html_url
stringlengths
48
51
title
stringlengths
5
268
comments
stringlengths
70
51.8k
body
stringlengths
0
29.8k
comment_length
int64
16
1.52k
text
stringlengths
164
54.1k
embeddings
sequence
https://github.com/huggingface/datasets/issues/228
Not able to access the XNLI dataset
Thanks for reporting this bug ! As it seems to be just a cache problem, I closed your PR. I think we might just need to clear and reload the `xnli` cache @srush ?
When I try to access the XNLI dataset, I get the following error. The option of plain_text get selected automatically and then I get the following error. ``` FileNotFoundError: [Errno 2] No such file or directory: '/home/sasha/.cache/huggingface/datasets/xnli/plain_text/1.0.0/dataset_info.json' Traceback: File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/ScriptRunner.py", line 322, in _run_script exec(code, module.__dict__) File "/home/sasha/nlp_viewer/run.py", line 86, in <module> dts, fail = get(str(option.id), str(conf_option.name) if conf_option else None) File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 591, in wrapped_func return get_or_create_cached_value() File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 575, in get_or_create_cached_value return_value = func(*args, **kwargs) File "/home/sasha/nlp_viewer/run.py", line 72, in get builder_instance = builder_cls(name=conf) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 610, in __init__ super(GeneratorBasedBuilder, self).__init__(*args, **kwargs) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 152, in __init__ self.info = DatasetInfo.from_directory(self._cache_dir) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/info.py", line 157, in from_directory with open(os.path.join(dataset_info_dir, DATASET_INFO_FILENAME), "r") as f: ``` Is it possible to see if the dataset_info.json is correctly placed?
34
Not able to access the XNLI dataset When I try to access the XNLI dataset, I get the following error. The option of plain_text get selected automatically and then I get the following error. ``` FileNotFoundError: [Errno 2] No such file or directory: '/home/sasha/.cache/huggingface/datasets/xnli/plain_text/1.0.0/dataset_info.json' Traceback: File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/ScriptRunner.py", line 322, in _run_script exec(code, module.__dict__) File "/home/sasha/nlp_viewer/run.py", line 86, in <module> dts, fail = get(str(option.id), str(conf_option.name) if conf_option else None) File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 591, in wrapped_func return get_or_create_cached_value() File "/home/sasha/.local/lib/python3.7/site-packages/streamlit/caching.py", line 575, in get_or_create_cached_value return_value = func(*args, **kwargs) File "/home/sasha/nlp_viewer/run.py", line 72, in get builder_instance = builder_cls(name=conf) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 610, in __init__ super(GeneratorBasedBuilder, self).__init__(*args, **kwargs) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/builder.py", line 152, in __init__ self.info = DatasetInfo.from_directory(self._cache_dir) File "/home/sasha/.local/lib/python3.7/site-packages/nlp/info.py", line 157, in from_directory with open(os.path.join(dataset_info_dir, DATASET_INFO_FILENAME), "r") as f: ``` Is it possible to see if the dataset_info.json is correctly placed? Thanks for reporting this bug ! As it seems to be just a cache problem, I closed your PR. I think we might just need to clear and reload the `xnli` cache @srush ?
[ -0.1408686488866806, -0.18925881385803223, -0.03353646397590637, 0.4693673253059387, 0.40544581413269043, 0.11200030893087387, 0.056036077439785004, 0.35449865460395813, -0.10948893427848816, 0.2927756607532501, -0.14041008055210114, 0.16212110221385956, 0.05022745206952095, -0.02804567478597164, 0.05054093897342682, -0.03532540053129196, -0.169126957654953, 0.15102019906044006, 0.1320728361606598, 0.005323596298694611, -0.1707707643508911, 0.045582592487335205, -0.26906365156173706, 0.25874191522598267, -0.28670650720596313, -0.08528319746255875, 0.0914461761713028, 0.38597795367240906, -0.30817025899887085, -0.5349300503730774, 0.045082367956638336, -0.14433495700359344, 0.24077430367469788, 0.19576658308506012, -0.0001110580051317811, 0.04933277517557144, 0.26291391253471375, -0.0573488250374794, -0.5286889672279358, -0.14541836082935333, -0.3568848669528961, -0.12314407527446747, -0.014699378982186317, -0.3132312595844269, -0.1287817358970642, -0.22096048295497894, 0.40224409103393555, -0.3689234256744385, 0.2718048691749573, 0.31527629494667053, 0.20390821993350983, 0.20004351437091827, 0.12182684242725372, -0.05471378192305565, 0.04933743178844452, 0.13519372045993805, -0.1314477175474167, -0.013969961553812027, 0.23928481340408325, 0.1932622492313385, 0.16366678476333618, 0.45788297057151794, -0.1711953580379486, -0.11164624989032745, 0.25520727038383484, 0.243842214345932, -0.030298396944999695, -0.16049769520759583, 0.32099449634552, 0.23313367366790771, 0.4437568187713623, -0.4332887530326843, -0.43349897861480713, -0.11148454993963242, -0.0038183475844562054, -0.26497241854667664, 0.24886426329612732, 0.10019523650407791, -0.31770798563957214, 0.15545983612537384, -0.11993886530399323, -0.36654940247535706, -0.36813732981681824, 0.1454017460346222, 0.2627983093261719, 0.3802306354045868, -0.059689246118068695, 0.06708643585443497, -0.061199504882097244, -0.19583767652511597, 0.05281435325741768, -0.08949437737464905, -0.16413260996341705, 0.3026691973209381, -0.22698481380939484, 0.12764689326286316, 0.13528817892074585, 0.05715329945087433, 0.11984620243310928, 0.36280232667922974, 0.2032434493303299, -0.1840134859085083, -0.37553951144218445, 0.0625774934887886, -0.2645849287509918, 0.4228106737136841, 0.2957010865211487, 0.14088551700115204, 0.38271546363830566, -0.008944224566221237, 0.2218949794769287, -0.36490070819854736, -0.2945745587348938, -0.16017666459083557, -0.11438265442848206, 0.2509685754776001, 0.3615801930427551, -0.1595371812582016, -0.2476469874382019, -0.0595758855342865, -0.1263955533504486, 0.01934809982776642, 0.20292793214321136, 0.564559280872345, 0.05364493280649185, 0.18948319554328918, 0.03513288497924805, 0.3217781186103821, 0.05370338261127472, -0.5332468748092651, -0.17430879175662994, 0.21654155850410461, -0.3087959587574005, 0.14569512009620667, 0.3105257749557495, -0.33132481575012207, 0.30016353726387024, -0.11575452983379364, -0.3032854199409485, 0.003235284239053726, 0.12947288155555725, -0.17491813004016876, 0.1020323857665062, 0.07796085625886917, 0.09318621456623077, 0.10305051505565643, 0.19042691588401794, -0.07996858656406403, 0.06550367921590805, -0.03327682614326477, -0.0471925213932991, -0.4852258861064911, 0.025506744161248207, 0.1306857317686081, 0.07165280729532242, -0.009505707770586014, -0.08786729723215103, -0.05430324003100395, 0.19623038172721863, -0.22172144055366516, -0.11173627525568008, 0.11289937049150467, -0.02556486427783966, -0.27134576439857483, 0.21264249086380005, 0.5540770292282104, -0.3136104345321655, 0.20926660299301147, -0.19852595031261444, -0.13489246368408203, 0.16316398978233337, 0.3628429174423218, -0.11672976613044739, 0.07758884876966476, -0.3848821818828583, 0.32663196325302124, 0.6447769403457642, -0.5202509164810181, -0.5076894164085388, 0.2816503047943115, -0.22235773503780365, -0.07156785577535629, 0.5067175030708313, 0.24071186780929565, -0.06453140825033188, -0.11318308115005493, 0.047000452876091, 0.11700893193483353, 0.04672098159790039, -0.08850011229515076, -0.11148910969495773, -0.2450886219739914, 0.34023767709732056, 0.11302006244659424, -0.030680183321237564, 0.132091224193573, -0.029109269380569458, 0.16131937503814697, 0.3716829717159271, -0.29108256101608276, 0.15118561685085297, 0.28406065702438354, 0.30056604743003845, 0.03720933198928833, -0.14257794618606567, -0.11960314959287643, -0.2946610152721405, 0.23998597264289856, -0.09203151613473892, -0.020116493105888367, -0.3088812828063965, 0.00025285035371780396, -0.11309009790420532, -0.04091973975300789, -0.3876267373561859, -0.2783125340938568, 0.19049659371376038, 0.4527055025100708, 0.22734439373016357, -0.019652556627988815, -0.2899417579174042, 0.3173922300338745, -0.4826750159263611, 0.0027163736522197723, -0.6354678273200989, 0.12806381285190582, -0.09318923950195312, -0.14903524518013, -0.16396383941173553, 0.16392536461353302, 0.023550454527139664, -0.002848624950274825, -0.09274837374687195, 0.2124704122543335, -0.2380763739347458, 0.08131269365549088, 0.26387062668800354, 0.10236126184463501, 0.03820338845252991, -0.5709119439125061, 0.3375822603702545, 0.4717596769332886, 0.22842147946357727, 0.08941132575273514, -0.2693713903427124, 0.11830680072307587, -0.08254849910736084, 0.09030067175626755, 0.043727852404117584, 0.01720271445810795, 0.4666250944137573, 0.014810655266046524, 0.06998185068368912, -0.21279391646385193, 0.3003252446651459, 0.21648988127708435, 0.46217799186706543, -0.08612561970949173, -0.5364863872528076, 0.086130291223526, 0.2796008586883545, 0.06051979586482048, -0.09864357858896255, -0.010936349630355835, -0.35100460052490234, -0.04163479432463646, 0.0550631582736969, 0.22897912561893463, 0.28139108419418335, 0.2278088480234146, -0.06535398960113525, 0.08311121165752411, 0.09837377071380615, -0.26217979192733765, 0.18072940409183502, -0.25792741775512695, 0.10123033076524734, 0.269871324300766, 0.19980132579803467, 0.032053712755441666, -0.03564509004354477, -0.5033626556396484, 0.07217030227184296, 0.4454902410507202, -0.27034348249435425, -0.17282292246818542, -0.1369045525789261, -0.6854813694953918, 0.06065024435520172, 0.07638534903526306, -0.11054055392742157, -0.34322434663772583, -0.02222895622253418, 0.18776102364063263, 0.2145620882511139, -0.07199481129646301, -0.09505246579647064, 0.07554078102111816, 0.06864779442548752, -0.10940844565629959, -0.41933244466781616, -0.31495964527130127, -0.20777547359466553, 0.04685630276799202, 0.2707954943180084, -0.022806549444794655, 0.24694311618804932, -0.2802799940109253, 0.0639284998178482, -0.41443806886672974, -0.025489725172519684, 0.0234772227704525, 0.029297329485416412, -0.00047362223267555237, -0.034068748354911804, 0.48188817501068115, 0.4775170087814331, -0.062324706465005875, 0.3475897014141083, 0.0415470227599144, -0.17423024773597717, 0.2578030526638031, -0.21523967385292053, -0.23448698222637177, -0.04180895537137985, -0.5249141454696655, -0.35637030005455017, -0.32827454805374146, 0.08275279402732849, -0.13988062739372253, 0.0803992748260498, 0.318090558052063, -0.14647728204727173, 0.2062612771987915, -0.038883764296770096, 0.19586814939975739, -0.07545912265777588, -0.22561395168304443, 0.3382496237754822, -0.16139717400074005, -0.21599532663822174, 0.23609322309494019, 0.06490501016378403, 0.2100887894630432, -0.04124322906136513, -0.23060056567192078, 0.1721261888742447, -0.22556985914707184, 0.28446367383003235, -0.13050413131713867, -0.15192082524299622, 0.34998250007629395, -0.010504120960831642, -0.0014594737440347672, 0.004380516707897186, 0.011881504207849503, 0.11780828982591629, 0.010026469826698303, 0.5716208815574646, 0.20394280552864075, 0.41344475746154785, 0.14451958239078522, 0.37235599756240845, 0.44866451621055603, 0.02675904519855976, 0.32839086651802063, -0.262139230966568, 0.06690724194049835, -0.0658818855881691, -0.3156571686267853, 0.11945406347513199, 0.17475613951683044, -0.17332357168197632, -0.10797254741191864, 0.13868209719657898, 0.3889438211917877, -0.33151957392692566, -0.13302285969257355, -0.28740715980529785, -0.08548354357481003, -0.1463780254125595, 0.020249327644705772, 0.011118672788143158, 0.12911128997802734, 0.20937472581863403, -0.34255439043045044, -0.3424656093120575, -0.11213222146034241, 0.4633781313896179, 0.002335883677005768, 0.15231482684612274, -0.27443593740463257, -0.28418031334877014, -0.5618230700492859, 0.4806254506111145, 0.1180061548948288, 0.21206675469875336, -0.27588602900505066, 0.12500113248825073, 0.26634669303894043, 0.022575749084353447, 0.32832586765289307, -0.3004830777645111, 