917647600174, 0.12138739228248596, 0.3563719093799591, 0.008664911612868309, -0.5445622205734253, 0.1948815882205963, 0.0784640908241272, -0.2436254322528839, 0.25166037678718567, -0.03253914788365364, -0.34858906269073486, -0.46765050292015076, -0.02082732319831848, -0.27371224761009216, -0.3687702417373657, 0.23967106640338898, 0.29409343004226685, 0.2099173218011856, 0.5754039883613586, 0.017711931839585304, 0.1252833753824234, 0.32062745094299316, 0.12391529232263565, 0.06834224611520767, -0.18604686856269836, 0.5255688428878784, -0.24559280276298523, -0.42524829506874084, 0.10256610810756683, -0.16349102556705475, 0.5601884722709656, -0.009719010442495346, -0.3501392602920532, -0.17843426764011383, 0.05947922542691231, 0.21830566227436066, -0.023407286033034325, -0.06744974106550217, 0.2789645195007324, -0.2049720734357834, 0.0612633042037487, -0.08193545043468475, -0.3369685411453247, 0.13447333872318268, 0.6249411702156067, 0.2558962106704712, -0.04228180646896362, 0.4419952630996704, 0.03808284550905228, 0.702817440032959, 0.10719973593950272, -0.33854779601097107, 0.4494321942329407, -0.038514576852321625, 0.18772773444652557, -0.13236775994300842, -0.3693930208683014, 0.06506077945232391, 0.22544583678245544, -0.061309121549129486, 0.13972721993923187, -0.022241167724132538, -0.10212083160877228, -0.4349319636821747, 0.23016828298568726, -0.2566096782684326, 0.048612866550683975, -0.005603456404060125, 0.1051207035779953, 0.05551037937402725, -0.06783249229192734, -0.21281284093856812, -0.3788439631462097, -0.4829709827899933, -0.16287334263324738, 0.274996280670166, -0.23916342854499817, 0.17479725182056427, -0.690329909324646, -0.09328947216272354, -0.660233199596405, 0.30017462372779846, -0.13872231543064117, 0.3314148187637329, -0.2023530751466751, 0.20676839351654053, 0.11336009949445724, -0.07665114849805832, 0.12803205847740173, -0.4650159478187561, -0.16797976195812225, -0.07704340666532516, 0.27476826310157776, -0.47235915064811707, -0.04191272705793381, -0.027604997158050537, 0.07491307705640793, 0.4365346133708954, 0.2350970208644867, 0.018999585881829262, 0.07955574244260788, -0.17663262784481049, -0.20306070148944855, -0.15370070934295654, -0.22335883975028992, -0.03689352795481682, -0.671158492565155, -0.35238781571388245, -0.44758138060569763, 0.06721347570419312, 0.1456538438796997, 0.14058217406272888, -0.037643156945705414, 0.1860169619321823, 0.11092223227024078, 0.5076406598091125, 0.04284252971410751, -0.04625982791185379, 0.6295244693756104, -0.09305112808942795, -0.45713672041893005, 0.4725044369697571, -0.009942784905433655, 0.38947999477386475, 0.0355948731303215, -0.1942717581987381, -0.1696307212114334, -0.2850334346294403, 0.45783311128616333, 0.17230325937271118, 0.08548005670309067, 0.21727736294269562, -0.1069505363702774, 0.02812555991113186, -0.15494033694267273, 0.46995556354522705, 0.33896535634994507, -0.09147336333990097, -0.41813501715660095, -0.18395139276981354, 0.44957318902015686, -0.17907258868217468, -0.03373924270272255, 0.2428112030029297, -0.1924963891506195, -0.2763461172580719, 0.2006179243326187, 0.11939814686775208, 1.0911482572555542, -0.0767856165766716, 0.37884509563446045, 0.12213306128978729, -0.04874815791845322, 0.7742531895637512, -0.14554929733276367, 0.045702993869781494, -0.23358850181102753, 0.10043316334486008, -0.1133071556687355, -0.058808714151382446, -0.06760459393262863, 0.14117109775543213, -0.23739980161190033, 0.0841975063085556, -0.04000546410679817, -0.11120779067277908, 0.0072577474638819695, 0.5523280501365662, 0.07270994037389755, -0.3941996693611145, -0.22025954723358154, 0.08384989202022552, -0.31129831075668335, 0.3093802332878113, -0.16995736956596375, -0.21965549886226654, -0.2857390344142914, -0.18248261511325836, -0.20725467801094055, 0.2760237157344818, -0.2514289319515228, 0.09939267486333847, 0.04887310042977333, -0.35540103912353516, 0.31920087337493896, 0.46335411071777344, 0.5242566466331482, -0.07650446146726608, -0.09716564416885376, 0.049891747534275055, -0.1302093267440796, -0.014176653698086739, -0.03877831622958183, 0.030591856688261032, 0.19029143452644348, -0.26959821581840515, 0.04275330528616905, 0.2530285120010376, -0.025740323588252068, -0.31226587295532227, -0.177422434091568, 0.14603470265865326, 0.04673486575484276, -0.16329139471054077, -0.3530452847480774, 0.08665294200181961, -0.1731015145778656, -0.044460032135248184, 0.027808673679828644, 0.02873469889163971, 0.07372542470693588, -0.06803662329912186, 0.09252864867448807, -0.2007802277803421, 0.08086942136287689, 0.10644633322954178, -0.11565113812685013, -0.058839522302150726, 0.7411609888076782, -0.06006137281656265, 0.07929185777902603, -0.08564344793558121, -0.19852125644683838, -0.31819090247154236, -0.1407570093870163, -0.08977312594652176, 0.27951017022132874, -0.030942494049668312, 0.08318286389112473, 0.6403139233589172, 0.3384789228439331, 0.10922110825777054, 0.16181790828704834, -0.14827963709831238, -0.3228500187397003, -0.30606070160865784, -0.043809112161397934, 0.1931377351284027, -0.10730196535587311, 0.1915976107120514, 0.20194734632968903, 0.29810234904289246, -0.22825747728347778, -0.1637887954711914, -0.2005677968263626, 0.16704167425632477, 0.7540920376777649, -0.01161434967070818, -0.06655433028936386, -0.036773163825273514, -0.01890006847679615, 0.002395035233348608, -0.1610787957906723, -0.11697705835103989, -0.381463885307312, 0.17557677626609802, 0.4809396266937256, -0.18953651189804077, -0.04248183220624924, -0.240421324968338, -0.09950734674930573, -0.141781285405159, -0.2967750132083893, 0.06756313145160675, -0.03371055796742439, 0.240889772772789, -0.26430991291999817, 0.33600613474845886, -0.09664478898048401, 0.21474190056324005, 0.006566926836967468, 0.17916470766067505, 0.08747515082359314, -0.022731004282832146, 0.19590197503566742, 0.07748415321111679, -0.37948134541511536, -0.038136936724185944, -0.20470303297042847, -0.06822069734334946, 0.028972098603844643, -0.4525884687900543, 0.13075189292430878, -0.11226863414049149, 0.04397566244006157, 0.03578454256057739, -0.4939001500606537, 0.1632535755634308, 0.39178842306137085, 0.10325771570205688, 0.014925550669431686, 0.11850062012672424, 0.16980460286140442, 0.2038426399230957, 0.020077433437108994, 0.2445470690727234, 0.034878578037023544, -0.2913967967033386, -0.051735661923885345, -0.20435793697834015, -0.0584847591817379, -0.2675394117832184, 0.12125473469495773, -0.09284712374210358, -0.3782728612422943, -0.04326389357447624, 0.15339526534080505, 0.18866707384586334, -0.1023828536272049, 0.19348889589309692, -0.20576736330986023, 0.12228444218635559, 0.0327523872256279, -0.07289689034223557, 0.21817176043987274, -0.27381035685539246, 0.11131168901920319, -0.06723932176828384, -0.02377392165362835, 0.3217805027961731, 0.14993244409561157, 0.6510176658630371, -0.35149484872817993, -0.005461457651108503, -0.2674219012260437, 0.110259048640728, 0.009168950840830803, 0.14395932853221893, -0.6295010447502136, -0.02550334855914116, 0.05436006933450699, -0.08338434249162674, 0.013343708589673042, -0.3383151888847351, -0.03636635094881058, 0.28414806723594666, -0.14115628600120544, 0.05001087114214897, -0.10629693418741226, 0.210164874792099, 0.1268288940191269, -0.15633128583431244, 0.4182301461696625, 0.01713680289685726, -0.10469288378953934, -0.04810405150055885, 0.07154310494661331, 0.5066620707511902, 0.5816252827644348, 0.13121724128723145, 0.2830183207988739, 0.1813768744468689, 0.03374316543340683, 0.06061118468642235, 0.19215203821659088, 0.4136764109134674, 0.29303279519081116, 0.14628487825393677, 0.029231693595647812, -0.11405081301927567, 0.07913728803396225, 0.04990001767873764, 0.057076435536146164, -0.31889382004737854, 0.28128981590270996, -0.031187495216727257, -0.3361920416355133, -0.16785842180252075, 0.3929539918899536, -0.4424988627433777, 0.011144530959427357, 0.26058071851730347, 0.22482050955295563, 0.353773832321167, -0.25676023960113525, 0.02439052425324917, -0.16960231959819794, 0.41601434350013733, 0.4884306490421295, -0.07816924154758453, -0.1432207226753235, -0.29344624280929565, -0.5263963937759399, 0.16106976568698883, -0.21628454327583313, -0.231517493724823, -0.3568781614303589, 0.18111124634742737, 0.05725966766476631, -0.09626953303813934, 0.05815539509057999, 0.14493902027606964, -0.23265387117862701, -0.004406028892844915, -0.4434366822242737, 0.014719784259796143, -0.15379782021045685, -0.01426580362021923, -0.038493312895298004, -0.16828307509422302, 0.02952456660568714, 0.09202312678098679, 0.018767833709716797, -0.006674290169030428, 0.22007077932357788, 0.13326260447502136, 0.3443257510662079, 0.47306180000305176, -0.09442444890737534, 0.7783324122428894, 0.0928485095500946, 0.004618179053068161, -0.006963617168366909, -0.06584227085113525, -0.16452108323574066, 0.025168526917696, 0.13609907031059265, 0.46698814630508423, -0.0853172242641449, 0.3731721341609955, -0.2298850268125534, 0.03581183776259422, -0.031407345086336136, -0.5221917629241943, 0.07673224806785583, 0.009969019331037998, -0.06821568310260773, 0.2494249939918518, 0.42410147190093994, 0.1379919946193695, 0.09841859340667725, -0.10654041916131973, -0.42790064215660095, -0.311104416847229, 0.41429758071899414, -0.5499927401542664, -0.33335110545158386, 0.19432814419269562, 0.22382156550884247, -0.05248158052563667, 0.02269347757101059, -0.2831406593322754, -0.07024017721414566, 0.1561148315668106, -0.2359621226787567, -0.019298378378152847, 0.305649071931839, -0.43189018964767456, 0.07046491652727127, -0.052911560982465744, 0.31289181113243103, 0.19471444189548492, -0.3801872730255127, -0.040246885269880295, -0.16324973106384277 ]
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
[ 0.010165240615606308, -0.014697768725454807, 0.06399791687726974, -0.04725101217627525, 0.13680754601955414, -0.14406216144561768, 0.37034836411476135, -0.03526405245065689, -0.1715516299009323, 0.15387865900993347, -0.10310790687799454, 0.2589789032936096, -0.23088803887367249, 0.3168736696243286, 0.37554576992988586, 0.09489147365093231, 0.01937245950102806, 0.2717735767364502, -0.020116714760661125, -0.09850385040044785, -0.22678795456886292, 0.07013359665870667, -0.0022228264715522528, -0.09574878960847855, -0.14261749386787415, -0.211626335978508, -0.0903276652097702, 0.2581227123737335, -0.23455414175987244, -0.34122946858406067, 0.48686009645462036, 0.23599067330360413, 0.34359243512153625, 0.5804954171180725, -0.00011795770842581987, -0.051996100693941116, 0.17183156311511993, -0.2157052755355835, -0.6393761038780212, -0.35548797249794006, -0.0061305672861635685, -0.1342170387506485, 0.06067762151360512, -0.2517407238483429, 0.19287487864494324, 0.027462197467684746, 0.03491726517677307, 0.03088015504181385, 0.48228883743286133, 0.4076222777366638, 0.14891524612903595, 0.14572235941886902, -0.1781214326620102, -0.15002456307411194, -0.00283222459256649, 0.1824355274438858, -0.058972034603357315, 0.25395670533180237, 0.358696311712265, -0.4081471860408783, 0.26444247364997864, -0.1995871663093567, 0.036781083792448044, -0.044240646064281464, 0.4874853789806366, -0.009249704889953136, -0.07491907477378845, -0.4059666395187378, -0.004839888773858547, 0.17713762819766998, 0.3769998848438263, -0.06002658233046532, -0.3945047855377197, -0.14569073915481567, 0.15505440533161163, -0.4790179133415222, 0.2537236213684082, 0.27886903285980225, -0.4247058629989624, -0.022530339658260345, -0.27115681767463684, 0.313744455575943, -0.40935614705085754, 0.4221656322479248, 0.10564938187599182, 0.35269659757614136, -0.0337098091840744, 0.06668226420879364, 0.24374648928642273, -0.1536940187215805, 0.011768690310418606, 0.06547364592552185, 0.02146664448082447, 0.27826330065727234, -0.21666431427001953, -0.19502942264080048, 0.0027933120727539062, -0.2405848652124405, -0.05818108096718788, 0.12915553152561188, 0.4047311544418335, -0.12968307733535767, -0.2442007213830948, 0.15063317120075226, 0.13368667662143707, 0.3241249620914459, 0.0014833194436505437, 0.12236350029706955, 0.12233434617519379, -0.005845759529620409, 0.2769153118133545, -0.0517423115670681, -0.2742709815502167, -0.050802670419216156, -0.11301040649414062, -0.06884700804948807, 0.2920278012752533, 0.06772636622190475, -0.005540801677852869, -0.1293468475341797, 0.09735879302024841, 0.01658647507429123, -0.10256070643663406, 0.34392380714416504, -0.0819462314248085, -0.26130321621894836, -0.053804393857717514, 0.23263372480869293, 0.1370869278907776, -0.29001033306121826, 0.05705771967768669, 0.2758285403251648, -0.1154928207397461, -0.154432475566864, 0.43567249178886414, 0.012411084957420826, 0.1956508904695511, -0.05445246398448944, 0.20219413936138153, -0.0660800114274025, 0.0722418874502182, 0.0119866281747818, -0.1628802865743637, 0.00868168007582426, 0.2195931077003479, 0.384787917137146, -0.06849811971187592, -0.5333331227302551, -0.14358177781105042, 0.026160957291722298, -0.35323747992515564, -0.3473283052444458, -0.45941483974456787, 0.0884740948677063, 0.0037648621946573257, -0.2817482650279999, -0.3244451582431793, -0.07150527834892273, 0.08965371549129486, -0.2517108917236328, -0.09521695226430893, -0.0463000126183033, -0.14219065010547638, -0.12248463928699493, 0.12956233322620392, 0.08803893625736237, -0.5273423194885254, 0.18487177789211273, -0.08601383119821548, 0.4892786741256714, 0.2712757885456085, 0.2785509526729584, -0.3086809515953064, 0.2438851147890091, -0.08011394739151001, 0.18975168466567993, 0.5997768640518188, -0.4807256758213043, -0.33468449115753174, 0.042785998433828354, -0.03731629624962807, -0.09002778679132462, 0.037278641015291214, 0.1934349089860916, 0.10801941156387329, -0.09636007249355316, 0.09529438614845276, 0.44792279601097107, 0.04567878320813179, 0.016692858189344406, -0.216155543923378, -0.16694770753383636, 0.2624109387397766, 0.040792349725961685, -0.15550221502780914, 0.07853050529956818, 0.2734293043613434, 0.08999689668416977, 0.024154871702194214, -0.3324548304080963, 0.1537163257598877, 0.1996031552553177, 0.3573831617832184, -0.17576821148395538, -0.13709259033203125, 0.2060297578573227, -0.46960514783859253, 0.21702037751674652, 0.18055644631385803, 0.34908387064933777, -0.1210264191031456, -0.018303420394659042, -0.4437152147293091, -0.10057171434164047, -0.1240309625864029, -0.058668725192546844, 0.049087703227996826, 0.11639878153800964, 0.054980918765068054, -0.17006053030490875, -0.2535091042518616, 0.18731683492660522, -0.27690744400024414, 0.061252981424331665, -0.21974235773086548, 0.04505937546491623, -0.2696878910064697, -0.12417373806238174, 0.25872766971588135, 0.18245269358158112, 0.16576695442199707, 0.006094505079090595, -0.1836494505405426, 0.10668756067752838, -0.47848543524742126, 0.26239877939224243, 0.033690717071294785, 0.1265527755022049, 0.04126238822937012, -0.06761747598648071, -0.1832844763994217, -0.1436624526977539, 0.17080573737621307, -0.14655129611492157, -0.40138426423072815, 0.14321783185005188, -0.027938373386859894, 0.004520689137279987, 0.1643313467502594, 0.00042179523734375834, 0.2795955538749695, -0.17933112382888794, 0.11594346910715103, 0.10681840032339096, 0.38520437479019165, 0.06651092320680618, 0.4020828306674957, -0.041343849152326584, -0.1261364072561264, -0.03238103538751602, 0.2109307199716568, 0.01925469934940338, 0.18925927579402924, 0.07365380227565765, 0.013792512938380241, -0.08014336228370667, 0.29153066873550415, 0.6511788964271545, 0.3804547190666199, 0.11107805371284485, -0.0320216529071331, -0.05627478286623955, -0.10829663276672363, -0.1881224364042282, -0.06301776319742203, -0.11955753713846207, 0.0715678259730339, 0.18021826446056366, 0.4236956834793091, 0.25452056527137756, -0.1388217955827713, -0.03602701425552368, 0.12983356416225433, 0.45807674527168274, -0.19086551666259766, 0.23381784558296204, -0.139661505818367, -0.4380410611629486, 0.24693123996257782, -0.22983141243457794, -0.09044498950242996, -0.16917316615581512, -0.2033490389585495, 0.13439378142356873, 0.14506076276302338, -0.08744094520807266, 0.11892320960760117, 0.337837815284729, -0.08473830670118332, -0.2890053987503052, -0.26066428422927856, -0.09300456941127777, -0.20018288493156433, 0.05327601730823517, 0.3222638964653015, -0.10707630962133408, 0.4432598650455475, 0.3364589512348175, 0.08638133853673935, -0.054918352514505386, -0.3232802152633667, 0.008572272025048733, -0.06074142083525658, 0.12722772359848022, 0.10323631763458252, 0.34343209862709045, -0.003099090652540326, -0.5138471126556396, 0.17997051775455475, 0.06420595943927765, -0.27024778723716736, 0.22155006229877472, -0.02242940664291382, -0.31593868136405945, -0.44382137060165405, -0.055347464978694916, -0.29095202684402466, -0.33917298913002014, 0.25748682022094727, 0.32961487770080566, 0.199835866689682, 0.5830566883087158, 0.027131997048854828, 0.13877148926258087, 0.32144829630851746, 0.10150294005870819, 0.049091972410678864, -0.20185287296772003, 0.514277994632721, -0.25085294246673584, -0.4668109714984894, 0.1272703856229782, -0.13542217016220093, 0.5388175845146179, -0.026185257360339165, -0.3549959361553192, -0.20387743413448334, 0.06519965082406998, 0.24580156803131104, -0.03664596006274223, -0.051929228007793427, 0.2837883532047272, -0.19145283102989197, 0.044388946145772934, -0.07229117304086685, -0.3318912088871002, 0.14527855813503265, 0.6041278839111328, 0.27121102809906006, -0.05854294076561928, 0.43397367000579834, 0.043424591422080994, 0.7070649862289429, 0.1327073574066162, -0.33110666275024414, 0.4485013484954834, -0.06909222155809402, 0.20655718445777893, -0.14011989533901215, -0.374726265668869, 0.13324417173862457, 0.23546576499938965, -0.033651091158390045, 0.11279858648777008, -0.030975205823779106, -0.08227454125881195, -0.41568055748939514, 0.2208867222070694, -0.2339077889919281, 0.04839266836643219, -0.016792137175798416, 0.11899584531784058, 0.04606027528643608, -0.08111140877008438, -0.2233678549528122, -0.3637562692165375, -0.5096312165260315, -0.1245180144906044, 0.24588903784751892, -0.2464720904827118, 0.2104235142469406, -0.6700915098190308, -0.05532330274581909, -0.6859091520309448, 0.28730279207229614, -0.14990417659282684, 0.308601051568985, -0.21619312465190887, 0.1993671953678131, 0.09411278367042542, -0.08231575042009354, 0.15053747594356537, -0.47544899582862854, -0.1588873416185379, -0.08539852499961853, 0.2848473787307739, -0.4748273491859436, -0.026424795389175415, -0.05132059007883072, 0.07265031337738037, 0.4124597907066345, 0.19209977984428406, 0.03950868919491768, 0.07596627622842789, -0.141396164894104, -0.21616971492767334, -0.17694474756717682, -0.22804735600948334, -0.04577120766043663, -0.6439386606216431, -0.3358875811100006, -0.46366724371910095, 0.056370608508586884, 0.13767080008983612, 0.163987934589386, -0.06866579502820969, 0.15550382435321808, 0.0947166234254837, 0.5056183338165283, 0.04792390391230583, -0.06624246388673782, 0.638486385345459, -0.09328273683786392, -0.42609381675720215, 0.48501643538475037, 0.003487482899799943, 0.38921913504600525, 0.02805565483868122, -0.1498478204011917, -0.1559254378080368, -0.29983288049697876, 0.47131356596946716, 0.13734617829322815, 0.08705977350473404, 0.18480399250984192, -0.09927671402692795, 0.028123168274760246, -0.1918715238571167, 0.5012048482894897, 0.3349218964576721, -0.08053108304738998, -0.4222894608974457, -0.18815910816192627, 0.4366784691810608, -0.16746816039085388, -0.002293181838467717, 0.2426333725452423, -0.18027819693088531, -0.2683976888656616, 0.17485585808753967, 0.14205577969551086, 1.0450083017349243, -0.09819275885820389, 0.35771864652633667, 0.11459758877754211, -0.03082357719540596, 0.7806406617164612, -0.1800106018781662, 0.045405708253383636, -0.2553688883781433, 0.09904003143310547, -0.11995553970336914, -0.04338901489973068, -0.045687925070524216, 0.13680113852024078, -0.25287312269210815, 0.08574365079402924, -0.029485363513231277, -0.14704401791095734, 0.00572573347017169, 0.5664043426513672, 0.061787668615579605, -0.39426034688949585, -0.2046336680650711, 0.10850322991609573, -0.31704944372177124, 0.2979930341243744, -0.16284997761249542, -0.20761966705322266, -0.28265663981437683, -0.1551133394241333, -0.22094187140464783, 0.3035309612751007, -0.23986954987049103, 0.07949504256248474, 0.058917660266160965, -0.3198927640914917, 0.33261823654174805, 0.4678569734096527, 0.5250081419944763, -0.08236940950155258, -0.10083149373531342, 0.03757582977414131, -0.167093425989151, -0.027054043486714363, 0.009089101105928421, 0.05029298737645149, 0.22911928594112396, -0.29069843888282776, 0.030211862176656723, 0.28145289421081543, -0.0009340427932329476, -0.3020414710044861, -0.1756669282913208, 0.13425414264202118, 0.07122772186994553, -0.16748587787151337, -0.30508026480674744, 0.10473180562257767, -0.13321736454963684, -0.03923353552818298, 0.040146101266145706, 0.047616634517908096, 0.06888847053050995, -0.02861354686319828, 0.07775545120239258, -0.1839817613363266, 0.057182762771844864, 0.11787638813257217, -0.11957657337188721, -0.0726916640996933, 0.752065896987915, -0.053994107991456985, 0.09325079619884491, -0.10994222015142441, -0.1821330338716507, -0.27367833256721497, -0.13815340399742126, -0.09949125349521637, 0.2768901586532593, -0.020947864279150963, 0.09042223542928696, 0.6383236050605774, 0.3213024437427521, 0.09763482958078384, 0.13834616541862488, -0.1498946100473404, -0.3523447513580322, -0.3358490765094757, -0.020538397133350372, 0.18586912751197815, -0.12017842382192612, 0.1458791345357895, 0.21818040311336517, 0.3104356825351715, -0.25146132707595825, -0.1950041502714157, -0.1998579055070877, 0.15835462510585785, 0.7369634509086609, -0.01766316033899784, -0.01290071476250887, -0.06114564463496208, 0.006987787317484617, -0.03010912798345089, -0.17383699119091034, -0.13647127151489258, -0.3693227767944336, 0.16456402838230133, 0.46165916323661804, -0.19141319394111633, -0.05623013526201248, -0.21856914460659027, -0.12847848236560822, -0.1337902843952179, -0.2693404257297516, 0.060402583330869675, -0.05756281316280365, 0.24150045216083527, -0.22885116934776306, 0.32027676701545715, -0.0882505476474762, 0.17375966906547546, 0.03331257402896881, 0.16573983430862427, 0.07580465078353882, -0.01964610442519188, 0.1906687617301941, 0.07480854541063309, -0.33523088693618774, -0.026200691238045692, -0.1774321347475052, -0.07269755750894547, 0.017780348658561707, -0.478832483291626, 0.1080644428730011, -0.1383734792470932, 0.0509609654545784, 0.02226053737103939, -0.49828410148620605, 0.13783934712409973, 0.3878341019153595, 0.11959420144557953, -0.002074282383546233, 0.11066970229148865, 0.20888058841228485, 0.202548548579216, 0.013452445156872272, 0.23718902468681335, 0.021878043189644814, -0.24977058172225952, -0.07447025179862976, -0.1981973946094513, -0.07921231538057327, -0.28159692883491516, 0.09921437501907349, -0.11246880143880844, -0.3867097795009613, -0.04920352250337601, 0.16625893115997314, 0.19723787903785706, -0.0932944118976593, 0.21389195322990417, -0.2051723599433899, 0.11387549340724945, 0.03207774832844734, -0.06195514276623726, 0.22813327610492706, -0.27972227334976196, 0.1353698968887329, -0.09851949661970139, 0.005144017282873392, 0.3340871334075928, 0.14589430391788483, 0.6321767568588257, -0.3162314295768738, -0.016112111508846283, -0.24867360293865204, 0.07460876554250717, 0.01759672351181507, 0.16787536442279816, -0.6061633825302124, -0.029132815077900887, 0.05296077951788902, -0.0852576494216919, 0.010777132585644722, -0.36782437562942505, -0.04822491481900215, 0.2759411633014679, -0.15788517892360687, 0.05883448198437691, -0.09756021201610565, 0.19536925852298737, 0.13777540624141693, -0.12739402055740356, 0.40262722969055176, -0.008652456104755402, -0.0919734314084053, -0.05486856400966644, 0.11255746334791183, 0.4699503183364868, 0.579823911190033, 0.1491088718175888, 0.3191978335380554, 0.20295317471027374, 0.03397463262081146, 0.06528588384389877, 0.18311414122581482, 0.3685247302055359, 0.26249200105667114, 0.13088896870613098, 0.04075699672102928, -0.10797836631536484, 0.05925663560628891, 0.027629010379314423, 0.07058052718639374, -0.3081677258014679, 0.2848705053329468, -0.01995880715548992, -0.3238847851753235, -0.17790165543556213, 0.3976590931415558, -0.42802727222442627, 0.00800661463290453, 0.305936723947525, 0.2120256870985031, 0.3255811631679535, -0.2719452679157257, 0.03361433371901512, -0.18920110166072845, 0.45011457800865173, 0.48201629519462585, -0.09710148721933365, -0.10160069167613983, -0.2517860233783722, -0.5138524770736694, 0.1598954200744629, -0.1947067230939865, -0.2622525095939636, -0.3417937755584717, 0.18984489142894745, 0.036116719245910645, -0.08899540454149246, 0.10701441019773483, 0.13900631666183472, -0.21020157635211945, -0.020036304369568825, -0.4624023735523224, 0.013604172505438328, -0.15904924273490906, 0.009875614196062088, -0.020205583423376083, -0.17562419176101685, 0.043981052935123444, 0.06408040970563889, 0.015800446271896362, -0.0002462867123540491, 0.18995250761508942, 0.11839306354522705, 0.32662031054496765, 0.49685928225517273, -0.10089108347892761, 0.758902370929718, 0.10915166884660721, -0.00811645109206438, -0.012253322638571262, -0.06070784851908684, -0.2116784155368805, 0.07008413970470428, 0.12792520225048065, 0.43845438957214355, -0.08241157978773117, 0.3603611886501312, -0.22094310820102692, 0.07509879767894745, -0.033126749098300934, -0.5146056413650513, 0.08451095223426819, 0.022846562787890434, -0.044427480548620224, 0.27646419405937195, 0.4149312674999237, 0.09248574823141098, 0.0941307544708252, -0.09509813785552979, -0.4262847304344177, -0.27461373805999756, 0.3977465331554413, -0.5510280728340149, -0.29202571511268616, 0.15320394933223724, 0.21421507000923157, -0.032668791711330414, -0.009405218064785004, -0.3166781961917877, -0.06710265576839447, 0.17874476313591003, -0.22816728055477142, -0.011713639833033085, 0.2962084412574768, -0.4524925649166107, 0.09072363376617432, -0.04688050225377083, 0.33023446798324585, 0.1773253232240677, -0.37144556641578674, -0.01620074175298214, -0.14145897328853607 ]
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>`
[ 0.07664868980646133, -0.0518912710249424, 0.05664343386888504, -0.013380457647144794, 0.162171870470047, -0.1869213581085205, 0.3156284689903259, -0.027821427211165428, -0.19460196793079376, 0.19591793417930603, -0.1271069496870041, 0.23585711419582367, -0.22072941064834595, 0.2926351726055145, 0.3533960282802582, 0.0726650059223175, 0.005575459450483322, 0.29510393738746643, 0.009005220606923103, -0.09385784715414047, -0.21986126899719238, 0.09480909258127213, 0.019349951297044754, -0.07987089455127716, -0.1348172277212143, -0.18814025819301605, -0.07919926941394806, 0.27564120292663574, -0.1972326636314392, -0.3347657024860382, 0.4486151337623596, 0.22274263203144073, 0.33252424001693726, 0.6002951860427856, -0.00011811569129349664, -0.05863655358552933, 0.16570234298706055, -0.22807447612285614, -0.6185939908027649, -0.37377285957336426, 0.07301405817270279, -0.14274387061595917, 0.0919017642736435, -0.23838476836681366, 0.18933576345443726, 0.037417326122522354, 0.016469420865178108, 0.0653846263885498, 0.5281583070755005, 0.3778298795223236, 0.14735278487205505, 0.10313776880502701, -0.1705758422613144, -0.09499390423297882, 0.02734413370490074, 0.1893177330493927, -0.0560426265001297, 0.29447853565216064, 0.338643342256546, -0.39923450350761414, 0.29277998208999634, -0.19877874851226807, -0.0034398380666971207, -0.060388773679733276, 0.4824240207672119, -0.027946293354034424, 0.00782924797385931, -0.41043806076049805, -0.029902243986725807, 0.1773935705423355, 0.301823228597641, -0.030178559944033623, -0.359175443649292, -0.11387854069471359, 0.16878537833690643, -0.4841402471065521, 0.25605517625808716, 0.27259188890457153, -0.4102139472961426, -0.038526471704244614, -0.2570941746234894, 0.3279765844345093, -0.4180815815925598, 0.44269421696662903, 0.11960644274950027, 0.3393746018409729, -0.022680839523673058, 0.07667829096317291, 0.25502854585647583, -0.127788707613945, 0.007880906574428082, 0.09094330668449402, 0.021747339516878128, 0.2719121277332306, -0.19632461667060852, -0.18465977907180786, -0.020689940080046654, -0.2897568345069885, -0.10948580503463745, 0.09983224421739578, 0.45074427127838135, -0.1309792846441269, -0.2473733127117157, 0.14947138726711273, 0.09313398599624634, 0.32917821407318115, -0.007071522064507008, 0.09553511440753937, 0.1282755732536316, 0.021282562986016273, 0.2753553092479706, -0.06610722839832306, -0.28716951608657837, -0.07212265580892563, -0.12615807354450226, -0.04732646048069, 0.2747192084789276, 0.08483279496431351, 0.011768665164709091, -0.09488501399755478, 0.0725313350558281, 0.060445673763751984, -0.08017151057720184, 0.3419951796531677, -0.1117427870631218, -0.24170131981372833, 0.0059720114804804325, 0.2359330803155899, 0.1392109990119934, -0.26889756321907043, 0.07524189352989197, 0.3023255169391632, -0.11196796596050262, -0.14454679191112518, 0.404755562543869, 0.008399560116231441, 0.19896051287651062, -0.046806301921606064, 0.18748697638511658, -0.1364452987909317, 0.06247745454311371, 0.006645039655268192, -0.1763249933719635, 0.033435266464948654, 0.2295500934123993, 0.38577964901924133, -0.06581900268793106, -0.5420822501182556, -0.1311398148536682, 0.03096657805144787, -0.3768734633922577, -0.3573203682899475, -0.4452537000179291, 0.08834988623857498, 0.04142205789685249, -0.2367769479751587, -0.3071371614933014, -0.1186675876379013, 0.09791096299886703, -0.2934204339981079, -0.0951496809720993, -0.04661884158849716, -0.12096823006868362, -0.13289515674114227, 0.11520522087812424, 0.04702838882803917, -0.4755053222179413, 0.21962401270866394, -0.08374593406915665, 0.5124321579933167, 0.3262300491333008, 0.2987666726112366, -0.31225261092185974, 0.2711575925350189, -0.06317327916622162, 0.17326472699642181, 0.5793401598930359, -0.4712461233139038, -0.27476492524147034, 0.027949534356594086, -0.0340612418949604, -0.10284299403429031, 0.054656997323036194, 0.17109662294387817, 0.0808371901512146, -0.09934047609567642, 0.0748930498957634, 0.4404069781303406, 0.07227008789777756, 0.04624664783477783, -0.23941275477409363, -0.16745546460151672, 0.258391797542572, 0.03397059813141823, -0.16789855062961578, 0.07635155320167542, 0.24125468730926514, 0.08969412744045258, 0.023809688165783882, -0.3316554129123688, 0.13372574746608734, 0.2027382105588913, 0.3554898798465729, -0.15696746110916138, -0.14370876550674438, 0.23550373315811157, -0.40783795714378357, 0.18856623768806458, 0.12364412099123001, 0.35227730870246887, -0.10161953419446945, -0.01019242126494646, -0.41855713725090027, -0.09694593399763107, -0.09851209819316864, -0.02375189960002899, 0.06040219962596893, 0.1085122674703598, 0.050504304468631744, -0.18064524233341217, -0.22650732100009918, 0.1404365599155426, -0.30637818574905396, 0.057290755212306976, -0.21424981951713562, 0.04307209700345993, -0.28264397382736206, -0.11205191910266876, 0.25318658351898193, 0.17547616362571716, 0.19479574263095856, -0.015051421709358692, -0.19718408584594727, 0.09308919310569763, -0.4591652750968933, 0.3227217495441437, -0.0023455191403627396, 0.13259641826152802, 0.030504202470183372, -0.02273811772465706, -0.2367989718914032, -0.09179305285215378, 0.17833521962165833, -0.16936469078063965, -0.33602121472358704, 0.13825161755084991, -0.02999313548207283, 0.016735803335905075, 0.15873326361179352, 0.0029886073898524046, 0.28925028443336487, -0.18946006894111633, 0.11126899719238281, 0.13147717714309692, 0.40158313512802124, 0.09714185446500778, 0.38154545426368713, -0.04512934386730194, -0.13328248262405396, -0.08486030995845795, 0.18308396637439728, 0.023189185187220573, 0.16448456048965454, 0.07439352571964264, 0.020963255316019058, -0.09218288213014603, 0.30198606848716736, 0.599858283996582, 0.36056649684906006, 0.07620036602020264, -0.05645541101694107, -0.06078992411494255, -0.11490523815155029, -0.22583211958408356, -0.03621252626180649, -0.08589238673448563, 0.06461313366889954, 0.21991629898548126, 0.43957409262657166, 0.26594147086143494, -0.13194304704666138, -0.011397837661206722, 0.10586456209421158, 0.4606073498725891, -0.19288812577724457, 0.24376694858074188, -0.12725606560707092, -0.41526421904563904, 0.24683043360710144, -0.27434298396110535, -0.0827629566192627, -0.1754634827375412, -0.1994417905807495, 0.1508503556251526, 0.09969300031661987, -0.11422137916088104, 0.1735134869813919, 0.32977738976478577, -0.10300758481025696, -0.26729583740234375, -0.286934494972229, -0.08144954591989517, -0.20950132608413696, 0.04389671981334686, 0.339133620262146, -0.1061054989695549, 0.42552557587623596, 0.33081144094467163, 0.05794107913970947, -0.031525325030088425, -0.3174517750740051, -0.012977947480976582, -0.038391850888729095, 0.09033861756324768, 0.08890721946954727, 0.3747076392173767, 0.011475229635834694, -0.5478419065475464, 0.19541186094284058, 0.05433361232280731, -0.2502206563949585, 0.2478957623243332, -0.03155764564871788, -0.31541720032691956, -0.4758594036102295, 0.0024254266172647476, -0.2676185965538025, -0.3631877899169922, 0.24924413859844208, 0.3301575779914856, 0.23138897120952606, 0.5911977887153625, 0.019704049453139305, 0.13168872892856598, 0.3197173476219177, 0.15069182217121124, 0.045514777302742004, -0.1869831383228302, 0.49666300415992737, -0.23607280850410461, -0.426547646522522, 0.10766643285751343, -0.14385592937469482, 0.5702797770500183, -0.044408418238162994, -0.36757585406303406, -0.1872582733631134, 0.08487427979707718, 0.17660591006278992, -0.07252094894647598, -0.0716763287782669, 0.29713714122772217, -0.1921670138835907, 0.0524093359708786, -0.05853385850787163, -0.29756608605384827, 0.13110122084617615, 0.6311604976654053, 0.26334044337272644, -0.05882002413272858, 0.4399246871471405, 0.017980076372623444, 0.6985477209091187, 0.10503315180540085, -0.3126603066921234, 0.4494112730026245, -0.03828611224889755, 0.18113109469413757, -0.14026722311973572, -0.34749284386634827, 0.0887269526720047, 0.2385050654411316, -0.04270147532224655, 0.14452488720417023, -0.04368490353226662, -0.0786023810505867, -0.4418254494667053, 0.24700261652469635, -0.26358747482299805, 0.058943815529346466, 0.009580853395164013, 0.10640182346105576, 0.03759393468499184, -0.10671734064817429, -0.2408447116613388, -0.3441416025161743, -0.5165196657180786, -0.12177073210477829, 0.2631755769252777, -0.2403482347726822, 0.18618828058242798, -0.6647049784660339, -0.12223073095083237, -0.6593774557113647, 0.30211141705513, -0.15353931486606598, 0.29570990800857544, -0.20639850199222565, 0.18639734387397766, 0.1276565045118332, -0.07750583440065384, 0.10924925655126572, -0.41989627480506897, -0.18771298229694366, -0.09577647596597672, 0.26937440037727356, -0.47796308994293213, -0.055133964866399765, -0.0379418283700943, 0.08147388696670532, 0.45575249195098877, 0.20855864882469177, 0.012649916112422943, 0.09506537765264511, -0.16743136942386627, -0.2288465052843094, -0.17120374739170074, -0.242479607462883, -0.047817107290029526, -0.6381924152374268, -0.3416769802570343, -0.4831194579601288, 0.048535868525505066, 0.1684761494398117, 0.1516401469707489, -0.07459689676761627, 0.18773460388183594, 0.09917932003736496, 0.49851417541503906, 0.021204466000199318, -0.05744658783078194, 0.6246474385261536, -0.11789756268262863, -0.44503387808799744, 0.48908868432044983, -0.03532375395298004, 0.3584001064300537, 0.06729511171579361, -0.14457647502422333, -0.17467674612998962, -0.2812870442867279, 0.44760602712631226, 0.18080000579357147, 0.10797969251871109, 0.21889828145503998, -0.11514120548963547, 0.0481148399412632, -0.1877969354391098, 0.4967006742954254, 0.3341119587421417, -0.09505461156368256, -0.3954765498638153, -0.15638883411884308, 0.4564092755317688, -0.197886660695076, -0.0036151697859168053, 0.25355666875839233, -0.22835582494735718, -0.28129127621650696, 0.20379817485809326, 0.16435492038726807, 1.138471245765686, -0.09754544496536255, 0.3732186257839203, 0.1278029829263687, -0.03744091838598251, 0.7614162564277649, -0.18077324330806732, 0.05255873128771782, -0.25659993290901184, 0.14564965665340424, -0.1063767522573471, -0.05241226404905319, -0.0832638069987297, 0.08725988864898682, -0.2380097210407257, 0.07538584619760513, -0.061703749001026154, -0.13530007004737854, -0.005105928517878056, 0.5711607336997986, 0.10040030628442764, -0.3970244526863098, -0.21205267310142517, 0.09203217178583145, -0.3342667818069458, 0.30825507640838623, -0.16845157742500305, -0.23394088447093964, -0.3076724708080292, -0.14399376511573792, -0.1809537410736084, 0.2873726189136505, -0.25357234477996826, 0.07502108812332153, 0.06265726685523987, -0.3308962881565094, 0.26308125257492065, 0.4393526017665863, 0.5037202835083008, -0.07716353982686996, -0.09232006222009659, 0.02560349367558956, -0.13528917729854584, -0.04990973323583603, -0.026900585740804672, 0.07493174076080322, 0.1911635845899582, -0.31154918670654297, 0.07954142242670059, 0.291578084230423, -0.02502903714776039, -0.30229300260543823, -0.18620504438877106, 0.15034280717372894, 0.0951572135090828, -0.14537633955478668, -0.330304890871048, 0.09154888987541199, -0.18308138847351074, -0.03186078369617462, 0.04029685631394386, 0.059759851545095444, 0.04360249266028404, -0.06491032987833023, 0.09557195007801056, -0.18126705288887024, 0.07073995471000671, 0.09366313368082047, -0.09329026937484741, -0.0547143779695034, 0.7555974125862122, -0.07153275609016418, 0.09503114968538284, -0.08538437634706497, -0.21003076434135437, -0.32625406980514526, -0.1590144783258438, -0.09987248480319977, 0.2788126766681671, -0.05178169533610344, 0.08019958436489105, 0.6546950340270996, 0.3361327350139618, 0.12761624157428741, 0.14098501205444336, -0.13919949531555176, -0.3442883789539337, -0.3336317241191864, -0.025999365374445915, 0.1786491870880127, -0.10025949031114578, 0.20802292227745056, 0.19337491691112518, 0.3369990289211273, -0.24393408000469208, -0.18857644498348236, -0.1968175321817398, 0.1518172323703766, 0.7084041237831116, -0.034368522465229034, -0.0673392117023468, -0.02548094093799591, -0.0026783200446516275, -0.025167925283312798, -0.1707734763622284, -0.12747353315353394, -0.3797087073326111, 0.16888122260570526, 0.48321565985679626, -0.18876101076602936, -0.06686556339263916, -0.21625451743602753, -0.10766861587762833, -0.13714717328548431, -0.3264840245246887, 0.035709016025066376, -0.04568510875105858, 0.20570190250873566, -0.22629836201667786, 0.3478628993034363, -0.09009262919425964, 0.20739369094371796, 0.01707553118467331, 0.18031854927539825, 0.058089207857847214, -0.04084552079439163, 0.17120887339115143, 0.062296971678733826, -0.35474345088005066, -0.037898916751146317, -0.18474070727825165, -0.10040660202503204, 0.025786183774471283, -0.43826916813850403, 0.10995157808065414, -0.15907256305217743, 0.043411094695329666, 0.032483939081430435, -0.503900945186615, 0.15125025808811188, 0.39393380284309387, 0.10969091206789017, 0.042538680136203766, 0.12714233994483948, 0.1619916558265686, 0.20843954384326935, -0.017391730099916458, 0.22560830414295197, 0.03903918340802193, -0.2658630609512329, -0.05208052694797516, -0.1688261479139328, -0.12271890044212341, -0.2723003923892975, 0.07296203821897507, -0.14059114456176758, -0.41853341460227966, -0.05919381603598595, 0.14822930097579956, 0.1680251806974411, -0.10553347319364548, 0.15312924981117249, -0.20561039447784424, 0.0897950753569603, 0.018970103934407234, -0.084957554936409, 0.23121879994869232, -0.26628378033638, 0.07315876334905624, -0.08859776705503464, -0.0327511727809906, 0.33546698093414307, 0.1424103081226349, 0.6379615068435669, -0.32824772596359253, -0.0250235665589571, -0.24423664808273315, 0.10706398636102676, 0.005122523754835129, 0.18222667276859283, -0.6148506999015808, -0.01935652457177639, 0.04265180230140686, -0.10466180741786957, 0.030356163159012794, -0.33526840806007385, -0.06210425868630409, 0.305653840303421, -0.17846325039863586, 0.09636448323726654, -0.1424015313386917, 0.2184900939464569, 0.16331425309181213, -0.16216027736663818, 0.40879136323928833, 0.034244559705257416, -0.10550766438245773, -0.08757507055997849, 0.08006265014410019, 0.49889031052589417, 0.5946882963180542, 0.11233997344970703, 0.30674800276756287, 0.21389825642108917, 0.03758174553513527, 0.0789858028292656, 0.20481674373149872, 0.43595919013023376, 0.2809090316295624, 0.11451315879821777, 0.04079470783472061, -0.12631756067276, 0.07522377371788025, 0.031144462525844574, 0.10622445493936539, -0.2927726209163666, 0.3019168972969055, -0.010781555436551571, -0.365828275680542, -0.17470870912075043, 0.3855646550655365, -0.46778783202171326, 0.023126905784010887, 0.24933665990829468, 0.21936675906181335, 0.3385201096534729, -0.258097380399704, 0.03067835606634617, -0.17712633311748505, 0.4321080148220062, 0.48943382501602173, -0.08823613822460175, -0.12464932352304459, -0.2862339913845062, -0.47932538390159607, 0.1886095255613327, -0.19280734658241272, -0.21018357574939728, -0.3527607321739197, 0.18883016705513, 0.04558240994811058, -0.10253987461328506, 0.07314882427453995, 0.09914784133434296, -0.22263698279857635, -0.029268166050314903, -0.45445019006729126, 0.04934417083859444, -0.17575675249099731, -0.033944062888622284, -0.028623422607779503, -0.16214442253112793, 0.07061192393302917, 0.059837255626916885, 0.03442050889134407, 0.020242400467395782, 0.22017008066177368, 0.11978647112846375, 0.3783564269542694, 0.4785473942756653, -0.09864486753940582, 0.7240330576896667, 0.08315030485391617, -0.00944886077195406, 0.02682553045451641, -0.05517898499965668, -0.17598706483840942, 0.03951163962483406, 0.12947797775268555, 0.4713503122329712, -0.07734821736812592, 0.35965675115585327, -0.2402815967798233, 0.07416460663080215, -0.05003417655825615, -0.5376128554344177, 0.08918233215808868, -0.004219116643071175, -0.04423782601952553, 0.26622432470321655, 0.42578303813934326, 0.11622224748134613, 0.11474061012268066, -0.1251278519630432, -0.4667598009109497, -0.283077210187912, 0.41881096363067627, -0.5340772867202759, -0.30159881711006165, 0.19961605966091156, 0.22690322995185852, -0.0685342326760292, 0.05043494701385498, -0.3018624782562256, -0.06729251146316528, 0.16313715279102325, -0.22798298299312592, -0.015588019043207169, 0.31345659494400024, -0.47143715620040894, 0.0794142559170723, -0.03801818564534187, 0.27980682253837585, 0.20968545973300934, -0.35224270820617676, -0.04154525324702263, -0.1549486219882965 ]
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)
[ 0.09444833546876907, -0.030215412378311157, 0.04769314453005791, -0.021573128178715706, 0.15865421295166016, -0.1987861692905426, 0.3093152940273285, -0.028614167124032974, -0.20045968890190125, 0.1795499324798584, -0.14158888161182404, 0.26486992835998535, -0.21903835237026215, 0.2876112163066864, 0.3653639853000641, 0.08437414467334747, -0.012491506524384022, 0.28119388222694397, -0.014132761396467686, -0.09843260049819946, -0.21194225549697876, 0.10448227822780609, 0.03202148899435997, -0.049356572329998016, -0.11954211443662643, -0.17534472048282623, -0.09639161825180054, 0.2744736075401306, -0.20591822266578674, -0.3228285610675812, 0.4480952322483063, 0.22257773578166962, 0.30274727940559387, 0.5770038962364197, -0.00011726107186404988, -0.042894262820482254, 0.14301028847694397, -0.22344984114170074, -0.6293323636054993, -0.3826368749141693, 0.07990775257349014, -0.1367635726928711, 0.07053135335445404, -0.25299206376075745, 0.1908077597618103, 0.041705552488565445, 0.0007065832614898682, 0.06395929306745529, 0.5112634897232056, 0.38380196690559387, 0.15527521073818207, 0.08335161954164505, -0.19444242119789124, -0.08382099121809006, 0.03142507001757622, 0.18943162262439728, -0.04444864019751549, 0.29738906025886536, 0.36469683051109314, -0.39997923374176025, 0.28799787163734436, -0.19582870602607727, -0.004944684449583292, -0.09528369456529617, 0.4774850308895111, -0.02443007566034794, 0.0022102727089077234, -0.40874117612838745, -0.004732964094728231, 0.1479024440050125, 0.27610254287719727, -0.044267989695072174, -0.34744545817375183, -0.1056399866938591, 0.1462828665971756, -0.4893597662448883, 0.2515539526939392, 0.3014516234397888, -0.42473965883255005, -0.029061710461974144, -0.2817879021167755, 0.32458627223968506, -0.413730263710022, 0.4369974434375763, 0.1153266504406929, 0.3545948266983032, -0.02137403003871441, 0.07461947947740555, 0.23075905442237854, -0.13691921532154083, -0.02224896103143692, 0.07378768175840378, 0.012209306471049786, 0.2807026505470276, -0.18891210854053497, -0.18761013448238373, -0.018634291365742683, -0.28084802627563477, -0.09787079691886902, 0.11185150593519211, 0.4561227560043335, -0.13783898949623108, -0.23979945480823517, 0.15594865381717682, 0.09088326245546341, 0.3234419524669647, -0.006549140904098749, 0.09779694676399231, 0.12052936851978302, 0.0003341523406561464, 0.28368979692459106, -0.060522742569446564, -0.27953991293907166, -0.07793846726417542, -0.14036300778388977, -0.03552103042602539, 0.26556676626205444, 0.0827498808503151, -0.0020337754394859076, -0.08893024921417236, 0.06566966325044632, 0.06152736023068428, -0.08327463269233704, 0.3429430425167084, -0.08083143830299377, -0.2568763792514801, -0.004063084255903959, 0.25254517793655396, 0.1472705900669098, -0.26811468601226807, 0.0705912634730339, 0.3221956193447113, -0.12113511562347412, -0.14960487186908722, 0.4027724862098694, 0.01671142689883709, 0.2068515419960022, -0.05475413799285889, 0.16550303995609283, -0.14268730580806732, 0.06256703287363052, 0.015680594369769096, -0.1528286337852478, 0.03313688561320305, 0.2373732030391693, 0.35915035009384155, -0.07919693738222122, -0.5343109369277954, -0.12540288269519806, 0.0493679940700531, -0.381190687417984, -0.35266658663749695, -0.45391926169395447, 0.09591703116893768, 0.04416889324784279, -0.22488152980804443, -0.32030725479125977, -0.10148510336875916, 0.09989181905984879, -0.2611238956451416, -0.07177607715129852, -0.02674815058708191, -0.11899209767580032, -0.12630967795848846, 0.12081444263458252, 0.03390286862850189, -0.46360495686531067, 0.2092762291431427, -0.046156611293554306, 0.5230318903923035, 0.2966090738773346, 0.28856387734413147, -0.3173559308052063, 0.2671298384666443, -0.06915769726037979, 0.18908894062042236, 0.5893788933753967, -0.48319557309150696, -0.25311359763145447, 0.050312772393226624, -0.03707096353173256, -0.11455785483121872, 0.02929728664457798, 0.16076205670833588, 0.08552628755569458, -0.0990605428814888, 0.07839163392782211, 0.4290335178375244, 0.07837426662445068, 0.045694947242736816, -0.24622957408428192, -0.18085859715938568, 0.22457733750343323, 0.03091498091816902, -0.17134487628936768, 0.07902851700782776, 0.2619319260120392, 0.056161124259233475, 0.028044216334819794, -0.3400057852268219, 0.14545108377933502, 0.1951558142900467, 0.33123913407325745, -0.16132643818855286, -0.14345785975456238, 0.24710281193256378, -0.4086434245109558, 0.18435809016227722, 0.11572956293821335, 0.3616275489330292, -0.14487527310848236, 0.0029635909013450146, -0.42881932854652405, -0.09242076426744461, -0.08533717691898346, -0.03767789155244827, 0.06950780749320984, 0.09063838422298431, 0.06854413449764252, -0.1673053354024887, -0.22225068509578705, 0.11700506508350372, -0.3020473122596741, 0.0560997799038887, -0.18788984417915344, 0.06197630614042282, -0.2840307056903839, -0.11223077774047852, 0.2280946522951126, 0.18706196546554565, 0.20362797379493713, -0.015009716153144836, -0.17897574603557587, 0.09054937958717346, -0.4720057249069214, 0.3271068036556244, 0.04591110721230507, 0.16948163509368896, 0.023193668574094772, -0.02609780803322792, -0.21184630692005157, -0.1186256930232048, 0.18914350867271423, -0.1595863550901413, -0.32982566952705383, 0.1566561609506607, -0.02223840169608593, 0.00816116388887167, 0.1695001721382141, 0.01233644038438797, 0.2830256521701813, -0.19770652055740356, 0.11460219323635101, 0.13609100878238678, 0.4101679027080536, 0.0970584824681282, 0.37629255652427673, -0.03750365972518921, -0.1397162824869156, -0.07505496591329575, 0.1551820933818817, 0.00046271836617961526, 0.1710677146911621, 0.07220102846622467, 0.025278232991695404, -0.09470286965370178, 0.28842294216156006, 0.5951279401779175, 0.3534688949584961, 0.07945828884840012, -0.052482202649116516, -0.04475077986717224, -0.11151335388422012, -0.2275596410036087, -0.016284175217151642, -0.07748278975486755, 0.04854608327150345, 0.21171732246875763, 0.4386913478374481, 0.26055291295051575, -0.13261154294013977, -0.01688237488269806, 0.11479976773262024, 0.45402824878692627, -0.1906822919845581, 0.2663327157497406, -0.09921876341104507, -0.4326513707637787, 0.2592712342739105, -0.2881089746952057, -0.08621691912412643, -0.182784304022789, -0.18338343501091003, 0.12537211179733276, 0.08592549711465836, -0.11140722036361694, 0.1523822396993637, 0.3510524034500122, -0.09527836740016937, -0.2716924548149109, -0.2880840003490448, -0.10307905077934265, -0.20516356825828552, 0.04893943667411804, 0.3304882347583771, -0.11064361780881882, 0.41046014428138733, 0.34332212805747986, 0.06626522541046143, -0.046241432428359985, -0.32678043842315674, -0.003214477328583598, -0.04425817355513573, 0.0937918946146965, 0.09139981120824814, 0.33954888582229614, 0.0049575152806937695, -0.544151246547699, 0.1870240718126297, 0.05857401341199875, -0.24736614525318146, 0.2597731351852417, -0.037839941680431366, -0.31450483202934265, -0.4980628490447998, 0.010082146152853966, -0.2758822739124298, -0.34603869915008545, 0.23943115770816803, 0.33021268248558044, 0.2363397777080536, 0.5939531326293945, 0.01818893663585186, 0.13191565871238708, 0.3163864314556122, 0.13424107432365417, 0.0335441529750824, -0.20578597486019135, 0.48493072390556335, -0.23606938123703003, -0.4347623586654663, 0.10699358582496643, -0.137483611702919, 0.5780161619186401, -0.04567144066095352, -0.39297571778297424, -0.1988610029220581, 0.09783361107110977, 0.2063962072134018, -0.0661422610282898, -0.08298236131668091, 0.2729759216308594, -0.19738003611564636, 0.03931650519371033, -0.06291234493255615, -0.29849177598953247, 0.15413431823253632, 0.6249304413795471, 0.23607304692268372, -0.07148337364196777, 0.4156576991081238, 0.05072728171944618, 0.6994938254356384, 0.08518064767122269, -0.31508761644363403, 0.45266783237457275, -0.03172579035162926, 0.17719560861587524, -0.1375226378440857, -0.34023913741111755, 0.09549224376678467, 0.23747237026691437, -0.040536068379879, 0.15654636919498444, -0.028311604633927345, -0.0831940695643425, -0.45006251335144043, 0.2251790314912796, -0.259337455034256, 0.049751199781894684, 0.014600427821278572, 0.08145259320735931, 0.03887324780225754, -0.09079224616289139, -0.24212270975112915, -0.3565264344215393, -0.5098559260368347, -0.12958136200904846, 0.2505680322647095, -0.23419034481048584, 0.1831371933221817, -0.6524078845977783, -0.10370760411024094, -0.6693298816680908, 0.2962002158164978, -0.14688852429389954, 0.275445818901062, -0.21646235883235931, 0.18662290275096893, 0.13347968459129333, -0.08640039712190628, 0.08926211297512054, -0.41608133912086487, -0.17348666489124298, -0.09920841455459595, 0.29537996649742126, -0.46098875999450684, -0.04646911844611168, -0.03333171457052231, 0.08325352519750595, 0.4845876693725586, 0.21635541319847107, 0.020407797768712044, 0.09795548021793365, -0.17281484603881836, -0.25559094548225403, -0.16451439261436462, -0.22752927243709564, -0.03326975181698799, -0.6557897925376892, -0.34370654821395874, -0.479366660118103, 0.07115891575813293, 0.1532273143529892, 0.1519201546907425, -0.07051830738782883, 0.2251535803079605, 0.1137520894408226, 0.49375736713409424, 0.017786385491490364, -0.0832151547074318, 0.6168341040611267, -0.12393522262573242, -0.46655601263046265, 0.49015718698501587, -0.062034472823143005, 0.3874390423297882, 0.07472100108861923, -0.13771234452724457, -0.15648919343948364, -0.30095174908638, 0.4412570595741272, 0.17833279073238373, 0.10966109484434128, 0.21357393264770508, -0.13211596012115479, 0.0521494522690773, -0.19166399538516998, 0.47335901856422424, 0.3384172022342682, -0.09716370701789856, -0.3771510124206543, -0.13973011076450348, 0.4467111825942993, -0.2157983034849167, -0.007921421900391579, 0.23950035870075226, -0.21200773119926453, -0.28229591250419617, 0.2108963429927826, 0.1703660637140274, 1.1364238262176514, -0.0774025246500969, 0.3717598617076874, 0.13204921782016754, -0.05455600842833519, 0.756166398525238, -0.18963627517223358, 0.024763673543930054, -0.2485024780035019, 0.1745956689119339, -0.10031919181346893, -0.04481510818004608, -0.08898072689771652, 0.11526481807231903, -0.22819103300571442, 0.057403650134801865, -0.04674175754189491, -0.1636371910572052, 0.007388418074697256, 0.541745126247406, 0.09941388666629791, -0.37804171442985535, -0.20321042835712433, 0.10226579010486603, -0.33484187722206116, 0.30063313245773315, -0.16814753413200378, -0.234699547290802, -0.32431986927986145, -0.1621614247560501, -0.1864948868751526, 0.3050607144832611, -0.23749375343322754, 0.08102425932884216, 0.05311482772231102, -0.3184690475463867, 0.25533413887023926, 0.43584567308425903, 0.5033401250839233, -0.10633108019828796, -0.07909468561410904, 0.03086078353226185, -0.1377715766429901, -0.029956288635730743, 0.00021440471755340695, 0.07229562103748322, 0.17591886222362518, -0.31243908405303955, 0.06527630984783173, 0.2966582775115967, -0.014302927069365978, -0.30962201952934265, -0.17401336133480072, 0.14716772735118866, 0.07626001536846161, -0.14346009492874146, -0.31247618794441223, 0.09920883178710938, -0.17771463096141815, -0.02474294975399971, 0.0412108413875103, 0.06617705523967743, 0.03593970835208893, -0.07256119698286057, 0.07392621040344238, -0.18735899031162262, 0.06512045115232468, 0.07385536283254623, -0.11527076363563538, -0.061693914234638214, 0.7477414608001709, -0.06718648970127106, 0.09660468995571136, -0.0992186963558197, -0.23556272685527802, -0.33968257904052734, -0.18870630860328674, -0.10450935363769531, 0.26849493384361267, -0.04515412449836731, 0.06955885142087936, 0.6650161147117615, 0.33289027214050293, 0.141617551445961, 0.16352130472660065, -0.15161098539829254, -0.33256784081459045, -0.3406681418418884, -0.03344973921775818, 0.16924771666526794, -0.08269954472780228, 0.19785884022712708, 0.20175224542617798, 0.3437700569629669, -0.2551347613334656, -0.17512620985507965, -0.21311795711517334, 0.16516347229480743, 0.7187831401824951, -0.054165903478860855, -0.06109914928674698, -0.01915791817009449, 0.003239512210711837, -0.044167228043079376, -0.16698209941387177, -0.13342908024787903, -0.37222734093666077, 0.16226878762245178, 0.47540825605392456, -0.2071847766637802, -0.07411118596792221, -0.20155294239521027, -0.1035783439874649, -0.12503717839717865, -0.3152001202106476, 0.04092053696513176, -0.058333441615104675, 0.21087126433849335, -0.24682369828224182, 0.35262563824653625, -0.08808566629886627, 0.19500517845153809, 0.019292518496513367, 0.18320320546627045, 0.07253221422433853, -0.02448059804737568, 0.16411755979061127, 0.05056824907660484, -0.3545888662338257, -0.031485073268413544, -0.2035149484872818, -0.0913093090057373, 0.04711707681417465, -0.44442635774612427, 0.11381721496582031, -0.16836649179458618, 0.04783881455659866, 0.02164548449218273, -0.5189083218574524, 0.1451278030872345, 0.38480260968208313, 0.12011392414569855, 0.048322197049856186, 0.12300321459770203, 0.1684437096118927, 0.21478116512298584, -0.010159051045775414, 0.23202195763587952, 0.06219123303890228, -0.27599048614501953, -0.05973257124423981, -0.17764456570148468, -0.10463611036539078, -0.26302623748779297, 0.07028678804636002, -0.12691263854503632, -0.40820831060409546, -0.0510936863720417, 0.13153265416622162, 0.17131730914115906, -0.11073257029056549, 0.14628790318965912, -0.2145094871520996, 0.09462959319353104, 0.02808256633579731, -0.08880054205656052, 0.22827261686325073, -0.26857805252075195, 0.07813674956560135, -0.10730600357055664, -0.04195268452167511, 0.3354441225528717, 0.14839155972003937, 0.6253678798675537, -0.3005789816379547, -0.02767838165163994, -0.24202127754688263, 0.14442552626132965, -0.0007230810006149113, 0.17657196521759033, -0.6177632808685303, -0.02887565828859806, 0.05070384964346886, -0.10464967787265778, 0.028326133266091347, -0.34961187839508057, -0.060821931809186935, 0.3098297715187073, -0.1774180680513382, 0.11428427696228027, -0.15285076200962067, 0.21429768204689026, 0.1484036147594452, -0.1513320356607437, 0.4299253523349762, 0.05962701886892319, -0.10391737520694733, -0.0821744054555893, 0.07708905637264252, 0.5147440433502197, 0.582910418510437, 0.11593953520059586, 0.3167453706264496, 0.21503429114818573, 0.028208231553435326, 0.07165063172578812, 0.21679039299488068, 0.43436092138290405, 0.2830864191055298, 0.12656116485595703, 0.047754447907209396, -0.12174839526414871, 0.07318650931119919, 0.05191691219806671, 0.12302333861589432, -0.27411773800849915, 0.3007916212081909, 0.0012677517952397466, -0.3716166615486145, -0.19212459027767181, 0.38253915309906006, -0.4631408154964447, 0.03657785430550575, 0.21874843537807465, 0.2108299732208252, 0.3433953523635864, -0.27728182077407837, 0.038886114954948425, -0.16345837712287903, 0.43623849749565125, 0.49675172567367554, -0.06773657351732254, -0.1250283122062683, -0.28918641805648804, -0.4506283700466156, 0.18131104111671448, -0.1986445188522339, -0.2299528867006302, -0.35685479640960693, 0.19610652327537537, 0.04848916828632355, -0.10085800290107727, 0.05383487418293953, 0.0999092236161232, -0.21706201136112213, -0.016973182559013367, -0.4612075686454773, 0.06106363609433174, -0.15878154337406158, -0.014815456233918667, -0.02872905135154724, -0.149317666888237, 0.06392442435026169, 0.07336704432964325, 0.04292551055550575, 0.02961539663374424, 0.2109256535768509, 0.10708624869585037, 0.3892146646976471, 0.4718272387981415, -0.10158520191907883, 0.7334184646606445, 0.0889010950922966, -0.0027491662185639143, 0.012416502460837364, -0.05101082846522331, -0.1712319701910019, 0.029724404215812683, 0.14156679809093475, 0.4371800422668457, -0.08350188285112381, 0.3541299104690552, -0.2191767394542694, 0.05322765186429024, -0.05195614695549011, -0.5288751125335693, 0.10355673730373383, -0.0058161355555057526, -0.03416019305586815, 0.2721138596534729, 0.4090966582298279, 0.11103543639183044, 0.1126570999622345, -0.1088172197341919, -0.4588083326816559, -0.28008827567100525, 0.3989613950252533, -0.5402223467826843, -0.301598459482193, 0.195817232131958, 0.2340400218963623, -0.060051657259464264, 0.047836579382419586, -0.3058382570743561, -0.05967315658926964, 0.16777817904949188, -0.21494092047214508, -0.009056543931365013, 0.3197662830352783, -0.4719263017177582, 0.07732557505369186, -0.03231992945075035, 0.27389785647392273, 0.20475371181964874, -0.3582354485988617, -0.03778165951371193, -0.16721458733081818 ]
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.
[ 0.10846643894910812, -0.04407990351319313, 0.05040095001459122, -0.02760525606572628, 0.1381322294473648, -0.18117071688175201, 0.3266914486885071, -0.009185025468468666, -0.1950228065252304, 0.18001538515090942, -0.12995536625385284, 0.23764540255069733, -0.2306763380765915, 0.29196545481681824, 0.36800000071525574, 0.0748562216758728, 0.0062184459529817104, 0.2890028953552246, 0.011848773807287216, -0.10124017298221588, -0.21110139787197113, 0.10817131400108337, 0.03599909693002701, -0.06653636693954468, -0.12294402718544006, -0.17405113577842712, -0.10020716488361359, 0.27096471190452576, -0.2011130154132843, -0.335197389125824, 0.4402619004249573, 0.22429488599300385, 0.32153749465942383, 0.5771902203559875, -0.00011768118565669283, -0.05328452214598656, 0.13914421200752258, -0.22553704679012299, -0.6261970400810242, -0.36731255054473877, 0.08496668189764023, -0.1735266000032425, 0.082673080265522, -0.25724655389785767, 0.18322902917861938, 0.025091679766774178, 0.008598617278039455, 0.07702258974313736, 0.5234442353248596, 0.36876630783081055, 0.15495122969150543, 0.09582117944955826, -0.18301618099212646, -0.09871847182512283, 0.02912258170545101, 0.1872798055410385, -0.042479708790779114, 0.2965948283672333, 0.3687248229980469, -0.3961844742298126, 0.30293771624565125, -0.18616315722465515, -0.008991989307105541, -0.08873263746500015, 0.4844440221786499, -0.02203051745891571, 0.006850486155599356, -0.40278059244155884, -0.00808301568031311, 0.15804347395896912, 0.2721642553806305, -0.0382772833108902, -0.34664636850357056, -0.1168702021241188, 0.15635870397090912, -0.4871675968170166, 0.24212117493152618, 0.2832774221897125, -0.421855628490448, -0.04651616886258125, -0.25824156403541565, 0.31141236424446106, -0.4107931852340698, 0.4234355092048645, 0.12531889975070953, 0.3503040671348572, -0.026141168549656868, 0.061015237122774124, 0.25930121541023254, -0.14530344307422638, -0.021243495866656303, 0.09257097542285919, 0.017637861892580986, 0.2819584012031555, -0.19193203747272491, -0.19057057797908783, -0.02157408744096756, -0.2899376451969147, -0.10397443175315857, 0.10910020023584366, 0.45452794432640076, -0.1308465451002121, -0.22314807772636414, 0.15165244042873383, 0.08718287944793701, 0.33124062418937683, 0.00862487219274044, 0.07960393279790878, 0.11092221736907959, 0.012732777744531631, 0.2880062162876129, -0.06930071115493774, -0.27576756477355957, -0.07849981635808945, -0.13718944787979126, -0.0488901361823082, 0.2654901444911957, 0.08655703812837601, 0.0027185112703591585, -0.08611767739057541, 0.0716390386223793, 0.039911285042762756, -0.07255179435014725, 0.33466634154319763, -0.09973805397748947, -0.2767598628997803, 0.008246522396802902, 0.24410685896873474, 0.12901811301708221, -0.2690318822860718, 0.07090899348258972, 0.3014501631259918, -0.12323852628469467, -0.141105517745018, 0.41318583488464355, 0.0026034817565232515, 0.190986767411232, -0.04976864531636238, 0.17416951060295105, -0.1157214492559433, 0.07259535789489746, 0.026877423748373985, -0.15458223223686218, 0.026117313653230667, 0.23558564484119415, 0.37500208616256714, -0.07226445525884628, -0.5426658987998962, -0.12554551661014557, 0.04687368869781494, -0.3762073218822479, -0.3481232821941376, -0.4494229555130005, 0.09366770088672638, 0.03533310443162918, -0.23435980081558228, -0.2903655469417572, -0.0844537764787674, 0.09920254349708557, -0.2737710773944855, -0.0880439504981041, -0.017819279804825783, -0.10808941721916199, -0.1402439922094345, 0.11493650078773499, 0.05398007482290268, -0.48611336946487427, 0.20890983939170837, -0.06266920268535614, 0.5137916207313538, 0.30137237906455994, 0.28101110458374023, -0.3043971359729767, 0.2599911689758301, -0.06236640736460686, 0.177378848195076, 0.5754059553146362, -0.4743398427963257, -0.25073695182800293, 0.034279100596904755, -0.03656109422445297, -0.1030130609869957, 0.05468667298555374, 0.16820448637008667, 0.0883626937866211, -0.10266004502773285, 0.07413917034864426, 0.4095366597175598, 0.07595815509557724, 0.030013155192136765, -0.2245142161846161, -0.1747814118862152, 0.25426870584487915, 0.012387234717607498, -0.15040357410907745, 0.07483933866024017, 0.2523055970668793, 0.08474497497081757, 0.030770523473620415, -0.3352094292640686, 0.1500527262687683, 0.20149943232536316, 0.335964560508728, -0.1352570503950119, -0.13722959160804749, 0.23809409141540527, -0.4048444628715515, 0.1965622752904892, 0.1271623969078064, 0.3596281409263611, -0.11083971709012985, 0.006236785091459751, -0.4200926721096039, -0.11388054490089417, -0.09573759138584137, -0.031772464513778687, 0.06666679680347443, 0.09637950360774994, 0.07029217481613159, -0.18559035658836365, -0.23020361363887787, 0.1288202852010727, -0.31538844108581543, 0.061058130115270615, -0.19712676107883453, 0.059170518070459366, -0.2863607406616211, -0.1116008386015892, 0.21818304061889648, 0.1753208190202713, 0.20732100307941437, -0.018019016832113266, -0.17255373299121857, 0.08949816972017288, -0.46536684036254883, 0.30873721837997437, 0.012427200563251972, 0.14155274629592896, 0.0335690900683403, -0.02158377878367901, -0.19962099194526672, -0.11788225173950195, 0.17727236449718475, -0.17059703171253204, -0.34310758113861084, 0.14997325837612152, -0.03314928337931633, 0.019478892907500267, 0.1668234020471573, -0.0024823786225169897, 0.2705676257610321, -0.19244840741157532, 0.1081031933426857, 0.11901303380727768, 0.3931276798248291, 0.1005677580833435, 0.37359169125556946, -0.062350206077098846, -0.12617242336273193, -0.0803188681602478, 0.17376309633255005, 0.014593850821256638, 0.15850791335105896, 0.06848897784948349, 0.03325919434428215, -0.07810261100530624, 0.2860078811645508, 0.6209805011749268, 0.3515610992908478, 0.09832153469324112, -0.048403818160295486, -0.05061190575361252, -0.11802094429731369, -0.222604900598526, -0.03233380615711212, -0.0865330919623375, 0.04504910111427307, 0.21368959546089172, 0.429272323846817, 0.2628836929798126, -0.1386045217514038, 0.002830345882102847, 0.11219191551208496, 0.441070020198822, -0.1806759387254715, 0.2472010999917984, -0.11076122522354126, -0.43094226717948914, 0.23817764222621918, -0.2824089527130127, -0.0893058255314827, -0.16441912949085236, -0.19531375169754028, 0.11842571198940277, 0.11353220045566559, -0.1269589513540268, 0.16630007326602936, 0.3391781747341156, -0.11040569841861725, -0.25449422001838684, -0.27940356731414795, -0.08828495442867279, -0.20852887630462646, 0.05295974016189575, 0.328294038772583, -0.11787298321723938, 0.42696431279182434, 0.3558059632778168, 0.06698241084814072, -0.04825431853532791, -0.34660619497299194, -0.012719879858195782, -0.06370160728693008, 0.07861372083425522, 0.08492806553840637, 0.34389033913612366, -0.0008335369057022035, -0.5517427325248718, 0.19058901071548462, 0.06303216516971588, -0.25250303745269775, 0.2460183948278427, -0.021942097693681717, -0.30630171298980713, -0.4941288232803345, -0.007458861917257309, -0.2783515453338623, -0.3438791036605835, 0.2504080533981323, 0.35123634338378906, 0.23518390953540802, 0.6135788559913635, 0.020829353481531143, 0.13948366045951843, 0.3170112371444702, 0.15296363830566406, 0.0331699401140213, -0.18742358684539795, 0.46579796075820923, -0.25305989384651184, -0.43299952149391174, 0.11046848446130753, -0.1329052597284317, 0.5748026967048645, -0.04160173982381821, -0.3659118413925171, -0.19975872337818146, 0.0914410725235939, 0.18816456198692322, -0.07725886255502701, -0.06613487005233765, 0.2856582999229431, -0.1821654587984085, 0.04210813716053963, -0.05911853909492493, -0.3117048144340515, 0.15016886591911316, 0.6340500116348267, 0.25497764348983765, -0.07658514380455017, 0.42161303758621216, 0.042422667145729065, 0.7029532194137573, 0.08945327252149582, -0.3097897469997406, 0.4549664855003357, -0.028219344094395638, 0.19479669630527496, -0.13296854496002197, -0.3579268753528595, 0.0959678515791893, 0.2346646934747696, -0.030786167830228806, 0.13775286078453064, -0.02721979469060898, -0.06776405870914459, -0.44664308428764343, 0.24750658869743347, -0.25212863087654114, 0.060791194438934326, -0.001671534962952137, 0.10914742201566696, 0.04245892912149429, -0.10231342166662216, -0.23655302822589874, -0.365654319524765, -0.5041986703872681, -0.12927499413490295, 0.24986600875854492, -0.25403234362602234, 0.1969151645898819, -0.6588920950889587, -0.12374038249254227, -0.6575385928153992, 0.2880604565143585, -0.14811073243618011, 0.25945526361465454, -0.2137509137392044, 0.17713764309883118, 0.13104258477687836, -0.08934591710567474, 0.11717302352190018, -0.4182169735431671, -0.18678325414657593, -0.0956520140171051, 0.2866273522377014, -0.47739705443382263, -0.05347135663032532, -0.043761592358350754, 0.08510797470808029, 0.4654034376144409, 0.2201540768146515, 0.0312618650496006, 0.0952751636505127, -0.15829230844974518, -0.2440684735774994, -0.1730845719575882, -0.2345629185438156, -0.04078040271997452, -0.6392326354980469, -0.33754411339759827, -0.48867669701576233, 0.049477726221084595, 0.15704037249088287, 0.1608155071735382, -0.07531017810106277, 0.2124437391757965, 0.11879844963550568, 0.49777543544769287, 0.022433524951338768, -0.0697203129529953, 0.6355296969413757, -0.12254391610622406, -0.44986554980278015, 0.48968538641929626, -0.053558703511953354, 0.37191230058670044, 0.05060388147830963, -0.1260620504617691, -0.18033045530319214, -0.2989444136619568, 0.46419766545295715, 0.18745729327201843, 0.11531223356723785, 0.21370379626750946, -0.1160859763622284, 0.04945126920938492, -0.18874643743038177, 0.4735020399093628, 0.34728625416755676, -0.09278995543718338, -0.3748313784599304, -0.16618874669075012, 0.44636300206184387, -0.18877370655536652, -0.0016305196331813931, 0.2298572063446045, -0.20053809881210327, -0.28849029541015625, 0.19627487659454346, 0.1790328472852707, 1.1350749731063843, -0.09616311639547348, 0.37867793440818787, 0.13602036237716675, -0.052425384521484375, 0.7858076691627502, -0.18570806086063385, 0.03413787856698036, -0.2420801967382431, 0.1655089259147644, -0.10266604274511337, -0.037637755274772644, -0.08714553713798523, 0.10502206534147263, -0.239122673869133, 0.06721816956996918, -0.05727590247988701, -0.1655508577823639, -0.00603439612314105, 0.5472075939178467, 0.09571060538291931, -0.3948436677455902, -0.19912925362586975, 0.1065409779548645, -0.33136844635009766, 0.29732823371887207, -0.18849727511405945, -0.2347911149263382, -0.2989427149295807, -0.15512244403362274, -0.17125003039836884, 0.30104008316993713, -0.2388865351676941, 0.08907832950353622, 0.05478125065565109, -0.3258427083492279, 0.24529245495796204, 0.42029955983161926, 0.5057652592658997, -0.0947704166173935, -0.07163017988204956, 0.015904534608125687, -0.12948505580425262, -0.058059316128492355, -0.016994044184684753, 0.07856115698814392, 0.2045135796070099, -0.3050251007080078, 0.06459306925535202, 0.28322598338127136, -0.004202884156256914, -0.2958986461162567, -0.19273832440376282, 0.1700761318206787, 0.10598578304052353, -0.14506222307682037, -0.32288122177124023, 0.10752256214618683, -0.19857680797576904, -0.02392820455133915, 0.043140217661857605, 0.05214371532201767, 0.02309333346784115, -0.06285279244184494, 0.07580920308828354, -0.18461143970489502, 0.0675206109881401, 0.09131079167127609, -0.08641175180673599, -0.0853634625673294, 0.7531558871269226, -0.06545611470937729, 0.10210257768630981, -0.09430177509784698, -0.23572029173374176, -0.3444412350654602, -0.16373024880886078, -0.10431704670190811, 0.288705050945282, -0.04541219770908356, 0.10916142165660858, 0.6310181617736816, 0.34796616435050964, 0.11863771080970764, 0.14224322140216827, -0.13004626333713531, -0.33897316455841064, -0.3504674732685089, -0.02279350534081459, 0.17645171284675598, -0.09358363598585129, 0.19203659892082214, 0.2066742330789566, 0.33863088488578796, -0.25417277216911316, -0.186846524477005, -0.18249276280403137, 0.16299647092819214, 0.7249686121940613, -0.0466189943253994, -0.06739460676908493, -0.020376473665237427, 0.0019187311409041286, -0.040756162256002426, -0.1685420572757721, -0.13428908586502075, -0.3807845115661621, 0.16128721833229065, 0.479666143655777, -0.2025005966424942, -0.0754031389951706, -0.20506322383880615, -0.09480379521846771, -0.11984402686357498, -0.30565744638442993, 0.04392362758517265, -0.048083823174238205, 0.22156156599521637, -0.23215943574905396, 0.33715251088142395, -0.05593423917889595, 0.18995916843414307, 0.030670499429106712, 0.17264918982982635, 0.07241874933242798, -0.03402110934257507, 0.1680602729320526, 0.057488687336444855, -0.3541560173034668, -0.03275204822421074, -0.1912260800600052, -0.10384628921747208, 0.029015010222792625, -0.44629907608032227, 0.10618835687637329, -0.17603521049022675, 0.04979933053255081, 0.022434255108237267, -0.48886486887931824, 0.13809172809123993, 0.38951030373573303, 0.11818034946918488, 0.03648423030972481, 0.11728765070438385, 0.16463901102542877, 0.21876654028892517, -0.018171390518546104, 0.22868035733699799, 0.042565248906612396, -0.26980361342430115, -0.06452685594558716, -0.17235441505908966, -0.12308423221111298, -0.2726885676383972, 0.08677515387535095, -0.11697171628475189, -0.39995527267456055, -0.06926877051591873, 0.15473812818527222, 0.1715613603591919, -0.09740512818098068, 0.14381466805934906, -0.20515547692775726, 0.10718613117933273, 0.033884890377521515, -0.08608301728963852, 0.22095489501953125, -0.2704428434371948, 0.0801362544298172, -0.08750227093696594, -0.02710726298391819, 0.34884393215179443, 0.14098073542118073, 0.6430830955505371, -0.3007907271385193, -0.03844214603304863, -0.23728397488594055, 0.1241435557603836, -0.0008916667429730296, 0.1930057555437088, -0.6118701100349426, -0.016115650534629822, 0.04017585515975952, -0.10860519856214523, 0.039865732192993164, -0.3541427552700043, -0.06759516149759293, 0.30947351455688477, -0.17172746360301971, 0.113868348300457, -0.16982096433639526, 0.20617768168449402, 0.14696456491947174, -0.14845171570777893, 0.3970652222633362, 0.05424059182405472, -0.10305633395910263, -0.07449612766504288, 0.08739224076271057, 0.5002298951148987, 0.6005567312240601, 0.11638975888490677, 0.32311058044433594, 0.20161010324954987, 0.038133371621370316, 0.07256476581096649, 0.20181408524513245, 0.42836955189704895, 0.27456167340278625, 0.10998047888278961, 0.047959405928850174, -0.1261918693780899, 0.0636170282959938, 0.047952551394701004, 0.09902054071426392, -0.29639679193496704, 0.29422900080680847, 0.001472344622015953, -0.3734475374221802, -0.19348116219043732, 0.38225844502449036, -0.488055944442749, 0.028025440871715546, 0.2486959546804428, 0.21553653478622437, 0.3356567323207855, -0.2708441913127899, 0.036630209535360336, -0.16707248985767365, 0.43749475479125977, 0.48262274265289307, -0.09399567544460297, -0.12574002146720886, -0.28346550464630127, -0.46568357944488525, 0.1705354005098343, -0.19895006716251373, -0.22324535250663757, -0.357099324464798, 0.20041121542453766, 0.0428692027926445, -0.12530548870563507, 0.07186221331357956, 0.09309978038072586, -0.2287142425775528, -0.058963678777217865, -0.472491979598999, 0.05648256465792656, -0.13548049330711365, -0.021281547844409943, -0.025280863046646118, -0.1377786099910736, 0.05076105520129204, 0.05550024285912514, 0.034974683076143265, 0.025941459462046623, 0.21243998408317566, 0.10449182242155075, 0.39724060893058777, 0.48885849118232727, -0.11699523776769638, 0.7418339848518372, 0.0934423953294754, -0.010376248508691788, 0.011939283460378647, -0.03746569901704788, -0.18987970054149628, 0.053903061896562576, 0.1236984133720398, 0.45055851340293884, -0.09263254702091217, 0.3479200303554535, -0.2200065702199936, 0.07006774097681046, -0.0405460000038147, -0.5381631851196289, 0.1111881285905838, -0.0031204470433294773, -0.031219761818647385, 0.2671128213405609, 0.4321093261241913, 0.12411201000213623, 0.11082860082387924, -0.10501456260681152, -0.4677984118461609, -0.27645838260650635, 0.4062139391899109, -0.533333420753479, -0.29643020033836365, 0.18773046135902405, 0.23693424463272095, -0.08291322737932205, 0.03613558039069176, -0.2994180917739868, -0.07017698138952255, 0.16325388848781586, -0.2294122278690338, -0.012820491567254066, 0.3217461407184601, -0.4839877784252167, 0.06313871592283249, -0.036177296191453934, 0.24127386510372162, 0.22410356998443604, -0.3566041588783264, -0.03303771838545799, -0.15160386264324188 ]
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.