0.2563643455505371, 0.08078660070896149, -0.09892930090427399, -0.23587936162948608, 0.1696384996175766, -0.1501612663269043, 0.19850066304206848, 0.33587101101875305, 0.19395439326763153, -0.16084104776382446, -0.11107290536165237, -0.016944006085395813, 0.10838137567043304, -0.17496953904628754, -0.2175944596529007, -0.017784323543310165, -0.0037048086524009705, -0.1634150892496109, -0.32848215103149414, -0.1715681403875351, 0.2472757250070572, 0.13296085596084595, 0.335891455411911, 0.04687159135937691, -0.07427210360765457, 0.17847181856632233, 0.22041521966457367, 0.3392874002456665, -0.0961354672908783, -0.15330930054187775, 0.16526304185390472, 0.2498362809419632, 0.07175793498754501, 0.49867454171180725, 0.28984537720680237, -0.36226436495780945, -0.014101285487413406, -0.22840815782546997, 0.12723538279533386, 0.29033511877059937, -0.2587491273880005, 0.10405394434928894, 0.03436683118343353, -0.007826589047908783, 0.061695680022239685, -0.09590508788824081, 0.2513183355331421, 0.031543757766485214, -0.12594200670719147, -0.398691326379776, 0.32729804515838623, 0.1011691689491272, 0.003973290324211121, 0.45400190353393555, -0.14631932973861694, -0.22761747241020203, 0.023735394701361656, 0.05723189190030098, 0.9369779825210571, 0.12992817163467407, 0.22923724353313446, 0.3930836021900177, 0.16831237077713013, 0.5259998440742493, -0.3035131096839905, 0.3688012957572937, -0.3871549367904663, -0.08039584755897522, -0.002947915345430374, -0.11370426416397095, -0.011663682758808136, -0.1774902492761612, -0.0803154706954956, 0.23407773673534393, -0.15633408725261688, 0.006111137568950653, -0.09859137237071991, 0.26289814710617065, -0.33494704961776733, -0.09046958386898041, -0.46061474084854126, 0.08475244790315628, -0.20800413191318512, 0.334980309009552, -0.10503844916820526, -0.1498464196920395, -0.385303258895874, 0.1126156598329544, -0.420208215713501, 0.3758964240550995, -0.08597476780414581, 0.12252435088157654, -0.07361695170402527, -0.3276987671852112, -0.05533352494239807, 0.6137100458145142, 0.5807362198829651, 0.05012687295675278, -0.3391721248626709, 0.007756650447845459, -0.2113005518913269, -0.023091383278369904, -0.0938933789730072, 0.15532897412776947, 0.33820483088493347, -0.04085526615381241, -0.30917805433273315, 0.5142857432365417, -0.11258649826049805, 0.01624208688735962, -0.23662205040454865, -0.25668203830718994, -0.24029575288295746, -0.04322406277060509, -0.12301504611968994, 0.01992608606815338, -0.12974657118320465, 0.025179501622915268, 0.09975022077560425, 0.012665413320064545, -0.23290427029132843, -0.24648424983024597, -0.13393037021160126, -0.1444866955280304, 0.18300418555736542, 0.32538139820098877, 0.047446150332689285, 0.1761953979730606, 0.6140016317367554, 0.22167591750621796, -0.050169289112091064, -0.24814751744270325, 0.06540375202894211, 0.15567605197429657, -0.25567328929901123, -0.2123810350894928, 0.10414337366819382, 0.1746923178434372, 0.09654483944177628, 0.27749502658843994, 0.1864689290523529, -0.006256610155105591, -0.14283424615859985, -0.4487021267414093, -0.542247474193573, 0.1517488956451416, -0.19254685938358307, 0.13901546597480774, -0.1516934037208557, 0.2596281170845032, 0.11874733865261078, -0.26524898409843445, -0.33702796697616577, -0.09570232778787613, -0.323411762714386, -0.005745381116867065, 0.39344823360443115, 0.059053368866443634, 0.196878120303154, 0.043673038482666016, 0.1371064931154251, -0.03669913858175278, -0.2849010229110718, -0.15982764959335327, 0.04887765273451805, 0.10002413392066956, 0.13717149198055267, -0.40278685092926025, -0.11658169329166412, -0.28826242685317993, -0.08812689781188965, -0.16044285893440247, 0.05028693005442619, 0.29853391647338867, 0.03162441775202751, 0.2532195746898651, -0.31334492564201355, 0.05800839141011238, -0.21368736028671265, 0.15737824141979218, -0.1528380811214447, 0.1841050684452057, 0.05986042693257332, 0.12613007426261902, 0.017675742506980896, 0.009883783757686615, -0.5507248044013977, -0.06257525831460953, -0.090798020362854, 0.18297600746154785, 0.1152569130063057, -0.4182227849960327, 0.22805601358413696, -0.14199551939964294, 0.17141956090927124, 0.1962626576423645, -0.03331763669848442, 0.2832372188568115, 0.181142657995224, 0.18628524243831635, -0.23866689205169678, 0.20860789716243744, 0.21577955782413483, 0.09463845193386078, -0.09388956427574158, 0.04307936131954193, -0.04945437237620354, -0.02566228061914444, 0.26920613646507263, -0.10244958102703094, 0.3867933452129364, -0.40722113847732544, -0.05248976871371269, 0.1194659024477005, 0.0713922530412674, -0.05252337083220482, 0.11530591547489166, 0.3662636876106262, 0.16459202766418457, 0.782757043838501, -0.24703267216682434, 0.14657151699066162, -0.18498504161834717, 0.32810965180397034, -0.2487824261188507, -0.33459770679473877, -0.3359786868095398, 0.11348118633031845, -0.29427021741867065, -0.1261742115020752, -0.1852845549583435, 0.29954174160957336, -0.1018143743276596, 0.14273476600646973, -0.23577992618083954, 0.3921544551849365, -0.148835688829422, -0.04539507254958153, -0.5625843405723572, -0.22663289308547974, 0.10585501790046692, -0.17970573902130127, 0.025090841576457024, 0.015994630753993988, -0.009149357676506042, 0.13094379007816315, -0.2652379870414734, -0.40185344219207764, 0.30937230587005615, 0.40477633476257324, -0.016356661915779114, -0.27932941913604736, 0.41067442297935486, 0.2905891239643097, 0.0659504160284996, 0.20533022284507751, 0.31929323077201843, 0.4276030361652374, 0.4022083282470703, 0.0665070116519928, 0.20000682771205902, -0.10693976283073425, -0.25286224484443665, -0.013881783932447433, 0.052523091435432434, 0.3014645278453827, 0.05845613032579422, 0.31253066658973694, 0.14970752596855164, -0.006234737113118172, 0.14302192628383636, 0.04569154232740402, 0.3462030589580536, 0.09222429990768433, -0.01502034068107605, -0.1465473622083664, -0.24552851915359497, -0.18946360051631927, -0.12856656312942505, -0.558147132396698, 0.13780371844768524, 0.21281994879245758, 0.2150428146123886, 0.2549576759338379, -0.0042096273973584175, 0.049009568989276886, -0.06657010316848755, 0.13284362852573395, 0.5250779986381531, -0.0340118333697319, -0.27033567428588867, -0.32096928358078003, -0.5026277303695679, 0.16537794470787048, -0.025284145027399063, -0.4208657443523407, -0.24853581190109253, 0.12115363776683807, -0.1755252182483673, -0.23555682599544525, 0.2410193681716919, 0.101404570043087, 0.06959471106529236, 0.3392389118671417, -0.22689945995807648, -0.33274638652801514, 0.26964038610458374, 0.05312708020210266, -0.1709013134241104, -0.17286834120750427, -0.09464499354362488, -0.14867743849754333, 0.04900039732456207, -0.046039119362831116, 0.3904174566268921, 0.00711417943239212, 0.2057296186685562, 0.19779503345489502, 0.19981157779693604, 0.48119598627090454, 0.0680660530924797, -0.04043828696012497, -0.3731038272380829, -0.1117364838719368, -0.20125921070575714, -0.041572049260139465, 0.3111209571361542, 0.5081038475036621, -0.3583824038505554, -0.00834466889500618, -0.3815689980983734, 0.12782993912696838, -0.1665935218334198, -0.017484026029706, -0.2955934703350067, 0.3703538477420807, -0.2835766077041626, 0.2372780591249466, 0.16417348384857178, -0.045550912618637085, -0.029033198952674866, 0.10304367542266846, -0.33316755294799805, 0.003624744713306427, 0.5232219696044922, -0.21019817888736725, -0.23654082417488098, 0.09007922559976578, 0.1933175027370453, 0.11350397765636444, 0.010346682742238045, -0.33786875009536743, 0.17566832900047302, 0.18242181837558746, -0.005352132022380829, -0.24062027037143707, -0.0816912055015564, 0.0652153268456459, 0.11355289816856384, -0.06532711535692215, -0.14270718395709991, 0.15939897298812866, -0.07285131514072418, -0.013610651716589928, -0.08653853833675385 ]
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 ?
[ -0.1408686488866806, -0.18925881385803223, -0.03353646397590637, 0.4693673253059387, 0.40544581413269043, 0.11200030893087387, 0.056036077439785004, 0.35449865460395813, -0.10948893427848816, 0.2927756607532501, -0.14041008055210114, 0.16212110221385956, 0.05022745206952095, -0.02804567478597164, 0.05054093897342682, -0.03532540053129196, -0.169126957654953, 0.15102019906044006, 0.1320728361606598, 0.005323596298694611, -0.1707707643508911, 0.045582592487335205, -0.26906365156173706, 0.25874191522598267, -0.28670650720596313, -0.08528319746255875, 0.0914461761713028, 0.38597795367240906, -0.30817025899887085, -0.5349300503730774, 0.045082367956638336, -0.14433495700359344, 0.24077430367469788, 0.19576658308506012, -0.0001110580051317811, 0.04933277517557144, 0.26291391253471375, -0.0573488250374794, -0.5286889672279358, -0.14541836082935333, -0.3568848669528961, -0.12314407527446747, -0.014699378982186317, -0.3132312595844269, -0.1287817358970642, -0.22096048295497894, 0.40224409103393555, -0.3689234256744385, 0.2718048691749573, 0.31527629494667053, 0.20390821993350983, 0.20004351437091827, 0.12182684242725372, -0.05471378192305565, 0.04933743178844452, 0.13519372045993805, -0.1314477175474167, -0.013969961553812027, 0.23928481340408325, 0.1932622492313385, 0.16366678476333618, 0.45788297057151794, -0.1711953580379486, -0.11164624989032745, 0.25520727038383484, 0.243842214345932, -0.030298396944999695, -0.16049769520759583, 0.32099449634552, 0.23313367366790771, 0.4437568187713623, -0.4332887530326843, -0.43349897861480713, -0.11148454993963242, -0.0038183475844562054, -0.26497241854667664, 0.24886426329612732, 0.10019523650407791, -0.31770798563957214, 0.15545983612537384, -0.11993886530399323, -0.36654940247535706, -0.36813732981681824, 0.1454017460346222, 0.2627983093261719, 0.3802306354045868, -0.059689246118068695, 0.06708643585443497, -0.061199504882097244, -0.19583767652511597, 0.05281435325741768, -0.08949437737464905, -0.16413260996341705, 0.3026691973209381, -0.22698481380939484, 0.12764689326286316, 0.13528817892074585, 0.05715329945087433, 0.11984620243310928, 0.36280232667922974, 0.2032434493303299, -0.1840134859085083, -0.37553951144218445, 0.0625774934887886, -0.2645849287509918, 0.4228106737136841, 0.2957010865211487, 0.14088551700115204, 0.38271546363830566, -0.008944224566221237, 0.2218949794769287, -0.36490070819854736, -0.2945745587348938, -0.16017666459083557, -0.11438265442848206, 0.2509685754776001, 0.3615801930427551, -0.1595371812582016, -0.2476469874382019, -0.0595758855342865, -0.1263955533504486, 0.01934809982776642, 0.20292793214321136, 0.564559280872345, 0.05364493280649185, 0.18948319554328918, 0.03513288497924805, 0.3217781186103821, 0.05370338261127472, -0.5332468748092651, -0.17430879175662994, 0.21654155850410461, -0.3087959587574005, 0.14569512009620667, 0.3105257749557495, -0.33132481575012207, 0.30016353726387024, -0.11575452983379364, -0.3032854199409485, 0.003235284239053726, 0.12947288155555725, -0.17491813004016876, 0.1020323857665062, 0.07796085625886917, 0.09318621456623077, 0.10305051505565643, 0.19042691588401794, -0.07996858656406403, 0.06550367921590805, -0.03327682614326477, -0.0471925213932991, -0.4852258861064911, 0.025506744161248207, 0.1306857317686081, 0.07165280729532242, -0.009505707770586014, -0.08786729723215103, -0.05430324003100395, 0.19623038172721863, -0.22172144055366516, -0.11173627525568008, 0.11289937049150467, -0.02556486427783966, -0.27134576439857483, 0.21264249086380005, 0.5540770292282104, -0.3136104345321655, 0.20926660299301147, -0.19852595031261444, -0.13489246368408203, 0.16316398978233337, 0.3628429174423218, -0.11672976613044739, 0.07758884876966476, -0.3848821818828583, 0.32663196325302124, 0.6447769403457642, -0.5202509164810181, -0.5076894164085388, 0.2816503047943115, -0.22235773503780365, -0.07156785577535629, 0.5067175030708313, 0.24071186780929565, -0.06453140825033188, -0.11318308115005493, 0.047000452876091, 0.11700893193483353, 0.04672098159790039, -0.08850011229515076, -0.11148910969495773, -0.2450886219739914, 0.34023767709732056, 0.11302006244659424, -0.030680183321237564, 0.132091224193573, -0.029109269380569458, 0.16131937503814697, 0.3716829717159271, -0.29108256101608276, 0.15118561685085297, 0.28406065702438354, 0.30056604743003845, 0.03720933198928833, -0.14257794618606567, -0.11960314959287643, -0.2946610152721405, 0.23998597264289856, -0.09203151613473892, -0.020116493105888367, -0.3088812828063965, 0.00025285035371780396, -0.11309009790420532, -0.04091973975300789, -0.3876267373561859, -0.2783125340938568, 0.19049659371376038, 0.4527055025100708, 0.22734439373016357, -0.019652556627988815, -0.2899417579174042, 0.3173922300338745, -0.4826750159263611, 0.0027163736522197723, -0.6354678273200989, 0.12806381285190582, -0.09318923950195312, -0.14903524518013, -0.16396383941173553, 0.16392536461353302, 0.023550454527139664, -0.002848624950274825, -0.09274837374687195, 0.2124704122543335, -0.2380763739347458, 0.08131269365549088, 0.26387062668800354, 0.10236126184463501, 0.03820338845252991, -0.5709119439125061, 0.3375822603702545, 0.4717596769332886, 0.22842147946357727, 0.08941132575273514, -0.2693713903427124, 0.11830680072307587, -0.08254849910736084, 0.09030067175626755, 0.043727852404117584, 0.01720271445810795, 0.4666250944137573, 0.014810655266046524, 0.06998185068368912, -0.21279391646385193, 0.3003252446651459, 0.21648988127708435, 0.46217799186706543, -0.08612561970949173, -0.5364863872528076, 0.086130291223526, 0.2796008586883545, 0.06051979586482048, -0.09864357858896255, -0.010936349630355835, -0.35100460052490234, -0.04163479432463646, 0.0550631582736969, 0.22897912561893463, 0.28139108419418335, 0.2278088480234146, -0.06535398960113525, 0.08311121165752411, 0.09837377071380615, -0.26217979192733765, 0.18072940409183502, -0.25792741775512695, 0.10123033076524734, 0.269871324300766, 0.19980132579803467, 0.032053712755441666, -0.03564509004354477, -0.5033626556396484, 0.07217030227184296, 0.4454902410507202, -0.27034348249435425, -0.17282292246818542, -0.1369045525789261, -0.6854813694953918, 0.06065024435520172, 0.07638534903526306, -0.11054055392742157, -0.34322434663772583, -0.02222895622253418, 0.18776102364063263, 0.2145620882511139, -0.07199481129646301, -0.09505246579647064, 0.07554078102111816, 0.06864779442548752, -0.10940844565629959, -0.41933244466781616, -0.31495964527130127, -0.20777547359466553, 0.04685630276799202, 0.2707954943180084, -0.022806549444794655, 0.24694311618804932, -0.2802799940109253, 0.0639284998178482, -0.41443806886672974, -0.025489725172519684, 0.0234772227704525, 0.029297329485416412, -0.00047362223267555237, -0.034068748354911804, 0.48188817501068115, 0.4775170087814331, -0.062324706465005875, 0.3475897014141083, 0.0415470227599144, -0.17423024773597717, 0.2578030526638031, -0.21523967385292053, -0.23448698222637177, -0.04180895537137985, -0.5249141454696655, -0.35637030005455017, -0.32827454805374146, 0.08275279402732849, -0.13988062739372253, 0.0803992748260498, 0.318090558052063, -0.14647728204727173, 0.2062612771987915, -0.038883764296770096, 0.19586814939975739, -0.07545912265777588, -0.22561395168304443, 0.3382496237754822, -0.16139717400074005, -0.21599532663822174, 0.23609322309494019, 0.06490501016378403, 0.2100887894630432, -0.04124322906136513, -0.23060056567192078, 0.1721261888742447, -0.22556985914707184, 0.28446367383003235, -0.13050413131713867, -0.15192082524299622, 0.34998250007629395, -0.010504120960831642, -0.0014594737440347672, 0.004380516707897186, 0.011881504207849503, 0.11780828982591629, 0.010026469826698303, 0.5716208815574646, 0.20394280552864075, 0.41344475746154785, 0.14451958239078522, 0.37235599756240845, 0.44866451621055603, 0.02675904519855976, 0.32839086651802063, -0.262139230966568, 0.06690724194049835, -0.0658818855881691, -0.3156571686267853, 0.11945406347513199, 0.17475613951683044, -0.17332357168197632, -0.10797254741191864, 0.13868209719657898, 0.3889438211917877, -0.33151957392692566, -0.13302285969257355, -0.28740715980529785, -0.08548354357481003, -0.1463780254125595, 0.020249327644705772, 0.011118672788143158, 0.12911128997802734, 0.20937472581863403, -0.34255439043045044, -0.3424656093120575, -0.11213222146034241, 0.4633781313896179, 0.002335883677005768, 0.15231482684612274, -0.27443593740463257, -0.28418031334877014, -0.5618230700492859, 0.4806254506111145, 0.1180061548948288, 0.21206675469875336, -0.27588602900505066, 0.12500113248825073, 0.26634669303894043, 0.022575749084353447, 0.32832586765289307, -0.3004830777645111, 0.2563643455505371, 0.08078660070896149, -0.09892930090427399, -0.23587936162948608, 0.1696384996175766, -0.1501612663269043, 0.19850066304206848, 0.33587101101875305, 0.19395439326763153, -0.16084104776382446, -0.11107290536165237, -0.016944006085395813, 0.10838137567043304, -0.17496953904628754, -0.2175944596529007, -0.017784323543310165, -0.0037048086524009705, -0.1634150892496109, -0.32848215103149414, -0.1715681403875351, 0.2472757250070572, 0.13296085596084595, 0.335891455411911, 0.04687159135937691, -0.07427210360765457, 0.17847181856632233, 0.22041521966457367, 0.3392874002456665, -0.0961354672908783, -0.15330930054187775, 0.16526304185390472, 0.2498362809419632, 0.07175793498754501, 0.49867454171180725, 0.28984537720680237, -0.36226436495780945, -0.014101285487413406, -0.22840815782546997, 0.12723538279533386, 0.29033511877059937, -0.2587491273880005, 0.10405394434928894, 0.03436683118343353, -0.007826589047908783, 0.061695680022239685, -0.09590508788824081, 0.2513183355331421, 0.031543757766485214, -0.12594200670719147, -0.398691326379776, 0.32729804515838623, 0.1011691689491272, 0.003973290324211121, 0.45400190353393555, -0.14631932973861694, -0.22761747241020203, 0.023735394701361656, 0.05723189190030098, 0.9369779825210571, 0.12992817163467407, 0.22923724353313446, 0.3930836021900177, 0.16831237077713013, 0.5259998440742493, -0.3035131096839905, 0.3688012957572937, -0.3871549367904663, -0.08039584755897522, -0.002947915345430374, -0.11370426416397095, -0.011663682758808136, -0.1774902492761612, -0.0803154706954956, 0.23407773673534393, -0.15633408725261688, 0.006111137568950653, -0.09859137237071991, 0.26289814710617065, -0.33494704961776733, -0.09046958386898041, -0.46061474084854126, 0.08475244790315628, -0.20800413191318512, 0.334980309009552, -0.10503844916820526, -0.1498464196920395, -0.385303258895874, 0.1126156598329544, -0.420208215713501, 0.3758964240550995, -0.08597476780414581, 0.12252435088157654, -0.07361695170402527, -0.3276987671852112, -0.05533352494239807, 0.6137100458145142, 0.5807362198829651, 0.05012687295675278, -0.3391721248626709, 0.007756650447845459, -0.2113005518913269, -0.023091383278369904, -0.0938933789730072, 0.15532897412776947, 0.33820483088493347, -0.04085526615381241, -0.30917805433273315, 0.5142857432365417, -0.11258649826049805, 0.01624208688735962, -0.23662205040454865, -0.25668203830718994, -0.24029575288295746, -0.04322406277060509, -0.12301504611968994, 0.01992608606815338, -0.12974657118320465, 0.025179501622915268, 0.09975022077560425, 0.012665413320064545, -0.23290427029132843, -0.24648424983024597, -0.13393037021160126, -0.1444866955280304, 0.18300418555736542, 0.32538139820098877, 0.047446150332689285, 0.1761953979730606, 0.6140016317367554, 0.22167591750621796, -0.050169289112091064, -0.24814751744270325, 0.06540375202894211, 0.15567605197429657, -0.25567328929901123, -0.2123810350894928, 0.10414337366819382, 0.1746923178434372, 0.09654483944177628, 0.27749502658843994, 0.1864689290523529, -0.006256610155105591, -0.14283424615859985, -0.4487021267414093, -0.542247474193573, 0.1517488956451416, -0.19254685938358307, 0.13901546597480774, -0.1516934037208557, 0.2596281170845032, 0.11874733865261078, -0.26524898409843445, -0.33702796697616577, -0.09570232778787613, -0.323411762714386, -0.005745381116867065, 0.39344823360443115, 0.059053368866443634, 0.196878120303154, 0.043673038482666016, 0.1371064931154251, -0.03669913858175278, -0.2849010229110718, -0.15982764959335327, 0.04887765273451805, 0.10002413392066956, 0.13717149198055267, -0.40278685092926025, -0.11658169329166412, -0.28826242685317993, -0.08812689781188965, -0.16044285893440247, 0.05028693005442619, 0.29853391647338867, 0.03162441775202751, 0.2532195746898651, -0.31334492564201355, 0.05800839141011238, -0.21368736028671265, 0.15737824141979218, -0.1528380811214447, 0.1841050684452057, 0.05986042693257332, 0.12613007426261902, 0.017675742506980896, 0.009883783757686615, -0.5507248044013977, -0.06257525831460953, -0.090798020362854, 0.18297600746154785, 0.1152569130063057, -0.4182227849960327, 0.22805601358413696, -0.14199551939964294, 0.17141956090927124, 0.1962626576423645, -0.03331763669848442, 0.2832372188568115, 0.181142657995224, 0.18628524243831635, -0.23866689205169678, 0.20860789716243744, 0.21577955782413483, 0.09463845193386078, -0.09388956427574158, 0.04307936131954193, -0.04945437237620354, -0.02566228061914444, 0.26920613646507263, -0.10244958102703094, 0.3867933452129364, -0.40722113847732544, -0.05248976871371269, 0.1194659024477005, 0.0713922530412674, -0.05252337083220482, 0.11530591547489166, 0.3662636876106262, 0.16459202766418457, 0.782757043838501, -0.24703267216682434, 0.14657151699066162, -0.18498504161834717, 0.32810965180397034, -0.2487824261188507, -0.33459770679473877, -0.3359786868095398, 0.11348118633031845, -0.29427021741867065, -0.1261742115020752, -0.1852845549583435, 0.29954174160957336, -0.1018143743276596, 0.14273476600646973, -0.23577992618083954, 0.3921544551849365, -0.148835688829422, -0.04539507254958153, -0.5625843405723572, -0.22663289308547974, 0.10585501790046692, -0.17970573902130127, 0.025090841576457024, 0.015994630753993988, -0.009149357676506042, 0.13094379007816315, -0.2652379870414734, -0.40185344219207764, 0.30937230587005615, 0.40477633476257324, -0.016356661915779114, -0.27932941913604736, 0.41067442297935486, 0.2905891239643097, 0.0659504160284996, 0.20533022284507751, 0.31929323077201843, 0.4276030361652374, 0.4022083282470703, 0.0665070116519928, 0.20000682771205902, -0.10693976283073425, -0.25286224484443665, -0.013881783932447433, 0.052523091435432434, 0.3014645278453827, 0.05845613032579422, 0.31253066658973694, 0.14970752596855164, -0.006234737113118172, 0.14302192628383636, 0.04569154232740402, 0.3462030589580536, 0.09222429990768433, -0.01502034068107605, -0.1465473622083664, -0.24552851915359497, -0.18946360051631927, -0.12856656312942505, -0.558147132396698, 0.13780371844768524, 0.21281994879245758, 0.2150428146123886, 0.2549576759338379, -0.0042096273973584175, 0.049009568989276886, -0.06657010316848755, 0.13284362852573395, 0.5250779986381531, -0.0340118333697319, -0.27033567428588867, -0.32096928358078003, -0.5026277303695679, 0.16537794470787048, -0.025284145027399063, -0.4208657443523407, -0.24853581190109253, 0.12115363776683807, -0.1755252182483673, -0.23555682599544525, 0.2410193681716919, 0.101404570043087, 0.06959471106529236, 0.3392389118671417, -0.22689945995807648, -0.33274638652801514, 0.26964038610458374, 0.05312708020210266, -0.1709013134241104, -0.17286834120750427, -0.09464499354362488, -0.14867743849754333, 0.04900039732456207, -0.046039119362831116, 0.3904174566268921, 0.00711417943239212, 0.2057296186685562, 0.19779503345489502, 0.19981157779693604, 0.48119598627090454, 0.0680660530924797, -0.04043828696012497, -0.3731038272380829, -0.1117364838719368, -0.20125921070575714, -0.041572049260139465, 0.3111209571361542, 0.5081038475036621, -0.3583824038505554, -0.00834466889500618, -0.3815689980983734, 0.12782993912696838, -0.1665935218334198, -0.017484026029706, -0.2955934703350067, 0.3703538477420807, -0.2835766077041626, 0.2372780591249466, 0.16417348384857178, -0.045550912618637085, -0.029033198952674866, 0.10304367542266846, -0.33316755294799805, 0.003624744713306427, 0.5232219696044922, -0.21019817888736725, -0.23654082417488098, 0.09007922559976578, 0.1933175027370453, 0.11350397765636444, 0.010346682742238045, -0.33786875009536743, 0.17566832900047302, 0.18242181837558746, -0.005352132022380829, -0.24062027037143707, -0.0816912055015564, 0.0652153268456459, 0.11355289816856384, -0.06532711535692215, -0.14270718395709991, 0.15939897298812866, -0.07285131514072418, -0.013610651716589928, -0.08653853833675385 ]
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!