[ 0.043388731777668, -0.19811096787452698, 0.0301857590675354, 0.21069127321243286, 0.10683344304561615, 0.08471887558698654, 0.5492756962776184, 0.19509580731391907, -0.01440933346748352, 0.152593731880188, 0.0023247934877872467, 0.4517519176006317, -0.27085113525390625, 0.325571745634079, 0.01907769963145256, -0.4061383903026581, -0.06452727317810059, 0.06130627542734146, 0.23302091658115387, 0.10228574275970459, 0.09767560660839081, 0.22732260823249817, -0.0382935032248497, 0.28144070506095886, -0.23800930380821228, -0.2281854748725891, -0.20587804913520813, -0.16539844870567322, 0.018803391605615616, -0.40434175729751587, 0.17547138035297394, 0.1728924959897995, -0.24170564115047455, 0.03427806869149208, -0.00011900685785803944, -0.1262153834104538, 0.21351443231105804, -0.009097998961806297, -0.1567247062921524, -0.17348851263523102, -0.5583308935165405, -0.49402689933776855, 0.441752165555954, -0.5981582999229431, -0.004088363144546747, -0.0666581466794014, -0.009363152086734772, 0.25334542989730835, 0.5818641185760498, -0.16109620034694672, 0.0953407883644104, -0.3728165328502655, -0.006457353010773659, 0.12188394367694855, 0.3031228482723236, 0.17062830924987793, -0.3206124007701874, -0.3328781723976135, -0.19216199219226837, 0.016420088708400726, -0.01434339303523302, 0.2186739146709442, 0.06726592034101486, -0.14753025770187378, 0.12505769729614258, 0.20889084041118622, -0.18132509291172028, -0.16720500588417053, -0.06251542270183563, 0.3708767890930176, -0.1658439189195633, -0.1862097531557083, -0.21817907691001892, -0.7446468472480774, 0.19280840456485748, -0.39910268783569336, 0.009476689621806145, 0.2820798456668854, -0.023257844150066376, 0.05690540373325348, 0.22809426486492157, -0.3026105463504791, -0.1733475923538208, 0.07937268912792206, 0.15809087455272675, 0.6143727898597717, 0.08293662965297699, 0.2064344882965088, 0.19808411598205566, 0.019645292311906815, 0.1997452676296234, 0.13025794923305511, -0.04197738319635391, 0.1836853325366974, -0.2608937919139862, -0.34557631611824036, -0.39284199476242065, -0.0823005810379982, -0.21530190110206604, 0.14773224294185638, 0.20480531454086304, -0.03726094216108322, 0.08624105155467987, 0.1348693072795868, 0.24606911838054657, 0.07950561493635178, 0.4300127327442169, 0.6358775496482849, 0.11269090324640274, -0.019237229600548744, -0.36080384254455566, -0.024251196533441544, -0.06769761443138123, 0.061923637986183167, -0.1763531118631363, 0.10095930099487305, -0.09962878376245499, 0.23635706305503845, -0.058111101388931274, -0.2936745584011078, -0.11544916033744812, -0.2922123074531555, 0.27775999903678894, -0.007407160475850105, 0.13234098255634308, 0.05559556186199188, -0.325080543756485, 0.13766522705554962, -0.1903276890516281, -0.40073299407958984, -0.10605233162641525, 0.12324108183383942, -0.3039294183254242, 0.285286009311676, -0.025047704577445984, 0.08957415819168091, 0.13974949717521667, -0.0239227507263422, -0.17118661105632782, -0.2618398368358612, 0.19998787343502045, -0.030773475766181946, 0.2510843575000763, 0.07993129640817642, 0.016892028972506523, 0.057535476982593536, -0.09818579256534576, -0.11058653146028519, -0.3132472634315491, -0.004342345055192709, -0.15221185982227325, -0.3459361791610718, 0.1437164545059204, 0.05959418788552284, 0.14884167909622192, 0.06756260991096497, 0.285625159740448, 0.621551513671875, 0.347048819065094, -0.11526218801736832, 0.1364315301179886, 0.0587196946144104, -0.22138525545597076, -0.24143952131271362, -0.020820066332817078, -0.05500065162777901, 0.04045460745692253, -0.4513779282569885, -0.2863803207874298, 0.10820052772760391, -0.10519055277109146, 0.09117943793535233, -0.22937317192554474, 0.11082300543785095, -0.02668602578341961, 0.6001889705657959, 0.7154176831245422, -0.2713157832622528, 0.0781719833612442, 0.3304651379585266, 0.0726044550538063, -0.34571951627731323, 0.4756770730018616, 0.08520403504371643, 0.13212943077087402, -0.05762225762009621, -0.39683762192726135, 0.18206678330898285, -0.38011759519577026, -0.17116214334964752, 0.20397405326366425, -0.5321475863456726, 0.1683947592973709, -0.09095202386379242, -0.0555260069668293, -0.3709448277950287, 0.020022103562951088, 0.2963101267814636, 0.4655570685863495, 0.034468431025743484, 0.23717822134494781, -0.17417491972446442, -0.07772985845804214, 0.28939035534858704, -0.17316140234470367, -0.36931225657463074, -0.2701145112514496, -0.010613551363348961, -0.173282653093338, 0.05678068473935127, 0.013177584856748581, -0.4073963165283203, 0.04309277608990669, -0.2719109356403351, -0.16464747488498688, -0.2850072681903839, 0.04664390906691551, -0.2125970721244812, -0.05866733938455582, -0.34646639227867126, -0.3343034088611603, 0.1712373048067093, -0.0637914389371872, 0.28379347920417786, -0.3155245780944824, -0.1585671603679657, 0.19284968078136444, 0.008334683254361153, -0.2399851530790329, 0.5741866827011108, 0.25061434507369995, -0.02533877268433571, 0.0648970752954483, 0.3517276346683502, 0.3894834518432617, -0.07242347300052643, 0.023375287652015686, -0.12618158757686615, 0.2539404034614563, -0.06732700765132904, -0.08393126726150513, -0.10601682215929031, -0.26386842131614685, -0.14377817511558533, -0.10177093744277954, 0.5019599795341492, 0.12187360972166061, 0.5909391641616821, 0.290153443813324, 0.058206330984830856, -0.06445088982582092, -0.25854021310806274, -0.19194462895393372, -0.4637325406074524, -0.2635042369365692, -0.3758361339569092, 0.08257997781038284, -0.04383235052227974, -0.2890101373195648, 0.24587547779083252, 0.44757288694381714, 0.14510799944400787, 0.016082242131233215, -0.23104320466518402, 0.11973905563354492, 0.17330706119537354, -0.2040422260761261, 0.7559658885002136, 0.19795012474060059, 0.16951479017734528, -0.09351564943790436, 0.05726437270641327, -0.042332831770181656, -0.18957479298114777, 0.2760685980319977, 0.28203046321868896, 0.07955443859100342, -0.027139578014612198, -0.169819638133049, -0.1431916058063507, 0.10303764790296555, -0.03333808854222298, 0.22930601239204407, -0.32859787344932556, -0.08614982664585114, -0.07723218947649002, -0.12500980496406555, -0.3374965488910675, -0.6012638807296753, -0.03178204596042633, -0.23478975892066956, -0.2417268604040146, 0.15794694423675537, -0.08859363198280334, -0.21088051795959473, 0.1417497843503952, -0.1397167593240738, 0.1332399845123291, -0.3738994300365448, 0.039972010999917984, 0.14731791615486145, -0.3808600902557373, -0.06988397240638733, 0.02591218613088131, 0.05599715933203697, 0.03493776172399521, 0.26271504163742065, 0.3256540894508362, -0.28088319301605225, 0.1503276228904724, -0.3359714448451996, 0.17685167491436005, -0.3270224928855896, -0.28942131996154785, 0.22199620306491852, -0.3101724088191986, -0.004219602327793837, -0.46895039081573486, 0.3010506331920624, -0.10614213347434998, -0.10014096647500992, -0.31769147515296936, 0.3218804895877838, 0.21165883541107178, -0.1271352767944336, -0.31906259059906006, -0.4289039075374603, -0.14181458950042725, 0.3455911874771118, -0.07541994750499725, 0.296516090631485, 0.09804391860961914, -0.42088595032691956, -0.029327617958188057, -0.0928119346499443, -0.11690017580986023, 0.04345478117465973, 0.14458249509334564, -0.08969900757074356, -0.4102131724357605, 0.1566271185874939, 0.052334014326334, -0.13682547211647034, 0.33396783471107483, -0.1761157512664795, -0.011461429297924042, 0.18883349001407623, -0.013530982658267021, -0.03493420407176018, 0.035726021975278854, 0.11508940905332565, 0.1559375375509262, 0.27365314960479736, 0.06651606410741806, -0.07385818660259247, -0.232402503490448, 0.434865802526474, 0.15245786309242249, 0.15153145790100098, -0.09006468951702118, 0.024158257991075516, -0.07993445545434952, 0.6646890044212341, 0.2682935297489166, 0.19962073862552643, -0.17779268324375153, -0.004739018622785807, -0.3417838513851166, 0.0782979279756546, -0.2114560902118683, 0.16598933935165405, -0.003022185992449522, -0.2993084788322449, 0.341092586517334, 0.03000732511281967, -0.061492856591939926, -0.2507544159889221, -0.06884315609931946, -0.03625015169382095, -0.5644198060035706, 0.06861457973718643, -0.18404856324195862, 0.21568171679973602, -0.19991739094257355, 0.05245448276400566, 0.0038426266983151436, 0.06245717778801918, -0.14635205268859863, -0.15383020043373108, -0.14873459935188293, -0.07147340476512909, -0.11421886831521988, -0.6203206777572632, -0.05381577089428902, -0.04224342107772827, 0.2887144982814789, -0.14281012117862701, -0.021646922454237938, -0.2228592336177826, 0.29551804065704346, 0.3017685115337372, 0.28602856397628784, -0.09012580662965775, -0.16662846505641937, 0.09333706647157669, 0.042149122804403305, 0.19225482642650604, -0.19830775260925293, 0.09100940078496933, 0.029229668900370598, 0.07483057677745819, 0.039358850568532944, -0.572121798992157, -0.29482296109199524, 0.5887435674667358, -0.2423548698425293, -0.33407774567604065, 0.03839486464858055, 0.0384373813867569, 0.13238516449928284, 0.07342905551195145, -0.04046778753399849, 0.10366377979516983, 0.12703461945056915, 0.13721337914466858, 0.08338861167430878, -0.03285423293709755, -0.23436816036701202, 0.08977613598108292, 0.13432160019874573, 0.021047048270702362, -0.05372081696987152, 0.08779164403676987, 0.2963988780975342, 0.2657298743724823, -0.3124952018260956, 0.4403763711452484, -0.4000658690929413, -0.09892062842845917, 0.009537799283862114, -0.21707412600517273, 0.5385265350341797, 0.33155620098114014, 0.1370117962360382, 0.44719985127449036, 0.005818965379148722, 0.08080262690782547, -0.20520934462547302, -0.06847020983695984, 0.2784639596939087, -0.17185603082180023, -0.4090304672718048, -0.7786412239074707, 0.3944488763809204, 0.0955309048295021, -0.1922817975282669, 0.22138717770576477, 0.13551382720470428, -0.40848293900489807, 0.5755680799484253, 0.01621149480342865, 0.8848461508750916, -0.07890180498361588, 0.21045677363872528, -0.052402425557374954, -0.2898177206516266, 0.6745566725730896, -0.09591887146234512, 0.13303451240062714, -0.36648881435394287, 0.29457151889801025, 0.05793565511703491, 0.10396681725978851, 0.23142316937446594, 0.48991286754608154, 0.006772751919925213, 0.23215791583061218, 0.08442983776330948, 0.26291361451148987, -0.33359795808792114, 0.33951878547668457, 0.14325372874736786, -0.11598335951566696, -0.08670647442340851, -0.048195451498031616, 0.001482380903325975, -0.24703125655651093, 0.2088811695575714, -0.19910140335559845, 0.0166256632655859, 0.17032137513160706, -0.2553980350494385, -0.00695032486692071, -0.12027207762002945, 0.21192005276679993, -0.12011606246232986, -0.4156661331653595, 0.3914091885089874, 0.17767205834388733, 0.23396164178848267, -0.16502228379249573, 0.01922636851668358, 0.4578540325164795, -0.04234327748417854, 0.09356636554002762, -0.17171676456928253, -0.012513539753854275, 0.30116787552833557, 0.05344660207629204, 0.027400396764278412, 0.031027765944600105, 0.018673833459615707, -0.0883808359503746, -0.2242787927389145, 0.2394072562456131, 0.4494664669036865, -0.5152653455734253, 0.2077825963497162, 0.031566351652145386, 0.0019562398083508015, -0.17893286049365997, 0.021222088485956192, -0.21939080953598022, -0.2493937909603119, 0.04010842740535736, -0.17997591197490692, -0.06843478232622147, 0.004961735103279352, -0.31928518414497375, 0.1781570464372635, 0.007905185222625732, 0.3882559835910797, -0.18148979544639587, 0.1604386419057846, -0.23996275663375854, 0.14632436633110046, 0.15137863159179688, -0.20615078508853912, 0.29052796959877014, 0.15306957066059113, -0.26518434286117554, 0.1352846324443817, -0.019382547587156296, 0.05809516832232475, 0.6687646508216858, -0.18001681566238403, 0.03185374289751053, -0.22082458436489105, -0.040196340531110764, -0.10357305407524109, 0.09732278436422348, 0.006829277612268925, 0.4960907995700836, -0.12097840756177902, 0.4065942168235779, -0.16641011834144592, -0.09653307497501373, 0.20132721960544586, 0.2539372742176056, -0.10842788964509964, 0.017032118514180183, 0.15272299945354462, 0.12926197052001953, 0.020380819216370583, -0.05497124046087265, 0.1774238497018814, -0.08827252686023712, -0.3262884020805359, 0.11676819622516632, 0.09450925141572952, 0.20274850726127625, -0.36953073740005493, -0.14116652309894562, 0.10343471169471741, -0.15560413897037506, 0.20738111436367035, 0.06280762702226639, 0.26048651337623596, 0.5012199878692627, 0.27072256803512573, 0.010650277137756348, -0.37880638241767883, 0.2169893980026245, -0.12480488419532776, 0.3370318114757538, 0.2525511085987091, 0.3109063506126404, -0.2106352001428604, 0.07289381325244904, 0.1331859976053238, -0.055775146931409836, 0.18610961735248566, -0.27954745292663574, -0.07307411730289459, 0.03606586530804634, -0.09714672714471817, -0.14056962728500366, 0.21550993621349335, 0.4737994074821472, 0.13721732795238495, -0.03496869280934334, 0.2236471325159073, 0.11047755926847458, -0.07229319214820862, 0.21312913298606873, 0.2800593972206116, -0.17977790534496307, 0.09948098659515381, 0.553147554397583, 0.11321590095758438, 0.2622584104537964, -0.4264288544654846, -0.0630769357085228, -0.10696731507778168, -0.03063422068953514, 0.5049611330032349, 0.8284040093421936, 0.21459436416625977, -0.06918115168809891, 0.46372485160827637, -0.18284419178962708, 0.2680799961090088, 0.45255979895591736, -0.181410551071167, 0.3613758683204651, 0.3433874547481537, -0.17254194617271423, 0.5454280972480774, 0.10762101411819458, -0.36955830454826355, 0.04664408788084984, -0.05818149819970131, -0.24876540899276733, -0.12223450839519501, 0.7737440466880798, 0.04387807101011276, -0.3321119546890259, 0.1711917221546173, 0.3155764043331146, 0.16809482872486115, -0.25017011165618896, -0.13615307211875916, 0.11348187923431396, -0.3541952073574066, 0.0006029580836184323, -0.05430023744702339, -0.22777578234672546, 0.35432401299476624, 0.08226720243692398, 0.3009105920791626, -0.09902658313512802, -0.06769805401563644, 0.21079160273075104, 0.4752943515777588, -0.35172614455223083, -0.1351248174905777, -0.017432937398552895, -0.29152214527130127, 0.15518023073673248, -0.07194460928440094, 0.008162492886185646, -0.134679913520813, -0.25497838854789734, -0.28856679797172546, 0.24606864154338837, -0.10548330098390579, 0.023284640163183212, -0.06977833062410355, 0.1218489557504654, 0.29119688272476196, 0.21185964345932007, -0.023083306849002838, -0.0354093462228775, -0.207214817404747, 0.2606797218322754, -0.018992913886904716, -0.05991332232952118, 0.5498867034912109, 0.23981556296348572, -0.187866672873497, -0.02599560283124447, 0.22880135476589203, -0.29212072491645813, -0.017803817987442017, 0.3589160740375519, -0.06609856337308884, 0.08894050121307373, -0.16790629923343658, -0.02272200398147106, -0.021172400563955307, 0.4855484068393707, -0.22243963181972504, -0.20367223024368286, -0.2754066288471222, -0.3702971339225769, -0.02241438254714012, -0.14031367003917694, -0.12521402537822723, -0.2741914391517639, -0.13138222694396973, 0.16895128786563873, -0.1570616364479065, -0.1679462194442749, 0.16192837059497833, 0.04043963924050331, 0.4329414367675781, 0.33852580189704895, -0.23166555166244507, 0.00576712004840374, -0.021587396040558815, 0.14577648043632507, 0.17384906113147736, 0.13576945662498474, -0.07442142069339752, 0.36563408374786377, -0.29715028405189514, 0.15661512315273285, -0.022885791957378387, 0.584936261177063, 0.1345309466123581, 0.0016774439718574286, -0.020736059173941612, 0.22753006219863892, 0.0298627857118845, 0.03711659461259842, 0.040696799755096436, 0.012698722071945667, -0.5554834604263306, 0.41142719984054565, 0.06802881509065628, 0.16265548765659332, -0.27482011914253235, 0.41206008195877075, -0.14148399233818054, 0.0037245426792651415, -0.23200520873069763, 0.10678321868181229, -0.013923829421401024, 0.19690777361392975, -0.02301151677966118, 0.2964518368244171, 0.37596169114112854, 0.2861540615558624, -0.11746016144752502, -0.06586798280477524, -0.13472895324230194, -0.37831565737724304, 0.1620430052280426, -0.036627378314733505, -0.09991870075464249, 0.08973991870880127, 0.33894968032836914, 0.19934053719043732, 0.11615035682916641, -0.7652567625045776, -0.2747171223163605, 0.1887979656457901, -0.24032483994960785, -0.024294720962643623, 0.14281313121318817, 0.28296005725860596, -0.16623690724372864, 0.01470609474927187, -0.13825108110904694, 0.2797047197818756, -0.06312903761863708, 0.030536239966750145, -0.143540620803833 ]
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?
[ 0.1045849546790123, -0.09824608266353607, 0.02759917639195919, 0.22247891128063202, 0.11074350774288177, 0.11211322993040085, 0.5969952940940857, 0.2325303852558136, -0.0358220711350441, 0.1119878888130188, 0.01879110187292099, 0.4395623207092285, -0.21577899158000946, 0.2508064806461334, -0.04807358980178833, -0.37081483006477356, -0.031474050134420395, 0.09958310425281525, 0.30093175172805786, 0.110613614320755, 0.07535490393638611, 0.24139268696308136, -0.04599487781524658, 0.27223172783851624, -0.21119630336761475, -0.1831187605857849, -0.23876817524433136, -0.1546238362789154, 0.028416195884346962, -0.4011644124984741, 0.19092689454555511, 0.1710233986377716, -0.297771155834198, 0.1150345429778099, -0.0001220013145939447, -0.14937621355056763, 0.2697802186012268, -0.008835426531732082, -0.17719940841197968, -0.2315707951784134, -0.5108581185340881, -0.47503459453582764, 0.4412979185581207, -0.6071305871009827, 0.006469717714935541, -0.16628816723823547, 0.00986267626285553, 0.23457123339176178, 0.5989478230476379, -0.14206351339817047, 0.07940222322940826, -0.39332544803619385, 0.004251692909747362, 0.059003546833992004, 0.2985042929649353, 0.2553159296512604, -0.3581681251525879, -0.3896654546260834, -0.13251174986362457, 0.019385745748877525, 0.013584992848336697, 0.19017548859119415, 0.048040300607681274, -0.15089279413223267, 0.05642225965857506, 0.1882016807794571, -0.2405909150838852, -0.10517501085996628, -0.031474869698286057, 0.3205353617668152, -0.14700567722320557, -0.1331748068332672, -0.22489891946315765, -0.7789201140403748, 0.12296402454376221, -0.44449713826179504, 0.0231789443641901, 0.30839911103248596, -0.08555471152067184, 0.0007130505400709808, 0.1598525047302246, -0.2577974796295166, -0.15868933498859406, 0.07997288554906845, 0.18729670345783234, 0.5755346417427063, 0.04268695041537285, 0.21303345263004303, 0.13065455853939056, 0.05146569013595581, 0.2459283173084259, 0.09748344123363495, 0.037074122577905655, 0.14819668233394623, -0.24022139608860016, -0.34626346826553345, -0.36292821168899536, -0.061227139085531235, -0.2478044182062149, 0.0727619156241417, 0.19479505717754364, -0.09304304420948029, 0.037920162081718445, 0.1094241663813591, 0.33250850439071655, 0.030752187594771385, 0.36637818813323975, 0.5928638577461243, 0.0970461517572403, 0.07390064746141434, -0.3302590847015381, -0.018299387767910957, -0.055822934955358505, 0.10930260270833969, -0.23530124127864838, 0.0842663049697876, -0.10003910213708878, 0.16355472803115845, -0.04588300362229347, -0.25977036356925964, -0.09913959354162216, -0.24285975098609924, 0.22444313764572144, 0.03780582547187805, 0.11807222664356232, 0.03906608372926712, -0.31292423605918884, 0.2033405303955078, -0.12280922383069992, -0.4388541579246521, -0.08893854916095734, 0.09969548135995865, -0.30024728178977966, 0.22964978218078613, 0.10692659765481949, 0.14246712625026703, 0.18240661919116974, 0.008428511209785938, -0.14765846729278564, -0.24537380039691925, 0.15634939074516296, 0.006643952801823616, 0.23175528645515442, 0.04239467903971672, 0.03246496617794037, 0.056230902671813965, -0.03774179145693779, -0.07237006723880768, -0.30212584137916565, 0.017727840691804886, -0.07856012135744095, -0.30994194746017456, 0.1863621473312378, 0.040432579815387726, 0.14403414726257324, 0.06193655729293823, 0.36202555894851685, 0.6019586324691772, 0.40557700395584106, -0.1647844910621643, 0.10734259337186813, 0.03777242451906204, -0.2596316933631897, -0.1934569925069809, -0.043695177882909775, -0.013956326059997082, -0.010729467496275902, -0.44869276881217957, -0.38156425952911377, 0.16156987845897675, -0.1280699223279953, 0.0727711021900177, -0.14765994250774384, 0.05844828858971596, -0.0028753280639648438, 0.6331417560577393, 0.5671983361244202, -0.22851821780204773, 0.08923426270484924, 0.336467444896698, 0.11590482294559479, -0.3352690637111664, 0.4838377833366394, 0.05015650391578674, 0.10745851695537567, -0.10666842758655548, -0.4239083528518677, 0.18046201765537262, -0.5068800449371338, -0.22542396187782288, 0.1737489402294159, -0.5565367341041565, 0.17674864828586578, -0.07062982022762299, -0.03899850696325302, -0.33048513531684875, -0.0028003675397485495, 0.29751551151275635, 0.4849635362625122, 0.0340145006775856, 0.2780067026615143, -0.193300262093544, -0.06715764105319977, 0.26429861783981323, -0.12160932272672653, -0.3572373688220978, -0.31178992986679077, -0.04289230331778526, -0.121290422976017, 0.031080026179552078, -0.03785647451877594, -0.3895494043827057, -0.012586976401507854, -0.2790810167789459, -0.16043052077293396, -0.27749618887901306, 0.031025076285004616, -0.22630879282951355, -0.06190367415547371, -0.33485132455825806, -0.3435286283493042, 0.11446848511695862, -0.08758202940225601, 0.2657090425491333, -0.3278542160987854, -0.168110191822052, 0.22280488908290863, -0.025595709681510925, -0.18210668861865997, 0.6047659516334534, 0.28620028495788574, -0.0080929109826684, 0.06207405775785446, 0.32744255661964417, 0.39119091629981995, -0.030648717656731606, 0.005734108854085207, -0.12919315695762634, 0.18323764204978943, -0.12301748245954514, -0.05245789512991905, -0.043942540884017944, -0.2799319624900818, -0.20842553675174713, -0.20216119289398193, 0.5459466576576233, 0.02758137322962284, 0.629275918006897, 0.2911069691181183, 0.04996936395764351, -0.012831170111894608, -0.3354654908180237, -0.17226038873195648, -0.5019679069519043, -0.2116856426000595, -0.4157891571521759, 0.11526723206043243, -0.13372965157032013, -0.27769961953163147, 0.2355109602212906, 0.43055251240730286, 0.17799296975135803, 0.033647146075963974, -0.23415637016296387, 0.13413676619529724, 0.14352886378765106, -0.19840008020401, 0.6893718242645264, 0.18910999596118927, 0.15224874019622803, -0.04077431187033653, 0.08033353835344315, 0.03301870450377464, -0.18132685124874115, 0.28554320335388184, 0.25729405879974365, 0.08319494873285294, -0.001740291714668274, -0.1671251654624939, -0.13735824823379517, 0.053016893565654755, 0.06068029999732971, 0.22533895075321198, -0.33681851625442505, -0.053813010454177856, -0.07482226938009262, -0.1295631378889084, -0.3124653995037079, -0.5548503398895264, -0.03342396765947342, -0.18639419972896576, -0.22622975707054138, 0.20903997123241425, 0.014331146143376827, -0.21730929613113403, 0.1576506793498993, -0.15920953452587128, 0.16644908487796783, -0.41059446334838867, 0.08151506632566452, 0.050145212560892105, -0.4121739864349365, -0.03657102212309837, 0.0229784045368433, 0.06295396387577057, 0.013407127000391483, 0.28876903653144836, 0.36025846004486084, -0.2948746383190155, 0.18425516784191132, -0.3726988732814789, 0.1845194548368454, -0.39978158473968506, -0.267551451921463, 0.22958461940288544, -0.3035169243812561, 0.014626649208366871, -0.42857247591018677, 0.3069131076335907, -0.16745948791503906, -0.13079631328582764, -0.33435043692588806, 0.32829388976097107, 0.1910603642463684, -0.17773818969726562, -0.31340304017066956, -0.40244102478027344, -0.14577074348926544, 0.39054152369499207, -0.06979208439588547, 0.2750597298145294, 0.07240902632474899, -0.47120168805122375, -0.028921328485012054, -0.0028674418572336435, -0.12118394672870636, 0.06887868791818619, 0.22075870633125305, -0.07444778829813004, -0.4078603982925415, 0.12988002598285675, 0.08459219336509705, -0.1246912032365799, 0.3680253028869629, -0.24733686447143555, -0.06267540901899338, 0.27278220653533936, 0.05128677189350128, -0.06166406348347664, 0.16137820482254028, 0.046861518174409866, 0.16062414646148682, 0.29216820001602173, 0.06726724654436111, -0.1720878779888153, -0.24137789011001587, 0.3609777092933655, 0.1215418130159378, 0.13401354849338531, -0.07184873521327972, 0.09102098643779755, -0.05060461908578873, 0.6642798185348511, 0.2965461313724518, 0.2599330544471741, -0.15238462388515472, 0.06831342726945877, -0.34164005517959595, 0.08138629794120789, -0.22828581929206848, 0.14147549867630005, -0.0514206700026989, -0.3744939863681793, 0.3664929270744324, 0.032048601657152176, -0.13655845820903778, -0.26294878125190735, -0.012271900661289692, -0.08253291249275208, -0.53285151720047, 0.008546235039830208, -0.18302984535694122, 0.2694477140903473, -0.13070566952228546, 0.06471651047468185, 0.032201189547777176, -0.03419843688607216, -0.14848755300045013, -0.19269603490829468, -0.0961197167634964, -0.14421388506889343, -0.13510818779468536, -0.5458654165267944, -0.05232841894030571, -0.013712217099964619, 0.24540738761425018, -0.1310090720653534, -0.04341272637248039, -0.19963489472866058, 0.29792720079421997, 0.2764440178871155, 0.28615865111351013, -0.07402978837490082, -0.1080964133143425, 0.14678682386875153, 0.0468648225069046, 0.19412098824977875, -0.15025007724761963, 0.07547260075807571, 0.02573367767035961, 0.04000896215438843, 0.02105630934238434, -0.5988473296165466, -0.27710795402526855, 0.5790029168128967, -0.2179688811302185, -0.3166866898536682, 0.02838323451578617, 0.05465530604124069, 0.1528545618057251, 0.08562318235635757, -0.028498493134975433, 0.12638428807258606, 0.10292913019657135, 0.06259778887033463, 0.1315196454524994, -0.037614330649375916, -0.18430204689502716, 0.12430355697870255, 0.12849855422973633, 0.00018755460041575134, -0.08899044245481491, 0.15551337599754333, 0.23965565860271454, 0.28416621685028076, -0.32764968276023865, 0.49771323800086975, -0.36366844177246094, -0.10032542794942856, 0.021687593311071396, -0.2143649011850357, 0.5283656716346741, 0.41589808464050293, 0.1421385407447815, 0.38511359691619873, 0.04732182249426842, 0.047360844910144806, -0.19367410242557526, -0.11665061116218567, 0.32908621430397034, -0.26012122631073, -0.4622158408164978, -0.7009152173995972, 0.41299328207969666, 0.07184483855962753, -0.2251182347536087, 0.24657076597213745, 0.15709906816482544, -0.3763376772403717, 0.5711115002632141, -0.0197212602943182, 0.8757331371307373, 0.0014642432797700167, 0.1962801069021225, 0.019914433360099792, -0.2845604717731476, 0.6315015554428101, -0.015095226466655731, 0.10312186926603317, -0.37773656845092773, 0.3019932806491852, 0.0610177256166935, 0.0848897323012352, 0.22855675220489502, 0.5355198979377747, 0.013650287874042988, 0.18368710577487946, 0.06374082714319229, 0.27864792943000793, -0.29285526275634766, 0.3380335867404938, 0.20744098722934723, -0.08355356752872467, -0.14133723080158234, -0.04221636429429054, 0.02688746526837349, -0.258587509393692, 0.19815722107887268, -0.18899017572402954, -0.022848578169941902, 0.141093447804451, -0.27698200941085815, 0.06596139073371887, -0.004031287506222725, 0.11964206397533417, -0.12575918436050415, -0.44830724596977234, 0.4320882558822632, 0.16505302488803864, 0.3249998092651367, -0.18011324107646942, -0.038074903190135956, 0.47148025035858154, -0.05818836763501167, 0.0914946123957634, -0.16750501096248627, 0.025093359872698784, 0.22142186760902405, 0.029510924592614174, 0.10628574341535568, 0.026310540735721588, -0.02527446486055851, -0.0637805163860321, -0.26449695229530334, 0.21506066620349884, 0.45295390486717224, -0.5559491515159607, 0.17467337846755981, -0.012984052300453186, 0.05249324068427086, -0.23507475852966309, 0.017459586262702942, -0.239332914352417, -0.2196994125843048, 0.016380349174141884, -0.18049490451812744, -0.04825229570269585, -0.026684382930397987, -0.29558905959129333, 0.2380407154560089, -0.015307963825762272, 0.4281602203845978, -0.1167140007019043, 0.16510877013206482, -0.232161283493042, 0.07606153190135956, 0.17264436185359955, -0.2626325190067291, 0.2844863533973694, 0.11356760561466217, -0.27455252408981323, 0.15061405301094055, -0.015604117885231972, 0.052263036370277405, 0.6398312449455261, -0.15392324328422546, 0.01546485349535942, -0.20868001878261566, 0.004173244349658489, -0.06663280725479126, 0.06332934647798538, -0.04518800973892212, 0.46698474884033203, -0.1915591061115265, 0.41386210918426514, -0.13755114376544952, -0.018592841923236847, 0.2807841897010803, 0.262392520904541, -0.15228039026260376, 0.00963304378092289, 0.1951940804719925, 0.07528989017009735, 0.005358344409614801, -0.006513390224426985, 0.20427770912647247, -0.06048789247870445, -0.3154314160346985, 0.12557700276374817, 0.1368718445301056, 0.1940063089132309, -0.39656922221183777, -0.16864928603172302, 0.07181937992572784, -0.15368948876857758, 0.17872723937034607, 0.06687215715646744, 0.2828254997730255, 0.4819812476634979, 0.3057697117328644, 0.11159984022378922, -0.3644503951072693, 0.18410365283489227, -0.12273786216974258, 0.3808995485305786, 0.21493041515350342, 0.2762942314147949, -0.16514000296592712, 0.09467661380767822, 0.1048569604754448, -0.04907926544547081, 0.19031324982643127, -0.3088618218898773, 0.013765405863523483, 0.03284619376063347, -0.030747033655643463, -0.14969192445278168, 0.2588709890842438, 0.567273736000061, 0.12643209099769592, -0.002587564056739211, 0.22976329922676086, 0.08614949882030487, -0.09101269394159317, 0.2563643455505371, 0.28159099817276, -0.15940631926059723, 0.08424223214387894, 0.5541465878486633, 0.12478671967983246, 0.2655104994773865, -0.4000857472419739, -0.03358311206102371, -0.05661614611744881, -0.044363103806972504, 0.5013855695724487, 0.8824548721313477, 0.206043541431427, -0.01995527558028698, 0.4297122061252594, -0.19077293574810028, 0.24227681756019592, 0.4915968179702759, -0.17309696972370148, 0.33046814799308777, 0.3147033452987671, -0.122627392411232, 0.5907780528068542, 0.19488202035427094, -0.40433844923973083, -0.0030669120606034994, -0.01677117869257927, -0.23411928117275238, -0.15214712917804718, 0.7617432475090027, -0.01158185675740242, -0.3254050016403198, 0.09411852806806564, 0.2618066966533661, 0.18014584481716156, -0.20263785123825073, -0.1622176170349121, 0.05505569651722908, -0.3555072546005249, 0.008617879822850227, -0.0889238715171814, -0.20625172555446625, 0.3511849045753479, 0.03732092306017876, 0.2648569345474243, -0.15261143445968628, -0.032472241669893265, 0.17793399095535278, 0.5029485821723938, -0.31820857524871826, -0.10205742716789246, -0.0024594180285930634, -0.293645441532135, 0.1971893459558487, -0.09273392707109451, -0.005650059785693884, -0.14058059453964233, -0.3290477693080902, -0.35682880878448486, 0.20284199714660645, -0.1465296745300293, -0.00009942549513652921, 0.036620646715164185, 0.12525320053100586, 0.30373483896255493, 0.1848142296075821, -0.04838896170258522, -0.026214780285954475, -0.15177983045578003, 0.281892329454422, -0.07489264011383057, -0.047114718705415726, 0.6482117176055908, 0.1358204483985901, -0.18381641805171967, 0.023075588047504425, 0.17055019736289978, -0.23164010047912598, -0.06651780009269714, 0.34107890725135803, -0.04825206845998764, 0.11722388863563538, -0.11295092850923538, -0.03360506519675255, -0.03483888506889343, 0.40514329075813293, -0.20681510865688324, -0.28671565651893616, -0.2614647150039673, -0.4266860783100128, -0.08076164126396179, -0.0739796832203865, -0.1493707299232483, -0.24868901073932648, -0.09933720529079437, 0.13811545073986053, -0.08726999163627625, -0.1610434651374817, 0.11898627877235413, -0.07627600431442261, 0.3919477164745331, 0.3623473048210144, -0.2248709797859192, 0.006812839303165674, -0.03206920623779297, 0.16071519255638123, 0.127204030752182, 0.11123882234096527, -0.10884104669094086, 0.3447790741920471, -0.32984745502471924, 0.17180854082107544, -0.07115079462528229, 0.5625975728034973, 0.20206698775291443, -0.05933590233325958, 0.012389561161398888, 0.23775726556777954, 0.029855815693736076, -0.06352149695158005, 0.03534272685647011, -0.0028577486518770456, -0.5393221378326416, 0.5002475380897522, 0.017259718850255013, 0.1763506978750229, -0.23865626752376556, 0.4447185695171356, -0.14618808031082153, -0.04464218392968178, -0.25948870182037354, 0.1652458906173706, -0.0692136213183403, 0.22428353130817413, -0.06041455268859863, 0.3944738507270813, 0.420529305934906, 0.2640943229198456, -0.12344487756490707, -0.025244109332561493, -0.2117692530155182, -0.3554522693157196, 0.23516441881656647, -0.07453874498605728, -0.13716080784797668, 0.10333750396966934, 0.3187505602836609, 0.17163394391536713, 0.016239121556282043, -0.7228701710700989, -0.26910263299942017, 0.1650715172290802, -0.2762211561203003, 0.020327286794781685, 0.03865714371204376, 0.23195233941078186, -0.2397080510854721, 0.013042972423136234, -0.045807451009750366, 0.2986750304698944, -0.07945689558982849, 0.06468239426612854, -0.11545092612504959 ]
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
[ -0.11933844536542892, 0.08862767368555069, -0.008977877907454967, 0.21019725501537323, 0.3692678213119507, 0.026059893891215324, 0.44878166913986206, 0.41310209035873413, 0.4330812692642212, 0.01256692036986351, 0.1238437220454216, 0.39542990922927856, -0.15054894983768463, 0.021962737664580345, 0.12090493738651276, -0.4289896786212921, -0.03373947739601135, 0.36367297172546387, -0.04146667942404747, 0.06297369301319122, -0.18768362700939178, 0.10851698368787766, -0.2651914656162262, 0.25500598549842834, -0.3057228922843933, -0.057025808840990067, 0.21592070162296295, 0.08206350356340408, -0.07891874760389328, -0.5590065121650696, 0.27708321809768677, -0.21704912185668945, 0.0644516572356224, 0.11474668979644775, -0.00010995571938110515, 0.20530638098716736, 0.33269718289375305, 0.030212869867682457, -0.5799456834793091, -0.18957771360874176, -0.47733166813850403, -0.11893873661756516, 0.2974317669868469, -0.24970759451389313, -0.028513627126812935, 0.05389740690588951, 0.03591078519821167, -0.07441216707229614, 0.2967468202114105, 0.531506359577179, 0.23788897693157196, 0.5236923098564148, 0.15454918146133423, 0.03449370339512825, -0.07856779545545578, -0.17057092487812042, -0.08211507648229599, 0.22396805882453918, 0.07387740164995193, -0.3915436565876007, -0.013373903930187225, 0.12692449986934662, 0.15117241442203522, 0.263701856136322, 0.5083708763122559, 0.01439398992806673, 0.24052372574806213, -0.15062718093395233, 0.014947105199098587, 0.07640064507722855, 0.3353829085826874, -0.1649331897497177, -0.2715999186038971, -0.2243589609861374, 0.24766606092453003, -0.33475616574287415, 0.23653045296669006, -0.036676667630672455, -0.16220439970493317, -0.00819715391844511, -0.2692277729511261, -0.02814740687608719, -0.24182094633579254, 0.35268712043762207, -0.0582139827311039, 0.4996492266654968, 0.04358334839344025, 0.14274729788303375, -0.05806553736329079, -0.07545780390501022, 0.1403771936893463, -0.0870911106467247, -0.12930051982402802, 0.25781315565109253, -0.38315534591674805, -0.036467500030994415, -0.03494661673903465, -0.11697590351104736, -0.03162477910518646, 0.263520210981369, 0.2199898362159729, -0.2248734086751938, 0.32383885979652405, 0.3154154121875763, 0.2229883372783661, 0.42046889662742615, 0.0727653056383133, 0.06359227746725082, -0.08280493319034576, 0.13602614402770996, -0.23372821509838104, -0.15548528730869293, -0.15553614497184753, -0.17655576765537262, 0.09289789199829102, -0.15460222959518433, 0.22602057456970215, -0.07111575454473495, -0.49847379326820374, 0.11472073942422867, -0.14340490102767944, 0.04629025608301163, 0.2051147073507309, 0.17948497831821442, -0.12763401865959167, 0.23223420977592468, 0.14412450790405273, 0.28256696462631226, -0.2519770562648773, -0.043135613203048706, -0.1662895381450653, 0.2492133229970932, -0.23438303172588348, -0.18224912881851196, 0.2427540421485901, 0.193400040268898, 0.3994782567024231, -0.10116763412952423, 0.11126401275396347, -0.20175547897815704, 0.22737567126750946, -0.04551948234438896, -0.1288793981075287, 0.2566531002521515, -0.21606101095676422, 0.15489795804023743, 0.20159123837947845, -0.2824235260486603, -0.08482388406991959, -0.10174447298049927, -0.27263563871383667, -0.30809205770492554, -0.12625139951705933, 0.28141555190086365, 0.14738212525844574, -0.24296484887599945, -0.15226320922374725, -0.06118768826127052, 0.14749260246753693, -0.36872363090515137, -0.11810611188411713, -0.46125850081443787, -0.2206392139196396, -0.023005584254860878, 0.3976435661315918, 0.20490935444831848, 0.13138674199581146, -0.10897259414196014, -0.11542075127363205, 0.07785411924123764, -0.03179730847477913, 0.1227133721113205, -0.438356876373291, 0.542163074016571, -0.061055924743413925, 0.118863046169281, 0.64134681224823, -0.3427872359752655, -0.45785459876060486, 0.12356872856616974, -0.01428011991083622, 0.209327831864357, -0.225368931889534, -0.016720708459615707, -0.013226387090981007, -0.18455132842063904, 0.2123136669397354, 0.6192132234573364, -0.015042993240058422, 0.0824793204665184, -0.12689702212810516, -0.015549133531749249, 0.500219464302063, 0.10985691845417023, 0.09252550452947617, -0.007928984239697456, -0.04170479252934456, 0.4966483414173126, -0.0013131386367604136, -0.12597055733203888, -0.014860047958791256, -0.013630194589495659, -0.06468609720468521, -0.10460405051708221, -0.21782632172107697, -0.07388683408498764, -0.27766451239585876, 0.3043040931224823, -0.11754775047302246, -0.058531954884529114, -0.04715004563331604, 0.1061866357922554, -0.3491132855415344, 0.16848868131637573, -0.4232693016529083, -0.37731894850730896, 0.17209340631961823, 0.15513364970684052, -0.010822752490639687, 0.23844130337238312, -0.20474109053611755, 0.219436913728714, -0.27402758598327637, 0.05550083518028259, -0.21892865002155304, 0.16128963232040405, -0.05588398873806, -0.042422182857990265, -0.0094792814925313, 0.3087964355945587, 0.038331370800733566, -0.01799662411212921, -0.3062887191772461, 0.28603023290634155, -0.12548455595970154, 0.08035627007484436, -0.03226484730839729, -0.14337030053138733, 0.06880644708871841, -0.07154319435358047, -0.17518453299999237, 0.20819184184074402, 0.331950843334198, -0.18012624979019165, -0.04314278066158295, 0.11287608742713928, 0.1272120177745819, 0.08142179995775223, 0.09886399656534195, 0.16397236287593842, 0.14399196207523346, -0.06582631915807724, 0.1012796014547348, 0.07743120193481445, 0.23370946943759918, -0.09498720616102219, 0.20471063256263733, -0.15645115077495575, -0.1671450287103653, -0.11880644410848618, 0.3131200075149536, 0.17005865275859833, 0.16741715371608734, 0.08148640394210815, -0.2752522826194763, -0.09487438946962357, 0.07855788618326187, 0.14604908227920532, 0.40096521377563477, -0.05183282122015953, -0.10834052413702011, -0.052570827305316925, -0.416596919298172, 0.0711255818605423, 0.1707078218460083, 0.03662902116775513, 0.20912893116474152, 0.30185675621032715, 0.041315339505672455, 0.08732044696807861, -0.29526710510253906, -0.35701707005500793, 0.28243377804756165, 0.440530002117157, -0.22848205268383026, 0.20788012444972992, -0.03053182363510132, -0.0025113264564424753, -0.06784813851118088, -0.33363381028175354, -0.18862831592559814, -0.34299373626708984, -0.12133599072694778, 0.4337753653526306, 0.024119378998875618, 0.1810045689344406, -0.39776164293289185, -0.022050954401493073, 0.19475102424621582, -0.11478555202484131, -0.031379494816064835, -0.31579792499542236, -0.26932892203330994, 0.045942142605781555, 0.3818606734275818, 0.09107666462659836, 0.35109788179397583, -0.06677065789699554, -0.09253539890050888, -0.0070348638109862804, -0.05841507017612457, -0.0755671039223671, -0.18372094631195068, 0.24404765665531158, 0.009521503932774067, 0.28558358550071716, -0.17702975869178772, -0.5220332145690918, 0.5017468929290771, -0.2557017505168915, -0.05778532475233078, 0.3151065707206726, 0.2711130976676941, -0.24919050931930542, -0.07947684079408646, -0.37468844652175903, -0.1264503300189972, -0.2806057631969452, -0.05946894735097885, 0.07482901960611343, 0.06202087178826332, 0.3974674344062805, 0.07085659354925156, 0.2160108983516693, -0.0025738144759088755, 0.11343798786401749, -0.048262786120176315, -0.09006398916244507, 0.18598473072052002, -0.31003379821777344, -0.3173184096813202, -0.12134277820587158, 0.009935948997735977, 0.3347616493701935, 0.11073566973209381, -0.3489592671394348, -0.21413171291351318, -0.10495787858963013, 0.08531288057565689, -0.2213253378868103, -0.04006680101156235, 0.18286007642745972, 0.0723033994436264, -0.06572244316339493, -0.0866440087556839, 0.035989612340927124, 0.07124002277851105, 0.15234600007534027, 0.13496200740337372, -0.0063046677969396114, 0.0892268493771553, -0.08488413691520691, 0.6008990406990051, 0.2186937928199768, -0.21437722444534302, 0.30028611421585083, -0.10327658802270889, 0.005144711118191481, -0.20868366956710815, -0.4398044943809509, 0.2641642689704895, -0.11523295938968658, 0.18292878568172455, 0.17231491208076477, -0.07949358969926834, -0.2247839868068695, -0.19255922734737396, 0.06885148584842682, -0.17844754457473755, -0.2727796733379364, 0.14616838097572327, -0.2173946499824524, 0.27871108055114746, -0.0688643679022789, 0.22253717482089996, -0.20522312819957733, -0.16798074543476105, 0.06985094398260117, 0.16693419218063354, -0.0905456691980362, 0.07967526465654373, -0.24400167167186737, -0.27610859274864197, -0.47995808720588684, -0.027839848771691322, 0.2410317063331604, 0.6437501311302185, -0.08638478815555573, 0.03718903288245201, 0.08072245866060257, 0.08174151927232742, 0.13737379014492035, -0.2301362156867981, 0.18226361274719238, 0.28913068771362305, 0.031824640929698944, -0.3197464346885681, -0.06433165818452835, -0.14011214673519135, 0.3580164909362793, 0.07665183395147324, -0.05443182960152626, -0.037231311202049255, -0.16765408217906952, 0.1398317664861679, 0.2599390745162964, -0.18729500472545624, -0.04411308094859123, -0.0917343720793724, -0.08981035649776459, -0.30776476860046387, -0.00615481473505497, -0.00994766503572464, 0.33566054701805115, 0.12296929955482483, 0.26572492718696594, -0.5462992191314697, 0.11665716022253036, 0.09016522765159607, 0.1745867282152176, 0.0959530621767044, 0.05794429033994675, 0.18724678456783295, -0.16663391888141632, 0.18320998549461365, -0.1459045261144638, 0.563380241394043, 0.14737606048583984, -0.33079883456230164, -0.31952935457229614, 0.021663231775164604, 0.1674334853887558, 0.1839803010225296, -0.018585484474897385, -0.14112089574337006, -0.2769453227519989, -0.061463501304388046, -0.008708842098712921, 0.3045703172683716, 0.5610411167144775, 0.002627668436616659, -0.3271808326244354, -0.2585405707359314, 0.48283934593200684, -0.2850031554698944, 0.11732912063598633, 0.3215573728084564, 0.05337723344564438, -0.31319230794906616, 0.08413907885551453, 0.07783256471157074, 0.6767755746841431, 0.16953879594802856, 0.3111465871334076, 0.3826707899570465, 0.18392826616764069, 0.41427063941955566, -0.043089501559734344, 0.2847144305706024, -0.3927355706691742, -0.13342593610286713, -0.08068574965000153, -0.06546270102262497, 0.10708170384168625, -0.008873440325260162, -0.025438690558075905, 0.15062108635902405, 0.15688933432102203, -0.004141780082136393, 0.057147782295942307, 0.4492528736591339, -0.03878398612141609, -0.19681917130947113, -0.3470250964164734, 0.15601642429828644, -0.21799980103969574, 0.2975102961063385, -0.09073382616043091, -0.06648409366607666, 0.05520455539226532, -0.06735203415155411, -0.25555863976478577, 0.2574385702610016, -0.29947468638420105, 0.4594579339027405, 0.05782169848680496, -0.475681334733963, 0.19841517508029938, 0.20709797739982605, 0.17566406726837158, 0.005698727443814278, -0.2995942533016205, 0.23877884447574615, -0.29008978605270386, -0.2218664139509201, -0.06728777289390564, 0.11194716393947601, 0.42007339000701904, -0.1725514531135559, -0.14637790620326996, 0.14794056117534637, -0.12065393477678299, -0.24150243401527405, 0.10395880788564682, 0.036866508424282074, 0.35248932242393494, -0.39257583022117615, -0.5152195692062378, -0.002420816570520401, -0.18207627534866333, -0.17491012811660767, 0.1745000183582306, 0.023450618609786034, -0.18903139233589172, 0.15189562737941742, 0.1498272567987442, -0.19897985458374023, 0.11640633642673492, 0.24804235994815826, -0.2495405375957489, 0.1462831199169159, 0.7637571096420288, 0.03394559025764465, 0.012805139645934105, -0.3138304054737091, -0.054288268089294434, 0.12508445978164673, -0.48199984431266785, -0.11712734401226044, 0.12550395727157593, 0.16912569105625153, -0.1442469209432602, 0.4983660876750946, 0.06302104145288467, -0.018844809383153915, 0.24718545377254486, -0.5424140691757202, -0.10621922463178635, -0.04302157834172249, 0.017753537744283676, 0.23777537047863007, -0.002624211832880974, 0.11488901823759079, 0.13550961017608643, 0.10338855534791946, -0.27263742685317993, 0.014254532754421234, -0.15762585401535034, 0.13799649477005005, 0.21783460676670074, -0.1565314084291458, 0.2232816368341446, -0.0769471824169159, 0.15638616681098938, -0.09584657102823257, -0.12303436547517776, -0.24829335510730743, -0.2548377513885498, 0.10011143982410431, 0.24556712806224823, -0.1645212471485138, -0.03067934699356556, -0.1023336872458458, 0.023677632212638855, -0.2311049848794937, -0.00846262276172638, -0.1177278533577919, -0.2933426797389984, 0.38943222165107727, 0.04389796033501625, 0.2783687710762024, -0.11169014126062393, 0.1263556182384491, -0.3007573187351227, 0.14912062883377075, -0.01849590614438057, -0.07522980868816376, 0.27581319212913513, 0.09365329146385193, -0.46588683128356934, 0.005420649889856577, -0.40260517597198486, 0.03714754059910774, 0.36982133984565735, -0.4551105201244354, 0.15995842218399048, -0.1316281408071518, 0.09546535462141037, 0.46568307280540466, -0.2714085280895233, -0.0036669739056378603, 0.3496656119823456, 0.2378426045179367, -0.297417551279068, -0.18926052749156952, 0.2399909794330597, 0.061345525085926056, 0.03245915472507477, 0.027238842099905014, 0.04666132479906082, -0.30374857783317566, -0.3212345540523529, 0.08927971869707108, -0.029340069741010666, -0.2656012177467346, -0.011407874524593353, 0.33932632207870483, -0.002364365616813302, -0.059514716267585754, -0.050049375742673874, 0.21102096140384674, 0.03539273142814636, 0.38269317150115967, -0.2926008105278015, 0.2069980949163437, 0.05926349014043808, 0.04391063377261162, -0.11114291101694107, -0.2192171812057495, 0.05030057206749916, -0.15784087777137756, 0.03383427858352661, 0.295282781124115, 0.09981466829776764, 0.15308073163032532, -0.2855035066604614, -0.12453679740428925, -0.20802298188209534, 0.22792452573776245, -0.1790422797203064, -0.048906464129686356, -0.24299554526805878, 0.021017009392380714, -0.025876479223370552, -0.09469880163669586, -0.14167028665542603, -0.03817872330546379, 0.3095560371875763, 0.2532220184803009, -0.1722707748413086, -0.4555346965789795, 0.10826654732227325, 0.34329426288604736, 0.1800471842288971, -0.24588419497013092, 0.3368530869483948, 0.1938154697418213, -0.06877834349870682, -0.014124344103038311, 0.3971104323863983, 0.4357682168483734, 0.3383936285972595, -0.36422058939933777, 0.08116783946752548, -0.2033807337284088, -0.056821923702955246, -0.023260261863470078, 0.2695125639438629, 0.10607320815324783, 0.11737607419490814, 0.462422639131546, 0.12148632109165192, -0.124727264046669, 0.04363897442817688, -0.08805680274963379, -0.26053252816200256, -0.262947142124176, 0.5191485285758972, 0.18579453229904175, -0.11438801884651184, -0.32065099477767944, 0.18993781507015228, -0.27498695254325867, -0.08091452717781067, 0.06285151094198227, 0.24058954417705536, 0.11315396428108215, -0.21382759511470795, 0.0705089196562767, 0.0003152035642415285, 0.7061733603477478, 0.4877219796180725, -0.01771940104663372, -0.3658052682876587, -0.11652376502752304, -0.5989871025085449, 0.2178361713886261, -0.3639681041240692, -0.2690182328224182, 0.2833365499973297, -0.0723314955830574, -0.2880823016166687, 0.24852751195430756, 0.12761974334716797, 0.15819577872753143, -0.05727637931704521, 0.09834792464971542, -0.4907650947570801, -0.14860443770885468, -0.03065044991672039, -0.007675084285438061, -0.08318091183900833, -0.257053017616272, 0.14532335102558136, -0.43499451875686646, 0.11610305309295654, 0.011985060758888721, 0.06934806704521179, -0.0002395449992036447, 0.09664040058851242, 0.3544246256351471, 0.16532395780086517, 0.5012481808662415, -0.03556109592318535, 0.0036573235411196947, -0.19115452468395233, -0.1566217541694641, -0.22899232804775238, 0.1318226009607315, 0.08762913942337036, 0.5133118629455566, -0.026107510551810265, -0.06854528188705444, -0.2640271484851837, 0.4031679034233093, -0.08100655674934387, 0.08831585943698883, -0.13508093357086182, 0.07731451839208603, -0.07034944742918015, 0.22758422791957855, 0.26436492800712585, 0.42454031109809875, -0.1783217340707779, 0.20220941305160522, -0.27229201793670654, -0.46024590730667114, 0.32893550395965576, -0.661480724811554, -0.5177905559539795, 0.0751223936676979, 0.1161927729845047, -0.14233405888080597, 0.058194272220134735, -0.6006340980529785, 0.07056694477796555, 0.24261268973350525, -0.11140575259923935, -0.3171198070049286, 0.21794509887695312, 0.01654207892715931, 0.08000141382217407, 0.04357372596859932, 0.44835659861564636, -0.14847296476364136, -0.1597675234079361, -0.0049345833249390125, -0.16325202584266663 ]
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).
[ -0.11933844536542892, 0.08862767368555069, -0.008977877907454967, 0.21019725501537323, 0.3692678213119507, 0.026059893891215324, 0.44878166913986206, 0.41310209035873413, 0.4330812692642212, 0.01256692036986351, 0.1238437220454216, 0.39542990922927856, -0.15054894983768463, 0.021962737664580345, 0.12090493738651276, -0.4289896786212921, -0.03373947739601135, 0.36367297172546387, -0.04146667942404747, 0.06297369301319122, -0.18768362700939178, 0.10851698368787766, -0.2651914656162262, 0.25500598549842834, -0.3057228922843933, -0.057025808840990067, 0.21592070162296295, 0.08206350356340408, -0.07891874760389328, -0.5590065121650696, 0.27708321809768677, -0.21704912185668945, 0.0644516572356224, 0.11474668979644775, -0.00010995571938110515, 0.20530638098716736, 0.33269718289375305, 0.030212869867682457, -0.5799456834793091, -0.18957771360874176, -0.47733166813850403, -0.11893873661756516, 0.2974317669868469, -0.24970759451389313, -0.028513627126812935, 0.05389740690588951, 0.03591078519821167, -0.07441216707229614, 0.2967468202114105, 0.531506359577179, 0.23788897693157196, 0.5236923098564148, 0.15454918146133423, 0.03449370339512825, -0.07856779545545578, -0.17057092487812042, -0.08211507648229599, 0.22396805882453918, 0.07387740164995193, -0.3915436565876007, -0.013373903930187225, 0.12692449986934662, 0.15117241442203522, 0.263701856136322, 0.5083708763122559, 0.01439398992806673, 0.24052372574806213, -0.15062718093395233, 0.014947105199098587, 0.07640064507722855, 0.3353829085826874, -0.1649331897497177, -0.2715999186038971, -0.2243589609861374, 0.24766606092453003, -0.33475616574287415, 0.23653045296669006, -0.036676667630672455, -0.16220439970493317, -0.00819715391844511, -0.2692277729511261, -0.02814740687608719, -0.24182094633579254, 0.35268712043762207, -0.0582139827311039, 0.4996492266654968, 0.04358334839344025, 0.14274729788303375, -0.05806553736329079, -0.07545780390501022, 0.1403771936893463, -0.0870911106467247, -0.12930051982402802, 0.25781315565109253, -0.38315534591674805, -0.036467500030994415, -0.03494661673903465, -0.11697590351104736, -0.03162477910518646, 0.263520210981369, 0.2199898362159729, -0.2248734086751938, 0.32383885979652405, 0.3154154121875763, 0.2229883372783661, 0.42046889662742615, 0.0727653056383133, 0.06359227746725082, -0.08280493319034576, 0.13602614402770996, -0.23372821509838104, -0.15548528730869293, -0.15553614497184753, -0.17655576765537262, 0.09289789199829102, -0.15460222959518433, 0.22602057456970215, -0.07111575454473495, -0.49847379326820374, 0.11472073942422867, -0.14340490102767944, 0.04629025608301163, 0.2051147073507309, 0.17948497831821442, -0.12763401865959167, 0.23223420977592468, 0.14412450790405273, 0.28256696462631226, -0.2519770562648773, -0.043135613203048706, -0.1662895381450653, 0.2492133229970932, -0.23438303172588348, -0.18224912881851196, 0.2427540421485901, 0.193400040268898, 0.3994782567024231, -0.10116763412952423, 0.11126401275396347, -0.20175547897815704, 0.22737567126750946, -0.04551948234438896, -0.1288793981075287, 0.2566531002521515, -0.21606101095676422, 0.15489795804023743, 0.20159123837947845, -0.2824235260486603, -0.08482388406991959, -0.10174447298049927, -0.27263563871383667, -0.30809205770492554, -0.12625139951705933, 0.28141555190086365, 0.14738212525844574, -0.24296484887599945, -0.15226320922374725, -0.06118768826127052, 0.14749260246753693, -0.36872363090515137, -0.11810611188411713, -0.46125850081443787, -0.2206392139196396, -0.023005584254860878, 0.3976435661315918, 0.20490935444831848, 0.13138674199581146, -0.10897259414196014, -0.11542075127363205, 0.07785411924123764, -0.03179730847477913, 0.1227133721113205, -0.438356876373291, 0.542163074016571, -0.061055924743413925, 0.118863046169281, 0.64134681224823, -0.3427872359752655, -0.45785459876060486, 0.12356872856616974, -0.01428011991083622, 0.209327831864357, -0.225368931889534, -0.016720708459615707, -0.013226387090981007, -0.18455132842063904, 0.2123136669397354, 0.6192132234573364, -0.015042993240058422, 0.0824793204665184, -0.12689702212810516, -0.015549133531749249, 0.500219464302063, 0.10985691845417023, 0.09252550452947617, -0.007928984239697456, -0.04170479252934456, 0.4966483414173126, -0.0013131386367604136, -0.12597055733203888, -0.014860047958791256, -0.013630194589495659, -0.06468609720468521, -0.10460405051708221, -0.21782632172107697, -0.07388683408498764, -0.27766451239585876, 0.3043040931224823, -0.11754775047302246, -0.058531954884529114, -0.04715004563331604, 0.1061866357922554, -0.3491132855415344, 0.16848868131637573, -0.4232693016529083, -0.37731894850730896, 0.17209340631961823, 0.15513364970684052, -0.010822752490639687, 0.23844130337238312, -0.20474109053611755, 0.219436913728714, -0.27402758598327637, 0.05550083518028259, -0.21892865002155304, 0.16128963232040405, -0.05588398873806, -0.042422182857990265, -0.0094792814925313, 0.3087964355945587, 0.038331370800733566, -0.01799662411212921, -0.3062887191772461, 0.28603023290634155, -0.12548455595970154, 0.08035627007484436, -0.03226484730839729, -0.14337030053138733, 0.06880644708871841, -0.07154319435358047, -0.17518453299999237, 0.20819184184074402, 0.331950843334198, -0.18012624979019165, -0.04314278066158295, 0.11287608742713928, 0.1272120177745819, 0.08142179995775223, 0.09886399656534195, 0.16397236287593842, 0.14399196207523346, -0.06582631915807724, 0.1012796014547348, 0.07743120193481445, 0.23370946943759918, -0.09498720616102219, 0.20471063256263733, -0.15645115077495575, -0.1671450287103653, -0.11880644410848618, 0.3131200075149536, 0.17005865275859833, 0.16741715371608734, 0.08148640394210815, -0.2752522826194763, -0.09487438946962357, 0.07855788618326187, 0.14604908227920532, 0.40096521377563477, -0.05183282122015953, -0.10834052413702011, -0.052570827305316925, -0.416596919298172, 0.0711255818605423, 0.1707078218460083, 0.03662902116775513, 0.20912893116474152, 0.30185675621032715, 0.041315339505672455, 0.08732044696807861, -0.29526710510253906, -0.35701707005500793, 0.28243377804756165, 0.440530002117157, -0.22848205268383026, 0.20788012444972992, -0.03053182363510132, -0.0025113264564424753, -0.06784813851118088, -0.33363381028175354, -0.18862831592559814, -0.34299373626708984, -0.12133599072694778, 0.4337753653526306, 0.024119378998875618, 0.1810045689344406, -0.39776164293289185, -0.022050954401493073, 0.19475102424621582, -0.11478555202484131, -0.031379494816064835, -0.31579792499542236, -0.26932892203330994, 0.045942142605781555, 0.3818606734275818, 0.09107666462659836, 0.35109788179397583, -0.06677065789699554, -0.09253539890050888, -0.0070348638109862804, -0.05841507017612457, -0.0755671039223671, -0.18372094631195068, 0.24404765665531158, 0.009521503932774067, 0.28558358550071716, -0.17702975869178772, -0.5220332145690918, 0.5017468929290771, -0.2557017505168915, -0.05778532475233078, 0.3151065707206726, 0.2711130976676941, -0.24919050931930542, -0.07947684079408646, -0.37468844652175903, -0.1264503300189972, -0.2806057631969452, -0.05946894735097885, 0.07482901960611343, 0.06202087178826332, 0.3974674344062805, 0.07085659354925156, 0.2160108983516693, -0.0025738144759088755, 0.11343798786401749, -0.048262786120176315, -0.09006398916244507, 0.18598473072052002, -0.31003379821777344, -0.3173184096813202, -0.12134277820587158, 0.009935948997735977, 0.3347616493701935, 0.11073566973209381, -0.3489592671394348, -0.21413171291351318, -0.10495787858963013, 0.08531288057565689, -0.2213253378868103, -0.04006680101156235, 0.18286007642745972, 0.0723033994436264, -0.06572244316339493, -0.0866440087556839, 0.035989612340927124, 0.07124002277851105, 0.15234600007534027, 0.13496200740337372, -0.0063046677969396114, 0.0892268493771553, -0.08488413691520691, 0.6008990406990051, 0.2186937928199768, -0.21437722444534302, 0.30028611421585083, -0.10327658802270889, 0.005144711118191481, -0.20868366956710815, -0.4398044943809509, 0.2641642689704895, -0.11523295938968658, 0.18292878568172455, 0.17231491208076477, -0.07949358969926834, -0.2247839868068695, -0.19255922734737396, 0.06885148584842682, -0.17844754457473755, -0.2727796733379364, 0.14616838097572327, -0.2173946499824524, 0.27871108055114746, -0.0688643679022789, 0.22253717482089996, -0.20522312819957733, -0.16798074543476105, 0.06985094398260117, 0.16693419218063354, -0.0905456691980362, 0.07967526465654373, -0.24400167167186737, -0.27610859274864197, -0.47995808720588684, -0.027839848771691322, 0.2410317063331604, 0.6437501311302185, -0.08638478815555573, 0.03718903288245201, 0.08072245866060257, 0.08174151927232742, 0.13737379014492035, -0.2301362156867981, 0.18226361274719238, 0.28913068771362305, 0.031824640929698944, -0.3197464346885681, -0.06433165818452835, -0.14011214673519135, 0.3580164909362793, 0.07665183395147324, -0.05443182960152626, -0.037231311202049255, -0.16765408217906952, 0.1398317664861679, 0.2599390745162964, -0.18729500472545624, -0.04411308094859123, -0.0917343720793724, -0.08981035649776459, -0.30776476860046387, -0.00615481473505497, -0.00994766503572464, 0.33566054701805115, 0.12296929955482483, 0.26572492718696594, -0.5462992191314697, 0.11665716022253036, 0.09016522765159607, 0.1745867282152176, 0.0959530621767044, 0.05794429033994675, 0.18724678456783295, -0.16663391888141632, 0.18320998549461365, -0.1459045261144638, 0.563380241394043, 0.14737606048583984, -0.33079883456230164, -0.31952935457229614, 0.021663231775164604, 0.1674334853887558, 0.1839803010225296, -0.018585484474897385, -0.14112089574337006, -0.2769453227519989, -0.061463501304388046, -0.008708842098712921, 0.3045703172683716, 0.5610411167144775, 0.002627668436616659, -0.3271808326244354, -0.2585405707359314, 0.48283934593200684, -0.2850031554698944, 0.11732912063598633, 0.3215573728084564, 0.05337723344564438, -0.31319230794906616, 0.08413907885551453, 0.07783256471157074, 0.6767755746841431, 0.16953879594802856, 0.3111465871334076, 0.3826707899570465, 0.18392826616764069, 0.41427063941955566, -0.043089501559734344, 0.2847144305706024, -0.3927355706691742, -0.13342593610286713, -0.08068574965000153, -0.06546270102262497, 0.10708170384168625, -0.008873440325260162, -0.025438690558075905, 0.15062108635902405, 0.15688933432102203, -0.004141780082136393, 0.057147782295942307, 0.4492528736591339, -0.03878398612141609, -0.19681917130947113, -0.3470250964164734, 0.15601642429828644, -0.21799980103969574, 0.2975102961063385, -0.09073382616043091, -0.06648409366607666, 0.05520455539226532, -0.06735203415155411, -0.25555863976478577, 0.2574385702610016, -0.29947468638420105, 0.4594579339027405, 0.05782169848680496, -0.475681334733963, 0.19841517508029938, 0.20709797739982605, 0.17566406726837158, 0.005698727443814278, -0.2995942533016205, 0.23877884447574615, -0.29008978605270386, -0.2218664139509201, -0.06728777289390564, 0.11194716393947601, 0.42007339000701904, -0.1725514531135559, -0.14637790620326996, 0.14794056117534637, -0.12065393477678299, -0.24150243401527405, 0.10395880788564682, 0.036866508424282074, 0.35248932242393494, -0.39257583022117615, -0.5152195692062378, -0.002420816570520401, -0.18207627534866333, -0.17491012811660767, 0.1745000183582306, 0.023450618609786034, -0.18903139233589172, 0.15189562737941742, 0.1498272567987442, -0.19897985458374023, 0.11640633642673492, 0.24804235994815826, -0.2495405375957489, 0.1462831199169159, 0.7637571096420288, 0.03394559025764465, 0.012805139645934105, -0.3138304054737091, -0.054288268089294434, 0.12508445978164673, -0.48199984431266785, -0.11712734401226044, 0.12550395727157593, 0.16912569105625153, -0.1442469209432602, 0.4983660876750946, 0.06302104145288467, -0.018844809383153915, 0.24718545377254486, -0.5424140691757202, -0.10621922463178635, -0.04302157834172249, 0.017753537744283676, 0.23777537047863007, -0.002624211832880974, 0.11488901823759079, 0.13550961017608643, 0.10338855534791946, -0.27263742685317993, 0.014254532754421234, -0.15762585401535034, 0.13799649477005005, 0.21783460676670074, -0.1565314084291458, 0.2232816368341446, -0.0769471824169159, 0.15638616681098938, -0.09584657102823257, -0.12303436547517776, -0.24829335510730743, -0.2548377513885498, 0.10011143982410431, 0.24556712806224823, -0.1645212471485138, -0.03067934699356556, -0.1023336872458458, 0.023677632212638855, -0.2311049848794937, -0.00846262276172638, -0.1177278533577919, -0.2933426797389984, 0.38943222165107727, 0.04389796033501625, 0.2783687710762024, -0.11169014126062393, 0.1263556182384491, -0.3007573187351227, 0.14912062883377075, -0.01849590614438057, -0.07522980868816376, 0.27581319212913513, 0.09365329146385193, -0.46588683128356934, 0.005420649889856577, -0.40260517597198486, 0.03714754059910774, 0.36982133984565735, -0.4551105201244354, 0.15995842218399048, -0.1316281408071518, 0.09546535462141037, 0.46568307280540466, -0.2714085280895233, -0.0036669739056378603, 0.3496656119823456, 0.2378426045179367, -0.297417551279068, -0.18926052749156952, 0.2399909794330597, 0.061345525085926056, 0.03245915472507477, 0.027238842099905014, 0.04666132479906082, -0.30374857783317566, -0.3212345540523529, 0.08927971869707108, -0.029340069741010666, -0.2656012177467346, -0.011407874524593353, 0.33932632207870483, -0.002364365616813302, -0.059514716267585754, -0.050049375742673874, 0.21102096140384674, 0.03539273142814636, 0.38269317150115967, -0.2926008105278015, 0.2069980949163437, 0.05926349014043808, 0.04391063377261162, -0.11114291101694107, -0.2192171812057495, 0.05030057206749916, -0.15784087777137756, 0.03383427858352661, 0.295282781124115, 0.09981466829776764, 0.15308073163032532, -0.2855035066604614, -0.12453679740428925, -0.20802298188209534, 0.22792452573776245, -0.1790422797203064, -0.048906464129686356, -0.24299554526805878, 0.021017009392380714, -0.025876479223370552, -0.09469880163669586, -0.14167028665542603, -0.03817872330546379, 0.3095560371875763, 0.2532220184803009, -0.1722707748413086, -0.4555346965789795, 0.10826654732227325, 0.34329426288604736, 0.1800471842288971, -0.24588419497013092, 0.3368530869483948, 0.1938154697418213, -0.06877834349870682, -0.014124344103038311, 0.3971104323863983, 0.4357682168483734, 0.3383936285972595, -0.36422058939933777, 0.08116783946752548, -0.2033807337284088, -0.056821923702955246, -0.023260261863470078, 0.2695125639438629, 0.10607320815324783, 0.11737607419490814, 0.462422639131546, 0.12148632109165192, -0.124727264046669, 0.04363897442817688, -0.08805680274963379, -0.26053252816200256, -0.262947142124176, 0.5191485285758972, 0.18579453229904175, -0.11438801884651184, -0.32065099477767944, 0.18993781507015228, -0.27498695254325867, -0.08091452717781067, 0.06285151094198227, 0.24058954417705536, 0.11315396428108215, -0.21382759511470795, 0.0705089196562767, 0.0003152035642415285, 0.7061733603477478, 0.4877219796180725, -0.01771940104663372, -0.3658052682876587, -0.11652376502752304, -0.5989871025085449, 0.2178361713886261, -0.3639681041240692, -0.2690182328224182, 0.2833365499973297, -0.0723314955830574, -0.2880823016166687, 0.24852751195430756, 0.12761974334716797, 0.15819577872753143, -0.05727637931704521, 0.09834792464971542, -0.4907650947570801, -0.14860443770885468, -0.03065044991672039, -0.007675084285438061, -0.08318091183900833, -0.257053017616272, 0.14532335102558136, -0.43499451875686646, 0.11610305309295654, 0.011985060758888721, 0.06934806704521179, -0.0002395449992036447, 0.09664040058851242, 0.3544246256351471, 0.16532395780086517, 0.5012481808662415, -0.03556109592318535, 0.0036573235411196947, -0.19115452468395233, -0.1566217541694641, -0.22899232804775238, 0.1318226009607315, 0.08762913942337036, 0.5133118629455566, -0.026107510551810265, -0.06854528188705444, -0.2640271484851837, 0.4031679034233093, -0.08100655674934387, 0.08831585943698883, -0.13508093357086182, 0.07731451839208603, -0.07034944742918015, 0.22758422791957855, 0.26436492800712585, 0.42454031109809875, -0.1783217340707779, 0.20220941305160522, -0.27229201793670654, -0.46024590730667114, 0.32893550395965576, -0.661480724811554, -0.5177905559539795, 0.0751223936676979, 0.1161927729845047, -0.14233405888080597, 0.058194272220134735, -0.6006340980529785, 0.07056694477796555, 0.24261268973350525, -0.11140575259923935, -0.3171198070049286, 0.21794509887695312, 0.01654207892715931, 0.08000141382217407, 0.04357372596859932, 0.44835659861564636, -0.14847296476364136, -0.1597675234079361, -0.0049345833249390125, -0.16325202584266663 ]
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.