[ -0.14572681486606598, -0.16974392533302307, -0.19722990691661835, 0.33974677324295044, -0.11949792504310608, -0.42426010966300964, -0.42039984464645386, 0.14767369627952576, -0.2700444161891937, 0.17250652611255646, -0.08286791294813156, 0.18038815259933472, 0.12033069878816605, 0.008093969896435738, -0.10038959980010986, -0.09797526895999908, 0.1656545251607895, 0.03882867097854614, 0.0924806147813797, -0.3720676898956299, 0.066559337079525, 0.5696004629135132, -0.03133261203765869, 0.21043483912944794, 0.05402218550443649, 0.2702386677265167, 0.2851133644580841, 0.30352964997291565, -0.1485796421766281, 0.028436988592147827, 0.12800218164920807, 0.030641354620456696, -0.02369792014360428, 0.22214922308921814, -0.00010886789095820859, -0.2347189486026764, 0.10260987281799316, -0.00316624715924263, -0.27572283148765564, -0.3026086688041687, -0.0345572903752327, 0.06436429917812347, -0.12292841076850891, -0.07039455324411392, 0.08584661036729813, -0.04489266499876976, 0.04877201467752457, 0.2095426619052887, 0.4293983578681946, 0.14757375419139862, 0.22625458240509033, -0.3024369776248932, -0.18534868955612183, 0.024081934243440628, 0.5763345956802368, 0.1626754105091095, -0.11448337137699127, 0.4472644627094269, 0.005699165165424347, -0.3526371419429779, -0.24019041657447815, -0.008564136922359467, 0.3430408537387848, -0.17255935072898865, 0.13255922496318817, 0.19315513968467712, 0.1485852599143982, -0.04594065248966217, -0.07953556627035141, 0.24472832679748535, -0.2702813148498535, 0.03221157193183899, -0.22717098891735077, -0.2778374254703522, -0.0763154923915863, -0.19976060092449188, 0.11689166724681854, -0.08144556730985641, 0.16428537666797638, 0.012247947975993156, 0.044132381677627563, 0.2954266369342804, 0.04963909834623337, 0.007164034992456436, -0.054055385291576385, -0.10784557461738586, 0.05084608495235443, 0.09846597909927368, 0.29527971148490906, -0.050543226301670074, -0.5163146257400513, 0.010772094130516052, -0.2593674659729004, 0.07496051490306854, -0.3022814691066742, -0.037855636328458786, 0.31449607014656067, 0.4659731090068817, -0.20087093114852905, 0.4996267557144165, 0.2691965401172638, -0.0934176966547966, -0.13220703601837158, -0.03944692760705948, -0.13782444596290588, 0.10813266038894653, -0.028946541249752045, 0.020693469792604446, 0.08811824768781662, 0.2246808260679245, -0.13358275592327118, -0.012008216232061386, 0.398222953081131, -0.5637694597244263, -0.47779831290245056, 0.10259105265140533, -0.2610342502593994, -0.43239712715148926, -0.6891927719116211, -0.06736093759536743, -0.25543132424354553, -0.11924251168966293, 0.09890831261873245, 0.22720104455947876, -0.1748848706483841, 0.29229775071144104, 0.046788617968559265, 0.1833907961845398, -0.28950414061546326, 0.019662335515022278, -0.18399952352046967, 0.06357942521572113, -0.3143925666809082, 0.13452425599098206, 0.17056594789028168, 0.11274605989456177, 0.3180578649044037, 0.2560538947582245, 0.27351248264312744, 0.03864358365535736, 0.3298662602901459, -0.28395089507102966, -0.022635363042354584, -0.21111632883548737, -0.21139101684093475, 0.19745370745658875, 0.027128886431455612, 0.09184889495372772, -0.008701689541339874, -0.07489361613988876, -0.18084082007408142, 0.14423638582229614, 0.46661776304244995, 0.20658016204833984, 0.1496511995792389, -0.11055412888526917, 0.0406920462846756, 0.1169372946023941, -0.24016308784484863, -0.5936912894248962, 0.20406046509742737, -0.13655175268650055, -0.12701159715652466, -0.05817777290940285, 0.10982012003660202, -0.03458242118358612, 0.09312103688716888, -0.11743132770061493, 0.3731987178325653, 0.06358182430267334, 0.26031294465065, 0.4534834623336792, -0.04083108901977539, 0.01186465471982956, -0.16285665333271027, -0.07088559120893478, 0.4665316641330719, -0.43853095173835754, -0.22229687869548798, 0.3354998826980591, -0.16507023572921753, -0.06518608331680298, 0.2390466332435608, 0.039275988936424255, -0.3886536955833435, -0.1660069078207016, 0.17745031416416168, 0.19752462208271027, 0.07779914885759354, -0.04839682951569557, -0.2970067858695984, 0.13495878875255585, 0.3166254758834839, 0.0654771476984024, -0.23182784020900726, -0.2993497848510742, -0.11140484362840652, -0.06439181417226791, 0.514197051525116, 0.13280721008777618, -0.2751377820968628, 0.10499174892902374, -0.2721591889858246, 0.2936933636665344, -0.05169518291950226, 0.1185731515288353, 0.4230370819568634, -0.02337534911930561, -0.7951251268386841, -0.16027691960334778, 0.43703293800354004, -0.05248725414276123, -0.2122921645641327, -0.47140875458717346, 0.24053315818309784, -0.23946473002433777, 0.22328287363052368, 0.055049411952495575, 0.18732194602489471, 0.03035423904657364, 0.02407148852944374, 0.09660723805427551, -0.23642680048942566, -0.15861909091472626, -0.17153595387935638, 0.20367786288261414, -0.22109945118427277, -0.007433035410940647, -0.010540366172790527, 0.3596077859401703, 0.31164130568504333, 0.08758017420768738, 0.03730412572622299, 0.26341819763183594, 0.050155505537986755, 0.0980498269200325, 0.3672504425048828, 0.1448051631450653, -0.0007640253752470016, 0.1889611780643463, -0.12266073375940323, 0.6879806518554688, 0.016462940722703934, 0.03225133568048477, -0.34003961086273193, 0.42591530084609985, -0.21703888475894928, 0.20069116353988647, 0.4093107581138611, -0.07173743844032288, -0.050097495317459106, 0.05778001621365547, -0.006799564231187105, -0.017152976244688034, 0.7467243671417236, 0.1952691674232483, 0.2502555549144745, 0.002956695854663849, -0.3220258355140686, 0.2596171498298645, 0.3991404175758362, -0.015342838130891323, -0.020852625370025635, -0.11995432525873184, 0.1937142014503479, -0.09338856488466263, -0.03672000765800476, 0.25073885917663574, 0.3810698390007019, 0.19513367116451263, 0.15570716559886932, -0.1311991810798645, -0.33823680877685547, -0.3438020944595337, 0.03513295203447342, -0.18210352957248688, 0.08744704723358154, 0.4963866174221039, 0.10160735994577408, 0.21628685295581818, -0.3020772635936737, -0.5422850847244263, -0.25364819169044495, -0.22231410443782806, 0.06132550165057182, 0.16164565086364746, -0.14391349256038666, -0.3678344488143921, -0.15366095304489136, -0.0919032171368599, -0.2918884754180908, -0.2732764482498169, 0.2300969362258911, 0.020874908193945885, -0.07249949127435684, -0.2189374715089798, 0.02323257550597191, 0.14874732494354248, -0.37039417028427124, 0.1783812940120697, -0.017374012619256973, -0.028042834252119064, -0.27953851222991943, 0.22281798720359802, -0.05679866671562195, -0.01930420473217964, -0.03147614002227783, -0.2791869640350342, -0.45083630084991455, 0.16968809068202972, -0.3762677013874054, -0.010960057377815247, 0.06464451551437378, -0.0032363757491111755, 0.05967199429869652, -0.2506176233291626, -0.07057526707649231, 0.14640071988105774, 0.09220800548791885, -0.1741306185722351, -0.35074377059936523, -0.08965122699737549, 0.03678468242287636, 0.014213903807103634, -0.19140614569187164, -0.27752187848091125, 0.07937449216842651, -0.14331994950771332, 0.6197946071624756, 0.3270139694213867, 0.005700867623090744, 0.3858785927295685, -0.2369554340839386, 0.00001177145168185234, 0.17174258828163147, 0.5837677121162415, -0.2605840563774109, -0.04442854970693588, 0.19860902428627014, -0.2997126579284668, -0.4421888589859009, -0.2730613946914673, 0.18830057978630066, 0.03232206031680107, -0.008429164066910744, -0.17861716449260712, 0.031004078686237335, -0.1309441179037094, -0.04899725317955017, -0.2130453884601593, -0.05581780523061752, 0.33782681822776794, -0.1836601346731186, -0.2421920895576477, -0.06447437405586243, 0.21673373878002167, 0.029940642416477203, 0.28654640913009644, 0.4599356949329376, -0.4840274751186371, 0.1646125614643097, -0.1400940716266632, 0.12755994498729706, 0.04408307000994682, 0.5117960572242737, 0.1533815860748291, 0.1332312375307083, 0.0008752048015594482, -0.14067688584327698, 0.09818042814731598, 0.4896940290927887, 0.056188441812992096, -0.14490284025669098, 0.3308490812778473, 0.20873358845710754, 0.24980415403842926, -0.11164984852075577, -0.04191560670733452, -0.07632186263799667, -0.00791233777999878, 0.10350729525089264, 0.07728653401136398, -0.05600050836801529, 0.017135661095380783, 0.07414527982473373, -0.0036139339208602905, -0.15379199385643005, 0.21246279776096344, 0.07286881655454636, 0.39130184054374695, -0.11124485731124878, -0.25944772362709045, -0.005773358047008514, -0.46897637844085693, -0.005584202706813812, 0.031853336840867996, -0.021542735397815704, -0.15841075778007507, -0.17286941409111023, 0.05907169356942177, -0.3485898971557617, 0.26234251260757446, 0.16069412231445312, 0.23712927103042603, 0.16907617449760437, -0.13415351510047913, -0.5272519588470459, -0.294148325920105, -0.3322208523750305, 0.18113453686237335, 0.596076488494873, 0.11514388769865036, -0.13822150230407715, -0.17282268404960632, 0.10999258607625961, -0.05185077711939812, -0.14824798703193665, 0.10066777467727661, -0.33127161860466003, 0.20241904258728027, 0.15132899582386017, -0.2806145250797272, 0.04583745449781418, 0.3753645718097687, -0.24381005764007568, 0.12480206787586212, 0.06615749001502991, -0.1499929428100586, 0.2123294323682785, 0.13495224714279175, 0.33930909633636475, -0.10402191430330276, -0.24397088587284088, -0.29319772124290466, 0.30135542154312134, -0.3514430522918701, 0.3701417148113251, 0.1212446391582489, 0.0003355368971824646, 0.30029138922691345, -0.09358739852905273, 0.10684163123369217, 0.3101273477077484, -0.18164154887199402, 0.1664067804813385, -0.6221522092819214, 0.11270289123058319, -0.44617894291877747, -0.2102094292640686, 0.36847415566444397, -0.07535791397094727, 0.2850716710090637, -0.1410284787416458, 0.2084536850452423, 0.23064105212688446, -0.47709769010543823, -0.07949845492839813, -0.16209423542022705, -0.19412687420845032, 0.2774372696876526, -0.02041754126548767, 0.7596277594566345, 0.4354732036590576, 0.5633991360664368, 0.23021692037582397, -0.1473512500524521, 0.38939782977104187, -0.5955854058265686, 0.19127970933914185, -0.34151583909988403, -0.37023502588272095, 0.09574323892593384, 0.10443316400051117, -0.06511325389146805, 0.21454881131649017, 0.20764991641044617, 0.2953759431838989, 0.39312154054641724, -0.09911176562309265, -0.10828026384115219, 0.17655311524868011, 0.12343072891235352, 0.18754130601882935, -0.2870902717113495, 0.1991366744041443, 0.34644365310668945, 0.004576649516820908, -0.05969361588358879, 0.0959683433175087, -0.7420675754547119, 0.02251587063074112, 0.04249122738838196, 0.10044645518064499, -0.09559989720582962, 0.35474252700805664, 0.11958041787147522, -0.2243277132511139, -0.04871850088238716, 0.3672390878200531, 0.15987490117549896, -0.015912584960460663, 0.00719662569463253, 0.21002444624900818, -0.3194081783294678, 0.32633107900619507, 0.242310032248497, -0.03880351036787033, 0.031407780945301056, -0.22021397948265076, -0.0787142813205719, -0.07076065987348557, 0.15572455525398254, -0.16848213970661163, -0.5484740734100342, 0.18215541541576385, -0.33470314741134644, 0.007207001093775034, -0.02991698868572712, 0.10068950057029724, -0.09960559010505676, -0.05343038588762283, 0.16725941002368927, -0.05686335265636444, -0.06265363842248917, -0.03488510102033615, 0.09307181090116501, -0.33285802602767944, -0.050463635474443436, 0.47442546486854553, -0.13673809170722961, -0.04600970447063446, 0.06487441807985306, -0.3305659294128418, 0.0352669283747673, -0.24317610263824463, 0.03892479091882706, -0.08075741678476334, -0.19715386629104614, -0.10430090129375458, -0.007580584846436977, -0.2291492074728012, 0.25272732973098755, 0.25843656063079834, 0.24956685304641724, 0.21303749084472656, 0.032852623611688614, -0.36308497190475464, -0.09832554310560226, 0.0810789093375206, 0.16799978911876678, -0.03473341092467308, -0.019618384540081024, -0.1595887839794159, 0.2588408589363098, 0.23063212633132935, -0.35463544726371765, -0.15017247200012207, -0.08997658640146255, 0.12681715190410614, -0.22102300822734833, -0.31873926520347595, 0.022486412897706032, 0.1468534916639328, 0.14308585226535797, -0.008150145411491394, -0.05782918632030487, -0.24771744012832642, -0.20516616106033325, 0.083225779235363, -0.04834723472595215, -0.09744089841842651, -0.13200169801712036, 0.21839922666549683, 0.1043625921010971, -0.14063486456871033, -0.2313222736120224, -0.13392606377601624, 0.3573988676071167, 0.12174542248249054, -0.08553825318813324, 0.1825152337551117, -0.1426304578781128, 0.1086663231253624, 0.01882341504096985, -0.02538543939590454, 0.07357704639434814, -0.026512712240219116, 0.10945796221494675, -0.20184220373630524, -0.2564111351966858, -0.07424164563417435, -0.28038954734802246, 0.14717337489128113, 0.10414642840623856, -0.05297304689884186, 0.008782848715782166, -0.18328946828842163, 0.2844215929508209, 0.36497512459754944, 0.11780503392219543, -0.09963665157556534, -0.013917770236730576, 0.22528287768363953, -0.22220194339752197, 0.09425365179777145, 0.4496179521083832, 0.043986815959215164, 0.06563544273376465, -0.028284482657909393, 0.15857842564582825, -0.20769783854484558, -0.018284879624843597, -0.05924851819872856, -0.274423748254776, -0.01816922426223755, 0.17280733585357666, -0.19119597971439362, 0.2515231966972351, 0.07363121956586838, 0.15538182854652405, -0.37914836406707764, 0.1871822029352188, 0.35076963901519775, -0.0025786757469177246, 