[ -0.11933844536542892, 0.08862767368555069, -0.008977877907454967, 0.21019725501537323, 0.3692678213119507, 0.026059893891215324, 0.44878166913986206, 0.41310209035873413, 0.4330812692642212, 0.01256692036986351, 0.1238437220454216, 0.39542990922927856, -0.15054894983768463, 0.021962737664580345, 0.12090493738651276, -0.4289896786212921, -0.03373947739601135, 0.36367297172546387, -0.04146667942404747, 0.06297369301319122, -0.18768362700939178, 0.10851698368787766, -0.2651914656162262, 0.25500598549842834, -0.3057228922843933, -0.057025808840990067, 0.21592070162296295, 0.08206350356340408, -0.07891874760389328, -0.5590065121650696, 0.27708321809768677, -0.21704912185668945, 0.0644516572356224, 0.11474668979644775, -0.00010995571938110515, 0.20530638098716736, 0.33269718289375305, 0.030212869867682457, -0.5799456834793091, -0.18957771360874176, -0.47733166813850403, -0.11893873661756516, 0.2974317669868469, -0.24970759451389313, -0.028513627126812935, 0.05389740690588951, 0.03591078519821167, -0.07441216707229614, 0.2967468202114105, 0.531506359577179, 0.23788897693157196, 0.5236923098564148, 0.15454918146133423, 0.03449370339512825, -0.07856779545545578, -0.17057092487812042, -0.08211507648229599, 0.22396805882453918, 0.07387740164995193, -0.3915436565876007, -0.013373903930187225, 0.12692449986934662, 0.15117241442203522, 0.263701856136322, 0.5083708763122559, 0.01439398992806673, 0.24052372574806213, -0.15062718093395233, 0.014947105199098587, 0.07640064507722855, 0.3353829085826874, -0.1649331897497177, -0.2715999186038971, -0.2243589609861374, 0.24766606092453003, -0.33475616574287415, 0.23653045296669006, -0.036676667630672455, -0.16220439970493317, -0.00819715391844511, -0.2692277729511261, -0.02814740687608719, -0.24182094633579254, 0.35268712043762207, -0.0582139827311039, 0.4996492266654968, 0.04358334839344025, 0.14274729788303375, -0.05806553736329079, -0.07545780390501022, 0.1403771936893463, -0.0870911106467247, -0.12930051982402802, 0.25781315565109253, -0.38315534591674805, -0.036467500030994415, -0.03494661673903465, -0.11697590351104736, -0.03162477910518646, 0.263520210981369, 0.2199898362159729, -0.2248734086751938, 0.32383885979652405, 0.3154154121875763, 0.2229883372783661, 0.42046889662742615, 0.0727653056383133, 0.06359227746725082, -0.08280493319034576, 0.13602614402770996, -0.23372821509838104, -0.15548528730869293, -0.15553614497184753, -0.17655576765537262, 0.09289789199829102, -0.15460222959518433, 0.22602057456970215, -0.07111575454473495, -0.49847379326820374, 0.11472073942422867, -0.14340490102767944, 0.04629025608301163, 0.2051147073507309, 0.17948497831821442, -0.12763401865959167, 0.23223420977592468, 0.14412450790405273, 0.28256696462631226, -0.2519770562648773, -0.043135613203048706, -0.1662895381450653, 0.2492133229970932, -0.23438303172588348, -0.18224912881851196, 0.2427540421485901, 0.193400040268898, 0.3994782567024231, -0.10116763412952423, 0.11126401275396347, -0.20175547897815704, 0.22737567126750946, -0.04551948234438896, -0.1288793981075287, 0.2566531002521515, -0.21606101095676422, 0.15489795804023743, 0.20159123837947845, -0.2824235260486603, -0.08482388406991959, -0.10174447298049927, -0.27263563871383667, -0.30809205770492554, -0.12625139951705933, 0.28141555190086365, 0.14738212525844574, -0.24296484887599945, -0.15226320922374725, -0.06118768826127052, 0.14749260246753693, -0.36872363090515137, -0.11810611188411713, -0.46125850081443787, -0.2206392139196396, -0.023005584254860878, 0.3976435661315918, 0.20490935444831848, 0.13138674199581146, -0.10897259414196014, -0.11542075127363205, 0.07785411924123764, -0.03179730847477913, 0.1227133721113205, -0.438356876373291, 0.542163074016571, -0.061055924743413925, 0.118863046169281, 0.64134681224823, -0.3427872359752655, -0.45785459876060486, 0.12356872856616974, -0.01428011991083622, 0.209327831864357, -0.225368931889534, -0.016720708459615707, -0.013226387090981007, -0.18455132842063904, 0.2123136669397354, 0.6192132234573364, -0.015042993240058422, 0.0824793204665184, -0.12689702212810516, -0.015549133531749249, 0.500219464302063, 0.10985691845417023, 0.09252550452947617, -0.007928984239697456, -0.04170479252934456, 0.4966483414173126, -0.0013131386367604136, -0.12597055733203888, -0.014860047958791256, -0.013630194589495659, -0.06468609720468521, -0.10460405051708221, -0.21782632172107697, -0.07388683408498764, -0.27766451239585876, 0.3043040931224823, -0.11754775047302246, -0.058531954884529114, -0.04715004563331604, 0.1061866357922554, -0.3491132855415344, 0.16848868131637573, -0.4232693016529083, -0.37731894850730896, 0.17209340631961823, 0.15513364970684052, -0.010822752490639687, 0.23844130337238312, -0.20474109053611755, 0.219436913728714, -0.27402758598327637, 0.05550083518028259, -0.21892865002155304, 0.16128963232040405, -0.05588398873806, -0.042422182857990265, -0.0094792814925313, 0.3087964355945587, 0.038331370800733566, -0.01799662411212921, -0.3062887191772461, 0.28603023290634155, -0.12548455595970154, 0.08035627007484436, -0.03226484730839729, -0.14337030053138733, 0.06880644708871841, -0.07154319435358047, -0.17518453299999237, 0.20819184184074402, 0.331950843334198, -0.18012624979019165, -0.04314278066158295, 0.11287608742713928, 0.1272120177745819, 0.08142179995775223, 0.09886399656534195, 0.16397236287593842, 0.14399196207523346, -0.06582631915807724, 0.1012796014547348, 0.07743120193481445, 0.23370946943759918, -0.09498720616102219, 0.20471063256263733, -0.15645115077495575, -0.1671450287103653, -0.11880644410848618, 0.3131200075149536, 0.17005865275859833, 0.16741715371608734, 0.08148640394210815, -0.2752522826194763, -0.09487438946962357, 0.07855788618326187, 0.14604908227920532, 0.40096521377563477, -0.05183282122015953, -0.10834052413702011, -0.052570827305316925, -0.416596919298172, 0.0711255818605423, 0.1707078218460083, 0.03662902116775513, 0.20912893116474152, 0.30185675621032715, 0.041315339505672455, 0.08732044696807861, -0.29526710510253906, -0.35701707005500793, 0.28243377804756165, 0.440530002117157, -0.22848205268383026, 0.20788012444972992, -0.03053182363510132, -0.0025113264564424753, -0.06784813851118088, -0.33363381028175354, -0.18862831592559814, -0.34299373626708984, -0.12133599072694778, 0.4337753653526306, 0.024119378998875618, 0.1810045689344406, -0.39776164293289185, -0.022050954401493073, 0.19475102424621582, -0.11478555202484131, -0.031379494816064835, -0.31579792499542236, -0.26932892203330994, 0.045942142605781555, 0.3818606734275818, 0.09107666462659836, 0.35109788179397583, -0.06677065789699554, -0.09253539890050888, -0.0070348638109862804, -0.05841507017612457, -0.0755671039223671, -0.18372094631195068, 0.24404765665531158, 0.009521503932774067, 0.28558358550071716, -0.17702975869178772, -0.5220332145690918, 0.5017468929290771, -0.2557017505168915, -0.05778532475233078, 0.3151065707206726, 0.2711130976676941, -0.24919050931930542, -0.07947684079408646, -0.37468844652175903, -0.1264503300189972, -0.2806057631969452, -0.05946894735097885, 0.07482901960611343, 0.06202087178826332, 0.3974674344062805, 0.07085659354925156, 0.2160108983516693, -0.0025738144759088755, 0.11343798786401749, -0.048262786120176315, -0.09006398916244507, 0.18598473072052002, -0.31003379821777344, -0.3173184096813202, -0.12134277820587158, 0.009935948997735977, 0.3347616493701935, 0.11073566973209381, -0.3489592671394348, -0.21413171291351318, -0.10495787858963013, 0.08531288057565689, -0.2213253378868103, -0.04006680101156235, 0.18286007642745972, 0.0723033994436264, -0.06572244316339493, -0.0866440087556839, 0.035989612340927124, 0.07124002277851105, 0.15234600007534027, 0.13496200740337372, -0.0063046677969396114, 0.0892268493771553, -0.08488413691520691, 0.6008990406990051, 0.2186937928199768, -0.21437722444534302, 0.30028611421585083, -0.10327658802270889, 0.005144711118191481, -0.20868366956710815, -0.4398044943809509, 0.2641642689704895, -0.11523295938968658, 0.18292878568172455, 0.17231491208076477, -0.07949358969926834, -0.2247839868068695, -0.19255922734737396, 0.06885148584842682, -0.17844754457473755, -0.2727796733379364, 0.14616838097572327, -0.2173946499824524, 0.27871108055114746, -0.0688643679022789, 0.22253717482089996, -0.20522312819957733, -0.16798074543476105, 0.06985094398260117, 0.16693419218063354, -0.0905456691980362, 0.07967526465654373, -0.24400167167186737, -0.27610859274864197, -0.47995808720588684, -0.027839848771691322, 0.2410317063331604, 0.6437501311302185, -0.08638478815555573, 0.03718903288245201, 0.08072245866060257, 0.08174151927232742, 0.13737379014492035, -0.2301362156867981, 0.18226361274719238, 0.28913068771362305, 0.031824640929698944, -0.3197464346885681, -0.06433165818452835, -0.14011214673519135, 0.3580164909362793, 0.07665183395147324, -0.05443182960152626, -0.037231311202049255, -0.16765408217906952, 0.1398317664861679, 0.2599390745162964, -0.18729500472545624, -0.04411308094859123, -0.0917343720793724, -0.08981035649776459, -0.30776476860046387, -0.00615481473505497, -0.00994766503572464, 0.33566054701805115, 0.12296929955482483, 0.26572492718696594, -0.5462992191314697, 0.11665716022253036, 0.09016522765159607, 0.1745867282152176, 0.0959530621767044, 0.05794429033994675, 0.18724678456783295, -0.16663391888141632, 0.18320998549461365, -0.1459045261144638, 0.563380241394043, 0.14737606048583984, -0.33079883456230164, -0.31952935457229614, 0.021663231775164604, 0.1674334853887558, 0.1839803010225296, -0.018585484474897385, -0.14112089574337006, -0.2769453227519989, -0.061463501304388046, -0.008708842098712921, 0.3045703172683716, 0.5610411167144775, 0.002627668436616659, -0.3271808326244354, -0.2585405707359314, 0.48283934593200684, -0.2850031554698944, 0.11732912063598633, 0.3215573728084564, 0.05337723344564438, -0.31319230794906616, 0.08413907885551453, 0.07783256471157074, 0.6767755746841431, 0.16953879594802856, 0.3111465871334076, 0.3826707899570465, 0.18392826616764069, 0.41427063941955566, -0.043089501559734344, 0.2847144305706024, -0.3927355706691742, -0.13342593610286713, -0.08068574965000153, -0.06546270102262497, 0.10708170384168625, -0.008873440325260162, -0.025438690558075905, 0.15062108635902405, 0.15688933432102203, -0.004141780082136393, 0.057147782295942307, 0.4492528736591339, -0.03878398612141609, -0.19681917130947113, -0.3470250964164734, 0.15601642429828644, -0.21799980103969574, 0.2975102961063385, -0.09073382616043091, -0.06648409366607666, 0.05520455539226532, -0.06735203415155411, -0.25555863976478577, 0.2574385702610016, -0.29947468638420105, 0.4594579339027405, 0.05782169848680496, -0.475681334733963, 0.19841517508029938, 0.20709797739982605, 0.17566406726837158, 0.005698727443814278, -0.2995942533016205, 0.23877884447574615, -0.29008978605270386, -0.2218664139509201, -0.06728777289390564, 0.11194716393947601, 0.42007339000701904, -0.1725514531135559, -0.14637790620326996, 0.14794056117534637, -0.12065393477678299, -0.24150243401527405, 0.10395880788564682, 0.036866508424282074, 0.35248932242393494, -0.39257583022117615, -0.5152195692062378, -0.002420816570520401, -0.18207627534866333, -0.17491012811660767, 0.1745000183582306, 0.023450618609786034, -0.18903139233589172, 0.15189562737941742, 0.1498272567987442, -0.19897985458374023, 0.11640633642673492, 0.24804235994815826, -0.2495405375957489, 0.1462831199169159, 0.7637571096420288, 0.03394559025764465, 0.012805139645934105, -0.3138304054737091, -0.054288268089294434, 0.12508445978164673, -0.48199984431266785, -0.11712734401226044, 0.12550395727157593, 0.16912569105625153, -0.1442469209432602, 0.4983660876750946, 0.06302104145288467, -0.018844809383153915, 0.24718545377254486, -0.5424140691757202, -0.10621922463178635, -0.04302157834172249, 0.017753537744283676, 0.23777537047863007, -0.002624211832880974, 0.11488901823759079, 0.13550961017608643, 0.10338855534791946, -0.27263742685317993, 0.014254532754421234, -0.15762585401535034, 0.13799649477005005, 0.21783460676670074, -0.1565314084291458, 0.2232816368341446, -0.0769471824169159, 0.15638616681098938, -0.09584657102823257, -0.12303436547517776, -0.24829335510730743, -0.2548377513885498, 0.10011143982410431, 0.24556712806224823, -0.1645212471485138, -0.03067934699356556, -0.1023336872458458, 0.023677632212638855, -0.2311049848794937, -0.00846262276172638, -0.1177278533577919, -0.2933426797389984, 0.38943222165107727, 0.04389796033501625, 0.2783687710762024, -0.11169014126062393, 0.1263556182384491, -0.3007573187351227, 0.14912062883377075, -0.01849590614438057, -0.07522980868816376, 0.27581319212913513, 0.09365329146385193, -0.46588683128356934, 0.005420649889856577, -0.40260517597198486, 0.03714754059910774, 0.36982133984565735, -0.4551105201244354, 0.15995842218399048, -0.1316281408071518, 0.09546535462141037, 0.46568307280540466, -0.2714085280895233, -0.0036669739056378603, 0.3496656119823456, 0.2378426045179367, -0.297417551279068, -0.18926052749156952, 0.2399909794330597, 0.061345525085926056, 0.03245915472507477, 0.027238842099905014, 0.04666132479906082, -0.30374857783317566, -0.3212345540523529, 0.08927971869707108, -0.029340069741010666, -0.2656012177467346, -0.011407874524593353, 0.33932632207870483, -0.002364365616813302, -0.059514716267585754, -0.050049375742673874, 0.21102096140384674, 0.03539273142814636, 0.38269317150115967, -0.2926008105278015, 0.2069980949163437, 0.05926349014043808, 0.04391063377261162, -0.11114291101694107, -0.2192171812057495, 0.05030057206749916, -0.15784087777137756, 0.03383427858352661, 0.295282781124115, 0.09981466829776764, 0.15308073163032532, -0.2855035066604614, -0.12453679740428925, -0.20802298188209534, 0.22792452573776245, -0.1790422797203064, -0.048906464129686356, -0.24299554526805878, 0.021017009392380714, -0.025876479223370552, -0.09469880163669586, -0.14167028665542603, -0.03817872330546379, 0.3095560371875763, 0.2532220184803009, -0.1722707748413086, -0.4555346965789795, 0.10826654732227325, 0.34329426288604736, 0.1800471842288971, -0.24588419497013092, 0.3368530869483948, 0.1938154697418213, -0.06877834349870682, -0.014124344103038311, 0.3971104323863983, 0.4357682168483734, 0.3383936285972595, -0.36422058939933777, 0.08116783946752548, -0.2033807337284088, -0.056821923702955246, -0.023260261863470078, 0.2695125639438629, 0.10607320815324783, 0.11737607419490814, 0.462422639131546, 0.12148632109165192, -0.124727264046669, 0.04363897442817688, -0.08805680274963379, -0.26053252816200256, -0.262947142124176, 0.5191485285758972, 0.18579453229904175, -0.11438801884651184, -0.32065099477767944, 0.18993781507015228, -0.27498695254325867, -0.08091452717781067, 0.06285151094198227, 0.24058954417705536, 0.11315396428108215, -0.21382759511470795, 0.0705089196562767, 0.0003152035642415285, 0.7061733603477478, 0.4877219796180725, -0.01771940104663372, -0.3658052682876587, -0.11652376502752304, -0.5989871025085449, 0.2178361713886261, -0.3639681041240692, -0.2690182328224182, 0.2833365499973297, -0.0723314955830574, -0.2880823016166687, 0.24852751195430756, 0.12761974334716797, 0.15819577872753143, -0.05727637931704521, 0.09834792464971542, -0.4907650947570801, -0.14860443770885468, -0.03065044991672039, -0.007675084285438061, -0.08318091183900833, -0.257053017616272, 0.14532335102558136, -0.43499451875686646, 0.11610305309295654, 0.011985060758888721, 0.06934806704521179, -0.0002395449992036447, 0.09664040058851242, 0.3544246256351471, 0.16532395780086517, 0.5012481808662415, -0.03556109592318535, 0.0036573235411196947, -0.19115452468395233, -0.1566217541694641, -0.22899232804775238, 0.1318226009607315, 0.08762913942337036, 0.5133118629455566, -0.026107510551810265, -0.06854528188705444, -0.2640271484851837, 0.4031679034233093, -0.08100655674934387, 0.08831585943698883, -0.13508093357086182, 0.07731451839208603, -0.07034944742918015, 0.22758422791957855, 0.26436492800712585, 0.42454031109809875, -0.1783217340707779, 0.20220941305160522, -0.27229201793670654, -0.46024590730667114, 0.32893550395965576, -0.661480724811554, -0.5177905559539795, 0.0751223936676979, 0.1161927729845047, -0.14233405888080597, 0.058194272220134735, -0.6006340980529785, 0.07056694477796555, 0.24261268973350525, -0.11140575259923935, -0.3171198070049286, 0.21794509887695312, 0.01654207892715931, 0.08000141382217407, 0.04357372596859932, 0.44835659861564636, -0.14847296476364136, -0.1597675234079361, -0.0049345833249390125, -0.16325202584266663 ]
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)
[ 0.1534196436405182, -0.037310294806957245, -0.06269726157188416, -0.12906987965106964, 0.1771615743637085, 0.08376257121562958, 0.2535591721534729, 0.23048622906208038, -0.21282312273979187, 0.3302721679210663, -0.07075786590576172, 0.7361390590667725, -0.2225145399570465, -0.2018117755651474, 0.023780278861522675, -0.0824756920337677, 0.24421846866607666, -0.21178558468818665, -0.16500338912010193, 0.02707301825284958, -0.11413010209798813, -0.38973110914230347, 0.02324465662240982, -0.14219339191913605, -0.19263310730457306, -0.04983311891555786, -0.13309884071350098, 0.047407105565071106, -0.24616175889968872, -0.2544477581977844, -0.4673667550086975, 0.7014351487159729, -0.15274126827716827, 0.44993725419044495, -0.00011275450378889218, -0.29292967915534973, 0.35525014996528625, 0.0323522575199604, -0.31195271015167236, 0.0747515931725502, -0.17222611606121063, -0.07203914225101471, 0.06607130169868469, -0.20177561044692993, 0.21404653787612915, 0.05777467414736748, 0.05699164420366287, -0.1298426240682602, 0.1646692156791687, -0.08943397551774979, 0.174724742770195, -0.09979445487260818, -0.3013957142829895, -0.0162506140768528, 0.5358917117118835, 0.5136167407035828, -0.15510813891887665, 0.11450318992137909, 0.5382506847381592, 0.3613970875740051, -0.18806718289852142, 0.15618129074573517, -0.0936751589179039, -0.07671938091516495, 0.48861485719680786, -0.29398173093795776, -0.5896618366241455, 0.04415915533900261, 0.12085754424333572, 0.1790134310722351, 0.6000990867614746, -0.2012621909379959, -0.41691854596138, -0.24042871594429016, 0.2678793966770172, -0.13337334990501404, 0.05386362224817276, -0.1913708597421646, -0.012277720496058464, -0.10944864153862, -0.13056567311286926, -0.1574682742357254, -0.005668409168720245, -0.058630067855119705, 0.49773091077804565, 0.5297004580497742, 0.12314765155315399, -0.019478395581245422, 0.11494004726409912, 0.0041554830968379974, 0.5591570734977722, -0.10236572474241257, 0.014495600014925003, 0.26156699657440186, -0.1102365255355835, -0.10021761804819107, -0.007673816289752722, 0.2563237249851227, 0.24947935342788696, 0.1132013350725174, 0.3054131269454956, 0.27013736963272095, -0.06867580115795135, 0.07461698353290558, -0.0694345086812973, -0.07609112560749054, 0.03268741816282272, -0.04483672231435776, 0.2888603210449219, -0.31532344222068787, 0.17996221780776978, -0.11023568361997604, -0.1513274908065796, -0.0501810684800148, 0.03848661482334137, 0.1283472627401352, -0.30147019028663635, -0.10199972242116928, 0.19806787371635437, -0.5291100740432739, 0.06318321079015732, -0.42724302411079407, 0.07826333492994308, 0.11914832890033722, 0.29036226868629456, 0.11539529263973236, -0.15124545991420746, 0.04944974556565285, 0.2924681007862091, -0.12642455101013184, -0.06676327437162399, 0.032525837421417236, -0.47437044978141785, 0.23198024928569794, 0.16124898195266724, -0.21715691685676575, 0.10984338819980621, 0.08951110392808914, 0.1500674933195114, 0.22949743270874023, 0.16365492343902588, 0.21711896359920502, 0.3288634419441223, -0.1775379329919815, -0.44875970482826233, -0.06935009360313416, -0.09622688591480255, 0.2949223518371582, -0.2965109050273895, 0.2620726227760315, -0.2736847698688507, -0.262124627828598, -0.06397727131843567, 0.12658850848674774, -0.025156322866678238, -0.10126537084579468, -0.23621203005313873, 0.20045344531536102, 0.1188172996044159, -0.3033382296562195, 0.3808518052101135, 0.1356823891401291, 0.1871735155582428, -0.10355512797832489, -0.030979732051491737, 0.2598656117916107, -0.09520698338747025, -0.020418601110577583, -0.338922381401062, -0.10294434428215027, 0.18109719455242157, 0.1123063936829567, -0.16153813898563385, 0.023971397429704666, -0.1227991133928299, -0.13101047277450562, 0.48198848962783813, 0.21702714264392853, -0.17160682380199432, -0.03480594605207443, -0.153493270277977, -0.16826403141021729, 0.5057480335235596, 0.3218178153038025, -0.07960861921310425, -0.25737959146499634, 0.28341493010520935, -0.09259998053312302, -0.41317111253738403, 0.03711938112974167, -0.23320914804935455, -0.13467738032341003, 0.38062310218811035, 0.23806463181972504, -0.09189716726541519, 0.0011277630692347884, 0.04510458558797836, -0.40925657749176025, 0.18579992651939392, -0.13240306079387665, -0.1216469407081604, -0.25277477502822876, -0.029326602816581726, 0.33917129039764404, -0.08386571705341339, -0.0944194495677948, -0.03489309176802635, -0.1526419222354889, 0.2866229712963104, -0.10843147337436676, 0.02840796299278736, -0.29122596979141235, 0.038068342953920364, -0.39275676012039185, -0.00456266850233078, -0.12355287373065948, 0.03388375788927078, 0.24871130287647247, -0.24041321873664856, -0.23405373096466064, -0.28614431619644165, -0.19298115372657776, -0.16904868185520172, -0.12849797308444977, -0.22818179428577423, -0.05662833899259567, -0.04576114937663078, -0.046052444726228714, -0.2763000726699829, 0.20862224698066711, -0.030198518186807632, -0.0818404033780098, -0.17129163444042206, 0.2613539695739746, 0.08897014707326889, -0.13293905556201935, 0.43658530712127686, 0.3457507789134979, -0.06262784451246262, 0.09248320013284683, 0.18023385107517242, -0.05637481436133385, -0.016703611239790916, -0.09874476492404938, -0.29944729804992676, 0.26037970185279846, -0.26657676696777344, 0.09091317653656006, -0.262186199426651, -0.3188469111919403, -0.047044847160577774, 0.37200790643692017, -0.2649485468864441, 0.16979554295539856, 0.11686816811561584, 0.20927320420742035, 0.35136842727661133, 0.1626695990562439, -0.3234310448169708, -0.13951793313026428, 0.264970064163208, -0.15998364984989166, -0.022034326568245888, 0.34199294447898865, 0.2032804787158966, -0.24940238893032074, 0.14528802037239075, 0.0685134083032608, 0.37311044335365295, 0.340804785490036, 0.012890176847577095, 0.07192332297563553, 0.2797624468803406, -0.1501164734363556, -0.048269402235746384, 0.16651776432991028, -0.2826218008995056, -0.011177689768373966, 0.1748599261045456, 0.12027431279420853, -0.15232916176319122, -0.3830469846725464, 0.07604912668466568, 0.032206904143095016, -0.03617879003286362, -0.07769857347011566, -0.018525078892707825, -0.07378366589546204, -0.20089362561702728, -0.19713100790977478, -0.33908456563949585, -0.26242780685424805, 0.26744765043258667, 0.058490872383117676, -0.2952107787132263, 0.3880602717399597, 0.18326818943023682, 0.5487815141677856, -0.2152481973171234, -0.10795985907316208, -0.35612183809280396, -0.44117629528045654, 0.12635700404644012, 0.04095584899187088, 0.22586026787757874, -0.13597220182418823, 0.7025216221809387, -0.05100472271442413, -0.07998783886432648, -0.38227954506874084, -0.46983015537261963, -0.08981521427631378, -0.007440574467182159, -0.2571098506450653, 0.3828088939189911, -0.3270595073699951, 0.08849538862705231, -0.4728187024593353, -0.06896984577178955, -0.2552247941493988, -0.010260583832859993, 0.007352877873927355, -0.010249421931803226, 0.00927678681910038, -0.18420103192329407, -0.29757729172706604, -0.13124607503414154, -0.4049797058105469, 0.07406428456306458, -0.631806492805481, -0.050029654055833817, 0.012037144042551517, -0.2868337631225586, -0.1324148029088974, -0.2205686718225479, 0.16137073934078217, -0.1366974264383316, -0.7596741318702698, 0.19291165471076965, -0.2022344022989273, 0.05757437273859978, -0.14220517873764038, -0.0697113573551178, 0.02889319695532322, 0.6446987986564636, 0.275482177734375, -0.16692982614040375, 0.14853490889072418, 0.20165486633777618, -0.0016146752750501037, 0.15614508092403412, 0.3470439612865448, 0.2847762107849121, -0.12292200326919556, 0.062341466546058655, -0.23541158437728882, -0.11293711513280869, -0.1072700023651123, 0.3154539465904236, 0.10719224810600281, 0.32110080122947693, 0.01664717122912407, 0.6867967844009399, -0.6602863669395447, -0.09058082103729248, 0.01934928074479103, 0.2002180516719818, 0.13724668323993683, -0.06796803325414658, -0.09096987545490265, 0.5424607396125793, 0.2456182986497879, -0.3071572780609131, 0.19216026365756989, -0.04839327558875084, -0.16012324392795563, 0.03776286914944649, 0.1311073899269104, -0.14749866724014282, -0.14253060519695282, 0.34279873967170715, -0.32730892300605774, 0.08625153452157974, -0.04006391763687134, 0.17369787395000458, -0.2951992154121399, 0.00426055770367384, 0.14684416353702545, 0.20264242589473724, 0.3288085162639618, -0.2704310417175293, -0.4360995590686798, -0.021179836243391037, -0.1265939474105835, 0.2958835959434509, 0.24746225774288177, 0.12883856892585754, -0.013604686595499516, -0.181803897023201, 0.12236522883176804, 0.25762295722961426, 0.4825330376625061, -0.24698933959007263, -0.2833612561225891, -0.005569362547248602, -0.016897061839699745, -0.11249478906393051, 0.015213693492114544, -0.07555755972862244, 0.1418469399213791, 0.0728941485285759, 0.09383472055196762, -0.22726120054721832, -0.17683008313179016, 0.11363405734300613, 0.19612278044223785, 0.1610931158065796, -0.33927232027053833, -0.08296406269073486, -0.16535842418670654, -0.1599021703004837, 0.1925913393497467, -0.02379539981484413, -0.30083733797073364, -0.055608440190553665, 0.058830611407756805, -0.033482130616903305, 0.3097730278968811, 0.026684368029236794, 0.12922775745391846, 0.03208724781870842, 0.12182526290416718, 0.044979412108659744, 0.01223849132657051, 0.37701913714408875, 0.42766544222831726, -0.20563451945781708, -0.48697635531425476, -0.1709444522857666, 0.014514414593577385, -0.5441598296165466, 0.24406971037387848, 0.2382853478193283, 0.2514321506023407, 0.29675233364105225, -0.14040762186050415, 0.17606574296951294, -0.2185063660144806, 0.07353869825601578, 0.052437473088502884, 0.007403991185128689, -0.5293941497802734, -0.9117096662521362, 0.5416221022605896, 0.3771866261959076, -0.322413831949234, 0.5554150938987732, 0.0005214220727793872, -0.26848965883255005, 0.41509297490119934, 0.2951991558074951, 1.2055479288101196, -0.08146850764751434, 0.1874581128358841, -0.02921690046787262, -0.376735121011734, 0.27660730481147766, -0.18090760707855225, 0.1493992805480957, -0.12933924794197083, -0.39364680647850037, 0.05398126691579819, -0.08897188305854797, 0.07474077492952347, 0.32798054814338684, -0.12489917129278183, 0.34531211853027344, -0.39326322078704834, 0.17307862639427185, 0.14458321034908295, 0.12142487615346909, -0.24394607543945312, -0.29141974449157715, -0.08227936178445816, 0.07282456010580063, -0.09159131348133087, -0.21026615798473358, -0.03348644822835922, -0.005069761071354151, -0.06700877845287323, 0.14088627696037292, -0.14437372982501984, -0.3057858347892761, 0.1522461473941803, -0.08004185557365417, -0.2814941704273224, -0.17594857513904572, 0.44333234429359436, 0.10858995467424393, 0.25358444452285767, 0.07985024899244308, 0.1667894572019577, 0.2920089662075043, 0.08756522089242935, 0.40555599331855774, -0.0955696627497673, -0.08928129822015762, 0.10746950656175613, 0.08212120085954666, -0.04872274026274681, 0.030048683285713196, 0.11088289320468903, -0.42860135436058044, -0.5298832654953003, -0.08485539257526398, 0.5617536902427673, -0.1269344538450241, 0.09261023998260498, 0.0187522042542696, -0.10598745942115784, -0.017499813809990883, 0.13704358041286469, 0.2787858545780182, -0.25865626335144043, 0.3623969852924347, -0.21821704506874084, 0.019990378990769386, -0.12006469070911407, 0.1049022525548935, -0.2883317172527313, -0.12774550914764404, 0.2179226279258728, -0.21949724853038788, -0.19734573364257812, -0.16577619314193726, 0.11748955398797989, 0.23314836621284485, -0.004313134588301182, 0.03345126658678055, -0.13740456104278564, -0.27819105982780457, -0.15487922728061676, -0.15670910477638245, 0.21951746940612793, 0.3490118682384491, -0.035353828221559525, 0.1103745698928833, -0.46019044518470764, 0.07673443853855133, 0.17053312063217163, 0.3398688733577728, -0.16756445169448853, -0.04664922505617142, -0.11511196941137314, -0.01923384889960289, -0.32862111926078796, -0.1494593620300293, 0.22030885517597198, 0.27189549803733826, 0.009397712536156178, 0.22101430594921112, -0.2014082819223404, -0.2405494898557663, 0.06552567332983017, 0.061104822903871536, -0.018643183633685112, -0.14807040989398956, -0.13654440641403198, 0.11171082407236099, -0.07000512629747391, 0.22942937910556793, 0.2037571221590042, -0.08181674033403397, -0.002820574911311269, -0.2503928244113922, 0.02648240514099598, 0.29474928975105286, 0.005087670870125294, 0.49682706594467163, 0.6184062361717224, 0.0818525105714798, -0.2173994928598404, 0.20662231743335724, -0.025502333417534828, 0.03144010156393051, 0.12063341587781906, 0.14772605895996094, -0.026562416926026344, -0.02038283832371235, 0.11907141655683517, 0.3216322064399719, 0.18480002880096436, 0.06783591955900192, -0.09052724391222, 0.2885022461414337, 0.19308574497699738, 0.4536781907081604, 0.09952207654714584, 0.4589994251728058, 0.0012196055613458157, -0.3575807511806488, 0.43149563670158386, 0.2065580040216446, -0.019029108807444572, -0.07362876087427139, -0.25354039669036865, -0.005759481806308031, 0.08054948598146439, -0.005397672299295664, -0.22964957356452942, 0.015039931051433086, -0.34495019912719727, -0.07759952545166016, 0.14512382447719574, -0.4788009226322174, 0.39619389176368713, 0.33247825503349304, -0.12170533835887909, -0.02795552648603916, 0.37087926268577576, 0.00944504700601101, -0.26959097385406494, 0.13246004283428192, 0.46284475922584534, 0.9553410410881042, 0.3267764747142792, 0.16063396632671356, 0.13742756843566895, 0.1492273509502411, 0.1605958491563797, 0.6927056312561035, 0.34117892384529114, 0.08903806656599045, 0.3850887715816498, 0.22731857001781464, 0.1477174460887909, -0.34027421474456787, 0.27715685963630676, 0.07764604687690735, 0.2560734450817108, 0.0304997768253088, -0.06635613739490509, -0.19421155750751495, -0.2519179880619049, -0.0834772139787674, 0.012616975232958794, 0.05153520032763481, 0.5799617171287537, 0.23499727249145508, 0.2591431140899658, -0.352833092212677, 0.12664292752742767, 0.27182695269584656, 0.44117605686187744, -0.15010906755924225, -0.4055155813694, 0.1420254111289978, -0.1715122014284134, 0.7451753616333008, -0.01977105624973774, -0.0012752799084410071, 0.24583163857460022, 0.029736962169408798, 0.11833067983388901, -0.3746950030326843, -0.03710133954882622, -0.015620497986674309, 0.12326820194721222, 0.14358174800872803, -0.17993003129959106, -0.012090394273400307, 0.17657993733882904, -0.14344371855258942, 0.0751895159482956, -0.07518414407968521, -0.046274520456790924, -0.1835743933916092, 0.34420838952064514, 0.4996166229248047, -0.3088761270046234, 0.10623560100793839, -0.08491872996091843, -0.19567659497261047, -0.152764230966568, 0.04055948555469513, 0.16102305054664612, -0.20532773435115814, 0.09134696424007416, 0.059067413210868835, 0.29307636618614197, 0.28379181027412415, 0.2766534090042114, -0.007268269546329975, -0.14981023967266083, -0.3939852714538574, -0.3894713222980499, -0.13016000390052795, 0.4760110676288605, 0.0036118850111961365, -0.035477686673402786, -0.1831999272108078, 0.03684115782380104, -0.06707921624183655, 0.14567507803440094, -0.12240853905677795, 0.05333220958709717, 0.19794218242168427, -0.35705775022506714, -0.12122122943401337, -0.1313951164484024, -0.022572437301278114, 0.1836550086736679, -0.06100436672568321, -0.30679699778556824, 0.41008293628692627, 0.04413175210356712, -0.30971184372901917, 0.144822895526886, 0.06988425552845001, 0.11397377401590347, 0.21174120903015137, -0.01955895498394966, 0.3880785405635834, -0.0024603435304015875, -0.18212997913360596, 0.026386026293039322, -0.14211656153202057, -0.5724018812179565, 0.1211908832192421, 0.007544351741671562, -0.011270610615611076, -0.03836646303534508, -0.20877577364444733, -0.4421340525150299, 0.5485701560974121, -0.35532131791114807, 0.08550751209259033, -0.7277727127075195, 0.15805083513259888, 0.24726340174674988, 0.08188675343990326, -0.04858015477657318, 0.11439969390630722, 0.05100652575492859, 0.2707606554031372, -0.2295469492673874, 0.1943163126707077, 0.5219243168830872, 0.16950620710849762, -0.24273920059204102, -0.3214069902896881, 0.2131146341562271, -0.08961259573698044, 0.029533307999372482, -0.31957292556762695, -0.08377379179000854, 0.2322506606578827, 0.0034356939140707254, 0.07984829694032669, 0.3839523494243622, 0.12845037877559662, -0.04665695130825043, -0.22277206182479858, -0.0014488812303170562, -0.12010572850704193, -0.030343806371092796, -0.1148550882935524, -0.28843310475349426 ]
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.
[ 0.0704682320356369, 0.016060734167695045, -0.08107712119817734, -0.13309316337108612, 0.18675071001052856, -0.01876540668308735, 0.36125361919403076, 0.24943794310092926, -0.1316787153482437, 0.3981238901615143, -0.11517088860273361, 0.7459422945976257, -0.32974615693092346, -0.21496960520744324, -0.04452408105134964, -0.07981587946414948, 0.2653794586658478, -0.10314806550741196, -0.20524297654628754, 0.056223832070827484, -0.12470748275518417, -0.34814876317977905, 0.023674415424466133, -0.14052928984165192, -0.0896880179643631, 0.01218586415052414, -0.11303745955228806, 0.07146775722503662, -0.2733202874660492, -0.2186136692762375, -0.44506892561912537, 0.7157350182533264, -0.10498201102018356, 0.5668811798095703, -0.00010911502613453194, -0.25887182354927063, 0.34147176146507263, 0.03292667120695114, -0.3560403287410736, 0.14131353795528412, -0.13315819203853607, -0.07831971347332001, 0.03868035972118378, -0.2729438245296478, 0.2220698595046997, -0.034398533403873444, 0.11589878797531128, -0.11278674751520157, 0.11099023371934891, -0.14737312495708466, 0.1819334626197815, -0.09936489909887314, -0.384681761264801, 0.0020912110339850187, 0.5785197019577026, 0.46103429794311523, -0.20912007987499237, 0.1824525147676468, 0.5745741724967957, 0.38539284467697144, -0.3338280916213989, 0.19154950976371765, -0.13094328343868256, -0.1019473671913147, 0.5779690146446228, -0.30015552043914795, -0.5396062731742859, 0.06476429849863052, -0.03939170390367508, 0.20368915796279907, 0.647619366645813, -0.19023607671260834, -0.4480343461036682, -0.15105858445167542, 0.1816704124212265, -0.04967699199914932, 0.030789922922849655, -0.24211882054805756, -0.019575783982872963, -0.08532565087080002, -0.03737858310341835, -0.11297054588794708, -0.0025628749281167984, -0.13636156916618347, 0.42179837822914124, 0.5597809553146362, 0.154982328414917, -0.028614362701773643, 0.23609454929828644, -0.022239690646529198, 0.5372040271759033, -0.06967486441135406, 0.07333265244960785, 0.23729822039604187, -0.21732133626937866, -0.18081369996070862, -0.015533479861915112, 0.21171033382415771, 0.18797330558300018, 0.10574378818273544, 0.3026638627052307, 0.27214017510414124, 0.01313999854028225, 0.10956476628780365, -0.11448922008275986, -0.038552526384592056, 0.03346487879753113, -0.1094348132610321, 0.37629517912864685, -0.22859592735767365, 0.1526605486869812, -0.08162547647953033, -0.26237425208091736, -0.017839230597019196, 0.10215314477682114, 0.14140655100345612, -0.41330859065055847, 0.05687246844172478, 0.17915132641792297, -0.4913693368434906, 0.15620656311511993, -0.42619946599006653, 0.10317131876945496, 0.18962305784225464, 0.18701577186584473, 0.13203969597816467, -0.15503764152526855, -0.012790262699127197, 0.21155120432376862, -0.18474088609218597, -0.1022360771894455, 0.04918672516942024, -0.5071455240249634, 0.2194851189851761, 0.10138185322284698, -0.2031903713941574, 0.11597630381584167, 0.03873908147215843, 0.1365172266960144, 0.34394508600234985, 0.15330371260643005, 0.20069310069084167, 0.19247911870479584, -0.255621999502182, -0.4391719400882721, -0.1295640915632248, -0.1443854719400406, 0.31523627042770386, -0.2803957462310791, 0.2423083335161209, -0.2930676341056824, -0.2366228699684143, -0.1138085201382637, 0.17173612117767334, -0.001863837125711143, -0.06199910491704941, -0.22932662069797516, 0.24639543890953064, 0.07400449365377426, -0.3056386113166809, 0.4155254065990448, 0.07851962745189667, 0.17123425006866455, -0.12333603203296661, -0.07254897058010101, 0.2618194818496704, -0.11127021163702011, -0.018436279147863388, -0.23025217652320862, -0.16228799521923065, 0.2056492269039154, 0.11422606557607651, -0.19200211763381958, 0.00789526104927063, -0.14195284247398376, -0.14744573831558228, 0.49070852994918823, 0.1812627613544464, -0.15881510078907013, -0.012354735285043716, -0.1144358366727829, -0.10257548838853836, 0.6056384444236755, 0.34671229124069214, 0.05649440363049507, -0.3341640532016754, 0.2879274785518646, -0.061200838536024094, -0.31021320819854736, 0.09815335273742676, -0.2247583568096161, -0.118413046002388, 0.38146331906318665, 0.2611493170261383, -0.12281814962625504, -0.009176775813102722, 0.005661121569573879, -0.3373619318008423, 0.19471292197704315, -0.17169396579265594, -0.0731256827712059, -0.23634685575962067, -0.08199552446603775, 0.30896881222724915, -0.015054668299853802, -0.0649888813495636, -0.06311707943677902, -0.11575747281312943, 0.2917585074901581, 0.017649617046117783, 0.18580938875675201, -0.28128278255462646, 0.029021013528108597, -0.29896023869514465, -0.015439607203006744, -0.17469657957553864, 0.061628591269254684, 0.26166343688964844, -0.22212819755077362, -0.24753651022911072, -0.2646147310733795, -0.17950399219989777, -0.27028515934944153, -0.16231748461723328, -0.2577844262123108, -0.10035737603902817, -0.04210999608039856, 0.02730180323123932, -0.2138466089963913, 0.22047041356563568, -0.13025030493736267, -0.07318172603845596, -0.15486456453800201, 0.2480490654706955, 0.1106286346912384, 0.0012297723442316055, 0.30390557646751404, 0.3766380250453949, -0.02301429957151413, 0.07872416079044342, 0.3076273500919342, -0.08393976092338562, 0.012866755947470665, -0.11360368132591248, -0.2761875092983246, 0.2542323172092438, -0.2328197956085205, 0.08788249641656876, -0.2713022828102112, -0.3496968746185303, -0.008261432871222496, 0.3596826493740082, -0.33686211705207825, 0.23720195889472961, 0.12871092557907104, 0.27883732318878174, 0.4074002802371979, 0.25084757804870605, -0.3290272653102875, -0.1328156590461731, 0.37403571605682373, -0.12329468876123428, -0.07980753481388092, 0.3697061538696289, 0.12849664688110352, -0.29496222734451294, 0.1104050725698471, 0.08891581743955612, 0.36991679668426514, 0.3470844626426697, 0.07254616171121597, -0.033813539892435074, 0.3311801254749298, -0.1331796497106552, -0.01005434151738882, 0.09691929817199707, -0.13696502149105072, -0.02087963931262493, 0.1016625240445137, 0.05811804533004761, -0.19984079897403717, -0.5052341818809509, 0.03356466814875603, 0.09178835153579712, 0.03976176679134369, -0.07727973163127899, 0.05568976700305939, -0.052659571170806885, -0.1414622813463211, -0.1758357137441635, -0.28012508153915405, -0.27149200439453125, 0.25312912464141846, 0.10748498141765594, -0.3382929563522339, 0.333686888217926, 0.1831435114145279, 0.5316724181175232, -0.15840718150138855, -0.058418549597263336, -0.3735503852367401, -0.4471944570541382, 0.15386372804641724, 0.07673252373933792, 0.22851338982582092, -0.1827123910188675, 0.6215429306030273, -0.035527605563402176, -0.06387956440448761, -0.35137543082237244, -0.3830471634864807, -0.062467653304338455, -0.004240075126290321, -0.22939221560955048, 0.3913436233997345, -0.22182050347328186, 0.1168418899178505, -0.49070876836776733, -0.005510801915079355, -0.3185173571109772, -0.10767776519060135, 0.02714838832616806, 0.025268081575632095, 0.013321153819561005, -0.1668286919593811, -0.24212759733200073, -0.17180240154266357, -0.48437127470970154, 0.03587125241756439, -0.5416005849838257, -0.0025579684879630804, 0.01305357925593853, -0.21188203990459442, -0.10356830805540085, -0.11099418252706528, 0.15742121636867523, -0.15449342131614685, -0.8162327408790588, 0.22808293998241425, -0.18251320719718933, 0.07005283236503601, -0.22892805933952332, -0.11542649567127228, 0.08359474688768387, 0.6077488660812378, 0.2217942476272583, -0.06396787613630295, 0.13452976942062378, 0.13789774477481842, 0.039365436881780624, 0.16842582821846008, 0.26220959424972534, 0.30263397097587585, -0.1816999763250351, 0.02077930048108101, -0.22535938024520874, -0.08990512043237686, -0.12235308438539505, 0.22594264149665833, 0.17040739953517914, 0.3607713580131531, -0.015084944665431976, 0.7392672300338745, -0.6775935292243958, -0.0537550188601017, -0.028204364702105522, 0.1690755933523178, 0.13554613292217255, -0.031218575313687325, -0.14977534115314484, 0.5597664713859558, 0.2193569391965866, -0.30563071370124817, 0.23087021708488464, -0.08712241798639297, -0.24886943399906158, -0.011342556215822697, 0.06998848170042038, -0.04159417375922203, -0.05780646204948425, 0.36301189661026, -0.30870139598846436, 0.013248959556221962, -0.061050549149513245, 0.12251812219619751, -0.326917827129364, 0.009557018987834454, 0.15242913365364075, 0.22981417179107666, 0.2655053436756134, -0.24354392290115356, -0.4098944664001465, 0.04422970116138458, -0.21469581127166748, 0.16246850788593292, 0.22661276161670685, 0.07951013743877411, 0.09746843576431274, -0.2776513397693634, 0.0666254386305809, 0.20896735787391663, 0.501102089881897, -0.12202269583940506, -0.29381856322288513, -0.018135186284780502, -0.03504689037799835, -0.08999644964933395, -0.024985546246170998, -0.11665762960910797, 0.1255786418914795, 0.02066105417907238, 0.09768424928188324, -0.2659021317958832, -0.16656826436519623, 0.07446415722370148, 0.13256391882896423, 0.07663951814174652, -0.3359167277812958, -0.1441166251897812, -0.0007815533317625523, -0.154881551861763, 0.17788617312908173, -0.09627953916788101, -0.28561335802078247, -0.06143779307603836, -0.02194374054670334, 0.036408234387636185, 0.2939795255661011, 0.07353495061397552, 0.21434882283210754, 0.05764332041144371, 0.07084504514932632, 0.04822852462530136, -0.06505811959505081, 0.28172892332077026, 0.43656718730926514, -0.14282438158988953, -0.4651130735874176, -0.10632552951574326, 0.1357995569705963, -0.4808531403541565, 0.24971464276313782, 0.11656617373228073, 0.21496319770812988, 0.1901756376028061, -0.14731937646865845, 0.21787656843662262, -0.2683401107788086, -0.010570209473371506, 0.061114780604839325, 0.02821219339966774, -0.488772988319397, -0.9323241114616394, 0.5213587880134583, 0.3647640347480774, -0.258247971534729, 0.5743233561515808, 0.004262708593159914, -0.20411081612110138, 0.41235747933387756, 0.30180302262306213, 1.1519231796264648, -0.09614540636539459, 0.2185087651014328, 0.10118726640939713, -0.21098670363426208, 0.2955073118209839, -0.22981369495391846, 0.15573902428150177, -0.15916039049625397, -0.3745523989200592, 0.06611084192991257, -0.13276307284832, 0.04989781603217125, 0.26341676712036133, -0.25604748725891113, 0.25369778275489807, -0.3084164261817932, 0.23445798456668854, 0.020436182618141174, 0.19144338369369507, -0.2874628007411957, -0.32861587405204773, -0.05082399770617485, 0.09983666986227036, -0.05817995220422745, -0.3000870645046234, -0.047375936061143875, -0.054685257375240326, -0.09370651841163635, 0.202875554561615, -0.10154698044061661, -0.2596941590309143, 0.1305817812681198, -0.059691477566957474, -0.26577749848365784, -0.09929962456226349, 0.36674028635025024, 0.034364085644483566, 0.1254948079586029, -0.014877182431519032, 0.15132878720760345, 0.2482612580060959, 0.03901895508170128, 0.4666067659854889, -0.04503754898905754, -0.13180333375930786, 0.2440776824951172, 0.06336645036935806, 0.005415190011262894, 0.045186515897512436, 0.11685154587030411, -0.48749634623527527, -0.4285191595554352, -0.0431751050055027, 0.5997397899627686, -0.15401989221572876, 0.11646635830402374, 0.049157094210386276, -0.19502632319927216, -0.0481613390147686, 0.15840952098369598, 0.26105907559394836, -0.23258619010448456, 0.2867872416973114, -0.12047946453094482, 0.04268622025847435, -0.12510398030281067, 0.08967858552932739, -0.16128075122833252, -0.11361159384250641, 0.1927422285079956, -0.23611818253993988, -0.28055888414382935, -0.19184032082557678, 0.15163621306419373, 0.21312813460826874, -0.02919427677989006, 0.11034593731164932, -0.2356518805027008, -0.37427812814712524, -0.07946981489658356, -0.19264331459999084, 0.232706680893898, 0.3861810564994812, -0.07074125856161118, 0.1231977641582489, -0.49343231320381165, 0.08805859088897705, 0.09681472927331924, 0.35483765602111816, -0.2407086342573166, -0.09278297424316406, -0.10960809141397476, -0.018454577773809433, -0.3788088858127594, -0.25162914395332336, 0.1554819941520691, 0.19931045174598694, 0.008189727552235126, 0.11616271734237671, -0.2287406325340271, -0.1067013368010521, 0.09685689210891724, 0.08086860924959183, -0.07961466163396835, -0.17176373302936554, -0.17849427461624146, 0.09644437581300735, -0.09953518956899643, 0.21651560068130493, 0.2164943814277649, -0.06108015775680542, -0.0931972786784172, -0.2493305504322052, 0.05122865363955498, 0.26158350706100464, -0.05960594117641449, 0.48680588603019714, 0.6710014939308167, 0.09724391996860504, -0.20602203905582428, 0.1600842922925949, 0.07366205006837845, 0.05982647463679314, 0.08756179362535477, 0.1277293711900711, -0.036417849361896515, -0.09296656399965286, 0.10473373532295227, 0.4397200345993042, 0.155014306306839, 0.13724631071090698, -0.002952560782432556, 0.2808763384819031, 0.19765019416809082, 0.35864368081092834, 0.18620380759239197, 0.4341937005519867, 0.04112089052796364, -0.3968983292579651, 0.4125514328479767, 0.23424577713012695, -0.03565313294529915, -0.13068056106567383, -0.20733660459518433, -0.11624662578105927, 0.13849037885665894, 0.010381389409303665, -0.14836116135120392, 0.006765497848391533, -0.21708600223064423, -0.13100187480449677, 0.0985453799366951, -0.4322226643562317, 0.3699086010456085, 0.24644221365451813, -0.1015395075082779, -0.07392793893814087, 0.4425276815891266, 0.017667576670646667, -0.1838264912366867, 0.1610756665468216, 0.3885458707809448, 0.8879133462905884, 0.38734063506126404, 0.14933855831623077, 0.09435474127531052, 0.17374706268310547, 0.144248366355896, 0.7236191630363464, 0.23297686874866486, 0.11240533739328384, 0.3068799674510956, 0.25330448150634766, 0.10200037807226181, -0.22233879566192627, 0.27811357378959656, 0.1487635374069214, 0.13228155672550201, 0.011391657404601574, -0.035711925476789474, -0.22587181627750397, -0.336729496717453, -0.09032927453517914, -0.01322777196764946, 0.08499010652303696, 0.5246772766113281, 0.1999402642250061, 0.2989276945590973, -0.2885640263557434, 0.1400093138217926, 0.2076197862625122, 0.4345974028110504, -0.21771666407585144, -0.429130882024765, 0.192712664604187, -0.11657308042049408, 0.673844575881958, 0.053588710725307465, -0.031151695176959038, 0.28961843252182007, 0.054473426192998886, 0.0748957097530365, -0.383704274892807, -0.03711391240358353, -0.06795186549425125, 0.10829351097345352, 0.10009680688381195, -0.23522518575191498, 0.04172050952911377, 0.17708906531333923, -0.19993330538272858, 0.016714802011847496, 0.035177990794181824, -0.14711067080497742, -0.12542790174484253, 0.233220174908638, 0.5211760401725769, -0.23741385340690613, 0.1050829142332077, -0.020695747807621956, -0.314187616109848, -0.15902067720890045, -0.029608743265271187, 0.07011167705059052, -0.16472910344600677, 0.04769108071923256, 0.08573616296052933, 0.24409623444080353, 0.3486332297325134, 0.27942225337028503, 0.009575002826750278, -0.11959215998649597, -0.41816818714141846, -0.40806251764297485, -0.13012897968292236, 0.44738849997520447, 0.02127690799534321, -0.02329108864068985, -0.03652399778366089, 0.03883344680070877, -0.03901761397719383, 0.16553105413913727, -0.13707801699638367, 0.01628255285322666, 0.11755883693695068, -0.3692340552806854, -0.10993658751249313, -0.09451664984226227, -0.052483849227428436, 0.18723344802856445, -0.1290990114212036, -0.27491408586502075, 0.31997203826904297, 0.10629631578922272, -0.249078169465065, 0.09010880440473557, 0.15677814185619354, 0.18383309245109558, 0.23136135935783386, -0.10574916005134583, 0.2222570925951004, -0.013971826992928982, -0.18763667345046997, 0.06642072647809982, -0.08607548475265503, -0.48912349343299866, 0.009830782189965248, 0.09424680471420288, 0.08463659137487411, -0.020324304699897766, -0.22575195133686066, -0.45795533061027527, 0.5146231055259705, -0.30500301718711853, 0.08379430323839188, -0.718244731426239, 0.21887975931167603, 0.23858310282230377, 0.05179145932197571, -0.15571334958076477, 0.1274535357952118, 0.10186196863651276, 0.2214876115322113, -0.3064771890640259, 0.1651531159877777, 0.44851773977279663, 0.10766983032226562, -0.2351183295249939, -0.28615710139274597, 0.16090813279151917, -0.043487317860126495, 0.1759278029203415, -0.3066343665122986, 0.002416900359094143, 0.21606068313121796, 0.011468209326267242, 0.10717847943305969, 0.4145713150501251, 0.05682278051972389, 0.0054260315373539925, -0.20953305065631866, 0.02015507034957409, -0.14458635449409485, -0.030570190399885178, -0.11404658108949661, -0.2860076129436493 ]
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
[ 0.11682155728340149, 0.04799531400203705, -0.008072509430348873, -0.21743646264076233, 0.12888161838054657, 0.028689714148640633, 0.13635879755020142, 0.3606432378292084, -0.264738529920578, 0.2979812026023865, 0.011635798960924149, 0.8511586785316467, -0.3536044955253601, -0.2320977747440338, 0.20139330625534058, -0.0705004408955574, 0.15841247141361237, -0.10356622189283371, -0.24149112403392792, 0.05236358568072319, -0.13913846015930176, -0.3828108310699463, -0.07822688668966293, -0.20003600418567657, -0.13133981823921204, -0.09400442242622375, -0.08385658264160156, 0.13230125606060028, -0.18401820957660675, -0.2935965359210968, -0.407964289188385, 0.635504961013794, -0.10218502581119537, 0.1419236809015274, -0.00010813646804308519, -0.2722446620464325, 0.3262632191181183, 0.13107794523239136, -0.4321395456790924, 0.04468610882759094, -0.08185338228940964, -0.09962773323059082, 0.008636760525405407, -0.15263865888118744, 0.06552903354167938, 0.1338963359594345, 0.16543425619602203, -0.14641165733337402, 0.24235907196998596, 0.05031067878007889, 0.20914511382579803, -0.03290887549519539, -0.19010016322135925, 0.15847449004650116, 0.5000158548355103, 0.40686890482902527, -0.10122667998075485, 0.02862830087542534, 0.4395236372947693, 0.22933238744735718, -0.26579898595809937, 0.2744673788547516, -0.06643908470869064, -0.1814919263124466, 0.3078673183917999, -0.28791874647140503, -0.568967878818512, 0.07428516447544098, 0.00254220818169415, 0.17878597974777222, 0.5037116408348083, -0.28533533215522766, -0.39036980271339417, -0.35262995958328247, 0.3302517831325531, -0.18849989771842957, 0.019990235567092896, -0.09641724824905396, -0.05681619420647621, -0.06574321538209915, -0.24524734914302826, -0.15989528596401215, 0.03680569306015968, -0.021753426641225815, 0.516143262386322, 0.6055987477302551, 0.20056715607643127, -0.0033854590728878975, 0.244998037815094, 0.08339651674032211, 0.5854984521865845, -0.00934140756726265, 0.03303343802690506, 0.29964470863342285, -0.14720937609672546, -0.20837336778640747, 0.05500360205769539, 0.2948368489742279, 0.23643526434898376, 0.1446126252412796, 0.3270719349384308, 0.28509995341300964, -0.041302140802145004, 0.03735360875725746, 0.005551229231059551, 0.03812026232481003, -0.05310114473104477, 0.05495556443929672, 0.28106918931007385, -0.30150800943374634, 0.23124811053276062, 0.003970597870647907, -0.0490373857319355, -0.0805973932147026, -0.010975543409585953, 0.13244427740573883, -0.1945452094078064, -0.0684988722205162, 0.03613012284040451, -0.5815837979316711, -0.04963265359401703, -0.32450786232948303, 0.14889656007289886, 0.1077943816781044, 0.2770358622074127, 0.1407211422920227, -0.10792279243469238, -0.04817398637533188, 0.18350952863693237, -0.23938733339309692, -0.10505363345146179, 0.02583879604935646, -0.4939267039299011, 0.17078779637813568, 0.17374026775360107, -0.07522222399711609, 0.13901489973068237, 0.11294354498386383, 0.06702785938978195, 0.23749452829360962, 0.07884728908538818, 0.26584652066230774, 0.40570104122161865, -0.20365078747272491, -0.4380703866481781, -0.1334306299686432, -0.10840364545583725, 0.28055936098098755, -0.37445271015167236, 0.252822607755661, -0.2730662524700165, -0.23034383356571198, -0.09223335236310959, 0.156523659825325, -0.09696948528289795, -0.269822359085083, -0.17591546475887299, 0.3530561327934265, 0.06264601647853851, -0.4234464764595032, 0.35462215542793274, 0.07788465917110443, 0.062279775738716125, -0.004048002418130636, 0.013162915594875813, 0.3176633417606354, 0.04695463925600052, 0.020499395206570625, -0.24038122594356537, 0.0018360089743509889, 0.23789848387241364, 0.23913870751857758, -0.19989807903766632, -0.02130218781530857, -0.16164134442806244, -0.06126963347196579, 0.5229777097702026, 0.18964558839797974, -0.20806993544101715, 0.013186002150177956, -0.2095041573047638, -0.09727854281663895, 0.40916773676872253, 0.3883783519268036, -0.12241676449775696, -0.3362310826778412, 0.20757317543029785, 0.1173953264951706, -0.41046634316444397, 0.1484217792749405, -0.3570764362812042, -0.07999246567487717, 0.4687560498714447, 0.31442928314208984, -0.24346326291561127, 0.0008655057172290981, 0.07365547120571136, -0.15947984158992767, 0.2940755784511566, -0.1337454617023468, -0.1370372474193573, -0.23566286265850067, -0.06926552951335907, 0.1285499632358551, -0.17178978025913239, -0.11686255782842636, -0.10402880609035492, -0.23739497363567352, 0.22362768650054932, -0.051780592650175095, 0.11644157767295837, -0.2859424352645874, 0.08745779097080231, -0.33856987953186035, 0.0740102156996727, -0.16828975081443787, 0.09142880886793137, 0.11526463180780411, -0.10391442477703094, -0.1696280688047409, -0.22858014702796936, -0.06513699889183044, -0.14694342017173767, -0.18812064826488495, -0.3230234980583191, 0.02998712658882141, -0.11075796186923981, -0.05123119801282883, -0.07592533528804779, 0.46679219603538513, -0.08376159518957138, -0.06239979714155197, -0.09606490284204483, 0.29552972316741943, -0.00012781273107975721, -0.2677120268344879, 0.3739857077598572, 0.29607221484184265, 0.03366595879197121, 0.03514617681503296, 0.09655135869979858, 0.03362156078219414, -0.05559786409139633, -0.002862672321498394, -0.3500952422618866, 0.3163100779056549, -0.16839592158794403, 0.028979802504181862, -0.18219533562660217, -0.2597624659538269, 0.06748025119304657, 0.2224579006433487, -0.33422815799713135, 0.16626115143299103, 0.21991674602031708, 0.21182870864868164, 0.28131020069122314, 0.22115717828273773, -0.47536152601242065, 0.06712370365858078, 0.4927191436290741, -0.0540447011590004, 0.018465720117092133, 0.34422391653060913, 0.2549563944339752, -0.3329377770423889, 0.12259748578071594, 0.12657155096530914, 0.27662360668182373, 0.4469536244869232, 0.05266129598021507, 0.14309270679950714, 0.09512905776500702, -0.16515357792377472, -0.044339485466480255, -0.06922032684087753, -0.1712166666984558, -0.02813546173274517, 0.2366146296262741, 0.14595533907413483, -0.28768280148506165, -0.3273473083972931, -0.04693558067083359, 0.1891975700855255, -0.11021775007247925, -0.058831699192523956, -0.1330818086862564, -0.3062010407447815, -0.3187926411628723, -0.09166981279850006, -0.3395281136035919, -0.30699580907821655, 0.3917927145957947, 0.010386633686721325, -0.3207392692565918, 0.4042350649833679, 0.14643117785453796, 0.5348382592201233, -0.2747848629951477, -0.011294024996459484, -0.2897605895996094, -0.43033847212791443, -0.08097262680530548, 0.07382325083017349, 0.2886367738246918, -0.04778197035193443, 0.7124970555305481, 0.07893375307321548, -0.13275255262851715, -0.30444130301475525, -0.5318077802658081, -0.14086173474788666, 0.031390685588121414, -0.21374160051345825, 0.3421265482902527, -0.39242318272590637, 0.09010981023311615, -0.3882695734500885, -0.18767431378364563, -0.3428921401500702, -0.08024100959300995, 0.12296085804700851, 0.0008934360812418163, 0.0223995391279459, -0.29992929100990295, -0.3448787033557892, -0.0723855271935463, -0.44976741075515747, 0.07709689438343048, -0.6393796801567078, 0.07205915451049805, 0.056613899767398834, -0.23647019267082214, -0.16641591489315033, -0.37279030680656433, 0.007765217684209347, -0.1279706209897995, -0.596966564655304, 0.16613708436489105, -0.1103721633553505, -0.06128225103020668, -0.25566282868385315, -0.056039854884147644, 0.013812823221087456, 0.6228567361831665, 0.1948453038930893, -0.13971494138240814, 0.10687892138957977, 0.07648121565580368, -0.08116404712200165, 0.15028268098831177, 0.43438124656677246, 0.16205047070980072, -0.12345719337463379, 0.11876938492059708, -0.18264278769493103, 0.04211527109146118, -0.015398133546113968, 0.32869139313697815, 0.03482571244239807, 0.2597433626651764, 0.132559135556221, 0.5310878753662109, -0.5287259817123413, -0.08033153414726257, 0.056856993585824966, 0.17602069675922394, 0.2583037316799164, -0.09530039131641388, 0.0022554812021553516, 0.538794755935669, 0.29065465927124023, -0.2538527846336365, 0.2548821270465851, -0.02697642520070076, -0.309207946062088, 0.12093067914247513, 0.11350788921117783, -0.0975208729505539, -0.15820005536079407, 0.46763983368873596, -0.3359433710575104, 0.02855132333934307, 0.054026778787374496, 0.12251213192939758, -0.26753947138786316, -0.16778570413589478, 0.007246167864650488, 0.044103674590587616, 0.3468674421310425, -0.13314171135425568, -0.42491504549980164, -0.04560152068734169, -0.29521650075912476, 0.19519013166427612, 0.262794554233551, 0.2744704484939575, -0.015354691073298454, -0.17939461767673492, 0.13794772326946259, 0.12701435387134552, 0.39910879731178284, -0.18520724773406982, -0.15685276687145233, -0.08336519449949265, 0.12594904005527496, -0.18887241184711456, 0.05366368591785431, -0.11359266936779022, 0.10668953508138657, 0.16241835057735443, 0.05929332599043846, -0.26366713643074036, -0.19537672400474548, 0.13751184940338135, 0.0862833559513092, 0.16331186890602112, -0.32343992590904236, -0.07841923087835312, -0.1965930163860321, -0.13586987555027008, 0.25778910517692566, -0.006182495504617691, -0.28980693221092224, -0.0326429083943367, 0.083587147295475, -0.05086014047265053, 0.22913531959056854, -0.008727198466658592, 0.17341569066047668, 0.05967763066291809, 0.210801899433136, -0.02444561757147312, -0.008577005937695503, 0.3802029490470886, 0.20151636004447937, -0.17109251022338867, -0.3553253710269928, 0.0014370860299095511, -0.0802207812666893, -0.3615972101688385, 0.2507229447364807, 0.21086853742599487, 0.2100779414176941, 0.26145613193511963, -0.1482923924922943, 0.15075692534446716, -0.23374517261981964, -0.030188648030161858, -0.06791025400161743, 0.007533748168498278, -0.5339681506156921, -0.8507199287414551, 0.5450387597084045, 0.324373334646225, -0.3054797649383545, 0.44373300671577454, -0.024501098319888115, -0.35627877712249756, 0.3782590925693512, 0.3299746811389923, 1.0034143924713135, -0.0732174664735794, 0.2827591598033905, 0.16532643139362335, -0.3735780417919159, 0.4298727214336395, -0.18536476790905, 0.29490453004837036, -0.199883833527565, -0.269877552986145, 0.10410721600055695, -0.0036481136921793222, -0.028021423146128654, 0.28348782658576965, -0.23576772212982178, 0.31240278482437134, -0.16876865923404694, 0.029424291104078293, 0.12953615188598633, 0.2141420543193817, -0.26988524198532104, -0.22724011540412903, -0.2418520450592041, 0.13649654388427734, -0.22734935581684113, -0.21245139837265015, 0.021457143127918243, -0.15386490523815155, 0.007626306265592575, 0.06421827524900436, -0.17730727791786194, -0.11582420766353607, 0.11999896168708801, -0.13377785682678223, -0.19474442303180695, -0.06841374188661575, 0.5051288604736328, 0.015424735844135284, 0.2500437796115875, 0.0793086588382721, 0.14135025441646576, 0.23059943318367004, 0.07139069586992264, 0.3418620824813843, -0.04602609574794769, -0.03625037893652916, 0.17303436994552612, 0.02574159763753414, 0.0010225859005004168, -0.05954594165086746, 0.13033851981163025, -0.6264159679412842, -0.47416719794273376, -0.008948001079261303, 0.3204629123210907, -0.025614704936742783, 0.1331462860107422, 0.23690783977508545, -0.07707684487104416, -0.08411940932273865, 0.15529409050941467, 0.2397758960723877, -0.24776588380336761, 0.3586961030960083, -0.11311723291873932, -0.03689295053482056, -0.07945557683706284, 0.15876643359661102, -0.2302314043045044, 0.015226062387228012, 0.3005436956882477, -0.23293353617191315, -0.1303839534521103, -0.23513083159923553, -0.019167596474289894, 0.23810617625713348, -0.06477062404155731, 0.15195618569850922, -0.13655045628547668, -0.2242705225944519, -0.06621591746807098, -0.13610877096652985, 0.052162013947963715, 0.28071829676628113, -0.054642368108034134, 0.013735885731875896, -0.3953648805618286, 0.020800147205591202, 0.0989733338356018, 0.30215853452682495, -0.11496814340353012, 0.1467445194721222, -0.15087012946605682, 0.005487203598022461, -0.37829846143722534, -0.24021053314208984, 0.11466765403747559, 0.319722443819046, 0.07316631078720093, 0.15754017233848572, -0.21056245267391205, -0.21844400465488434, 0.06116172671318054, 0.04334093630313873, -0.04984681308269501, -0.2375105619430542, -0.13929511606693268, 0.06823792308568954, -0.0031457676086574793, 0.20376445353031158, 0.15675577521324158, -0.12760601937770844, -0.03966988995671272, -0.20938262343406677, -0.1145479753613472, 0.17652060091495514, -0.019262196496129036, 0.5665623545646667, 0.5050376057624817, 0.10235250741243362, 0.009776461869478226, 0.18582408130168915, -0.1405273675918579, -0.11232739686965942, 0.019576584920287132, 0.2767694592475891, -0.013695317320525646, 0.05107109621167183, 0.02432303875684738, 0.2786233425140381, 0.1418539434671402, 0.08936867862939835, -0.08146517723798752, 0.19317317008972168, 0.10954715311527252, 0.4665200114250183, 0.061829715967178345, 0.35824066400527954, 0.08235473930835724, -0.48556673526763916, 0.4337496757507324, 0.2181694656610489, -0.16130036115646362, -0.0022095523308962584, -0.173440620303154, 0.07349132746458054, 0.16542163491249084, -0.0031536463648080826, -0.10503245145082474, -0.0687323734164238, -0.28761211037635803, -0.07359521836042404, 0.14936892688274384, -0.5486273765563965, 0.28238943219184875, 0.2594434916973114, -0.12507213652133942, -0.16082531213760376, 0.217079758644104, -0.09827984124422073, -0.16002942621707916, 0.16166959702968597, 0.3601122796535492, 1.004159927368164, 0.3350103497505188, 0.21191076934337616, 0.024365659803152084, 0.14911019802093506, 0.09338062256574631, 0.6433221697807312, 0.37874460220336914, 0.14951620995998383, 0.4294309616088867, 0.2312847077846527, 0.1897776573896408, -0.30213481187820435, 0.26007258892059326, 0.16610713303089142, 0.1658436805009842, 0.06580910086631775, -0.11546791344881058, -0.11934860050678253, -0.26598143577575684, -0.007739689666777849, -0.05381818860769272, 0.10528258234262466, 0.5309206247329712, 0.18618077039718628, 0.16036458313465118, -0.41128966212272644, 0.20345550775527954, 0.2584473788738251, 0.47005727887153625, -0.18684881925582886, -0.24186141788959503, 0.025070015341043472, -0.21610885858535767, 0.7542917728424072, 0.041959211230278015, 0.02450696751475334, 0.2809271514415741, 0.04929617792367935, 0.032787639647722244, -0.41480574011802673, 0.009417206048965454, -0.15027841925621033, 0.17937573790550232, 0.2596067190170288, -0.12145703285932541, 0.021811291575431824, 0.13871459662914276, -0.11151226609945297, 0.20139597356319427, -0.04670346900820732, -0.07099395990371704, -0.08586213737726212, 0.31809765100479126, 0.5659240484237671, -0.2598892152309418, 0.040796004235744476, 0.029231498017907143, -0.13603146374225616, -0.06514059007167816, 0.006684291176497936, 0.14587584137916565, -0.17006824910640717, -0.0568634457886219, 0.09269672632217407, 0.3524606227874756, 0.40319985151290894, 0.2684459686279297, -0.10742657631635666, -0.11384572833776474, -0.3688149154186249, -0.5009534358978271, -0.1890980750322342, 0.35712480545043945, -0.05158931389451027, -0.02819443866610527, -0.02227320708334446, -0.018851567059755325, -0.0188995860517025, -0.044772710651159286, 0.03546379879117012, 0.013845215551555157, 0.07077246159315109, -0.365939736366272, -0.013891619630157948, -0.20218130946159363, -0.07325322180986404, 0.14310459792613983, -0.12609434127807617, -0.2787931263446808, 0.3147028684616089, 0.04463818669319153, -0.22550302743911743, 0.2509474754333496, -0.08069885522127151, 0.14602084457874298, 0.03393829986453056, 0.13528800010681152, 0.18925458192825317, 0.049447763711214066, -0.11370112746953964, -0.1537807732820511, -0.10955603420734406, -0.6963850855827332, 0.005130944773554802, 0.058589380234479904, 0.054867301136255264, -0.06912807375192642, -0.10537657141685486, -0.41169610619544983, 0.50682532787323, -0.2618788480758667, 0.08110775798559189, -0.582264244556427, 0.17405715584754944, 0.19211417436599731, 0.1904580295085907, -0.147505521774292, 0.2194097340106964, 0.09773124009370804, 0.1874556988477707, -0.23057447373867035, 0.1007898673415184, 0.4633551239967346, 0.021655917167663574, -0.17914269864559174, -0.21649980545043945, 0.24227572977542877, 0.012541387230157852, -0.14014913141727448, -0.3644132614135742, -0.17112553119659424, 0.17598265409469604, -0.04166668653488159, 0.2560970187187195, 0.4652128517627716, 0.046046044677495956, -0.046316877007484436, -0.18348297476768494, 0.0035770623944699764, -0.06137705594301224, -0.04535584896802902, 0.005076654255390167, -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
[ -0.25224706530570984, -0.2662774920463562, -0.1281619668006897, -0.38261544704437256, -0.4156891107559204, -0.10087782144546509, 0.019268035888671875, 0.21927271783351898, -0.08714980632066727, 0.38095977902412415, 0.05702590569853783, 0.3763355016708374, -0.015546144917607307, 0.38528192043304443, -0.07647734880447388, 0.14052651822566986, 0.08472870290279388, 0.36588814854621887, -0.08661056309938431, -0.13832901418209076, -0.20194722712039948, 0.1104140654206276, -0.06942588835954666, -0.05444721132516861, 0.14053840935230255, 0.1708446592092514, -0.1972423940896988, -0.0724431648850441, -0.22427019476890564, -0.20771373808383942, 0.3550550043582916, 0.4555104374885559, 0.22513939440250397, 0.06983289867639542, -0.00010584224219201133, -0.23818418383598328, 0.40260031819343567, -0.07296498119831085, -0.3339161276817322, -0.23844744265079498, -0.4179013967514038, -0.3771299421787262, 0.06569415330886841, -0.09597811847925186, 0.06359268724918365, 0.09290829300880432, 0.16502171754837036, -0.1966177076101303, 0.06413626670837402, 0.011999454349279404, 0.3026268184185028, 0.17412489652633667, -0.08758427947759628, -0.34514912962913513, 0.33494833111763, 0.4491066336631775, -0.35373738408088684, -0.17945297062397003, 0.3167716860771179, 0.01306353323161602, -0.06734225153923035, 0.4110815227031708, 0.22772838175296783, -0.1817958652973175, -0.025220684707164764, -0.06428272277116776, 0.1812652051448822, 0.00741971842944622, 0.14173556864261627, 0.11687707155942917, 0.5061972737312317, 0.12980134785175323, -0.4155389070510864, 0.2899840772151947, 0.004279250744730234, -0.49353092908859253, 0.26659294962882996, -0.10623881220817566, -0.029614422470331192, 0.06423865258693695, 0.1872217357158661, -0.33883580565452576, -0.4957720637321472, 0.2065143883228302, -0.07795720547437668, 0.09057173132896423, -0.15790343284606934, -0.15367045998573303, 0.12449891120195389, 0.10767853260040283, -0.04348260909318924, 0.04136791080236435, -0.17636631429195404, 0.2043864130973816, -0.24313992261886597, -0.25906282663345337, 0.28339555859565735, 0.0010784141486510634, 0.45887577533721924, 0.08281127363443375, 0.15400449931621552, 0.04369572177529335, -0.0718052089214325, 0.10415859520435333, 0.15376785397529602, 0.20156995952129364, 0.46074289083480835, -0.05985700711607933, 0.39925825595855713, 0.2871793210506439, 0.47682806849479675, 0.042006246745586395, 0.15518613159656525, 0.03422975912690163, -0.06027251109480858, -0.03193383291363716, 0.07255369424819946, -0.5815816521644592, -0.0960402637720108, 0.05002456530928612, -0.01638941653072834, 0.016601091250777245, -0.03814186155796051, 0.09351354837417603, 0.26915350556373596, 0.22098171710968018, 0.29162850975990295, -0.06439059972763062, -0.03568042814731598, -0.48720136284828186, -0.2953397333621979, -0.07494328171014786, -0.24456514418125153, 0.29936912655830383, 0.3277590274810791, -0.42228972911834717, 0.13820594549179077, -0.31309938430786133, 0.28337621688842773, 0.04933447763323784, -0.180079385638237, 0.3390873372554779, -0.22517737746238708, 0.12475220113992691, 0.057131897658109665, 0.08531034737825394, -0.27858278155326843, 0.5125008225440979, -0.1971522569656372, 0.12465895712375641, -0.33828333020210266, -0.4485531449317932, -0.38689664006233215, 0.2870851755142212, -0.0740518644452095, -0.1555136740207672, -0.15182098746299744, 0.3666447401046753, -0.23981423676013947, -0.16440413892269135, -0.08050496131181717, 0.30108094215393066, -0.27182331681251526, -0.12088774144649506, 0.027140580117702484, 0.14325271546840668, -0.12814576923847198, -0.04009607061743736, -0.5082142353057861, -0.09364745765924454, -0.1425669938325882, -0.029935723170638084, -0.12085287272930145, -0.2735487222671509, -0.16993111371994019, -0.17352014780044556, 0.3994160294532776, -0.3332189917564392, -0.1463918387889862, 0.23286618292331696, -0.028698895126581192, -0.2497420310974121, 0.14834274351596832, -0.15650637447834015, 0.0467190146446228, -0.2703949809074402, -0.22819823026657104, 0.04900611191987991, -0.14814770221710205, -0.10404177755117416, -0.14730137586593628, -0.4412732720375061, -0.18274016678333282, 0.2603656053543091, 0.24740341305732727, 0.033463869243860245, 0.22792409360408783, -0.399613618850708, 0.09001302719116211, -0.16994056105613708, 0.1756184846162796, -0.019763948395848274, 0.7601091265678406, 0.03279290348291397, 0.1584608107805252, 0.0639222264289856, -0.3966703712940216, 0.07201750576496124, 0.09055466204881668, 0.169218510389328, 0.22833280265331268, -0.3766861855983734, -0.035117290914058685, -0.0492091029882431, 0.09248863905668259, -0.0949602946639061, 0.20418104529380798, 0.12047625333070755, -0.13210277259349823, -0.0152665376663208, -0.028222646564245224, 0.2592668831348419, -0.24022415280342102, -0.15056923031806946, -0.07579459995031357, 0.1669909507036209, -0.18648114800453186, 0.03914568945765495, -0.18251177668571472, 0.12863761186599731, -0.07770580798387527, -0.0659889280796051, -0.1271277517080307, 0.3903428912162781, -0.09755893051624298, 0.23098556697368622, 0.05519198998808861, 0.32341456413269043, 0.14523309469223022, -0.17544782161712646, 0.0759681686758995, -0.1623837649822235, -0.10134578496217728, 0.1879151314496994, -0.22592777013778687, 0.38410696387290955, 0.0692225769162178, -0.15123489499092102, 0.009792744182050228, -0.008774356916546822, 0.06245538592338562, -0.09672950953245163, -0.3642009496688843, -0.11718852072954178, 0.1875603348016739, 0.11140111088752747, -0.44378724694252014, -0.28570082783699036, -0.3088611960411072, -0.0880218967795372, 0.03273797035217285, -0.06471414864063263, 0.11028388887643814, 0.202970951795578, -0.0012312150793150067, -0.23741556704044342, 0.30815061926841736, -0.10598643869161606, 0.273464560508728, 0.38543739914894104, -0.1669514775276184, -0.021352306008338928, -0.2048148810863495, 0.11400625109672546, 0.138396218419075, 0.2815909683704376, 0.04978889971971512, 0.1022426187992096, 0.3348847031593323, 0.09427323937416077, -0.350233256816864, -0.30894431471824646, -0.04541853442788124, 0.057697996497154236, -0.15502667427062988, -0.19572319090366364, -0.03369702398777008, 0.07117698341608047, 0.0014657102292403579, 0.14784599840641022, 0.16372084617614746, -0.19144901633262634, 0.40738844871520996, 0.2692485451698303, -0.22280722856521606, 0.3243142068386078, 0.0367613360285759, 0.8047559857368469, -0.12738479673862457, 0.18538588285446167, -0.244889497756958, 0.05357452481985092, -0.31329864263534546, 0.25463640689849854, -0.22676579654216766, -0.22785387933254242, 0.4608537554740906, -0.10080944001674652, 0.2458382397890091, -0.11814744025468826, -0.7101554274559021, 0.09934356808662415, -0.10333850979804993, -0.09342791140079498, 0.030385173857212067, -0.21468646824359894, -0.07387538999319077, 0.019448628649115562, 0.059018027037382126, -0.44859611988067627, -0.482538640499115, -0.10999584943056107, -0.2480621337890625, 0.1305108517408371, -0.36499282717704773, -0.4868319034576416, 0.14573654532432556, -0.26064586639404297, -0.2076714038848877, -0.0529680959880352, 0.16154150664806366, 0.26981121301651, 0.12356329709291458, -0.0581197626888752, 0.06886541098356247, 0.05131300166249275, -0.3860308825969696, 0.04582163318991661, 0.3424372971057892, -0.3188912272453308, -0.4158431887626648, 0.12177399545907974, 0.011761395260691643, 0.15939757227897644, 0.1757788509130478, -0.3779390752315521, -0.39495596289634705, 0.2752779722213745, 0.12953302264213562, -0.14654022455215454, 0.06901098042726517, 0.5344861149787903, 0.06983812898397446, -0.21253934502601624, -0.07770729809999466, -0.4181554913520813, 0.18073098361492157, 0.46797892451286316, 0.14321157336235046, -0.29487934708595276, 0.2655492126941681, 0.19312967360019684, 0.3202052116394043, 0.16470855474472046, -0.07552196830511093, 0.12141634523868561, 0.1337214708328247, 0.21813704073429108, -0.2059163898229599, 0.010713852010667324, 0.016583023592829704, 0.05266087129712105, 0.17323456704616547, 0.5349008440971375, 0.19639062881469727, -0.4072144627571106, -0.37295612692832947, -0.22823147475719452, 0.09779666364192963, 0.0007368475198745728, -0.1938171535730362, 0.16490212082862854, 0.44048231840133667, -0.06908144056797028, -0.15247796475887299, -0.3290402293205261, -0.10789789259433746, 0.09796718508005142, 0.341805636882782, 0.15771354734897614, -0.08698372542858124, 0.09859788417816162, 0.17732582986354828, -0.2839728593826294, 0.35610511898994446, 0.1472279578447342, -0.0044860453344881535, 0.005454570055007935, -0.10187935829162598, 0.15186411142349243, 0.019838934764266014, 0.1859121173620224, -0.2136528342962265, -0.16913339495658875, -0.1400270313024521, -0.08662858605384827, 0.08970767259597778, -0.06182194501161575, -0.11171083152294159, -0.2805306911468506, -0.09050513058900833, 0.05194632336497307, 0.02726135216653347, -0.14532765746116638, 0.6831697225570679, -0.17081421613693237, 0.010883476585149765, -0.2852771282196045, 0.05051574110984802, -0.23356260359287262, -0.0009022200247272849, -0.017939334735274315, -0.19018055498600006, 0.0005651781684719026, -0.10173071920871735, -0.15057457983493805, 0.2393195629119873, 0.013857969082891941, -0.006709704641252756, 0.14193840324878693, -0.14012378454208374, 0.01141413301229477, 0.11479334533214569, 0.0812954232096672, 0.04784291237592697, 0.37294790148735046, 0.12629111111164093, 0.22759857773780823, -0.26654568314552307, 0.17605413496494293, -0.1937771886587143, 0.06537327170372009, 0.24919407069683075, 0.03014046512544155, -0.06442665308713913, -0.29717183113098145, -0.22939181327819824, -0.12345857173204422, -0.2008921504020691, 0.2619435489177704, 0.041544437408447266, -0.3726586699485779, 0.028384743258357048, 0.32451099157333374, 0.02175331301987171, -0.272900253534317, 0.2598053216934204, 0.21533317863941193, -0.24798914790153503, 0.33155831694602966, 0.000659460318274796, 0.7347703576087952, 0.24774326384067535, -0.06344851106405258, 0.23774388432502747, -0.04784760996699333, 0.49948814511299133, -0.15336962044239044, 0.013707512989640236, 0.07889294624328613, -0.13275223970413208, -0.09385038167238235, 0.29843801259994507, 0.37705063819885254, -0.0920611247420311, -0.30580031871795654, 0.17827291786670685, 0.3676147758960724, 0.24887479841709137, 0.08150078356266022, 0.18558509647846222, 0.3359358310699463, -0.30991461873054504, 0.1831713318824768, 0.1883806586265564, 0.007580730132758617, 0.16981655359268188, 0.000819022417999804, -0.036344654858112335, 0.014058265835046768, -0.3261232376098633, -0.22565685212612152, -0.14075137674808502, 0.34428584575653076, -0.009299780242145061, -0.03576219826936722, -0.07222994416952133, 0.4353812634944916, -0.3726440966129303, 0.20798252522945404, 0.19100753962993622, -0.289655864238739, 0.12577462196350098, 0.16576622426509857, 0.053724367171525955, -0.06959031522274017, -0.094485342502594, 0.07702063769102097, -0.16798236966133118, -0.09140515327453613, 0.189395472407341, 0.21163713932037354, -0.15778563916683197, -0.0802491307258606, -0.14002512395381927, -0.032127268612384796, 0.048934001475572586, 0.36480703949928284, 0.5734384655952454, 0.19827164709568024, -0.2803203761577606, 0.24688148498535156, 0.22918559610843658, -0.1116873100399971, -0.07789988070726395, 0.15205608308315277, -0.10007689148187637, -0.2093641757965088, 0.4151040315628052, 0.4110594689846039, -0.4059256911277771, -0.04722736403346062, -0.08289742469787598, -0.08956559747457504, -0.27430734038352966, -0.046448029577732086, 0.12019757181406021, -0.5605069994926453, -0.19973479211330414, 0.06941748410463333, -0.20039395987987518, 0.18112991750240326, 0.7872426509857178, 0.1409822255373001, 0.17524504661560059, -0.05200938135385513, -0.01895678974688053, -0.18221913278102875, 0.0649246945977211, -0.43132078647613525, 0.3021855652332306, -0.026819266378879547, -0.1548498421907425, -0.11379795521497726, 0.03981051966547966, -0.4568730592727661, -0.19220364093780518, 0.26230743527412415, -0.13239428400993347, 0.37219324707984924, 0.31053993105888367, 0.2645421326160431, 0.0035356192383915186, 0.21191640198230743, -0.14141181111335754, -0.333881139755249, -0.20281733572483063, 0.06465177983045578, 0.09212342649698257, 0.066771499812603, 0.1733732372522354, -0.038243938237428665, -0.11493132263422012, -0.26429882645606995, 0.03255194053053856, 0.2622591257095337, -0.029773158952593803, -0.1581728607416153, 0.1851058453321457, -0.24483951926231384, -0.07198569923639297, 0.15404565632343292, 0.3823603093624115, 0.18938875198364258, 0.08267559856176376, -0.06526102870702744, 0.4259438216686249, -0.007294213864952326, -0.11266492307186127, 0.046603430062532425, 0.37811312079429626, -0.2774091362953186, 0.30024397373199463, 0.01867472194135189, -0.024585021659731865, -0.08572158962488174, -0.03785385936498642, 0.13398221135139465, -0.4113802909851074, -0.07079371064901352, 0.02890712581574917, 0.033668648451566696, 0.3719574511051178, -0.28011345863342285, 0.05250160023570061, -0.058175042271614075, -0.055172450840473175, 0.2176857441663742, 0.19678087532520294, 0.347512423992157, 0.015417419373989105, 0.7643712759017944, 0.032585255801677704, 0.012450192123651505, -0.05361108481884003, 0.45255008339881897, 0.10950759798288345, 0.04519696161150932, 0.092800073325634, 0.08463887870311737, 0.06753087043762207, 0.17613516747951508, -0.06873729825019836, 0.11972779035568237, -0.08243212103843689, 0.4220133125782013, 0.5082911252975464, 0.009178519248962402, -0.33825504779815674, 0.07828788459300995, 0.28291958570480347, 0.08967132121324539, 0.03318355232477188, 0.30095845460891724, -0.16590328514575958, 0.011287071742117405, 0.07916275411844254, -0.7445873022079468, -0.1521623283624649, 0.020007845014333725, -0.15537327527999878, -0.15849360823631287, -0.13477882742881775, -0.40616244077682495, -0.06609372794628143, 0.16439220309257507, 0.007546603679656982, 0.1547902673482895, -0.19950610399246216, 0.0156902726739645, -0.10653707385063171, -0.007983309216797352, 0.11110591143369675, 0.22404581308364868, 0.026626797392964363, 0.3685724139213562, -0.49482420086860657, -0.2772012948989868, 0.30086541175842285, 0.3290161192417145, -0.01646137610077858, -0.07080608606338501, 0.38324427604675293, 0.12766486406326294, 0.08333270996809006, -0.10145274549722672, 0.04092055931687355, 0.08495263755321503, 0.09764688462018967, 0.14666281640529633, 0.20908553898334503, 0.20400592684745789, -0.13698941469192505, 0.24193722009658813, -0.07487769424915314, -0.1394316703081131, -0.28207167983055115, 0.11750311404466629, 0.16903309524059296, 0.09013192355632782, 0.03028200939297676, 0.16643118858337402, 0.041157688945531845, -0.3709057569503784, -0.07294447720050812, -0.057605814188718796, -0.1295691877603531, -0.16902805864810944, 0.13554269075393677, -0.16934321820735931, -0.048529867082834244, 0.22995789349079132, 0.007201621308922768, -0.013600761070847511, -0.402497261762619, 0.0531436912715435, 0.3580961227416992, 0.14343705773353577, -0.0768822655081749, 0.0670187771320343, 0.22650203108787537, 0.25970277190208435, -0.1298811137676239, 0.15739211440086365, -0.2579095959663391, 0.3699952960014343, 0.09451120346784592, -0.3097194731235504, 0.16183723509311676, 0.45948633551597595, -0.028369247913360596, -0.11244422942399979, -0.1568193882703781, -0.10699564963579178, 0.04783365875482559, 0.0735948458313942, -0.24181491136550903, 0.07018910348415375, 0.13543981313705444, -0.001816423493437469, 0.5354022979736328, 0.0257039126008749, 0.17436641454696655, -0.2132638394832611, -0.04778827726840973, -0.44986972212791443, -0.2057502567768097, -0.08338746428489685, 0.02801678143441677, 0.08583550155162811, 0.08767634630203247, -0.004705298226326704, 0.05368807911872864, -0.03178027272224426, 0.2798509895801544, -0.014120880514383316, 0.00022543963859789073, -0.24943330883979797, 0.47038906812667847, -0.37131351232528687, -0.2708434462547302, -0.049185220152139664, -0.15531040728092194, 0.11308012902736664, -0.17197838425636292, -0.10465776175260544, -0.17666897177696228, 0.2870733439922333, -0.071136474609375, 0.05178253725171089, -0.18545517325401306, 0.32384487986564636, 0.3122841715812683, -0.0000627957415417768, -0.33270496129989624, 0.06873562932014465, 0.1956883668899536, 0.17945174872875214, -0.13641569018363953, 0.26792481541633606, -0.07479515671730042, 0.02350030280649662, -0.18617725372314453, -0.011183531023561954, -0.19011545181274414, -0.06747455894947052, 0.03285990282893181, -0.37482622265815735 ]
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
[ -0.02323281578719616, -0.18676303327083588, -0.1784633845090866, -0.3313103914260864, -0.5435771942138672, 0.0657171905040741, -0.17831721901893616, 0.12594059109687805, -0.34406140446662903, 0.4689723551273346, 0.23541775345802307, 0.12405291199684143, 0.07490067929029465, 0.4212982952594757, 0.01564185880124569, 0.0894116684794426, 0.11703021079301834, 0.40649157762527466, 0.15641163289546967, -0.07623111456632614, -0.04696584865450859, 0.32866430282592773, -0.2688596546649933, -0.01200912706553936, 0.16894924640655518, 0.11247768253087997, -0.32557398080825806, -0.26800379157066345, -0.14709429442882538, -0.16988062858581543, 0.4301484227180481, 0.28739723563194275, 0.3542039692401886, -0.008296303451061249, -0.00011804488894995302, -0.24944931268692017, 0.2678099572658539, -0.06388570368289948, -0.22480154037475586, -0.2884773015975952, -0.3987264037132263, -0.37567535042762756, -0.1216314435005188, -0.015980413183569908, 0.13137918710708618, 0.058061666786670685, 0.20080380141735077, 0.01567564159631729, -0.12592117488384247, 0.10718570649623871, 0.18949460983276367, 0.18076005578041077, 0.14113673567771912, -0.3795875906944275, 0.17057523131370544, 0.31918057799339294, -0.2734740078449249, -0.2285310924053192, 0.2699374854564667, 0.02585633099079132, 0.3077669143676758, 0.4816245138645172, 0.2874336838722229, -0.3409464657306671, -0.17208921909332275, -0.013270976021885872, 0.11673465371131897, -0.12508049607276917, 0.27835148572921753, -0.045266292989254, 0.5306006073951721, 0.17983469367027283, -0.38571861386299133, 0.41310515999794006, 0.04533279314637184, -0.4731229245662689, 0.26737481355667114, -0.013816172257065773, 0.06820400059223175, 0.11895722150802612, 0.23744791746139526, -0.305306613445282, -0.4384269416332245, 0.16605907678604126, -0.16326308250427246, 0.005223647225648165, -0.11229757219552994, -0.18681839108467102, 0.0656309649348259, 0.045105691999197006, -0.04875468835234642, 0.18406322598457336, -0.250978022813797, 0.26039397716522217, -0.22966527938842773, -0.383053719997406, 0.2817981541156769, 0.021808361634612083, 0.3297159969806671, -0.034621939063072205, 0.15862037241458893, -0.2194433957338333, -0.2475138157606125, 0.0007928619161248207, 0.12062596529722214, 0.3228638172149658, 0.35899588465690613, 0.05294784903526306, 0.14591911435127258, 0.23537763953208923, 0.44059988856315613, -0.04040248319506645, 0.07127875834703445, 0.01292092353105545, -0.16717562079429626, 0.056895069777965546, 0.06021241098642349, -0.604914665222168, -0.1965150535106659, -0.004578383639454842, -0.2337338626384735, 0.06281173974275589, -0.14380207657814026, 0.03192567080259323, 0.4180449843406677, 0.2180628925561905, 0.2200229912996292, -0.19355085492134094, -0.009806999936699867, -0.615911602973938, -0.2765553593635559, -0.014677828177809715, -0.06726319342851639, 0.3063664436340332, 0.21016402542591095, -0.36181536316871643, 0.22055070102214813, -0.3328659236431122, 0.2242400348186493, 0.05530818551778793, -0.1585884392261505, 0.4364674985408783, -0.22660861909389496, 0.1858072578907013, -0.022986486554145813, 0.06077186390757561, -0.28177499771118164, 0.49024420976638794, -0.2170695662498474, 0.1842242330312729, -0.3824470341205597, -0.4308209717273712, -0.5014219880104065, 0.14038513600826263, 0.05070982873439789, -0.19810688495635986, 0.03673606738448143, 0.24654091894626617, -0.2765115201473236, -0.17123226821422577, 0.02272341586649418, 0.2964550256729126, -0.17472514510154724, -0.13263124227523804, -0.00553283654153347, 0.17904718220233917, -0.26312288641929626, -0.06219698116183281, -0.3929761052131653, 0.07705941796302795, -0.11440914869308472, 0.058697137981653214, -0.06112329289317131, -0.31643974781036377, -0.14322926104068756, -0.25974997878074646, 0.3386320471763611, -0.6128677725791931, -0.5238037109375, 0.14918272197246552, -0.03131807968020439, -0.2901611626148224, 0.21898776292800903, 0.0052811154164373875, 0.007203897926956415, -0.1621968150138855, -0.20951786637306213, -0.03626938536763191, -0.2873406708240509, -0.3331367075443268, 0.021973565220832825, -0.4311453402042389, -0.07969625294208527, 0.10742981731891632, 0.09247319400310516, 0.09353889524936676, 0.23239237070083618, -0.6215120553970337, 0.14444591104984283, -0.01956053264439106, 0.010354234836995602, 0.06936413794755936, 0.6898936629295349, 0.07786291092634201, 0.26113954186439514, 0.051612675189971924, -0.31839659810066223, 0.056199997663497925, -0.2152480185031891, 0.06954531371593475, 0.2679485082626343, -0.3772900700569153, -0.11547408998012543, 0.0034696152433753014, 0.16920092701911926, -0.2824362516403198, 0.04788864031434059, 0.048852499574422836, 0.10802231729030609, 0.0012774725910276175, 0.013500632718205452, 0.33671054244041443, -0.15173639357089996, -0.1799667477607727, -0.10303991287946701, 0.1771983653306961, -0.19508013129234314, 0.008516116067767143, 0.016285469755530357, 0.2220592200756073, 0.10666127502918243, 0.019013136625289917, -0.08943860977888107, 0.19991247355937958, -0.16442209482192993, 0.2517150044441223, 0.25995251536369324, 0.43010544776916504, 0.22773388028144836, -0.22602134943008423, 0.06586539000272751, -0.09690932929515839, -0.1978672593832016, 0.3507431745529175, -0.01078734453767538, 0.25207388401031494, 0.14984849095344543, -0.07520993798971176, -0.007431892212480307, 0.20284220576286316, 0.031593725085258484, 0.04165693372488022, -0.20685194432735443, -0.11898760497570038, 0.2147410809993744, 0.10000596195459366, -0.348251074552536, -0.38111934065818787, -0.20631013810634613, 0.1608160138130188, 0.13865163922309875, -0.11609699577093124, -0.0007915304158814251, 0.11394789814949036, -0.02068360522389412, -0.08092039823532104, 0.3460270166397095, -0.10363270342350006, 0.09568571299314499, 0.28695350885391235, -0.2408590018749237, -0.18608057498931885, -0.00962131842970848, 0.16023384034633636, 0.15206803381443024, 0.27990028262138367, -0.2412092536687851, -0.1741049587726593, 0.37920963764190674, 0.1788070797920227, -0.12696203589439392, -0.23240917921066284, -0.014871442690491676, 0.05603490024805069, 0.028574738651514053, -0.09625121206045151, -0.13883979618549347, 0.005258546210825443, -0.06992295384407043, -0.059411391615867615, 0.18317662179470062, -0.19776903092861176, 0.27721211314201355, 0.1787654608488083, 0.07255777716636658, 0.2747688889503479, -0.0001998098596232012, 0.7763861417770386, -0.2226073294878006, 0.30331775546073914, -0.07584984600543976, 0.13949930667877197, -0.42412352561950684, 0.13569270074367523, -0.2917243242263794, -0.1730983853340149, 0.41069939732551575, -0.09878098964691162, 0.3334304392337799, -0.0885261595249176, -0.8045455813407898, -0.0008277685847133398, -0.10774475336074829, -0.10973869264125824, 0.11922883987426758, -0.3133365213871002, -0.09177757799625397, 0.024535227566957474, -0.025782432407140732, -0.2734338343143463, -0.4853475093841553, -0.1646532267332077, -0.3983466327190399, 0.08785327523946762, -0.17175859212875366, -0.3822392523288727, -0.04387316107749939, -0.07672450691461563, -0.12428029626607895, -0.12437798827886581, 0.26520493626594543, 0.12538963556289673, -0.008540192618966103, -0.07314744591712952, 0.17700523138046265, -0.0466655008494854, -0.1972220540046692, 0.20906874537467957, 0.3373955488204956, -0.33221402764320374, -0.3296048939228058, 0.3415127694606781, -0.021702377125620842, 0.2813383936882019, 0.011478010565042496, -0.4279918968677521, -0.15206891298294067, 0.06688906252384186, 0.13095754384994507, 0.051632486283779144, -0.03655092418193817, 0.7915471792221069, -0.11600492894649506, -0.07611781358718872, -0.027554595842957497, -0.5734845995903015, 0.10225136578083038, 0.5704914927482605, 0.24227403104305267, -0.1116476058959961, 0.31742674112319946, 0.15854166448116302, 0.5127478837966919, 0.35929906368255615, 0.010425462387502193, 0.022734301164746284, 0.17875592410564423, 0.04132753610610962, -0.1243182122707367, -0.23321433365345, 0.030881822109222412, 0.08355056494474411, 0.0810873806476593, 0.3579850494861603, 0.17127445340156555, -0.2575511634349823, -0.23423998057842255, -0.4313468039035797, -0.048631273210048676, 0.1353088766336441, -0.3467608690261841, 0.08997053653001785, 0.6666930913925171, -0.05727139487862587, -0.1794433444738388, -0.48417314887046814, -0.052042484283447266, 0.19989366829395294, 0.4482506215572357, 0.31226596236228943, -0.0023317551240324974, 0.15901756286621094, -0.06324784457683563, -0.2990370988845825, 0.38223665952682495, 0.09644591063261032, 0.07628446072340012, -0.02362111210823059, 0.057053014636039734, 0.16065505146980286, 0.003988371230661869, 0.20746296644210815, -0.11883365362882614, -0.20663690567016602, -0.08955513685941696, -0.04198884963989258, 0.32143864035606384, -0.032358571887016296, 0.03982272371649742, -0.24847453832626343, -0.10593555867671967, -0.18759500980377197, 0.19501085579395294, -0.031700219959020615, 0.661746621131897, -0.2799379229545593, -0.17864881455898285, -0.11406147480010986, -0.05655478313565254, -0.07455010712146759, 0.045680951327085495, -0.05885607376694679, -0.26100632548332214, -0.01762528531253338, -0.03928031772375107, -0.17231270670890808, 0.21923212707042694, -0.2052270919084549, 0.007135676220059395, 0.048595596104860306, -0.06891925632953644, -0.01868927851319313, 0.08036024123430252, 0.10861817747354507, -0.028513725847005844, 0.2341812402009964, 0.0030013923533260822, 0.3989971876144409, -0.3178151249885559, 0.17326252162456512, -0.04418456181883812, 0.22356350719928741, 0.15306679904460907, -0.03164377808570862, 0.07378613948822021, -0.17802105844020844, -0.31108027696609497, -0.22394557297229767, -0.41653263568878174, 0.3540853261947632, 0.0946246013045311, -0.20846574008464813, 0.08699792623519897, 0.2448112815618515, 0.09113167971372604, -0.2714417278766632, 0.33740517497062683, 0.39424195885658264, -0.01379085797816515, 0.178623229265213, -0.04098159074783325, 0.7828720211982727, 0.38780492544174194, -0.11633848398923874, 0.08909568935632706, -0.1653100997209549, 0.30338603258132935, 0.027321895584464073, 0.10152041912078857, 0.16568328440189362, -0.1250874102115631, -0.13380064070224762, 0.19046860933303833, 0.3938691020011902, -0.19231148064136505, -0.27615976333618164, 0.3090684115886688, 0.29632294178009033, 0.37913796305656433, 0.18517246842384338, 0.046606484800577164, 0.3087807595729828, -0.01217486709356308, 0.22883759438991547, 0.08569752424955368, 0.17001819610595703, 0.4446307420730591, -0.015017342753708363, -0.1496933400630951, -0.0736471563577652, -0.24480481445789337, -0.3190770745277405, -0.242390975356102, 0.10532345622777939, -0.12534372508525848, 0.06371274590492249, -0.13501378893852234, 0.4795578122138977, -0.18676002323627472, 0.43321555852890015, 0.01976674236357212, -0.3183548152446747, 0.114285908639431, 0.24963229894638062, -0.24592357873916626, 0.0507211908698082, -0.03206101059913635, -0.2735678553581238, -0.19744344055652618, -0.12815308570861816, 0.4675575792789459, 0.2643011212348938, -0.1820860654115677, -0.29033800959587097, -0.1792747676372528, -0.009771398268640041, 0.16735336184501648, 0.3781091272830963, 0.49319449067115784, 0.21740108728408813, -0.23251906037330627, 0.13885167241096497, 0.14542095363140106, -0.12462029606103897, -0.3335355222225189, -0.10845143347978592, 0.12606151401996613, -0.15928465127944946, 0.20721028745174408, 0.4298137426376343, -0.3224421739578247, 0.07744607329368591, -0.12384442239999771, 0.03547762334346771, -0.13220760226249695, -0.2717389762401581, 0.030616771429777145, -0.43419989943504333, -0.15630215406417847, 0.03174484148621559, -0.17103591561317444, 0.19185861945152283, 0.9666602611541748, 0.06149487942457199, 0.20077861845493317, -0.18781894445419312, -0.015118432231247425, -0.20959652960300446, 0.1748201996088028, -0.5250983834266663, 0.21809229254722595, -0.17597238719463348, -0.07979463040828705, -0.16315676271915436, 0.26689285039901733, -0.2849743366241455, -0.22890225052833557, 0.17995238304138184, -0.15777553617954254, 0.332746297121048, 0.289823055267334, 0.17223280668258667, 0.11206067353487015, 0.04333651810884476, -0.13135673105716705, -0.14088326692581177, -0.04713466763496399, 0.1000949963927269, 0.12241390347480774, 0.08467119932174683, 0.3393619656562805, -0.037495195865631104, -0.18758808076381683, -0.19110724329948425, 0.19971665740013123, 0.24985700845718384, -0.09309850633144379, 0.0867888554930687, 0.2476036548614502, -0.2960224747657776, -0.1283234804868698, -0.08523381501436234, 0.5458859205245972, 0.24163506925106049, 0.13734957575798035, 0.1575447916984558, 0.3756001591682434, -0.0907992273569107, -0.03890293464064598, -0.07435854524374008, 0.44656896591186523, -0.3893502950668335, 0.25679096579551697, 0.03685351461172104, 0.09243882447481155, -0.1318696290254593, 0.038229480385780334, -0.2625074088573456, -0.5318466424942017, -0.012972341850399971, 0.19005700945854187, -0.05723995715379715, 0.20679323375225067, -0.3643629252910614, 0.022032788023352623, -0.26102837920188904, -0.0993008092045784, -0.029650526121258736, 0.21123574674129486, 0.3775576651096344, 0.07074476033449173, 0.8590599298477173, 0.11811657249927521, -0.11029167473316193, -0.0721648707985878, 0.4866432845592499, 0.22771117091178894, 0.390080988407135, 0.1389111578464508, 0.09025737643241882, 0.0017621266888454556, 0.10719453543424606, -0.165870800614357, -0.07671059668064117, -0.13972154259681702, 0.4415702819824219, 0.4549821615219116, 0.3275452256202698, -0.20021550357341766, -0.20803415775299072, 0.24367780983448029, 0.18482717871665955, 0.014333994127810001, 0.2330917865037918, -0.08115053921937943, 0.12781989574432373, 0.09431018680334091, -0.7621111273765564, -0.09891969710588455, 0.07419556379318237, -0.16120022535324097, -0.1580151617527008, -0.12280523031949997, -0.26581400632858276, -0.11310786008834839, 0.3679637610912323, 0.09795967489480972, 0.2267370969057083, -0.2759629786014557, -0.14949998259544373, -0.05910405516624451, -0.20407916605472565, 0.28761768341064453, 0.259488970041275, 0.06134825944900513, 0.4843769669532776, -0.5644541382789612, -0.3007172644138336, 0.4297739267349243, 0.21083767712116241, 0.019137809053063393, -0.2560523450374603, 0.29161855578422546, 0.05553995445370674, -0.02176821231842041, -0.2667056918144226, 0.23909786343574524, 0.18271808326244354, -0.14026693999767303, 0.11083662509918213, -0.01465611718595028, 0.062215372920036316, 0.04342527687549591, -0.058801013976335526, 0.012008771300315857, -0.08062668144702911, -0.4757100045681, 0.13423487544059753, -0.20503267645835876, 0.1652188003063202, -0.12193163484334946, 0.4493809640407562, 0.08370702713727951, -0.3802337348461151, -0.2866358757019043, -0.044458724558353424, -0.22272084653377533, 0.014352958649396896, 0.055022258311510086, -0.4033523499965668, -0.16033092141151428, 0.38487374782562256, 0.1999768316745758, -0.12101326882839203, -0.6013122200965881, 0.06163432076573372, 0.2573508620262146, 0.1335514932870865, -0.13102884590625763, 0.11507105082273483, 0.26026177406311035, 0.1477556675672531, -0.25710946321487427, 0.23596131801605225, -0.014066509902477264, 0.2910212278366089, 0.09884415566921234, -0.1629011332988739, 0.08296121656894684, 0.33615177869796753, 0.003927461802959442, -0.241536483168602, -0.01370282657444477, -0.1254633218050003, 0.28384727239608765, -0.1250927895307541, -0.14821316301822662, -0.029125139117240906, 0.09563417732715607, 0.12166106700897217, 0.46083855628967285, 0.09020748734474182, 0.021488890051841736, -0.06572095304727554, -0.14753438532352448, -0.4012153446674347, -0.060062576085329056, 0.032888349145650864, -0.13261793553829193, 0.01420643925666809, 0.05187319219112396, -0.13590918481349945, 0.3132944405078888, -0.00924934446811676, 0.43694430589675903, 0.1265650987625122, -0.02437763102352619, -0.14826248586177826, 0.4578132629394531, -0.4662644565105438, -0.16879823803901672, 0.28169527649879456, -0.11046450585126877, 0.0842432901263237, -0.11323761940002441, -0.10509444773197174, -0.01785537786781788, 0.17952172458171844, -0.15866152942180634, 0.22237536311149597, -0.20011016726493835, 0.40533655881881714, 0.2907789647579193, -0.031212477013468742, -0.35276690125465393, -0.07439649105072021, 0.1831320822238922, 0.27748629450798035, -0.017893442884087563, 0.3830200135707855, 0.17758183181285858, 0.059884212911129, -0.12405876070261002, 0.2663041353225708, -0.1834818422794342, -0.004384451080113649, 0.04803779348731041, -0.2990565299987793 ]
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.
[ -0.2492513507604599, -0.1036457046866417, -0.13912150263786316, -0.4210846424102783, -0.5307989716529846, -0.06874822080135345, -0.07875005155801773, 0.0790654867887497, -0.21976330876350403, 0.43905583024024963, 0.1513829380273819, 0.29369914531707764, 0.0653417557477951, 0.4375327527523041, 0.006778002250939608, 0.2519570291042328, 0.154427632689476, 0.398141473531723, -0.03590301424264908, -0.08111084997653961, -0.09899748116731644, 0.30109986662864685, -0.2269812822341919, -0.05949539318680763, 0.06923174113035202, -0.020368149504065514, -0.2728619873523712, -0.16568756103515625, -0.15396250784397125, -0.15395794808864594, 0.41139307618141174, 0.3626806139945984, 0.20779547095298767, 0.004311579279601574, -0.0001105137198464945, -0.21097858250141144, 0.2824910581111908, -0.1277121603488922, -0.23003865778446198, -0.12380574643611908, -0.33321255445480347, -0.3251551687717438, 0.00864006020128727, 0.002827170304954052, 0.08679933845996857, 0.136275053024292, 0.12899619340896606, -0.10482455044984818, 0.0015904665924608707, 0.17070777714252472, 0.26446279883384705, 0.26653367280960083, -0.005565534345805645, -0.39218854904174805, 0.16724227368831635, 0.2768290638923645, -0.31645849347114563, -0.1491539180278778, 0.30685749650001526, -0.017131377011537552, 0.08296898007392883, 0.5250523686408997, 0.26207658648490906, -0.17759644985198975, -0.12935487926006317, -0.05383513867855072, 0.12355474382638931, -0.11475832760334015, 0.3260294497013092, 0.055493343621492386, 0.5153299570083618, 0.14862285554409027, -0.42044371366500854, 0.38333144783973694, -0.0019987288396805525, -0.38956838846206665, 0.2576102018356323, -0.05857376754283905, 0.006151999346911907, 0.05968880280852318, 0.14559221267700195, -0.1512787789106369, -0.3991319239139557, 0.14248180389404297, -0.17090865969657898, 0.03124726004898548, -0.1322280615568161, -0.21965524554252625, 0.08026741445064545, 0.09614317119121552, -0.1033070832490921, 0.2098437398672104, -0.258256733417511, 0.18841110169887543, -0.2771470844745636, -0.2824537754058838, 0.3214116394519806, 0.06597021222114563, 0.41316038370132446, -0.062351662665605545, 0.1519732028245926, -0.09951052814722061, -0.11116885393857956, 0.14640198647975922, 0.05179968848824501, 0.2504030466079712, 0.235428124666214, -0.053371917456388474, 0.32314833998680115, 0.22540897130966187, 0.48004665970802307, -0.03999730572104454, 0.16129685938358307, 0.04168177396059036, -0.21857784688472748, 0.012427395209670067, -0.026502054184675217, -0.675119936466217, -0.27946555614471436, 0.0029018407221883535, -0.07580024749040604, 0.05903168022632599, -0.07884777337312698, 0.14038971066474915, 0.38729918003082275, 0.11087030917406082, 0.23839271068572998, -0.03799464926123619, -0.06861920654773712, -0.5039108395576477, -0.31144461035728455, 0.04990041255950928, -0.08856603503227234, 0.2114315927028656, 0.29866087436676025, -0.22972559928894043, 0.3060237169265747, -0.33083871006965637, 0.36734291911125183, 0.11446765810251236, -0.20475688576698303, 0.36665451526641846, -0.19268490374088287, 0.28552624583244324, 0.06034734100103378, 0.04286833852529526, -0.22038298845291138, 0.3571576476097107, -0.22282858192920685, 0.09572034329175949, -0.2677159011363983, -0.42350247502326965, -0.4708488881587982, 0.19991572201251984, 0.05588306486606598, -0.22937507927417755, 0.049219727516174316, 0.32721787691116333, -0.2058468610048294, -0.17716984450817108, -0.025696886703372, 0.3373243510723114, -0.10757417231798172, -0.12294255197048187, 0.11131129413843155, 0.18788522481918335, -0.290587842464447, -0.009797886945307255, -0.36860230565071106, 0.04430405795574188, -0.04414789006114006, 0.04356708005070686, -0.1005471721291542, -0.22311870753765106, -0.11662527173757553, -0.234968364238739, 0.30609145760536194, -0.3721947968006134, -0.4113188087940216, 0.13090983033180237, 0.05861206725239754, -0.34151026606559753, 0.12815356254577637, -0.15334750711917877, 0.01939241588115692, -0.19477084279060364, -0.17797912657260895, 0.10281573981046677, -0.1804133504629135, -0.16533176600933075, -0.02364679053425789, -0.41427046060562134, -0.1249915286898613, 0.20059582591056824, 0.12091831117868423, 0.06461630016565323, 0.2701733112335205, -0.6116053462028503, 0.05576060712337494, -0.13917377591133118, 0.11108893156051636, 0.0837540253996849, 0.7464856505393982, -0.04615722596645355, 0.18957947194576263, 0.14584112167358398, -0.31344014406204224, 0.023550845682621002, -0.2664191722869873, 0.1402662694454193, 0.11664490401744843, -0.3408263623714447, -0.09279391914606094, 0.002391715534031391, 0.06440763175487518, -0.33980438113212585, 0.1386115550994873, 0.09465963393449783, 0.05083655193448067, 0.09796702861785889, 0.12586458027362823, 0.30949774384498596, -0.19311819970607758, -0.22526568174362183, 0.001833481714129448, 0.14263980090618134, -0.26999393105506897, -0.10479746758937836, 0.1443323940038681, 0.28665733337402344, 0.02502942457795143, -0.04019114747643471, -0.16687524318695068, 0.26575028896331787, -0.16327393054962158, 0.27370139956474304, 0.09590210020542145, 0.38308003544807434, 0.14369435608386993, -0.1427636295557022, 0.0734449177980423, -0.11707872897386551, -0.12005744874477386, 0.3190229833126068, -0.11595192551612854, 0.3486648201942444, 0.08622607588768005, -0.08114539831876755, -0.03529399260878563, 0.18887214362621307, 0.18004757165908813, -0.03710181638598442, -0.21818651258945465, -0.10062821209430695, 0.19473578035831451, 0.09217849373817444, -0.3050864636898041, -0.36676180362701416, -0.24421091377735138, 0.05284702405333519, 0.1815319061279297, -0.008825973607599735, 0.09464174509048462, 0.1603843718767166, -0.03851761296391487, -0.16388629376888275, 0.4497126042842865, -0.09222956746816635, 0.10848800092935562, 0.30496329069137573, -0.2005971223115921, -0.15271040797233582, -0.14874950051307678, 0.04772251099348068, 0.12991151213645935, 0.26417914032936096, -0.0851147398352623, -0.04902438446879387, 0.3393489420413971, 0.14000371098518372, -0.35330459475517273, -0.22798289358615875, -0.058396413922309875, 0.028281457722187042, 0.0008942059357650578, -0.0952368974685669, -0.13379205763339996, 0.006520122289657593, 0.03023415245115757, 0.0011706569930538535, 0.13668008148670197, -0.23664964735507965, 0.3674170672893524, 0.14765097200870514, -0.11774065345525742, 0.3822777271270752, -0.007208812981843948, 0.750099778175354, -0.11174754053354263, 0.30824875831604004, -0.18575984239578247, 0.12192963063716888, -0.423503041267395, 0.22970548272132874, -0.28927677869796753, -0.12858949601650238, 0.49558523297309875, -0.09794559329748154, 0.471207857131958, -0.10985421389341354, -0.7647413015365601, 0.06731845438480377, -0.12290873378515244, -0.01593669503927231, 0.014705386012792587, -0.2569577395915985, -0.15914715826511383, -0.01935533806681633, -0.049181826412677765, -0.3315581679344177, -0.4355301856994629, -0.05172495171427727, -0.4801786541938782, -0.005570515990257263, -0.3519700765609741, -0.4595111012458801, -0.021146252751350403, -0.21627852320671082, -0.1986929029226303, -0.21795080602169037, 0.17308765649795532, 0.2513792812824249, 0.025441674515604973, 0.050335194915533066, -0.015435825102031231, 0.007242949213832617, -0.2543773651123047, 0.12977926433086395, 0.49869662523269653, -0.3865443170070648, -0.5030508637428284, 0.2911558747291565, -0.04333087056875229, 0.3502500653266907, -0.006293013691902161, -0.35988572239875793, -0.26416605710983276, 0.0008944234577938914, 0.1479327231645584, -0.009468025527894497, 0.038268450647592545, 0.7621445059776306, -0.11581346392631531, -0.19106876850128174, -0.11626650393009186, -0.5392231941223145, 0.195543572306633, 0.49827247858047485, 0.24280884861946106, -0.27890506386756897, 0.381069153547287, 0.20991499722003937, 0.3193809688091278, 0.35116875171661377, -0.05651888623833656, 0.12163636833429337, 0.16339369118213654, 0.14771617949008942, -0.19633936882019043, -0.016693875193595886, 0.14289216697216034, 0.07544203847646713, 0.14860863983631134, 0.5334334969520569, 0.12728251516819, -0.3797416687011719, -0.2531399726867676, -0.2787446081638336, -0.05657920613884926, 0.05245794728398323, -0.29205578565597534, 0.015738386660814285, 0.5925154685974121, -0.027571655809879303, -0.1643771529197693, -0.29470542073249817, -0.024257855489850044, 0.18267175555229187, 0.40327081084251404, 0.22562986612319946, -0.011568686924874783, 0.08430559933185577, -0.07599975168704987, -0.3604991137981415, 0.38195040822029114, 0.05675148963928223, 0.10147249698638916, -0.12819631397724152, 0.0615396723151207, 0.1697922796010971, 0.010395820252597332, 0.23315389454364777, -0.1957228034734726, -0.2707689106464386, -0.09753895550966263, -0.0332343690097332, 0.20889709889888763, -0.07565904408693314, -0.06040547043085098, -0.1760564148426056, -0.05845481529831886, -0.17866337299346924, 0.03458104282617569, -0.09470617026090622, 0.5656788349151611, -0.281123548746109, -0.04575740545988083, -0.09697021543979645, -0.01756092719733715, -0.2522437870502472, -0.05329691618680954, 0.01899961568415165, -0.1461750566959381, -0.029696712270379066, -0.06324408948421478, -0.2030782252550125, 0.3560207188129425, -0.13447655737400055, 0.005193465389311314, 0.20223785936832428, 0.013026390224695206, 0.05348801240324974, 0.06713107973337173, 0.038212310522794724, 0.031947098672389984, 0.20771417021751404, 0.022881265729665756, 0.36382195353507996, -0.14250944554805756, 0.26192325353622437, -0.2598850131034851, 0.25790277123451233, 0.17956392467021942, -0.0531545951962471, 0.047247301787137985, -0.2090415060520172, -0.2319691777229309, -0.11197464168071747, -0.2654051184654236, 0.28373515605926514, 0.036032821983098984, -0.14099271595478058, 0.12018971890211105, 0.2442871630191803, 0.026496756821870804, -0.24268589913845062, 0.23383940756320953, 0.34146368503570557, -0.10660894960165024, 0.3187497854232788, 0.014410682022571564, 0.7144323587417603, 0.28476589918136597, -0.15133036673069, 0.17770220339298248, -0.09784683585166931, 0.28467896580696106, -0.08520586788654327, 0.03084203600883484, 0.14362406730651855, -0.029414881020784378, -0.1080327183008194, 0.22408486902713776, 0.47331008315086365, -0.18071293830871582, -0.40747833251953125, 0.23934638500213623, 0.37372472882270813, 0.25265419483184814, 0.179173082113266, 0.025824788957834244, 0.27924466133117676, -0.1941184252500534, 0.22286485135555267, 0.16126206517219543, 0.11865909397602081, 0.24784404039382935, -0.00793006457388401, -0.12176423519849777, -0.10707315057516098, -0.33475151658058167, -0.22227029502391815, -0.15550272166728973, 0.17588695883750916, -0.13710139691829681, 0.12958721816539764, -0.13968022167682648, 0.5002409219741821, -0.21630269289016724, 0.35122835636138916, 0.07694317400455475, -0.3939014673233032, 0.19849863648414612, 0.12374666333198547, -0.09371854364871979, -0.04847061634063721, 0.011539848521351814, -0.09307023882865906, -0.3150058686733246, -0.040459416806697845, 0.2745908796787262, 0.13504841923713684, -0.229490265250206, -0.15048819780349731, -0.15534135699272156, -0.05174742266535759, 0.19557896256446838, 0.3932056725025177, 0.49676477909088135, 0.20558051764965057, -0.2776391804218292, 0.2118620127439499, 0.19321781396865845, -0.09961574524641037, -0.09029293805360794, 0.001679231645539403, 0.05238170549273491, -0.2097773402929306, 0.20240306854248047, 0.34476226568222046, -0.20526762306690216, 0.11227019876241684, -0.06963343918323517, -0.07564254850149155, -0.23330840468406677, -0.32038724422454834, -0.01394434180110693, -0.5064800381660461, -0.21499545872211456, 0.04626109078526497, -0.2646346986293793, 0.1625894159078598, 0.9665590524673462, 0.019336046651005745, 0.08566281199455261, -0.15412995219230652, -0.12778057157993317, -0.2720561921596527, 0.09145692735910416, -0.6348620653152466, 0.3025994896888733, -0.09700286388397217, -0.010685187764465809, -0.058876533061265945, 0.2275681495666504, -0.4005643427371979, -0.23105886578559875, 0.08580855280160904, -0.07548953592777252, 0.17271456122398376, 0.18505536019802094, 0.300216406583786, 0.20005753636360168, 0.10802095383405685, -0.03642885759472847, -0.29785341024398804, -0.1409248560667038, 0.06706114858388901, 0.09022770822048187, 0.0851747989654541, 0.2651062607765198, 0.015573673881590366, -0.09612862765789032, -0.36367666721343994, 0.1402253955602646, 0.31589746475219727, -0.2554151713848114, -0.0834077000617981, 0.21389135718345642, -0.29568302631378174, -0.15318594872951508, 0.07409202307462692, 0.268174946308136, 0.2032589167356491, 0.00018322282994631678, 0.022653166204690933, 0.4017670154571533, -0.07402580231428146, -0.12795914709568024, 0.06694348156452179, 0.3990152180194855, -0.3243759870529175, 0.23266412317752838, 0.05787014216184616, 0.021776895970106125, -0.0746772363781929, -0.02741912007331848, -0.11537866294384003, -0.5985844135284424, -0.05257910490036011, 0.08984842151403427, 0.02726883441209793, 0.274113267660141, -0.40830355882644653, 0.10977047681808472, -0.23368923366069794, -0.06540714204311371, 0.17333252727985382, 0.12680713832378387, 0.5387256145477295, -0.04353973641991615, 0.9180060625076294, 0.10789885371923447, -0.07989369332790375, -0.07315406948328018, 0.3522469103336334, 0.09844844043254852, 0.19314135611057281, 0.20261448621749878, 0.16971679031848907, 0.01739216037094593, 0.08312473446130753, -0.06407185643911362, 0.010315862484276295, -0.09518471360206604, 0.32112690806388855, 0.4013400077819824, 0.15263208746910095, -0.31373369693756104, 0.03715776652097702, 0.16140064597129822, 0.19645005464553833, 0.01642380654811859, 0.2140069156885147, -0.09558118134737015, 0.1256731003522873, 0.11404751241207123, -0.8520112037658691, -0.14799943566322327, -0.00028115283930674195, -0.1519279032945633, -0.15474063158035278, -0.1215364933013916, -0.26129472255706787, -0.10409193485975266, 0.29159078001976013, 0.10824941098690033, 0.29016730189323425, -0.29568859934806824, -0.08855625241994858, -0.07152854651212692, -0.08061777800321579, 0.14213576912879944, 0.1428743600845337, 0.08947974443435669, 0.5019300580024719, -0.49050286412239075, -0.25601011514663696, 0.3050372898578644, 0.30416804552078247, 0.16102740168571472, -0.20795680582523346, 0.2815847098827362, 0.18219353258609772, 0.01964338682591915, -0.26642152667045593, 0.1677819788455963, 0.263054758310318, -0.001191830146126449, 0.1215171292424202, 0.10152959078550339, 0.11109840869903564, -0.06895507872104645, 0.019005170091986656, -0.110426165163517, -0.004637938924133778, -0.29690900444984436, 0.14319093525409698, -0.07266207039356232, 0.12516354024410248, -0.08303319662809372, 0.2422429621219635, 0.060601405799388885, -0.33828553557395935, -0.23554354906082153, -0.14563825726509094, -0.21853677928447723, -0.1448456197977066, 0.11242606490850449, -0.2741743326187134, 0.020280269905924797, 0.24042855203151703, 0.25404390692710876, -0.026827706024050713, -0.5908100008964539, 0.0029731381218880415, 0.2741870582103729, 0.13339382410049438, 0.033587243407964706, 0.08341857045888901, 0.25750410556793213, 0.20823241770267487, -0.26136961579322815, 0.2973886728286743, -0.14023049175739288, 0.3001268208026886, -0.008812090381979942, -0.3560074269771576, 0.06686931848526001, 0.26842185854911804, -0.008376008830964565, -0.1806323081254959, -0.04371897131204605, -0.11608517915010452, 0.24436362087726593, 0.006212041713297367, -0.2181481271982193, 0.01431509293615818, 0.08671735227108002, 0.04670063778758049, 0.47451040148735046, 0.2091078758239746, 0.1296912133693695, -0.14030086994171143, -0.13859006762504578, -0.4177078306674957, -0.09693166613578796, 0.040200021117925644, -0.16448312997817993, 0.07427632808685303, 0.2251129001379013, -0.0425955094397068, 0.1396816223859787, -0.003688798053190112, 0.4698244333267212, 0.07916233688592911, -0.08481240272521973, -0.12159675359725952, 0.5496421456336975, -0.41634684801101685, -0.17564964294433594, 0.12344091385602951, -0.13445450365543365, 0.16413378715515137, -0.18473336100578308, -0.14443321526050568, -0.10510902106761932, 0.2785085439682007, -0.22749796509742737, 0.0910428911447525, -0.0843522772192955, 0.31331202387809753, 0.11957121640443802, -0.010894513688981533, -0.3383719027042389, 0.041730623692274094, 0.11309710144996643, 0.23142509162425995, -0.085887610912323, 0.444174587726593, 0.10862895846366882, 0.011421849951148033, -0.14661389589309692, 0.36385494470596313, -0.15389764308929443, -0.04098913446068764, 0.04374954104423523, -0.31981804966926575 ]
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 !
[ -0.2524738311767578, -0.2647852599620819, -0.11422257125377655, -0.4864540994167328, -0.37947604060173035, -0.20802776515483856, 0.08028701692819595, 0.18927013874053955, 0.11779671162366867, 0.3005794286727905, -0.1557752639055252, 0.5226977467536926, -0.14769266545772552, 0.4125770032405853, 0.14620788395404816, 0.10172685980796814, 0.03274613246321678, 0.36265671253204346, -0.14609384536743164, -0.20175479352474213, -0.24998606741428375, 0.22434024512767792, -0.17047972977161407, -0.0546531081199646, 0.21928048133850098, 0.07695106416940689, -0.29594308137893677, 0.07935602962970734, -0.06169789656996727, -0.19533969461917877, 0.3291567862033844, 0.3292328119277954, 0.18306970596313477, 0.06467748433351517, -0.0001042149233398959, -0.23515945672988892, 0.2574850022792816, -0.12841558456420898, -0.28690019249916077, -0.18150971829891205, -0.1894628256559372, -0.47377657890319824, 0.22126221656799316, -0.21766071021556854, 0.10919701308012009, 0.12459298223257065, 0.20315435528755188, -0.12437304854393005, 0.12705568969249725, 0.22815097868442535, 0.3232874870300293, 0.40828680992126465, -0.034265510737895966, -0.18497470021247864, 0.20779332518577576, 0.2706397771835327, -0.20863200724124908, -0.012949143536388874, 0.15916481614112854, -0.27545619010925293, -0.19682584702968597, 0.40579551458358765, 0.14011330902576447, -0.09584835916757584, -0.13215944170951843, -0.06573261320590973, 0.16082273423671722, 0.02158612199127674, 0.010288055054843426, 0.03543519601225853, 0.23504425585269928, -0.04971456900238991, -0.3959871232509613, 0.1401749849319458, 0.028912870213389397, -0.5668572783470154, 0.09331749379634857, -0.08714385330677032, -0.1245134025812149, -0.14305627346038818, 0.20747773349285126, -0.1623469889163971, -0.4726060926914215, 0.2572449743747711, 0.004830068442970514, 0.2947580814361572, -0.0636480301618576, -0.201700359582901, 0.22664368152618408, 0.04250158369541168, -0.0799819827079773, 0.15150265395641327, -0.03282984346151352, 0.3442157506942749, -0.5325772166252136, -0.28025975823402405, 0.3405754268169403, 0.09036559611558914, 0.26787158846855164, 0.04260241612792015, 0.29281583428382874, 0.14166145026683807, -0.003866547252982855, 0.1931775063276291, 0.030217591673135757, 0.2401256263256073, 0.29266491532325745, -0.04129847511649132, 0.4281787872314453, 0.2653532028198242, 0.3137470483779907, 0.05636424198746681, 0.07044050097465515, -0.10898221284151077, -0.14499960839748383, -0.06849265098571777, 0.008154265582561493, -0.4685887396335602, -0.2638152539730072, -0.05357510596513748, -0.05584701895713806, 0.1455128937959671, -0.013918544165790081, 0.2732975482940674, 0.17012737691402435, 0.2448023557662964, 0.43004360795021057, 0.009215272963047028, -0.2248593270778656, -0.3951990008354187, -0.2916809916496277, 0.0973813608288765, -0.31904566287994385, 0.08400415629148483, 0.3731897473335266, -0.20060113072395325, 0.27423250675201416, -0.43340110778808594, 0.25433430075645447, 0.07114069163799286, -0.19919106364250183, 0.23665952682495117, -0.11447160691022873, 0.061437468975782394, 0.22511331737041473, -0.04297833889722824, -0.23835983872413635, 0.2062339037656784, -0.208388090133667, -0.0808207094669342, -0.31519266963005066, -0.3829514980316162, -0.40688356757164, 0.27562111616134644, -0.05516853928565979, -0.3148968815803528, -0.1343497335910797, 0.3234364688396454, -0.1739307940006256, -0.18654756247997284, -0.036192961037158966, 0.13921798765659332, -0.25417572259902954, -0.12859772145748138, 0.09105677902698517, 0.11252575367689133, 0.05460629239678383, -0.13686442375183105, -0.5214048624038696, 0.0013879557373002172, -0.010002112947404385, 0.049277134239673615, 0.0218989048153162, -0.12459671497344971, -0.1309000551700592, -0.13091003894805908, 0.5791589617729187, -0.36789417266845703, -0.06391140818595886, 0.12525290250778198, 0.0537506639957428, -0.2881126403808594, 0.09406545758247375, -0.17142540216445923, -0.1428021490573883, -0.10669810324907303, -0.1012190505862236, 0.3418886661529541, -0.07049549371004105, 0.09957309812307358, -0.1165209412574768, -0.3035583794116974, -0.012261515483260155, 0.3013629615306854, 0.2912398874759674, -0.0810905322432518, 0.2330091893672943, -0.1553657352924347, 0.20717059075832367, -0.27829211950302124, 0.0341414250433445, 0.05451255291700363, 0.5859829187393188, 0.046290259808301926, 0.09907408803701401, -0.05174323543906212, -0.3171607255935669, 0.0166789498180151, -0.11427313834428787, 0.2951972782611847, 0.11283311247825623, -0.29259777069091797, -0.032230351120233536, 0.05407179519534111, 0.11803607642650604, -0.11149363964796066, 0.2614138126373291, 0.11125187575817108, 0.02320450358092785, 0.0792037770152092, 0.07408283650875092, 0.370496928691864, -0.14236293733119965, -0.13941872119903564, -0.25196877121925354, 0.19473256170749664, -0.2114020735025406, -0.09095368534326553, 0.09032333642244339, 0.30762800574302673, -0.014147036708891392, -0.016491519287228584, -0.09004618972539902, 0.3420785963535309, -0.1977245807647705, 0.3473963141441345, 0.06613973528146744, 0.11991211026906967, 0.03165197744965553, -0.18702732026576996, 0.10462112724781036, 0.020478181540966034, 0.1050189658999443, 0.19834399223327637, -0.18742062151432037, 0.4752040207386017, 0.09016674757003784, -0.059459175914525986, 0.09390714019536972, 0.032954420894384384, 0.1509227305650711, -0.23204603791236877, -0.3025614321231842, 0.01277228444814682, 0.3870747983455658, 0.11589404195547104, -0.36801931262016296, -0.373274028301239, -0.3721950948238373, -0.0663413256406784, 0.15661008656024933, 0.005696083884686232, 0.21773898601531982, 0.16886994242668152, -0.022963739931583405, -0.24440787732601166, 0.23325873911380768, -0.09595499187707901, 0.31295421719551086, 0.3339039385318756, -0.08845502883195877, 0.036614567041397095, -0.3495223820209503, 0.03818896412849426, 0.18476954102516174, 0.21045918762683868, -0.0037190252915024757, 0.18175466358661652, 0.43975594639778137, -0.03354663774371147, -0.4896511137485504, -0.4122563898563385, -0.22404496371746063, 0.04827713221311569, -0.24939954280853271, -0.012751425616443157, -0.06169269233942032, -0.14514487981796265, -0.05505142733454704, 0.09441986680030823, 0.09586165100336075, -0.3651599884033203, 0.2836414575576782, 0.33103182911872864, -0.2519606947898865, 0.1996849626302719, -0.01126947533339262, 0.6269262433052063, -0.04075061157345772, 0.12832388281822205, -0.2383575439453125, -0.05031690001487732, -0.24117931723594666, 0.26183509826660156, -0.24112819135189056, 0.031092802062630653, 0.483078271150589, -0.1881498098373413, 0.26334211230278015, 0.051257967948913574, -0.6950430274009705, 0.019227707758545876, -0.12177256494760513, -0.06724286824464798, 0.10748769342899323, -0.31135663390159607, -0.1011442020535469, -0.06949954479932785, 0.043600983917713165, -0.636491060256958, -0.3448212444782257, -0.01351488009095192, -0.23367318511009216, 0.16604377329349518, -0.40941816568374634, -0.4348345100879669, -0.10463443398475647, -0.37184280157089233, -0.1279684454202652, 0.09901562333106995, 0.18223023414611816, 0.390127032995224, -0.005850309040397406, 0.17498701810836792, 0.0610402375459671, 0.05882662534713745, -0.31342649459838867, 0.17934054136276245, 0.42067214846611023, -0.40306001901626587, -0.5246383547782898, 0.008749074302613735, -0.09246595203876495, 0.3225846588611603, 0.09169075638055801, -0.11638537049293518, -0.39315295219421387, 0.17101822793483734, 0.03368808329105377, -0.1324877142906189, 0.27087679505348206, 0.4044116735458374, -0.03216136246919632, -0.21041572093963623, -0.17659389972686768, -0.24583308398723602, 0.15085183084011078, 0.4518739879131317, 0.13134147226810455, -0.24895279109477997, 0.2793442904949188, 0.0862233117222786, 0.1133275255560875, 0.10454671084880829, -0.16755004227161407, 0.2823275327682495, -0.027697620913386345, 0.20288464426994324, -0.0016181454993784428, -0.00043098884634673595, 0.07328079640865326, 0.07297271490097046, 0.2253543734550476, 0.5441485643386841, 0.14013344049453735, -0.41670721769332886, -0.3592550754547119, -0.3044956922531128, -0.03829644247889519, -0.0881940945982933, -0.05756686255335808, 0.2055095136165619, 0.35952043533325195, -0.028068920597434044, -0.19928358495235443, -0.4128097891807556, -0.20246312022209167, -0.012766271829605103, 0.23737101256847382, 0.004086978267878294, -0.06508511304855347, 0.0963447242975235, 0.03865635395050049, -0.4979262351989746, 0.381961852312088, 0.11513081938028336, 0.07316949218511581, 0.03802061080932617, -0.0845394879579544, 0.06384734064340591, -0.10432152450084686, 0.18705981969833374, -0.11142297834157944, 0.00545673631131649, -0.05644489452242851, -0.1029587984085083, -0.17706289887428284, 0.0564187653362751, -0.378465473651886, -0.19693315029144287, 0.07580850273370743, -0.009619357995688915, -0.0009150568512268364, -0.1784559190273285, 0.6600261926651001, -0.08007243275642395, -0.0571611151099205, -0.19711542129516602, -0.012537786737084389, -0.41267839074134827, -0.07509077340364456, -0.003672925755381584, -0.15633085370063782, 0.2179160714149475, -0.12022332847118378, -0.0820920467376709, 0.17336654663085938, -0.017646921798586845, 0.1186627447605133, 0.28284525871276855, -0.07113151997327805, 0.12168458104133606, -0.015217063017189503, -0.025802332907915115, 0.07832583039999008, 0.11322902143001556, 0.11903863400220871, 0.18738578259944916, -0.06267867237329483, 0.050076380372047424, -0.05787045881152153, 0.29155218601226807, 0.18741926550865173, 0.09294849634170532, 0.022943777963519096, -0.3587073087692261, -0.0909990593791008, -0.08881597965955734, -0.006468827370554209, 0.28515785932540894, 0.23027543723583221, -0.2992843985557556, 0.018360573798418045, 0.40848520398139954, 0.11278790980577469, -0.2255142331123352, 0.1382581740617752, 0.03543655201792717, -0.4326167404651642, 0.5216684937477112, -0.04760744422674179, 0.6814640164375305, 0.26648518443107605, 0.0666789710521698, 0.40899035334587097, 0.019803138449788094, 0.619162917137146, -0.12821847200393677, 0.0904494896531105, 0.008613606914877892, -0.074935182929039, -0.02831864170730114, 0.25809162855148315, 0.28022435307502747, -0.14534054696559906, -0.3026863634586334, 0.35803526639938354, 0.47414830327033997, -0.02822880633175373, 0.06608006358146667, 0.19065892696380615, 0.26750096678733826, -0.3572765290737152, -0.17591634392738342, 0.19538570940494537, -0.057400405406951904, 0.1983063668012619, -0.009514211677014828, -0.07632476836442947, -0.03352508693933487, -0.34462064504623413, -0.09483600407838821, -0.004620157182216644, 0.4087621867656708, 0.06272778660058975, -0.013134665787220001, 0.00018247614207211882, 0.3545762300491333, -0.251066654920578, 0.20681533217430115, 0.12703624367713928, -0.23320059478282928, 0.1341613531112671, 0.030415287241339684, 0.0334748774766922, -0.05756489187479019, 0.06580767035484314, 0.17728284001350403, -0.21683083474636078, -0.0924762710928917, 0.13309335708618164, 0.2801991105079651, -0.22764967381954193, 0.07511942088603973, -0.05992459878325462, -0.0967978835105896, 0.09913453459739685, 0.27097418904304504, 0.5199607014656067, 0.042326949536800385, -0.29324328899383545, 0.2494107484817505, 0.21730072796344757, -0.24735428392887115, 0.125958651304245, 0.1490776687860489, -0.1499183624982834, -0.12614017724990845, 0.42955100536346436, 0.32879096269607544, -0.1759682446718216, 0.16236115992069244, 0.028476305305957794, -0.06598140299320221, -0.2654573321342468, -0.24508419632911682, -0.09568382799625397, -0.5750572085380554, -0.06528225541114807, 0.11142846941947937, -0.18903671205043793, 0.20219171047210693, 0.7269651889801025, -0.03920881450176239, 0.20746353268623352, -0.10957226902246475, -0.2568225562572479, -0.2760741412639618, 0.05067762732505798, -0.4521556496620178, 0.22447267174720764, 0.14675383269786835, 0.057489413768053055, -0.057873133569955826, 0.08658542484045029, -0.4668600559234619, -0.22374427318572998, 0.07354415953159332, -0.011188030242919922, 0.36665621399879456, 0.08408282697200775, 0.33175262808799744, 0.07002580910921097, 0.2591492235660553, 0.017092356458306313, -0.18132847547531128, -0.27145421504974365, -0.12523697316646576, 0.07885976135730743, 0.02188052423298359, 0.17113512754440308, 0.09806942939758301, -0.05675091966986656, -0.2730421721935272, -0.059700340032577515, 0.1882498860359192, -0.08539959788322449, -0.186428040266037, 0.3759686350822449, -0.2097957283258438, -0.1313132792711258, 0.3740188777446747, 0.299776554107666, 0.2019176334142685, 0.013176181353628635, -0.10522720217704773, 0.4919646680355072, -0.06271221488714218, -0.10722258687019348, 0.04128642380237579, 0.19946125149726868, -0.21136514842510223, 0.1068032905459404, 0.08347336202859879, 0.0005004917620681226, -0.2596767544746399, -0.20306354761123657, 0.21257968246936798, -0.2958911955356598, 0.015859710052609444, 0.06989355385303497, 0.07521281391382217, 0.3328958749771118, -0.43360835313796997, 0.15516604483127594, -0.06994188576936722, -0.1500576138496399, 0.3807324171066284, 0.16599556803703308, 0.2593381404876709, -0.14491227269172668, 0.6386488676071167, -0.011185579001903534, -0.019398722797632217, 0.06352221965789795, 0.2700665295124054, 0.007332423236221075, 0.08318693935871124, 0.007492111064493656, 0.06729616224765778, 0.0025866348296403885, 0.26146379113197327, -0.024469969794154167, 0.043507665395736694, -0.026396358385682106, 0.1955518275499344, 0.3980350196361542, -0.004164998885244131, -0.21858011186122894, -0.04063577577471733, 0.2692480981349945, 0.1335304230451584, 0.22033314406871796, 0.12551972270011902, 0.013996214605867863, -0.026474546641111374, 0.07339495420455933, -0.6518138647079468, 0.10224496573209763, 0.04633382707834244, -0.267489492893219, -0.192404106259346, -0.060147058218717575, -0.5283677577972412, -0.03477143496274948, 0.16022390127182007, -0.09570473432540894, 0.11657214164733887, -0.05166498199105263, 0.07509466260671616, -0.17832209169864655, -0.06448644399642944, 0.1521919071674347, 0.20974574983119965, -0.07242792099714279, 0.2600595951080322, -0.31430959701538086, -0.13356724381446838, 0.11316914856433868, 0.3317931294441223, 0.2536149322986603, -0.07983924448490143, 0.25138118863105774, 0.1893155574798584, 0.0781901404261589, -0.06341107189655304, 0.015542170032858849, 0.12749645113945007, 0.17518633604049683, 0.15679244697093964, 0.34565526247024536, 0.19967436790466309, -0.2353813648223877, 0.10904031991958618, 0.018087051808834076, -0.18517586588859558, -0.18787141144275665, 0.29476144909858704, 0.25069111585617065, 0.12793800234794617, -0.1549278348684311, 0.056655216962099075, -0.2226928174495697, -0.3151545226573944, -0.05655490607023239, -0.04162634164094925, -0.06579135358333588, -0.26920464634895325, 0.14210392534732819, -0.10145176202058792, 0.24289849400520325, 0.20207805931568146, 0.30529624223709106, 0.07524057477712631, -0.43133583664894104, -0.1671849638223648, 0.21204809844493866, 0.18553702533245087, -0.23350784182548523, -0.04269701987504959, 0.16694140434265137, 0.14643624424934387, -0.24126248061656952, 0.10201063752174377, -0.13331879675388336, 0.4249040186405182, -0.001524023013189435, -0.39028066396713257, 0.1590537279844284, 0.14354099333286285, -0.09171497076749802, -0.037332337349653244, -0.2592243552207947, -0.30493348836898804, -0.13834229111671448, 0.1372605562210083, -0.2919715940952301, 0.0967421680688858, 0.006600793916732073, 0.13372188806533813, 0.4191933572292328, 0.02971341274678707, 0.277936726808548, -0.33394157886505127, -0.006887862458825111, -0.3844676911830902, -0.25012093782424927, -0.06227722018957138, 0.04657304286956787, 0.0962267741560936, 0.1779748648405075, 0.020027192309498787, 0.1866796761751175, 0.04949426278471947, 0.39270496368408203, 0.03392896428704262, -0.2314755618572235, -0.04233105108141899, 0.5437836050987244, -0.3518792986869812, -0.20718631148338318, -0.07825575023889542, -0.004872821271419525, 0.039167728275060654, -0.13390211760997772, -0.05519884452223778, -0.30653658509254456, 0.3257247805595398, -0.3351879417896271, -0.05862843617796898, -0.07195766270160675, 0.3476722240447998, 0.15555118024349213, -0.006516837049275637, -0.33956602215766907, 0.16316469013690948, 0.14248697459697723, 0.16235297918319702, -0.03403221070766449, 0.3986179828643799, -0.14811229705810547, -0.01205473206937313, -0.21766532957553864, -0.16810859739780426, -0.06432843208312988, -0.08481377363204956, 0.13156437873840332, -0.43430906534194946 ]
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).