0.1941520869731903, -0.2856433689594269, 0.14951317012310028, -0.02367575466632843, 0.09032042324542999, -0.2703600227832794, 0.34511232376098633, 0.0853983536362648, -0.02180592715740204, 0.08363571763038635, 0.16912135481834412, -0.10833384096622467, -0.3376132547855377, -0.004882588982582092, -0.08809277415275574, -0.07047510892152786, -0.544400155544281, -0.4664246141910553, -0.22646990418434143, -0.394583523273468, 0.0710502415895462, -0.16874070465564728, 0.3817867636680603, 0.02951917052268982, 0.24772146344184875, -0.09672556817531586, -0.18267740309238434, -0.46176448464393616, 0.01024707779288292, 0.4208807945251465, -0.04675701633095741, 0.17743246257305145, 0.21671605110168457, 0.41431334614753723, 0.35717928409576416, 0.1616116166114807, 0.2387458086013794, 0.1888272762298584, -0.23361444473266602, -0.2279309630393982, 0.03246691823005676, 0.08932045102119446, -0.19615799188613892, 0.17421609163284302, 0.25997745990753174, 0.043390270322561264, 0.3002178966999054, 0.1322067826986313, -0.2410818189382553, 0.07742059230804443, 0.26498717069625854, -0.003574555739760399, -0.21413002908229828, 0.29473719000816345, -0.03916162997484207, 0.018681995570659637, -0.3314741849899292, 0.3797486424446106, -0.5077441334724426, -0.08658870309591293, -0.5439391136169434, 0.15140368044376373, 0.17864581942558289, -0.12435458600521088, 0.10393225401639938, 0.26889368891716003, 0.41731470823287964, -0.12723563611507416, 0.33877450227737427, -0.28413742780685425, -0.4613204002380371, -0.2653636336326599, -0.04584982246160507, 0.03950878977775574, 0.27052927017211914, -0.4309377372264862, 0.28873559832572937, -0.29741546511650085, 0.051449984312057495, 0.27456027269363403, 0.4283304512500763, 0.24691930413246155, 0.06649784743785858, -0.33822938799858093, 0.1559208780527115, 0.19840247929096222, -0.08824513852596283, -0.11956954002380371, 0.1306329071521759, -0.17390617728233337, 0.019816752523183823, 0.10971526056528091, 0.0855555534362793, -0.3465362787246704, -0.28865334391593933, 0.4207552969455719, 0.3385229706764221, 0.275066614151001, 0.2680865228176117, -0.37258201837539673, -0.3853943943977356, 0.031004125252366066, -0.12375257909297943, -0.06270965933799744, 0.06228291243314743, -0.06584185361862183, 0.38764387369155884, 0.02506408654153347, 0.07477395981550217, 0.07386114448308945, 0.5041284561157227, 0.055955611169338226, 0.0017196849221363664, -0.22702951729297638, -0.026206813752651215, 0.09364626556634903, 0.0634227991104126, -0.6129947304725647, 0.407443642616272, 0.0005168020725250244, 0.18714407086372375, -0.23370829224586487, -0.25390002131462097, 0.3750455975532532, -0.35842156410217285, 0.2720431089401245, -0.0997019112110138, 0.34275418519973755, 0.22176191210746765, 0.29680517315864563, 0.04679156094789505, -0.10589517652988434, 0.07175469398498535, 0.29574763774871826, 0.1622447669506073, 0.23034629225730896, -0.2334200143814087, -0.08380287140607834, -0.0016767457127571106, 0.29279518127441406, 0.06492644548416138, -0.29197126626968384, 0.41343367099761963, -0.18410754203796387 ]
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 ^^
[ -0.14572681486606598, -0.16974392533302307, -0.19722990691661835, 0.33974677324295044, -0.11949792504310608, -0.42426010966300964, -0.42039984464645386, 0.14767369627952576, -0.2700444161891937, 0.17250652611255646, -0.08286791294813156, 0.18038815259933472, 0.12033069878816605, 0.008093969896435738, -0.10038959980010986, -0.09797526895999908, 0.1656545251607895, 0.03882867097854614, 0.0924806147813797, -0.3720676898956299, 0.066559337079525, 0.5696004629135132, -0.03133261203765869, 0.21043483912944794, 0.05402218550443649, 0.2702386677265167, 0.2851133644580841, 0.30352964997291565, -0.1485796421766281, 0.028436988592147827, 0.12800218164920807, 0.030641354620456696, -0.02369792014360428, 0.22214922308921814, -0.00010886789095820859, -0.2347189486026764, 0.10260987281799316, -0.00316624715924263, -0.27572283148765564, -0.3026086688041687, -0.0345572903752327, 0.06436429917812347, -0.12292841076850891, -0.07039455324411392, 0.08584661036729813, -0.04489266499876976, 0.04877201467752457, 0.2095426619052887, 0.4293983578681946, 0.14757375419139862, 0.22625458240509033, -0.3024369776248932, -0.18534868955612183, 0.024081934243440628, 0.5763345956802368, 0.1626754105091095, -0.11448337137699127, 0.4472644627094269, 0.005699165165424347, -0.3526371419429779, -0.24019041657447815, -0.008564136922359467, 0.3430408537387848, -0.17255935072898865, 0.13255922496318817, 0.19315513968467712, 0.1485852599143982, -0.04594065248966217, -0.07953556627035141, 0.24472832679748535, -0.2702813148498535, 0.03221157193183899, -0.22717098891735077, -0.2778374254703522, -0.0763154923915863, -0.19976060092449188, 0.11689166724681854, -0.08144556730985641, 0.16428537666797638, 0.012247947975993156, 0.044132381677627563, 0.2954266369342804, 0.04963909834623337, 0.007164034992456436, -0.054055385291576385, -0.10784557461738586, 0.05084608495235443, 0.09846597909927368, 0.29527971148490906, -0.050543226301670074, -0.5163146257400513, 0.010772094130516052, -0.2593674659729004, 0.07496051490306854, -0.3022814691066742, -0.037855636328458786, 0.31449607014656067, 0.4659731090068817, -0.20087093114852905, 0.4996267557144165, 0.2691965401172638, -0.0934176966547966, -0.13220703601837158, -0.03944692760705948, -0.13782444596290588, 0.10813266038894653, -0.028946541249752045, 0.020693469792604446, 0.08811824768781662, 0.2246808260679245, -0.13358275592327118, -0.012008216232061386, 0.398222953081131, -0.5637694597244263, -0.47779831290245056, 0.10259105265140533, -0.2610342502593994, -0.43239712715148926, -0.6891927719116211, -0.06736093759536743, -0.25543132424354553, -0.11924251168966293, 0.09890831261873245, 0.22720104455947876, -0.1748848706483841, 0.29229775071144104, 0.046788617968559265, 0.1833907961845398, -0.28950414061546326, 0.019662335515022278, -0.18399952352046967, 0.06357942521572113, -0.3143925666809082, 0.13452425599098206, 0.17056594789028168, 0.11274605989456177, 0.3180578649044037, 0.2560538947582245, 0.27351248264312744, 0.03864358365535736, 0.3298662602901459, -0.28395089507102966, -0.022635363042354584, -0.21111632883548737, -0.21139101684093475, 0.19745370745658875, 0.027128886431455612, 0.09184889495372772, -0.008701689541339874, -0.07489361613988876, -0.18084082007408142, 0.14423638582229614, 0.46661776304244995, 0.20658016204833984, 0.1496511995792389, -0.11055412888526917, 0.0406920462846756, 0.1169372946023941, -0.24016308784484863, -0.5936912894248962, 0.20406046509742737, -0.13655175268650055, -0.12701159715652466, -0.05817777290940285, 0.10982012003660202, -0.03458242118358612, 0.09312103688716888, -0.11743132770061493, 0.3731987178325653, 0.06358182430267334, 0.26031294465065, 0.4534834623336792, -0.04083108901977539, 0.01186465471982956, -0.16285665333271027, -0.07088559120893478, 0.4665316641330719, -0.43853095173835754, -0.22229687869548798, 0.3354998826980591, -0.16507023572921753, -0.06518608331680298, 0.2390466332435608, 0.039275988936424255, -0.3886536955833435, -0.1660069078207016, 0.17745031416416168, 0.19752462208271027, 0.07779914885759354, -0.04839682951569557, -0.2970067858695984, 0.13495878875255585, 0.3166254758834839, 0.0654771476984024, -0.23182784020900726, -0.2993497848510742, -0.11140484362840652, -0.06439181417226791, 0.514197051525116, 0.13280721008777618, -0.2751377820968628, 0.10499174892902374, -0.2721591889858246, 0.2936933636665344, -0.05169518291950226, 0.1185731515288353, 0.4230370819568634, -0.02337534911930561, -0.7951251268386841, -0.16027691960334778, 0.43703293800354004, -0.05248725414276123, -0.2122921645641327, -0.47140875458717346, 0.24053315818309784, -0.23946473002433777, 0.22328287363052368, 0.055049411952495575, 0.18732194602489471, 0.03035423904657364, 0.02407148852944374, 0.09660723805427551, -0.23642680048942566, -0.15861909091472626, -0.17153595387935638, 0.20367786288261414, -0.22109945118427277, -0.007433035410940647, -0.010540366172790527, 0.3596077859401703, 0.31164130568504333, 0.08758017420768738, 0.03730412572622299, 0.26341819763183594, 0.050155505537986755, 0.0980498269200325, 0.3672504425048828, 0.1448051631450653, -0.0007640253752470016, 0.1889611780643463, -0.12266073375940323, 0.6879806518554688, 0.016462940722703934, 0.03225133568048477, -0.34003961086273193, 0.42591530084609985, -0.21703888475894928, 0.20069116353988647, 0.4093107581138611, -0.07173743844032288, -0.050097495317459106, 0.05778001621365547, -0.006799564231187105, -0.017152976244688034, 0.7467243671417236, 0.1952691674232483, 0.2502555549144745, 0.002956695854663849, -0.3220258355140686, 0.2596171498298645, 0.3991404175758362, -0.015342838130891323, -0.020852625370025635, -0.11995432525873184, 0.1937142014503479, -0.09338856488466263, -0.03672000765800476, 0.25073885917663574, 0.3810698390007019, 0.19513367116451263, 0.15570716559886932, -0.1311991810798645, -0.33823680877685547, -0.3438020944595337, 0.03513295203447342, -0.18210352957248688, 0.08744704723358154, 0.4963866174221039, 0.10160735994577408, 0.21628685295581818, -0.3020772635936737, -0.5422850847244263, -0.25364819169044495, -0.22231410443782806, 0.06132550165057182, 0.16164565086364746, -0.14391349256038666, -0.3678344488143921, -0.15366095304489136, -0.0919032171368599, -0.2918884754180908, -0.2732764482498169, 0.2300969362258911, 0.020874908193945885, -0.07249949127435684, -0.2189374715089798, 0.02323257550597191, 0.14874732494354248, -0.37039417028427124, 0.1783812940120697, -0.017374012619256973, -0.028042834252119064, -0.27953851222991943, 0.22281798720359802, -0.05679866671562195, -0.01930420473217964, -0.03147614002227783, -0.2791869640350342, -0.45083630084991455, 0.16968809068202972, -0.3762677013874054, -0.010960057377815247, 0.06464451551437378, -0.0032363757491111755, 0.05967199429869652, -0.2506176233291626, -0.07057526707649231, 0.14640071988105774, 0.09220800548791885, -0.1741306185722351, -0.35074377059936523, -0.08965122699737549, 0.03678468242287636, 0.014213903807103634, -0.19140614569187164, -0.27752187848091125, 0.07937449216842651, -0.14331994950771332, 0.6197946071624756, 0.3270139694213867, 0.005700867623090744, 0.3858785927295685, -0.2369554340839386, 0.00001177145168185234, 0.17174258828163147, 0.5837677121162415, -0.2605840563774109, -0.04442854970693588, 0.19860902428627014, -0.2997126579284668, -0.4421888589859009, -0.2730613946914673, 0.18830057978630066, 0.03232206031680107, -0.008429164066910744, -0.17861716449260712, 0.031004078686237335, -0.1309441179037094, -0.04899725317955017, -0.2130453884601593, -0.05581780523061752, 0.33782681822776794, -0.1836601346731186, -0.2421920895576477, -0.06447437405586243, 0.21673373878002167, 0.029940642416477203, 0.28654640913009644, 0.4599356949329376, -0.4840274751186371, 0.1646125614643097, -0.1400940716266632, 0.12755994498729706, 0.04408307000994682, 0.5117960572242737, 0.1533815860748291, 0.1332312375307083, 0.0008752048015594482, -0.14067688584327698, 0.09818042814731598, 0.4896940290927887, 0.056188441812992096, -0.14490284025669098, 0.3308490812778473, 0.20873358845710754, 0.24980415403842926, -0.11164984852075577, -0.04191560670733452, -0.07632186263799667, -0.00791233777999878, 0.10350729525089264, 0.07728653401136398, -0.05600050836801529, 0.017135661095380783, 0.07414527982473373, -0.0036139339208602905, -0.15379199385643005, 0.21246279776096344, 0.07286881655454636, 0.39130184054374695, -0.11124485731124878, -0.25944772362709045, -0.005773358047008514, -0.46897637844085693, -0.005584202706813812, 0.031853336840867996, -0.021542735397815704, -0.15841075778007507, -0.17286941409111023, 0.05907169356942177, -0.3485898971557617, 0.26234251260757446, 0.16069412231445312, 0.23712927103042603, 0.16907617449760437, -0.13415351510047913, -0.5272519588470459, -0.294148325920105, -0.3322208523750305, 0.18113453686237335, 0.596076488494873, 0.11514388769865036, -0.13822150230407715, -0.17282268404960632, 0.10999258607625961, -0.05185077711939812, -0.14824798703193665, 0.10066777467727661, -0.33127161860466003, 0.20241904258728027, 0.15132899582386017, -0.2806145250797272, 0.04583745449781418, 0.3753645718097687, -0.24381005764007568, 0.12480206787586212, 0.06615749001502991, -0.1499929428100586, 0.2123294323682785, 0.13495224714279175, 0.33930909633636475, -0.10402191430330276, -0.24397088587284088, -0.29319772124290466, 0.30135542154312134, -0.3514430522918701, 0.3701417148113251, 0.1212446391582489, 0.0003355368971824646, 0.30029138922691345, -0.09358739852905273, 0.10684163123369217, 0.3101273477077484, -0.18164154887199402, 0.1664067804813385, -0.6221522092819214, 0.11270289123058319, -0.44617894291877747, -0.2102094292640686, 0.36847415566444397, -0.07535791397094727, 0.2850716710090637, -0.1410284787416458, 0.2084536850452423, 0.23064105212688446, -0.47709769010543823, -0.07949845492839813, -0.16209423542022705, -0.19412687420845032, 0.2774372696876526, -0.02041754126548767, 0.7596277594566345, 0.4354732036590576, 0.5633991360664368, 0.23021692037582397, -0.1473512500524521, 0.38939782977104187, -0.5955854058265686, 0.19127970933914185, -0.34151583909988403, -0.37023502588272095, 0.09574323892593384, 0.10443316400051117, -0.06511325389146805, 0.21454881131649017, 0.20764991641044617, 0.2953759431838989, 