[ -0.22634349763393402, -0.28592008352279663, -0.11841712892055511, -0.49191898107528687, -0.4530395567417145, -0.2508593499660492, 0.21630561351776123, 0.22147013247013092, -0.03563842177391052, 0.3164645731449127, 0.07013283669948578, 0.4571438133716583, 0.03245719522237778, 0.3533177971839905, -0.1624484360218048, 0.4214377701282501, 0.11718575656414032, 0.2206689715385437, -0.030111903324723244, -0.255616694688797, -0.2271455079317093, 0.13091059029102325, -0.1685999035835266, -0.12527340650558472, 0.16298635303974152, 0.2887752950191498, -0.2544238865375519, -0.09574782103300095, -0.0414617545902729, -0.1833728700876236, 0.40287676453590393, 0.34786322712898254, 0.17024527490139008, 0.06462860107421875, -0.00011004399857483804, -0.13638828694820404, 0.45679688453674316, -0.0639549046754837, -0.27537044882774353, -0.1536129117012024, -0.3556925654411316, -0.36505308747291565, 0.07075218111276627, -0.07159381359815598, 0.21727681159973145, 0.16272689402103424, 0.14945925772190094, -0.031100958585739136, 0.04020832106471062, 0.1662721037864685, 0.26914650201797485, 0.28607985377311707, -0.031676582992076874, -0.33285123109817505, 0.21256211400032043, 0.47362667322158813, -0.36004501581192017, -0.07488920539617538, 0.04369533807039261, -0.09709255397319794, -0.14182771742343903, 0.38200196623802185, 0.22545598447322845, -0.17204232513904572, -0.26014798879623413, -0.03067469410598278, 0.21465106308460236, 0.1305542141199112, 0.26056674122810364, 0.056902796030044556, 0.3529529869556427, 0.13865041732788086, -0.264924019575119, 0.25002291798591614, -0.05883804336190224, -0.42920053005218506, 0.22503983974456787, -0.1623154729604721, 0.02110522985458374, 0.0047462391667068005, 0.3541201949119568, -0.18666324019432068, -0.5334498286247253, 0.15618234872817993, -0.21667078137397766, 0.17143075168132782, -0.2124953716993332, -0.14087367057800293, 0.09421994537115097, 0.10761155933141708, -0.09298378229141235, 0.11195176839828491, -0.2070818394422531, 0.18042755126953125, -0.34744325280189514, -0.10264928638935089, 0.3085196018218994, -0.06392962485551834, 0.42677056789398193, 0.15892189741134644, 0.19507107138633728, 0.1905672550201416, -0.01569860801100731, 0.15070199966430664, 0.011564132757484913, 0.368662565946579, 0.5664451718330383, 0.06401557475328445, 0.4030506908893585, 0.14254871010780334, 0.4047582149505615, 0.016547903418540955, 0.07813003659248352, 0.021182335913181305, -0.0022472194395959377, -0.16156648099422455, 0.11037593334913254, -0.5184800624847412, -0.18017198145389557, -0.03196050599217415, -0.1015343889594078, 0.09271789342164993, 0.004587937146425247, 0.2791183292865753, 0.2888186275959015, 0.19537875056266785, 0.3214777112007141, -0.02211964875459671, 0.016755197197198868, -0.28218892216682434, -0.2884853184223175, -0.08126705139875412, -0.24854698777198792, 0.20286910235881805, 0.1990642100572586, -0.3987305760383606, 0.20908957719802856, -0.36445584893226624, 0.40940061211586, -0.06413407623767853, -0.2289317101240158, 0.3046197295188904, -0.03571649640798569, 0.2634499967098236, 0.038811665028333664, -0.017299367114901543, -0.09945603460073471, 0.3986690938472748, -0.16598185896873474, 0.11357780545949936, -0.21756260097026825, -0.48887714743614197, -0.26995527744293213, 0.2650912404060364, -0.1804889440536499, -0.2526894211769104, -0.24294321238994598, 0.4173618257045746, -0.24704475700855255, -0.007607513107359409, 0.013684321194887161, 0.14105524122714996, -0.3214115500450134, -0.20329101383686066, -0.08240274339914322, 0.19734489917755127, -0.04359940066933632, -0.08860617876052856, -0.6579611897468567, -0.09929916262626648, 0.01547541469335556, -0.18410451710224152, -0.0183941051363945, -0.30888548493385315, -0.24293756484985352, -0.09758492559194565, 0.40919962525367737, -0.540909469127655, -0.08889667689800262, 0.3007030785083771, -0.05289776250720024, -0.21085965633392334, 0.1997181475162506, -0.30009976029396057, -0.15993176400661469, -0.28480520844459534, -0.311284601688385, 0.0520436055958271, -0.03445470333099365, -0.05105556920170784, -0.0711141973733902, -0.34644243121147156, -0.3489779531955719, 0.258571058511734, 0.3637619912624359, 0.03293479606509209, 0.3567889928817749, -0.4188338816165924, 0.1504986733198166, -0.05560458078980446, 0.13320352137088776, -0.025527598336338997, 0.6704122424125671, 0.1037113294005394, 0.19556774199008942, 0.13810685276985168, -0.3721383810043335, 0.0969766154885292, 0.1827079802751541, 0.15224672853946686, 0.20965880155563354, -0.36246708035469055, -0.15044844150543213, 0.1442914754152298, -0.006205354817211628, -0.19019721448421478, 0.18903611600399017, 0.16746042668819427, -0.06673936545848846, -0.0394475944340229, -0.0403214693069458, 0.41204607486724854, -0.33331194519996643, -0.0893506184220314, 0.16916677355766296, 0.06708631664514542, -0.05909042805433273, -0.007216480560600758, -0.05223170295357704, 0.09109999239444733, -0.0032765318173915148, 0.008714244700968266, -0.151804581284523, 0.39785104990005493, -0.014699592255055904, 0.13733486831188202, 0.07569364458322525, 0.2027662694454193, 0.12991935014724731, -0.11296960711479187, 0.01564052328467369, 0.0311867855489254, -0.06054399162530899, 0.10886549204587936, -0.21376876533031464, 0.3678346872329712, -0.044001463800668716, 0.07868467271327972, -0.003053507534787059, -0.08954411745071411, 0.026837356388568878, -0.07286697626113892, -0.5226488709449768, -0.21800456941127777, 0.24215558171272278, 0.10695771127939224, -0.4233154058456421, -0.23976384103298187, -0.2204616665840149, -0.08050723373889923, 0.17624559998512268, -0.10011545568704605, 0.15004439651966095, 0.3162190914154053, 0.1081685870885849, -0.1911517232656479, 0.2767573893070221, -0.1665320098400116, 0.40634220838546753, 0.31514105200767517, -0.10612723976373672, -0.04802081361413002, -0.30542609095573425, -0.053228311240673065, 0.13648007810115814, 0.2142147421836853, 0.12367849797010422, 0.12573492527008057, 0.4169749319553375, -0.14952272176742554, -0.5176542401313782, -0.41359326243400574, -0.0627230778336525, -0.1665928214788437, -0.19666177034378052, -0.16656166315078735, -0.06399580836296082, 0.09422239661216736, -0.10740865021944046, 0.03863611817359924, 0.08334238827228546, -0.236199289560318, 0.2636566758155823, 0.22466324269771576, -0.061354249715805054, 0.23576240241527557, 0.025401392951607704, 0.6541306376457214, -0.022526388987898827, 0.13648198544979095, -0.15872693061828613, 0.06660047918558121, -0.4203682243824005, 0.19410811364650726, -0.5060151219367981, -0.22600549459457397, 0.43392136693000793, -0.007591398898512125, 0.30970650911331177, -0.060099489986896515, -0.693355143070221, 0.07365769147872925, -0.19269123673439026, -0.1378568857908249, 0.06659690290689468, -0.288351833820343, -0.11153613030910492, 0.04072213172912598, 0.08749247342348099, -0.39441975951194763, -0.35635921359062195, -0.08151999861001968, -0.23026946187019348, 0.27909964323043823, -0.3458874225616455, -0.4851737320423126, 0.23486746847629547, -0.33899596333503723, -0.11091107130050659, -0.08392266184091568, 0.1287865787744522, 0.5446904897689819, 0.13814017176628113, 0.08742541819810867, 0.12513822317123413, -0.014643410220742226, -0.4106563627719879, 0.23893669247627258, 0.34010204672813416, -0.3349621295928955, -0.40006372332572937, 0.10895273089408875, 0.09359128773212433, 0.15444834530353546, 0.2097408026456833, -0.18663428723812103, -0.5588260889053345, 0.18018373847007751, 0.15571695566177368, -0.20452992618083954, 0.1585502028465271, 0.3840808570384979, -0.01926814578473568, -0.19568859040737152, -0.1868504136800766, -0.3143168091773987, 0.22574366629123688, 0.3081618547439575, 0.12987065315246582, -0.16425448656082153, 0.26787999272346497, 0.16116441786289215, 0.14897942543029785, 0.08203332126140594, 0.01912017911672592, 0.050538934767246246, 0.1108473613858223, 0.351094514131546, -0.025693649426102638, 0.055278800427913666, 0.1333310902118683, 0.0819227546453476, 0.20691190659999847, 0.4673640727996826, 0.1389228105545044, -0.2309902012348175, -0.2874262034893036, -0.2523173391819, 0.16300897300243378, 0.031660858541727066, -0.16662479937076569, 0.1882852464914322, 0.43356141448020935, -0.12905941903591156, -0.09658844023942947, -0.3269830346107483, 0.032860737293958664, 0.06107475981116295, 0.42140087485313416, 0.09682809561491013, -0.17435337603092194, -0.004774642642587423, 0.14035235345363617, -0.32818493247032166, 0.3598346710205078, 0.10829276591539383, 0.051345694810152054, -0.0035427461843937635, -0.20926474034786224, 0.21159036457538605, -0.029142966493964195, 0.4310411810874939, -0.1808997541666031, 0.013701523654162884, -0.11118058115243912, -0.07789874076843262, -0.09302481263875961, -0.16937604546546936, -0.11786647886037827, -0.3060426115989685, -0.03530876711010933, -0.03556106239557266, -0.05281476303935051, -0.029260344803333282, 0.576262891292572, -0.26637834310531616, 0.07449094206094742, -0.2874302566051483, -0.20820041000843048, -0.4295230507850647, -0.10088329762220383, 0.00454550888389349, -0.05962261185050011, -0.0953708291053772, -0.2208271473646164, -0.13561567664146423, 0.2697882056236267, -0.2274513691663742, -0.14196445047855377, 0.12684495747089386, -0.2542560398578644, -0.0007774035912007093, -0.03305470198392868, 0.10001487284898758, -0.11615116894245148, 0.3154968321323395, 0.016632242128252983, 0.2869827449321747, -0.3732561469078064, 0.10811784863471985, -0.18032854795455933, 0.2171064019203186, 0.27198326587677, 0.001340661896392703, 0.14543330669403076, -0.4411616325378418, -0.2380370944738388, -0.013946136459708214, -0.08445221185684204, 0.2103116810321808, -0.004943088628351688, -0.20481328666210175, -0.04624127224087715, 0.39969974756240845, 0.05420341342687607, -0.2666933834552765, 0.27243226766586304, 0.14732800424098969, -0.28837451338768005, 0.44838595390319824, 0.01125964056700468, 0.6613936424255371, 0.2898109555244446, -0.21789254248142242, 0.29653027653694153, -0.21648351848125458, 0.5671088695526123, -0.18582509458065033, 0.13965469598770142, 0.0983457937836647, -0.1416415274143219, 0.024118805304169655, 0.2629130184650421, 0.3644730746746063, -0.04198655113577843, -0.32919684052467346, 0.24835875630378723, 0.21675218641757965, 0.3008902072906494, 0.212734192609787, 0.11631892621517181, 0.08536699414253235, -0.30704614520072937, 0.0013403140474110842, 0.1875009834766388, 0.08409211039543152, 0.270291268825531, -0.041175663471221924, 0.09877128154039383, 0.02851025015115738, -0.3716997802257538, -0.22798888385295868, 0.03163028508424759, 0.4609609842300415, 0.1189180314540863, 0.012707741931080818, 0.06544401496648788, 0.2390555739402771, -0.32361093163490295, 0.2711414098739624, 0.10313189774751663, -0.08830984681844711, 0.29943177103996277, 0.3193259537220001, 0.01049504242837429, -0.17676083743572235, -0.15102694928646088, 0.02432101219892502, -0.15881365537643433, -0.2020001858472824, 0.22722779214382172, 0.17419664561748505, -0.23866313695907593, -0.20172902941703796, 0.03390911594033241, -0.054471634328365326, 0.1423092633485794, 0.5334224700927734, 0.43454352021217346, 0.10620968788862228, -0.1701265126466751, 0.2046642154455185, 0.1269294172525406, -0.11475131660699844, 0.0031178707722574472, 0.15291020274162292, -0.19069091975688934, -0.18325136601924896, 0.5322156548500061, 0.2791721224784851, -0.13204549252986908, -0.06292683631181717, -0.11571285128593445, 0.016296280547976494, -0.21505871415138245, -0.18904048204421997, 0.14092501997947693, -0.39371079206466675, -0.12774966657161713, 0.16813978552818298, -0.2827528119087219, 0.24938488006591797, 0.809527575969696, 0.07201100885868073, 0.2018945962190628, -0.22826363146305084, -0.12816613912582397, -0.025408152490854263, -0.07955557852983475, -0.47719627618789673, 0.18120895326137543, -0.020505890250205994, -0.05071517825126648, -0.08789913356304169, -0.012811357155442238, -0.3956335484981537, -0.24205562472343445, 0.1952885389328003, -0.09695495665073395, 0.40214818716049194, 0.2122165858745575, 0.4604620635509491, 0.18588772416114807, 0.2123551070690155, 0.015105297788977623, -0.2088433802127838, -0.17874522507190704, -0.01365647278726101, 0.10726439952850342, 0.08703921735286713, 0.2987453043460846, -0.19903306663036346, -0.18296363949775696, -0.2923518121242523, 0.022287430241703987, 0.33509308099746704, 0.10991951078176498, -0.19674448668956757, 0.08827044069766998, -0.45101413130760193, 0.010249345563352108, 0.3422631621360779, 0.18725669384002686, 0.17638689279556274, 0.09170447289943695, -0.06608203798532486, 0.37622925639152527, -0.12612789869308472, -0.026595162227749825, 0.1228456124663353, 0.3449654281139374, -0.2599084973335266, 0.12928426265716553, 0.008575891144573689, 0.08216602355241776, -0.19888155162334442, -0.17354613542556763, 0.0164008941501379, -0.3018588721752167, 0.006897104438394308, -0.05663783848285675, -0.16726353764533997, 0.30348315834999084, -0.239481121301651, 0.06966462731361389, -0.27626943588256836, -0.16598942875862122, 0.20793922245502472, 0.2772291898727417, 0.2561793625354767, -0.09345095604658127, 0.7590213418006897, 0.10688679665327072, 0.14143723249435425, -0.07472973316907883, 0.22079786658287048, 0.16938231885433197, 0.0426347479224205, 0.1686861217021942, 0.2758140563964844, -0.08789416402578354, 0.3604968786239624, -0.0775500163435936, 0.30085062980651855, 0.036803945899009705, 0.2739917039871216, 0.5524517893791199, 0.05506047606468201, -0.31329724192619324, 0.10778644680976868, 0.4623662233352661, 0.051164962351322174, 0.05843339487910271, 0.1863643229007721, -0.22566872835159302, -0.020918242633342743, 0.18213814496994019, -0.7633130550384521, -0.17677104473114014, 0.0875551849603653, -0.2771151065826416, -0.19742587208747864, -0.10568828880786896, -0.47327932715415955, -0.006776300258934498, 0.29948869347572327, 0.0507827065885067, -0.016121232882142067, -0.20815593004226685, 0.00819282978773117, -0.06037769839167595, -0.2778101861476898, 0.07377651333808899, 0.2569637596607208, 0.006414117757230997, 0.20321528613567352, -0.4178900122642517, -0.13610754907131195, 0.3770047724246979, 0.1393405944108963, 0.13725171983242035, -0.08849239349365234, 0.5053784847259521, 0.11966430395841599, 0.10259051620960236, -0.09160067141056061, -0.009294182062149048, 0.18863531947135925, 0.1232810765504837, 0.1045759841799736, 0.18037033081054688, 0.19206105172634125, -0.18656572699546814, 0.15649479627609253, -0.03523559123277664, -0.24298477172851562, -0.24431516230106354, 0.18434305489063263, 0.10635493695735931, 0.15036258101463318, -0.08765791356563568, 0.22046947479248047, 0.032329436391592026, -0.3462272882461548, -0.06211283802986145, 0.02057473175227642, -0.1276412159204483, -0.14604659378528595, 0.11986333876848221, -0.06510601937770844, -0.12488351762294769, 0.16994766891002655, -0.11498904973268509, 0.06258691847324371, -0.3730802536010742, -0.004313156008720398, 0.36594322323799133, 0.18773028254508972, 0.07066261768341064, 0.05478579178452492, 0.14213618636131287, 0.1427461951971054, -0.2538003921508789, 0.12222755700349808, -0.43262815475463867, 0.5907478928565979, -0.08773446828126907, -0.1828019767999649, 0.1407758891582489, 0.4150145947933197, -0.11227630823850632, -0.1277855485677719, -0.2960691452026367, -0.21165794134140015, 0.06730363517999649, 0.10888475179672241, -0.33292505145072937, -0.12729686498641968, 0.18393932282924652, 0.037889670580625534, 0.44412943720817566, 0.04809120297431946, 0.21306101977825165, -0.1285381019115448, -0.07115907967090607, -0.5014456510543823, -0.13466621935367584, -0.023916104808449745, 0.09259093552827835, -0.13384462893009186, -0.029838230460882187, 0.043160393834114075, 0.1195140928030014, 0.02203390933573246, 0.3425886631011963, 0.046129900962114334, -0.09075770527124405, -0.0950559601187706, 0.42447608709335327, -0.5147945284843445, -0.30801668763160706, -0.10602685809135437, -0.10277251899242401, 0.06682411581277847, -0.12948055565357208, -0.04089654982089996, -0.24251851439476013, 0.3495895564556122, -0.22609186172485352, -0.08503812551498413, -0.0861118957400322, 0.4343441426753998, 0.20386965572834015, 0.08677313476800919, -0.22049081325531006, 0.09851104021072388, 0.22752763330936432, 0.19771930575370789, -0.27606436610221863, 0.1508471816778183, -0.044821642339229584, 0.01777292974293232, -0.22074343264102936, 0.02358340285718441, -0.17097176611423492, 0.015043104998767376, 0.15334273874759674, -0.32380443811416626 ]
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)
[ -0.16037388145923615, -0.1905597299337387, -0.04093000665307045, -0.5251522064208984, -0.5301500558853149, -0.20913481712341309, 0.07916968315839767, 0.3027108311653137, -0.08486776798963547, 0.2993153929710388, 0.09133685380220413, 0.5883415937423706, -0.12738902866840363, 0.3007548749446869, 0.14482663571834564, 0.19229452311992645, 0.0493808351457119, 0.3573557734489441, 0.016091464087367058, -0.14633303880691528, -0.28475213050842285, 0.2668963372707367, -0.15900416672229767, 0.02206628955900669, 0.17621323466300964, 0.18231701850891113, -0.3217965066432953, -0.02632552944123745, -0.06103837117552757, -0.2656818926334381, 0.4338560104370117, 0.29688435792922974, 0.24196375906467438, -0.1544303297996521, -0.00011306988017167896, -0.24875323474407196, 0.5088642835617065, -0.062202952802181244, -0.429745614528656, -0.3075883984565735, -0.3339640498161316, -0.518495500087738, 0.16902515292167664, -0.09877879917621613, 0.1858435422182083, 0.1506466567516327, 0.3687630593776703, -0.03254236280918121, -0.0012722763931378722, 0.1081598699092865, 0.22739464044570923, 0.28239166736602783, -0.08409292250871658, -0.29824087023735046, 0.11302664130926132, 0.3750302493572235, -0.30863311886787415, -0.16344061493873596, -0.04228092357516289, -0.2713524401187897, -0.05244668200612068, 0.3673297166824341, 0.2016817331314087, -0.2252694070339203, -0.1958562433719635, -0.04784805700182915, 0.14524711668491364, 0.04428547993302345, 0.07859806716442108, 0.09422238171100616, 0.34997421503067017, 0.04661249741911888, -0.2632305324077606, 0.04236666113138199, 0.07684119790792465, -0.596377432346344, 0.28150659799575806, -0.06955305486917496, -0.1712326854467392, -0.07540766149759293, 0.2659726142883301, -0.2741575837135315, -0.42526859045028687, 0.3804297149181366, -0.06080375611782074, 0.2607448697090149, -0.06658735126256943, -0.06787658482789993, 0.23503398895263672, 0.08089529722929001, -0.10468418151140213, 0.027629202231764793, -0.14414867758750916, 0.2545183002948761, -0.4161166846752167, -0.14212217926979065, 0.40127938985824585, 0.04141957312822342, 0.2769641876220703, 0.14983652532100677, 0.2790851891040802, 0.01285003125667572, -0.05144145339727402, 0.18539775907993317, 0.08551684021949768, 0.31466537714004517, 0.6022413969039917, 0.028707554563879967, 0.26106035709381104, 0.09194567054510117, 0.5128620862960815, 0.1002446636557579, 0.07537593692541122, 0.020593984052538872, -0.16388972103595734, -0.12788309156894684, 0.11135813593864441, -0.3812929689884186, -0.12265577167272568, -0.09454666078090668, -0.09384271502494812, 0.06300088763237, 0.02480912022292614, 0.19082994759082794, 0.2591688632965088, 0.2565682530403137, 0.34428563714027405, -0.0727609172463417, -0.1324228197336197, -0.38469770550727844, -0.24910196661949158, -0.05333348736166954, -0.3931252360343933, 0.27025628089904785, 0.35871729254722595, -0.23135314881801605, 0.28058430552482605, -0.4026280641555786, 0.19323459267616272, 0.016766542568802834, -0.2188311368227005, 0.3100361227989197, -0.1656475067138672, 0.07989101856946945, 0.046882424503564835, 0.01372385025024414, -0.30024245381355286, 0.26143157482147217, -0.2551455497741699, 0.09527451545000076, -0.2514301836490631, -0.5217939019203186, -0.2898750901222229, 0.2131015658378601, -0.21021588146686554, -0.339721143245697, -0.16480211913585663, 0.4888755679130554, -0.22434598207473755, -0.2618385851383209, -0.019775930792093277, 0.07405801862478256, -0.371123731136322, -0.1326882541179657, -0.08441714197397232, 0.1536417454481125, 0.03227527067065239, -0.2395111471414566, -0.6589330434799194, 0.023701194673776627, 0.050404760986566544, -0.10603699088096619, -0.10561704635620117, -0.2657807767391205, -0.14965106546878815, -0.05989725887775421, 0.722976565361023, -0.4373762309551239, -0.11550039798021317, 0.2689363956451416, -0.06760767102241516, -0.2052617222070694, 0.10484587401151657, -0.13712166249752045, -0.23001404106616974, -0.15859808027744293, -0.1731010228395462, 0.25630760192871094, -0.05131038278341293, -0.10487193614244461, -0.08494435250759125, -0.3723754286766052, -0.08402247726917267, 0.24752339720726013, 0.2598855197429657, 0.01023820973932743, 0.26540133357048035, -0.17162631452083588, 0.20318880677223206, -0.06941086798906326, 0.14847148954868317, 0.03821396082639694, 0.44097697734832764, 0.04296781122684479, 0.09988930821418762, 0.10250699520111084, -0.5050916075706482, 0.07856345176696777, -0.04806401953101158, 0.2747190594673157, 0.32303062081336975, -0.21497687697410583, -0.04981773719191551, -0.05138131231069565, 0.06278396397829056, -0.24295863509178162, 0.12956157326698303, 0.2916697561740875, -0.04538189247250557, -0.02740434557199478, -0.08810258656740189, 0.4771677553653717, -0.21470007300376892, -0.1701606661081314, -0.22500821948051453, 0.07947145402431488, -0.13179583847522736, -0.03842926025390625, 0.001159653882496059, 0.3066427409648895, -0.00575024401769042, 0.031460970640182495, -0.0858968123793602, 0.3410399258136749, -0.1707218438386917, 0.2327120155096054, 0.0020120181143283844, 0.051782723516225815, 0.15532894432544708, -0.17103660106658936, 0.07149914652109146, -0.038749124854803085, -0.0591483972966671, 0.17416730523109436, -0.23587451875209808, 0.40878504514694214, 0.055551592260599136, -0.0049431161023676395, 0.015664083883166313, -0.025185374543070793, 0.02458135411143303, -0.21304407715797424, -0.5066877007484436, -0.16084517538547516, 0.3907112777233124, 0.09037645906209946, -0.41927769780158997, -0.2182738035917282, -0.42194899916648865, -0.05391254648566246, 0.26591917872428894, 0.010921400040388107, 0.06860531121492386, 0.34612032771110535, 0.19898222386837006, -0.25830382108688354, 0.28767600655555725, -0.033756136894226074, 0.5213518142700195, 0.3559092879295349, -0.14374245703220367, 0.055378805845975876, -0.4970718324184418, -0.034334201365709305, 0.19374391436576843, 0.22271728515625, 0.16498343646526337, 0.0345129668712616, 0.47993749380111694, 0.024719323962926865, -0.41517606377601624, -0.4975740611553192, 0.03008289635181427, -0.06677840650081635, -0.1709928810596466, -0.07678594440221786, -0.23566098511219025, -0.2485530525445938, -0.1343785971403122, 0.1527375876903534, -0.006642441730946302, -0.28226959705352783, 0.2530544400215149, 0.22281832993030548, -0.0968485102057457, 0.27046480774879456, -0.08819720149040222, 0.7165157794952393, -0.17323870956897736, 0.13250192999839783, -0.21116554737091064, -0.01973300613462925, -0.43862268328666687, 0.14359325170516968, -0.3673212230205536, -0.12818694114685059, 0.4306376278400421, 0.03370533138513565, 0.24169108271598816, 0.10172418504953384, -0.656622052192688, 0.07370667904615402, -0.16742001473903656, -0.0849752128124237, 0.07531515508890152, -0.3813069462776184, -0.17965354025363922, -0.09716573357582092, 0.09173287451267242, -0.34687796235084534, -0.3567788004875183, -0.0608406737446785, -0.17782975733280182, 0.19364377856254578, -0.3861486613750458, -0.5043302178382874, -0.002892016666010022, -0.38540250062942505, -0.06033296138048172, -0.03287303075194359, 0.2645634412765503, 0.7287463545799255, -0.02927178516983986, 0.06091081723570824, -0.018814289942383766, 0.05397895351052284, -0.3222584128379822, 0.41111984848976135, 0.24176251888275146, -0.3663308918476105, -0.32564133405685425, -0.09408039599657059, -0.009010884910821915, 0.2759546935558319, 0.13772441446781158, -0.20072083175182343, -0.37605276703834534, 0.1728864163160324, 0.09563286602497101, -0.1895500123500824, 0.16957533359527588, 0.49473246932029724, 0.035685282200574875, -0.14348937571048737, -0.11991340667009354, -0.39781278371810913, 0.22548769414424896, 0.5513763427734375, 0.3134695887565613, -0.09973730891942978, 0.237979456782341, 0.12890557944774628, 0.0546545647084713, 0.18605999648571014, -0.031547706574201584, 0.21331413090229034, 0.07510702311992645, 0.30275553464889526, -0.06819276511669159, -0.0488397553563118, 0.18620440363883972, 0.1191440299153328, 0.32899248600006104, 0.4975132346153259, 0.10180936753749847, -0.49265387654304504, -0.3609597086906433, -0.24205853044986725, 0.12460783123970032, 0.003417834872379899, -0.06890643388032913, 0.06073231250047684, 0.40878644585609436, 0.026039866730570793, -0.26158958673477173, -0.4135235548019409, -0.17632894217967987, -0.10904944688081741, 0.4381903409957886, 0.032993242144584656, 0.012865313328802586, -0.08266300708055496, 0.026466086506843567, -0.4243904948234558, 0.23217682540416718, 0.06501646339893341, 0.284856915473938, -0.05420095473527908, -0.13267427682876587, 0.11208400130271912, 0.08531039208173752, 0.26006919145584106, -0.2370326817035675, -0.12139769643545151, -0.01105495449155569, 0.12417022138834, -0.1962534338235855, -0.06323396414518356, -0.1607275903224945, -0.17356663942337036, 0.07299992442131042, 0.016527099534869194, 0.01241595670580864, -0.18408221006393433, 0.5894101858139038, -0.1919884830713272, 0.02772826887667179, -0.3377680778503418, -0.15445536375045776, -0.29206809401512146, -0.09544756263494492, -0.08875241875648499, -0.1869194507598877, 0.14225392043590546, -0.20191363990306854, -0.00533294165506959, 0.22868745028972626, -0.09486312419176102, -0.005437770392745733, 0.22884440422058105, -0.24832817912101746, 0.13451343774795532, -0.09508774429559708, 0.017488442361354828, 0.09789492189884186, 0.03977174684405327, -0.04592752456665039, 0.2397378534078598, -0.09042980521917343, 0.18363958597183228, 0.04171434044837952, 0.34280669689178467, 0.29112496972084045, 0.07612723857164383, 0.1502438336610794, -0.38553011417388916, -0.25055772066116333, -0.042446549981832504, -0.0807965025305748, 0.24254873394966125, -0.0665721446275711, -0.33067652583122253, -0.09095700830221176, 0.4912468194961548, 0.06609174609184265, -0.3306335508823395, 0.3240831196308136, 0.08614302426576614, -0.337843656539917, 0.40923941135406494, -0.12891997396945953, 0.749197244644165, 0.3950468599796295, 0.037740446627140045, 0.32539886236190796, -0.0398080013692379, 0.8740343451499939, -0.03195716068148613, 0.2768107056617737, 0.07144162803888321, 0.05260846018791199, 0.03373277559876442, 0.24986213445663452, 0.37605005502700806, -0.015048959292471409, -0.31148967146873474, 0.3749997913837433, 0.5791001319885254, 0.0683809444308281, 0.19329139590263367, 0.3513813316822052, 0.14927451312541962, -0.45527225732803345, -0.1680750846862793, 0.1207701787352562, -0.10806284844875336, 0.3852909803390503, -0.014131398871541023, -0.006578296888619661, 0.016940664499998093, -0.3202211260795593, -0.21057531237602234, 0.04029101878404617, 0.392930269241333, 0.05890288203954697, 0.07368319481611252, 0.02825707197189331, 0.4237455427646637, -0.23016299307346344, 0.1826673150062561, 0.16094546020030975, -0.2040279060602188, 0.16447226703166962, 0.13954250514507294, -0.018579889088869095, -0.07313830405473709, -0.1439746469259262, 0.08395533263683319, -0.17628918588161469, -0.2379300892353058, 0.10621494799852371, 0.19553440809249878, -0.2047402709722519, -0.11868378520011902, -0.007554020266979933, -0.038433391600847244, 0.24524562060832977, 0.2921876609325409, 0.5004428625106812, 0.10781090706586838, -0.2376682609319687, 0.17232926189899445, 0.09804809093475342, -0.10802943259477615, 0.0239799153059721, 0.20026245713233948, -0.15120664238929749, -0.1128026619553566, 0.5301059484481812, 0.21330489218235016, -0.18939799070358276, 0.10306416451931, -0.14518485963344574, 0.07328268885612488, -0.19599448144435883, -0.1548786759376526, -0.032985247671604156, -0.5144608616828918, -0.19930768013000488, 0.1846107840538025, -0.23774094879627228, 0.30920684337615967, 0.7213867902755737, 0.01655501313507557, 0.24946637451648712, -0.1088579073548317, -0.18847277760505676, -0.05346197262406349, 0.027733899652957916, -0.5827199220657349, 0.16098636388778687, 0.1386856585741043, 0.05973067507147789, -0.039665352553129196, 0.07866829633712769, -0.3642282783985138, -0.29292258620262146, 0.12223710119724274, 0.04758669063448906, 0.5287964344024658, 0.1722341775894165, 0.2284759283065796, 0.08470907062292099, 0.13575126230716705, -0.10030167549848557, -0.18807877600193024, -0.15801219642162323, -0.13643503189086914, 0.11905048787593842, 0.17177554965019226, 0.17103168368339539, -0.17962627112865448, -0.19826345145702362, -0.20716480910778046, -0.02411828748881817, 0.1777467131614685, -0.07091812044382095, -0.19365093111991882, 0.3347189128398895, -0.4681258499622345, -0.045132704079151154, 0.38918548822402954, 0.3572055399417877, 0.16981029510498047, 0.04913058504462242, -0.02882910519838333, 0.5353574156761169, -0.03685219958424568, -0.0511413998901844, -0.08615139871835709, 0.24315862357616425, -0.3270842432975769, 0.24757340550422668, -0.09059056639671326, -0.007488401606678963, -0.22181124985218048, -0.2281968742609024, -0.030340127646923065, -0.30786874890327454, 0.03573714196681976, -0.044831354171037674, 0.06395255029201508, 0.2379174679517746, -0.3148874342441559, 0.055222924798727036, -0.035948414355516434, -0.029131945222616196, 0.18411917984485626, 0.21877340972423553, 0.32171398401260376, -0.11545762419700623, 0.6415477991104126, -0.04418725520372391, 0.014910897240042686, -0.18163006007671356, 0.28369414806365967, 0.1905105710029602, 0.16893945634365082, 0.08880256861448288, 0.18636417388916016, 0.008465265855193138, 0.3330291509628296, -0.06245890632271767, 0.22023776173591614, -0.048584770411252975, 0.3439483344554901, 0.5547140836715698, 0.07117712497711182, -0.35833027958869934, -0.054302338510751724, 0.38938507437705994, 0.14579102396965027, 0.1239122599363327, 0.13414667546749115, 0.007062012795358896, -0.07100220769643784, 0.18138770759105682, -0.6420043110847473, 0.05338063836097717, 0.03196076303720474, -0.3519982695579529, -0.38255995512008667, -0.09260966628789902, -0.5104485750198364, -0.062152307480573654, 0.14173418283462524, -0.007492905016988516, 0.03294220194220543, -0.07868869602680206, 0.023906685411930084, -0.20594532787799835, -0.05458948016166687, 0.22706565260887146, 0.16826966404914856, -0.023408498615026474, 0.3689238429069519, -0.5310978889465332, -0.21128730475902557, 0.19069629907608032, 0.24770766496658325, 0.20929519832134247, -0.03900136426091194, 0.40585803985595703, 0.11202472448348999, 0.14556220173835754, -0.07076343894004822, -0.059819597750902176, 0.1610332727432251, 0.3544894754886627, 0.04199710860848427, 0.26132214069366455, 0.09808699786663055, -0.09979218244552612, 0.14007267355918884, 0.0434308685362339, -0.29908287525177, -0.21188056468963623, 0.32497167587280273, 0.1753530204296112, 0.18041376769542694, -0.02329993061721325, 0.37575462460517883, -0.13017165660858154, -0.2383156716823578, 0.08112339675426483, 0.10263878852128983, -0.108948715031147, -0.2417629063129425, 0.09944391995668411, -0.07116585969924927, 0.007066245656460524, 0.1841975748538971, -0.03055766224861145, 0.02113470435142517, -0.41406765580177307, -0.12865160405635834, 0.2505955100059509, 0.11373458057641983, -0.12676304578781128, 0.09774211049079895, 0.15977345407009125, 0.04668572172522545, -0.2836064100265503, -0.12935735285282135, -0.050163883715867996, 0.5626107454299927, -0.019448649138212204, -0.21125924587249756, 0.1509997844696045, 0.35848814249038696, -0.02639123797416687, -0.12816977500915527, -0.18867050111293793, -0.28754138946533203, 0.043899085372686386, -0.008159461431205273, -0.3241664469242096, 0.011525612324476242, 0.04369822517037392, 0.23319730162620544, 0.3404276967048645, 0.1010020449757576, 0.27757009863853455, -0.1553436666727066, -0.085016630589962, -0.5489782094955444, -0.18481045961380005, -0.056379616260528564, 0.06647171825170517, -0.09563827514648438, 0.052345070987939835, -0.03479740768671036, 0.343843013048172, -0.010122096166014671, 0.3170663118362427, -0.02081405371427536, -0.0981980413198471, -0.03269492834806442, 0.4011123478412628, -0.474880188703537, -0.15526585280895233, -0.021511351689696312, -0.017494842410087585, 0.03464554250240326, -0.06058502197265625, -0.07859868556261063, -0.2389971911907196, 0.3525122404098511, -0.16081716120243073, -0.04476692155003548, -0.020304448902606964, 0.42433613538742065, 0.25574564933776855, 0.07693671435117722, -0.25946563482284546, -0.01638062112033367, 0.21026985347270966, 0.1324426680803299, -0.13598944246768951, 0.25627580285072327, -0.15202336013317108, -0.025489507243037224, -0.20269417762756348, -0.10972960293292999, -0.11596863716840744, -0.15834087133407593, 0.11542036384344101, -0.34984612464904785 ]
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? :-)
[ -0.29844120144844055, 0.2715890109539032, -0.025401653721928596, -0.17361219227313995, 0.11133779585361481, -0.03249095380306244, 0.34969308972358704, 0.3046948313713074, 0.09639139473438263, 0.13277669250965118, 0.1502356380224228, 0.18139135837554932, -0.2963349521160126, 0.05446114391088486, 0.16967593133449554, 0.030123505741357803, -0.0018446823814883828, 0.2378808706998825, -0.10852643847465515, 0.2644939124584198, -0.1570071280002594, 0.025941213592886925, -0.08346940577030182, 0.03472060337662697, -0.007924212142825127, 0.1446915864944458, -0.15397828817367554, 0.015508268028497696, -0.3736419677734375, -0.9120109677314758, 0.4226648211479187, 0.08653827011585236, 0.18527039885520935, 0.40165120363235474, -0.00012305592827033252, -0.07805012911558151, 0.5848662257194519, -0.08126469701528549, -0.6614242196083069, -0.24752658605575562, 0.1799677312374115, -0.4401649534702301, 0.3309338092803955, -0.030891763046383858, 0.05642978101968765, -0.1428368091583252, 0.2256724238395691, 0.0363924540579319, 0.06255213916301727, 0.39391690492630005, 0.14157573878765106, 0.2784302830696106, 0.10396715998649597, -0.22894790768623352, -0.1139814704656601, 0.06613500416278839, -0.13955551385879517, 0.5098288059234619, 0.3375474810600281, 0.08086442947387695, -0.02614494226872921, -0.11726607382297516, 0.07583936303853989, 0.4826371967792511, 0.1798248142004013, -0.020105307921767235, 0.45073404908180237, -0.02641773782670498, -0.19099833071231842, 0.000986358500085771, 0.8557760715484619, -0.5572368502616882, -0.28589072823524475, 0.02064567431807518, 0.09727935492992401, -0.16944199800491333, 0.21763058006763458, -0.20579729974269867, -0.3246438205242157, 0.017004728317260742, -0.19486911594867706, -0.33706390857696533, -0.09025231003761292, 0.31063970923423767, -0.21952824294567108, 0.1833125799894333, 0.0680326446890831, 0.21227164566516876, 0.05599785968661308, -0.09097370505332947, 0.23839090764522552, -0.022679463028907776, -0.006122705060988665, 0.25748273730278015, -0.08425585180521011, -0.24596458673477173, -0.3158252239227295, -0.2029242366552353, 0.1599837988615036, 0.05829690396785736, -0.48271408677101135, -0.1835443675518036, 0.0428902693092823, 0.21460536122322083, 0.11195956915616989, 0.3177332282066345, 0.34393373131752014, 0.2845931351184845, 0.19506828486919403, 0.17445608973503113, 0.24242228269577026, 0.15372471511363983, -0.07111621648073196, -0.3969818949699402, 0.10665556788444519, 0.08801104128360748, 0.24009375274181366, -0.28065332770347595, -0.1707744151353836, -0.2619643211364746, -0.3381376564502716, 0.009782202541828156, -0.07732301205396652, 0.2922811210155487, -0.09014904499053955, -0.23386307060718536, 0.04907851666212082, 0.4467690587043762, -0.1982865333557129, -0.15629220008850098, -0.02728947624564171, 0.10530085116624832, -0.21463532745838165, 0.04991351068019867, 0.32521167397499084, 0.014613106846809387, 0.40427473187446594, -0.2409370243549347, -0.15935160219669342, -0.04674369841814041, 0.10865794867277145, 0.05923201143741608, -0.2635219693183899, 0.30100691318511963, -0.07218001782894135, 0.08340995758771896, 0.04866299778223038, -0.27629363536834717, -0.21490858495235443, 0.23098906874656677, 0.06791452318429947, -0.49536123871803284, -0.023342734202742577, 0.11333981901407242, -0.5093955397605896, -0.05985885113477707, -0.06913233548402786, -0.16433438658714294, 0.10568123310804367, -0.392935574054718, -0.03081454709172249, -0.4348367154598236, -0.37226438522338867, -0.11169157922267914, -0.03351213037967682, 0.2969772219657898, -0.3369506895542145, -0.2700839936733246, -0.4456768035888672, -0.382372111082077, 0.21316732466220856, -0.2579704225063324, -0.07269939035177231, 0.20810386538505554, -0.41918888688087463, -0.029214022681117058, 0.5602771639823914, -0.5683712959289551, -0.36130163073539734, 0.44704410433769226, -0.41898664832115173, -0.2590131461620331, 0.18650418519973755, -0.0008597858250141144, 0.06650570034980774, -0.3295418322086334, -0.1430530548095703, 0.06461155414581299, 0.04237373545765877, -0.13622692227363586, -0.1102428212761879, -0.2676573693752289, 0.19914422929286957, 0.2632274925708771, -0.09678249061107635, 0.13125523924827576, 0.14763672649860382, 0.10275793075561523, 0.3266957700252533, 0.14259035885334015, 0.17181022465229034, -0.08835872262716293, 0.23017242550849915, -0.22566960752010345, -0.14068520069122314, -0.10932894796133041, -0.609980583190918, 0.23831862211227417, -0.06623590737581253, 0.20779447257518768, -0.1417090743780136, 0.02335844747722149, -0.6258836388587952, 0.015870006754994392, -0.05917185917496681, -0.22863855957984924, 0.006851940881460905, 0.19095498323440552, 0.11202678829431534, 0.01821458525955677, 0.037107907235622406, 0.06265643239021301, -0.2197255939245224, 0.17596407234668732, 0.010264958254992962, -0.0012475220719352365, -0.10338165611028671, 0.03828238695859909, 0.23613698780536652, 0.22019854187965393, 0.08650313317775726, -0.31085503101348877, -0.04248672351241112, 0.28622907400131226, 0.274850994348526, 0.20396468043327332, 0.19296391308307648, -0.07467364519834518, 0.13896988332271576, 0.15383516252040863, 0.1740686148405075, 0.21679824590682983, 0.08699439465999603, -0.2102240025997162, 0.08055448532104492, 0.1106918603181839, 0.19249220192432404, -0.02169334888458252, -0.17491643130779266, 0.026053965091705322, 0.40494304895401, 0.12913766503334045, -0.026237299665808678, -0.382283478975296, 0.10046029090881348, 0.2953745126724243, 0.6366948485374451, -0.04710427671670914, 0.36636224389076233, -0.10841920226812363, -0.19691762328147888, -0.3486177325248718, 0.0082011753693223, 0.16020703315734863, -0.15804874897003174, -0.09225582331418991, -0.00574048375710845, 0.2688913941383362, 0.8805612921714783, -0.02649058774113655, 0.171645388007164, 0.14539216458797455, -0.11143907159566879, -0.20710623264312744, 0.18835227191448212, 0.07700632512569427, 0.048256367444992065, 0.14204104244709015, -0.01417927909642458, -0.38311076164245605, -0.08877197653055191, -0.05308203771710396, 0.4589439332485199, 0.48439303040504456, -0.3677517771720886, -0.1962980329990387, -0.24734267592430115, -0.3671190142631531, -0.4130915403366089, -0.32595229148864746, -0.15261603891849518, -0.3070302903652191, -0.12076380848884583, 0.27662521600723267, 0.13867105543613434, 0.22709313035011292, -0.251585990190506, 0.18580666184425354, -0.23863708972930908, -0.31146040558815, 0.09594015032052994, -0.1713581085205078, -0.34025323390960693, 0.030635083094239235, 0.2242276519536972, 0.03132116049528122, 0.41515204310417175, -0.14686712622642517, -0.21823176741600037, -0.251299113035202, -0.244547501206398, 0.12566019594669342, 0.08897522836923599, 0.5211694836616516, 0.4599177837371826, -0.027609296143054962, 0.45716771483421326, 0.117194764316082, 0.034604527056217194, -0.26613324880599976, -0.4669431447982788, -0.1576690971851349, -0.019804026931524277, 0.0831540897488594, -0.03829829394817352, -0.4752779006958008, -0.5595126748085022, -0.07798308879137039, 0.017075948417186737, 0.10753898322582245, 0.026405155658721924, 0.044131964445114136, -0.09966818243265152, 0.20519140362739563, 0.025717174634337425, -0.11663949489593506, -0.1441696286201477, 0.0730142816901207, 0.020061355084180832, -0.17576855421066284, -0.17963047325611115, 0.11319729685783386, -0.06287635862827301, 0.08196175843477249, -0.15483389794826508, -0.41781702637672424, -0.3486686944961548, 0.07465797662734985, 0.1838570386171341, 0.280537873506546, -0.27233701944351196, 0.28959473967552185, 0.24983306229114532, 0.1481553167104721, -0.12449976801872253, -0.3925282657146454, 0.15622222423553467, 0.01943698339164257, 0.2505042254924774, 0.14687803387641907, 0.48256418108940125, -0.1921388804912567, 0.7506556510925293, 0.11803855746984482, -0.16311046481132507, 0.02059563808143139, -0.2556125819683075, 0.041471853852272034, -0.0539502277970314, -0.12413882464170456, 0.3178105354309082, -0.13975438475608826, -0.24665462970733643, 0.19574549794197083, -0.2231135070323944, -0.0524621345102787, -0.09552223980426788, 0.22133146226406097, -0.06543813645839691, -0.025436243042349815, -0.0534769706428051, -0.20936022698879242, 0.05761292204260826, 0.037889041006565094, -0.07096292823553085, -0.19440782070159912, -0.21401964128017426, 0.004863300360739231, 0.3135368227958679, 0.004929548129439354, 0.1602725088596344, -0.30051466822624207, 0.018574286252260208, -0.4462011754512787, 0.3039744794368744, 0.2712737023830414, 0.5984059572219849, 0.08295068889856339, 0.07675141096115112, 0.12854960560798645, 0.10277166217565536, 0.21404898166656494, -0.028820736333727837, 0.12243352830410004, 0.2516895532608032, -0.13396118581295013, -0.08915529400110245, -0.24068503081798553, -0.06588214635848999, 0.2125190645456314, -0.029730752110481262, 0.09589412063360214, -0.3500882685184479, -0.089262455701828, 0.26529309153556824, 0.1409558206796646, -0.08590085804462433, -0.35511109232902527, -0.12330783903598785, -0.07265663892030716, -0.15032240748405457, -0.21273303031921387, -0.17580485343933105, 0.10044082999229431, -0.1292785406112671, 0.282055526971817, 0.014954011887311935, -0.09018872678279877, -0.08661637455224991, 0.11567702144384384, 0.14001037180423737, -0.18445229530334473, 0.33762672543525696, 0.17730817198753357, 0.046880707144737244, 0.40233907103538513, 0.31333476305007935, -0.10161526501178741, -0.2746429443359375, 0.026888970285654068, -0.006902288179844618, 0.3866167366504669, 0.12779264152050018, 0.11361660063266754, 0.04185843840241432, -0.2697533369064331, -0.14895153045654297, -0.2984605133533478, 0.19843843579292297, 0.3825056850910187, -0.09436489641666412, -0.051382165402173996, -0.3670404851436615, 0.6012393832206726, 0.20579013228416443, -0.1570291668176651, 0.47727280855178833, -0.2519432008266449, -0.2803330421447754, -0.05517932027578354, 0.07219129055738449, 0.8358826637268066, 0.12051711231470108, 0.16542373597621918, 0.377535879611969, -0.2624789774417877, 0.3626869320869446, 0.09037939459085464, 0.16148507595062256, -0.5544984936714172, -0.08480037748813629, 0.024787679314613342, -0.23059771955013275, 0.24379876255989075, -0.1052330806851387, -0.23620764911174774, 0.37085604667663574, 0.04794016107916832, 0.283141553401947, 0.2545764744281769, 0.35597464442253113, -0.11178461462259293, -0.044524919241666794, -0.10924962162971497, 0.10165836662054062, -0.1658712774515152, 0.40507498383522034, -0.2347889542579651, 0.08953949809074402, -0.08646851778030396, -0.012371635995805264, -0.2658649682998657, -0.11775975674390793, -0.14358916878700256, 0.0613333098590374, 0.07406347990036011, -0.14600135385990143, 0.10794626921415329, 0.19950465857982635, 0.1334606111049652, -0.16214707493782043, -0.16512370109558105, 0.1512046605348587, -0.15446172654628754, 0.10339865833520889, 0.3309915363788605, 0.05736226588487625, 0.23977375030517578, -0.3374669551849365, -0.17271959781646729, 0.13349278271198273, -0.238130584359169, -0.13328112661838531, -0.09552451223134995, -0.18657712638378143, 0.10564952343702316, -0.35079923272132874, -0.32003968954086304, -0.14020808041095734, 0.07292859256267548, -0.14415039122104645, 0.040004901587963104, 0.2728864252567291, -0.37908753752708435, -0.00606553815305233, -0.12707988917827606, -0.3006232678890228, 0.016579188406467438, 0.4735439121723175, -0.061930812895298004, -0.16933615505695343, 0.7641564607620239, -0.1777879148721695, 0.0686742439866066, -0.1367507129907608, 0.0628247857093811, 0.011606607586145401, -0.13863545656204224, 0.02574157528579235, -0.048890821635723114, -0.18072548508644104, 0.10184184461832047, 0.5352265238761902, 0.3329120874404907, 0.004821674432605505, 0.0025996509939432144, -0.46171140670776367, -0.442257821559906, -0.09019425511360168, -0.15435589849948883, 0.1846054196357727, -0.07652480900287628, 0.13444079458713531, 0.21724897623062134, 0.10431482642889023, -0.20098237693309784, -0.1766953021287918, 0.007592774461954832, 0.2428709864616394, -0.12657582759857178, 0.04392238333821297, -0.02672380581498146, 0.08731298893690109, 0.031429801136255264, -0.027037255465984344, -0.33485147356987, -0.07218395173549652, -0.06499511748552322, 0.20209327340126038, 0.09678614139556885, -0.028630727902054787, 0.07220373302698135, -0.1510540246963501, -0.02545309066772461, 0.17819719016551971, -0.02685387246310711, 0.0856492817401886, -0.09723051637411118, 0.15172509849071503, 0.32388028502464294, -0.07167699933052063, -0.050032760947942734, 0.21732299029827118, 0.05624875798821449, 0.16201576590538025, -0.0630064532160759, 0.2864287197589874, -0.13695983588695526, 0.05460163950920105, 0.20251114666461945, 0.10919026285409927, 0.008270323276519775, 0.25383976101875305, 0.11799754202365875, -0.16820859909057617, 0.14139597117900848, -0.04990166053175926, 0.355796217918396, 0.7263289093971252, -0.21952444314956665, -0.12481629103422165, 0.4329032301902771, 0.0563654862344265, -0.2547045648097992, -0.04399596154689789, 0.3098618686199188, 0.1377621442079544, 0.04623548313975334, 0.2116360068321228, 0.0015202092472463846, -0.03864780440926552, 0.16678832471370697, 0.13611532747745514, 0.12866833806037903, -0.2262921929359436, -0.05903869867324829, 0.44832172989845276, -0.22031773626804352, 0.16248928010463715, 0.3900027573108673, 0.06479004770517349, 0.213303342461586, 0.2186940610408783, 0.037032514810562134, 0.6142653822898865, 0.05644666776061058, 0.017884409055113792, 0.0840257853269577, -0.04976397380232811, -0.000603065884206444, 0.27663639187812805, -0.030039489269256592, -0.0860842689871788, 0.32754382491111755, 0.5549077987670898, -0.3502870500087738, -0.159073606133461, -0.26408183574676514, -0.08660200983285904, 0.03199806064367294, -0.14423878490924835, -0.28509464859962463, 0.08155794441699982, -0.21613909304141998, -0.13049784302711487, -0.11887709051370621, -0.3093903660774231, 0.16270646452903748, 0.017042696475982666, -0.39833709597587585, -0.2444036900997162, -0.2909780740737915, 0.1936604082584381, -0.10940326750278473, -0.2699061632156372, 0.22087472677230835, -0.013040215708315372, 0.03285606950521469, 0.2911534607410431, 0.32487058639526367, 0.4699462354183197, 0.16254085302352905, -0.31814828515052795, -0.23251184821128845, -0.12140605598688126, 0.09721093624830246, 0.035079892724752426, 0.38069814443588257, -0.04988781735301018, -0.10703174769878387, 0.46314844489097595, 0.0726584866642952, 0.03713405877351761, 0.21864303946495056, 0.43211278319358826, -0.48619702458381653, -0.17150579392910004, 0.7934106588363647, -0.07552909851074219, -0.1411667764186859, -0.12957747280597687, 0.2552591562271118, -0.2604227364063263, 0.13434675335884094, 0.0995590090751648, -0.07344663143157959, 0.13018882274627686, 0.025473011657595634, 0.03947748616337776, -0.10640545934438705, 0.3546757102012634, 0.31101417541503906, -0.043816789984703064, -0.36947277188301086, -0.2513991594314575, -0.7986483573913574, 0.227462500333786, -0.11356871575117111, -0.32369786500930786, -0.06485449522733688, -0.2603071928024292, 0.18936939537525177, 0.048111118376255035, -0.14149145781993866, 0.19518284499645233, 0.0548703670501709, 0.1351766586303711, -0.18994483351707458, -0.3179088830947876, -0.28872278332710266, -0.023948414251208305, -0.10388311743736267, -0.38692671060562134, 0.11374453455209732, -0.3304969072341919, -0.07958601415157318, 0.11841181665658951, -0.009781562723219395, 0.4779796600341797, -0.005791524890810251, 0.36548948287963867, -0.11248205602169037, 0.7752768993377686, 0.18795086443424225, 0.0632830262184143, -0.1649438738822937, 0.12217654287815094, -0.12064743041992188, 0.13483719527721405, -0.07033577561378479, 0.1044105514883995, -0.11896292865276337, 0.5212558507919312, -0.44372132420539856, 0.4605461657047272, -0.028103237971663475, 0.07408104091882706, -0.3021466135978699, 0.04564795643091202, -0.2554609477519989, 0.12183521687984467, 0.3444594442844391, 0.14114558696746826, -0.09161016345024109, 0.3550700843334198, -0.2149428427219391, -0.1903918981552124, 0.3962547481060028, -0.10348962247371674, -0.09398669749498367, 0.021515732631087303, 0.2646273970603943, -0.051723506301641464, -0.14009112119674683, -0.43938711285591125, 0.19632801413536072, 0.2907603085041046, -0.0937122330069542, 0.11838947981595993, 0.2602342963218689, 0.28375184535980225, 0.13402220606803894, 0.01832745596766472, 0.9827133417129517, -0.05974356457591057, 0.17604413628578186, -0.10538424551486969, -0.22785966098308563 ]
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
[ -0.19621717929840088, 0.4000371992588043, -0.0314687043428421, -0.1644129604101181, 0.04428338259458542, -0.13329292833805084, 0.2794848084449768, 0.37617406249046326, 0.009489261545240879, 0.08292654156684875, 0.20643186569213867, 0.15806883573532104, -0.3005681335926056, 0.0680416002869606, 0.1369239091873169, 0.001200942788273096, -0.05646305903792381, 0.1698962152004242, -0.09795276820659637, 0.21250583231449127, -0.1272204965353012, 0.07482492923736572, 0.020457642152905464, 0.10806220024824142, -0.027999628335237503, 0.08875580877065659, -0.21589472889900208, 0.03439030423760414, -0.3809467852115631, -0.7858074307441711, 0.3818265199661255, 0.07057245820760727, 0.18520081043243408, 0.3587058484554291, -0.0001226895401487127, -0.08625618368387222, 0.5880789160728455, -0.05607704073190689, -0.7038208246231079, -0.2316201627254486, 0.2454066127538681, -0.4804557263851166, 0.30185335874557495, -0.021508481353521347, 0.18851667642593384, -0.13163875043392181, 0.2442077249288559, 0.10435047745704651, 0.1617514044046402, 0.32244840264320374, 0.11873669177293777, 0.2776325047016144, 0.05146842449903488, -0.17972369492053986, -0.10094425827264786, 0.08600422739982605, -0.08926767855882645, 0.502880871295929, 0.36486220359802246, 0.03112071380019188, -0.11236335337162018, -0.13111881911754608, 0.09226356446743011, 0.4931356608867645, 0.12502175569534302, -0.022293858230113983, 0.501173734664917, 0.019112180918455124, -0.21289405226707458, 0.0410485565662384, 0.7716662883758545, -0.5064083933830261, -0.37548545002937317, -0.07860947400331497, 0.05649561062455177, -0.13751477003097534, 0.2607558071613312, -0.1600870043039322, -0.3946448266506195, 0.019820082932710648, -0.16637548804283142, -0.28037238121032715, -0.0671222060918808, 0.2935130298137665, -0.2830924987792969, 0.2682918310165405, 0.07774831354618073, 0.17557889223098755, 0.08950256556272507, -0.10085028409957886, 0.2573443055152893, -0.061836861073970795, 0.024404756724834442, 0.2348816692829132, -0.030215810984373093, -0.223875030875206, -0.2610195279121399, -0.2221411019563675, 0.19221366941928864, 0.08718405663967133, -0.4626268446445465, -0.19067595899105072, -0.03120706044137478, 0.1495988965034485, 0.1416134536266327, 0.4171294867992401, 0.3384239971637726, 0.23890098929405212, 0.24328820407390594, 0.16279198229312897, 0.24343900382518768, 0.26342490315437317, -0.059813082218170166, -0.37045860290527344, -0.032496582716703415, 0.09223354607820511, 0.1482526808977127, -0.32870447635650635, -0.15333983302116394, -0.22707119584083557, -0.38531744480133057, -0.04631436988711357, -0.054925587028265, 0.23764845728874207, -0.10888964682817459, -0.14659403264522552, 0.04901217296719551, 0.41204357147216797, -0.20630070567131042, -0.1511349380016327, -0.02948884479701519, 0.04452644661068916, -0.19153964519500732, -0.012613998726010323, 0.3086452782154083, 0.05575784668326378, 0.3780360221862793, -0.36712178587913513, -0.11766080558300018, -0.03431221842765808, 0.11314677447080612, 0.18523536622524261, -0.26896557211875916, 0.289959192276001, -0.15554536879062653, 0.05203242599964142, 0.07850492000579834, -0.2927016019821167, -0.22106754779815674, 0.16356559097766876, 0.12039411813020706, -0.4569106996059418, 0.03789442405104637, 0.09306164085865021, -0.6060147881507874, -0.0646645724773407, 0.01272648200392723, -0.11572955548763275, 0.0913872942328453, -0.33262190222740173, 0.026575805619359016, -0.39136916399002075, -0.20526589453220367, -0.0672704204916954, -0.09972582012414932, 0.23910433053970337, -0.2667432427406311, -0.33723148703575134, -0.4481423795223236, -0.32592353224754333, 0.12920185923576355, -0.2248053103685379, -0.0771235004067421, 0.21498148143291473, -0.37946197390556335, -0.00851595401763916, 0.5497832298278809, -0.5216637253761292, -0.3586077094078064, 0.4454956352710724, -0.4292171895503998, -0.2467309832572937, 0.23419709503650665, -0.04898141324520111, -0.002735686022788286, -0.3621470332145691, -0.08949652314186096, 0.08227434754371643, 0.0496206097304821, -0.1781887263059616, -0.18848401308059692, -0.3439171314239502, 0.1136094406247139, 0.2200978547334671, -0.039881639182567596, 0.08876359462738037, 0.09676407277584076, 0.05409688875079155, 0.3513943552970886, 0.11439776420593262, 0.15504485368728638, -0.14401818811893463, 0.17716756463050842, -0.2378762662410736, -0.17119497060775757, -0.05177273228764534, -0.5992653369903564, 0.2523058354854584, -0.11843118071556091, 0.2130230814218521, -0.04645254835486412, -0.01118173822760582, -0.5071437358856201, 0.003305845195427537, 0.01983991265296936, -0.2721047103404999, 0.0020706679206341505, 0.20442242920398712, 0.12961800396442413, -0.04795818030834198, -0.01060548983514309, 0.032223671674728394, -0.21072918176651, 0.19253285229206085, -0.009809551760554314, -0.02028551511466503, -0.046190254390239716, 0.062408044934272766, 0.24959442019462585, 0.2555881440639496, 0.0547189898788929, -0.34098246693611145, 0.007952791638672352, 0.28200289607048035, 0.20832459628582, 0.18946506083011627, 0.18919211626052856, -0.07011068612337112, 0.17034904658794403, 0.26392465829849243, 0.1645798236131668, 0.2629547417163849, -0.015182062052190304, -0.21230486035346985, -0.08616418391466141, 0.24468593299388885, 0.16804054379463196, 0.04419169947504997, -0.12836302816867828, 0.06571392714977264, 0.3238115906715393, 0.11381257325410843, -0.13485535979270935, -0.22805830836296082, 0.19604426622390747, 0.2852373719215393, 0.6019861102104187, -0.05045652016997337, 0.3259406089782715, -0.047923069447278976, -0.27811819314956665, -0.37065553665161133, 0.017936676740646362, 0.1138995885848999, -0.09621269255876541, -0.0562751330435276, 0.09414384514093399, 0.3609091639518738, 0.8490375876426697, -0.03925875574350357, 0.27091628313064575, 0.08667974919080734, -0.1844286471605301, -0.23165786266326904, 0.15094101428985596, 0.11443629860877991, -0.02646540105342865, 0.1556093841791153, 0.0009207495022565126, -0.401189923286438, -0.08902247250080109, -0.00013077813491690904, 0.4331217110157013, 0.39993196725845337, -0.34717369079589844, -0.109804667532444, -0.302101194858551, -0.4445493817329407, -0.39858973026275635, -0.36305129528045654, -0.20808124542236328, -0.3813641369342804, -0.061455726623535156, 0.22124448418617249, -0.00008055925718508661, 0.23552894592285156, -0.19494959712028503, 0.2832927703857422, -0.3094019591808319, -0.33439627289772034, 0.11647383868694305, -0.2011946588754654, -0.4004579484462738, 0.03807689622044563, 0.28132152557373047, 0.07048767805099487, 0.442693293094635, -0.10782022029161453, -0.297360897064209, -0.21326033771038055, -0.3188502788543701, 0.1328551322221756, 0.06095751374959946, 0.5699265599250793, 0.476179301738739, -0.014883671887218952, 0.42041015625, 0.07882949709892273, 0.04204424470663071, -0.23354877531528473, -0.43352946639060974, -0.23417872190475464, 0.09182917326688766, 0.09691415727138519, -0.13966430723667145, -0.3972931504249573, -0.52669358253479, -0.08559008687734604, 0.17034344375133514, 0.12703529000282288, 0.07320814579725266, 0.0939827412366867, -0.11142731457948685, 0.2511829435825348, 0.08683266490697861, -0.18744918704032898, -0.11685238778591156, 0.07108856737613678, -0.009083881974220276, -0.1429181694984436, -0.1871636062860489, 0.0666741356253624, -0.1160891130566597, 0.0868183821439743, -0.15720684826374054, -0.48865655064582825, -0.44232669472694397, 0.15424863994121552, 0.23099461197853088, 0.39259207248687744, -0.20678263902664185, 0.2817327082157135, 0.20180164277553558, 0.12314781546592712, -0.10951650142669678, -0.3401251435279846, 0.16022445261478424, 0.0937713235616684, 0.31101879477500916, 0.13703419268131256, 0.342370867729187, -0.11723069846630096, 0.7626632452011108, 0.11879247426986694, -0.09411514550447464, -0.03337504714727402, -0.13440799713134766, 0.015389545820653439, -0.01784064993262291, -0.04059600085020065, 0.3350592851638794, -0.20015400648117065, -0.3104262948036194, 0.20721714198589325, -0.11684462428092957, -0.03705434873700142, -0.13537591695785522, 0.289180189371109, -0.017413048073649406, -0.06730537861585617, 0.003846805775538087, -0.2191673219203949, 0.00807683914899826, 0.09175163507461548, -0.028248872607946396, -0.19920827448368073, -0.15857328474521637, 0.03104431740939617, 0.3056078851222992, 0.009298847988247871, 0.17336410284042358, -0.25022366642951965, 0.006672019604593515, -0.3804590106010437, 0.3209517300128937, 0.3083265721797943, 0.4791109561920166, 0.08994165807962418, 0.0405026376247406, 0.128177210688591, 0.02376876398921013, 0.22550968825817108, 0.07952553033828735, 0.16413156688213348, 0.2530301809310913, -0.13103342056274414, -0.04519311338663101, -0.24799948930740356, -0.06108873337507248, 0.06379713118076324, 0.0043015084229409695, 0.09554512053728104, -0.40971311926841736, -0.09873327612876892, 0.1882556825876236, 0.08911500126123428, -0.07824201136827469, -0.2351612001657486, -0.07499716430902481, -0.08631077408790588, -0.11132063716650009, -0.21148724853992462, -0.2218320667743683, 0.01701837033033371, -0.21502208709716797, 0.4063698351383209, -0.0038532966282218695, -0.08811710774898529, -0.060154661536216736, 0.09694603085517883, 0.08987788110971451, -0.2820564806461334, 0.2924055755138397, 0.26512569189071655, 0.07484521716833115, 0.30377933382987976, 0.3153017461299896, -0.09081123769283295, -0.21888956427574158, 0.025368615984916687, 0.05520337447524071, 0.4211718440055847, 0.35064882040023804, 0.10434013605117798, -0.05036785081028938, -0.23994180560112, -0.20737224817276, -0.23432552814483643, 0.19697536528110504, 0.28577351570129395, -0.03091188333928585, 0.07850944995880127, -0.3184972405433655, 0.7194925546646118, 0.22033104300498962, -0.25014519691467285, 0.4313560724258423, -0.21806475520133972, -0.2628091871738434, -0.06578298658132553, 0.12758295238018036, 0.903336763381958, 0.14794974029064178, 0.23772449791431427, 0.4199577867984772, -0.26912352442741394, 0.4643268287181854, 0.07282882183790207, 0.15510959923267365, -0.4947304129600525, -0.00016532746667508036, 0.019681045785546303, -0.18237507343292236, 0.20952031016349792, -0.09575735032558441, -0.2247014343738556, 0.3749464750289917, 0.14652331173419952, 0.2810671329498291, 0.3701073229312897, 0.30880028009414673, -0.2517704665660858, -0.03366036340594292, -0.1858089566230774, 0.09027866274118423, -0.1498018503189087, 0.40192824602127075, -0.2816436290740967, -0.00031441307510249317, -0.14423689246177673, -0.044248200953006744, -0.25294551253318787, -0.11223705857992172, -0.17343932390213013, 0.06214240565896034, 0.15585291385650635, -0.13206006586551666, -0.0661604106426239, 0.18333199620246887, 0.08269117027521133, -0.21432539820671082, -0.14978258311748505, 0.125441774725914, -0.14331720769405365, 0.16847188770771027, 0.38182613253593445, 0.04532894119620323, 0.32668188214302063, -0.3169657588005066, -0.05816745385527611, 0.06710822135210037, -0.2803245186805725, -0.12472721189260483, -0.13294479250907898, -0.1724696010351181, 0.1416565328836441, -0.29994383454322815, -0.14256222546100616, -0.10251215100288391, -0.009897514246404171, -0.11590301990509033, 0.022395465523004532, 0.281062513589859, -0.35529693961143494, -0.06101086735725403, -0.11896082013845444, -0.2299092411994934, 0.009612989611923695, 0.47477036714553833, -0.06723783165216446, -0.13764384388923645, 0.6765257716178894, -0.21037650108337402, 0.0658826008439064, -0.1299179494380951, 0.06201278045773506, -0.10107384622097015, -0.11184277385473251, 0.0767831802368164, -0.16967418789863586, -0.3008667826652527, 0.043714553117752075, 0.585026741027832, 0.44189453125, 0.012407045811414719, -0.0704069659113884, -0.46289244294166565, -0.4116658866405487, -0.0003568688698578626, -0.14149734377861023, 0.20699259638786316, -0.21216470003128052, 0.22073911130428314, 0.16328179836273193, 0.04603215679526329, -0.18724986910820007, -0.201705664396286, 0.04946436733007431, 0.20594926178455353, -0.13847608864307404, 0.04342867061495781, -0.015562719665467739, 0.14493201673030853, -0.03392535820603371, 0.06349065899848938, -0.3615333139896393, -0.07694461941719055, -0.10113620012998581, 0.2054293006658554, 0.1118430346250534, -0.03894906863570213, 0.15560276806354523, -0.11734660714864731, 0.039721958339214325, 0.21439768373966217, -0.06621690094470978, -0.04835868254303932, -0.09268875420093536, 0.11267220228910446, 0.3458282947540283, 0.09795032441616058, -0.00695122079923749, 0.30149000883102417, 0.10003796219825745, 0.17635118961334229, -0.1161336600780487, 0.2669965922832489, -0.23393209278583527, 0.08855469524860382, 0.20248867571353912, 0.12920895218849182, -0.01943167857825756, 0.26001304388046265, 0.13960129022598267, -0.11417288333177567, 0.08041925728321075, -0.16252903640270233, 0.3586396872997284, 0.8415077924728394, -0.17826025187969208, -0.13109157979488373, 0.4703674018383026, 0.03272818773984909, -0.19339071214199066, 0.027116786688566208, 0.22767911851406097, 0.20254084467887878, 0.039652734994888306, 0.26613739132881165, 0.010690607130527496, -0.13747429847717285, 0.196112260222435, 0.1407436579465866, 0.09474264830350876, -0.1786995530128479, -0.07266828417778015, 0.5383135080337524, -0.25530382990837097, 0.16207417845726013, 0.3502660393714905, 0.0004748752689920366, 0.21568593382835388, 0.08506105095148087, 0.025451570749282837, 0.5338094234466553, -0.012156331911683083, 0.058235444128513336, 0.11353041976690292, 0.01912621036171913, -0.04424289986491203, 0.25939202308654785, 0.05144336074590683, -0.011776112020015717, 0.3216550648212433, 0.6301922798156738, -0.43068012595176697, -0.1747276782989502, -0.27786174416542053, -0.049408238381147385, 0.0822819322347641, -0.14998450875282288, -0.39644932746887207, 0.043598249554634094, -0.22130972146987915, -0.07336201518774033, -0.14795057475566864, -0.31852343678474426, 0.229910746216774, 0.003296385984867811, -0.355591356754303, -0.3461935818195343, -0.3967621922492981, 0.1408468633890152, -0.035808224231004715, -0.27528682351112366, 0.2508656084537506, 0.029407447203993797, 0.01853197254240513, 0.2829645574092865, 0.28391730785369873, 0.444156676530838, 0.08628561347723007, -0.438735693693161, -0.2540844678878784, -0.19133083522319794, 0.13583217561244965, -0.07762159407138824, 0.3995724022388458, -0.059817031025886536, -0.16802285611629486, 0.49920526146888733, 0.058867864310741425, 0.043532274663448334, 0.2664313316345215, 0.37248632311820984, -0.5206149220466614, -0.1689813733100891, 0.8208153247833252, -0.06794863939285278, -0.1667519509792328, -0.027103718370199203, 0.17767955362796783, -0.23198485374450684, 0.08259589970111847, -0.035704534500837326, -0.047500114887952805, 0.18776071071624756, 0.11940514296293259, 0.04107683524489403, -0.1780634969472885, 0.39927202463150024, 0.3388811945915222, -0.0715000107884407, -0.34252864122390747, -0.21393318474292755, -0.8131757974624634, 0.08850467205047607, -0.10057631134986877, -0.3737140893936157, -0.0635504424571991, -0.2374129593372345, 0.2535680830478668, 0.10478401929140091, -0.2574351727962494, 0.17715521156787872, -0.0015825501177459955, 0.12594352662563324, -0.26279616355895996, -0.3659880459308624, -0.21408528089523315, -0.025764524936676025, -0.0297405906021595, -0.40254831314086914, 0.07842455059289932, -0.31337136030197144, -0.09926480799913406, 0.22371716797351837, -0.03939524665474892, 0.4940952658653259, 0.07308303564786911, 0.25377750396728516, -0.012197490781545639, 0.6868563294410706, 0.24270302057266235, -0.031302742660045624, -0.24734275043010712, 0.1787668913602829, -0.13330835103988647, 0.1253570318222046, -0.10701523721218109, 0.13508670032024384, -0.17025801539421082, 0.614316999912262, -0.414183109998703, 0.4506014883518219, 0.07495110481977463, 0.07818938046693802, -0.28697699308395386, 0.04797082021832466, -0.34138694405555725, 0.14387965202331543, 0.3280654549598694, 0.3049732744693756, -0.11956630647182465, 0.34439799189567566, -0.26570072770118713, -0.15193794667720795, 0.4117482006549835, -0.24894623458385468, -0.0583859421312809, 0.05723512917757034, 0.2190214991569519, 0.0009546805522404611, -0.15867112576961517, -0.4435001611709595, 0.1210443526506424, 0.2671409547328949, -0.1325509399175644, 0.08481620252132416, 0.14585541188716888, 0.2846495807170868, 0.026441916823387146, -0.0357406921684742, 0.9360101819038391, -0.07236693799495697, 0.1544700413942337, -0.018570519983768463, -0.24318115413188934 ]
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?
[ -0.10087020695209503, 0.5024334192276001, -0.03979721665382385, -0.08978375792503357, 0.09680269658565521, -0.12240693718194962, 0.28728705644607544, 0.3135531544685364, -0.054385244846343994, 0.11574174463748932, 0.046181775629520416, 0.16493380069732666, -0.27445492148399353, -0.00596386706456542, 0.16595663130283356, -0.0370168462395668, -0.07802143692970276, 0.25511327385902405, -0.0911446288228035, 0.21499690413475037, -0.23069186508655548, 0.12446112185716629, -0.07093735784292221, 0.0745798721909523, 0.1365300416946411, 0.1996983140707016, -0.16580645740032196, 0.05431871861219406, -0.316034734249115, -0.8764683604240417, 0.347922146320343, 0.18099501729011536, 0.24238252639770508, 0.23878885805606842, -0.00012480560690164566, -0.15380287170410156, 0.5954885482788086, -0.12650667130947113, -0.6724078059196472, -0.17283138632774353, 0.24162212014198303, -0.4436884820461273, 0.2643412947654724, -0.07162002474069595, 0.27901872992515564, -0.16837641596794128, 0.18214714527130127, 0.11028321087360382, 0.074053093791008, 0.38633596897125244, 0.11946219205856323, 0.24075792729854584, 0.0858573168516159, -0.12019337713718414, -0.058449503034353256, 0.024083545431494713, -0.00082276517059654, 0.6254268884658813, 0.40664729475975037, 0.021436531096696854, 0.021082628518342972, -0.010083942674100399, -0.061721377074718475, 0.4501315653324127, 0.24951216578483582, -0.04781978949904442, 0.4997137784957886, -0.07006870955228806, -0.16319669783115387, -0.013614890165627003, 0.6971591711044312, -0.5866326093673706, -0.3620380461215973, -0.0547504760324955, 0.15433409810066223, -0.2042607069015503, 0.13494256138801575, -0.2809629440307617, -0.439496785402298, 0.058052800595760345, -0.10192985087633133, -0.5488119721412659, 0.017848556861281395, 0.25423648953437805, -0.30119749903678894, 0.2867899239063263, 0.11398310214281082, 0.14544890820980072, 0.24242764711380005, -0.0527881383895874, 0.21367676556110382, -0.022138435393571854, -0.08245576173067093, 0.1698237508535385, 0.0019517162581905723, -0.30449897050857544, -0.31498029828071594, -0.2155274599790573, 0.09691508859395981, 0.13492164015769958, -0.3737146258354187, -0.14380383491516113, -0.05633988603949547, 0.1341736912727356, 0.0918913334608078, 0.4165337383747101, 0.3381829559803009, 0.2206364870071411, 0.10820475220680237, 0.22823025286197662, 0.2354670614004135, 0.12068713456392288, 0.03867361322045326, -0.36110830307006836, 0.01983506605029106, 0.1914501041173935, 0.13478787243366241, -0.19446922838687897, -0.08906855434179306, -0.2402390092611313, -0.4562418460845947, -0.0577772855758667, -0.17793451249599457, 0.2095634490251541, -0.08753703534603119, -0.08982712030410767, 0.10732913017272949, 0.4088948369026184, -0.12385629117488861, -0.27147436141967773, 0.0032227053306996822, 0.19675244390964508, -0.09802448749542236, 0.11117058992385864, 0.2187105268239975, 0.02494209259748459, 0.33839651942253113, -0.37400153279304504, -0.0343971885740757, -0.03799381107091904, 0.10087798535823822, -0.013926914893090725, -0.1752641499042511, 0.2979563772678375, -0.13329321146011353, 0.03545450419187546, 0.020310943946242332, -0.3518683910369873, -0.208586186170578, 0.24709345400333405, 0.22873254120349884, -0.45718780159950256, -0.09473888576030731, 0.05991038680076599, -0.5005057454109192, -0.02239154651761055, -0.05310032516717911, -0.07679621875286102, -0.07516732066869736, -0.5091546177864075, 0.156699076294899, -0.3630223572254181, -0.2893633246421814, -0.05024617165327072, -0.05941978096961975, 0.37651270627975464, -0.2985677719116211, -0.29981210827827454, -0.5405725240707397, -0.30797699093818665, 0.1497212052345276, -0.24499884247779846, -0.16400115191936493, 0.26623764634132385, -0.34657198190689087, 0.043165139853954315, 0.6672208309173584, -0.6826149225234985, -0.381733238697052, 0.3879486322402954, -0.2968747019767761, -0.2016323208808899, 0.13723108172416687, -0.08174243569374084, 0.07141726464033127, -0.3140595853328705, -0.20285403728485107, -0.005927060265094042, 0.09159844368696213, -0.21299773454666138, -0.13809294998645782, -0.47395309805870056, 0.13925640285015106, 0.22118891775608063, 0.00078498141374439, 0.24563638865947723, 0.03358962759375572, 0.1470761001110077, 0.3634043335914612, 0.10033781826496124, 0.13167646527290344, -0.15881597995758057, 0.24994681775569916, -0.18818029761314392, -0.18373267352581024, -0.08811618387699127, -0.5891329646110535, 0.2165801227092743, -0.16063688695430756, 0.22971302270889282, -0.07063723355531693, -0.0018129912205040455, -0.4160647392272949, -0.03989436849951744, 0.09383713454008102, -0.2353094220161438, -0.025886772200465202, 0.19697043299674988, 0.1955232173204422, 0.0044663529843091965, -0.015730272978544235, 0.06437632441520691, -0.16206392645835876, 0.23343215882778168, 0.07400725781917572, -0.051650065928697586, -0.022029241546988487, 0.14289583265781403, 0.2040296494960785, 0.24361878633499146, 0.030902333557605743, -0.2502630650997162, -0.0366775318980217, 0.31348931789398193, 0.14573073387145996, 0.27442288398742676, 0.10608712583780289, -0.04028899595141411, 0.18484197556972504, 0.28031063079833984, 0.1604781299829483, 0.2589096426963806, 0.11254432052373886, -0.12400847673416138, 0.08150605112314224, 0.11401201784610748, 0.30637645721435547, 0.03690572828054428, -0.10141030699014664, 0.04883810132741928, 0.25260061025619507, 0.06619589030742645, -0.12181849777698517, -0.3277985155582428, -0.024630144238471985, 0.2636629045009613, 0.6258193850517273, -0.09830475598573685, 0.3067034184932709, -0.10810405761003494, -0.2765680253505707, -0.37293481826782227, 0.02712651528418064, 0.1354561746120453, -0.24313892424106598, -0.05934790149331093, 0.005910019390285015, 0.41281387209892273, 0.8891434669494629, -0.08765062689781189, 0.21271070837974548, 0.2576932907104492, 0.09946395456790924, -0.14763516187667847, 0.3102176785469055, 0.2731751501560211, 0.03670462593436241, 0.15607129037380219, 0.017295710742473602, -0.2442476600408554, -0.03242715448141098, -0.09209553152322769, 0.33091604709625244, 0.35919851064682007, -0.30033421516418457, -0.10634081810712814, -0.3063659965991974, -0.5244473814964294, -0.3691823184490204, -0.3438993990421295, -0.2026212364435196, -0.33673498034477234, -0.16702476143836975, 0.25610968470573425, -0.0168010164052248, 0.251079261302948, -0.17695975303649902, 0.23437507450580597, -0.3581812381744385, -0.26189225912094116, 0.014088939875364304, -0.06981243938207626, -0.28657805919647217, 0.0055031077936291695, 0.23285916447639465, 0.16187751293182373, 0.2496270388364792, -0.2545627951622009, -0.21234293282032013, -0.3464905321598053, -0.18361692130565643, 0.19877409934997559, 0.24772775173187256, 0.413788765668869, 0.5279135704040527, -0.07021380215883255, 0.35673263669013977, 0.0594496950507164, -0.06608276069164276, -0.1920316368341446, -0.37169933319091797, -0.30342352390289307, -0.0005276899901218712, 0.27848020195961, -0.08362525701522827, -0.31373143196105957, -0.6029025912284851, -0.14122095704078674, 0.03423202782869339, 0.07534781843423843, 0.125221386551857, 0.030590953305363655, -0.24102644622325897, 0.22648747265338898, -0.06759881228208542, -0.23341898620128632, -0.20004764199256897, 0.15993013978004456, -0.009144792333245277, -0.1631017029285431, -0.04546220973134041, 0.07400721311569214, -0.10867724567651749, 0.05650431662797928, -0.23117831349372864, -0.43902674317359924, -0.49937334656715393, 0.2032100260257721, -0.011472409591078758, 0.14936910569667816, -0.18084725737571716, 0.22669410705566406, 0.24462661147117615, 0.1441558301448822, 0.007774782832711935, -0.4050804674625397, 0.12961573898792267, 0.1733417809009552, 0.2596011757850647, 0.20252127945423126, 0.470723032951355, -0.18957823514938354, 0.6809960007667542, 0.08019028604030609, -0.06540752947330475, -0.006914707366377115, -0.19380629062652588, 0.12983430922031403, 0.09109865128993988, -0.0736021175980568, 0.21883228421211243, -0.19008880853652954, -0.2723950147628784, 0.2607298493385315, -0.36413395404815674, -0.0907403826713562, -0.15291056036949158, 0.21480895578861237, -0.18263714015483856, -0.16377519071102142, 0.07783850282430649, -0.2045089602470398, 0.0020859313663095236, 0.14842146635055542, 0.00008275919390143827, -0.2190665453672409, -0.05846767872571945, 0.10795805603265762, 0.2922643721103668, 0.09092120826244354, 0.11736846715211868, -0.28290241956710815, -0.007356392685323954, -0.4521167576313019, 0.35754692554473877, 0.3472752869129181, 0.4865120053291321, 0.21581467986106873, 0.039204370230436325, 0.00824748631566763, 0.040773678570985794, 0.2137785106897354, -0.029948169365525246, 0.013004876673221588, 0.255681574344635, -0.13648797571659088, 0.061477258801460266, -0.22550569474697113, -0.0926547721028328, 0.07190167158842087, 0.08172453939914703, -0.07073019444942474, -0.31187528371810913, -0.15123584866523743, 0.3328525722026825, 0.15579360723495483, -0.11175574362277985, -0.24236740171909332, 0.008509926497936249, -0.15314169228076935, -0.19700324535369873, -0.19007037580013275, -0.2455592006444931, 0.1381574273109436, -0.12386270612478256, 0.447030246257782, 0.013527008704841137, -0.09884022921323776, 0.087062306702137, 0.024769239127635956, 0.18710507452487946, -0.15448355674743652, 0.13126251101493835, 0.17856891453266144, 0.002818757202476263, 0.3041512370109558, 0.2868058681488037, -0.02494513988494873, -0.2621815502643585, 0.024909837171435356, 0.11324093490839005, 0.3784663677215576, 0.3422502875328064, 0.19741098582744598, 0.06382720917463303, -0.2843545079231262, -0.1820516139268875, -0.14466549456119537, 0.1674746870994568, 0.35075610876083374, -0.10339299589395523, 0.051271505653858185, -0.23811379075050354, 0.6501339077949524, 0.1520877331495285, -0.2175855189561844, 0.2924673557281494, -0.19579976797103882, -0.251296728849411, -0.12040609121322632, 0.1437394767999649, 0.977217972278595, 0.17870554327964783, 0.19215720891952515, 0.3799651563167572, -0.2701801359653473, 0.5282027721405029, 0.12167201936244965, 0.1143144816160202, -0.49746766686439514, -0.03680960834026337, 0.01277187280356884, -0.21958330273628235, 0.226039856672287, -0.15513774752616882, -0.23718324303627014, 0.3542986214160919, -0.010540386661887169, 0.2823977768421173, 0.3365882933139801, 0.3203302323818207, -0.22276349365711212, -0.026248745620250702, -0.16136066615581512, 0.04255669564008713, -0.0497179739177227, 0.4177896976470947, -0.2992788851261139, 0.009875716641545296, -0.08630738407373428, -0.045527052134275436, -0.32561591267585754, -0.2006007879972458, -0.29272347688674927, 0.01719260960817337, 0.0827983096241951, -0.13996829092502594, -0.010616450570523739, 0.1779652088880539, 0.08981616050004959, -0.216007262468338, -0.09981653839349747, 0.09243496507406235, -0.1403931975364685, 0.2634115219116211, 0.38699325919151306, 0.09648367762565613, 0.2686980664730072, -0.34610801935195923, -0.09964431077241898, 0.30464765429496765, -0.23491984605789185, -0.24699541926383972, -0.041801001876592636, -0.20667646825313568, 0.13844741880893707, -0.19164036214351654, -0.3267771601676941, -0.025652840733528137, 0.0002864466223400086, -0.13440848886966705, -0.0031315407250076532, 0.1647346317768097, -0.3638900816440582, -0.008035240694880486, -0.2035222351551056, -0.25346139073371887, 0.04944063350558281, 0.4804391860961914, -0.02904575690627098, -0.22069190442562103, 0.6655134558677673, -0.26307475566864014, 0.11766959726810455, -0.12205168604850769, -0.007447111885994673, -0.09300627559423447, -0.10929881781339645, 0.07604224979877472, -0.1761193871498108, -0.1763356626033783, -0.010683232918381691, 0.5820416808128357, 0.3484698534011841, 0.08524666726589203, -0.15745826065540314, -0.4427753686904907, -0.40297091007232666, 0.017590384930372238, -0.06714548170566559, 0.1379370242357254, -0.18047980964183807, 0.18105769157409668, 0.12914854288101196, 0.08300849050283432, -0.17618049681186676, -0.10881161689758301, 0.02771729603409767, 0.10157670825719833, -0.13341473042964935, 0.1775045543909073, 0.005330807529389858, 0.14194735884666443, -0.06230195611715317, 0.17821501195430756, -0.2920292317867279, -0.052992310374975204, -0.1140306293964386, 0.2268325239419937, 0.023889265954494476, 0.060892973095178604, 0.1892106533050537, -0.04796650633215904, 0.015474878251552582, 0.24556539952754974, -0.10414794832468033, -0.03355497866868973, -0.05473880469799042, 0.2197341024875641, 0.3419259190559387, 0.04210347682237625, -0.13635236024856567, 0.23928789794445038, 0.2469775527715683, 0.21957458555698395, -0.1397828310728073, 0.37925583124160767, -0.1171855479478836, 0.0650605857372284, 0.2873701751232147, 0.04528270661830902, -0.0799846351146698, 0.2675742208957672, 0.09384787827730179, 0.03071211278438568, 0.11529325693845749, -0.07018136233091354, 0.22991909086704254, 0.8423089385032654, -0.24265415966510773, -0.11675290018320084, 0.4015064239501953, 0.009781548753380775, -0.23323599994182587, -0.08031415939331055, 0.2138211876153946, 0.06446272134780884, 0.00968229304999113, 0.294380784034729, 0.049018945544958115, -0.017301369458436966, 0.31864213943481445, 0.09393884241580963, 0.08764542639255524, -0.14573411643505096, -0.07335060834884644, 0.5390332937240601, -0.26027193665504456, 0.21228520572185516, 0.39230507612228394, 0.13994279503822327, 0.3845628798007965, 0.15995639562606812, -0.06687816232442856, 0.5641663074493408, -0.06525655835866928, 0.007685437798500061, 0.20442354679107666, -0.08264785259962082, -0.08976011723279953, 0.3919523358345032, 0.02877570502460003, 0.022416604682803154, 0.30100804567337036, 0.3995402455329895, -0.34889495372772217, -0.16715607047080994, -0.20334205031394958, 0.030661925673484802, 0.12445730715990067, -0.1107570230960846, -0.23745104670524597, 0.005827812012284994, -0.2603165805339813, -0.058111440390348434, -0.18110883235931396, -0.3672564923763275, 0.2537263035774231, -0.051365774124860764, -0.39613714814186096, -0.17723986506462097, -0.5898517966270447, 0.02973223850131035, -0.10691597312688828, -0.26929721236228943, 0.05167815834283829, 0.03170264884829521, 0.07564802467823029, 0.2259105145931244, 0.44081342220306396, 0.36183953285217285, 0.10118669271469116, -0.3247075080871582, -0.31040719151496887, -0.2057202011346817, 0.1689281016588211, -0.0008589329663664103, 0.2647930085659027, 0.06462588906288147, -0.1833549439907074, 0.4019775390625, 0.03411140665411949, 0.05196738615632057, 0.1816425770521164, 0.4109753668308258, -0.4785666763782501, -0.2713506519794464, 0.9588332176208496, 0.01858782209455967, -0.0753745511174202, -0.10514386743307114, 0.2092442512512207, -0.24594128131866455, 0.15358959138393402, -0.054069966077804565, -0.02507866732776165, 0.1652688831090927, 0.20228450000286102, 0.011583523824810982, -0.04843883588910103, 0.28346461057662964, 0.29296818375587463, -0.06524429470300674, -0.2998436391353607, -0.3237364590167999, -0.8680163025856018, 0.1610240340232849, 0.0929507315158844, -0.27846941351890564, -0.019421817734837532, -0.28995683789253235, 0.17587050795555115, 0.1876770555973053, -0.24990339577198029, 0.2563641667366028, -0.007288945373147726, -0.03545403853058815, -0.28572240471839905, -0.289183109998703, -0.2412431687116623, -0.023864371702075005, -0.022986765950918198, -0.39517971873283386, 0.19543127715587616, -0.4377289116382599, -0.1360102891921997, 0.16049228608608246, -0.051106762140989304, 0.4187006652355194, 0.07762985676527023, 0.44485777616500854, -0.03417625278234482, 0.5052465796470642, 0.15348993241786957, -0.050153717398643494, -0.04782102257013321, 0.25087103247642517, -0.1411207914352417, 0.11360593140125275, -0.07329358160495758, 0.09428734332323074, -0.17538884282112122, 0.6784104704856873, -0.4178822338581085, 0.4013877213001251, 0.011650527827441692, 0.12361020594835281, -0.4917759597301483, -0.002498371759429574, -0.17377296090126038, 0.14623793959617615, 0.37973955273628235, 0.253579318523407, -0.17763745784759521, 0.36763089895248413, -0.4226216971874237, 0.02618410810828209, 0.504698634147644, -0.16471503674983978, -0.0771922841668129, -0.04225714132189751, 0.31683632731437683, -0.046242352575063705, -0.20655162632465363, -0.5365793108940125, 0.06525743007659912, 0.24328750371932983, -0.06352650374174118, 0.14099588990211487, 0.20620769262313843, 0.17869806289672852, 0.07907188683748245, -0.0381837822496891, 0.822300136089325, -0.0023480707313865423, 0.11937950551509857, -0.19420164823532104, -0.2611883878707886 ]
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
[ 0.05404609441757202, 0.07422395050525665, 0.009660082869231701, 0.17164768278598785, 0.018456941470503807, -0.18276670575141907, 0.40755581855773926, 0.1807366907596588, 0.03177383169531822, 0.24988555908203125, 0.038060180842876434, 0.0683055967092514, -0.06527305394411087, -0.25279682874679565, 0.01663317158818245, 0.15957821905612946, 0.1347912847995758, 0.044665589928627014, -0.21993470191955566, -0.005511440336704254, -0.3515840470790863, 0.2071038782596588, -0.1873503178358078, 0.1459728479385376, -0.0956173986196518, 0.27064040303230286, -0.22023676335811615, 0.2536098062992096, -0.19941842555999756, -0.5675514936447144, 0.1467324048280716, 0.0295448899269104, 0.16571645438671112, 0.574688732624054, -0.00013265380403026938, -0.07714195549488068, 0.5401034951210022, -0.1227799728512764, -0.6687530875205994, -0.23048801720142365, 0.28792858123779297, -0.15296025574207306, 0.20211315155029297, -0.08125124126672745, 0.2706156373023987, -0.17419052124023438, 0.20436598360538483, -0.06535616517066956, 0.12489988654851913, 0.352989137172699, 0.019501974806189537, 0.1126968041062355, -0.010982497595250607, 0.07280845940113068, 0.09343899041414261, 0.25168946385383606, -0.09757187962532043, 0.4664694666862488, 0.2114461064338684, 0.0007745109032839537, -0.04591601341962814, -0.015806561335921288, 0.019108431413769722, 0.4609079957008362, 0.3064594268798828, 0.18696483969688416, 0.46506136655807495, 0.10373448580503464, -0.012513532303273678, -0.03786543384194374, 0.6569064259529114, -0.4838958978652954, -0.43441078066825867, -0.08118182420730591, -0.12728406488895416, -0.20948481559753418, 0.3831119239330292, -0.10318626463413239, -0.3200289309024811, 0.09504386782646179, 0.014491116628050804, -0.26167526841163635, -0.03538685664534569, 0.2144232541322708, -0.3736017048358917, 0.39959239959716797, 0.03471045941114426, 0.1933126449584961, 0.07282914221286774, -0.0495954193174839, 0.30319875478744507, -0.3424917757511139, 0.04692396521568298, 0.16503411531448364, 0.09963259845972061, -0.004732954315841198, -0.16642402112483978, -0.06243468075990677, -0.14533306658267975, 0.03446194902062416, -0.4331531524658203, -0.13929206132888794, -0.18535862863063812, 0.2131970375776291, 0.18500934541225433, 0.4487713575363159, 0.2529619634151459, 0.4930517375469208, 0.178958460688591, 0.1321784257888794, 0.13476142287254333, 0.18852676451206207, -0.007200551684945822, -0.44622159004211426, -0.0033644414506852627, 0.1070360392332077, 0.21821117401123047, -0.2600003182888031, -0.263174444437027, -0.2860173285007477, -0.46615469455718994, -0.03614896163344383, -0.05770577862858772, 0.2854340374469757, -0.2194119244813919, -0.11122121661901474, 0.06850847601890564, 0.38588759303092957, -0.19950072467327118, 0.05821707099676132, 0.0005525426822714508, 0.03904839605093002, -0.2579004466533661, 0.036313507705926895, 0.09758543223142624, 0.11273565143346786, 0.2136290967464447, -0.3273884952068329, 0.13679760694503784, -0.14557074010372162, 0.07292045652866364, -0.0004480592906475067, 0.11817514896392822, 0.35788652300834656, -0.1497725248336792, 0.19033083319664001, 0.10121169686317444, -0.45273321866989136, -0.25621846318244934, 0.29938071966171265, -0.10430818051099777, -0.4040023386478424, 0.09910067170858383, -0.005106224212795496, -0.4261918067932129, 0.06087484955787659, 0.0788298100233078, -0.3964565396308899, 0.19351842999458313, -0.5260318517684937, 0.06719458848237991, -0.2752557694911957, -0.28047803044319153, -0.12022757530212402, -0.09893929213285446, 0.48066288232803345, -0.3474731147289276, -0.2877001464366913, -0.3806164264678955, -0.0432821661233902, 0.12351839989423752, -0.07172437012195587, -0.12091962993144989, 0.04395836591720581, -0.562542200088501, 0.01875755563378334, 0.43581637740135193, -0.5859131217002869, -0.3239838778972626, 0.3623732328414917, -0.5026475787162781, -0.19662758708000183, 0.4976199269294739, -0.11810917407274246, -0.019310563802719116, -0.4362891614437103, 0.00029172777431085706, 0.04905928298830986, 0.06989682465791702, -0.04112544283270836, -0.1771133989095688, -0.20519767701625824, -0.02500290237367153, 0.08768301457166672, -0.1762070208787918, 0.10934789478778839, 0.04550109803676605, -0.10576528310775757, 0.4786716103553772, 0.07372603565454483, 0.09647493064403534, -0.04458683356642723, 0.2020455002784729, 0.05049119517207146, -0.04118293151259422, 0.25568580627441406, -0.5695755481719971, 0.14952592551708221, -0.15539132058620453, 0.1262124627828598, -0.24375773966312408, -0.08564534038305283, -0.3687710165977478, 0.06484155356884003, -0.1458721160888672, -0.15639923512935638, -0.10608262568712234, 0.5268374681472778, -0.016501430422067642, -0.06277436017990112, 0.04309353977441788, 0.08777246624231339, -0.14425550401210785, 0.24223831295967102, -0.11042921245098114, 0.005361850839108229, -0.05164015293121338, 0.08585337549448013, -0.07348965853452682, 0.030897891148924828, 0.2242913395166397, -0.40818125009536743, 0.033208467066287994, 0.3498998284339905, 0.2747839093208313, 0.46276599168777466, -0.02255517989397049, -0.13486960530281067, 0.1559962034225464, 0.24798265099525452, 0.3091241121292114, 0.3445506989955902, 0.08198758214712143, -0.28991881012916565, -0.01919381134212017, 0.2700665593147278, 0.16893000900745392, 0.31897249817848206, -0.03178844600915909, -0.04940670728683472, 0.041294340044260025, 0.13392695784568787, -0.0626228079199791, -0.2984601557254791, 0.1893182098865509, 0.16936105489730835, 0.5996586084365845, 0.07128755003213882, 0.32798266410827637, 0.025674529373645782, -0.17272047698497772, -0.27904754877090454, -0.0857865959405899, -0.00003843488229904324, -0.18876491487026215, -0.08370199799537659, 0.09133265912532806, 0.4056796729564667, 0.9681352972984314, -0.09375651180744171, 0.332884281873703, 0.033049486577510834, -0.10807810723781586, -0.2747821509838104, 0.2888104319572449, 0.2276335060596466, -0.07494214177131653, 0.17336972057819366, -0.0313984714448452, -0.318402498960495, -0.040105074644088745, -0.05030503496527672, 0.5000061392784119, 0.14260078966617584, -0.5107771158218384, 0.22430644929409027, -0.25457602739334106, -0.30979371070861816, -0.6035182476043701, -0.31199508905410767, -0.060537807643413544, -0.5021188259124756, -0.09551544487476349, 0.39612290263175964, -0.030482972040772438, 0.20940719544887543, -0.07476459443569183, 0.0845911055803299, -0.44329190254211426, -0.44742313027381897, -0.014093589037656784, 0.00826400425285101, -0.48457759618759155, -0.1461126208305359, 0.3326222002506256, -0.08755498379468918, 0.22779256105422974, -0.23387664556503296, -0.20161932706832886, -0.23572476208209991, -0.29723620414733887, 0.12632699310779572, 0.15794751048088074, 0.4473173916339874, 0.23237906396389008, 0.00020792857685592026, 0.4607328474521637, 0.0031124260276556015, 0.12738220393657684, -0.2060607522726059, -0.3891749083995819, -0.1575378030538559, 0.12728096544742584, 0.13422873616218567, -0.1593405157327652, -0.3914782702922821, -0.4792371392250061, -0.0361468531191349, -0.09529062360525131, 0.05918173864483833, 0.1256534457206726, 0.24842175841331482, -0.16251930594444275, 0.1452769637107849, 0.06382469832897186, -0.09572581946849823, -0.033040065318346024, 0.08951997011899948, 0.03843207657337189, -0.23889976739883423, -0.0015532771358266473, 0.04427354782819748, 0.007728737313300371, 0.05663514882326126, -0.09045357257127762, -0.46226149797439575, -0.4703611433506012, 0.1932879537343979, 0.05560008063912392, 0.18048302829265594, -0.10260498523712158, 0.45524054765701294, 0.29743990302085876, 0.13758741319179535, -0.0904003381729126, -0.34502050280570984, 0.024569131433963776, 0.08051776885986328, 0.3078869879245758, 0.15233784914016724, 0.3881365656852722, -0.19256757199764252, 0.8443019390106201, 0.12629885971546173, 0.03526840731501579, 0.08385485410690308, -0.15377208590507507, 0.18944503366947174, -0.08218682557344437, -0.03154938295483589, 0.1504143625497818, -0.08360805362462997, -0.5522130727767944, 0.3674585521221161, -0.057122182101011276, -0.2313217669725418, -0.06647651642560959, 0.2273530662059784, -0.14406514167785645, -0.15885142982006073, 0.07420063018798828, -0.1906692236661911, -0.01243407279253006, 0.16560135781764984, 0.002517790300771594, -0.26383131742477417, -0.21585756540298462, 0.012425120919942856, 0.30228421092033386, 0.027334822341799736, 0.08392296731472015, -0.3305441439151764, -0.12841558456420898, -0.14755140244960785, 0.3084337115287781, 0.33007603883743286, 0.4101948142051697, -0.015941977500915527, -0.11342453956604004, 0.2813287377357483, 0.007648499216884375, 0.3506704568862915, 0.06734642386436462, 0.21243351697921753, 0.19533105194568634, -0.0024001137353479862, -0.11793357133865356, -0.1861051768064499, -0.07736071944236755, 0.14718523621559143, 0.10939383506774902, 0.1820797175168991, -0.2384730577468872, -0.26433053612709045, 0.20997917652130127, 0.19632458686828613, -0.15524879097938538, -0.21320834755897522, -0.20942974090576172, -0.12904751300811768, -0.12419740855693817, -0.345580130815506, -0.13917407393455505, -0.0629340410232544, -0.09446372836828232, 0.387828528881073, 0.01809650845825672, -0.18357950448989868, 0.05521735921502113, 0.0047636935487389565, 0.30929064750671387, -0.18026484549045563, 0.15318964421749115, 0.36620134115219116, 0.10744716972112656, 0.40602561831474304, 0.3897082507610321, -0.1806139349937439, -0.12755848467350006, 0.025755397975444794, 0.09579354524612427, 0.39173397421836853, 0.6246480345726013, 0.04617780074477196, -0.18961071968078613, -0.2906065285205841, -0.21935446560382843, -0.23982897400856018, 0.09601417928934097, 0.4089130163192749, 0.009106334298849106, 0.10071633011102676, -0.28545665740966797, 0.5554724335670471, 0.20811249315738678, -0.35643690824508667, 0.5225512385368347, -0.00355436815880239, -0.2918553054332733, 0.11549779027700424, -0.05118603631854057, 1.1163864135742188, 0.2942344844341278, 0.42396605014801025, 0.36086007952690125, -0.2088659554719925, 0.09465111792087555, 0.11679020524024963, 0.1752215325832367, -0.46067431569099426, -0.02153201773762703, -0.019033221527934074, -0.33305877447128296, 0.23628020286560059, -0.03547033667564392, -0.09988828003406525, 0.40644264221191406, -0.04007876664400101, 0.3567257523536682, 0.19364900887012482, 0.15737156569957733, -0.29641982913017273, 0.028264818713068962, 0.026144489645957947, -0.04065540060400963, 0.20728260278701782, 0.45894938707351685, -0.16293080151081085, 0.03178032860159874, -0.26329663395881653, -0.07733353227376938, -0.030731867998838425, -0.07309769093990326, -0.13402140140533447, 0.07461819052696228, 0.02224455028772354, -0.251910001039505, -0.2052024006843567, 0.3793691098690033, 0.1928965449333191, -0.22925755381584167, 0.022599415853619576, 0.17318683862686157, -0.17188309133052826, 0.3654875159263611, 0.3521055579185486, -0.04134288430213928, 0.3046315312385559, -0.13484062254428864, -0.13243822753429413, 0.10160686820745468, -0.28353267908096313, -0.19922994077205658, -0.24217480421066284, -0.09257839620113373, 0.22591692209243774, -0.3246316909790039, -0.1483447253704071, -0.1495775729417801, -0.0864618644118309, -0.10077989101409912, -0.052191708236932755, 0.14122462272644043, -0.28765591979026794, 0.010074022226035595, -0.26084959506988525, -0.4288972318172455, -0.0010193543275818229, 0.5187016129493713, -0.12068278342485428, -0.055858906358480453, 0.8927492499351501, -0.2553897202014923, 0.07004079222679138, 0.03783557936549187, 0.16071440279483795, -0.2221294641494751, -0.2651761472225189, 0.06714926660060883, -0.18132175505161285, -0.24391543865203857, 0.02168780006468296, 0.5689193606376648, 0.32607367634773254, 0.08179114013910294, -0.3220890164375305, -0.46930259466171265, -0.24841611087322235, 0.006860717665404081, -0.1358150690793991, 0.1807660013437271, -0.20048141479492188, 0.028381092473864555, 0.18213126063346863, -0.11743344366550446, -0.10768835991621017, -0.019980765879154205, 0.1194516271352768, 0.22964100539684296, -0.1515393853187561, 0.05161181464791298, 0.06271813064813614, 0.03996075317263603, -0.08371783792972565, 0.10385242104530334, -0.2190839797258377, 0.056734584271907806, -0.08646655082702637, 0.2645622789859772, 0.0287005752325058, 0.004550891928374767, 0.006185134407132864, -0.32587993144989014, 0.1301831305027008, 0.06811775267124176, -0.023158850148320198, 0.028891706839203835, -0.050951194018125534, 0.042670220136642456, 0.3008289039134979, 0.06145782396197319, 0.029404647648334503, 0.38852688670158386, 0.05780842900276184, 0.2583293914794922, -0.1853509098291397, 0.21913090348243713, -0.2305954545736313, 0.020766349509358406, 0.040415260940790176, 0.10006487369537354, -0.01876332238316536, 0.20925168693065643, 0.13397367298603058, -0.18022282421588898, 0.11828558892011642, -0.10023929923772812, 0.363948255777359, 0.917977511882782, -0.14689354598522186, -0.23009678721427917, 0.4129929840564728, -0.09078711271286011, -0.05701126903295517, -0.12697502970695496, 0.07886944711208344, 0.13585875928401947, -0.05844684690237045, 0.3512888252735138, -0.05280636250972748, -0.1174372136592865, 0.3474317491054535, 0.21134528517723083, 0.20100338757038116, -0.23741693794727325, -0.10164041072130203, 0.40461185574531555, -0.305597722530365, 0.4028898775577545, 0.47488754987716675, 0.06230856105685234, 0.2003495693206787, 0.32515010237693787, 0.08612178266048431, 0.5218544006347656, -0.14537319540977478, -0.028906842693686485, 0.12090805172920227, -0.10828753560781479, -0.029849298298358917, 0.2552216351032257, -0.005976909305900335, 0.04436124116182327, 0.35601699352264404, 0.5558950304985046, -0.4765010178089142, -0.10528743267059326, -0.23334403336048126, -0.13780395686626434, 0.016764026135206223, -0.2931421101093292, -0.218849778175354, -0.03580227494239807, -0.13725592195987701, -0.03259795531630516, -0.19874653220176697, -0.19986774027347565, 0.28982943296432495, 0.05615776777267456, -0.2658580541610718, -0.22861966490745544, -0.3638242185115814, 0.019897572696208954, 0.039980459958314896, -0.34758296608924866, 0.25455915927886963, -0.0002536192478146404, 0.13587935268878937, 0.2975616157054901, 0.13145554065704346, 0.4875790476799011, 0.19343812763690948, -0.2696225643157959, -0.2617153227329254, -0.28797098994255066, -0.001829315209761262, -0.07789875566959381, 0.3554929792881012, -0.02986961416900158, -0.09149472415447235, 0.35833123326301575, -0.015709245577454567, -0.005666792392730713, 0.22060009837150574, 0.4076075851917267, -0.2764076292514801, -0.14980630576610565, 0.8099430799484253, -0.24034635722637177, -0.21010665595531464, -0.0757312998175621, 0.03368872404098511, -0.4592144787311554, 0.12718135118484497, 0.011043603532016277, -0.14476530253887177, 0.18752071261405945, 0.11137880384922028, -0.039163827896118164, -0.09055307507514954, 0.34215009212493896, 0.40220582485198975, -0.1118353083729744, -0.3300440013408661, -0.1995307058095932, -0.8534042239189148, 0.2650274634361267, -0.05009031668305397, -0.14365532994270325, -0.20968367159366608, -0.07952211052179337, 0.21403388679027557, -0.017104575410485268, -0.11020110547542572, 0.1108066514134407, 0.05849335342645645, 0.1832571178674698, -0.30739831924438477, -0.37756311893463135, -0.18099553883075714, 0.039649978280067444, -0.00847790390253067, -0.5915560126304626, 0.1994037628173828, -0.41751715540885925, -0.15232011675834656, 0.161967471241951, -0.07171511650085449, 0.3689970076084137, 0.20901598036289215, 0.36451995372772217, -0.10860451310873032, 0.6161584258079529, 0.1779773086309433, -0.1468055248260498, -0.262819766998291, 0.05785923823714256, -0.09725837409496307, 0.018361615017056465, -0.08927461504936218, 0.17368991672992706, -0.32429012656211853, 0.41180843114852905, -0.3782767951488495, 0.22013433277606964, 0.08509602397680283, -0.0034855755511671305, -0.2222464233636856, 0.18447372317314148, -0.3588894009590149, 0.17099492251873016, 0.3244398534297943, 0.32722920179367065, 0.04074658080935478, 0.3295932412147522, -0.3035624623298645, -0.24247664213180542, 0.640537440776825, -0.099056176841259, -0.16434797644615173, 0.13678689301013947, 0.2945442497730255, -0.011159390211105347, -0.2771061360836029, -0.5036940574645996, 0.17269374430179596, 0.24136312305927277, 0.06933799386024475, 0.02412627637386322, 0.05193420499563217, 0.25292181968688965, 0.017193133011460304, -0.06566262990236282, 0.9180787801742554, 0.031080778688192368, 0.07061402499675751, 0.09902803599834442, -0.20347753167152405 ]
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?