0.39312154054641724, -0.09911176562309265, -0.10828026384115219, 0.17655311524868011, 0.12343072891235352, 0.18754130601882935, -0.2870902717113495, 0.1991366744041443, 0.34644365310668945, 0.004576649516820908, -0.05969361588358879, 0.0959683433175087, -0.7420675754547119, 0.02251587063074112, 0.04249122738838196, 0.10044645518064499, -0.09559989720582962, 0.35474252700805664, 0.11958041787147522, -0.2243277132511139, -0.04871850088238716, 0.3672390878200531, 0.15987490117549896, -0.015912584960460663, 0.00719662569463253, 0.21002444624900818, -0.3194081783294678, 0.32633107900619507, 0.242310032248497, -0.03880351036787033, 0.031407780945301056, -0.22021397948265076, -0.0787142813205719, -0.07076065987348557, 0.15572455525398254, -0.16848213970661163, -0.5484740734100342, 0.18215541541576385, -0.33470314741134644, 0.007207001093775034, -0.02991698868572712, 0.10068950057029724, -0.09960559010505676, -0.05343038588762283, 0.16725941002368927, -0.05686335265636444, -0.06265363842248917, -0.03488510102033615, 0.09307181090116501, -0.33285802602767944, -0.050463635474443436, 0.47442546486854553, -0.13673809170722961, -0.04600970447063446, 0.06487441807985306, -0.3305659294128418, 0.0352669283747673, -0.24317610263824463, 0.03892479091882706, -0.08075741678476334, -0.19715386629104614, -0.10430090129375458, -0.007580584846436977, -0.2291492074728012, 0.25272732973098755, 0.25843656063079834, 0.24956685304641724, 0.21303749084472656, 0.032852623611688614, -0.36308497190475464, -0.09832554310560226, 0.0810789093375206, 0.16799978911876678, -0.03473341092467308, -0.019618384540081024, -0.1595887839794159, 0.2588408589363098, 0.23063212633132935, -0.35463544726371765, -0.15017247200012207, -0.08997658640146255, 0.12681715190410614, -0.22102300822734833, -0.31873926520347595, 0.022486412897706032, 0.1468534916639328, 0.14308585226535797, -0.008150145411491394, -0.05782918632030487, -0.24771744012832642, -0.20516616106033325, 0.083225779235363, -0.04834723472595215, -0.09744089841842651, -0.13200169801712036, 0.21839922666549683, 0.1043625921010971, -0.14063486456871033, -0.2313222736120224, -0.13392606377601624, 0.3573988676071167, 0.12174542248249054, -0.08553825318813324, 0.1825152337551117, -0.1426304578781128, 0.1086663231253624, 0.01882341504096985, -0.02538543939590454, 0.07357704639434814, -0.026512712240219116, 0.10945796221494675, -0.20184220373630524, -0.2564111351966858, -0.07424164563417435, -0.28038954734802246, 0.14717337489128113, 0.10414642840623856, -0.05297304689884186, 0.008782848715782166, -0.18328946828842163, 0.2844215929508209, 0.36497512459754944, 0.11780503392219543, -0.09963665157556534, -0.013917770236730576, 0.22528287768363953, -0.22220194339752197, 0.09425365179777145, 0.4496179521083832, 0.043986815959215164, 0.06563544273376465, -0.028284482657909393, 0.15857842564582825, -0.20769783854484558, -0.018284879624843597, -0.05924851819872856, -0.274423748254776, -0.01816922426223755, 0.17280733585357666, -0.19119597971439362, 0.2515231966972351, 0.07363121956586838, 0.15538182854652405, -0.37914836406707764, 0.1871822029352188, 0.35076963901519775, -0.0025786757469177246, 0.1941520869731903, -0.2856433689594269, 0.14951317012310028, -0.02367575466632843, 0.09032042324542999, -0.2703600227832794, 0.34511232376098633, 0.0853983536362648, -0.02180592715740204, 0.08363571763038635, 0.16912135481834412, -0.10833384096622467, -0.3376132547855377, -0.004882588982582092, -0.08809277415275574, -0.07047510892152786, -0.544400155544281, -0.4664246141910553, -0.22646990418434143, -0.394583523273468, 0.0710502415895462, -0.16874070465564728, 0.3817867636680603, 0.02951917052268982, 0.24772146344184875, -0.09672556817531586, -0.18267740309238434, -0.46176448464393616, 0.01024707779288292, 0.4208807945251465, -0.04675701633095741, 0.17743246257305145, 0.21671605110168457, 0.41431334614753723, 0.35717928409576416, 0.1616116166114807, 0.2387458086013794, 0.1888272762298584, -0.23361444473266602, -0.2279309630393982, 0.03246691823005676, 0.08932045102119446, -0.19615799188613892, 0.17421609163284302, 0.25997745990753174, 0.043390270322561264, 0.3002178966999054, 0.1322067826986313, -0.2410818189382553, 0.07742059230804443, 0.26498717069625854, -0.003574555739760399, -0.21413002908229828, 0.29473719000816345, -0.03916162997484207, 0.018681995570659637, -0.3314741849899292, 0.3797486424446106, -0.5077441334724426, -0.08658870309591293, -0.5439391136169434, 0.15140368044376373, 0.17864581942558289, -0.12435458600521088, 0.10393225401639938, 0.26889368891716003, 0.41731470823287964, -0.12723563611507416, 0.33877450227737427, -0.28413742780685425, -0.4613204002380371, -0.2653636336326599, -0.04584982246160507, 0.03950878977775574, 0.27052927017211914, -0.4309377372264862, 0.28873559832572937, -0.29741546511650085, 0.051449984312057495, 0.27456027269363403, 0.4283304512500763, 0.24691930413246155, 0.06649784743785858, -0.33822938799858093, 0.1559208780527115, 0.19840247929096222, -0.08824513852596283, -0.11956954002380371, 0.1306329071521759, -0.17390617728233337, 0.019816752523183823, 0.10971526056528091, 0.0855555534362793, -0.3465362787246704, -0.28865334391593933, 0.4207552969455719, 0.3385229706764221, 0.275066614151001, 0.2680865228176117, -0.37258201837539673, -0.3853943943977356, 0.031004125252366066, -0.12375257909297943, -0.06270965933799744, 0.06228291243314743, -0.06584185361862183, 0.38764387369155884, 0.02506408654153347, 0.07477395981550217, 0.07386114448308945, 0.5041284561157227, 0.055955611169338226, 0.0017196849221363664, -0.22702951729297638, -0.026206813752651215, 0.09364626556634903, 0.0634227991104126, -0.6129947304725647, 0.407443642616272, 0.0005168020725250244, 0.18714407086372375, -0.23370829224586487, -0.25390002131462097, 0.3750455975532532, -0.35842156410217285, 0.2720431089401245, -0.0997019112110138, 0.34275418519973755, 0.22176191210746765, 0.29680517315864563, 0.04679156094789505, -0.10589517652988434, 0.07175469398498535, 0.29574763774871826, 0.1622447669506073, 0.23034629225730896, -0.2334200143814087, -0.08380287140607834, -0.0016767457127571106, 0.29279518127441406, 0.06492644548416138, -0.29197126626968384, 0.41343367099761963, -0.18410754203796387 ]
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!
[ -0.14572681486606598, -0.16974392533302307, -0.19722990691661835, 0.33974677324295044, -0.11949792504310608, -0.42426010966300964, -0.42039984464645386, 0.14767369627952576, -0.2700444161891937, 0.17250652611255646, -0.08286791294813156, 0.18038815259933472, 0.12033069878816605, 0.008093969896435738, -0.10038959980010986, -0.09797526895999908, 0.1656545251607895, 0.03882867097854614, 0.0924806147813797, -0.3720676898956299, 0.066559337079525, 0.5696004629135132, -0.03133261203765869, 0.21043483912944794, 0.05402218550443649, 0.2702386677265167, 0.2851133644580841, 0.30352964997291565, -0.1485796421766281, 0.028436988592147827, 0.12800218164920807, 0.030641354620456696, -0.02369792014360428, 0.22214922308921814, -0.00010886789095820859, -0.2347189486026764, 0.10260987281799316, -0.00316624715924263, -0.27572283148765564, -0.3026086688041687, -0.0345572903752327, 0.06436429917812347, -0.12292841076850891, -0.07039455324411392, 0.08584661036729813, -0.04489266499876976, 0.04877201467752457, 0.2095426619052887, 0.4293983578681946, 0.14757375419139862, 0.22625458240509033, -0.3024369776248932, -0.18534868955612183, 0.024081934243440628, 0.5763345956802368, 0.1626754105091095, -0.11448337137699127, 0.4472644627094269, 0.005699165165424347, -0.3526371419429779, -0.24019041657447815, -0.008564136922359467, 0.3430408537387848, -0.17255935072898865, 0.13255922496318817, 0.19315513968467712, 0.1485852599143982, -0.04594065248966217, -0.07953556627035141, 0.24472832679748535, -0.2702813148498535, 0.03221157193183899, -0.22717098891735077, -0.2778374254703522, -0.0763154923915863, -0.19976060092449188, 0.11689166724681854, -0.08144556730985641, 0.16428537666797638, 0.012247947975993156, 0.044132381677627563, 0.2954266369342804, 0.04963909834623337, 0.007164034992456436, -0.054055385291576385, -0.10784557461738586, 0.05084608495235443, 0.09846597909927368, 0.29527971148490906, -0.050543226301670074, -0.5163146257400513, 0.010772094130516052, -0.2593674659729004, 0.07496051490306854, -0.3022814691066742, -0.037855636328458786, 0.31449607014656067, 0.4659731090068817, -0.20087093114852905, 0.4996267557144165, 0.2691965401172638, -0.0934176966547966, -0.13220703601837158, -0.03944692760705948, -0.13782444596290588, 0.10813266038894653, -0.028946541249752045, 0.020693469792604446, 0.08811824768781662, 0.2246808260679245, -0.13358275592327118, -0.012008216232061386, 0.398222953081131, -0.5637694597244263, -0.47779831290245056, 0.10259105265140533, -0.2610342502593994, -0.43239712715148926, -0.6891927719116211, -0.06736093759536743, -0.25543132424354553, -0.11924251168966293, 0.09890831261873245, 0.22720104455947876, -0.1748848706483841, 0.29229775071144104, 0.046788617968559265, 0.1833907961845398, -0.28950414061546326, 0.019662335515022278, -0.18399952352046967, 0.06357942521572113, -0.3143925666809082, 0.13452425599098206, 0.17056594789028168, 0.11274605989456177, 0.3180578649044037, 0.2560538947582245, 0.27351248264312744, 0.03864358365535736, 0.3298662602901459, -0.28395089507102966, -0.022635363042354584, -0.21111632883548737, -0.21139101684093475, 0.19745370745658875, 0.027128886431455612, 0.09184889495372772, -0.008701689541339874, -0.07489361613988876, -0.18084082007408142, 0.14423638582229614, 0.46661776304244995, 0.20658016204833984, 0.1496511995792389, -0.11055412888526917, 0.0406920462846756, 0.1169372946023941, -0.24016308784484863, -0.5936912894248962, 0.20406046509742737, -0.13655175268650055, -0.12701159715652466, -0.05817777290940285, 0.10982012003660202, -0.03458242118358612, 0.09312103688716888, -0.11743132770061493, 0.3731987178325653, 0.06358182430267334, 0.26031294465065, 0.4534834623336792, -0.04083108901977539, 0.01186465471982956, -0.16285665333271027, -0.07088559120893478, 0.4665316641330719, -0.43853095173835754, -0.22229687869548798, 0.3354998826980591, -0.16507023572921753, -0.06518608331680298, 0.2390466332435608, 0.039275988936424255, -0.3886536955833435, -0.1660069078207016, 0.17745031416416168, 0.19752462208271027, 0.07779914885759354, -0.04839682951569557, -0.2970067858695984, 0.13495878875255585, 0.3166254758834839, 0.0654771476984024, -0.23182784020900726, -0.2993497848510742, -0.11140484362840652, -0.06439181417226791, 0.514197051525116, 0.13280721008777618, -0.2751377820968628, 0.10499174892902374, -0.2721591889858246, 0.2936933636665344, -0.05169518291950226, 0.1185731515288353, 0.4230370819568634, -0.02337534911930561, -0.7951251268386841, -0.16027691960334778, 0.43703293800354004, -0.05248725414276123, -0.2122921645641327, -0.47140875458717346, 0.24053315818309784, -0.23946473002433777, 0.22328287363052368, 0.055049411952495575, 0.18732194602489471, 0.03035423904657364, 0.02407148852944374, 0.09660723805427551, -0.23642680048942566, -0.15861909091472626, -0.17153595387935638, 0.20367786288261414, -0.22109945118427277, -0.007433035410940647, -0.010540366172790527, 0.3596077859401703, 0.31164130568504333, 0.08758017420768738, 0.03730412572622299, 0.26341819763183594, 0.050155505537986755, 0.0980498269200325, 0.3672504425048828, 0.1448051631450653, -0.0007640253752470016, 0.1889611780643463, -0.12266073375940323, 0.6879806518554688, 0.016462940722703934, 0.03225133568048477, -0.34003961086273193, 0.42591530084609985, -0.21703888475894928, 0.20069116353988647, 0.4093107581138611, -0.07173743844032288, -0.050097495317459106, 0.05778001621365547, -0.006799564231187105, -0.017152976244688034, 0.7467243671417236, 0.1952691674232483, 0.2502555549144745, 0.002956695854663849, -0.3220258355140686, 0.2596171498298645, 0.3991404175758362, -0.015342838130891323, -0.020852625370025635, -0.11995432525873184, 0.1937142014503479, -0.09338856488466263, -0.03672000765800476, 0.25073885917663574, 0.3810698390007019, 0.19513367116451263, 0.15570716559886932, -0.1311991810798645, -0.33823680877685547, -0.3438020944595337, 0.03513295203447342, -0.18210352957248688, 0.08744704723358154, 0.4963866174221039, 0.10160735994577408, 0.21628685295581818, -0.3020772635936737, -0.5422850847244263, -0.25364819169044495, -0.22231410443782806, 0.06132550165057182, 0.16164565086364746, -0.14391349256038666, -0.3678344488143921, -0.15366095304489136, -0.0919032171368599, -0.2918884754180908, -0.2732764482498169, 0.2300969362258911, 0.020874908193945885, -0.07249949127435684, -0.2189374715089798, 0.02323257550597191, 0.14874732494354248, -0.37039417028427124, 0.1783812940120697, -0.017374012619256973, -0.028042834252119064, -0.27953851222991943, 0.22281798720359802, -0.05679866671562195, -0.01930420473217964, -0.03147614002227783, -0.2791869640350342, -0.45083630084991455, 0.16968809068202972, -0.3762677013874054, -0.010960057377815247, 0.06464451551437378, -0.0032363757491111755, 0.05967199429869652, -0.2506176233291626, -0.07057526707649231, 0.14640071988105774, 0.09220800548791885, -0.1741306185722351, -0.35074377059936523, -0.08965122699737549, 0.03678468242287636, 0.014213903807103634, -0.19140614569187164, -0.27752187848091125, 0.07937449216842651, -0.14331994950771332, 0.6197946071624756, 0.3270139694213867, 