[ -0.07890302687883377, 0.28551438450813293, 0.00024860110715962946, 0.052755165845155716, 0.05063405632972717, -0.260754257440567, 0.21438190340995789, 0.27597564458847046, -0.05403408408164978, 0.08017336577177048, 0.2072276473045349, 0.22357390820980072, -0.2214677929878235, -0.03458124026656151, 0.12442827969789505, 0.054476212710142136, -0.10658468306064606, 0.0847870409488678, -0.22504180669784546, 0.1550331562757492, -0.24932251870632172, 0.17533214390277863, 0.003421714762225747, 0.1672305166721344, 0.015413636341691017, 0.14610639214515686, -0.22377106547355652, 0.20182134211063385, -0.41943028569221497, -0.7930229902267456, 0.3121378719806671, 0.03021336905658245, 0.23527850210666656, 0.38400545716285706, -0.00012361556582618505, -0.09185246378183365, 0.5897998213768005, -0.1559891700744629, -0.7334057688713074, -0.42152419686317444, 0.24337323009967804, -0.5072745680809021, 0.27730071544647217, -0.15510356426239014, 0.18571291863918304, -0.06349453330039978, 0.2981804311275482, 0.20960934460163116, 0.18837349116802216, 0.40039685368537903, 0.07241328805685043, 0.3003309965133667, 0.05199890583753586, -0.11267975717782974, 0.008521407842636108, 0.17452745139598846, -0.05216110497713089, 0.5873817205429077, 0.20252273976802826, 0.07431744784116745, -0.15627048909664154, -0.07906939089298248, 0.10531937330961227, 0.502699077129364, 0.1522771716117859, 0.06298096477985382, 0.39901185035705566, 0.04652860388159752, -0.26758769154548645, 0.08663006126880646, 0.6675556898117065, -0.5247458219528198, -0.3587067723274231, -0.07098351418972015, -0.07901199907064438, -0.05743113532662392, 0.3807906210422516, -0.1414186954498291, -0.4126582145690918, 0.09209895133972168, -0.10644984245300293, -0.1712896078824997, -0.03199552744626999, 0.39670464396476746, -0.29836413264274597, 0.40517210960388184, -0.018465669825673103, 0.1622788906097412, 0.15625673532485962, 0.05911780893802643, 0.25665172934532166, -0.17506061494350433, 0.17345495522022247, 0.18063056468963623, 0.10605969280004501, -0.20942503213882446, -0.28958410024642944, -0.27422618865966797, 0.1452915221452713, 0.11145225912332535, -0.3082638084888458, -0.19240982830524445, -0.06261329352855682, 0.22528693079948425, 0.2307129055261612, 0.324116051197052, 0.17963656783103943, 0.26923829317092896, 0.2589323818683624, 0.3053898811340332, 0.27003681659698486, 0.16195181012153625, 0.007208313327282667, -0.33093100786209106, -0.12101811170578003, 0.15665869414806366, 0.20481866598129272, -0.2120765596628189, -0.19729717075824738, -0.24859653413295746, -0.40486738085746765, -0.12655949592590332, -0.06877494603395462, 0.31161004304885864, -0.1900971233844757, -0.1030607670545578, 0.09242871403694153, 0.3832481801509857, -0.1280381679534912, -0.07070960104465485, 0.028685377910733223, 0.04314432293176651, -0.24823197722434998, 0.028636908158659935, 0.2963486611843109, 0.22021248936653137, 0.4854155480861664, -0.29047420620918274, -0.16111870110034943, -0.05845465511083603, 0.12964698672294617, 0.043460216373205185, -0.047212421894073486, 0.21914930641651154, -0.07588888704776764, 0.1827896535396576, 0.09156720340251923, -0.30373260378837585, -0.2998288869857788, 0.17252537608146667, -0.07229047268629074, -0.40235957503318787, 0.10723471641540527, 0.0654560998082161, -0.5473222732543945, 0.021734289824962616, 0.09589467197656631, -0.16957348585128784, -0.057539671659469604, -0.3571426570415497, 0.04796435311436653, -0.23688262701034546, -0.2540796995162964, -0.15955203771591187, -0.013156778179109097, 0.333122581243515, -0.16873471438884735, -0.3116174042224884, -0.4004579186439514, -0.2978609502315521, 0.21164460480213165, -0.1955411434173584, -0.1178334578871727, 0.2728707790374756, -0.49670082330703735, 0.06526114791631699, 0.6163497567176819, -0.4722353219985962, -0.299718976020813, 0.5806671380996704, -0.45252543687820435, -0.06820155680179596, 0.26381435990333557, -0.036954063922166824, -0.22122478485107422, -0.2977391183376312, -0.08449379354715347, 0.13490734994411469, 0.026060890406370163, -0.15002305805683136, -0.19787004590034485, -0.35361677408218384, 0.14594611525535583, 0.23883672058582306, -0.08798091858625412, 0.0926896408200264, 0.053085099905729294, -0.07039504498243332, 0.4528985619544983, 0.10668683797121048, 0.06376813352108002, -0.061137713491916656, 0.12078893184661865, -0.19681982696056366, -0.10716599971055984, 0.026188118383288383, -0.5845913887023926, 0.12513156235218048, -0.3148135542869568, 0.19763021171092987, -0.18850240111351013, -0.0742160826921463, -0.5055440068244934, -0.018278587609529495, 0.022848855704069138, -0.2564987540245056, 0.009734353981912136, 0.3550097644329071, 0.09080709517002106, -0.04917643964290619, 0.03225114569067955, 0.01558596733957529, -0.20264917612075806, 0.2942972183227539, -0.01093258149921894, 0.06413184851408005, -0.08776215463876724, 0.11537975072860718, 0.24681350588798523, 0.2504462003707886, 0.16802041232585907, -0.40396809577941895, 0.03069000504910946, 0.3176621198654175, 0.1910523772239685, 0.183998242020607, 0.20126758515834808, -0.06688780337572098, 0.16839487850666046, 0.19109980762004852, 0.08350725471973419, 0.46942803263664246, 0.047558918595314026, -0.129586860537529, 0.045802608132362366, 0.20797207951545715, 0.1549786478281021, -0.002130033215507865, -0.07428238540887833, 0.07928673177957535, 0.34484007954597473, 0.022174546495079994, -0.10134486109018326, -0.25451672077178955, 0.1040521040558815, 0.33493462204933167, 0.6334260702133179, -0.006708210334181786, 0.2735130488872528, -0.10974109172821045, -0.14043641090393066, -0.3109171390533447, 0.04697338119149208, 0.14044460654258728, -0.18512630462646484, -0.1090429499745369, 0.0009038840653374791, 0.37554624676704407, 0.8084439635276794, -0.008677445352077484, 0.3378496468067169, 0.09863553941249847, -0.163443461060524, -0.2317846119403839, 0.24697159230709076, 0.17262253165245056, -0.09168842434883118, 0.12587019801139832, -0.15991602838039398, -0.3018523156642914, -0.14503931999206543, 0.019934969022870064, 0.4233739674091339, 0.31813257932662964, -0.4547545909881592, -0.056838784366846085, -0.29740625619888306, -0.5187102556228638, -0.5051659345626831, -0.31656914949417114, -0.20915921032428741, -0.3562679886817932, -0.05608318746089935, 0.2834828197956085, -0.13092844188213348, 0.1668250411748886, -0.11503621935844421, 0.25020647048950195, -0.3294556140899658, -0.305985689163208, 0.05414086952805519, -0.09253209084272385, -0.43069514632225037, 0.0013972087763249874, 0.40533071756362915, 0.09085648506879807, 0.2507202625274658, -0.2317570298910141, -0.2034362107515335, -0.26207900047302246, -0.27703648805618286, 0.13477779924869537, 0.1804635226726532, 0.556559681892395, 0.44621291756629944, -0.0005804250249639153, 0.41545507311820984, 0.023446349427103996, 0.021319299936294556, -0.17712341248989105, -0.3594783842563629, -0.16657453775405884, 0.11461659520864487, 0.11637777090072632, -0.1748236119747162, -0.4352657198905945, -0.5285281538963318, -0.05852999910712242, 0.029240112751722336, 0.08637495338916779, 0.15409404039382935, 0.09188544750213623, -0.17155224084854126, 0.18889690935611725, 0.04282170906662941, -0.14498664438724518, -0.15512079000473022, -0.07493051141500473, 0.026332801207900047, -0.14079639315605164, -0.17391923069953918, 0.008674859069287777, -0.2095598578453064, 0.05793795734643936, -0.16223233938217163, -0.4062245786190033, -0.5565922856330872, 0.14096714556217194, 0.18127132952213287, 0.25342312455177307, -0.15074816346168518, 0.37731093168258667, 0.18244114518165588, 0.09180489927530289, -0.132671058177948, -0.33633488416671753, 0.24598191678524017, 0.1070009395480156, 0.40160319209098816, 0.20600655674934387, 0.2957025468349457, -0.02632216364145279, 0.7256444096565247, 0.14918839931488037, -0.07229574769735336, 0.09985034167766571, -0.18606211245059967, 0.10383373498916626, 0.15628527104854584, 0.005900402553379536, 0.1845201998949051, -0.158172145485878, -0.26709094643592834, 0.31861916184425354, -0.197682186961174, -0.04666416719555855, -0.1197163537144661, 0.2846110761165619, -0.2596029043197632, -0.12736395001411438, 0.12735620141029358, -0.05860144644975662, -0.04352826252579689, 0.09806595742702484, -0.04892711713910103, -0.16124720871448517, -0.061358459293842316, 0.1211145892739296, 0.35151782631874084, 0.033287715166807175, 0.1457095444202423, -0.3123936057090759, -0.04856237769126892, -0.3304588496685028, 0.2873315215110779, 0.3822646737098694, 0.429168164730072, 0.058980733156204224, 0.013043897226452827, 0.09527935087680817, 0.001286008395254612, 0.28689947724342346, 0.01905209757387638, 0.15850669145584106, 0.13101910054683685, -0.15485486388206482, 0.024732084944844246, -0.15523311495780945, -0.0956423282623291, 0.09846913814544678, 0.0075195650570094585, 0.15079323947429657, -0.4683271646499634, -0.13879448175430298, 0.1901916265487671, 0.11472011357545853, -0.058022379875183105, -0.2220773994922638, -0.01593135856091976, -0.12120787054300308, -0.04730749502778053, -0.22213146090507507, -0.14633381366729736, -0.009872944094240665, -0.10142967104911804, 0.2902648448944092, 0.07743430882692337, -0.14815443754196167, -0.01483830064535141, 0.08857021480798721, 0.2045021504163742, -0.3280986249446869, 0.11342329531908035, 0.2718534469604492, 0.19918057322502136, 0.2238496094942093, 0.40062451362609863, -0.10000590234994888, -0.21022135019302368, 0.03399193659424782, 0.05974175035953522, 0.3214838206768036, 0.40066125988960266, 0.10434781014919281, -0.039059609174728394, -0.25135987997055054, -0.15662217140197754, -0.3356707990169525, 0.22357980906963348, 0.4120948314666748, -0.10309459269046783, 0.1445939689874649, -0.22911982238292694, 0.6646548509597778, 0.13159409165382385, -0.23838545382022858, 0.5089073777198792, -0.28388920426368713, -0.2806603014469147, -0.07736442238092422, 0.046777043491601944, 1.02817702293396, 0.24669434130191803, 0.2519366145133972, 0.4927612543106079, -0.2603977918624878, 0.2928634583950043, 0.0694463923573494, 0.0985812395811081, -0.43218109011650085, 0.01908106915652752, -0.06405007839202881, -0.20193001627922058, 0.25413885712623596, 0.03236907720565796, -0.2334105521440506, 0.33858874440193176, 0.132755845785141, 0.38942983746528625, 0.24295948445796967, 0.4002091884613037, -0.2431943118572235, -0.02153211086988449, -0.2314874529838562, 0.04449516907334328, -0.029154596850275993, 0.35789698362350464, -0.2018136829137802, -0.08174538612365723, -0.2016921043395996, -0.13223761320114136, -0.0011774834711104631, -0.08184613287448883, -0.3922838866710663, 0.02882993593811989, 0.06671180576086044, -0.10534244030714035, -0.23339541256427765, 0.20267951488494873, 0.1094052791595459, -0.27650630474090576, -0.1312546283006668, 0.14214500784873962, -0.31567662954330444, 0.2580554783344269, 0.32302427291870117, -0.006119505502283573, 0.24911470711231232, -0.33159470558166504, -0.0262776929885149, 0.07291920483112335, -0.32363563776016235, -0.1923999786376953, 0.0774712935090065, -0.21757866442203522, 0.07555209845304489, -0.31039899587631226, -0.2757740318775177, -0.13168898224830627, -0.08827878534793854, -0.11784470081329346, -0.004206700716167688, 0.2224295288324356, -0.3529711365699768, -0.07808299362659454, -0.06195702403783798, -0.26087629795074463, -0.00546250119805336, 0.4155064523220062, -0.2048795372247696, -0.048417095094919205, 0.6307446360588074, -0.4179270565509796, 0.093370221555233, -0.07985344529151917, 0.13359001278877258, -0.1575952172279358, -0.11624106764793396, 0.05989886820316315, -0.08217968791723251, -0.314584881067276, -0.05027598515152931, 0.5699489116668701, 0.42291882634162903, 0.15990449488162994, -0.18029305338859558, -0.4978507459163666, -0.3021978735923767, 0.04070747643709183, -0.1916135847568512, 0.192902609705925, -0.0863233134150505, 0.21488814055919647, 0.2381758838891983, 0.0008389143040403724, -0.17753928899765015, -0.11061957478523254, -0.044823504984378815, 0.18705302476882935, -0.1764594316482544, 0.028117723762989044, 0.087776780128479, 0.1109379380941391, -0.05144146829843521, 0.07127981632947922, -0.26637980341911316, -0.03337728604674339, -0.10417585074901581, 0.22602678835391998, 0.15439240634441376, -0.0742286741733551, 0.1739833652973175, -0.14891166985034943, 0.028481600806117058, 0.1557396948337555, -0.13240185379981995, -0.051179803907871246, -0.030449457466602325, 0.11647067964076996, 0.23405788838863373, 0.04433012753725052, -0.20257164537906647, 0.2824554443359375, -0.009379423223435879, 0.15508577227592468, -0.09460099041461945, 0.20901906490325928, -0.3223203718662262, 0.03160146623849869, 0.17144255340099335, 0.09846799075603485, -0.06700026243925095, 0.2769372761249542, 0.1365860402584076, 0.0017017460195347667, 0.0024579225573688745, -0.047430362552404404, 0.3074740171432495, 0.747338593006134, -0.2027827799320221, -0.1288806051015854, 0.4639030396938324, -0.0035371938720345497, -0.211854487657547, 0.06120779365301132, 0.1913791298866272, 0.004370989743620157, 0.07066092640161514, 0.2556614875793457, 0.04539068043231964, 0.07980195432901382, 0.25364449620246887, 0.05678645893931389, 0.1590503454208374, -0.06254474818706512, -0.05567091703414917, 0.5163019299507141, -0.24758435785770416, 0.34521743655204773, 0.34482505917549133, -0.02677176333963871, 0.31721264123916626, 0.04373558983206749, -0.022251520305871964, 0.5483088493347168, -0.005632554180920124, 0.007812442723661661, 0.1248817890882492, 0.00736754247918725, -0.10260381549596786, 0.20615404844284058, 0.09833581000566483, 0.05027981847524643, 0.2542817294597626, 0.6443101763725281, -0.530163586139679, -0.1942591816186905, -0.2695655822753906, -0.0984446257352829, 0.08751729875802994, -0.1178487092256546, -0.45354077219963074, -0.008798843249678612, -0.3257633149623871, -0.14106622338294983, -0.13782067596912384, -0.34438979625701904, 0.25596413016319275, 0.003879873314872384, -0.3412185311317444, -0.25168558955192566, -0.4334021210670471, 0.13744868338108063, -0.04456019401550293, -0.3266587257385254, 0.20112894475460052, 0.033438265323638916, 0.031245745718479156, 0.2748129665851593, 0.29906925559043884, 0.4600580930709839, 0.07712918519973755, -0.28695201873779297, -0.34785836935043335, -0.19405251741409302, 0.0755385234951973, -0.07693590223789215, 0.2965199947357178, 0.006350375711917877, -0.16871410608291626, 0.5234033465385437, 0.04897024855017662, 0.01559342723339796, 0.21652773022651672, 0.4239833652973175, -0.4900082051753998, -0.14891736209392548, 0.7677968740463257, -0.07951386272907257, -0.08629511296749115, -0.053112514317035675, -0.012460045516490936, -0.271552711725235, 0.06507037580013275, -0.1014523133635521, -0.13405345380306244, 0.2226327657699585, 0.1248546913266182, 0.026110300794243813, -0.059313151985406876, 0.2937096953392029, 0.3687489330768585, -0.1291072815656662, -0.4022064507007599, -0.2841159701347351, -0.9730969071388245, 0.07774335891008377, 0.019219394773244858, -0.3939652740955353, -0.07160518318414688, -0.27665695548057556, 0.21714937686920166, 0.11051176488399506, -0.21781712770462036, 0.2072969228029251, 0.2407177984714508, 0.04598325863480568, -0.26249322295188904, -0.31379956007003784, -0.24278908967971802, 0.036388054490089417, -0.04189567640423775, -0.42559242248535156, 0.11773175746202469, -0.2274351716041565, -0.14005209505558014, 0.28931406140327454, -0.04074511304497719, 0.3404313325881958, 0.07629311829805374, 0.28897252678871155, -0.07201769202947617, 0.6170161366462708, 0.10390694439411163, 0.010898228734731674, -0.11753685772418976, 0.2104000598192215, -0.11691698431968689, 0.06729161739349365, -0.07970216870307922, 0.20908330380916595, -0.18039117753505707, 0.6635019779205322, -0.5068442821502686, 0.43745550513267517, 0.04022597149014473, 0.08135269582271576, -0.26135608553886414, 0.10217384248971939, -0.3644084632396698, 0.13158947229385376, 0.3581804931163788, 0.28291067481040955, -0.12177626043558121, 0.30615198612213135, -0.13787060976028442, -0.1978144645690918, 0.3982645869255066, -0.1261209398508072, -0.04153156653046608, 0.04977935552597046, 0.2727915346622467, -0.09283066540956497, -0.06110864505171776, -0.4452347755432129, 0.05422287434339523, 0.2737286388874054, 0.02921048365533352, 0.112928606569767, 0.09593446552753448, 0.15679363906383514, 0.11953739821910858, -0.062240466475486755, 0.8743889927864075, 0.04461148381233215, 0.017599137499928474, 0.0011316041927784681, -0.24424104392528534 ]
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.
[ -0.16123469173908234, 0.2551891505718231, 0.016922514885663986, -0.06889128684997559, -0.018733125180006027, -0.07994208484888077, 0.4035908579826355, 0.33871880173683167, 0.07540584355592728, 0.04867968708276749, 0.1928054243326187, 0.11071844398975372, -0.2993963062763214, 0.10281572490930557, 0.10059234499931335, -0.11661422252655029, 0.022554922848939896, 0.22912077605724335, -0.17938731610774994, 0.22091640532016754, -0.25348156690597534, 0.10033520311117172, -0.11082926392555237, 0.17837145924568176, -0.011252306401729584, 0.15829148888587952, -0.28702643513679504, 0.0593031644821167, -0.3393077254295349, -0.7205919027328491, 0.35620057582855225, 0.07208022475242615, 0.25911787152290344, 0.49416422843933105, -0.00012244142999406904, -0.07432948052883148, 0.5197003483772278, -0.14471912384033203, -0.6755970120429993, -0.23255515098571777, 0.24015560746192932, -0.3556254804134369, 0.3359568119049072, 0.0014606605982407928, 0.0567571260035038, -0.22579842805862427, 0.24793405830860138, 0.1120186373591423, 0.2297283411026001, 0.3637811541557312, 0.13691164553165436, 0.29282698035240173, 0.054873302578926086, -0.1232556626200676, -0.14173883199691772, -0.011339879594743252, -0.11851324141025543, 0.5441452860832214, 0.3419768214225769, 0.03315950185060501, 0.007737561594694853, -0.054533105343580246, 0.08724573254585266, 0.39128613471984863, 0.19234420359134674, 0.00016610170132480562, 0.5235510468482971, -0.04455597326159477, -0.11648385226726532, 0.09077635407447815, 0.7365921139717102, -0.5787573456764221, -0.3857159912586212, -0.00001621146839170251, 0.05161529779434204, -0.11142762750387192, 0.17219944298267365, -0.10055891424417496, -0.29932597279548645, 0.05937644839286804, -0.1098719909787178, -0.22395332157611847, -0.10024043917655945, 0.2735373079776764, -0.2138683944940567, 0.23304490745067596, 0.060238275676965714, 0.22848206758499146, 0.0010028486140072346, -0.10714864730834961, 0.19523385167121887, -0.13000643253326416, -0.056149356067180634, 0.2684307098388672, -0.09157221019268036, -0.18151724338531494, -0.344387024641037, -0.16297568380832672, 0.13957536220550537, -0.10119596123695374, -0.3182563781738281, -0.116225965321064, 0.04142558574676514, 0.1830257624387741, 0.15644407272338867, 0.3665720224380493, 0.27113908529281616, 0.20313182473182678, 0.12128373980522156, 0.14138710498809814, 0.2718665599822998, 0.20314235985279083, -0.06695923209190369, -0.3533778488636017, -0.012509217485785484, 0.09310371428728104, 0.20516276359558105, -0.2799244821071625, -0.16181974112987518, -0.21086330711841583, -0.36214479804039, 0.011534592136740685, -0.12015803903341293, 0.32975900173187256, -0.192222461104393, -0.18938544392585754, 0.07900412380695343, 0.5592828392982483, -0.20093844830989838, -0.21602347493171692, -0.014422508887946606, 0.024940894916653633, -0.21503762900829315, -0.0653628408908844, 0.23748525977134705, 0.009567680768668652, 0.4028279483318329, -0.2557929754257202, -0.2903884947299957, -0.006051949691027403, 0.23149867355823517, 0.02016107738018036, -0.19140711426734924, 0.2558794915676117, -0.0500890277326107, -0.024773042649030685, 0.10018891096115112, -0.332697331905365, -0.22785712778568268, 0.2212538868188858, 0.10084763169288635, -0.4667082130908966, -0.21180196106433868, 0.09065300226211548, -0.45038554072380066, 0.02167459949851036, 0.017005877569317818, -0.046402283012866974, 0.09855184704065323, -0.3617471754550934, -0.08257295936346054, -0.42291557788848877, -0.29159101843833923, -0.11434758454561234, -0.05740787833929062, 0.32801353931427, -0.4315403699874878, -0.2729126811027527, -0.40979182720184326, -0.2997347116470337, 0.10333647578954697, -0.26148858666419983, -0.07541261613368988, 0.27065980434417725, -0.3903684616088867, -0.1074509471654892, 0.5073695182800293, -0.5006062984466553, -0.25220510363578796, 0.44504258036613464, -0.3928949236869812, -0.1923522651195526, 0.17964528501033783, 0.03588736802339554, 0.0022716878447681665, -0.2983285188674927, -0.12099183350801468, 0.07886011898517609, 0.08520595729351044, -0.13364531099796295, -0.16009750962257385, -0.36979398131370544, 0.13916605710983276, 0.27015915513038635, -0.08963312208652496, 0.12558971345424652, 0.021427147090435028, -0.011060327291488647, 0.33645501732826233, 0.08014420419931412, 0.18637977540493011, -0.08931325376033783, 0.11018544435501099, -0.1883333921432495, -0.09345078468322754, -0.011489448137581348, -0.6761327981948853, 0.21053017675876617, -0.2046804130077362, 0.3002508282661438, -0.014470888301730156, 0.03403792157769203, -0.6020739078521729, 0.018952032551169395, -0.08776948601007462, -0.24328812956809998, 0.04227837920188904, 0.1985483169555664, 0.13480249047279358, 0.05542037636041641, 0.005591148976236582, 0.13680925965309143, -0.15025204420089722, 0.25498682260513306, -0.07851354777812958, -0.028988609090447426, -0.1147482842206955, 0.09138856828212738, 0.30210477113723755, 0.28191590309143066, 0.0884697437286377, -0.3099205195903778, 0.05243707448244095, 0.28284212946891785, 0.23467174172401428, 0.14354699850082397, 0.2016162872314453, -0.11179038882255554, 0.08561423420906067, 0.22559989988803864, 0.26149848103523254, 0.3159290850162506, 0.10255779325962067, -0.21182097494602203, -0.001660871901549399, 0.24940195679664612, 0.24553607404232025, -0.004479784052819014, -0.2242429554462433, 0.058573056012392044, 0.2959596812725067, 0.10568228363990784, -0.10605474561452866, -0.31116706132888794, 0.09343535453081131, 0.3330497145652771, 0.5936193466186523, -0.09914620965719223, 0.19831407070159912, -0.03524594381451607, -0.26504430174827576, -0.32063809037208557, 0.11222020536661148, 0.11663506180047989, -0.283305287361145, -0.1018625795841217, 0.0325123593211174, 0.27163228392601013, 0.8015955090522766, -0.055096495896577835, 0.059599000960588455, 0.11370692402124405, -0.15358583629131317, -0.19464080035686493, 0.19420088827610016, 0.10261137038469315, -0.07959944754838943, 0.11133997142314911, -0.04581402614712715, -0.34773877263069153, -0.03733551502227783, -0.009878520853817463, 0.41767418384552, 0.4067237377166748, -0.40734046697616577, -0.05698336660861969, -0.15354451537132263, -0.3752639591693878, -0.38578373193740845, -0.3960627317428589, -0.2093992829322815, -0.37914860248565674, -0.06100843846797943, 0.34465765953063965, 0.018216244876384735, 0.25030508637428284, -0.09401354193687439, 0.2637220621109009, -0.2055864930152893, -0.3096761405467987, 0.16009965538978577, -0.13870471715927124, -0.4587656557559967, 0.030165407806634903, 0.2429065704345703, 0.16651375591754913, 0.36580023169517517, -0.26891201734542847, -0.21224737167358398, -0.20119889080524445, -0.3196375072002411, 0.14715661108493805, -0.01063154824078083, 0.6500741839408875, 0.4507485330104828, 0.05853559076786041, 0.2979262173175812, 0.11121866106987, 0.003967353142797947, -0.13633903861045837, -0.3817053735256195, -0.08096244186162949, -0.02113957330584526, 0.06345321238040924, -0.11353873461484909, -0.43938931822776794, -0.5979090332984924, -0.028065398335456848, 0.07417379319667816, 0.11462879925966263, 0.12124437093734741, 0.13008536398410797, -0.015364635735750198, 0.18101924657821655, 0.031908899545669556, -0.19629231095314026, -0.19380241632461548, 0.053100209683179855, 0.08232906460762024, -0.17546892166137695, -0.28174203634262085, 0.16572806239128113, -0.21765847504138947, 0.08913887292146683, -0.26760950684547424, -0.463716596364975, -0.4648182988166809, 0.24480947852134705, 0.16363246738910675, 0.30453333258628845, -0.2045295685529709, 0.29216691851615906, 0.25186437368392944, 0.17327281832695007, -0.12148870527744293, -0.35631003975868225, 0.20366652309894562, 0.01565190963447094, 0.3223893642425537, 0.19924822449684143, 0.3970184326171875, -0.2766508460044861, 0.7669867873191833, 0.18207107484340668, -0.08122340589761734, -0.03907801955938339, -0.21187752485275269, 0.09913458675146103, -0.08708599954843521, -0.057647719979286194, 0.23883278667926788, -0.1382332742214203, -0.20207801461219788, 0.31310534477233887, -0.17024579644203186, -0.009244069457054138, -0.12770071625709534, 0.2386336624622345, -0.1114010363817215, -0.06311237066984177, 0.05958012863993645, -0.16654349863529205, 0.03780542314052582, 0.09868082404136658, -0.014956257306039333, -0.2617679238319397, -0.1944642961025238, -0.052546631544828415, 0.33535370230674744, 0.002153830137103796, 0.11820787936449051, -0.31880077719688416, 0.011236653663218021, -0.426546186208725, 0.2786056101322174, 0.23024237155914307, 0.5973358154296875, 0.12704841792583466, 0.08186546713113785, 0.10294009000062943, 0.10305611044168472, 0.21146512031555176, -0.007445289753377438, 0.09969271719455719, 0.2119990438222885, -0.09975704550743103, -0.2154729962348938, -0.24166728556156158, -0.0882355347275734, 0.15256533026695251, -0.04120654985308647, 0.12435609102249146, -0.4224526882171631, -0.06276466697454453, 0.2315380573272705, 0.1961362510919571, -0.18541821837425232, -0.3014693260192871, -0.19508326053619385, -0.01329123042523861, -0.10634052008390427, -0.22760486602783203, -0.23696132004261017, 0.07546564191579819, -0.19839823246002197, 0.4346012473106384, 0.026548141613602638, -0.09503377974033356, -0.009467562660574913, 0.2204018235206604, 0.2185642421245575, -0.22556573152542114, 0.34710660576820374, 0.21265456080436707, 0.04422237351536751, 0.19378289580345154, 0.2733420729637146, 0.04669993743300438, -0.26991793513298035, 0.07171867787837982, -0.05596558749675751, 0.3882954716682434, 0.23093856871128082, 0.07861834764480591, -0.000054110627388581634, -0.22865737974643707, -0.1555214375257492, -0.2668573260307312, 0.11353752762079239, 0.3650883138179779, -0.012089059688150883, -0.057567041367292404, -0.3237791657447815, 0.5772802829742432, 0.12792930006980896, -0.22235463559627533, 0.3620975613594055, -0.22821207344532013, -0.23417487740516663, 0.17707756161689758, 0.058152470737695694, 0.8921769261360168, 0.16323097050189972, 0.26067519187927246, 0.4901677072048187, -0.2221510112285614, 0.3511625826358795, 0.21474815905094147, 0.11311934143304825, -0.6522518396377563, -0.0027110478840768337, 0.03967943787574768, -0.14634227752685547, 0.2141655832529068, -0.14346681535243988, -0.33664727210998535, 0.3871508240699768, 0.004389516077935696, 0.2776857614517212, 0.28608790040016174, 0.3106580674648285, -0.1787114441394806, -0.02707960084080696, -0.1167604997754097, 0.0829826146364212, -0.0814642384648323, 0.3729133605957031, -0.23302620649337769, -0.07518498599529266, -0.26928508281707764, -0.062462545931339264, -0.19180411100387573, -0.13464783132076263, -0.16934460401535034, -0.052010297775268555, 0.1199265718460083, -0.1107712984085083, -0.04196763411164284, 0.2312607318162918, 0.2426557093858719, -0.1447327733039856, -0.22735722362995148, 0.05623316392302513, -0.26526325941085815, 0.07546386867761612, 0.3290456235408783, 0.061140693724155426, 0.25031226873397827, -0.27121174335479736, -0.17359347641468048, 0.10917002707719803, -0.24852511286735535, -0.06863456219434738, -0.149458646774292, -0.20029288530349731, 0.16526983678340912, -0.30027440190315247, -0.34574997425079346, -0.11714136600494385, -0.04909713566303253, -0.09219103306531906, 0.02859162911772728, 0.22841808199882507, -0.3406803011894226, -0.005799603182822466, -0.11406496167182922, -0.26731911301612854, -0.028432782739400864, 0.45415565371513367, -0.02947232313454151, -0.013764605857431889, 0.8067978620529175, -0.09696390479803085, 0.1509828269481659, -0.12489808350801468, 0.06405561417341232, -0.0021408319007605314, -0.20873384177684784, 0.0074975015595555305, -0.07276037335395813, -0.3054572641849518, 0.06333993375301361, 0.6002621650695801, 0.3502773642539978, -0.006405259016901255, -0.015813319012522697, -0.5503165125846863, -0.47054189443588257, -0.017437636852264404, -0.1337224394083023, 0.10882795602083206, -0.09000848978757858, 0.19828680157661438, 0.25997644662857056, 0.030203085392713547, -0.20928727090358734, -0.13878832757472992, -0.02385551668703556, 0.1949308067560196, -0.15603135526180267, 0.015625854954123497, 0.040912505239248276, 0.10926039516925812, -0.015660692006349564, 0.01699761673808098, -0.39186593890190125, -0.07138170301914215, -0.10062697529792786, 0.2022780478000641, 0.11084690690040588, -0.11428555846214294, 0.06330281496047974, -0.12285293638706207, -0.11143913865089417, 0.19947606325149536, -0.0942072719335556, 0.01717548444867134, -0.10651468485593796, 0.18531453609466553, 0.3287246823310852, 0.051338762044906616, -0.09047828614711761, 0.25120505690574646, 0.10262138396501541, 0.18970024585723877, -0.11579453200101852, 0.21868295967578888, -0.2577763497829437, 0.04440455511212349, 0.12394807487726212, 0.13644637167453766, -0.06430702656507492, 0.31939178705215454, 0.18316546082496643, -0.1417764574289322, 0.14575529098510742, -0.12332585453987122, 0.3411394953727722, 0.762729287147522, -0.2596036195755005, -0.16030213236808777, 0.43502870202064514, 0.042088307440280914, -0.23856453597545624, 0.0053892056457698345, 0.20815490186214447, 0.1026712954044342, 0.05342177674174309, 0.26197218894958496, 0.0547143891453743, -0.08130950480699539, 0.17981910705566406, 0.14861799776554108, 0.11086545884609222, -0.190028116106987, -0.08867544680833817, 0.5046725869178772, -0.2913787066936493, 0.23321571946144104, 0.3728407919406891, 0.04240838438272476, 0.16424502432346344, 0.16576345264911652, 0.052263617515563965, 0.47337597608566284, 0.05527505651116371, 0.051702745258808136, 0.11520887911319733, -0.02602476254105568, -0.004513641353696585, 0.39915093779563904, -0.009431731887161732, 0.008123252540826797, 0.311003178358078, 0.5946425199508667, -0.40880000591278076, -0.15509112179279327, -0.3245684504508972, -0.104848213493824, 0.035217978060245514, -0.1329294890165329, -0.3802723288536072, 0.07117965817451477, -0.2537851631641388, -0.15523599088191986, -0.15357694029808044, -0.3261699378490448, 0.11912330240011215, 0.0025212992914021015, -0.3347843587398529, -0.23041357100009918, -0.4620949625968933, 0.296138197183609, -0.10243222862482071, -0.2105662226676941, 0.24565617740154266, 0.06732213497161865, 0.13905878365039825, 0.3560231924057007, 0.3868749439716339, 0.41622281074523926, 0.1502307504415512, -0.4203753173351288, -0.17785894870758057, -0.15420076251029968, 0.10862664878368378, 0.0784469023346901, 0.3482343554496765, -0.01889944262802601, -0.18846866488456726, 0.551433801651001, 0.06358038634061813, 0.041565172374248505, 0.2617349624633789, 0.43246617913246155, -0.4136829674243927, -0.19127339124679565, 0.7280192971229553, -0.1477660834789276, -0.12769801914691925, -0.11568349599838257, 0.1740320324897766, -0.33913251757621765, 0.1881859004497528, 0.055040717124938965, -0.011934816837310791, 0.2017202377319336, 0.00427025742828846, 0.03773863986134529, -0.17049357295036316, 0.42981836199760437, 0.3963537812232971, -0.030322326347231865, -0.4015502631664276, -0.24162371456623077, -0.826250433921814, 0.19436655938625336, -0.09572457522153854, -0.31783270835876465, -0.03783004358410835, -0.23111914098262787, 0.19117677211761475, 0.14636509120464325, -0.25047582387924194, 0.23906350135803223, -0.015760639682412148, 0.16460031270980835, -0.3513512909412384, -0.386923223733902, -0.2707550823688507, 0.1131700873374939, -0.08452508598566055, -0.4501943290233612, 0.17498204112052917, -0.24662648141384125, -0.09763592481613159, 0.16377043724060059, -0.019515525549650192, 0.4524601101875305, 0.10025583207607269, 0.3120701313018799, -0.04562406614422798, 0.5762509703636169, 0.17790578305721283, -0.040750328451395035, -0.232215017080307, 0.19575998187065125, -0.14506040513515472, 0.09771554172039032, 0.0491047278046608, 0.15912660956382751, -0.12989234924316406, 0.5070652365684509, -0.36369603872299194, 0.43635937571525574, 0.08669737726449966, 0.12016666680574417, -0.45000457763671875, 0.17375853657722473, -0.22239550948143005, 0.1384895145893097, 0.24804353713989258, 0.2696765661239624, -0.10547009110450745, 0.41415756940841675, -0.18960261344909668, -0.05892520397901535, 0.42898717522621155, -0.20631740987300873, -0.10361848771572113, 0.06802312284708023, 0.19691026210784912, -0.1153518557548523, -0.17309722304344177, -0.4120732843875885, 0.2724257707595825, 0.2450026571750641, -0.062201227992773056, 0.14686501026153564, 0.16532820463180542, 0.26175686717033386, 0.09281085431575775, -0.01229468360543251, 0.8619529008865356, -0.04305988550186157, 0.12368973344564438, -0.07380129396915436, -0.23250684142112732 ]
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?
[ -0.17673911154270172, 0.21117544174194336, -0.009480178356170654, -0.13179261982440948, 0.07344277948141098, -0.14275361597537994, 0.4176614582538605, 0.25736004114151, 0.049972958862781525, 0.11905588209629059, 0.22333267331123352, 0.0696350559592247, -0.2887464463710785, 0.08405549079179764, 0.09769392758607864, -0.007682157680392265, 0.06393381953239441, 0.14133629202842712, -0.10444147884845734, 0.20766907930374146, -0.2553173005580902, -0.005777726415544748, -0.05395077168941498, 0.1049625501036644, -0.021724600344896317, 0.12302418798208237, -0.20951852202415466, 0.10340232402086258, -0.43610069155693054, -0.8527146577835083, 0.38024279475212097, 0.10695000737905502, 0.288836807012558, 0.3949591815471649, -0.00012843437434639782, -0.057478029280900955, 0.5988144874572754, -0.10965356975793839, -0.6252241730690002, -0.14960186183452606, 0.278555691242218, -0.3912251591682434, 0.3590850532054901, -0.03522174432873726, 0.15007907152175903, -0.18746738135814667, 0.2187315821647644, 0.041837651282548904, 0.14435066282749176, 0.28827840089797974, 0.0764792338013649, 0.28133058547973633, 0.06819649040699005, -0.21769124269485474, -0.11556920409202576, 0.05953773856163025, -0.0826352909207344, 0.5680218935012817, 0.4420578181743622, 0.1645021140575409, -0.04790715500712395, -0.13474895060062408, 0.09925627708435059, 0.45589521527290344, 0.18909738957881927, -0.04519218951463699, 0.6421706080436707, 0.017217114567756653, -0.14175628125667572, -0.0005506889428943396, 0.9479652047157288, -0.5009497404098511, -0.3974156677722931, 0.050296034663915634, -0.011666894890367985, -0.12136594206094742, 0.23012861609458923, -0.19277577102184296, -0.29002588987350464, 0.04725101590156555, -0.17219503223896027, -0.3062463402748108, -0.1238466203212738, 0.30030012130737305, -0.3025883436203003, 0.30529963970184326, 0.10334384441375732, 0.24936053156852722, 0.01884394697844982, -0.11828628182411194, 0.2843591272830963, -0.036351125687360764, -0.014641765505075455, 0.18634478747844696, 0.07852458208799362, -0.2858280837535858, -0.34592512249946594, -0.20418091118335724, 0.11312146484851837, -0.04123446345329285, -0.42331263422966003, -0.1887901872396469, -0.034955840557813644, 0.1762140691280365, 0.22279851138591766, 0.2793654203414917, 0.3226895332336426, 0.1964389532804489, 0.2440313845872879, 0.07674857974052429, 0.3281914293766022, 0.2670007348060608, -0.07985193282365799, -0.4030311703681946, 0.01620848849415779, 0.05976123735308647, 0.21140821278095245, -0.35215795040130615, -0.08716574311256409, -0.26098179817199707, -0.3343002200126648, -0.04781255125999451, -0.13767479360103607, 0.2543802261352539, -0.11698365211486816, -0.09509852528572083, 0.030406631529331207, 0.4238699972629547, -0.15298622846603394, -0.15912993252277374, -0.01079873088747263, 0.014280279166996479, -0.16778883337974548, -0.018229397013783455, 0.21012499928474426, 0.08893667161464691, 0.31579455733299255, -0.2774042785167694, -0.2017739713191986, 0.003569229505956173, -0.015546680428087711, 0.09739143401384354, -0.19562110304832458, 0.2981414794921875, -0.24754206836223602, -0.021107975393533707, 0.043806903064250946, -0.2993892729282379, -0.20705439150333405, 0.2779181897640228, 0.09650807082653046, -0.4538351595401764, -0.017878489568829536, 0.04868262633681297, -0.5866075754165649, -0.0063497149385511875, -0.018574092537164688, -0.11635686457157135, 0.11945272237062454, -0.382311075925827, 0.013696291483938694, -0.44516515731811523, -0.37953710556030273, -0.10745594650506973, -0.15624317526817322, 0.22268806397914886, -0.3308179974555969, -0.2761635482311249, -0.3116798996925354, -0.3430010676383972, 0.15970611572265625, -0.24245966970920563, -0.09673310071229935, 0.09309789538383484, -0.35627588629722595, -0.21318872272968292, 0.5165300369262695, -0.6141241788864136, -0.31686392426490784, 0.47944414615631104, -0.3933298885822296, -0.3192737102508545, 0.24876071512699127, -0.020787721499800682, 0.09857188165187836, -0.43709471821784973, -0.12158027291297913, -0.01584029383957386, -0.018847446888685226, -0.16953273117542267, -0.21318617463111877, -0.34320876002311707, 0.15771731734275818, 0.2769617736339569, -0.06715579330921173, 0.14599771797657013, 0.16151903569698334, 0.1352691352367401, 0.3999654948711395, 0.19973228871822357, 0.1501392275094986, -0.17016705870628357, 0.24233393371105194, -0.2819077968597412, -0.03480294719338417, -0.06494942307472229, -0.6823078393936157, 0.20384760200977325, -0.08313970267772675, 0.16828672587871552, 0.01756279729306698, -0.010356965474784374, -0.5097718834877014, -0.004371626302599907, 0.0005457354127429426, -0.18655143678188324, -0.07908067852258682, 0.11807814240455627, 0.17097045481204987, -0.06011898070573807, -0.029049217700958252, 0.008818240836262703, -0.21767589449882507, 0.19490455090999603, 0.06572461873292923, 0.02183309942483902, -0.04672496020793915, 0.05511999875307083, 0.20559023320674896, 0.2732926309108734, 0.05788034945726395, -0.34605276584625244, -0.007602643221616745, 0.2738025188446045, 0.37539657950401306, 0.24678650498390198, 0.14574921131134033, 0.03405608981847763, 0.20076198875904083, 0.2431771159172058, 0.2701937258243561, 0.24104581773281097, 0.013896987773478031, -0.18850606679916382, 0.04307826980948448, 0.19351908564567566, 0.16598349809646606, -0.03977653756737709, -0.16236189007759094, -0.028106771409511566, 0.2738693356513977, 0.1649010181427002, -0.1608114242553711, -0.3247181475162506, 0.016357501968741417, 0.3211115002632141, 0.6261169910430908, 0.021360039710998535, 0.22093906998634338, -0.10340315848588943, -0.24081888794898987, -0.3683859705924988, -0.03585400804877281, 0.1324319988489151, -0.14667390286922455, -0.13123919069766998, 0.034572891891002655, 0.2909606099128723, 0.8139574527740479, -0.05824136361479759, 0.14619776606559753, 0.09321155399084091, -0.08680208772420883, -0.21096478402614594, 0.13557134568691254, 0.13352099061012268, 0.0800887942314148, 0.13520997762680054, 0.024450499564409256, -0.3316609263420105, -0.012322175316512585, 0.05064711719751358, 0.42544788122177124, 0.3760313391685486, -0.38500383496284485, -0.1450604498386383, -0.18982531130313873, -0.4163709878921509, -0.46484336256980896, -0.26002493500709534, -0.2401759922504425, -0.32249778509140015, -0.11344315856695175, 0.3259071111679077, 0.017790691927075386, 0.20503295958042145, -0.1692851483821869, 0.29844972491264343, -0.29495877027511597, -0.27069270610809326, 0.1276620626449585, -0.20652711391448975, -0.37653347849845886, -0.005978101398795843, 0.2203708291053772, -0.0590674914419651, 0.42291030287742615, -0.15566523373126984, -0.3003937900066376, -0.15991675853729248, -0.32177412509918213, 0.14323222637176514, 0.16493992507457733, 0.5009920597076416, 0.4258585274219513, -0.019074568524956703, 0.46502479910850525, 0.18072986602783203, 0.029476579278707504, -0.19941362738609314, -0.45201021432876587, -0.20943664014339447, -0.013931074179708958, 0.0920100286602974, -0.09755640476942062, -0.3627035319805145, -0.6268528699874878, -0.04943504557013512, 0.0924636721611023, 0.046491798013448715, 0.08684590458869934, 0.07008551806211472, -0.17430385947227478, 0.12629644572734833, 0.0648910254240036, -0.14631466567516327, -0.13924884796142578, 0.1288541853427887, -0.06379901617765427, -0.11443964391946793, -0.2529374361038208, 0.0870978981256485, -0.09722974896430969, -0.02210492268204689, -0.0692761167883873, -0.41635215282440186, -0.49807921051979065, 0.18607720732688904, 0.12020499259233475, 0.3934658169746399, -0.30545684695243835, 0.3518293499946594, 0.22900041937828064, 0.19614936411380768, -0.07248470932245255, -0.34929344058036804, 0.2188248187303543, 0.11768502742052078, 0.2108534425497055, 0.31401345133781433, 0.3968403935432434, -0.17006205022335052, 0.6948562264442444, 0.13153190910816193, -0.17439696192741394, -0.08181396871805191, -0.1846349686384201, 0.02651890739798546, -0.0716574639081955, 0.02526930347084999, 0.28924667835235596, -0.12184963375329971, -0.3842003047466278, 0.2331882119178772, -0.17590048909187317, -0.04469316452741623, -0.18192723393440247, 0.30313512682914734, 0.04550117254257202, -0.023568296805024147, -0.017659317702054977, -0.18742913007736206, 0.062108781188726425, 0.0636761412024498, -0.03822570666670799, -0.1691713184118271, -0.19289416074752808, 0.0061613814905285835, 0.3494342863559723, 0.0028034807182848454, 0.20179879665374756, -0.2370384782552719, 0.0679619312286377, -0.32495540380477905, 0.32468414306640625, 0.3511452376842499, 0.6161119341850281, 0.12353891879320145, 0.006404577754437923, 0.12947894632816315, 0.04525177553296089, 0.16851778328418732, 0.01586620695888996, 0.20254448056221008, 0.19419027864933014, -0.15339328348636627, -0.1033148318529129, -0.2794736623764038, -0.05562492832541466, 0.10736342519521713, -0.014489459805190563, 0.20647266507148743, -0.3730700612068176, -0.16325263679027557, 0.2161453366279602, 0.13574638962745667, -0.10049261152744293, -0.33303189277648926, -0.08526643365621567, -0.01718946546316147, -0.032943159341812134, -0.2879052758216858, -0.2201477587223053, -0.003518399316817522, -0.17048852145671844, 0.3609248995780945, -0.009871484711766243, -0.15480662882328033, -0.04507419839501381, 0.12206512689590454, 0.14888499677181244, -0.28252148628234863, 0.30649593472480774, 0.2818535566329956, 0.10912101715803146, 0.42920413613319397, 0.2410028725862503, -0.07917188107967377, -0.20654112100601196, 0.062447912991046906, -0.008029818534851074, 0.3869709074497223, 0.36695730686187744, 0.101006880402565, -0.04046180844306946, -0.191474050283432, -0.1860707551240921, -0.3599006235599518, 0.10493282973766327, 0.31456083059310913, -0.07839667052030563, -0.08095976710319519, -0.3749857544898987, 0.647111713886261, 0.18893463909626007, -0.28201550245285034, 0.4349672496318817, -0.29321226477622986, -0.24159428477287292, 0.016749808564782143, 0.11472064256668091, 0.9402284622192383, 0.1011531800031662, 0.17937569320201874, 0.4074013829231262, -0.25691235065460205, 0.27772751450538635, 0.20907960832118988, 0.09308316558599472, -0.5795203447341919, 0.036909300833940506, 0.006617447827011347, -0.18128737807273865, 0.2285756766796112, -0.10842627286911011, -0.1412527859210968, 0.3231202960014343, 0.06759390234947205, 0.34467002749443054, 0.2996441125869751, 0.32671666145324707, -0.13545860350131989, 0.03873373568058014, -0.17634396255016327, 0.014816293492913246, -0.16323575377464294, 0.3315982520580292, -0.24513894319534302, 0.020707568153738976, -0.20245018601417542, -0.06340140849351883, -0.2631102204322815, -0.20407195389270782, -0.14579348266124725, 0.029360070824623108, 0.058056242763996124, -0.1723589450120926, -0.003578634699806571, 0.2685675621032715, 0.18531276285648346, -0.2835439145565033, -0.11798539012670517, 0.06572907418012619, -0.1478344202041626, 0.12417249381542206, 0.3541322350502014, 0.09129989147186279, 0.32435905933380127, -0.3461558520793915, -0.041704293340444565, 0.0784570649266243, -0.31472769379615784, -0.11034668236970901, -0.1682371348142624, -0.1697041392326355, 0.17790353298187256, -0.35208582878112793, -0.2943045496940613, -0.14130598306655884, 0.02850194461643696, -0.052879419177770615, -0.021204598248004913, 0.2687153220176697, -0.24451051652431488, -0.01123104989528656, -0.1383839249610901, -0.24387359619140625, 0.014475860632956028, 0.4472575783729553, 0.03498523309826851, -0.1098531037569046, 0.6268345713615417, -0.20007368922233582, 0.07830175757408142, -0.08096442371606827, 0.04233311116695404, -0.030404554679989815, -0.1384967416524887, 0.07945141941308975, -0.06528545916080475, -0.30925342440605164, 0.030999628826975822, 0.6317124962806702, 0.45381903648376465, -0.022200845181941986, -0.0184402447193861, -0.41377198696136475, -0.46844619512557983, 0.06022888049483299, -0.1337687373161316, 0.15529011189937592, -0.14459748566150665, 0.1882135570049286, 0.10624808073043823, -0.0461602583527565, -0.1483945995569229, -0.1558256596326828, 0.1758178472518921, 0.22977682948112488, -0.1876084804534912, 0.04707973450422287, -0.020073402673006058, 0.04233904555439949, -0.063949815928936, -0.004503414500504732, -0.3734772205352783, -0.00042919788393191993, -0.06305219978094101, 0.2435455471277237, 0.10599253326654434, -0.10039080679416656, 0.10395444929599762, -0.19135864078998566, -0.013094725087285042, 0.2529706656932831, -0.042604535818099976, 0.09640425443649292, -0.08035597205162048, 0.10699306428432465, 0.34468013048171997, 0.154820054769516, -0.008971737697720528, 0.3556567132472992, 0.08667532354593277, 0.3026207387447357, -0.07837759703397751, 0.3181910514831543, -0.25813028216362, 0.07906077802181244, 0.18990519642829895, 0.15776771306991577, -0.0170417670160532, 0.20191439986228943, 0.11414910107851028, -0.17145679891109467, 0.14508673548698425, -0.0436178520321846, 0.4042547941207886, 0.8169978857040405, -0.2331593632698059, -0.20942072570323944, 0.39549368619918823, -0.006465123035013676, -0.20162183046340942, 0.0467279888689518, 0.27429771423339844, 0.16639956831932068, -0.024561749771237373, 0.3101126551628113, 0.09837249666452408, -0.0165251437574625, 0.13899216055870056, 0.2003037929534912, 0.17520666122436523, -0.2545609474182129, -0.014315994456410408, 0.436180979013443, -0.19119292497634888, 0.23482514917850494, 0.39259886741638184, 0.10050836950540543, 0.19480371475219727, 0.09433556348085403, 0.0811319574713707, 0.5245577096939087, 0.12521147727966309, 0.10351601243019104, 0.17616017162799835, -0.013916284777224064, 0.026047736406326294, 0.401174396276474, -0.0035630997736006975, -0.04608476534485817, 0.40405771136283875, 0.6079491376876831, -0.38512495160102844, -0.20143236219882965, -0.29688215255737305, -0.1906670182943344, 0.051356710493564606, -0.13128790259361267, -0.25092726945877075, -0.021365327760577202, -0.1806849092245102, -0.06044841185212135, -0.24239729344844818, -0.2871663272380829, 0.231474831700325, -0.004475134890526533, -0.2735222578048706, -0.28109535574913025, -0.36880752444267273, 0.1331091821193695, -0.029771355912089348, -0.2417406141757965, 0.2890283763408661, -0.010914106853306293, 0.04701468348503113, 0.3937858045101166, 0.25173327326774597, 0.40285128355026245, 0.1396004855632782, -0.35749927163124084, -0.2865707278251648, -0.20618568360805511, 0.05763344466686249, 0.045157089829444885, 0.34859344363212585, -0.003062981180846691, -0.13039466738700867, 0.47411981225013733, 0.029031790792942047, 0.10622138530015945, 0.22748379409313202, 0.3550853133201599, -0.46599528193473816, -0.10022183507680893, 0.8348633646965027, -0.08329663425683975, -0.20123893022537231, 0.04727884382009506, 0.17994125187397003, -0.2012372761964798, 0.04935791343450546, -0.06658075004816055, 0.008687359280884266, 0.1635264903306961, 0.14323116838932037, -0.000718126364517957, -0.09857743233442307, 0.2886529862880707, 0.36612027883529663, 0.02545391581952572, -0.28435567021369934, -0.2156105935573578, -0.7466065287590027, 0.19427816569805145, -0.05663437023758888, -0.30991148948669434, -0.021590132266283035, -0.2513846755027771, 0.23229767382144928, 0.06197988614439964, -0.17731770873069763, 0.1805734932422638, 0.09646055102348328, 0.10784216970205307, -0.19800592958927155, -0.2759745717048645, -0.21555311977863312, 0.0227828286588192, -0.1038484126329422, -0.4249189496040344, 0.11890020966529846, -0.3571050465106964, -0.15924252569675446, 0.0888780876994133, 0.017661770805716515, 0.4647805392742157, 0.005589638836681843, 0.2521132826805115, -0.08387422561645508, 0.7441831231117249, 0.2783361077308655, -0.0673995167016983, -0.16776937246322632, 0.12654726207256317, -0.14369526505470276, 0.17402991652488708, -0.12162169069051743, 0.08444184809923172, -0.18292030692100525, 0.5333900451660156, -0.4123540222644806, 0.49777406454086304, -0.12797123193740845, 0.06901825964450836, -0.3403320014476776, 0.05988267436623573, -0.2637890577316284, 0.031390249729156494, 0.3083663880825043, 0.19540035724639893, -0.037802763283252716, 0.30247583985328674, -0.22645705938339233, -0.12371481209993362, 0.4578658938407898, -0.17120184004306793, -0.032308150082826614, 0.08505775779485703, 0.2551971971988678, 0.05695366859436035, -0.1403316706418991, -0.4825669527053833, 0.1895676702260971, 0.2474292814731598, -0.09605926275253296, 0.12113308906555176, 0.16393163800239563, 0.25603315234184265, 0.014003678224980831, -0.02363726869225502, 1.0229696035385132, -0.11149588227272034, 0.08293846249580383, -0.14923937618732452, -0.23327741026878357 ]
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 ?
[ -0.07840733975172043, 0.2176990807056427, -0.007994581945240498, -0.21871644258499146, -0.016750426962971687, -0.15013821423053741, 0.4548572599887848, 0.3005518615245819, 0.017230574041604996, 0.06836659461259842, 0.24650812149047852, 0.0794769898056984, -0.3443647623062134, 0.09023873507976532, 0.06199247017502785, 0.004056716803461313, 0.11398115754127502, 0.0871872752904892, -0.06385980546474457, 0.15320898592472076, -0.23857009410858154, -0.046687137335538864, 0.058539897203445435, 0.18519869446754456, 0.010334555990993977, 0.11389544606208801, -0.2475726306438446, 0.0872463807463646, -0.4137381315231323, -0.8389007449150085, 0.33150291442871094, 0.16535553336143494, 0.3825397491455078, 0.42129871249198914, -0.0001269610074814409, -0.09459535777568817, 0.6248884201049805, -0.020554399117827415, -0.6024013161659241, -0.16927850246429443, 0.3652232587337494, -0.3112028241157532, 0.41104578971862793, 0.024779697880148888, 0.1619839370250702, -0.20271039009094238, 0.1706407070159912, 0.16206811368465424, 0.1313151717185974, 0.13344086706638336, 0.06785380095243454, 0.22151784598827362, 0.052437469363212585, -0.14496329426765442, -0.19909518957138062, 0.05197041109204292, -0.06580492854118347, 0.6167842149734497, 0.5382672548294067, 0.17721574008464813, -0.021481065079569817, -0.0021341778337955475, 0.14147183299064636, 0.3280442953109741, 0.2071109563112259, -0.07337606698274612, 0.7348248362541199, 0.06136085093021393, -0.13098599016666412, 0.008156134746968746, 0.9527406096458435, -0.4113695025444031, -0.5182245373725891, 0.055669624358415604, 0.04883908852934837, -0.13101543486118317, 0.18289075791835785, -0.16959086060523987, -0.19738315045833588, 0.10677902400493622, -0.1812894195318222, -0.3202578127384186, -0.2214033305644989, 0.28122881054878235, -0.2812548577785492, 0.22530823945999146, 0.08337712287902832, 0.24231493473052979, -0.10925688594579697, -0.17242568731307983, 0.2602318227291107, -0.09966807812452316, -0.018706925213336945, 0.19757521152496338, 0.1821487694978714, -0.3591312766075134, -0.3890571594238281, -0.110869400203228, 0.1458977609872818, -0.10912870615720749, -0.3968172073364258, -0.2765679657459259, -0.051240794360637665, 0.09083912521600723, 0.25132495164871216, 0.2249232679605484, 0.41865605115890503, 0.11730164289474487, 0.23796424269676208, 0.0282039362937212, 0.385678768157959, 0.27554237842559814, -0.03827497735619545, -0.3110353648662567, 0.030388519167900085, 0.07220054417848587, 0.18404822051525116, -0.40609800815582275, 0.04705101251602173, -0.23230351507663727, -0.34174275398254395, -0.13881917297840118, -0.11908897757530212, 0.1863192468881607, -0.11593019217252731, 0.00033129012444987893, -0.00722667807713151, 0.37839213013648987, -0.1264200508594513, -0.18677878379821777, -0.029239289462566376, -0.057438790798187256, -0.13046741485595703, -0.020339027047157288, 0.19088761508464813, 0.03255537524819374, 0.28977516293525696, -0.2891445755958557, -0.18417224287986755, 0.02045205980539322, -0.08764704316854477, 0.12490960955619812, -0.2652384638786316, 0.2683083117008209, -0.40122634172439575, -0.015373065136373043, -0.07192295044660568, -0.36129674315452576, -0.15017589926719666, 0.28826919198036194, 0.07951928675174713, -0.4344099462032318, -0.015946345403790474, 0.052783116698265076, -0.485525906085968, -0.01224830374121666, -0.02927076630294323, -0.058224868029356, 0.1487240344285965, -0.3259883224964142, 0.043539609760046005, -0.4276154339313507, -0.3264017701148987, -0.12506118416786194, -0.20648092031478882, 0.2373216450214386, -0.3779641091823578, -0.24657811224460602, -0.21628721058368683, -0.3262248635292053, 0.051763828843832016, -0.20617173612117767, -0.09009300917387009, 0.03439401462674141, -0.14357252418994904, -0.24087366461753845, 0.5255959033966064, -0.6463608145713806, -0.2875416576862335, 0.5019088387489319, -0.27390560507774353, -0.2755509614944458, 0.25255343317985535, -0.02460145391523838, 0.15832319855690002, -0.5408315062522888, -0.15352264046669006, -0.10466581583023071, -0.03884044662117958, -0.1779830902814865, -0.3315534293651581, -0.42146801948547363, 0.20058292150497437, 0.33454370498657227, -0.14667847752571106, 0.10944339632987976, 0.09240774810314178, 0.0771399512887001, 0.5037009716033936, 0.18341265618801117, 0.05109666660428047, -0.21777045726776123, 0.3656325936317444, -0.2329491674900055, -0.07816821336746216, -0.014533604495227337, -0.7206131219863892, 0.1898140162229538, -0.10229065269231796, 0.09577054530382156, 0.13708701729774475, 0.005196834448724985, -0.4359968304634094, -0.023731974884867668, -0.013888614252209663, -0.2627042829990387, -0.081575408577919, -0.061651356518268585, 0.05206909403204918, -0.029485521838068962, -0.010522870346903801, 0.022949429228901863, -0.20240570604801178, 0.22217680513858795, 0.08579126000404358, -0.0024011279456317425, -0.09218493849039078, 0.04781798645853996, 0.15576091408729553, 0.29392772912979126, 0.07502374798059464, -0.30252036452293396, -0.02034229040145874, 0.3204719126224518, 0.3806018531322479, 0.24521303176879883, 0.11209177225828171, 0.10057686269283295, 0.2997724711894989, 0.3446670472621918, 0.3244912028312683, 0.14844390749931335, -0.043833568692207336, -0.09661417454481125, -0.0020461424719542265, 0.1990577131509781, 0.1992998570203781, -0.15083448588848114, -0.24920310080051422, -0.08458071947097778, 0.18606233596801758, 0.20408250391483307, -0.22944805026054382, -0.2984946072101593, -0.022397654131054878, 0.33672550320625305, 0.5842092633247375, 0.01033463142812252, 0.1890861988067627, -0.035991519689559937, -0.22549141943454742, -0.3813839256763458, -0.1191440001130104, 0.1591704934835434, -0.1531737595796585, -0.1768583506345749, 0.03739101067185402, 0.2746085822582245, 0.6558879613876343, -0.05080952122807503, 0.06169281527400017, 0.07968903332948685, -0.09127514064311981, -0.15820515155792236, 0.13297711312770844, 0.1951862871646881, 0.0797487422823906, 0.074482761323452, 0.02701414003968239, -0.22329115867614746, 0.05465954914689064, 0.07418333739042282, 0.296488881111145, 0.30123239755630493, -0.30336427688598633, -0.1087808609008789, -0.14369478821754456, -0.35501810908317566, -0.4484526813030243, -0.2716017961502075, -0.36121875047683716, -0.2839105725288391, -0.044409457594156265, 0.3055325746536255, -0.00804908201098442, 0.16161374747753143, -0.08846750855445862, 0.35138940811157227, -0.30625665187835693, -0.08021791279315948, 0.13091735541820526, -0.1687154769897461, -0.40213847160339355, -0.007545642554759979, 0.16932882368564606, -0.14140662550926208, 0.39903175830841064, -0.07358195632696152, -0.3104705512523651, -0.07312043011188507, -0.4179687201976776, 0.07629478722810745, 0.13683463633060455, 0.5680723786354065, 0.44378867745399475, -0.015857845544815063, 0.4559708833694458, 0.2141161561012268, 0.02653992734849453, -0.18789534270763397, -0.3747599422931671, -0.24569809436798096, 0.002451384672895074, 0.06221310421824455, -0.06288836151361465, -0.341156005859375, -0.6747945547103882, 0.015952667221426964, 0.14174172282218933, 0.04154043272137642, 0.09176699817180634, 0.09137053787708282, -0.20751479268074036, 0.009894351474940777, -0.033491432666778564, -0.12606921792030334, -0.09700166434049606, 0.1426057666540146, -0.11858964711427689, -0.1148613840341568, -0.21240973472595215, 0.09414661675691605, 0.050037942826747894, -0.04971381276845932, -0.15218950808048248, -0.3812304139137268, -0.5387279391288757, 0.27524417638778687, 0.09593963623046875, 0.44321733713150024, -0.3750835657119751, 0.31686413288116455, 0.2574607729911804, 0.21060171723365784, -0.030558010563254356, -0.3124098777770996, 0.22272084653377533, 0.2024611085653305, 0.23872724175453186, 0.34162360429763794, 0.4278687834739685, -0.17156077921390533, 0.6987606883049011, 0.26078978180885315, -0.2120242565870285, -0.23117181658744812, -0.005481533240526915, -0.12854988873004913, -0.09365098178386688, 0.03923323005437851, 0.2602946162223816, -0.11454684287309647, -0.36592674255371094, 0.28301024436950684, -0.07185600697994232, 0.000910446688067168, -0.17153973877429962, 0.23263606429100037, 0.0482664629817009, -0.025847911834716797, -0.01699170470237732, -0.1041966900229454, 0.09018270671367645, 0.03616925701498985, -0.06196339800953865, -0.16355231404304504, -0.24948865175247192, 0.07056637108325958, 0.45116743445396423, -0.00716620497405529, 0.2389383167028427, -0.14188571274280548, 0.14389030635356903, -0.3231486678123474, 0.29190754890441895, 0.3646891415119171, 0.5911163091659546, 0.20548394322395325, -0.06498752534389496, 0.16462978720664978, 0.0394924022257328, 0.16430601477622986, 0.06811739504337311, 0.22538651525974274, 0.08400459587574005, -0.23833215236663818, -0.08051207661628723, -0.32152679562568665, 0.001393120619468391, 0.04301424324512482, 0.06709840893745422, 0.20781594514846802, -0.3719227612018585, -0.22546426951885223, 0.1916746348142624, 0.11681441962718964, -0.14998026192188263, -0.32392826676368713, -0.045550305396318436, 0.1384337991476059, 0.0778772234916687, -0.3222934901714325, -0.3112131357192993, -0.0988616943359375, -0.23056119680404663, 0.3555450439453125, -0.003292020410299301, -0.15369579195976257, -0.06030827760696411, 0.17602160573005676, 0.12314106523990631, -0.35025742650032043, 0.3231487572193146, 0.30365246534347534, 0.19759118556976318, 0.41991448402404785, 0.11987707763910294, -0.12025188654661179, -0.10920175909996033, -0.03899136185646057, -0.09179151058197021, 0.3305353820323944, 0.4466265141963959, 0.029408618807792664, -0.07550712674856186, -0.15196453034877777, -0.21442191302776337, -0.36957892775535583, -0.030504893511533737, 0.2852321267127991, -0.0497964508831501, -0.06767596304416656, -0.3444429934024811, 0.6544525623321533, 0.1329582780599594, -0.3539952039718628, 0.3894944489002228, -0.2603277266025543, -0.20492447912693024, 0.01446898840367794, 0.1230357363820076, 0.9271782636642456, 0.1357526332139969, 0.2261447012424469, 0.35611435770988464, -0.2628800868988037, 0.1910613477230072, 0.20730815827846527, -0.021173957735300064, -0.5951223969459534, 0.07956129312515259, -0.010454600676894188, -0.11937642097473145, 0.2597513198852539, -0.06818687170743942, -0.06725700944662094, 0.3064366281032562, 0.11517395824193954, 0.3037261366844177, 0.27472928166389465, 0.3166326880455017, -0.07457040995359421, 0.12467680126428604, -0.1457645446062088, -0.0027337572537362576, -0.19871336221694946, 0.22155626118183136, -0.2173815220594406, -0.05912114307284355, -0.13456153869628906, -0.13533838093280792, -0.19719459116458893, -0.27885571122169495, -0.14172963798046112, 0.002587545895949006, 0.0993952825665474, -0.2422437071800232, 0.09887183457612991, 0.20783916115760803, 0.21243159472942352, -0.2769320011138916, -0.10583128035068512, -0.006890191230922937, -0.09415440261363983, 0.19183051586151123, 0.23260459303855896, 0.1145530641078949, 0.4135673940181732, -0.36565208435058594, -0.02983371540904045, -0.002150309504941106, -0.30152428150177, -0.07882047444581985, -0.24475951492786407, -0.2292691171169281, 0.20691950619220734, -0.3499620854854584, -0.2433503121137619, -0.07171531021595001, 0.060952525585889816, -0.04963540658354759, -0.016589459031820297, 0.2987673878669739, -0.16764627397060394, 0.0024614299181848764, -0.06528221815824509, -0.14836636185646057, 0.0035220470745116472, 0.4236816465854645, 0.035341836512088776, -0.12862008810043335, 0.545073390007019, -0.1776934564113617, -0.012218748219311237, -0.06524506956338882, 0.042220331728458405, -0.05693398788571358, -0.26405829191207886, 0.024849427863955498, -0.08269258588552475, -0.2541992962360382, 0.026647474616765976, 0.6445194482803345, 0.5140349864959717, -0.08816099166870117, 0.008511189371347427, -0.31155475974082947, -0.43375876545906067, 0.11887901276350021, -0.14966146647930145, 0.12794055044651031, -0.20163486897945404, 0.19356249272823334, 0.089628204703331, -0.08079355955123901, -0.16443048417568207, -0.1689247488975525, 0.28728950023651123, 0.16615435481071472, -0.20746596157550812, 0.05575072020292282, -0.05666794255375862, 0.00035758785088546574, -0.07731951773166656, -0.09144674241542816, -0.43206503987312317, -0.009923174045979977, -0.06818363070487976, 0.22385187447071075, 0.0970194935798645, -0.09937220066785812, 0.07501237094402313, -0.18205678462982178, 0.03296229615807533, 0.33374619483947754, -0.029270146042108536, 0.057746101170778275, 0.040878523141145706, 0.21472656726837158, 0.31241950392723083, 0.24340099096298218, -0.026759712025523186, 0.413670152425766, 0.05288239195942879, 0.27785634994506836, -0.03120395354926586, 0.3052886128425598, -0.23807896673679352, 0.037635914981365204, 0.07074975222349167, 0.192378968000412, -0.18816927075386047, 0.3230148255825043, 0.044290367513895035, -0.19937770068645477, 0.17659622430801392, 0.008234268985688686, 0.38792794942855835, 0.7847015261650085, -0.19644685089588165, -0.3009434640407562, 0.4536358118057251, 0.02780822291970253, -0.12452956289052963, 0.0831708088517189, 0.262359619140625, 0.1925949901342392, -0.04848919063806534, 0.3369877338409424, 0.14231346547603607, 0.03786316514015198, 0.10244564712047577, 0.2327052801847458, 0.11558367311954498, -0.30714139342308044, -0.014102927409112453, 0.485095351934433, -0.19111378490924835, 0.2585494816303253, 0.42612993717193604, 0.2196153998374939, 0.042595162987709045, 0.017360961064696312, 0.08998827636241913, 0.4031573534011841, 0.20004743337631226, 0.04172375798225403, 0.2126077562570572, 0.07285197079181671, -0.052007149904966354, 0.3586236238479614, 0.010091923177242279, -0.07322868704795837, 0.4328913688659668, 0.6263275146484375, -0.24576961994171143, -0.2735120952129364, -0.2652326226234436, -0.2573288679122925, 0.11760266125202179, -0.15073959529399872, -0.32535263895988464, -0.031385939568281174, -0.17232447862625122, -0.06642254441976547, -0.39282557368278503, -0.2323376089334488, 0.2464112639427185, 0.06092778965830803, -0.2492005079984665, -0.3521990478038788, -0.4377784729003906, 0.04229312390089035, 0.06692028045654297, -0.14867763221263885, 0.24843792617321014, 0.00396771403029561, 0.0939524695277214, 0.40526556968688965, 0.2469262033700943, 0.24045316874980927, 0.09698624163866043, -0.3979857563972473, -0.33229950070381165, -0.24257342517375946, 0.07602300494909286, 0.16924093663692474, 0.28834548592567444, 0.02029639668762684, -0.16908402740955353, 0.43928292393684387, 0.004951465874910355, 0.1724262237548828, 0.3671841323375702, 0.30420100688934326, -0.41586989164352417, -0.11310701817274094, 0.8669891953468323, -0.010363766923546791, -0.2329448163509369, 0.14805132150650024, 0.1992991715669632, -0.207626610994339, 0.12630239129066467, -0.04110443964600563, 0.05794181302189827, 0.12840141355991364, 0.15669365227222443, 0.012644628994166851, -0.03198810666799545, 0.1883249282836914, 0.3991010785102844, -0.01953895390033722, -0.23692449927330017, -0.12392809987068176, -0.5654715299606323, 0.07696294039487839, 0.0513409785926342, -0.2140647917985916, -0.018772630020976067, -0.2449738085269928, 0.20464442670345306, 0.06571204960346222, -0.11546415835618973, 0.02215912751853466, -0.050805553793907166, 0.14864717423915863, -0.17918595671653748, -0.19621124863624573, -0.13138937950134277, 0.11998564004898071, -0.11988760530948639, -0.30400800704956055, 0.05137846618890762, -0.3177837133407593, -0.19125224649906158, 0.018191443756222725, 0.047366250306367874, 0.4646492600440979, 0.038575343787670135, 0.13066112995147705, -0.06797140836715698, 0.5171794295310974, 0.3847469091415405, -0.17758962512016296, -0.1369655430316925, 0.10956775397062302, -0.1280331015586853, 0.18178777396678925, -0.13699191808700562, 0.15139365196228027, -0.12285551428794861, 0.42535218596458435, -0.33312827348709106, 0.418358713388443, -0.22756369411945343, 0.00016266130842268467, -0.4170818328857422, 0.10477396100759506, -0.11555349081754684, -0.11772508919239044, 0.2918640971183777, 0.27572396397590637, -0.0717189833521843, 0.26153188943862915, -0.2811012864112854, -0.0743652880191803, 0.4302864372730255, -0.2299855798482895, 0.047566380351781845, 0.09626500308513641, 0.29810604453086853, 0.11599428951740265, -0.05385642498731613, -0.5310579538345337, 0.12829171121120453, 0.2365073561668396, -0.14698557555675507, 0.022672118619084358, 0.22228986024856567, 0.311658650636673, -0.06905549019575119, -0.039132989943027496, 1.0794568061828613, -0.17351488769054413, 0.1126842051744461, -0.15644317865371704, -0.19749972224235535 ]
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
[ -0.14118066430091858, 0.28056493401527405, -0.010229108855128288, -0.17343832552433014, 0.09962044656276703, -0.05622611194849014, 0.3951389193534851, 0.3567787706851959, 0.13251592218875885, 0.16331014037132263, 0.25443193316459656, 0.11689899116754532, -0.26364338397979736, -0.032295871526002884, 0.08826706558465958, -0.022581255063414574, 0.024319814518094063, 0.1539965569972992, -0.06924246996641159, 0.25380879640579224, -0.1577264666557312, 0.027644138783216476, -0.07621072232723236, 0.0034686201252043247, -0.032067347317934036, 0.14956481754779816, -0.12435536086559296, -0.02923269383609295, -0.3976534605026245, -0.883568525314331, 0.38515710830688477, 0.08030342310667038, 0.1571994423866272, 0.45871952176094055, -0.0001239301636815071, -0.09218361973762512, 0.6603642702102661, -0.09545938670635223, -0.7126159071922302, -0.11028527468442917, 0.21409551799297333, -0.4431755244731903, 0.3148983120918274, -0.12270469963550568, 0.07477673888206482, -0.12571564316749573, 0.24232162535190582, 0.002949959132820368, 0.0423441119492054, 0.2779587209224701, 0.11370956152677536, 0.20985451340675354, 0.1447022259235382, -0.2050851583480835, -0.06326283514499664, 0.07268218696117401, -0.125519260764122, 0.5561835765838623, 0.29881545901298523, 0.0897962674498558, -0.011444438248872757, -0.08872735500335693, 0.08542568981647491, 0.4745652973651886, 0.20247608423233032, -0.025697477161884308, 0.5517066717147827, 0.010823695920407772, -0.16354906558990479, -0.002150510670617223, 0.9606096744537354, -0.5900261402130127, -0.28566619753837585, 0.05598325654864311, 0.061815645545721054, -0.044158224016427994, 0.22932514548301697, -0.19335760176181793, -0.37108850479125977, 0.005593345034867525, -0.11386454850435257, -0.3460592031478882, -0.13907594978809357, 0.38633793592453003, -0.2816321849822998, 0.3920342028141022, 0.12043488770723343, 0.20180344581604004, 0.029666446149349213, -0.1227109283208847, 0.1766301393508911, -0.037354111671447754, 0.05294116586446762, 0.2122684270143509, 0.05316194146871567, -0.29137104749679565, -0.3612183928489685, -0.23289959132671356, 0.1176925078034401, -0.0717005580663681, -0.5114049911499023, -0.2140529900789261, 0.060136161744594574, 0.21360693871974945, 0.11943667382001877, 0.2886669337749481, 0.3164392113685608, 0.26326683163642883, 0.26908665895462036, 0.1591002196073532, 0.23070327937602997, 0.22681626677513123, -0.0012361377011984587, -0.39059728384017944, 0.007878215052187443, 0.04351721704006195, 0.2372445911169052, -0.27922511100769043, -0.0018135776044800878, -0.22336560487747192, -0.35482582449913025, -0.037689756602048874, -0.1549629122018814, 0.27120161056518555, -0.2122439295053482, -0.21905986964702606, 0.035491686314344406, 0.3669431507587433, -0.19033756852149963, -0.08297476917505264, -0.030062654986977577, 0.10269169509410858, -0.2094659060239792, 0.04134001582860947, 0.27360138297080994, 0.1198733001947403, 0.38156476616859436, -0.22116583585739136, -0.17527370154857635, -0.06266413629055023, 0.04889081418514252, 0.1650390923023224, -0.2584548890590668, 0.31381168961524963, -0.17547889053821564, 0.09148246049880981, 0.07570208609104156, -0.398679256439209, -0.2606751620769501, 0.19058091938495636, 0.08797695487737656, -0.3984541893005371, -0.07537229359149933, 0.0986795574426651, -0.5322718620300293, -0.047282926738262177, 0.03677930682897568, -0.10876619070768356, 0.1694987565279007, -0.4394988715648651, 0.03998028114438057, -0.3296203017234802, -0.327553927898407, -0.1206929013133049, -0.04477451369166374, 0.22451503574848175, -0.33446282148361206, -0.3448534905910492, -0.39960041642189026, -0.2620833218097687, 0.19108642637729645, -0.27225595712661743, -0.05161505937576294, 0.1798933446407318, -0.34467560052871704, -0.09361156821250916, 0.5451645851135254, -0.5609612464904785, -0.36663374304771423, 0.4103248119354248, -0.3971414864063263, -0.3093593716621399, 0.21600046753883362, -0.0019479896873235703, 0.07213959097862244, -0.29543232917785645, -0.13503095507621765, 0.02669873647391796, 0.05479004979133606, -0.12957759201526642, -0.17708131670951843, -0.3129168748855591, 0.15713398158550262, 0.24047550559043884, -0.05776771903038025, 0.1276276856660843, 0.1767682582139969, 0.22123204171657562, 0.3395687937736511, 0.12165817618370056, 0.19460849463939667, -0.08480003476142883, 0.20399193465709686, -0.17558155953884125, -0.07350818812847137, -0.0987354964017868, -0.6286861300468445, 0.21739508211612701, -0.04047953709959984, 0.2884497344493866, -0.07102853804826736, -0.05127634108066559, -0.6030596494674683, -0.028186215087771416, 0.0071443310007452965, -0.25438499450683594, 0.0009690880542621017, 0.21090400218963623, 0.1257525086402893, 0.07182269543409348, 0.08742771297693253, -0.05625343695282936, -0.20948579907417297, 0.1899573653936386, -0.03993655741214752, -0.01607614941895008, -0.11674633622169495, 0.11367618292570114, 0.12232311815023422, 0.20108260214328766, 0.03238670155405998, -0.2665899693965912, -0.011400003917515278, 0.1885949969291687, 0.3017379343509674, 0.26696479320526123, 0.20126529037952423, -0.028640784323215485, 0.18482019007205963, 0.1433064192533493, 0.23561520874500275, 0.31909412145614624, 0.07314863055944443, -0.2795489728450775, 0.000763388816267252, 0.13875515758991241, 0.21099039912223816, 0.0036713057197630405, -0.22517907619476318, -0.008010994642972946, 0.3341837525367737, 0.12936261296272278, -0.12944406270980835, -0.3318285644054413, 0.09353760629892349, 0.34653738141059875, 0.5930150151252747, -0.030656397342681885, 0.3187483549118042, -0.17704986035823822, -0.196113720536232, -0.33191585540771484, 0.01057212334126234, 0.14727658033370972, -0.14844293892383575, -0.13274461030960083, 0.004782307893037796, 0.26078271865844727, 0.7473019361495972, -0.017134331166744232, 0.21024365723133087, 0.08281276375055313, -0.1943524032831192, -0.1811455935239792, 0.1961624175310135, 0.14606782793998718, 0.058496639132499695, 0.11732533574104309, 0.008353599347174168, -0.3501940667629242, -0.04217192903161049, -0.010637049563229084, 0.461648553609848, 0.4212026000022888, -0.3261986970901489, -0.12541860342025757, -0.20866438746452332, -0.41110870242118835, -0.450446754693985, -0.40198686718940735, -0.3150009512901306, -0.35423216223716736, -0.10665455460548401, 0.270224004983902, 0.08679602295160294, 0.28298383951187134, -0.22471579909324646, 0.30379197001457214, -0.25691157579421997, -0.3586897850036621, 0.0704030767083168, -0.2922022342681885, -0.3351595103740692, 0.011352064087986946, 0.23351672291755676, 0.0032300183083862066, 0.45929455757141113, -0.1246679499745369, -0.3249702751636505, -0.2268066257238388, -0.36631810665130615, 0.10803461074829102, 0.06434576213359833, 0.5244491696357727, 0.41656580567359924, 0.040302008390426636, 0.4771682620048523, 0.04909313842654228, 0.03497670590877533, -0.18719090521335602, -0.4240945875644684, -0.23610304296016693, -0.01263449341058731, 0.17698727548122406, -0.09271983057260513, -0.37856367230415344, -0.5373613238334656, -0.10313158482313156, 0.08804342150688171, 0.0672018751502037, 0.006694730371236801, 0.15287841856479645, -0.10761261731386185, 0.1514066755771637, 0.10712627321481705, -0.05570575222373009, -0.133965864777565, 0.1255894899368286, 0.020721806213259697, -0.16886918246746063, -0.1097414419054985, 0.10617504268884659, -0.11159298568964005, 0.03736865147948265, -0.13643255829811096, -0.45776939392089844, -0.4050000011920929, 0.1082661971449852, 0.1651182472705841, 0.4095953404903412, -0.252712219953537, 0.2943410277366638, 0.2326204776763916, 0.14786070585250854, -0.09710408747196198, -0.3855842053890228, 0.18258561193943024, 0.03584698587656021, 0.2548632025718689, 0.23821207880973816, 0.4276253581047058, -0.18805338442325592, 0.8638370037078857, 0.1257786750793457, -0.09333620220422745, -0.0604630745947361, -0.22055989503860474, 0.0037175603210926056, -0.059210192412137985, -0.04446578770875931, 0.3229115307331085, -0.11662008613348007, -0.32144486904144287, 0.2759305536746979, -0.20463065803050995, -0.1540335714817047, -0.1347918063402176, 0.2869543731212616, -0.006108600180596113, -0.06751160323619843, -0.044665396213531494, -0.14884786307811737, 0.04530717059969902, 0.0540638267993927, -0.0495029017329216, -0.2289304882287979, -0.25206321477890015, 0.04696149006485939, 0.31666627526283264, -0.1577635556459427, 0.18858616054058075, -0.2466302067041397, 0.03164246305823326, -0.4474696218967438, 0.2941853106021881, 0.3331857919692993, 0.5381500720977783, 0.10278936475515366, -0.015300100669264793, 0.11338116973638535, 0.10491451621055603, 0.20852701365947723, -0.01775924116373062, 0.13445258140563965, 0.2412898689508438, -0.09606357663869858, -0.12461602687835693, -0.2597489655017853, -0.008720080368220806, 0.18970325589179993, -0.004294935613870621, 0.13426728546619415, -0.4184403419494629, -0.07045486569404602, 0.26349595189094543, 0.19597162306308746, -0.09762411564588547, -0.32441073656082153, -0.08676259219646454, -0.02829734794795513, -0.08013059198856354, -0.23945491015911102, -0.168500617146492, 0.10118233412504196, -0.07583745568990707, 0.3622828423976898, 0.03684867173433304, -0.06702408939599991, -0.08236654102802277, 0.15544459223747253, 0.14671307802200317, -0.20712395012378693, 0.26483258605003357, 0.22875158488750458, 0.027473557740449905, 0.3402649462223053, 0.24137155711650848, -0.11826849728822708, -0.2586960792541504, -0.04757877066731453, 0.00721639534458518, 0.3554089069366455, 0.23109307885169983, 0.18838483095169067, -0.01939685456454754, -0.16011203825473785, -0.14804206788539886, -0.34913772344589233, 0.19839338958263397, 0.3260376751422882, -0.09454560279846191, -0.081577368080616, -0.3744668960571289, 0.6699296832084656, 0.21066710352897644, -0.22166112065315247, 0.5757380723953247, -0.21588179469108582, -0.3613104820251465, -0.020962581038475037, 0.09729383885860443, 0.9390981793403625, 0.12304749339818954, 0.22184528410434723, 0.3271252512931824, -0.20360592007637024, 0.40682461857795715, 0.11397702991962433, 0.1631506383419037, -0.5522783994674683, 0.03558992221951485, 0.04169232398271561, -0.18077509105205536, 0.18681390583515167, -0.12864045798778534, -0.27663466334342957, 0.3458256423473358, 0.09094027429819107, 0.3162320554256439, 0.2804926335811615, 0.3736044466495514, -0.06282458454370499, -0.04738723114132881, -0.16777853667736053, 0.06752444803714752, -0.1731283813714981, 0.39323803782463074, -0.25016361474990845, 0.09912217408418655, -0.19187681376934052, -0.03845306858420372, -0.31208106875419617, -0.19377578794956207, -0.12246160954236984, 0.12016241997480392, 0.0544203482568264, -0.16867966949939728, -0.02505224011838436, 0.2270255833864212, 0.12529364228248596, -0.16949120163917542, -0.17709822952747345, 0.14197902381420135, -0.1813592165708542, 0.0641481801867485, 0.30314627289772034, 0.0513925664126873, 0.302105188369751, -0.347411572933197, -0.09088604152202606, 0.12196376919746399, -0.284130334854126, -0.20523597300052643, -0.12637683749198914, -0.1708374321460724, 0.08427341282367706, -0.32675784826278687, -0.28286051750183105, -0.13302044570446014, 0.04689616337418556, -0.14486835896968842, 0.026161782443523407, 0.22552159428596497, -0.31772467494010925, 0.0012359123211354017, -0.03404853492975235, -0.3448731303215027, 0.018304940313100815, 0.44678616523742676, -0.08538512885570526, -0.07896119356155396, 0.7237139940261841, -0.16906118392944336, 0.03817709535360336, -0.11864373832941055, 0.02304743602871895, -0.18729428946971893, -0.1536630243062973, 0.049601003527641296, -0.05159326642751694, -0.272928386926651, 0.010612351819872856, 0.5785099864006042, 0.4408789277076721, 0.04504666104912758, -0.02072545699775219, -0.45889103412628174, -0.4163408875465393, 0.007794680539518595, -0.18109843134880066, 0.2008616030216217, -0.0820298120379448, 0.189322829246521, 0.1570991426706314, 0.015954917296767235, -0.1874680072069168, -0.16336198151111603, 0.040294844657182693, 0.2436019480228424, -0.1337098479270935, 0.02252829074859619, -0.07007412612438202, 0.08745723217725754, 0.0018108010990545154, -0.08830499649047852, -0.2857421040534973, -0.05251790210604668, -0.1365344375371933, 0.22036156058311462, 0.06873489171266556, -0.0029469819273799658, 0.0809541717171669, -0.193817138671875, -0.005588103085756302, 0.16855359077453613, -0.05514583736658096, 0.10774040222167969, -0.09697446972131729, 0.19977805018424988, 0.334000825881958, 0.04802209138870239, -0.019415950402617455, 0.29507070779800415, 0.11993749439716339, 0.2520233988761902, -0.04795214906334877, 0.3378596901893616, -0.1911020129919052, 0.06257524341344833, 0.17886318266391754, 0.20285235345363617, -0.08617794513702393, 0.17287015914916992, 0.08450309187173843, -0.10913454741239548, 0.10387429594993591, 0.045503754168748856, 0.36438554525375366, 0.7686899900436401, -0.26971736550331116, -0.14736343920230865, 0.4336664378643036, 0.0431382916867733, -0.20703481137752533, -0.06151585280895233, 0.3559981882572174, 0.09675057232379913, 0.059565167874097824, 0.24292246997356415, -0.0034860826563090086, -0.019756434485316277, 0.15557125210762024, 0.16882438957691193, 0.08332826942205429, -0.22388125956058502, -0.014182246290147305, 0.3811241090297699, -0.2043299674987793, 0.16130268573760986, 0.3652932643890381, 0.12268368899822235, 0.25588366389274597, 0.11702356487512589, 0.12785881757736206, 0.49282413721084595, 0.06652938574552536, 0.03491088002920151, 0.07122964411973953, 0.07629617303609848, 0.06142061576247215, 0.26965659856796265, -0.09182754904031754, -0.021174924448132515, 0.3710108697414398, 0.5512712001800537, -0.3462454676628113, -0.131628155708313, -0.2795214056968689, -0.14258615672588348, 0.04645569995045662, -0.1809135228395462, -0.24025413393974304, 0.013755792751908302, -0.1374763399362564, -0.11026796698570251, -0.15492673218250275, -0.35871633887290955, 0.2438008040189743, 0.00835169106721878, -0.3695175349712372, -0.30901584029197693, -0.21209479868412018, 0.17734560370445251, -0.0065534464083611965, -0.31453657150268555, 0.18164996802806854, -0.060157738626003265, 0.009301644749939442, 0.2860337495803833, 0.2964818775653839, 0.4562511146068573, 0.08334755897521973, -0.4143468141555786, -0.25419455766677856, -0.18059150874614716, 0.06383511424064636, -0.027550609782338142, 0.34688517451286316, 0.03700839728116989, -0.19906127452850342, 0.49579212069511414, 0.06062591075897217, 0.035857148468494415, 0.13404998183250427, 0.44267943501472473, -0.5077687501907349, -0.1259593963623047, 0.6570249795913696, -0.06691065430641174, -0.17114907503128052, -0.014400038868188858, 0.2682858109474182, -0.3131226599216461, 0.09172582626342773, -0.0007792114629410207, -0.04264109209179878, 0.14234837889671326, 0.07496245950460434, 0.025010135024785995, -0.17087219655513763, 0.40392717719078064, 0.3193940222263336, 0.019597329199314117, -0.2624475657939911, -0.25915491580963135, -0.7384797930717468, 0.19107277691364288, -0.11629625409841537, -0.34262755513191223, -0.07228410989046097, -0.19995799660682678, 0.24219225347042084, 0.04681173712015152, -0.1436450034379959, 0.10192439705133438, 0.0297587551176548, 0.14397841691970825, -0.1850316971540451, -0.3121947944164276, -0.200699120759964, -0.022073809057474136, -0.06889423727989197, -0.40647271275520325, 0.11685915291309357, -0.3669162392616272, -0.09841712564229965, 0.0744376927614212, 0.002644643886014819, 0.532345175743103, 0.07270567864179611, 0.29426658153533936, -0.1021232008934021, 0.8140294551849365, 0.17402304708957672, -0.04890269413590431, -0.17330341041088104, 0.11475573480129242, -0.10631567239761353, 0.16286814212799072, -0.127364844083786, 0.1119912788271904, -0.20888230204582214, 0.4868564009666443, -0.3890765309333801, 0.6018154621124268, -0.017861654981970787, 0.08193845301866531, -0.2825641930103302, 0.011469092220067978, -0.2549222409725189, 0.08918764442205429, 0.29561367630958557, 0.1952594816684723, -0.07665416598320007, 0.3548060953617096, -0.1530630886554718, -0.25798770785331726, 0.417568564414978, -0.16063256561756134, -0.11916570365428925, 0.031318001449108124, 0.22145995497703552, -0.012549388222396374, -0.09062878042459488, -0.43135038018226624, 0.19755998253822327, 0.3064979314804077, -0.06527061015367508, 0.035290442407131195, 0.19647514820098877, 0.22549037635326385, 0.0979534164071083, -0.005873504094779491, 0.9912327527999878, -0.10930541157722473, 0.11685292422771454, -0.16476839780807495, -0.2565416395664215 ]
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.
[ -0.08672048896551132, 0.27374622225761414, 0.011248765513300896, -0.14076030254364014, 0.06271112710237503, -0.048166874796152115, 0.3938750624656677, 0.3804904818534851, 0.08933456242084503, 0.2072926014661789, 0.2584838271141052, 0.1472856104373932, -0.23994091153144836, -0.015442528761923313, 0.04751153662800789, -0.0067649041302502155, -0.05851535499095917, 0.2121698409318924, -0.02903229370713234, 0.2738306522369385, -0.08489841967821121, -0.011124231852591038, 0.07523142546415329, 0.09017889946699142, -0.044721536338329315, 0.11217693984508514, -0.12271828949451447, -0.06276879459619522, -0.263785183429718, -0.9248582124710083, 0.34500616788864136, 0.124372698366642, 0.1424262672662735, 0.5053982138633728, -0.00012234803580213338, -0.05318588763475418, 0.6260417699813843, -0.10737151652574539, -0.69929039478302, -0.1502019464969635, 0.23082512617111206, -0.46198564767837524, 0.2858988046646118, -0.18482643365859985, 0.034542266279459, -0.13203316926956177, 0.23104465007781982, 0.050337255001068115, 0.09132004529237747, 0.14613577723503113, 0.10933604836463928, 0.1425984650850296, 0.11538046598434448, -0.16234838962554932, -0.10149697959423065, 0.027807360514998436, -0.09675858169794083, 0.5679476261138916, 0.25440484285354614, 0.13260599970817566, 0.016347231343388557, -0.01072574034333229, 0.06912650913000107, 0.45414209365844727, 0.13088612258434296, -0.05509122088551521, 0.5264002084732056, 0.05460205301642418, -0.127457857131958, 0.13458846509456635, 0.9771099090576172, -0.594192385673523, -0.34010767936706543, 0.009946764446794987, 0.14042027294635773, -0.0033939231652766466, 0.24677728116512299, -0.17631296813488007, -0.3652084767818451, 0.00855911523103714, -0.09733173996210098, -0.2902492582798004, -0.19435934722423553, 0.3475719094276428, -0.2534480392932892, 0.4243468940258026, 0.13284410536289215, 0.21705615520477295, -0.050475072115659714, -0.23413123190402985, 0.14442595839500427, -0.047585681080818176, 0.06462270766496658, 0.12047569453716278, 0.06505173444747925, -0.31930020451545715, -0.4705292880535126, -0.32817307114601135, 0.1123797595500946, -0.17740857601165771, -0.40035656094551086, -0.24110467731952667, 0.0657077506184578, 0.20268885791301727, 0.1215643510222435, 0.1992536336183548, 0.38518956303596497, 0.22244979441165924, 0.2539425492286682, 0.14471566677093506, 0.1746019572019577, 0.2078220248222351, 0.023900305852293968, -0.4101088047027588, 0.002597258659079671, 0.03073343262076378, 0.28490954637527466, -0.2973403334617615, 0.017244908958673477, -0.12513384222984314, -0.36313799023628235, -0.02633252926170826, -0.21659749746322632, 0.358425498008728, -0.25124818086624146, -0.22439974546432495, 0.08188824355602264, 0.3014506697654724, -0.10217811167240143, -0.07199306786060333, -0.023294862359762192, 0.02457551658153534, -0.12462344765663147, 0.11025375872850418, 0.24170352518558502, 0.08670324087142944, 0.4142285883426666, -0.2295181155204773, -0.15621817111968994, -0.12500232458114624, -0.003739673411473632, 0.211934432387352, -0.23602980375289917, 0.31187018752098083, -0.17207059264183044, 0.09038297086954117, 0.02592657506465912, -0.4385862648487091, -0.30044975876808167, 0.19515301287174225, 0.08729155361652374, -0.3440072536468506, -0.11515507102012634, 0.09680043160915375, -0.55954909324646, -0.08119650930166245, 0.16798873245716095, -0.12056028097867966, 0.13395193219184875, -0.4694448411464691, 0.03654409945011139, -0.2592836320400238, -0.30312806367874146, -0.15861913561820984, -0.0653032660484314, 0.27007022500038147, -0.31470462679862976, -0.3589428961277008, -0.33123233914375305, -0.20738761126995087, 0.0995786190032959, -0.3180847465991974, -0.03698370233178139, 0.1321748048067093, -0.23593920469284058, -0.07427401840686798, 0.5720570683479309, -0.5332985520362854, -0.3166048228740692, 0.4164280891418457, -0.3450375497341156, -0.2490309774875641, 0.2598499059677124, 0.02415752224624157, 0.1698985993862152, -0.2648543417453766, -0.12308835238218307, -0.050398681312799454, 0.12207385897636414, -0.10380496829748154, -0.19102099537849426, -0.41373616456985474, 0.0998164564371109, 0.27179476618766785, -0.10852338373661041, 0.10879956930875778, 0.12809593975543976, 0.19002825021743774, 0.4033347964286804, 0.08722389489412308, 0.12643881142139435, -0.08443894237279892, 0.19085732102394104, -0.09954720735549927, -0.15517720580101013, -0.037928685545921326, -0.6501653790473938, 0.28278523683547974, -0.02253638207912445, 0.22074216604232788, 0.02934495359659195, -0.11142092943191528, -0.5947407484054565, -0.07118990272283554, -0.014745841734111309, -0.22981144487857819, 0.038247477263212204, 0.2451982945203781, 0.04038107395172119, 0.028742438182234764, 0.1323024183511734, -0.06919823586940765, -0.17278465628623962, 0.15835611522197723, -0.05750095471739769, -0.03594747930765152, -0.10883574187755585, 0.10054495185613632, 0.14131931960582733, 0.1920613795518875, 0.08820649981498718, -0.20974355936050415, -0.008258403278887272, 0.19893237948417664, 0.26645544171333313, 0.2994747459888458, 0.17117832601070404, -0.03975282609462738, 0.11774793267250061, 0.16865266859531403, 0.22219640016555786, 0.2807433307170868, 0.0006242480594664812, -0.2647783160209656, 0.012452122755348682, 0.18017512559890747, 0.2809547781944275, -0.0212848037481308, -0.3174169361591339, -0.028278708457946777, 0.319345086812973, 0.063719242811203, -0.16110455989837646, -0.2894609272480011, 0.045685138553380966, 0.3787928521633148, 0.558189332485199, -0.05286308005452156, 0.15000592172145844, -0.24770133197307587, -0.24770930409431458, -0.3083847463130951, 0.014664016664028168, 0.09064112603664398, -0.17407074570655823, -0.025309525430202484, 0.05884459614753723, 0.26641160249710083, 0.637290894985199, 0.0469006672501564, 0.14139916002750397, 0.08129165321588516, -0.2351590096950531, -0.12353920936584473, 0.17031626403331757, 0.18290166556835175, 0.12640762329101562, 0.1413463056087494, -0.009156841784715652, -0.2856316566467285, 0.027415456250309944, -0.03308270126581192, 0.4702058732509613, 0.3248136043548584, -0.3375096321105957, -0.10117151588201523, -0.23019550740718842, -0.39344772696495056, -0.5377247929573059, -0.43573400378227234, -0.41512221097946167, -0.3019927442073822, -0.03800147399306297, 0.2989155650138855, 0.021401604637503624, 0.3244091272354126, -0.2172331064939499, 0.29080939292907715, -0.27205801010131836, -0.4005090594291687, 0.05991336330771446, -0.2860785126686096, -0.3161654770374298, 0.00833261851221323, 0.24271610379219055, -0.04014505073428154, 0.4774545431137085, -0.09524811804294586, -0.3994390666484833, -0.24949169158935547, -0.3282950520515442, 0.08962168544530869, 0.08339371532201767, 0.5332791805267334, 0.41089367866516113, 0.11481615900993347, 0.5304831862449646, -0.03753087297081947, 0.017749475315213203, -0.07441186904907227, -0.35614123940467834, -0.23213955760002136, -0.014169537462294102, 0.21951958537101746, -0.055851344019174576, -0.33744144439697266, -0.5183429718017578, -0.11137231439352036, 0.08031538128852844, 0.05806170031428337, 0.07833413034677505, 0.21258775889873505, -0.09202129393815994, 0.10448270291090012, 0.06396204233169556, 0.012659781612455845, -0.13171133399009705, 0.12492002546787262, 0.04393640533089638, -0.16999368369579315, -0.03391962870955467, 0.12768378853797913, -0.06895232200622559, 0.04899897426366806, -0.14158327877521515, -0.43566039204597473, -0.4000312387943268, 0.16557326912879944, 0.17159414291381836, 0.36952999234199524, -0.29879215359687805, 0.2539084255695343, 0.21395757794380188, 0.12991242110729218, -0.10910612344741821, -0.4278598725795746, 0.22798621654510498, 0.10129064321517944, 0.3282497823238373, 0.27787476778030396, 0.36871102452278137, -0.19837863743305206, 0.8586010336875916, 0.20692935585975647, -0.2106233388185501, -0.14095428586006165, -0.24552248418331146, -0.06579769402742386, 0.03593473508954048, 0.013525295071303844, 0.21881991624832153, -0.027355510741472244, -0.31299540400505066, 0.30828824639320374, -0.23085933923721313, -0.13247287273406982, -0.12156091630458832, 0.20832249522209167, -0.02576570212841034, -0.11315495520830154, 0.06008422374725342, -0.14215587079524994, 0.060876939445734024, 0.008019084110856056, -0.0666147917509079, -0.2422122210264206, -0.25388818979263306, 0.028734354302287102, 0.33859941363334656, -0.18707042932510376, 0.19989916682243347, -0.15251681208610535, 0.018473321571946144, -0.4077703058719635, 0.33840200304985046, 0.3403618037700653, 0.5474084615707397, 0.1201217770576477, -0.03538355603814125, 0.1302526593208313, 0.14203637838363647, 0.2074132263660431, -0.0937039777636528, 0.14432428777217865, 0.2877649664878845, -0.07581381499767303, -0.13675835728645325, -0.3179689049720764, 0.036600641906261444, 0.18235066533088684, 0.03332753852009773, 0.12281156331300735, -0.4709644317626953, -0.09846274554729462, 0.28636619448661804, 0.1944669485092163, -0.1631324291229248, -0.33340057730674744, -0.10241498798131943, 0.048791442066431046, -0.07874776422977448, -0.21630644798278809, -0.24037282168865204, 0.044299714267253876, -0.03885626047849655, 0.30861982703208923, 0.008862534537911415, -0.03533400222659111, -0.061219144612550735, 0.26634854078292847, 0.203140527009964, -0.20016618072986603, 0.2893237769603729, 0.2616366446018219, 0.08416706323623657, 0.31533539295196533, 0.20172274112701416, -0.08268649131059647, -0.2403508722782135, -0.08250290900468826, 0.01741078682243824, 0.30566394329071045, 0.30834633111953735, 0.24035508930683136, -0.008557828143239021, -0.18761910498142242, -0.14486688375473022, -0.28925856947898865, 0.16906338930130005, 0.2935924530029297, -0.1301058828830719, -0.10033601522445679, -0.328279972076416, 0.7530413866043091, 0.17228135466575623, -0.2961040139198303, 0.5668507814407349, -0.23936688899993896, -0.41096052527427673, -0.013560386374592781, 0.10263330489397049, 0.9836556911468506, 0.07682210952043533, 0.21855399012565613, 0.22350849211215973, -0.196955606341362, 0.3902524411678314, 0.16065633296966553, 0.06136449798941612, -0.5651342868804932, 0.08804205060005188, 0.029159659519791603, -0.09895415604114532, 0.2444816380739212, -0.06411735713481903, -0.3517037034034729, 0.3407967984676361, 0.08218179643154144, 0.29725393652915955, 0.25180643796920776, 0.3703334629535675, -0.046204786747694016, -0.0392565093934536, -0.2405530959367752, 0.05103296414017677, -0.22437438368797302, 0.3719373345375061, -0.2197384387254715, 0.03416544198989868, -0.1888171136379242, -0.07785965502262115, -0.32790815830230713, -0.2395802140235901, -0.15013854205608368, 0.0694781020283699, 0.019197087734937668, -0.18754315376281738, -0.02779516950249672, 0.23903660476207733, 0.06307201832532883, -0.12049292027950287, -0.1902085542678833, 0.11415348947048187, -0.20278052985668182, 0.08895596861839294, 0.2375626266002655, -0.017683524638414383, 0.32868754863739014, -0.3449383080005646, -0.11815748363733292, 0.1530843824148178, -0.18728302419185638, -0.2327251434326172, -0.08339548110961914, -0.1945822387933731, 0.04582109674811363, -0.34027644991874695, -0.24017827212810516, -0.07632223516702652, 0.027845138683915138, -0.10413619875907898, 0.03824279084801674, 0.16137798130512238, -0.3257570266723633, 0.017180662602186203, 0.05420560762286186, -0.33293431997299194, 0.04696368798613548, 0.36871370673179626, -0.08637315779924393, -0.0008771973080001771, 0.6735668778419495, -0.2475472092628479, 0.08735302835702896, -0.10800735652446747, 0.00033757989876903594, -0.3327270448207855, -0.13322947919368744, -0.04113689810037613, -0.043933335691690445, -0.2875957190990448, -0.06571585685014725, 0.5596664547920227, 0.5117821097373962, 0.10100823640823364, 0.06241867318749428, -0.3830115795135498, -0.3422270119190216, 0.07540882378816605, -0.12961703538894653, 0.26565855741500854, -0.07393582165241241, 0.2888076603412628, 0.09227890521287918, 0.044827744364738464, -0.20386137068271637, -0.14279790222644806, 0.01740514114499092, 0.23781272768974304, -0.16256766021251678, 0.12307856976985931, -0.11569207161664963, 0.028434082865715027, -0.014393133111298084, -0.10343123227357864, -0.26048508286476135, -0.04680192470550537, -0.16584151983261108, 0.225122332572937, 0.03080480918288231, 0.08343669772148132, 0.012772651389241219, -0.1899290829896927, 0.09936899691820145, 0.1725461781024933, 0.004947731737047434, 0.11314801126718521, -0.06754770129919052, 0.27282944321632385, 0.32850661873817444, 0.11845222115516663, -0.12467202544212341, 0.27501195669174194, 0.10709872841835022, 0.21082834899425507, -0.006238408386707306, 0.3608352839946747, -0.2627810835838318, 0.03662109375, 0.1833149790763855, 0.18572686612606049, -0.20792990922927856, 0.13544856011867523, 0.01638094149529934, -0.04313182830810547, 0.03326157107949257, 0.14998763799667358, 0.3344949781894684, 0.7781230211257935, -0.31009286642074585, -0.17035150527954102, 0.4529954195022583, 0.06668002158403397, -0.192729651927948, -0.05105968192219734, 0.33120474219322205, 0.040324617177248, 0.07683911919593811, 0.26023226976394653, 0.011645044200122356, -0.06465482711791992, 0.1997525542974472, 0.17046508193016052, -0.010798650793731213, -0.2592611312866211, 0.08984921872615814, 0.3426205515861511, -0.18214915692806244, 0.16570426523685455, 0.3916596472263336, 0.1482623964548111, 0.2784763276576996, 0.007383354473859072, 0.13599735498428345, 0.4200589060783386, 0.04143046587705612, -0.07995546609163284, 0.09235884249210358, 0.1196613684296608, 0.10311779379844666, 0.2714880108833313, -0.11963316053152084, 0.03871641680598259, 0.4046541154384613, 0.5331206321716309, -0.2811601459980011, -0.14102178812026978, -0.34136176109313965, -0.15611936151981354, 0.06328184902667999, -0.2039298564195633, -0.3100762963294983, 0.028065113350749016, -0.12455876171588898, -0.12496023625135422, -0.17175205051898956, -0.3785262405872345, 0.26906606554985046, 0.06951303780078888, -0.37244653701782227, -0.3337211608886719, -0.21656209230422974, 0.24834316968917847, 0.09707966446876526, -0.355741024017334, 0.15877223014831543, -0.02369745448231697, -0.03222399204969406, 0.2971590459346771, 0.337347149848938, 0.4456101357936859, -0.01678173430263996, -0.3285939395427704, -0.2467663437128067, -0.1693970412015915, 0.09490405768156052, 0.03207847476005554, 0.2031138688325882, 0.010795323178172112, -0.30086663365364075, 0.5508683919906616, 0.07143072783946991, 0.05135613679885864, 0.12386081367731094, 0.3914082646369934, -0.43770965933799744, -0.17570674419403076, 0.5832920074462891, -0.017454417422413826, -0.182017982006073, 0.005244074854999781, 0.2448505312204361, -0.32897523045539856, 0.18183258175849915, 0.02182474173605442, 0.04144676402211189, 0.1453244686126709, 0.08675617724657059, 0.021519968286156654, -0.11364676058292389, 0.4138198792934418, 0.2832196354866028, -0.06735663115978241, -0.2639589309692383, -0.2231563925743103, -0.7593228816986084, 0.18017226457595825, -0.13770219683647156, -0.3707537353038788, 0.0003888763312716037, -0.21278631687164307, 0.246369406580925, 0.03339150920510292, -0.11697361618280411, -0.017687788233160973, -0.00454599829390645, 0.1856839656829834, -0.23927196860313416, -0.3589063286781311, -0.13322195410728455, 0.014740189537405968, 0.01359924953430891, -0.36586934328079224, 0.11531183868646622, -0.2808384597301483, -0.11550085991621017, 0.061025869101285934, 0.012041742913424969, 0.5453345775604248, 0.05366500839591026, 0.24777215719223022, -0.16881850361824036, 0.7727639675140381, 0.1596136838197708, -0.0645497664809227, -0.15310201048851013, 0.09544996917247772, -0.13104106485843658, 0.14249755442142487, -0.0821685865521431, 0.12933450937271118, -0.16671274602413177, 0.49387991428375244, -0.3666074275970459, 0.5406872630119324, -0.09922464936971664, 0.08082845062017441, -0.22181886434555054, -0.0370963029563427, -0.15141664445400238, 0.11986321955919266, 0.22769077122211456, 0.2716069221496582, -0.10734829306602478, 0.3570466637611389, -0.2341272532939911, -0.35361194610595703, 0.48149460554122925, -0.10037002712488174, -0.1550842523574829, 0.021686933934688568, 0.32063138484954834, -0.09507087618112564, 0.07444220781326294, -0.4297243356704712, 0.23768439888954163, 0.3241075277328491, -0.04060264676809311, 0.008704984560608864, 0.21143370866775513, 0.2268538624048233, 0.08315180987119675, 0.001767108915373683, 1.0600348711013794, -0.10923667997121811, 0.02344803512096405, -0.2541804313659668, -0.3114486634731293 ]
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.
[ -0.15275150537490845, 0.21056921780109406, 0.005356146022677422, -0.15212182700634003, 0.1695309430360794, -0.12297341227531433, 0.3444475531578064, 0.37722504138946533, 0.013979529030621052, 0.22968514263629913, 0.1677788943052292, 0.015225007198750973, -0.33424168825149536, 0.03130737692117691, 0.022818395867943764, -0.10024451464414597, -0.05941534787416458, 0.1970064640045166, -0.07303550094366074, 0.34329548478126526, -0.0601121187210083, -0.07316172122955322, 0.001971533289179206, 0.19343027472496033, 0.010817328467965126, 0.14580456912517548, -0.1153508722782135, -0.01875404641032219, -0.38851669430732727, -0.858357310295105, 0.13438116014003754, 0.09185036271810532, 0.23209969699382782, 0.4368850886821747, -0.00012612353020813316, -0.0739654153585434, 0.5866604447364807, -0.024042129516601562, -0.7674410343170166, -0.16982930898666382, 0.3446428179740906, -0.3756410479545593, 0.2535513639450073, -0.06578025221824646, 0.11371959000825882, -0.1838432252407074, 0.18135349452495575, 0.07539600878953934, 0.1361459344625473, 0.24556030333042145, 0.09919442236423492, 0.31069278717041016, 0.028131229802966118, -0.11110047250986099, -0.059569377452135086, 0.12354497611522675, -0.14380046725273132, 0.514221727848053, 0.48955073952674866, 0.24067158997058868, -0.07860221713781357, -0.0708247721195221, 0.06261499226093292, 0.4359807074069977, 0.13713961839675903, -0.03264545649290085, 0.5384782552719116, 0.04423699155449867, -0.1900421380996704, 0.1604461818933487, 0.8945167660713196, -0.508608341217041, -0.4655519127845764, -0.06761329621076584, 0.017939787358045578, -0.10367343574762344, 0.1076582819223404, -0.19388841092586517, -0.37744948267936707, 0.20891010761260986, -0.16695833206176758, -0.3321853280067444, -0.09904033690690994, 0.28892990946769714, -0.17620734870433807, 0.231237530708313, 0.12704668939113617, 0.251554012298584, -0.02958354726433754, -0.06738026440143585, 0.15492850542068481, -0.05209415778517723, 0.015268944203853607, 0.13405336439609528, 0.08804340660572052, -0.3319090008735657, -0.4149753153324127, -0.29810580611228943, 0.1995212286710739, 0.02563527226448059, -0.5338376760482788, -0.18990309536457062, -0.06855519860982895, 0.17624151706695557, 0.08776311576366425, 0.2744107246398926, 0.3668699562549591, 0.22054284811019897, 0.311817467212677, -0.010775198228657246, 0.28586405515670776, 0.21933381259441376, -0.005532244220376015, -0.27972203493118286, -0.002387609099969268, 0.07539398223161697, 0.1061914935708046, -0.21187777817249298, -0.02138405106961727, -0.25974956154823303, -0.422829806804657, -0.02772076241672039, -0.09297060966491699, 0.23292586207389832, -0.17843395471572876, -0.14660950005054474, -0.019972611218690872, 0.33416399359703064, -0.10098529607057571, -0.19496288895606995, -0.00010904225200647488, -0.02032029815018177, -0.24842841923236847, -0.06086070463061333, 0.22468295693397522, -0.13178780674934387, 0.359870970249176, -0.3554438054561615, -0.14415745437145233, -0.03814078867435455, -0.022520434111356735, 0.030894123017787933, -0.17065323889255524, 0.2882438600063324, -0.23571866750717163, 0.017072271555662155, 0.05582601577043533, -0.30029070377349854, -0.21881400048732758, 0.21825045347213745, -0.00323013449087739, -0.43608358502388, -0.13643647730350494, 0.06006932631134987, -0.5637034177780151, 0.02172395959496498, 0.04728560894727707, -0.16956357657909393, 0.18883579969406128, -0.2967650890350342, 0.14172428846359253, -0.2503824830055237, -0.36074402928352356, -0.15156862139701843, -0.17838171124458313, 0.2280636429786682, -0.3107917904853821, -0.2898619472980499, -0.229727640748024, -0.38468170166015625, 0.07798092812299728, -0.13051730394363403, -0.07232090085744858, 0.20862674713134766, -0.33610519766807556, -0.03591899201273918, 0.6512777805328369, -0.5617481470108032, -0.28832465410232544, 0.6501520276069641, -0.3996002972126007, -0.25953537225723267, 0.2737455666065216, -0.07269962131977081, 0.14991028606891632, -0.44446325302124023, -0.13253694772720337, 0.00422174297273159, -0.04450634866952896, -0.11314094066619873, -0.08715606480836868, -0.42463457584381104, 0.2098805159330368, 0.20373673737049103, -0.06479045003652573, 0.22341397404670715, 0.1218448132276535, 0.10671240836381912, 0.44717109203338623, 0.03521319478750229, 0.16927513480186462, -0.21021628379821777, 0.20478938519954681, -0.12286790460348129, -0.10746625810861588, -0.13562634587287903, -0.5853859782218933, 0.24203677475452423, -0.04733669012784958, 0.15349699556827545, -0.006374368444085121, -0.06405384093523026, -0.4417176842689514, 0.028502406552433968, 0.1057552620768547, -0.21586720645427704, -0.05539104714989662, 0.14584705233573914, 0.025066383183002472, -0.03469594940543175, -0.026319218799471855, -0.007540535181760788, -0.29132288694381714, 0.1625141054391861, 0.20891229808330536, -0.1484832912683487, -0.0833222046494484, 0.09957931190729141, 0.15043024718761444, 0.25229009985923767, -0.06979327648878098, -0.30531662702560425, 0.0206984244287014, 0.3233891725540161, 0.2586061358451843, 0.14943844079971313, 0.07985381782054901, 0.050534434616565704, 0.18621470034122467, 0.2668219804763794, 0.1463746577501297, 0.2067103385925293, 0.024877075105905533, -0.27857062220573425, 0.023999130353331566, 0.15529471635818481, 0.15474042296409607, -0.02467263862490654, -0.21239587664604187, -0.042884550988674164, 0.29667001962661743, 0.08589226007461548, -0.15950694680213928, -0.28861546516418457, 0.10071874409914017, 0.2132439762353897, 0.627162754535675, 0.013069916516542435, 0.2344062328338623, -0.16070429980754852, -0.19119088351726532, -0.4197954833507538, 0.04762929305434227, 0.2504351735115051, -0.08639678359031677, -0.06156589835882187, -0.03870211914181709, 0.245597705245018, 0.8113179206848145, 0.007162152323871851, 0.22367329895496368, 0.059366174042224884, -0.062219470739364624, -0.2031267136335373, 0.21539625525474548, 0.18741610646247864, -0.019954949617385864, 0.14958535134792328, -0.009534570388495922, -0.21307161450386047, 0.011150955222547054, -0.029949363321065903, 0.37450698018074036, 0.38911503553390503, -0.4598504304885864, -0.07126249372959137, -0.24342282116413116, -0.463784784078598, -0.35133451223373413, -0.3214045763015747, -0.28302067518234253, -0.28866177797317505, -0.07606270164251328, 0.3205259144306183, 0.042951785027980804, 0.2218059003353119, -0.27010655403137207, 0.2553025186061859, -0.26923397183418274, -0.29005882143974304, 0.03959788754582405, -0.138365238904953, -0.37830352783203125, -0.01013457216322422, 0.20340538024902344, -0.12987159192562103, 0.49237075448036194, -0.04010777547955513, -0.2758660614490509, -0.21922950446605682, -0.3464270532131195, 0.17859600484371185, -0.008019930683076382, 0.5385331511497498, 0.5283277034759521, 0.011747214011847973, 0.49220019578933716, 0.04509775713086128, -0.10305839031934738, -0.28246697783470154, -0.44600310921669006, -0.19963915646076202, 0.04542338103055954, 0.1237611472606659, -0.22519958019256592, -0.24200619757175446, -0.5144344568252563, -0.1097060814499855, 0.14452889561653137, -0.034998420625925064, 0.12034610658884048, 0.27052798867225647, -0.1861664354801178, 0.13806472718715668, -0.006836252752691507, -0.061954375356435776, -0.1589338481426239, 0.07213596999645233, -0.09341350197792053, -0.1512037068605423, -0.15034742653369904, 0.05138729512691498, 0.009144775569438934, -0.02612355351448059, -0.03587004914879799, -0.3401898443698883, -0.4658867120742798, 0.1552971750497818, 0.13913817703723907, 0.3482474982738495, -0.22888308763504028, 0.19969870150089264, 0.23388059437274933, 0.16123874485492706, -0.06107718124985695, -0.34528324007987976, 0.14796283841133118, 0.04752764478325844, 0.24498234689235687, 0.26928913593292236, 0.41952982544898987, -0.01958627998828888, 0.7755737900733948, 0.14332544803619385, -0.09589071571826935, -0.07831799238920212, -0.2671539783477783, 0.006282673217356205, -0.05424756184220314, -0.005567406304180622, 0.3108569383621216, -0.10074681788682938, -0.2828224301338196, 0.31856876611709595, -0.18947573006153107, 0.05149194225668907, -0.09044608473777771, 0.26561543345451355, -0.06809309870004654, -0.04419058933854103, 0.050073470920324326, -0.28195005655288696, 0.03934032469987869, -0.012657828629016876, -0.03752124309539795, -0.13320256769657135, -0.21860474348068237, 0.09952650964260101, 0.36235520243644714, -0.01867963746190071, 0.29034310579299927, -0.11583908647298813, 0.1568782478570938, -0.3423961400985718, 0.2570708096027374, 0.4411434233188629, 0.6185429096221924, 0.0831977054476738, 0.09134894609451294, 0.18326643109321594, 0.12001955509185791, 0.2273234874010086, -0.03647727146744728, 0.23378092050552368, 0.06326095759868622, -0.23603394627571106, -0.01786557398736477, -0.22646212577819824, 0.0005946849705651402, 0.09605279564857483, -0.07571954280138016, 0.17361819744110107, -0.47740060091018677, -0.23207798600196838, 0.21772125363349915, 0.1053234189748764, -0.07665842771530151, -0.38931044936180115, -0.0650530532002449, 0.03828468546271324, -0.05286571756005287, -0.3129257261753082, -0.1361313909292221, -0.005222050007432699, -0.08902187645435333, 0.43324968218803406, -0.10226185619831085, -0.14556938409805298, -0.005133246071636677, 0.10432886332273483, 0.04470985010266304, -0.19614458084106445, 0.20612114667892456, 0.1978483349084854, 0.1746518462896347, 0.2170514613389969, 0.1883532702922821, -0.08402597904205322, -0.15616196393966675, -0.03743556886911392, -0.05730266869068146, 0.3986189365386963, 0.3663228750228882, 0.10062583535909653, 0.042771704494953156, -0.21017588675022125, -0.1530294269323349, -0.3422596752643585, 0.2205570936203003, 0.28233081102371216, -0.1639905422925949, -0.051029957830905914, -0.284090131521225, 0.6471734046936035, 0.14356432855129242, -0.24567492306232452, 0.45120638608932495, -0.15048593282699585, -0.30368855595588684, -0.02545827440917492, 0.12860316038131714, 0.9857158660888672, 0.04336363449692726, 0.228694885969162, 0.38514620065689087, -0.3389439284801483, 0.4281971752643585, 0.14748509228229523, 0.026662638410925865, -0.5198193192481995, -0.05517949163913727, -0.05943053960800171, -0.1654731035232544, 0.2741527259349823, -0.15859319269657135, -0.20023596286773682, 0.3456084132194519, 0.0879504531621933, 0.3736132085323334, 0.4137502610683441, 0.39923733472824097, -0.10544481873512268, -0.0578831247985363, -0.16087795794010162, 0.022064870223402977, -0.18493811786174774, 0.2843340039253235, -0.21571725606918335, -0.02065260522067547, -0.18660974502563477, -0.02854018285870552, -0.24675121903419495, -0.23737679421901703, -0.12552998960018158, 0.017985118553042412, 0.1404607594013214, -0.10921979695558548, 0.05769842118024826, 0.16871491074562073, 0.18284764885902405, -0.09496632218360901, -0.13536006212234497, 0.025371426716446877, -0.23170380294322968, 0.2188091278076172, 0.30446043610572815, 0.028276942670345306, 0.3999039828777313, -0.28152602910995483, -0.07076135277748108, 0.1486913412809372, -0.27996036410331726, -0.2452803999185562, -0.17162904143333435, -0.21622014045715332, 0.16090387105941772, -0.35440149903297424, -0.2652539610862732, -0.18462562561035156, -0.020434681326150894, -0.036143217235803604, -0.03854867443442345, 0.2329091727733612, -0.33000391721725464, -0.013231867924332619, -0.03742232546210289, -0.3219470679759979, 0.03071586787700653, 0.4821368455886841, -0.04637717083096504, -0.17012323439121246, 0.6307935118675232, -0.2415556162595749, 0.028976138681173325, -0.08974552154541016, 0.032223571091890335, -0.18571774661540985, -0.18621821701526642, 0.038522880524396896, -0.07007024437189102, -0.32104843854904175, -0.03493758291006088, 0.6218964457511902, 0.5002495646476746, 0.02622940205037594, -0.07093346863985062, -0.42394766211509705, -0.4768604338169098, -0.019827622920274734, -0.21596616506576538, 0.14384187757968903, -0.2070552259683609, 0.3047220706939697, 0.09505709260702133, 0.03632986173033714, -0.16911859810352325, -0.23322783410549164, 0.12708957493305206, 0.261135071516037, -0.2023114711046219, 0.03588296100497246, -0.010400906205177307, 0.04883389174938202, -0.07755542546510696, 0.05452270060777664, -0.3443521559238434, -0.022740541025996208, -0.22121800482273102, 0.24034817516803741, 0.12901587784290314, 0.015883522108197212, 0.10097530484199524, -0.12922509014606476, -0.030205244198441505, 0.19713298976421356, -0.03883175179362297, 0.14353664219379425, -0.08869221806526184, 0.26557838916778564, 0.31918033957481384, 0.16784854233264923, -0.0894203931093216, 0.29042306542396545, 0.08136001974344254, 0.2633010745048523, -0.037363383919000626, 0.3029986321926117, -0.2278713881969452, 0.036965735256671906, 0.10371316224336624, 0.2465100884437561, -0.018953152000904083, 0.24331055581569672, 0.11851643025875092, -0.0831337571144104, 0.09952311217784882, 0.0027662022039294243, 0.3833658993244171, 0.7761162519454956, -0.18420854210853577, -0.31482014060020447, 0.4969842731952667, 0.01780625246465206, -0.10703206807374954, 0.006450393702834845, 0.24761874973773956, 0.16800163686275482, 0.025308171287178993, 0.3604944944381714, 0.030094275251030922, -0.035909172147512436, 0.09229403734207153, 0.10623396188020706, 0.08866777271032333, -0.2641749083995819, -0.2379094362258911, 0.5210915207862854, -0.22591117024421692, 0.24124936759471893, 0.35357728600502014, 0.13580664992332458, 0.2959987223148346, 0.0376608744263649, 0.0756630226969719, 0.5019505620002747, -0.010012992657721043, 0.043126750737428665, 0.14046449959278107, 0.13297909498214722, 0.014500248245894909, 0.2870313227176666, 0.0750098004937172, 0.01374475471675396, 0.3375609815120697, 0.5905367732048035, -0.20217737555503845, -0.16447226703166962, -0.17271223664283752, -0.16748769581317902, 0.07695972174406052, -0.22807033360004425, -0.2735541760921478, 0.12884899973869324, -0.2124818116426468, -0.10302963852882385, -0.2302057296037674, -0.29267197847366333, 0.21165403723716736, 0.052750878036022186, -0.33862540125846863, -0.27144673466682434, -0.40145108103752136, 0.14914578199386597, 0.05067789927124977, -0.3459288775920868, 0.12932954728603363, -0.07099087536334991, 0.11562038213014603, 0.3956111669540405, 0.2540733814239502, 0.3397335708141327, 0.06733046472072601, -0.30972784757614136, -0.3134117126464844, -0.32432126998901367, 0.15070973336696625, 0.008740011602640152, 0.39771589636802673, 0.017714744433760643, -0.11327949911355972, 0.6163250207901001, 0.004213014617562294, 0.10033023357391357, 0.2640596926212311, 0.3659212589263916, -0.5369965434074402, -0.1176462173461914, 0.7859680652618408, 0.0799904391169548, -0.1935427486896515, 0.07940418273210526, 0.22868479788303375, -0.1913021206855774, 0.17380355298519135, 0.04644022509455681, -0.11265183985233307, 0.18436302244663239, 0.11147178709506989, 0.015183702111244202, -0.05229572579264641, 0.34844404458999634, 0.4192717373371124, -0.0645555704832077, -0.2744944095611572, -0.01557890884578228, -0.6457433700561523, 0.18133974075317383, -0.0532815083861351, -0.25216230750083923, 0.004350471310317516, -0.24237245321273804, 0.1777762472629547, 0.11598416417837143, -0.2468877136707306, 0.13935714960098267, -0.08880768716335297, 0.08666245639324188, -0.14721152186393738, -0.3679317533969879, -0.1788339614868164, 0.08051315695047379, -0.07546975463628769, -0.4362682104110718, 0.2002209722995758, -0.4206125736236572, -0.12240849435329437, 0.07628058642148972, 0.015164745040237904, 0.5335346460342407, -0.019309718161821365, 0.29269614815711975, -0.054952219128608704, 0.533309817314148, 0.19510102272033691, 0.006029760465025902, -0.18519791960716248, 0.1820680946111679, -0.19926540553569794, 0.13084809482097626, -0.04895918071269989, 0.06954223662614822, -0.14772464334964752, 0.5281544327735901, -0.4214254915714264, 0.5098819136619568, -0.05077604204416275, 0.16153916716575623, -0.3290116786956787, 0.10240166634321213, -0.2403387427330017, 0.08635560423135757, 0.28884053230285645, 0.21862661838531494, -0.11910427361726761, 0.3717104494571686, -0.25404420495033264, -0.20215846598148346, 0.477955162525177, -0.18597476184368134, -0.1001986563205719, 0.15425093472003937, 0.3032083809375763, 0.19870327413082123, -0.18652042746543884, -0.5294913649559021, 0.15762922167778015, 0.32636627554893494, -0.10436984896659851, -0.04502720758318901, 0.22184590995311737, 0.20797044038772583, 0.06086600944399834, 0.007538784760981798, 0.9585250020027161, -0.13304029405117035, 0.08896134048700333, -0.06882631778717041, -0.24809624254703522 ]
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?
[ 0.057603031396865845, 0.24249684810638428, -0.0005984987365081906, -0.016872894018888474, -0.02622312307357788, -0.10918811708688736, 0.3703443706035614, 0.27449798583984375, 0.0017392392037436366, -0.07336346060037613, 0.19066597521305084, 0.10673833638429642, -0.259065717458725, 0.05271518602967262, 0.06369071453809738, -0.17159421741962433, 0.0611245222389698, 0.2052566409111023, -0.0903208777308464, 0.19936907291412354, -0.10669586062431335, -0.013935642316937447, -0.021723810583353043, 0.2025269716978073, -0.09186001867055893, -0.044857654720544815, -0.2902246117591858, 0.06356658041477203, -0.35639965534210205, -0.9518454670906067, 0.2559971511363983, 0.16132839024066925, 0.21488942205905914, 0.3902356028556824, -0.00012299601803533733, -0.13422220945358276, 0.5136100649833679, -0.030448010191321373, -0.7506924867630005, -0.19761863350868225, 0.32741355895996094, -0.43662112951278687, 0.3908314108848572, -0.08335299044847488, 0.20270659029483795, -0.2667604982852936, 0.3141605257987976, 0.207979217171669, 0.1440407633781433, 0.21452786028385162, 0.07082146406173706, 0.2529841661453247, 0.029718715697526932, -0.09841018170118332, -0.04017733410000801, 0.11929956078529358, -0.034478019922971725, 0.5099807977676392, 0.39377081394195557, 0.12621475756168365, -0.16321873664855957, 0.1172778531908989, 0.14007438719272614, 0.398349404335022, 0.22597910463809967, -0.017953379079699516, 0.5039623975753784, -0.04759890213608742, -0.2023797482252121, 0.05471792072057724, 0.8494492769241333, -0.5009222626686096, -0.4326479732990265, -0.14169369637966156, 0.0322897806763649, -0.16478882730007172, 0.21547117829322815, -0.10920216143131256, -0.15507255494594574, 0.057124797254800797, -0.24445419013500214, -0.11133302003145218, -0.15746982395648956, 0.33689171075820923, -0.22511886060237885, 0.2915147840976715, -0.01036517508327961, 0.196104496717453, -0.030724596232175827, -0.18355897068977356, 0.057907700538635254, -0.2011939138174057, 0.07630181312561035, 0.17462003231048584, 0.016124248504638672, -0.28860238194465637, -0.2943037748336792, -0.18726441264152527, 0.1245485246181488, 0.014857195317745209, -0.44811877608299255, -0.17962484061717987, -0.03736487776041031, 0.2117084115743637, 0.2502307593822479, 0.34051209688186646, 0.39323949813842773, 0.19038569927215576, 0.1532820463180542, 0.08018222451210022, 0.15144778788089752, 0.15082016587257385, -0.04188396409153938, -0.22569631040096283, -0.011925511993467808, 0.0255417637526989, 0.1661313772201538, -0.282471626996994, -0.05167897790670395, -0.12572962045669556, -0.5409553050994873, -0.07574271410703659, -0.13150538504123688, 0.23289436101913452, -0.17989909648895264, -0.0450725257396698, 0.03212342783808708, 0.41973376274108887, -0.1784876585006714, -0.15913519263267517, -0.01568870060145855, 0.0746147483587265, -0.25813594460487366, -0.04566468670964241, 0.17432168126106262, -0.09063214808702469, 0.4508667588233948, -0.21950915455818176, -0.33825036883354187, -0.17939616739749908, 0.06953271478414536, 0.013141985982656479, -0.092068612575531, 0.23776043951511383, -0.19513803720474243, 0.06764183938503265, 0.08036789298057556, -0.4109775424003601, -0.2782209813594818, 0.0108876321464777, 0.05627641826868057, -0.3433370590209961, -0.027284763753414154, 0.06280619651079178, -0.4833715856075287, -0.010563389398157597, 0.12420543283224106, 0.09027823805809021, 0.14917360246181488, -0.3166504204273224, 0.10074396431446075, -0.41337764263153076, -0.2347722053527832, -0.09195569902658463, -0.18899689614772797, 0.30426025390625, -0.28402915596961975, -0.27550452947616577, -0.16645711660385132, -0.31274059414863586, 0.09089355170726776, -0.18867778778076172, -0.07610412687063217, 0.2745271921157837, -0.19024626910686493, 0.030188988894224167, 0.5996156334877014, -0.5389537811279297, -0.208756223320961, 0.5325260758399963, -0.27516645193099976, -0.05293510481715202, 0.2893843650817871, -0.07397853583097458, 0.11017443984746933, -0.3594624996185303, -0.10819654911756516, 0.05834299325942993, 0.003969159442931414, -0.19246144592761993, -0.21571461856365204, -0.2680111825466156, 0.1802842617034912, 0.2851310074329376, -0.1920585185289383, 0.01870235800743103, 0.0654810443520546, 0.1850583404302597, 0.3273561894893646, 0.09518381208181381, 0.1283266693353653, -0.18868350982666016, -0.055134810507297516, -0.07846126705408096, -0.12363836914300919, -0.016171490773558617, -0.737356424331665, 0.15254534780979156, -0.10349711030721664, 0.20631952583789825, 0.04400278255343437, 0.003577522234991193, -0.46556249260902405, -0.049690261483192444, -0.0746120736002922, -0.30068254470825195, 0.03034386970102787, 0.23752416670322418, 0.07605670392513275, -0.003968010190874338, 0.0089028999209404, -0.012331632897257805, -0.30985939502716064, 0.24812142550945282, -0.04932263121008873, -0.18346978724002838, -0.08366940915584564, 0.1499592363834381, 0.20539920032024384, 0.25154611468315125, 0.03397839888930321, -0.275988906621933, 0.1046687662601471, 0.3256835639476776, 0.23714859783649445, 0.24148520827293396, 0.17153841257095337, 0.06900230050086975, 0.23371607065200806, 0.3315989375114441, 0.30703678727149963, 0.165581613779068, 0.08676355332136154, -0.26936155557632446, -0.09193151444196701, 0.1549011617898941, 0.24279727041721344, 0.004586389288306236, -0.16184182465076447, -0.06739956885576248, 0.33853983879089355, 0.12129128724336624, -0.23096582293510437, -0.26559656858444214, 0.012635068036615849, 0.2951555550098419, 0.6402700543403625, 0.06318027526140213, 0.086736761033535, -0.034494198858737946, -0.11137327551841736, -0.4682064652442932, -0.023292429745197296, 0.09991350024938583, -0.271459698677063, -0.1655067354440689, -0.019511448219418526, 0.25677603483200073, 0.7403607368469238, 0.014357035048305988, 0.2837745249271393, 0.04305184260010719, -0.26060423254966736, -0.21312537789344788, 0.19014260172843933, 0.29233258962631226, -0.07247406244277954, 0.14260810613632202, 0.0594441257417202, -0.22141094505786896, -0.011801712214946747, -0.14425192773342133, 0.3677428960800171, 0.32906365394592285, -0.4422619044780731, 0.011568523943424225, -0.22433887422084808, -0.4463372826576233, -0.5092366337776184, -0.3774477243423462, -0.26062434911727905, -0.3482596278190613, -0.054240647703409195, 0.22306641936302185, -0.09885343164205551, 0.2085065245628357, -0.04445434361696243, 0.3514874279499054, -0.3339146077632904, -0.06189103424549103, 0.18121939897537231, -0.2874465882778168, -0.43042632937431335, -0.02127765864133835, 0.3774441182613373, 0.010301336646080017, 0.36990195512771606, -0.16296157240867615, -0.3067916929721832, -0.1718650758266449, -0.35774677991867065, 0.17577321827411652, -0.051336295902729034, 0.4902479648590088, 0.43741652369499207, 0.05680619552731514, 0.42201197147369385, -0.0797569528222084, -0.005827446933835745, -0.06160556897521019, -0.35395559668540955, -0.1985042542219162, 0.008053592406213284, 0.011529703624546528, -0.22299189865589142, -0.46764835715293884, -0.5795465707778931, -0.0671650618314743, 0.23849734663963318, 0.02373994141817093, 0.13625791668891907, 0.35848933458328247, -0.23525846004486084, 0.23094302415847778, -0.11682116240262985, 0.061752721667289734, -0.1307677924633026, 0.048422977328300476, 0.033469658344984055, -0.1410066932439804, -0.14650937914848328, 0.08306869864463806, -0.0692908838391304, -0.05623940750956535, -0.1251676082611084, -0.40929535031318665, -0.46038922667503357, 0.19253315031528473, -0.01987825706601143, 0.38545799255371094, -0.11975408345460892, 0.19413025677204132, 0.22859132289886475, 0.1526886373758316, -0.09221981465816498, -0.2539873421192169, 0.3085111677646637, 0.04816817864775658, 0.2682611346244812, 0.266781210899353, 0.3099624514579773, -0.0824957937002182, 0.8548855781555176, 0.23433233797550201, -0.22297149896621704, -0.17823854088783264, -0.028072040528059006, -0.07432743161916733, 0.008990345522761345, 0.03905418887734413, 0.24598316848278046, -0.19373948872089386, -0.25766152143478394, 0.45392850041389465, -0.114717036485672, -0.07941077649593353, -0.012890457175672054, 0.28912872076034546, -0.10928487032651901, -0.005735121667385101, 0.19463348388671875, -0.1576356142759323, 0.023471355438232422, 0.010373285040259361, -0.03370669484138489, -0.15833383798599243, -0.06846459209918976, -0.03319685161113739, 0.34193718433380127, -0.1114078089594841, 0.21972525119781494, -0.2726879417896271, -0.06407061964273453, -0.2533634305000305, 0.2327742874622345, 0.3978850543498993, 0.4647171199321747, 0.1509024053812027, -0.03166869655251503, 0.21203112602233887, 0.20424804091453552, 0.08792249113321304, -0.03190029039978981, 0.10637728124856949, 0.18112285435199738, -0.22042028605937958, -0.030371354892849922, -0.3028002083301544, -0.22541546821594238, 0.042755309492349625, 0.019967298954725266, 0.17918391525745392, -0.4684465229511261, -0.24010694026947021, 0.32111090421676636, 0.062143951654434204, -0.05001577362418175, -0.23638132214546204, 0.023509353399276733, 0.0056046550162136555, -0.04983547702431679, -0.17327041923999786, -0.1989143192768097, -0.013889351859688759, -0.1176556721329689, 0.4240594804286957, -0.038367871195077896, 0.04531456157565117, -0.021976131945848465, 0.2474602460861206, 0.09659852087497711, -0.14922766387462616, 0.14238084852695465, 0.2692522406578064, 0.14296649396419525, 0.27166348695755005, 0.18352940678596497, -0.13022014498710632, -0.2218180149793625, 0.05195283889770508, -0.06875796616077423, 0.3754236698150635, 0.44151338934898376, 0.20746366679668427, 0.031254373490810394, -0.24363255500793457, -0.17298321425914764, -0.3376982510089874, 0.277704119682312, 0.3622962236404419, -0.12266848981380463, -0.018863769248127937, -0.39845767617225647, 0.763932466506958, 0.11514189094305038, -0.3186497986316681, 0.3197757303714752, -0.17126931250095367, -0.27541038393974304, 0.06674200296401978, 0.04492722824215889, 0.9632915258407593, 0.18758025765419006, 0.26972949504852295, 0.335610032081604, -0.3050228953361511, 0.5088146924972534, 0.07729941606521606, 0.08329837024211884, -0.5266809463500977, -0.004951012786477804, 0.011899054050445557, -0.08526358753442764, 0.2526700794696808, -0.09175106883049011, -0.290208101272583, 0.3714599907398224, 0.15412476658821106, 0.3369050621986389, 0.2666240334510803, 0.39048609137535095, -0.10265123099088669, -0.08184856176376343, -0.31482934951782227, 0.032741472125053406, -0.1812625527381897, 0.294230580329895, -0.2555634677410126, -0.054535187780857086, -0.19459009170532227, -0.04891155660152435, -0.16635648906230927, -0.23210124671459198, -0.17980769276618958, 0.044044654816389084, 0.09738380461931229, -0.24454793334007263, 0.007083280943334103, 0.2463042438030243, 0.13479113578796387, -0.30617713928222656, -0.08299602568149567, -0.0007393055129796267, -0.20389775931835175, 0.19736719131469727, 0.2699480950832367, -0.052063800394535065, 0.45949187874794006, -0.26142504811286926, 0.004677867516875267, 0.062005817890167236, -0.2999454438686371, -0.01149108074605465, -0.07748068869113922, -0.1792585849761963, 0.18972744047641754, -0.297353595495224, -0.2757723927497864, 0.034674838185310364, 0.01572929322719574, -0.01693558506667614, -0.013131990097463131, 0.1991761028766632, -0.2707010507583618, 0.016309307888150215, -0.09080835431814194, -0.23815378546714783, -0.012713671661913395, 0.34332630038261414, -0.08797547966241837, -0.07147157937288284, 0.7205211520195007, -0.1423327922821045, -0.013048007152974606, -0.08775774389505386, 0.07892696559429169, -0.07721301913261414, -0.21021488308906555, -0.024668319150805473, -0.0939796194434166, -0.3097339868545532, -0.11627881228923798, 0.6664245128631592, 0.5130176544189453, 0.08737719804048538, -0.012764586135745049, -0.391679584980011, -0.33617329597473145, 0.1102447509765625, -0.21333107352256775, 0.20463062822818756, -0.15378157794475555, 0.23010995984077454, 0.20599886775016785, 0.09173551946878433, -0.19896350800991058, -0.21286627650260925, -0.05241968110203743, 0.32197800278663635, -0.20810827612876892, 0.008621268905699253, -0.022832270711660385, 0.07475634664297104, -0.07202237099409103, -0.0303916335105896, -0.36013880372047424, -0.036116380244493484, -0.16067586839199066, 0.23085978627204895, 0.16787376999855042, -0.10099497437477112, 0.051194943487644196, -0.056667596101760864, 0.0830276757478714, 0.15836568176746368, -0.06596110761165619, -0.06922005861997604, -0.046982504427433014, 0.31562039256095886, 0.22112755477428436, 0.26535165309906006, -0.06126149743795395, 0.15333591401576996, 0.07793194055557251, 0.1418081820011139, 0.08331627398729324, 0.18333099782466888, -0.3230111598968506, 0.013775057159364223, 0.07649023830890656, 0.29494181275367737, -0.1126130223274231, 0.2398706078529358, 0.11625227332115173, -0.03209947049617767, 0.1410961151123047, 0.08694125711917877, 0.2844574451446533, 0.84383225440979, -0.1847463846206665, -0.2881440222263336, 0.41060060262680054, 0.017603576183319092, -0.042186688631772995, 0.0829828679561615, 0.2660801112651825, 0.114603690803051, 0.022272920235991478, 0.28549620509147644, 0.05585028603672981, -0.00353293982334435, 0.03952416032552719, 0.16859540343284607, 0.19243672490119934, -0.21285246312618256, -0.0687088891863823, 0.414207398891449, -0.2447899430990219, 0.2616272568702698, 0.3799397945404053, 0.05843619629740715, 0.21717144548892975, -0.09376535564661026, 0.06660273671150208, 0.4720230996608734, 0.1034340187907219, 0.07247654348611832, 0.14047156274318695, 0.10956514626741409, -0.05472443252801895, 0.461446613073349, 0.027293263003230095, 0.019859718158841133, 0.28616175055503845, 0.6132856011390686, -0.3343772888183594, -0.3217318654060364, -0.17343226075172424, -0.07680433243513107, 0.10721215605735779, -0.18119576573371887, -0.5276592969894409, 0.014906941913068295, -0.2530689239501953, -0.07244852185249329, -0.27018454670906067, -0.3557169437408447, 0.3199385106563568, 0.041180867701768875, -0.3105320632457733, -0.41609370708465576, -0.360491544008255, 0.08546419441699982, 0.12812167406082153, -0.32149261236190796, 0.2644851505756378, 0.2310389131307602, 0.06905602663755417, 0.40392908453941345, 0.3324470520019531, 0.40230703353881836, 0.09167826920747757, -0.38616615533828735, -0.3036429286003113, -0.2539190948009491, 0.10707513988018036, 0.12506259977817535, 0.278357595205307, -0.04276977851986885, -0.21499799191951752, 0.5813985466957092, 0.01551705040037632, 0.07906582206487656, 0.2922722101211548, 0.40926888585090637, -0.5363081693649292, -0.24184061586856842, 0.6709061861038208, -0.0029571945779025555, -0.24788883328437805, 0.10846411436796188, 0.15607409179210663, -0.36080631613731384, 0.08438090234994888, 0.0010096946498379111, 0.07150529325008392, 0.19846242666244507, 0.09360726177692413, 0.02608860470354557, -0.009924471378326416, 0.49392178654670715, 0.49144265055656433, -0.10784617066383362, -0.34578385949134827, -0.10265512019395828, -0.7086463570594788, 0.02320404350757599, -0.024922534823417664, -0.30694249272346497, -0.02865140698850155, -0.13132339715957642, 0.1805088073015213, 0.09593189507722855, -0.2593301236629486, 0.05915505439043045, -0.034586481750011444, 0.07855796068906784, -0.31683021783828735, -0.2308543622493744, -0.16345837712287903, 0.011753891594707966, -0.11217959225177765, -0.24734245240688324, 0.11470603197813034, -0.39097902178764343, -0.14848636090755463, 0.22859196364879608, 0.04866290092468262, 0.4706302583217621, 0.04398013651371002, 0.16804145276546478, -0.014945035800337791, 0.5717008113861084, 0.27138981223106384, -0.1270381659269333, -0.22806526720523834, 0.2154112011194229, -0.1466306447982788, 0.17723962664604187, -0.12875765562057495, 0.36341142654418945, -0.1993665099143982, 0.5129455327987671, -0.20585566759109497, 0.47028860449790955, -0.04468398168683052, 0.05702589824795723, -0.33190685510635376, 0.08545426279306412, -0.2292839139699936, 0.0452086441218853, 0.33929353952407837, 0.3558688461780548, -0.06925654411315918, 0.2981768250465393, -0.26029494404792786, -0.2253071665763855, 0.32007938623428345, -0.31884053349494934, -0.11840415745973587, 0.14298276603221893, 0.2893073558807373, 0.14046335220336914, 0.02453642152249813, -0.6277335286140442, 0.07381168007850647, 0.2311152070760727, -0.17052224278450012, -0.09328329563140869, 0.17623229324817657, 0.23813961446285248, -0.06259177625179291, -0.03592388331890106, 0.9138442873954773, -0.075088731944561, -0.0677710622549057, -0.10770953446626663, -0.33352911472320557 ]
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?
[ -0.1263234168291092, -0.22070905566215515, -0.10063228756189346, 0.2780112326145172, 0.24547505378723145, -0.20522351562976837, 0.1372111439704895, 0.5128591656684875, 0.1882859617471695, -0.062335360795259476, 0.1278190314769745, 0.38913482427597046, 0.0037764159496873617, -0.1606176644563675, -0.01074370089918375, 0.02421225793659687, 0.09608051925897598, 0.23505258560180664, -0.14556263387203217, -0.16482320427894592, 0.0818943902850151, 0.2129843682050705, -0.04241016134619713, 0.11652717739343643, -0.3903499245643616, 0.04128958284854889, 0.21235674619674683, -0.23089076578617096, 0.06845791637897491, -0.2570658028125763, 0.1678502857685089, -0.09324300289154053, 0.02583426982164383, 0.16631819307804108, -0.000109738131868653, 0.0067979744635522366, -0.015387536957859993, -0.13856972754001617, -0.007274676114320755, 0.04019525647163391, 0.1181342601776123, -0.2117292284965515, -0.1091829240322113, -0.3121323585510254, -0.3320121765136719, -0.0376615971326828, 0.01780911348760128, -0.3556598424911499, 0.31285756826400757, 0.14745281636714935, 0.2607105076313019, 0.22490935027599335, -0.2538118362426758, 0.21113692224025726, 0.14512760937213898, -0.02932742051780224, -0.17419736087322235, -0.04230478033423424, -0.043978799134492874, -0.3344639539718628, -0.3767643868923187, 0.5254572629928589, -0.07658200711011887, 0.17014172673225403, 0.21466556191444397, 0.2250560075044632, 0.17260614037513733, 0.03593749552965164, 0.1429366171360016, -0.1139628142118454, 0.011095208115875721, -0.22691704332828522, -0.027116188779473305, -0.3733711242675781, -0.4186401069164276, -0.07852597534656525, 0.13762468099594116, 0.181045800447464, -0.1363930106163025, -0.06371636688709259, -0.43193674087524414, 0.23928946256637573, 0.06524856388568878, -0.012788441963493824, 0.1195276528596878, 0.2951611876487732, 0.04380587115883827, 0.19522327184677124, 0.12944640219211578, -0.06973284482955933, -0.02255813591182232, -0.21451735496520996, -0.05897236615419388, 0.22640195488929749, -0.43598562479019165, 0.07208240777254105, 0.43151915073394775, -0.14041432738304138, -0.053242821246385574, 0.030637066811323166, 0.3919302225112915, 0.07921385765075684, 0.11681852489709854, -0.008491439744830132, -0.1041610911488533, 0.5898671746253967, 0.044322144240140915, 0.1711844950914383, -0.014970467425882816, -0.1982778012752533, -0.35601741075515747, 0.0049875471740961075, 0.2880174517631531, -0.2866014838218689, 0.42841267585754395, 0.402360200881958, -0.04225970804691315, -0.12152931094169617, -0.03010164014995098, 0.0940236821770668, -0.3308537006378174, 0.05566929280757904, 0.058729760348796844, 0.31395789980888367, -0.2254507839679718, 0.14926527440547943, -0.17376787960529327, -0.10335493087768555, -0.38496968150138855, 0.12735134363174438, -0.33784639835357666, -0.033323660492897034, -0.4962180256843567, 0.001516752759926021, 0.2853839695453644, -0.0028605787083506584, 0.285352885723114, -0.03844567388296127, -0.13722026348114014, -0.30198997259140015, 0.10317455977201462, -0.3985126316547394, 0.5702307224273682, -0.10722927004098892, -0.19701692461967468, 0.24102142453193665, 0.13037806749343872, 0.08526397496461868, -0.2976876199245453, -0.19837690889835358, -0.16980776190757751, -0.20906922221183777, 0.25985991954803467, 0.24224543571472168, -0.07942714542150497, 0.14139650762081146, 0.09248153120279312, 0.07767834514379501, 0.34325864911079407, -0.23285464942455292, 0.10284031927585602, -0.20148888230323792, -0.28198033571243286, -0.20435941219329834, 0.1625692993402481, 0.26230981945991516, 0.06677709519863129, -0.1373722106218338, 0.40179747343063354, 0.23718252778053284, 0.14729131758213043, 0.4138065278530121, -0.1965337097644806, 0.30949482321739197, -0.4334923326969147, 0.3486187756061554, 0.5064778327941895, -0.32673120498657227, -0.49634432792663574, 0.09944815188646317, -0.15631528198719025, 0.10028795152902603, 0.10244916379451752, 0.15137308835983276, 0.02908792719244957, 0.01441967859864235, 0.12853601574897766, 0.018702292814850807, -0.0005142427398823202, 0.20842033624649048, -0.31692802906036377, -0.021933671087026596, 0.5500645637512207, -0.25246718525886536, 0.028472382575273514, -0.1312367022037506, -0.05504239350557327, 0.021932199597358704, 0.0067685069516301155, -0.1121898964047432, 0.1218591183423996, 0.11460506170988083, -0.05687858536839485, -0.15818439424037933, -0.10643845796585083, -0.02469293214380741, -0.28826266527175903, 0.23830117285251617, -0.14981962740421295, -0.09996888786554337, 0.061725009232759476, -0.09704943746328354, -0.1620030701160431, -0.12927453219890594, -0.1809171438217163, -0.27306488156318665, 0.20130188763141632, 0.23052427172660828, 0.32024043798446655, -0.07290686666965485, 0.22520951926708221, 0.36808186769485474, 0.04836326465010643, -0.21607212722301483, -0.4300347864627838, 0.08568006008863449, -0.17773543298244476, -0.21882851421833038, -0.07484656572341919, 0.2230692207813263, 0.20902669429779053, -0.04792690649628639, -0.1424286812543869, 0.1482759416103363, -0.222085103392601, 0.4189184010028839, -0.13702847063541412, 0.269452303647995, -0.06965663284063339, 0.02930617146193981, 0.01666727475821972, 0.2540055215358734, 0.058820389211177826, -0.08980126678943634, -0.07270590960979462, 0.4742339253425598, 0.18140152096748352, 0.17011570930480957, -0.04807102307677269, -0.21626847982406616, 0.05198436602950096, -0.17831741273403168, -0.12091562896966934, -0.1188163310289383, 0.1364070028066635, -0.08057195693254471, 0.14888811111450195, 0.15232902765274048, -0.10698776692152023, 0.3120073974132538, 0.7600834369659424, 0.1969190239906311, -0.011016630567610264, 0.015740377828478813, -0.05721341073513031, -0.2177916318178177, 0.3338569700717926, -0.05868323892354965, 0.3103885352611542, 0.10473150759935379, 0.12716469168663025, -0.04620937630534172, -0.2196495234966278, -0.15443581342697144, 0.12464867532253265, -0.21436883509159088, -0.19197611510753632, 0.30803436040878296, 0.28208714723587036, -0.15836739540100098, -0.35378298163414, 0.09401589632034302, -0.012764276005327702, 0.0261914674192667, -0.15800099074840546, 0.21723708510398865, -0.16652946174144745, -0.2780843675136566, -0.28641605377197266, 0.13615699112415314, -0.12220324575901031, -0.0921652764081955, -0.04126395657658577, 0.42542997002601624, -0.036653950810432434, 0.19515201449394226, -0.3149716258049011, 0.08951807022094727, 0.22660399973392487, -0.2663272023200989, -0.07168542593717575, -0.19028738141059875, -0.39492031931877136, 0.04223180189728737, -0.09523077309131622, -0.09299586713314056, 0.37832993268966675, -0.04731646925210953, -0.06795979291200638, -0.28814661502838135, -0.17254012823104858, 0.10717806965112686, -0.09261105209589005, -0.15536722540855408, 0.16724833846092224, -0.03133821487426758, -0.1011589840054512, -0.298745721578598, 0.2937973439693451, -0.3053131401538849, -0.2991430461406708, -0.0519869327545166, 0.09890599548816681, 0.030109098181128502, -0.2803472578525543, -0.42862468957901, 0.2735773026943207, -0.31724420189857483, 0.07546748220920563, -0.14892889559268951, 0.1862272024154663, 0.3309342861175537, 0.008954071439802647, -0.014561185613274574, -0.06404047459363937, 0.16439427435398102, -0.46740975975990295, -0.14095474779605865, 0.32257840037345886, 0.018760429695248604, -0.21630436182022095, -0.32459110021591187, -0.2757244408130646, 0.34772419929504395, 0.1923881620168686, -0.5803956985473633, -0.21442270278930664, -0.04890016093850136, 0.24803172051906586, -0.051259271800518036, -0.046495221555233, 0.5745418071746826, 0.019725333899259567, -0.17946326732635498, -0.16727812588214874, -0.33183929324150085, 0.35973072052001953, 0.41926220059394836, 0.24948781728744507, -0.011158456094563007, 0.1778150349855423, 0.20092009007930756, 0.6707469820976257, 0.20192037522792816, -0.2590579390525818, 0.3317311108112335, 0.10723166167736053, 0.07499769330024719, -0.20349697768688202, -0.31302309036254883, 0.18576005101203918, -0.1816031038761139, 0.01973230578005314, 0.27198368310928345, 0.006250523962080479, -0.2801840007305145, 0.017742546275258064, 0.2188185304403305, -0.2253447324037552, -0.3280579745769501, 0.2597457468509674, -0.3424278795719147, 0.24094857275485992, 0.03636978566646576, 0.31579962372779846, -0.5820139050483704, -0.2427240014076233, 0.1946277767419815, -0.3359457850456238, 0.27984464168548584, 0.04449823498725891, -0.5836009383201599, 0.1214703917503357, -0.5380211472511292, 0.18194159865379333, 0.31462812423706055, 0.27240779995918274, -0.09202467650175095, -0.24652515351772308, 0.04334753006696701, -0.14713141322135925, 0.57114177942276, 0.04097321629524231, 0.40857937932014465, 0.1699264943599701, -0.13024675846099854, -0.19947350025177002, -0.22471316158771515, 0.017235223203897476, 0.2801368832588196, 0.3265001177787781, 0.4008474349975586, -0.34524449706077576, -0.3632737994194031, -0.265765905380249, 0.007363766897469759, -0.0461815744638443, 0.007106977980583906, -0.12583452463150024, 0.18938513100147247, 0.016485881060361862, 0.17434586584568024, -0.01498030498623848, 0.11830329149961472, 0.016039730980992317, -0.13654667139053345, -0.03913423791527748, -0.2194267064332962, 0.1282399445772171, 0.22643792629241943, 0.47359511256217957, 0.15032468736171722, 0.03246146813035011, 0.03321428596973419, 0.21515770256519318, -0.5671080350875854, 0.2659573256969452, -0.08884955942630768, -0.004131757188588381, 0.0746438056230545, 0.14443759620189667, -0.030301183462142944, 0.33789676427841187, -0.042525336146354675, 0.1457137167453766, -0.12862981855869293, 0.008957206271588802, -0.5539681911468506, 0.4854546785354614, 0.387459397315979, 0.13246704638004303, -0.0448889285326004, -0.24685737490653992, 0.1734629124403, 0.07412078231573105, -0.27653083205223083, 0.1322595775127411, -0.43069255352020264, -0.43199506402015686, 0.18579724431037903, 0.20948676764965057, 0.9232465028762817, 0.332022100687027, 0.37195777893066406, 0.10070342570543289, 0.103580541908741, 0.2642515301704407, -0.23411165177822113, 0.4092499911785126, -0.3693643808364868, -0.046317894011735916, 0.024538729339838028, -0.16673347353935242, -0.22738294303417206, -0.010426540859043598, -0.010993897914886475, 0.30259546637535095, 0.32404956221580505, 0.5999361276626587, -0.11212006956338882, 0.16670002043247223, 0.21569199860095978, -0.15810944139957428, -0.24923092126846313, 0.2010732889175415, -0.29230424761772156, 0.3152405619621277, -0.11240841448307037, -0.07410834729671478, -0.016508914530277252, -0.3513421416282654, 0.06714001297950745, 0.06094168499112129, -0.33651697635650635, 0.30484193563461304, 0.1334618479013443, -0.2726665437221527, -0.2969878315925598, 0.2550783157348633, 0.11198701709508896, 0.08252796530723572, -0.04859941452741623, -0.035185184329748154, 0.4006448984146118, 0.3504580557346344, -0.006891822908073664, -0.20158866047859192, 0.1864124983549118, 0.06387507170438766, -0.01289470586925745, 0.030242804437875748, -0.05640558898448944, -0.30158281326293945, -0.2525118589401245, 0.14820460975170135, 0.003372971899807453, -0.4129619300365448, -0.055318888276815414, 0.06144449859857559, -0.06678875535726547, -0.09413730353116989, 0.15088044106960297, -0.02606821060180664, -0.10794124007225037, 0.31515446305274963, 0.11733822524547577, -0.24304218590259552, 0.05235825479030609, 0.47483208775520325, -0.04039623588323593, 0.031818028539419174, 0.41542848944664, -0.15281835198402405, -0.10380636155605316, -0.2553911507129669, -0.2841723561286926, -0.10206018388271332, -0.4488404393196106, 0.11220814287662506, -0.2194056361913681, -0.40191105008125305, 0.044232070446014404, 0.28081876039505005, 0.014706442132592201, 0.31849661469459534, -0.011618257500231266, -0.12495818734169006, -0.3769727349281311, -0.10379086434841156, -0.10658226162195206, 0.26717832684516907, 0.2795191705226898, 0.284347802400589, -0.29586678743362427, 0.24922697246074677, -0.38310956954956055, -0.10714221000671387, -0.18717068433761597, 0.43791306018829346, -0.039137572050094604, -0.36558258533477783, -0.009147626347839832, 0.0018550800159573555, 0.1299709975719452, 0.17754508554935455, -0.6386584043502808, -0.29261913895606995, -0.0909920185804367, 0.0818607434630394, -0.007587989326566458, -0.15982191264629364, -0.23928239941596985, -0.017214275896549225, 0.18238499760627747, -0.42740264534950256, 0.09215671569108963, 0.15121649205684662, 0.04703198000788689, 0.2371872216463089, -0.1610299050807953, 0.12145564705133438, 0.28816723823547363, -0.1865275353193283, -0.18414904177188873, -0.07622885704040527, 0.08048073947429657, 0.25701698660850525, -0.06946761906147003, 0.02047896571457386, -0.21040348708629608, 0.03047192096710205, 0.08927451074123383, 0.1427948772907257, 0.2369290292263031, -0.024752648547291756, -0.21566133201122284, 0.0005506349843926728, 0.18279911577701569, 0.11411940306425095, -0.07855328917503357, -0.32897576689720154, 0.28022339940071106, 0.2783651351928711, -0.11751823127269745, -0.010237314738333225, 0.16770750284194946, -0.0993887260556221, 0.2528513967990875, 0.20715337991714478, 0.10758770257234573, 0.05261155217885971, 0.2696119546890259, 0.1264941692352295, 0.4050402343273163, -0.3483290374279022, 0.08942107856273651, 0.06441547721624374, 0.10164907574653625, 0.037778209894895554, 0.3812154531478882, -0.15324409306049347, 0.08992163836956024, 0.14838311076164246, 0.18833674490451813, 0.20253996551036835, 0.2628902196884155, 0.21354077756404877, -0.18786455690860748, -0.03412069007754326, -0.24108663201332092, 0.1436367630958557, -0.1427268236875534, 0.41565975546836853, 0.15059226751327515, 0.19285550713539124, -0.11485161632299423, -0.46193909645080566, -0.24626745283603668, 0.23536424338817596, -0.324541836977005, -0.3092094957828522, -0.04867883026599884, -0.0801846906542778, 0.055929236114025116, 0.19456082582473755, -0.07928955554962158, 0.20385949313640594, 0.3428383469581604, -0.008659737184643745, -0.013750121928751469, -0.08435364067554474, -0.17061713337898254, 0.11397942155599594, 0.2748646140098572, -0.0817389041185379, 0.446936696767807, 0.010801088996231556, -0.05156935378909111, -0.23447194695472717, 0.41141754388809204, 0.28612878918647766, 0.19494746625423431, -0.16886235773563385, 0.0639532282948494, 0.20064730942249298, 0.03061334788799286, -0.3546886742115021, 0.32865145802497864, 0.151144340634346, -0.29302841424942017, 0.36800533533096313, 0.1237955093383789, -0.2321559190750122, 0.2764036953449249, -0.10419564694166183, -0.20146705210208893, -0.04988894239068031, 0.3224369287490845, 0.09970726817846298, -0.039932768791913986, -0.1949632614850998, 0.11112037301063538, -0.144571453332901, -0.28533878922462463, 0.15934637188911438, -0.023898735642433167, 0.46766093373298645, -0.27567628026008606, 0.11720526963472366, -0.06182510778307915, 0.6220861673355103, -0.029227983206510544, 0.1243293285369873, -0.3501380383968353, 0.10264355689287186, -0.33316680788993835, 0.19841012358665466, -0.08381723612546921, 0.15374304354190826, -0.3975839614868164, 0.20426513254642487, -0.11549928039312363, -0.023466292768716812, -0.11804226040840149, -0.007886320352554321, 0.1474074274301529, 0.17893218994140625, -0.38691800832748413, 0.15548168122768402, -0.08625979721546173, -0.0062670656479895115, 0.29177364706993103, -0.3766753375530243, 0.2219911813735962, 0.04247850924730301, 0.20691059529781342, 0.01098591648042202, -0.0217235516756773, -0.05480630695819855, 0.3365738093852997, 0.6232410669326782, 0.3470291495323181, 0.14255070686340332, -0.10102760046720505, 0.08155444264411926, -0.3474372327327728, 0.014104297384619713, -0.06088805943727493, -0.20964989066123962, 0.09756883978843689, 0.525830864906311, -0.2256278097629547, 0.13194623589515686, -0.17804619669914246, 0.23669403791427612, 0.38227635622024536, -0.15272943675518036, 0.06180020794272423, 0.04822545871138573, -0.259840190410614, -0.0051995194517076015, -0.12586642801761627, 0.4537962079048157, 0.007182563189417124, 0.32383066415786743, -0.35929134488105774, -0.5417627096176147, 0.4329098165035248, -0.45112600922584534, -0.3025253415107727, 0.12327741086483002, 0.23878581821918488, 0.34956854581832886, -0.06000017747282982, -0.5436674952507019, -0.17249032855033875, 0.2691141366958618, -0.030661141499876976, -0.15704505145549774, 0.20768523216247559, -0.14864318072795868, -0.04107455909252167, -0.11675450205802917, 0.17646948993206024, 0.10152691602706909, -0.2914120852947235, 0.24369898438453674, -0.37010377645492554 ]
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.