0.005700867623090744, 0.3858785927295685, -0.2369554340839386, 0.00001177145168185234, 0.17174258828163147, 0.5837677121162415, -0.2605840563774109, -0.04442854970693588, 0.19860902428627014, -0.2997126579284668, -0.4421888589859009, -0.2730613946914673, 0.18830057978630066, 0.03232206031680107, -0.008429164066910744, -0.17861716449260712, 0.031004078686237335, -0.1309441179037094, -0.04899725317955017, -0.2130453884601593, -0.05581780523061752, 0.33782681822776794, -0.1836601346731186, -0.2421920895576477, -0.06447437405586243, 0.21673373878002167, 0.029940642416477203, 0.28654640913009644, 0.4599356949329376, -0.4840274751186371, 0.1646125614643097, -0.1400940716266632, 0.12755994498729706, 0.04408307000994682, 0.5117960572242737, 0.1533815860748291, 0.1332312375307083, 0.0008752048015594482, -0.14067688584327698, 0.09818042814731598, 0.4896940290927887, 0.056188441812992096, -0.14490284025669098, 0.3308490812778473, 0.20873358845710754, 0.24980415403842926, -0.11164984852075577, -0.04191560670733452, -0.07632186263799667, -0.00791233777999878, 0.10350729525089264, 0.07728653401136398, -0.05600050836801529, 0.017135661095380783, 0.07414527982473373, -0.0036139339208602905, -0.15379199385643005, 0.21246279776096344, 0.07286881655454636, 0.39130184054374695, -0.11124485731124878, -0.25944772362709045, -0.005773358047008514, -0.46897637844085693, -0.005584202706813812, 0.031853336840867996, -0.021542735397815704, -0.15841075778007507, -0.17286941409111023, 0.05907169356942177, -0.3485898971557617, 0.26234251260757446, 0.16069412231445312, 0.23712927103042603, 0.16907617449760437, -0.13415351510047913, -0.5272519588470459, -0.294148325920105, -0.3322208523750305, 0.18113453686237335, 0.596076488494873, 0.11514388769865036, -0.13822150230407715, -0.17282268404960632, 0.10999258607625961, -0.05185077711939812, -0.14824798703193665, 0.10066777467727661, -0.33127161860466003, 0.20241904258728027, 0.15132899582386017, -0.2806145250797272, 0.04583745449781418, 0.3753645718097687, -0.24381005764007568, 0.12480206787586212, 0.06615749001502991, -0.1499929428100586, 0.2123294323682785, 0.13495224714279175, 0.33930909633636475, -0.10402191430330276, -0.24397088587284088, -0.29319772124290466, 0.30135542154312134, -0.3514430522918701, 0.3701417148113251, 0.1212446391582489, 0.0003355368971824646, 0.30029138922691345, -0.09358739852905273, 0.10684163123369217, 0.3101273477077484, -0.18164154887199402, 0.1664067804813385, -0.6221522092819214, 0.11270289123058319, -0.44617894291877747, -0.2102094292640686, 0.36847415566444397, -0.07535791397094727, 0.2850716710090637, -0.1410284787416458, 0.2084536850452423, 0.23064105212688446, -0.47709769010543823, -0.07949845492839813, -0.16209423542022705, -0.19412687420845032, 0.2774372696876526, -0.02041754126548767, 0.7596277594566345, 0.4354732036590576, 0.5633991360664368, 0.23021692037582397, -0.1473512500524521, 0.38939782977104187, -0.5955854058265686, 0.19127970933914185, -0.34151583909988403, -0.37023502588272095, 0.09574323892593384, 0.10443316400051117, -0.06511325389146805, 0.21454881131649017, 0.20764991641044617, 0.2953759431838989, 0.39312154054641724, -0.09911176562309265, -0.10828026384115219, 0.17655311524868011, 0.12343072891235352, 0.18754130601882935, -0.2870902717113495, 0.1991366744041443, 0.34644365310668945, 0.004576649516820908, -0.05969361588358879, 0.0959683433175087, -0.7420675754547119, 0.02251587063074112, 0.04249122738838196, 0.10044645518064499, -0.09559989720582962, 0.35474252700805664, 0.11958041787147522, -0.2243277132511139, -0.04871850088238716, 0.3672390878200531, 0.15987490117549896, -0.015912584960460663, 0.00719662569463253, 0.21002444624900818, -0.3194081783294678, 0.32633107900619507, 0.242310032248497, -0.03880351036787033, 0.031407780945301056, -0.22021397948265076, -0.0787142813205719, -0.07076065987348557, 0.15572455525398254, -0.16848213970661163, -0.5484740734100342, 0.18215541541576385, -0.33470314741134644, 0.007207001093775034, -0.02991698868572712, 0.10068950057029724, -0.09960559010505676, -0.05343038588762283, 0.16725941002368927, -0.05686335265636444, -0.06265363842248917, -0.03488510102033615, 0.09307181090116501, -0.33285802602767944, -0.050463635474443436, 0.47442546486854553, -0.13673809170722961, -0.04600970447063446, 0.06487441807985306, -0.3305659294128418, 0.0352669283747673, -0.24317610263824463, 0.03892479091882706, -0.08075741678476334, -0.19715386629104614, -0.10430090129375458, -0.007580584846436977, -0.2291492074728012, 0.25272732973098755, 0.25843656063079834, 0.24956685304641724, 0.21303749084472656, 0.032852623611688614, -0.36308497190475464, -0.09832554310560226, 0.0810789093375206, 0.16799978911876678, -0.03473341092467308, -0.019618384540081024, -0.1595887839794159, 0.2588408589363098, 0.23063212633132935, -0.35463544726371765, -0.15017247200012207, -0.08997658640146255, 0.12681715190410614, -0.22102300822734833, -0.31873926520347595, 0.022486412897706032, 0.1468534916639328, 0.14308585226535797, -0.008150145411491394, -0.05782918632030487, -0.24771744012832642, -0.20516616106033325, 0.083225779235363, -0.04834723472595215, -0.09744089841842651, -0.13200169801712036, 0.21839922666549683, 0.1043625921010971, -0.14063486456871033, -0.2313222736120224, -0.13392606377601624, 0.3573988676071167, 0.12174542248249054, -0.08553825318813324, 0.1825152337551117, -0.1426304578781128, 0.1086663231253624, 0.01882341504096985, -0.02538543939590454, 0.07357704639434814, -0.026512712240219116, 0.10945796221494675, -0.20184220373630524, -0.2564111351966858, -0.07424164563417435, -0.28038954734802246, 0.14717337489128113, 0.10414642840623856, -0.05297304689884186, 0.008782848715782166, -0.18328946828842163, 0.2844215929508209, 0.36497512459754944, 0.11780503392219543, -0.09963665157556534, -0.013917770236730576, 0.22528287768363953, -0.22220194339752197, 0.09425365179777145, 0.4496179521083832, 0.043986815959215164, 0.06563544273376465, -0.028284482657909393, 0.15857842564582825, -0.20769783854484558, -0.018284879624843597, -0.05924851819872856, -0.274423748254776, -0.01816922426223755, 0.17280733585357666, -0.19119597971439362, 0.2515231966972351, 0.07363121956586838, 0.15538182854652405, -0.37914836406707764, 0.1871822029352188, 0.35076963901519775, -0.0025786757469177246, 0.1941520869731903, -0.2856433689594269, 0.14951317012310028, -0.02367575466632843, 0.09032042324542999, -0.2703600227832794, 0.34511232376098633, 0.0853983536362648, -0.02180592715740204, 0.08363571763038635, 0.16912135481834412, -0.10833384096622467, -0.3376132547855377, -0.004882588982582092, -0.08809277415275574, -0.07047510892152786, -0.544400155544281, -0.4664246141910553, -0.22646990418434143, -0.394583523273468, 0.0710502415895462, -0.16874070465564728, 0.3817867636680603, 0.02951917052268982, 0.24772146344184875, -0.09672556817531586, -0.18267740309238434, -0.46176448464393616, 0.01024707779288292, 0.4208807945251465, -0.04675701633095741, 0.17743246257305145, 0.21671605110168457, 0.41431334614753723, 0.35717928409576416, 0.1616116166114807, 0.2387458086013794, 0.1888272762298584, -0.23361444473266602, -0.2279309630393982, 0.03246691823005676, 0.08932045102119446, -0.19615799188613892, 0.17421609163284302, 0.25997745990753174, 0.043390270322561264, 0.3002178966999054, 0.1322067826986313, -0.2410818189382553, 0.07742059230804443, 0.26498717069625854, -0.003574555739760399, -0.21413002908229828, 0.29473719000816345, -0.03916162997484207, 0.018681995570659637, -0.3314741849899292, 0.3797486424446106, -0.5077441334724426, -0.08658870309591293, -0.5439391136169434, 0.15140368044376373, 0.17864581942558289, -0.12435458600521088, 0.10393225401639938, 0.26889368891716003, 0.41731470823287964, -0.12723563611507416, 0.33877450227737427, -0.28413742780685425, -0.4613204002380371, -0.2653636336326599, -0.04584982246160507, 0.03950878977775574, 0.27052927017211914, -0.4309377372264862, 0.28873559832572937, -0.29741546511650085, 0.051449984312057495, 0.27456027269363403, 0.4283304512500763, 0.24691930413246155, 0.06649784743785858, -0.33822938799858093, 0.1559208780527115, 0.19840247929096222, -0.08824513852596283, -0.11956954002380371, 0.1306329071521759, -0.17390617728233337, 0.019816752523183823, 0.10971526056528091, 0.0855555534362793, -0.3465362787246704, -0.28865334391593933, 0.4207552969455719, 0.3385229706764221, 0.275066614151001, 0.2680865228176117, -0.37258201837539673, -0.3853943943977356, 0.031004125252366066, -0.12375257909297943, -0.06270965933799744, 0.06228291243314743, -0.06584185361862183, 0.38764387369155884, 0.02506408654153347, 0.07477395981550217, 0.07386114448308945, 0.5041284561157227, 0.055955611169338226, 0.0017196849221363664, -0.22702951729297638, -0.026206813752651215, 0.09364626556634903, 0.0634227991104126, -0.6129947304725647, 0.407443642616272, 0.0005168020725250244, 0.18714407086372375, -0.23370829224586487, -0.25390002131462097, 0.3750455975532532, -0.35842156410217285, 0.2720431089401245, -0.0997019112110138, 0.34275418519973755, 0.22176191210746765, 0.29680517315864563, 0.04679156094789505, -0.10589517652988434, 0.07175469398498535, 0.29574763774871826, 0.1622447669506073, 0.23034629225730896, -0.2334200143814087, -0.08380287140607834, -0.0016767457127571106, 0.29279518127441406, 0.06492644548416138, -0.29197126626968384, 0.41343367099761963, -0.18410754203796387 ]
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?
[ 0.0004875287413597107, -0.19439971446990967, 0.0977800041437149, 0.14103549718856812, 0.28600651025772095, -0.21947556734085083, 0.29816094040870667, 0.17576207220554352, 0.16847142577171326, 0.15842986106872559, -0.22391526401042938, 0.20470955967903137, -0.47158321738243103, 0.24365346133708954, -0.03945283219218254, -0.09374237060546875, -0.18991327285766602, -0.03363172709941864, 0.1052035465836525, -0.015584971755743027, -0.09069801867008209, 0.2365892231464386, 0.04565010964870453, -0.02871144562959671, -0.31746068596839905, 0.35048094391822815, 0.05521952360868454, 0.055927153676748276, -0.14533692598342896, -0.032764554023742676, 0.3463771343231201, -0.1924692839384079, 0.2557666003704071, 0.7926076650619507, -0.00012355622311588377, -0.2130509316921234, 0.06719064712524414, -0.011289679445326328, -0.27556076645851135, -0.30580538511276245, 0.4654071033000946, -0.2523364722728729, 0.0014711632393300533, -0.2047473043203354, -0.027092352509498596, -0.06222647801041603, -0.2024286985397339, -0.08055980503559113, 0.008652262389659882, 0.3162081837654114, 0.040686558932065964, 0.2029191106557846, 0.0005703046917915344, 0.07062731683254242, 0.09595900774002075, -0.10697981715202332, -0.11937206983566284, 0.5899260640144348, 0.19050449132919312, -0.039017993956804276, -0.19934436678886414, 0.16511176526546478, 0.1557428538799286, 0.22187592089176178, 0.781131386756897, 0.03963211923837662, -0.10514868050813675, 0.007861129939556122, 0.051743365824222565, 0.35052239894866943, -0.0570165291428566, -0.23309963941574097, -0.24215416610240936, 0.18319877982139587, 0.068288154900074, -0.2815703749656677, -0.25224000215530396, 0.052199751138687134, 0.04698941484093666, -0.27747398614883423, -0.3826291263103485, -0.3548019528388977, -0.25620725750923157, 0.13442924618721008, -0.26187700033187866, 0.39998865127563477, -0.0648723617196083, -0.05895049870014191, 0.18283896148204803, 0.05480562150478363, 0.20971029996871948, -0.012297403067350388, 0.357552707195282, 0.3185586929321289, 0.045000024139881134, -0.22733232378959656, 0.15663954615592957, -0.14627161622047424, 0.0655384361743927, -0.24303464591503143, -0.2348577231168747, 0.12916642427444458, -0.31121769547462463, 0.009082503616809845, -0.005182210355997086, 0.4701380431652069, 0.038915686309337616, 0.37499338388442993, 0.350391149520874, 0.2842978835105896, -0.029754575341939926, -0.07869093120098114, -0.06778586655855179, -0.25944027304649353, 0.28838038444519043, 0.16139480471611023, -0.21710926294326782, -0.3110698163509369, -0.1479407250881195, -0.058148615062236786, -0.1836259663105011, -0.21360017359256744, 0.38544076681137085, 0.24474015831947327, -0.3576725423336029, -0.11658535152673721, 0.5851033926010132, 0.05813033878803253, -0.2100522369146347, 0.0859038382768631, -0.09296082705259323, -0.09709036350250244, -0.29787325859069824, 0.08617281913757324, 0.20618440210819244, -0.1500415951013565, 0.20931978523731232, -0.10786916315555573, 0.6836860179901123, 0.13581722974777222, 0.18622775375843048, 0.333433598279953, 0.08329427987337112, -0.1664721518754959, -0.24997058510780334, -0.40275388956069946, 0.25848880410194397, 0.09364005923271179, -0.17199425399303436, 0.06458362936973572, -0.11431395262479782, -0.3340190649032593, -0.2537508010864258, 0.0032773101702332497, -0.29372066259384155, -0.13937141001224518, 0.04673631861805916, 0.3141668736934662, 0.0852745771408081, -0.29998788237571716, 0.021874941885471344, -0.2990492582321167, -0.25199851393699646, 0.09707091003656387, 0.2996886968612671, -0.3436623215675354, -0.3732588589191437, -0.4258340895175934, 0.11342962086200714, 0.3820705711841583, 0.22695305943489075, 0.14589886367321014, -0.15026432275772095, -0.11249729245901108, 0.180483877658844, -0.3285110592842102, 0.3918129503726959, 0.03860662505030632, -0.1812959760427475, -0.23008078336715698, 0.10167472064495087, -0.2862394154071808, -0.21849629282951355, 