[ -0.1263234168291092, -0.22070905566215515, -0.10063228756189346, 0.2780112326145172, 0.24547505378723145, -0.20522351562976837, 0.1372111439704895, 0.5128591656684875, 0.1882859617471695, -0.062335360795259476, 0.1278190314769745, 0.38913482427597046, 0.0037764159496873617, -0.1606176644563675, -0.01074370089918375, 0.02421225793659687, 0.09608051925897598, 0.23505258560180664, -0.14556263387203217, -0.16482320427894592, 0.0818943902850151, 0.2129843682050705, -0.04241016134619713, 0.11652717739343643, -0.3903499245643616, 0.04128958284854889, 0.21235674619674683, -0.23089076578617096, 0.06845791637897491, -0.2570658028125763, 0.1678502857685089, -0.09324300289154053, 0.02583426982164383, 0.16631819307804108, -0.000109738131868653, 0.0067979744635522366, -0.015387536957859993, -0.13856972754001617, -0.007274676114320755, 0.04019525647163391, 0.1181342601776123, -0.2117292284965515, -0.1091829240322113, -0.3121323585510254, -0.3320121765136719, -0.0376615971326828, 0.01780911348760128, -0.3556598424911499, 0.31285756826400757, 0.14745281636714935, 0.2607105076313019, 0.22490935027599335, -0.2538118362426758, 0.21113692224025726, 0.14512760937213898, -0.02932742051780224, -0.17419736087322235, -0.04230478033423424, -0.043978799134492874, -0.3344639539718628, -0.3767643868923187, 0.5254572629928589, -0.07658200711011887, 0.17014172673225403, 0.21466556191444397, 0.2250560075044632, 0.17260614037513733, 0.03593749552965164, 0.1429366171360016, -0.1139628142118454, 0.011095208115875721, -0.22691704332828522, -0.027116188779473305, -0.3733711242675781, -0.4186401069164276, -0.07852597534656525, 0.13762468099594116, 0.181045800447464, -0.1363930106163025, -0.06371636688709259, -0.43193674087524414, 0.23928946256637573, 0.06524856388568878, -0.012788441963493824, 0.1195276528596878, 0.2951611876487732, 0.04380587115883827, 0.19522327184677124, 0.12944640219211578, -0.06973284482955933, -0.02255813591182232, -0.21451735496520996, -0.05897236615419388, 0.22640195488929749, -0.43598562479019165, 0.07208240777254105, 0.43151915073394775, -0.14041432738304138, -0.053242821246385574, 0.030637066811323166, 0.3919302225112915, 0.07921385765075684, 0.11681852489709854, -0.008491439744830132, -0.1041610911488533, 0.5898671746253967, 0.044322144240140915, 0.1711844950914383, -0.014970467425882816, -0.1982778012752533, -0.35601741075515747, 0.0049875471740961075, 0.2880174517631531, -0.2866014838218689, 0.42841267585754395, 0.402360200881958, -0.04225970804691315, -0.12152931094169617, -0.03010164014995098, 0.0940236821770668, -0.3308537006378174, 0.05566929280757904, 0.058729760348796844, 0.31395789980888367, -0.2254507839679718, 0.14926527440547943, -0.17376787960529327, -0.10335493087768555, -0.38496968150138855, 0.12735134363174438, -0.33784639835357666, -0.033323660492897034, -0.4962180256843567, 0.001516752759926021, 0.2853839695453644, -0.0028605787083506584, 0.285352885723114, -0.03844567388296127, -0.13722026348114014, -0.30198997259140015, 0.10317455977201462, -0.3985126316547394, 0.5702307224273682, -0.10722927004098892, -0.19701692461967468, 0.24102142453193665, 0.13037806749343872, 0.08526397496461868, -0.2976876199245453, -0.19837690889835358, -0.16980776190757751, -0.20906922221183777, 0.25985991954803467, 0.24224543571472168, -0.07942714542150497, 0.14139650762081146, 0.09248153120279312, 0.07767834514379501, 0.34325864911079407, -0.23285464942455292, 0.10284031927585602, -0.20148888230323792, -0.28198033571243286, -0.20435941219329834, 0.1625692993402481, 0.26230981945991516, 0.06677709519863129, -0.1373722106218338, 0.40179747343063354, 0.23718252778053284, 0.14729131758213043, 0.4138065278530121, -0.1965337097644806, 0.30949482321739197, -0.4334923326969147, 0.3486187756061554, 0.5064778327941895, -0.32673120498657227, -0.49634432792663574, 0.09944815188646317, -0.15631528198719025, 0.10028795152902603, 0.10244916379451752, 0.15137308835983276, 0.02908792719244957, 0.01441967859864235, 0.12853601574897766, 0.018702292814850807, -0.0005142427398823202, 0.20842033624649048, -0.31692802906036377, -0.021933671087026596, 0.5500645637512207, -0.25246718525886536, 0.028472382575273514, -0.1312367022037506, -0.05504239350557327, 0.021932199597358704, 0.0067685069516301155, -0.1121898964047432, 0.1218591183423996, 0.11460506170988083, -0.05687858536839485, -0.15818439424037933, -0.10643845796585083, -0.02469293214380741, -0.28826266527175903, 0.23830117285251617, -0.14981962740421295, -0.09996888786554337, 0.061725009232759476, -0.09704943746328354, -0.1620030701160431, -0.12927453219890594, -0.1809171438217163, -0.27306488156318665, 0.20130188763141632, 0.23052427172660828, 0.32024043798446655, -0.07290686666965485, 0.22520951926708221, 0.36808186769485474, 0.04836326465010643, -0.21607212722301483, -0.4300347864627838, 0.08568006008863449, -0.17773543298244476, -0.21882851421833038, -0.07484656572341919, 0.2230692207813263, 0.20902669429779053, -0.04792690649628639, -0.1424286812543869, 0.1482759416103363, -0.222085103392601, 0.4189184010028839, -0.13702847063541412, 0.269452303647995, -0.06965663284063339, 0.02930617146193981, 0.01666727475821972, 0.2540055215358734, 0.058820389211177826, -0.08980126678943634, -0.07270590960979462, 0.4742339253425598, 0.18140152096748352, 0.17011570930480957, -0.04807102307677269, -0.21626847982406616, 0.05198436602950096, -0.17831741273403168, -0.12091562896966934, -0.1188163310289383, 0.1364070028066635, -0.08057195693254471, 0.14888811111450195, 0.15232902765274048, -0.10698776692152023, 0.3120073974132538, 0.7600834369659424, 0.1969190239906311, -0.011016630567610264, 0.015740377828478813, -0.05721341073513031, -0.2177916318178177, 0.3338569700717926, -0.05868323892354965, 0.3103885352611542, 0.10473150759935379, 0.12716469168663025, -0.04620937630534172, -0.2196495234966278, -0.15443581342697144, 0.12464867532253265, -0.21436883509159088, -0.19197611510753632, 0.30803436040878296, 0.28208714723587036, -0.15836739540100098, -0.35378298163414, 0.09401589632034302, -0.012764276005327702, 0.0261914674192667, -0.15800099074840546, 0.21723708510398865, -0.16652946174144745, -0.2780843675136566, -0.28641605377197266, 0.13615699112415314, -0.12220324575901031, -0.0921652764081955, -0.04126395657658577, 0.42542997002601624, -0.036653950810432434, 0.19515201449394226, -0.3149716258049011, 0.08951807022094727, 0.22660399973392487, -0.2663272023200989, -0.07168542593717575, -0.19028738141059875, -0.39492031931877136, 0.04223180189728737, -0.09523077309131622, -0.09299586713314056, 0.37832993268966675, -0.04731646925210953, -0.06795979291200638, -0.28814661502838135, -0.17254012823104858, 0.10717806965112686, -0.09261105209589005, -0.15536722540855408, 0.16724833846092224, -0.03133821487426758, -0.1011589840054512, -0.298745721578598, 0.2937973439693451, -0.3053131401538849, -0.2991430461406708, -0.0519869327545166, 0.09890599548816681, 0.030109098181128502, -0.2803472578525543, -0.42862468957901, 0.2735773026943207, -0.31724420189857483, 0.07546748220920563, -0.14892889559268951, 0.1862272024154663, 0.3309342861175537, 0.008954071439802647, -0.014561185613274574, -0.06404047459363937, 0.16439427435398102, -0.46740975975990295, -0.14095474779605865, 0.32257840037345886, 0.018760429695248604, -0.21630436182022095, -0.32459110021591187, -0.2757244408130646, 0.34772419929504395, 0.1923881620168686, -0.5803956985473633, -0.21442270278930664, -0.04890016093850136, 0.24803172051906586, -0.051259271800518036, -0.046495221555233, 0.5745418071746826, 0.019725333899259567, -0.17946326732635498, -0.16727812588214874, -0.33183929324150085, 0.35973072052001953, 0.41926220059394836, 0.24948781728744507, -0.011158456094563007, 0.1778150349855423, 0.20092009007930756, 0.6707469820976257, 0.20192037522792816, -0.2590579390525818, 0.3317311108112335, 0.10723166167736053, 0.07499769330024719, -0.20349697768688202, -0.31302309036254883, 0.18576005101203918, -0.1816031038761139, 0.01973230578005314, 0.27198368310928345, 0.006250523962080479, -0.2801840007305145, 0.017742546275258064, 0.2188185304403305, -0.2253447324037552, -0.3280579745769501, 0.2597457468509674, -0.3424278795719147, 0.24094857275485992, 0.03636978566646576, 0.31579962372779846, -0.5820139050483704, -0.2427240014076233, 0.1946277767419815, -0.3359457850456238, 0.27984464168548584, 0.04449823498725891, -0.5836009383201599, 0.1214703917503357, -0.5380211472511292, 0.18194159865379333, 0.31462812423706055, 0.27240779995918274, -0.09202467650175095, -0.24652515351772308, 0.04334753006696701, -0.14713141322135925, 0.57114177942276, 0.04097321629524231, 0.40857937932014465, 0.1699264943599701, -0.13024675846099854, -0.19947350025177002, -0.22471316158771515, 0.017235223203897476, 0.2801368832588196, 0.3265001177787781, 0.4008474349975586, -0.34524449706077576, -0.3632737994194031, -0.265765905380249, 0.007363766897469759, -0.0461815744638443, 0.007106977980583906, -0.12583452463150024, 0.18938513100147247, 0.016485881060361862, 0.17434586584568024, -0.01498030498623848, 0.11830329149961472, 0.016039730980992317, -0.13654667139053345, -0.03913423791527748, -0.2194267064332962, 0.1282399445772171, 0.22643792629241943, 0.47359511256217957, 0.15032468736171722, 0.03246146813035011, 0.03321428596973419, 0.21515770256519318, -0.5671080350875854, 0.2659573256969452, -0.08884955942630768, -0.004131757188588381, 0.0746438056230545, 0.14443759620189667, -0.030301183462142944, 0.33789676427841187, -0.042525336146354675, 0.1457137167453766, -0.12862981855869293, 0.008957206271588802, -0.5539681911468506, 0.4854546785354614, 0.387459397315979, 0.13246704638004303, -0.0448889285326004, -0.24685737490653992, 0.1734629124403, 0.07412078231573105, -0.27653083205223083, 0.1322595775127411, -0.43069255352020264, -0.43199506402015686, 0.18579724431037903, 0.20948676764965057, 0.9232465028762817, 0.332022100687027, 0.37195777893066406, 0.10070342570543289, 0.103580541908741, 0.2642515301704407, -0.23411165177822113, 0.4092499911785126, -0.3693643808364868, -0.046317894011735916, 0.024538729339838028, -0.16673347353935242, -0.22738294303417206, -0.010426540859043598, -0.010993897914886475, 0.30259546637535095, 0.32404956221580505, 0.5999361276626587, -0.11212006956338882, 0.16670002043247223, 0.21569199860095978, -0.15810944139957428, -0.24923092126846313, 0.2010732889175415, -0.29230424761772156, 0.3152405619621277, -0.11240841448307037, -0.07410834729671478, -0.016508914530277252, -0.3513421416282654, 0.06714001297950745, 0.06094168499112129, -0.33651697635650635, 0.30484193563461304, 0.1334618479013443, -0.2726665437221527, -0.2969878315925598, 0.2550783157348633, 0.11198701709508896, 0.08252796530723572, -0.04859941452741623, -0.035185184329748154, 0.4006448984146118, 0.3504580557346344, -0.006891822908073664, -0.20158866047859192, 0.1864124983549118, 0.06387507170438766, -0.01289470586925745, 0.030242804437875748, -0.05640558898448944, -0.30158281326293945, -0.2525118589401245, 0.14820460975170135, 0.003372971899807453, -0.4129619300365448, -0.055318888276815414, 0.06144449859857559, -0.06678875535726547, -0.09413730353116989, 0.15088044106960297, -0.02606821060180664, -0.10794124007225037, 0.31515446305274963, 0.11733822524547577, -0.24304218590259552, 0.05235825479030609, 0.47483208775520325, -0.04039623588323593, 0.031818028539419174, 0.41542848944664, -0.15281835198402405, -0.10380636155605316, -0.2553911507129669, -0.2841723561286926, -0.10206018388271332, -0.4488404393196106, 0.11220814287662506, -0.2194056361913681, -0.40191105008125305, 0.044232070446014404, 0.28081876039505005, 0.014706442132592201, 0.31849661469459534, -0.011618257500231266, -0.12495818734169006, -0.3769727349281311, -0.10379086434841156, -0.10658226162195206, 0.26717832684516907, 0.2795191705226898, 0.284347802400589, -0.29586678743362427, 0.24922697246074677, -0.38310956954956055, -0.10714221000671387, -0.18717068433761597, 0.43791306018829346, -0.039137572050094604, -0.36558258533477783, -0.009147626347839832, 0.0018550800159573555, 0.1299709975719452, 0.17754508554935455, -0.6386584043502808, -0.29261913895606995, -0.0909920185804367, 0.0818607434630394, -0.007587989326566458, -0.15982191264629364, -0.23928239941596985, -0.017214275896549225, 0.18238499760627747, -0.42740264534950256, 0.09215671569108963, 0.15121649205684662, 0.04703198000788689, 0.2371872216463089, -0.1610299050807953, 0.12145564705133438, 0.28816723823547363, -0.1865275353193283, -0.18414904177188873, -0.07622885704040527, 0.08048073947429657, 0.25701698660850525, -0.06946761906147003, 0.02047896571457386, -0.21040348708629608, 0.03047192096710205, 0.08927451074123383, 0.1427948772907257, 0.2369290292263031, -0.024752648547291756, -0.21566133201122284, 0.0005506349843926728, 0.18279911577701569, 0.11411940306425095, -0.07855328917503357, -0.32897576689720154, 0.28022339940071106, 0.2783651351928711, -0.11751823127269745, -0.010237314738333225, 0.16770750284194946, -0.0993887260556221, 0.2528513967990875, 0.20715337991714478, 0.10758770257234573, 0.05261155217885971, 0.2696119546890259, 0.1264941692352295, 0.4050402343273163, -0.3483290374279022, 0.08942107856273651, 0.06441547721624374, 0.10164907574653625, 0.037778209894895554, 0.3812154531478882, -0.15324409306049347, 0.08992163836956024, 0.14838311076164246, 0.18833674490451813, 0.20253996551036835, 0.2628902196884155, 0.21354077756404877, -0.18786455690860748, -0.03412069007754326, -0.24108663201332092, 0.1436367630958557, -0.1427268236875534, 0.41565975546836853, 0.15059226751327515, 0.19285550713539124, -0.11485161632299423, -0.46193909645080566, -0.24626745283603668, 0.23536424338817596, -0.324541836977005, -0.3092094957828522, -0.04867883026599884, -0.0801846906542778, 0.055929236114025116, 0.19456082582473755, -0.07928955554962158, 0.20385949313640594, 0.3428383469581604, -0.008659737184643745, -0.013750121928751469, -0.08435364067554474, -0.17061713337898254, 0.11397942155599594, 0.2748646140098572, -0.0817389041185379, 0.446936696767807, 0.010801088996231556, -0.05156935378909111, -0.23447194695472717, 0.41141754388809204, 0.28612878918647766, 0.19494746625423431, -0.16886235773563385, 0.0639532282948494, 0.20064730942249298, 0.03061334788799286, -0.3546886742115021, 0.32865145802497864, 0.151144340634346, -0.29302841424942017, 0.36800533533096313, 0.1237955093383789, -0.2321559190750122, 0.2764036953449249, -0.10419564694166183, -0.20146705210208893, -0.04988894239068031, 0.3224369287490845, 0.09970726817846298, -0.039932768791913986, -0.1949632614850998, 0.11112037301063538, -0.144571453332901, -0.28533878922462463, 0.15934637188911438, -0.023898735642433167, 0.46766093373298645, -0.27567628026008606, 0.11720526963472366, -0.06182510778307915, 0.6220861673355103, -0.029227983206510544, 0.1243293285369873, -0.3501380383968353, 0.10264355689287186, -0.33316680788993835, 0.19841012358665466, -0.08381723612546921, 0.15374304354190826, -0.3975839614868164, 0.20426513254642487, -0.11549928039312363, -0.023466292768716812, -0.11804226040840149, -0.007886320352554321, 0.1474074274301529, 0.17893218994140625, -0.38691800832748413, 0.15548168122768402, -0.08625979721546173, -0.0062670656479895115, 0.29177364706993103, -0.3766753375530243, 0.2219911813735962, 0.04247850924730301, 0.20691059529781342, 0.01098591648042202, -0.0217235516756773, -0.05480630695819855, 0.3365738093852997, 0.6232410669326782, 0.3470291495323181, 0.14255070686340332, -0.10102760046720505, 0.08155444264411926, -0.3474372327327728, 0.014104297384619713, -0.06088805943727493, -0.20964989066123962, 0.09756883978843689, 0.525830864906311, -0.2256278097629547, 0.13194623589515686, -0.17804619669914246, 0.23669403791427612, 0.38227635622024536, -0.15272943675518036, 0.06180020794272423, 0.04822545871138573, -0.259840190410614, -0.0051995194517076015, -0.12586642801761627, 0.4537962079048157, 0.007182563189417124, 0.32383066415786743, -0.35929134488105774, -0.5417627096176147, 0.4329098165035248, -0.45112600922584534, -0.3025253415107727, 0.12327741086483002, 0.23878581821918488, 0.34956854581832886, -0.06000017747282982, -0.5436674952507019, -0.17249032855033875, 0.2691141366958618, -0.030661141499876976, -0.15704505145549774, 0.20768523216247559, -0.14864318072795868, -0.04107455909252167, -0.11675450205802917, 0.17646948993206024, 0.10152691602706909, -0.2914120852947235, 0.24369898438453674, -0.37010377645492554 ]
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
[ -0.1263234168291092, -0.22070905566215515, -0.10063228756189346, 0.2780112326145172, 0.24547505378723145, -0.20522351562976837, 0.1372111439704895, 0.5128591656684875, 0.1882859617471695, -0.062335360795259476, 0.1278190314769745, 0.38913482427597046, 0.0037764159496873617, -0.1606176644563675, -0.01074370089918375, 0.02421225793659687, 0.09608051925897598, 0.23505258560180664, -0.14556263387203217, -0.16482320427894592, 0.0818943902850151, 0.2129843682050705, -0.04241016134619713, 0.11652717739343643, -0.3903499245643616, 0.04128958284854889, 0.21235674619674683, -0.23089076578617096, 0.06845791637897491, -0.2570658028125763, 0.1678502857685089, -0.09324300289154053, 0.02583426982164383, 0.16631819307804108, -0.000109738131868653, 0.0067979744635522366, -0.015387536957859993, -0.13856972754001617, -0.007274676114320755, 0.04019525647163391, 0.1181342601776123, -0.2117292284965515, -0.1091829240322113, -0.3121323585510254, -0.3320121765136719, -0.0376615971326828, 0.01780911348760128, -0.3556598424911499, 0.31285756826400757, 0.14745281636714935, 0.2607105076313019, 0.22490935027599335, -0.2538118362426758, 0.21113692224025726, 0.14512760937213898, -0.02932742051780224, -0.17419736087322235, -0.04230478033423424, -0.043978799134492874, -0.3344639539718628, -0.3767643868923187, 0.5254572629928589, -0.07658200711011887, 0.17014172673225403, 0.21466556191444397, 0.2250560075044632, 0.17260614037513733, 0.03593749552965164, 0.1429366171360016, -0.1139628142118454, 0.011095208115875721, -0.22691704332828522, -0.027116188779473305, -0.3733711242675781, -0.4186401069164276, -0.07852597534656525, 0.13762468099594116, 0.181045800447464, -0.1363930106163025, -0.06371636688709259, -0.43193674087524414, 0.23928946256637573, 0.06524856388568878, -0.012788441963493824, 0.1195276528596878, 0.2951611876487732, 0.04380587115883827, 0.19522327184677124, 0.12944640219211578, -0.06973284482955933, -0.02255813591182232, -0.21451735496520996, -0.05897236615419388, 0.22640195488929749, -0.43598562479019165, 0.07208240777254105, 0.43151915073394775, -0.14041432738304138, -0.053242821246385574, 0.030637066811323166, 0.3919302225112915, 0.07921385765075684, 0.11681852489709854, -0.008491439744830132, -0.1041610911488533, 0.5898671746253967, 0.044322144240140915, 0.1711844950914383, -0.014970467425882816, -0.1982778012752533, -0.35601741075515747, 0.0049875471740961075, 0.2880174517631531, -0.2866014838218689, 0.42841267585754395, 0.402360200881958, -0.04225970804691315, -0.12152931094169617, -0.03010164014995098, 0.0940236821770668, -0.3308537006378174, 0.05566929280757904, 0.058729760348796844, 0.31395789980888367, -0.2254507839679718, 0.14926527440547943, -0.17376787960529327, -0.10335493087768555, -0.38496968150138855, 0.12735134363174438, -0.33784639835357666, -0.033323660492897034, -0.4962180256843567, 0.001516752759926021, 0.2853839695453644, -0.0028605787083506584, 0.285352885723114, -0.03844567388296127, -0.13722026348114014, -0.30198997259140015, 0.10317455977201462, -0.3985126316547394, 0.5702307224273682, -0.10722927004098892, -0.19701692461967468, 0.24102142453193665, 0.13037806749343872, 0.08526397496461868, -0.2976876199245453, -0.19837690889835358, -0.16980776190757751, -0.20906922221183777, 0.25985991954803467, 0.24224543571472168, -0.07942714542150497, 0.14139650762081146, 0.09248153120279312, 0.07767834514379501, 0.34325864911079407, -0.23285464942455292, 0.10284031927585602, -0.20148888230323792, -0.28198033571243286, -0.20435941219329834, 0.1625692993402481, 0.26230981945991516, 0.06677709519863129, -0.1373722106218338, 0.40179747343063354, 0.23718252778053284, 0.14729131758213043, 0.4138065278530121, -0.1965337097644806, 0.30949482321739197, -0.4334923326969147, 0.3486187756061554, 0.5064778327941895, -0.32673120498657227, -0.49634432792663574, 0.09944815188646317, -0.15631528198719025, 0.10028795152902603, 0.10244916379451752, 0.15137308835983276, 0.02908792719244957, 0.01441967859864235, 0.12853601574897766, 0.018702292814850807, -0.0005142427398823202, 0.20842033624649048, -0.31692802906036377, -0.021933671087026596, 0.5500645637512207, -0.25246718525886536, 0.028472382575273514, -0.1312367022037506, -0.05504239350557327, 0.021932199597358704, 0.0067685069516301155, -0.1121898964047432, 0.1218591183423996, 0.11460506170988083, -0.05687858536839485, -0.15818439424037933, -0.10643845796585083, -0.02469293214380741, -0.28826266527175903, 0.23830117285251617, -0.14981962740421295, -0.09996888786554337, 0.061725009232759476, -0.09704943746328354, -0.1620030701160431, -0.12927453219890594, -0.1809171438217163, -0.27306488156318665, 0.20130188763141632, 0.23052427172660828, 0.32024043798446655, -0.07290686666965485, 0.22520951926708221, 0.36808186769485474, 0.04836326465010643, -0.21607212722301483, -0.4300347864627838, 0.08568006008863449, -0.17773543298244476, -0.21882851421833038, -0.07484656572341919, 0.2230692207813263, 0.20902669429779053, -0.04792690649628639, -0.1424286812543869, 0.1482759416103363, -0.222085103392601, 0.4189184010028839, -0.13702847063541412, 0.269452303647995, -0.06965663284063339, 0.02930617146193981, 0.01666727475821972, 0.2540055215358734, 0.058820389211177826, -0.08980126678943634, -0.07270590960979462, 0.4742339253425598, 0.18140152096748352, 0.17011570930480957, -0.04807102307677269, -0.21626847982406616, 0.05198436602950096, -0.17831741273403168, -0.12091562896966934, -0.1188163310289383, 0.1364070028066635, -0.08057195693254471, 0.14888811111450195, 0.15232902765274048, -0.10698776692152023, 0.3120073974132538, 0.7600834369659424, 0.1969190239906311, -0.011016630567610264, 0.015740377828478813, -0.05721341073513031, -0.2177916318178177, 0.3338569700717926, -0.05868323892354965, 0.3103885352611542, 0.10473150759935379, 0.12716469168663025, -0.04620937630534172, -0.2196495234966278, -0.15443581342697144, 0.12464867532253265, -0.21436883509159088, -0.19197611510753632, 0.30803436040878296, 0.28208714723587036, -0.15836739540100098, -0.35378298163414, 0.09401589632034302, -0.012764276005327702, 0.0261914674192667, -0.15800099074840546, 0.21723708510398865, -0.16652946174144745, -0.2780843675136566, -0.28641605377197266, 0.13615699112415314, -0.12220324575901031, -0.0921652764081955, -0.04126395657658577, 0.42542997002601624, -0.036653950810432434, 0.19515201449394226, -0.3149716258049011, 0.08951807022094727, 0.22660399973392487, -0.2663272023200989, -0.07168542593717575, -0.19028738141059875, -0.39492031931877136, 0.04223180189728737, -0.09523077309131622, -0.09299586713314056, 0.37832993268966675, -0.04731646925210953, -0.06795979291200638, -0.28814661502838135, -0.17254012823104858, 0.10717806965112686, -0.09261105209589005, -0.15536722540855408, 0.16724833846092224, -0.03133821487426758, -0.1011589840054512, -0.298745721578598, 0.2937973439693451, -0.3053131401538849, -0.2991430461406708, -0.0519869327545166, 0.09890599548816681, 0.030109098181128502, -0.2803472578525543, -0.42862468957901, 0.2735773026943207, -0.31724420189857483, 0.07546748220920563, -0.14892889559268951, 0.1862272024154663, 0.3309342861175537, 0.008954071439802647, -0.014561185613274574, -0.06404047459363937, 0.16439427435398102, -0.46740975975990295, -0.14095474779605865, 0.32257840037345886, 0.018760429695248604, -0.21630436182022095, -0.32459110021591187, -0.2757244408130646, 0.34772419929504395, 0.1923881620168686, -0.5803956985473633, -0.21442270278930664, -0.04890016093850136, 0.24803172051906586, -0.051259271800518036, -0.046495221555233, 0.5745418071746826, 0.019725333899259567, -0.17946326732635498, -0.16727812588214874, -0.33183929324150085, 0.35973072052001953, 0.41926220059394836, 0.24948781728744507, -0.011158456094563007, 0.1778150349855423, 0.20092009007930756, 0.6707469820976257, 0.20192037522792816, -0.2590579390525818, 0.3317311108112335, 0.10723166167736053, 0.07499769330024719, -0.20349697768688202, -0.31302309036254883, 0.18576005101203918, -0.1816031038761139, 0.01973230578005314, 0.27198368310928345, 0.006250523962080479, -0.2801840007305145, 0.017742546275258064, 0.2188185304403305, -0.2253447324037552, -0.3280579745769501, 0.2597457468509674, -0.3424278795719147, 0.24094857275485992, 0.03636978566646576, 0.31579962372779846, -0.5820139050483704, -0.2427240014076233, 0.1946277767419815, -0.3359457850456238, 0.27984464168548584, 0.04449823498725891, -0.5836009383201599, 0.1214703917503357, -0.5380211472511292, 0.18194159865379333, 0.31462812423706055, 0.27240779995918274, -0.09202467650175095, -0.24652515351772308, 0.04334753006696701, -0.14713141322135925, 0.57114177942276, 0.04097321629524231, 0.40857937932014465, 0.1699264943599701, -0.13024675846099854, -0.19947350025177002, -0.22471316158771515, 0.017235223203897476, 0.2801368832588196, 0.3265001177787781, 0.4008474349975586, -0.34524449706077576, -0.3632737994194031, -0.265765905380249, 0.007363766897469759, -0.0461815744638443, 0.007106977980583906, -0.12583452463150024, 0.18938513100147247, 0.016485881060361862, 0.17434586584568024, -0.01498030498623848, 0.11830329149961472, 0.016039730980992317, -0.13654667139053345, -0.03913423791527748, -0.2194267064332962, 0.1282399445772171, 0.22643792629241943, 0.47359511256217957, 0.15032468736171722, 0.03246146813035011, 0.03321428596973419, 0.21515770256519318, -0.5671080350875854, 0.2659573256969452, -0.08884955942630768, -0.004131757188588381, 0.0746438056230545, 0.14443759620189667, -0.030301183462142944, 0.33789676427841187, -0.042525336146354675, 0.1457137167453766, -0.12862981855869293, 0.008957206271588802, -0.5539681911468506, 0.4854546785354614, 0.387459397315979, 0.13246704638004303, -0.0448889285326004, -0.24685737490653992, 0.1734629124403, 0.07412078231573105, -0.27653083205223083, 0.1322595775127411, -0.43069255352020264, -0.43199506402015686, 0.18579724431037903, 0.20948676764965057, 0.9232465028762817, 0.332022100687027, 0.37195777893066406, 0.10070342570543289, 0.103580541908741, 0.2642515301704407, -0.23411165177822113, 0.4092499911785126, -0.3693643808364868, -0.046317894011735916, 0.024538729339838028, -0.16673347353935242, -0.22738294303417206, -0.010426540859043598, -0.010993897914886475, 0.30259546637535095, 0.32404956221580505, 0.5999361276626587, -0.11212006956338882, 0.16670002043247223, 0.21569199860095978, -0.15810944139957428, -0.24923092126846313, 0.2010732889175415, -0.29230424761772156, 0.3152405619621277, -0.11240841448307037, -0.07410834729671478, -0.016508914530277252, -0.3513421416282654, 0.06714001297950745, 0.06094168499112129, -0.33651697635650635, 0.30484193563461304, 0.1334618479013443, -0.2726665437221527, -0.2969878315925598, 0.2550783157348633, 0.11198701709508896, 0.08252796530723572, -0.04859941452741623, -0.035185184329748154, 0.4006448984146118, 0.3504580557346344, -0.006891822908073664, -0.20158866047859192, 0.1864124983549118, 0.06387507170438766, -0.01289470586925745, 0.030242804437875748, -0.05640558898448944, -0.30158281326293945, -0.2525118589401245, 0.14820460975170135, 0.003372971899807453, -0.4129619300365448, -0.055318888276815414, 0.06144449859857559, -0.06678875535726547, -0.09413730353116989, 0.15088044106960297, -0.02606821060180664, -0.10794124007225037, 0.31515446305274963, 0.11733822524547577, -0.24304218590259552, 0.05235825479030609, 0.47483208775520325, -0.04039623588323593, 0.031818028539419174, 0.41542848944664, -0.15281835198402405, -0.10380636155605316, -0.2553911507129669, -0.2841723561286926, -0.10206018388271332, -0.4488404393196106, 0.11220814287662506, -0.2194056361913681, -0.40191105008125305, 0.044232070446014404, 0.28081876039505005, 0.014706442132592201, 0.31849661469459534, -0.011618257500231266, -0.12495818734169006, -0.3769727349281311, -0.10379086434841156, -0.10658226162195206, 0.26717832684516907, 0.2795191705226898, 0.284347802400589, -0.29586678743362427, 0.24922697246074677, -0.38310956954956055, -0.10714221000671387, -0.18717068433761597, 0.43791306018829346, -0.039137572050094604, -0.36558258533477783, -0.009147626347839832, 0.0018550800159573555, 0.1299709975719452, 0.17754508554935455, -0.6386584043502808, -0.29261913895606995, -0.0909920185804367, 0.0818607434630394, -0.007587989326566458, -0.15982191264629364, -0.23928239941596985, -0.017214275896549225, 0.18238499760627747, -0.42740264534950256, 0.09215671569108963, 0.15121649205684662, 0.04703198000788689, 0.2371872216463089, -0.1610299050807953, 0.12145564705133438, 0.28816723823547363, -0.1865275353193283, -0.18414904177188873, -0.07622885704040527, 0.08048073947429657, 0.25701698660850525, -0.06946761906147003, 0.02047896571457386, -0.21040348708629608, 0.03047192096710205, 0.08927451074123383, 0.1427948772907257, 0.2369290292263031, -0.024752648547291756, -0.21566133201122284, 0.0005506349843926728, 0.18279911577701569, 0.11411940306425095, -0.07855328917503357, -0.32897576689720154, 0.28022339940071106, 0.2783651351928711, -0.11751823127269745, -0.010237314738333225, 0.16770750284194946, -0.0993887260556221, 0.2528513967990875, 0.20715337991714478, 0.10758770257234573, 0.05261155217885971, 0.2696119546890259, 0.1264941692352295, 0.4050402343273163, -0.3483290374279022, 0.08942107856273651, 0.06441547721624374, 0.10164907574653625, 0.037778209894895554, 0.3812154531478882, -0.15324409306049347, 0.08992163836956024, 0.14838311076164246, 0.18833674490451813, 0.20253996551036835, 0.2628902196884155, 0.21354077756404877, -0.18786455690860748, -0.03412069007754326, -0.24108663201332092, 0.1436367630958557, -0.1427268236875534, 0.41565975546836853, 0.15059226751327515, 0.19285550713539124, -0.11485161632299423, -0.46193909645080566, -0.24626745283603668, 0.23536424338817596, -0.324541836977005, -0.3092094957828522, -0.04867883026599884, -0.0801846906542778, 0.055929236114025116, 0.19456082582473755, -0.07928955554962158, 0.20385949313640594, 0.3428383469581604, -0.008659737184643745, -0.013750121928751469, -0.08435364067554474, -0.17061713337898254, 0.11397942155599594, 0.2748646140098572, -0.0817389041185379, 0.446936696767807, 0.010801088996231556, -0.05156935378909111, -0.23447194695472717, 0.41141754388809204, 0.28612878918647766, 0.19494746625423431, -0.16886235773563385, 0.0639532282948494, 0.20064730942249298, 0.03061334788799286, -0.3546886742115021, 0.32865145802497864, 0.151144340634346, -0.29302841424942017, 0.36800533533096313, 0.1237955093383789, -0.2321559190750122, 0.2764036953449249, -0.10419564694166183, -0.20146705210208893, -0.04988894239068031, 0.3224369287490845, 0.09970726817846298, -0.039932768791913986, -0.1949632614850998, 0.11112037301063538, -0.144571453332901, -0.28533878922462463, 0.15934637188911438, -0.023898735642433167, 0.46766093373298645, -0.27567628026008606, 0.11720526963472366, -0.06182510778307915, 0.6220861673355103, -0.029227983206510544, 0.1243293285369873, -0.3501380383968353, 0.10264355689287186, -0.33316680788993835, 0.19841012358665466, -0.08381723612546921, 0.15374304354190826, -0.3975839614868164, 0.20426513254642487, -0.11549928039312363, -0.023466292768716812, -0.11804226040840149, -0.007886320352554321, 0.1474074274301529, 0.17893218994140625, -0.38691800832748413, 0.15548168122768402, -0.08625979721546173, -0.0062670656479895115, 0.29177364706993103, -0.3766753375530243, 0.2219911813735962, 0.04247850924730301, 0.20691059529781342, 0.01098591648042202, -0.0217235516756773, -0.05480630695819855, 0.3365738093852997, 0.6232410669326782, 0.3470291495323181, 0.14255070686340332, -0.10102760046720505, 0.08155444264411926, -0.3474372327327728, 0.014104297384619713, -0.06088805943727493, -0.20964989066123962, 0.09756883978843689, 0.525830864906311, -0.2256278097629547, 0.13194623589515686, -0.17804619669914246, 0.23669403791427612, 0.38227635622024536, -0.15272943675518036, 0.06180020794272423, 0.04822545871138573, -0.259840190410614, -0.0051995194517076015, -0.12586642801761627, 0.4537962079048157, 0.007182563189417124, 0.32383066415786743, -0.35929134488105774, -0.5417627096176147, 0.4329098165035248, -0.45112600922584534, -0.3025253415107727, 0.12327741086483002, 0.23878581821918488, 0.34956854581832886, -0.06000017747282982, -0.5436674952507019, -0.17249032855033875, 0.2691141366958618, -0.030661141499876976, -0.15704505145549774, 0.20768523216247559, -0.14864318072795868, -0.04107455909252167, -0.11675450205802917, 0.17646948993206024, 0.10152691602706909, -0.2914120852947235, 0.24369898438453674, -0.37010377645492554 ]
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
[ -0.1263234168291092, -0.22070905566215515, -0.10063228756189346, 0.2780112326145172, 0.24547505378723145, -0.20522351562976837, 0.1372111439704895, 0.5128591656684875, 0.1882859617471695, -0.062335360795259476, 0.1278190314769745, 0.38913482427597046, 0.0037764159496873617, -0.1606176644563675, -0.01074370089918375, 0.02421225793659687, 0.09608051925897598, 0.23505258560180664, -0.14556263387203217, -0.16482320427894592, 0.0818943902850151, 0.2129843682050705, -0.04241016134619713, 0.11652717739343643, -0.3903499245643616, 0.04128958284854889, 0.21235674619674683, -0.23089076578617096, 0.06845791637897491, -0.2570658028125763, 0.1678502857685089, -0.09324300289154053, 0.02583426982164383, 0.16631819307804108, -0.000109738131868653, 0.0067979744635522366, -0.015387536957859993, -0.13856972754001617, -0.007274676114320755, 0.04019525647163391, 0.1181342601776123, -0.2117292284965515, -0.1091829240322113, -0.3121323585510254, -0.3320121765136719, -0.0376615971326828, 0.01780911348760128, -0.3556598424911499, 0.31285756826400757, 0.14745281636714935, 0.2607105076313019, 0.22490935027599335, -0.2538118362426758, 0.21113692224025726, 0.14512760937213898, -0.02932742051780224, -0.17419736087322235, -0.04230478033423424, -0.043978799134492874, -0.3344639539718628, -0.3767643868923187, 0.5254572629928589, -0.07658200711011887, 0.17014172673225403, 0.21466556191444397, 0.2250560075044632, 0.17260614037513733, 0.03593749552965164, 0.1429366171360016, -0.1139628142118454, 0.011095208115875721, -0.22691704332828522, -0.027116188779473305, -0.3733711242675781, -0.4186401069164276, -0.07852597534656525, 0.13762468099594116, 0.181045800447464, -0.1363930106163025, -0.06371636688709259, -0.43193674087524414, 0.23928946256637573, 0.06524856388568878, -0.012788441963493824, 0.1195276528596878, 0.2951611876487732, 0.04380587115883827, 0.19522327184677124, 0.12944640219211578, -0.06973284482955933, -0.02255813591182232, -0.21451735496520996, -0.05897236615419388, 0.22640195488929749, -0.43598562479019165, 0.07208240777254105, 0.43151915073394775, -0.14041432738304138, -0.053242821246385574, 0.030637066811323166, 0.3919302225112915, 0.07921385765075684, 0.11681852489709854, -0.008491439744830132, -0.1041610911488533, 0.5898671746253967, 0.044322144240140915, 0.1711844950914383, -0.014970467425882816, -0.1982778012752533, -0.35601741075515747, 0.0049875471740961075, 0.2880174517631531, -0.2866014838218689, 0.42841267585754395, 0.402360200881958, -0.04225970804691315, -0.12152931094169617, -0.03010164014995098, 0.0940236821770668, -0.3308537006378174, 0.05566929280757904, 0.058729760348796844, 0.31395789980888367, -0.2254507839679718, 0.14926527440547943, -0.17376787960529327, -0.10335493087768555, -0.38496968150138855, 0.12735134363174438, -0.33784639835357666, -0.033323660492897034, -0.4962180256843567, 0.001516752759926021, 0.2853839695453644, -0.0028605787083506584, 0.285352885723114, -0.03844567388296127, -0.13722026348114014, -0.30198997259140015, 0.10317455977201462, -0.3985126316547394, 0.5702307224273682, -0.10722927004098892, -0.19701692461967468, 0.24102142453193665, 0.13037806749343872, 0.08526397496461868, -0.2976876199245453, -0.19837690889835358, -0.16980776190757751, -0.20906922221183777, 0.25985991954803467, 0.24224543571472168, -0.07942714542150497, 0.14139650762081146, 0.09248153120279312, 0.07767834514379501, 0.34325864911079407, -0.23285464942455292, 0.10284031927585602, -0.20148888230323792, -0.28198033571243286, -0.20435941219329834, 0.1625692993402481, 0.26230981945991516, 0.06677709519863129, -0.1373722106218338, 0.40179747343063354, 0.23718252778053284, 0.14729131758213043, 0.4138065278530121, -0.1965337097644806, 0.30949482321739197, -0.4334923326969147, 0.3486187756061554, 0.5064778327941895, -0.32673120498657227, -0.49634432792663574, 0.09944815188646317, -0.15631528198719025, 0.10028795152902603, 0.10244916379451752, 0.15137308835983276, 0.02908792719244957, 0.01441967859864235, 0.12853601574897766, 0.018702292814850807, -0.0005142427398823202, 0.20842033624649048, -0.31692802906036377, -0.021933671087026596, 0.5500645637512207, -0.25246718525886536, 0.028472382575273514, -0.1312367022037506, -0.05504239350557327, 0.021932199597358704, 0.0067685069516301155, -0.1121898964047432, 0.1218591183423996, 0.11460506170988083, -0.05687858536839485, -0.15818439424037933, -0.10643845796585083, -0.02469293214380741, -0.28826266527175903, 0.23830117285251617, -0.14981962740421295, -0.09996888786554337, 0.061725009232759476, -0.09704943746328354, -0.1620030701160431, -0.12927453219890594, -0.1809171438217163, -0.27306488156318665, 0.20130188763141632, 0.23052427172660828, 0.32024043798446655, -0.07290686666965485, 0.22520951926708221, 0.36808186769485474, 0.04836326465010643, -0.21607212722301483, -0.4300347864627838, 0.08568006008863449, -0.17773543298244476, -0.21882851421833038, -0.07484656572341919, 0.2230692207813263, 0.20902669429779053, -0.04792690649628639, -0.1424286812543869, 0.1482759416103363, -0.222085103392601, 0.4189184010028839, -0.13702847063541412, 0.269452303647995, -0.06965663284063339, 0.02930617146193981, 0.01666727475821972, 0.2540055215358734, 0.058820389211177826, -0.08980126678943634, -0.07270590960979462, 0.4742339253425598, 0.18140152096748352, 0.17011570930480957, -0.04807102307677269, -0.21626847982406616, 0.05198436602950096, -0.17831741273403168, -0.12091562896966934, -0.1188163310289383, 0.1364070028066635, -0.08057195693254471, 0.14888811111450195, 0.15232902765274048, -0.10698776692152023, 0.3120073974132538, 0.7600834369659424, 0.1969190239906311, -0.011016630567610264, 0.015740377828478813, -0.05721341073513031, -0.2177916318178177, 0.3338569700717926, -0.05868323892354965, 0.3103885352611542, 0.10473150759935379, 0.12716469168663025, -0.04620937630534172, -0.2196495234966278, -0.15443581342697144, 0.12464867532253265, -0.21436883509159088, -0.19197611510753632, 0.30803436040878296, 0.28208714723587036, -0.15836739540100098, -0.35378298163414, 0.09401589632034302, -0.012764276005327702, 0.0261914674192667, -0.15800099074840546, 0.21723708510398865, -0.16652946174144745, -0.2780843675136566, -0.28641605377197266, 0.13615699112415314, -0.12220324575901031, -0.0921652764081955, -0.04126395657658577, 0.42542997002601624, -0.036653950810432434, 0.19515201449394226, -0.3149716258049011, 0.08951807022094727, 0.22660399973392487, -0.2663272023200989, -0.07168542593717575, -0.19028738141059875, -0.39492031931877136, 0.04223180189728737, -0.09523077309131622, -0.09299586713314056, 0.37832993268966675, -0.04731646925210953, -0.06795979291200638, -0.28814661502838135, -0.17254012823104858, 0.10717806965112686, -0.09261105209589005, -0.15536722540855408, 0.16724833846092224, -0.03133821487426758, -0.1011589840054512, -0.298745721578598, 0.2937973439693451, -0.3053131401538849, -0.2991430461406708, -0.0519869327545166, 0.09890599548816681, 0.030109098181128502, -0.2803472578525543, -0.42862468957901, 0.2735773026943207, -0.31724420189857483, 0.07546748220920563, -0.14892889559268951, 0.1862272024154663, 0.3309342861175537, 0.008954071439802647, -0.014561185613274574, -0.06404047459363937, 0.16439427435398102, -0.46740975975990295, -0.14095474779605865, 0.32257840037345886, 0.018760429695248604, -0.21630436182022095, -0.32459110021591187, -0.2757244408130646, 0.34772419929504395, 0.1923881620168686, -0.5803956985473633, -0.21442270278930664, -0.04890016093850136, 0.24803172051906586, -0.051259271800518036, -0.046495221555233, 0.5745418071746826, 0.019725333899259567, -0.17946326732635498, -0.16727812588214874, -0.33183929324150085, 0.35973072052001953, 0.41926220059394836, 0.24948781728744507, -0.011158456094563007, 0.1778150349855423, 0.20092009007930756, 0.6707469820976257, 0.20192037522792816, -0.2590579390525818, 0.3317311108112335, 0.10723166167736053, 0.07499769330024719, -0.20349697768688202, -0.31302309036254883, 0.18576005101203918, -0.1816031038761139, 0.01973230578005314, 0.27198368310928345, 0.006250523962080479, -0.2801840007305145, 0.017742546275258064, 0.2188185304403305, -0.2253447324037552, -0.3280579745769501, 0.2597457468509674, -0.3424278795719147, 0.24094857275485992, 0.03636978566646576, 0.31579962372779846, -0.5820139050483704, -0.2427240014076233, 0.1946277767419815, -0.3359457850456238, 0.27984464168548584, 0.04449823498725891, -0.5836009383201599, 0.1214703917503357, -0.5380211472511292, 0.18194159865379333, 0.31462812423706055, 0.27240779995918274, -0.09202467650175095, -0.24652515351772308, 0.04334753006696701, -0.14713141322135925, 0.57114177942276, 0.04097321629524231, 0.40857937932014465, 0.1699264943599701, -0.13024675846099854, -0.19947350025177002, -0.22471316158771515, 0.017235223203897476, 0.2801368832588196, 0.3265001177787781, 0.4008474349975586, -0.34524449706077576, -0.3632737994194031, -0.265765905380249, 0.007363766897469759, -0.0461815744638443, 0.007106977980583906, -0.12583452463150024, 0.18938513100147247, 0.016485881060361862, 0.17434586584568024, -0.01498030498623848, 0.11830329149961472, 0.016039730980992317, -0.13654667139053345, -0.03913423791527748, -0.2194267064332962, 0.1282399445772171, 0.22643792629241943, 0.47359511256217957, 0.15032468736171722, 0.03246146813035011, 0.03321428596973419, 0.21515770256519318, -0.5671080350875854, 0.2659573256969452, -0.08884955942630768, -0.004131757188588381, 0.0746438056230545, 0.14443759620189667, -0.030301183462142944, 0.33789676427841187, -0.042525336146354675, 0.1457137167453766, -0.12862981855869293, 0.008957206271588802, -0.5539681911468506, 0.4854546785354614, 0.387459397315979, 0.13246704638004303, -0.0448889285326004, -0.24685737490653992, 0.1734629124403, 0.07412078231573105, -0.27653083205223083, 0.1322595775127411, -0.43069255352020264, -0.43199506402015686, 0.18579724431037903, 0.20948676764965057, 0.9232465028762817, 0.332022100687027, 0.37195777893066406, 0.10070342570543289, 0.103580541908741, 0.2642515301704407, -0.23411165177822113, 0.4092499911785126, -0.3693643808364868, -0.046317894011735916, 0.024538729339838028, -0.16673347353935242, -0.22738294303417206, -0.010426540859043598, -0.010993897914886475, 0.30259546637535095, 0.32404956221580505, 0.5999361276626587, -0.11212006956338882, 0.16670002043247223, 0.21569199860095978, -0.15810944139957428, -0.24923092126846313, 0.2010732889175415, -0.29230424761772156, 0.3152405619621277, -0.11240841448307037, -0.07410834729671478, -0.016508914530277252, -0.3513421416282654, 0.06714001297950745, 0.06094168499112129, -0.33651697635650635, 0.30484193563461304, 0.1334618479013443, -0.2726665437221527, -0.2969878315925598, 0.2550783157348633, 0.11198701709508896, 0.08252796530723572, -0.04859941452741623, -0.035185184329748154, 0.4006448984146118, 0.3504580557346344, -0.006891822908073664, -0.20158866047859192, 0.1864124983549118, 0.06387507170438766, -0.01289470586925745, 0.030242804437875748, -0.05640558898448944, -0.30158281326293945, -0.2525118589401245, 0.14820460975170135, 0.003372971899807453, -0.4129619300365448, -0.055318888276815414, 0.06144449859857559, -0.06678875535726547, -0.09413730353116989, 0.15088044106960297, -0.02606821060180664, -0.10794124007225037, 0.31515446305274963, 0.11733822524547577, -0.24304218590259552, 0.05235825479030609, 0.47483208775520325, -0.04039623588323593, 0.031818028539419174, 0.41542848944664, -0.15281835198402405, -0.10380636155605316, -0.2553911507129669, -0.2841723561286926, -0.10206018388271332, -0.4488404393196106, 0.11220814287662506, -0.2194056361913681, -0.40191105008125305, 0.044232070446014404, 0.28081876039505005, 0.014706442132592201, 0.31849661469459534, -0.011618257500231266, -0.12495818734169006, -0.3769727349281311, -0.10379086434841156, -0.10658226162195206, 0.26717832684516907, 0.2795191705226898, 0.284347802400589, -0.29586678743362427, 0.24922697246074677, -0.38310956954956055, -0.10714221000671387, -0.18717068433761597, 0.43791306018829346, -0.039137572050094604, -0.36558258533477783, -0.009147626347839832, 0.0018550800159573555, 0.1299709975719452, 0.17754508554935455, -0.6386584043502808, -0.29261913895606995, -0.0909920185804367, 0.0818607434630394, -0.007587989326566458, -0.15982191264629364, -0.23928239941596985, -0.017214275896549225, 0.18238499760627747, -0.42740264534950256, 0.09215671569108963, 0.15121649205684662, 0.04703198000788689, 0.2371872216463089, -0.1610299050807953, 0.12145564705133438, 0.28816723823547363, -0.1865275353193283, -0.18414904177188873, -0.07622885704040527, 0.08048073947429657, 0.25701698660850525, -0.06946761906147003, 0.02047896571457386, -0.21040348708629608, 0.03047192096710205, 0.08927451074123383, 0.1427948772907257, 0.2369290292263031, -0.024752648547291756, -0.21566133201122284, 0.0005506349843926728, 0.18279911577701569, 0.11411940306425095, -0.07855328917503357, -0.32897576689720154, 0.28022339940071106, 0.2783651351928711, -0.11751823127269745, -0.010237314738333225, 0.16770750284194946, -0.0993887260556221, 0.2528513967990875, 0.20715337991714478, 0.10758770257234573, 0.05261155217885971, 0.2696119546890259, 0.1264941692352295, 0.4050402343273163, -0.3483290374279022, 0.08942107856273651, 0.06441547721624374, 0.10164907574653625, 0.037778209894895554, 0.3812154531478882, -0.15324409306049347, 0.08992163836956024, 0.14838311076164246, 0.18833674490451813, 0.20253996551036835, 0.2628902196884155, 0.21354077756404877, -0.18786455690860748, -0.03412069007754326, -0.24108663201332092, 0.1436367630958557, -0.1427268236875534, 0.41565975546836853, 0.15059226751327515, 0.19285550713539124, -0.11485161632299423, -0.46193909645080566, -0.24626745283603668, 0.23536424338817596, -0.324541836977005, -0.3092094957828522, -0.04867883026599884, -0.0801846906542778, 0.055929236114025116, 0.19456082582473755, -0.07928955554962158, 0.20385949313640594, 0.3428383469581604, -0.008659737184643745, -0.013750121928751469, -0.08435364067554474, -0.17061713337898254, 0.11397942155599594, 0.2748646140098572, -0.0817389041185379, 0.446936696767807, 0.010801088996231556, -0.05156935378909111, -0.23447194695472717, 0.41141754388809204, 0.28612878918647766, 0.19494746625423431, -0.16886235773563385, 0.0639532282948494, 0.20064730942249298, 0.03061334788799286, -0.3546886742115021, 0.32865145802497864, 0.151144340634346, -0.29302841424942017, 0.36800533533096313, 0.1237955093383789, -0.2321559190750122, 0.2764036953449249, -0.10419564694166183, -0.20146705210208893, -0.04988894239068031, 0.3224369287490845, 0.09970726817846298, -0.039932768791913986, -0.1949632614850998, 0.11112037301063538, -0.144571453332901, -0.28533878922462463, 0.15934637188911438, -0.023898735642433167, 0.46766093373298645, -0.27567628026008606, 0.11720526963472366, -0.06182510778307915, 0.6220861673355103, -0.029227983206510544, 0.1243293285369873, -0.3501380383968353, 0.10264355689287186, -0.33316680788993835, 0.19841012358665466, -0.08381723612546921, 0.15374304354190826, -0.3975839614868164, 0.20426513254642487, -0.11549928039312363, -0.023466292768716812, -0.11804226040840149, -0.007886320352554321, 0.1474074274301529, 0.17893218994140625, -0.38691800832748413, 0.15548168122768402, -0.08625979721546173, -0.0062670656479895115, 0.29177364706993103, -0.3766753375530243, 0.2219911813735962, 0.04247850924730301, 0.20691059529781342, 0.01098591648042202, -0.0217235516756773, -0.05480630695819855, 0.3365738093852997, 0.6232410669326782, 0.3470291495323181, 0.14255070686340332, -0.10102760046720505, 0.08155444264411926, -0.3474372327327728, 0.014104297384619713, -0.06088805943727493, -0.20964989066123962, 0.09756883978843689, 0.525830864906311, -0.2256278097629547, 0.13194623589515686, -0.17804619669914246, 0.23669403791427612, 0.38227635622024536, -0.15272943675518036, 0.06180020794272423, 0.04822545871138573, -0.259840190410614, -0.0051995194517076015, -0.12586642801761627, 0.4537962079048157, 0.007182563189417124, 0.32383066415786743, -0.35929134488105774, -0.5417627096176147, 0.4329098165035248, -0.45112600922584534, -0.3025253415107727, 0.12327741086483002, 0.23878581821918488, 0.34956854581832886, -0.06000017747282982, -0.5436674952507019, -0.17249032855033875, 0.2691141366958618, -0.030661141499876976, -0.15704505145549774, 0.20768523216247559, -0.14864318072795868, -0.04107455909252167, -0.11675450205802917, 0.17646948993206024, 0.10152691602706909, -0.2914120852947235, 0.24369898438453674, -0.37010377645492554 ]
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).
[ -0.1263234168291092, -0.22070905566215515, -0.10063228756189346, 0.2780112326145172, 0.24547505378723145, -0.20522351562976837, 0.1372111439704895, 0.5128591656684875, 0.1882859617471695, -0.062335360795259476, 0.1278190314769745, 0.38913482427597046, 0.0037764159496873617, -0.1606176644563675, -0.01074370089918375, 0.02421225793659687, 0.09608051925897598, 0.23505258560180664, -0.14556263387203217, -0.16482320427894592, 0.0818943902850151, 0.2129843682050705, -0.04241016134619713, 0.11652717739343643, -0.3903499245643616, 0.04128958284854889, 0.21235674619674683, -0.23089076578617096, 0.06845791637897491, -0.2570658028125763, 0.1678502857685089, -0.09324300289154053, 0.02583426982164383, 0.16631819307804108, -0.000109738131868653, 0.0067979744635522366, -0.015387536957859993, -0.13856972754001617, -0.007274676114320755, 0.04019525647163391, 0.1181342601776123, -0.2117292284965515, -0.1091829240322113, -0.3121323585510254, -0.3320121765136719, -0.0376615971326828, 0.01780911348760128, -0.3556598424911499, 0.31285756826400757, 0.14745281636714935, 0.2607105076313019, 0.22490935027599335, -0.2538118362426758, 0.21113692224025726, 0.14512760937213898, -0.02932742051780224, -0.17419736087322235, -0.04230478033423424, -0.043978799134492874, -0.3344639539718628, -0.3767643868923187, 0.5254572629928589, -0.07658200711011887, 0.17014172673225403, 0.21466556191444397, 0.2250560075044632, 0.17260614037513733, 0.03593749552965164, 0.1429366171360016, -0.1139628142118454, 0.011095208115875721, -0.22691704332828522, -0.027116188779473305, -0.3733711242675781, -0.4186401069164276, -0.07852597534656525, 0.13762468099594116, 0.181045800447464, -0.1363930106163025, -0.06371636688709259, -0.43193674087524414, 0.23928946256637573, 0.06524856388568878, -0.012788441963493824, 0.1195276528596878, 0.2951611876487732, 0.04380587115883827, 0.19522327184677124, 0.12944640219211578, -0.06973284482955933, -0.02255813591182232, -0.21451735496520996, -0.05897236615419388, 0.22640195488929749, -0.43598562479019165, 0.07208240777254105, 0.43151915073394775, -0.14041432738304138, -0.053242821246385574, 0.030637066811323166, 0.3919302225112915, 0.07921385765075684, 0.11681852489709854, -0.008491439744830132, -0.1041610911488533, 0.5898671746253967, 0.044322144240140915, 0.1711844950914383, -0.014970467425882816, -0.1982778012752533, -0.35601741075515747, 0.0049875471740961075, 0.2880174517631531, -0.2866014838218689, 0.42841267585754395, 0.402360200881958, -0.04225970804691315, -0.12152931094169617, -0.03010164014995098, 0.0940236821770668, -0.3308537006378174, 0.05566929280757904, 0.058729760348796844, 0.31395789980888367, -0.2254507839679718, 0.14926527440547943, -0.17376787960529327, -0.10335493087768555, -0.38496968150138855, 0.12735134363174438, -0.33784639835357666, -0.033323660492897034, -0.4962180256843567, 0.001516752759926021, 0.2853839695453644, -0.0028605787083506584, 0.285352885723114, -0.03844567388296127, -0.13722026348114014, -0.30198997259140015, 0.10317455977201462, -0.3985126316547394, 0.5702307224273682, -0.10722927004098892, -0.19701692461967468, 0.24102142453193665, 0.13037806749343872, 0.08526397496461868, -0.2976876199245453, -0.19837690889835358, -0.16980776190757751, -0.20906922221183777, 0.25985991954803467, 0.24224543571472168, -0.07942714542150497, 0.14139650762081146, 0.09248153120279312, 0.07767834514379501, 0.34325864911079407, -0.23285464942455292, 0.10284031927585602, -0.20148888230323792, -0.28198033571243286, -0.20435941219329834, 0.1625692993402481, 0.26230981945991516, 0.06677709519863129, -0.1373722106218338, 0.40179747343063354, 0.23718252778053284, 0.14729131758213043, 0.4138065278530121, -0.1965337097644806, 0.30949482321739197, -0.4334923326969147, 0.3486187756061554, 0.5064778327941895, -0.32673120498657227, -0.49634432792663574, 0.09944815188646317, -0.15631528198719025, 0.10028795152902603, 0.10244916379451752, 0.15137308835983276, 0.02908792719244957, 0.01441967859864235, 0.12853601574897766, 0.018702292814850807, -0.0005142427398823202, 0.20842033624649048, -0.31692802906036377, -0.021933671087026596, 0.5500645637512207, -0.25246718525886536, 0.028472382575273514, -0.1312367022037506, -0.05504239350557327, 0.021932199597358704, 0.0067685069516301155, -0.1121898964047432, 0.1218591183423996, 0.11460506170988083, -0.05687858536839485, -0.15818439424037933, -0.10643845796585083, -0.02469293214380741, -0.28826266527175903, 0.23830117285251617, -0.14981962740421295, -0.09996888786554337, 0.061725009232759476, -0.09704943746328354, -0.1620030701160431, -0.12927453219890594, -0.1809171438217163, -0.27306488156318665, 0.20130188763141632, 0.23052427172660828, 0.32024043798446655, -0.07290686666965485, 0.22520951926708221, 0.36808186769485474, 0.04836326465010643, -0.21607212722301483, -0.4300347864627838, 0.08568006008863449, -0.17773543298244476, -0.21882851421833038, -0.07484656572341919, 0.2230692207813263, 0.20902669429779053, -0.04792690649628639, -0.1424286812543869, 0.1482759416103363, -0.222085103392601, 0.4189184010028839, -0.13702847063541412, 0.269452303647995, -0.06965663284063339, 0.02930617146193981, 0.01666727475821972, 0.2540055215358734, 0.058820389211177826, -0.08980126678943634, -0.07270590960979462, 0.4742339253425598, 0.18140152096748352, 0.17011570930480957, -0.04807102307677269, -0.21626847982406616, 0.05198436602950096, -0.17831741273403168, -0.12091562896966934, -0.1188163310289383, 0.1364070028066635, -0.08057195693254471, 0.14888811111450195, 0.15232902765274048, -0.10698776692152023, 0.3120073974132538, 0.7600834369659424, 0.1969190239906311, -0.011016630567610264, 0.015740377828478813, -0.05721341073513031, -0.2177916318178177, 0.3338569700717926, -0.05868323892354965, 0.3103885352611542, 0.10473150759935379, 0.12716469168663025, -0.04620937630534172, -0.2196495234966278, -0.15443581342697144, 0.12464867532253265, -0.21436883509159088, -0.19197611510753632, 0.30803436040878296, 0.28208714723587036, -0.15836739540100098, -0.35378298163414, 0.09401589632034302, -0.012764276005327702, 0.0261914674192667, -0.15800099074840546, 0.21723708510398865, -0.16652946174144745, -0.2780843675136566, -0.28641605377197266, 0.13615699112415314, -0.12220324575901031, -0.0921652764081955, -0.04126395657658577, 0.42542997002601624, -0.036653950810432434, 0.19515201449394226, -0.3149716258049011, 0.08951807022094727, 0.22660399973392487, -0.2663272023200989, -0.07168542593717575, -0.19028738141059875, -0.39492031931877136, 0.04223180189728737, -0.09523077309131622, -0.09299586713314056, 0.37832993268966675, -0.04731646925210953, -0.06795979291200638, -0.28814661502838135, -0.17254012823104858, 0.10717806965112686, -0.09261105209589005, -0.15536722540855408, 0.16724833846092224, -0.03133821487426758, -0.1011589840054512, -0.298745721578598, 0.2937973439693451, -0.3053131401538849, -0.2991430461406708, -0.0519869327545166, 0.09890599548816681, 0.030109098181128502, -0.2803472578525543, -0.42862468957901, 0.2735773026943207, -0.31724420189857483, 0.07546748220920563, -0.14892889559268951, 0.1862272024154663, 0.3309342861175537, 0.008954071439802647, -0.014561185613274574, -0.06404047459363937, 0.16439427435398102, -0.46740975975990295, -0.14095474779605865, 0.32257840037345886, 0.018760429695248604, -0.21630436182022095, -0.32459110021591187, -0.2757244408130646, 0.34772419929504395, 0.1923881620168686, -0.5803956985473633, -0.21442270278930664, -0.04890016093850136, 0.24803172051906586, -0.051259271800518036, -0.046495221555233, 0.5745418071746826, 0.019725333899259567, -0.17946326732635498, -0.16727812588214874, -0.33183929324150085, 0.35973072052001953, 0.41926220059394836, 0.24948781728744507, -0.011158456094563007, 0.1778150349855423, 0.20092009007930756, 0.6707469820976257, 0.20192037522792816, -0.2590579390525818, 0.3317311108112335, 0.10723166167736053, 0.07499769330024719, -0.20349697768688202, -0.31302309036254883, 0.18576005101203918, -0.1816031038761139, 0.01973230578005314, 0.27198368310928345, 0.006250523962080479, -0.2801840007305145, 0.017742546275258064, 0.2188185304403305, -0.2253447324037552, -0.3280579745769501, 0.2597457468509674, -0.3424278795719147, 0.24094857275485992, 0.03636978566646576, 0.31579962372779846, -0.5820139050483704, -0.2427240014076233, 0.1946277767419815, -0.3359457850456238, 0.27984464168548584, 0.04449823498725891, -0.5836009383201599, 0.1214703917503357, -0.5380211472511292, 0.18194159865379333, 0.31462812423706055, 0.27240779995918274, -0.09202467650175095, -0.24652515351772308, 0.04334753006696701, -0.14713141322135925, 0.57114177942276, 0.04097321629524231, 0.40857937932014465, 0.1699264943599701, -0.13024675846099854, -0.19947350025177002, -0.22471316158771515, 0.017235223203897476, 0.2801368832588196, 0.3265001177787781, 0.4008474349975586, -0.34524449706077576, -0.3632737994194031, -0.265765905380249, 0.007363766897469759, -0.0461815744638443, 0.007106977980583906, -0.12583452463150024, 0.18938513100147247, 0.016485881060361862, 0.17434586584568024, -0.01498030498623848, 0.11830329149961472, 0.016039730980992317, -0.13654667139053345, -0.03913423791527748, -0.2194267064332962, 0.1282399445772171, 0.22643792629241943, 0.47359511256217957, 0.15032468736171722, 0.03246146813035011, 0.03321428596973419, 0.21515770256519318, -0.5671080350875854, 0.2659573256969452, -0.08884955942630768, -0.004131757188588381, 0.0746438056230545, 0.14443759620189667, -0.030301183462142944, 0.33789676427841187, -0.042525336146354675, 0.1457137167453766, -0.12862981855869293, 0.008957206271588802, -0.5539681911468506, 0.4854546785354614, 0.387459397315979, 0.13246704638004303, -0.0448889285326004, -0.24685737490653992, 0.1734629124403, 0.07412078231573105, -0.27653083205223083, 0.1322595775127411, -0.43069255352020264, -0.43199506402015686, 0.18579724431037903, 0.20948676764965057, 0.9232465028762817, 0.332022100687027, 0.37195777893066406, 0.10070342570543289, 0.103580541908741, 0.2642515301704407, -0.23411165177822113, 0.4092499911785126, -0.3693643808364868, -0.046317894011735916, 0.024538729339838028, -0.16673347353935242, -0.22738294303417206, -0.010426540859043598, -0.010993897914886475, 0.30259546637535095, 0.32404956221580505, 0.5999361276626587, -0.11212006956338882, 0.16670002043247223, 0.21569199860095978, -0.15810944139957428, -0.24923092126846313, 0.2010732889175415, -0.29230424761772156, 0.3152405619621277, -0.11240841448307037, -0.07410834729671478, -0.016508914530277252, -0.3513421416282654, 0.06714001297950745, 0.06094168499112129, -0.33651697635650635, 0.30484193563461304, 0.1334618479013443, -0.2726665437221527, -0.2969878315925598, 0.2550783157348633, 0.11198701709508896, 0.08252796530723572, -0.04859941452741623, -0.035185184329748154, 0.4006448984146118, 0.3504580557346344, -0.006891822908073664, -0.20158866047859192, 0.1864124983549118, 0.06387507170438766, -0.01289470586925745, 0.030242804437875748, -0.05640558898448944, -0.30158281326293945, -0.2525118589401245, 0.14820460975170135, 0.003372971899807453, -0.4129619300365448, -0.055318888276815414, 0.06144449859857559, -0.06678875535726547, -0.09413730353116989, 0.15088044106960297, -0.02606821060180664, -0.10794124007225037, 0.31515446305274963, 0.11733822524547577, -0.24304218590259552, 0.05235825479030609, 0.47483208775520325, -0.04039623588323593, 0.031818028539419174, 0.41542848944664, -0.15281835198402405, -0.10380636155605316, -0.2553911507129669, -0.2841723561286926, -0.10206018388271332, -0.4488404393196106, 0.11220814287662506, -0.2194056361913681, -0.40191105008125305, 0.044232070446014404, 0.28081876039505005, 0.014706442132592201, 0.31849661469459534, -0.011618257500231266, -0.12495818734169006, -0.3769727349281311, -0.10379086434841156, -0.10658226162195206, 0.26717832684516907, 0.2795191705226898, 0.284347802400589, -0.29586678743362427, 0.24922697246074677, -0.38310956954956055, -0.10714221000671387, -0.18717068433761597, 0.43791306018829346, -0.039137572050094604, -0.36558258533477783, -0.009147626347839832, 0.0018550800159573555, 0.1299709975719452, 0.17754508554935455, -0.6386584043502808, -0.29261913895606995, -0.0909920185804367, 0.0818607434630394, -0.007587989326566458, -0.15982191264629364, -0.23928239941596985, -0.017214275896549225, 0.18238499760627747, -0.42740264534950256, 0.09215671569108963, 0.15121649205684662, 0.04703198000788689, 0.2371872216463089, -0.1610299050807953, 0.12145564705133438, 0.28816723823547363, -0.1865275353193283, -0.18414904177188873, -0.07622885704040527, 0.08048073947429657, 0.25701698660850525, -0.06946761906147003, 0.02047896571457386, -0.21040348708629608, 0.03047192096710205, 0.08927451074123383, 0.1427948772907257, 0.2369290292263031, -0.024752648547291756, -0.21566133201122284, 0.0005506349843926728, 0.18279911577701569, 0.11411940306425095, -0.07855328917503357, -0.32897576689720154, 0.28022339940071106, 0.2783651351928711, -0.11751823127269745, -0.010237314738333225, 0.16770750284194946, -0.0993887260556221, 0.2528513967990875, 0.20715337991714478, 0.10758770257234573, 0.05261155217885971, 0.2696119546890259, 0.1264941692352295, 0.4050402343273163, -0.3483290374279022, 0.08942107856273651, 0.06441547721624374, 0.10164907574653625, 0.037778209894895554, 0.3812154531478882, -0.15324409306049347, 0.08992163836956024, 0.14838311076164246, 0.18833674490451813, 0.20253996551036835, 0.2628902196884155, 0.21354077756404877, -0.18786455690860748, -0.03412069007754326, -0.24108663201332092, 0.1436367630958557, -0.1427268236875534, 0.41565975546836853, 0.15059226751327515, 0.19285550713539124, -0.11485161632299423, -0.46193909645080566, -0.24626745283603668, 0.23536424338817596, -0.324541836977005, -0.3092094957828522, -0.04867883026599884, -0.0801846906542778, 0.055929236114025116, 0.19456082582473755, -0.07928955554962158, 0.20385949313640594, 0.3428383469581604, -0.008659737184643745, -0.013750121928751469, -0.08435364067554474, -0.17061713337898254, 0.11397942155599594, 0.2748646140098572, -0.0817389041185379, 0.446936696767807, 0.010801088996231556, -0.05156935378909111, -0.23447194695472717, 0.41141754388809204, 0.28612878918647766, 0.19494746625423431, -0.16886235773563385, 0.0639532282948494, 0.20064730942249298, 0.03061334788799286, -0.3546886742115021, 0.32865145802497864, 0.151144340634346, -0.29302841424942017, 0.36800533533096313, 0.1237955093383789, -0.2321559190750122, 0.2764036953449249, -0.10419564694166183, -0.20146705210208893, -0.04988894239068031, 0.3224369287490845, 0.09970726817846298, -0.039932768791913986, -0.1949632614850998, 0.11112037301063538, -0.144571453332901, -0.28533878922462463, 0.15934637188911438, -0.023898735642433167, 0.46766093373298645, -0.27567628026008606, 0.11720526963472366, -0.06182510778307915, 0.6220861673355103, -0.029227983206510544, 0.1243293285369873, -0.3501380383968353, 0.10264355689287186, -0.33316680788993835, 0.19841012358665466, -0.08381723612546921, 0.15374304354190826, -0.3975839614868164, 0.20426513254642487, -0.11549928039312363, -0.023466292768716812, -0.11804226040840149, -0.007886320352554321, 0.1474074274301529, 0.17893218994140625, -0.38691800832748413, 0.15548168122768402, -0.08625979721546173, -0.0062670656479895115, 0.29177364706993103, -0.3766753375530243, 0.2219911813735962, 0.04247850924730301, 0.20691059529781342, 0.01098591648042202, -0.0217235516756773, -0.05480630695819855, 0.3365738093852997, 0.6232410669326782, 0.3470291495323181, 0.14255070686340332, -0.10102760046720505, 0.08155444264411926, -0.3474372327327728, 0.014104297384619713, -0.06088805943727493, -0.20964989066123962, 0.09756883978843689, 0.525830864906311, -0.2256278097629547, 0.13194623589515686, -0.17804619669914246, 0.23669403791427612, 0.38227635622024536, -0.15272943675518036, 0.06180020794272423, 0.04822545871138573, -0.259840190410614, -0.0051995194517076015, -0.12586642801761627, 0.4537962079048157, 0.007182563189417124, 0.32383066415786743, -0.35929134488105774, -0.5417627096176147, 0.4329098165035248, -0.45112600922584534, -0.3025253415107727, 0.12327741086483002, 0.23878581821918488, 0.34956854581832886, -0.06000017747282982, -0.5436674952507019, -0.17249032855033875, 0.2691141366958618, -0.030661141499876976, -0.15704505145549774, 0.20768523216247559, -0.14864318072795868, -0.04107455909252167, -0.11675450205802917, 0.17646948993206024, 0.10152691602706909, -0.2914120852947235, 0.24369898438453674, -0.37010377645492554 ]