0.06395968049764633, -0.1890784204006195, -0.1219702959060669, 0.32159721851348877, 0.2692631781101227, -0.15223830938339233, -0.11927391588687897, -0.23923271894454956, -0.2757665812969208, 0.23823878169059753, 0.1066085696220398, 0.3859720230102539, -0.018501952290534973, 0.15808379650115967, 0.5193025469779968, 0.018723219633102417, -0.4093892276287079, -0.04577060043811798, -0.13258491456508636, 0.3818323016166687, -0.2926695942878723, 0.39089125394821167, -0.320676326751709, -0.12131329625844955, 0.23700359463691711, -0.22047315537929535, 0.3617454767227173, 0.19992277026176453, 0.11520010232925415, -0.10204403847455978, -0.04495716840028763, 0.34095436334609985, 0.03917757049202919, -0.11112780123949051, -0.14549946784973145, -0.23571977019309998, -0.19972245395183563, -0.20017972588539124, -0.33275479078292847, -0.2919325530529022, -0.0543326810002327, 0.4235488176345825, 0.2041720747947693, -0.03850071132183075, -0.08839352428913116, -0.08542560786008835, 0.42085832357406616, 0.13108524680137634, 0.21697238087654114, -0.12121257930994034, -0.20755881071090698, 0.13657215237617493, -0.28143829107284546, 0.24335289001464844, 0.2878012955188751, 0.34533360600471497, -0.08362184464931488, 0.1704852432012558, 0.2630351781845093, -0.1229206994175911, -0.0553075447678566, -0.05954970419406891, 0.23119059205055237, 0.29891395568847656, 0.2667691111564636, 0.2532308101654053, 0.07388072460889816, -0.28169938921928406, 0.05441199615597725, 0.012851081788539886, 0.012367136776447296, 0.4185680150985718, -0.2691675126552582, -0.14407937228679657, -0.19352592527866364, 0.20795512199401855, 0.065392404794693, 0.3745553493499756, -0.026033086702227592, 0.042966097593307495, 0.3506179749965668, -0.29047855734825134, -0.21643570065498352, -0.12244248390197754, -0.023048128932714462, 0.1722012311220169, 0.10130808502435684, 0.4216386377811432, 0.12463176250457764, -0.10538530349731445, 0.07629331946372986, -0.05406550318002701, 0.11428824067115784, -0.1998913437128067, -0.10426612943410873, 0.04114816337823868, 0.08948369324207306, 0.16858136653900146, -0.019060129299759865, 0.06301812827587128, 0.1298634260892868, -0.2475425750017166, 0.368917316198349, 0.033474162220954895, 0.3673948645591736, -0.1088714599609375, -0.3008711338043213, -0.25574710965156555, -0.18844883143901825, 0.008497297763824463, 0.253911554813385, 0.09166291356086731, -0.1420186161994934, 0.0540216900408268, 0.46498996019363403, -0.0766250267624855, -0.14190179109573364, -0.029270004481077194, -0.36523035168647766, -0.34538906812667847, 0.00930764526128769, 0.04194915294647217, 0.042097557336091995, 0.0964585542678833, 0.08965092897415161, -0.03133619576692581, -0.23312805593013763, -0.7229450941085815, 0.004387848079204559, -0.006775561720132828, 0.2805350720882416, 0.21111741662025452, -0.08313450217247009, -0.33911240100860596, 0.004088832065463066, 0.05235666781663895, -0.5004477500915527, -0.005801443941891193, 0.33237186074256897, -0.3564043939113617, 0.08894755691289902, -0.0786409080028534, -0.10567044466733932, -0.09344393759965897, -0.07754145562648773, -0.0709226131439209, 0.1984402984380722, 0.05370702967047691, -0.058387286961078644, -0.020465310662984848, 0.4024657607078552, 0.5086860656738281, -0.048032745718955994, -0.1685408502817154, 0.24041202664375305, 0.3423587679862976, -0.3283785879611969, -0.3037557899951935, -0.0282772034406662, 0.06588000804185867, 0.02901761792600155, 0.005568805616348982, -0.3014434576034546, -0.4165246784687042, -0.10350775718688965, -0.07528005540370941, 0.0514252707362175, 0.20570337772369385, 0.024882882833480835, 0.024753913283348083, 0.07889006286859512, -0.29824864864349365, -0.20026525855064392, -0.019337251782417297, 0.14835575222969055, 0.7090576887130737, -0.15822094678878784, -0.1775602400302887, 0.14850440621376038, 0.8652079701423645, 0.01294020563364029, -0.45980194211006165, 0.10806562006473541, 0.17440694570541382, 0.09269344806671143, 0.015705883502960205, -0.21939322352409363, 0.2692299783229828, -0.3033927083015442, -0.17277409136295319, 0.3044235408306122, 0.10612741112709045, -0.12558645009994507, -0.27443021535873413, -0.12407653033733368, 0.5246573686599731, -0.13078711926937103, 0.0432763397693634, 0.08086720108985901, 0.027793578803539276, -0.05913827195763588, -0.13328993320465088, -0.335997611284256, -0.11665208637714386, 0.2886207401752472, 0.08492168039083481, -0.016899973154067993, 0.04680157080292702, 0.4622916579246521, -0.1196925938129425, -0.5215777158737183, 0.4779226779937744, 0.14861798286437988, -0.6269400119781494, 0.03034314140677452, -0.09077312797307968, -0.01497764140367508, -0.004339289385825396, -0.30573010444641113, 0.08299004286527634, 0.02205529808998108, -0.02300373651087284, 0.047544971108436584, -0.2639550566673279, 0.11107394099235535, -0.13539643585681915, -0.3203109800815582, 0.4749372601509094, 0.2835904061794281, -0.3579861521720886, 0.060252055525779724, 0.2661556303501129, 0.3918269872665405, -0.16820651292800903, -0.044259700924158096, -0.38714534044265747, -0.24752110242843628, 0.05117461085319519, -0.16052791476249695, 0.05983017385005951, 0.26893991231918335, 0.06399001181125641, 0.15757091343402863, 0.14858344197273254, -0.05238517373800278, 0.15332116186618805, 0.06338642537593842, 0.06765417754650116, -0.03543105721473694, 0.036598119884729385, 0.15093505382537842, -0.05382128059864044, 0.16688156127929688, 0.311791330575943, -0.4149140417575836, -0.4968370199203491, 0.13977214694023132, 0.2661862373352051, 0.21740731596946716, 0.03490883857011795, 0.07774905860424042, -0.10712722688913345, 0.3252790570259094, 0.3136734962463379, 0.17274144291877747, 0.3225984573364258, 0.15361593663692474, 0.31555381417274475, 0.11419904977083206, -0.1838606595993042, 0.3391721546649933, -0.11920882761478424, 0.12232592701911926, 0.34898805618286133, 0.029947329312562943, -0.027535466477274895, 0.06459630280733109, 0.5371322631835938, 0.9507374167442322, 0.07198921591043472, 0.10562898218631744, 0.3453916609287262, -0.21591255068778992, 0.33187928795814514, -0.13465917110443115, -0.054771460592746735, -0.20902936160564423, -0.27686482667922974, -0.028736935928463936, -0.00940077006816864, -0.41684117913246155, -0.30249524116516113, -0.24513211846351624, 0.17041343450546265, -0.37227344512939453, -0.2893790900707245, 0.27391788363456726, 0.2895249128341675, -0.012750349938869476, 0.026801928877830505, 0.2843709886074066, 0.07212065160274506, -0.24628673493862152, 0.3137531578540802, -0.06891941279172897, -0.13711774349212646, -0.12174910306930542, 0.11128251254558563, -0.34428173303604126, 0.18231087923049927, -0.14361527562141418, -0.043186455965042114, -0.1379406601190567, -0.2727839946746826, 0.04846426472067833, 0.04304114356637001, 0.7600718140602112, 0.019665725529193878, -0.3796805143356323, 0.2225048840045929, -0.06933467835187912, 0.18749548494815826, 0.12327985465526581, -0.11628518253564835, 0.3619278073310852, -0.21870329976081848, -0.08358802646398544, -0.1212911531329155, 0.23607277870178223, -0.4464300274848938, 0.16874206066131592, -0.1505315899848938, -0.16223563253879547, 0.01151440292596817, 0.2456081658601761, 0.23653148114681244, 0.11821115016937256, -0.1011229157447815, 0.05867795646190643, 0.14740237593650818, -0.23920926451683044, 0.032710086554288864, 0.04361782222986221, -0.3119707703590393, 0.029099272564053535, -0.04143796116113663, -0.07832875847816467, -0.009219510480761528, 0.22514592111110687, 0.18119826912879944, -0.08502624928951263, 0.2331380546092987, -0.5456886887550354, 0.48003166913986206, -0.39778536558151245, 0.16316473484039307, 0.08292648196220398, -0.033607445657253265, 0.08675375580787659, 0.7250291705131531, 0.4419584274291992, 0.13836205005645752, -0.4551559388637543, 0.10337215662002563, -0.38534581661224365, 0.05547995865345001, -0.18755152821540833, 0.3016723394393921, -0.21308249235153198, 0.16312822699546814, -0.05166450887918472, 0.19557636976242065, -0.19422325491905212, -0.019207777455449104, 0.1287049949169159, -0.1361647993326187, 0.05789325013756752, -0.36982864141464233, 0.18015706539154053, 0.07099314033985138, -0.2820511758327484, -0.12980441749095917, 0.07917620986700058, -0.0011406578123569489, -0.49501511454582214, 0.1621682494878769, 0.23005986213684082, -0.015350095927715302, -0.32313966751098633, -0.4455387592315674, -0.057451631873846054, -0.10607446730136871, 0.021053496748209, 0.14171037077903748, 0.04478997737169266, 0.3292349874973297, -0.05912426486611366, -0.04719097912311554, -0.10356050729751587, -0.05455074459314346, 0.46705278754234314, 0.22028081119060516, 0.11532847583293915, 0.012021519243717194, 0.34375685453414917, 0.0014599710702896118, 0.13209125399589539, 0.02976076677441597, 0.006227478384971619, 0.06046120077371597, 0.2978818416595459, 0.14499559998512268, 0.1928529441356659, -0.3699391782283783, 0.229014053940773, 0.14820793271064758, 0.12221960723400116, -0.1619810312986374, 0.17190229892730713, 0.03464827314019203, -0.2363877147436142, -0.20888032019138336, -0.31774207949638367, 0.08207541704177856, -0.17358574271202087, 0.13484278321266174, -0.2804161012172699, 0.4191932678222656, 0.3138980269432068, -0.09825557470321655, -0.24227765202522278, -0.22451098263263702, 0.068136066198349, 0.1616407185792923, -0.18701015412807465, 0.31334683299064636, 0.40764638781547546, 0.11467598378658295, 0.20449267327785492, -0.4011955261230469, -0.14249026775360107, 0.06567273288965225, -0.42258644104003906, -0.08957509696483612, 0.3227171003818512, 0.11632966995239258, -0.053208302706480026, -0.1357431709766388, 0.17436088621616364, -0.07144565135240555, -0.19191372394561768, 0.395449161529541, -0.3155260682106018, -0.21161703765392303, -0.08746601641178131, 0.35621920228004456, 0.1253102868795395, 0.09618903696537018, 0.10229560732841492, -0.08662346750497818, 0.20117151737213135, -0.05531684309244156, 0.033217210322618484, 0.15262728929519653, -0.15838593244552612, 0.12446343898773193, -0.04102522134780884, 0.26512834429740906, -0.74640291929245, 0.2926129698753357, 0.2878526747226715, 0.14978192746639252, -0.21726614236831665, 0.11766484379768372, -0.013434730470180511, -0.053702328354120255, 0.31686997413635254, 0.03299371153116226, 0.19358165562152863, -0.37752828001976013, -0.036859020590782166, 0.21869458258152008, 0.00196247361600399, 0.16003358364105225, 0.16048268973827362, 0.4990955889225006, -0.18967550992965698, 0.05753704532980919, -0.10264881700277328, -0.14954744279384613, 0.4168829023838043, -0.39484140276908875, 0.16702069342136383, -0.061144061386585236, 0.872622013092041, 0.28352856636047363, -0.15029379725456238, -0.2694929242134094, -0.08044754713773727, -0.2182770073413849, 0.0842762440443039, 0.08718106150627136, -0.12014051526784897, 0.07406158000230789, -0.05157264694571495, 0.006864530965685844, 0.1870855689048767, 0.18445222079753876, 0.24675045907497406, 0.3651888966560364, -0.27606475353240967, 0.31659793853759766, -0.3954797089099884, 0.15575823187828064, 0.17428436875343323, 0.06074602156877518, -0.04848933592438698, -0.14108151197433472, 0.023486055433750153, -0.11504451930522919, 0.2646304965019226, -0.22524169087409973, 0.051122941076755524, -0.18277771770954132, -0.23931050300598145, -0.34198781847953796, -0.05003689229488373, 0.038902245461940765, -0.031736794859170914, -0.1246156319975853, -0.06375440955162048, -0.09632004052400589, -0.12807002663612366, -0.050185658037662506, -0.29143333435058594, 0.39763709902763367, 0.3708544075489044, 0.15045663714408875, -0.0725288838148117, 0.7684551477432251, 0.20320600271224976, 0.006717095151543617, 0.30119818449020386, 0.20140664279460907, -0.11478064954280853, -0.20222163200378418, -0.037507813423871994, 0.19382283091545105, -0.25180089473724365, -0.26196107268333435, -0.4261299967765808, 0.6186966896057129, 0.10939067602157593, -0.47760042548179626, -0.0888892114162445, -0.22655119001865387, -0.22605806589126587, -0.13082444667816162, 0.2619841396808624, 0.4220671057701111, 0.07908143103122711, 0.032229527831077576, -0.39357441663742065, 0.07538671791553497, 0.5604331493377686, -0.42825835943222046, -0.19135086238384247, -0.01269717700779438, 0.18095439672470093, 0.239191472530365, -0.30848655104637146, -0.767350435256958, -0.12440449744462967, 0.1188991442322731, 0.1689777672290802, -0.16665229201316833, 0.23522740602493286, 0.13245025277137756, 0.2679494023323059, -0.22290945053100586, 0.5535248517990112, -0.07835764437913895, -0.13849210739135742, -0.41262978315353394, -0.09996578842401505 ]
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
[ 0.03960353508591652, -0.294548898935318, 0.0695950984954834, 0.16438624262809753, 0.27279844880104065, -0.281116247177124, 0.3806678056716919, 0.14022669196128845, 0.14548735320568085, 0.1657414734363556, -0.21056005358695984, 0.2934771180152893, -0.4552498459815979, 0.22415979206562042, 0.06959349662065506, -0.10988396406173706, -0.05097750574350357, 0.03715774416923523, 0.10635839402675629, -0.13577419519424438, -0.10876131802797318, 0.16499070823192596, 0.05934944748878479, -0.1525467485189438, -0.2916155159473419, 0.3176150619983673, 0.019248859956860542, 0.025706477463245392, -0.20944362878799438, -0.05774233117699623, 0.2948092520236969, -0.06041407212615013, 0.2058347761631012, 0.7945738434791565, -0.00012249436986166984, -0.267738401889801, 0.13256409764289856, -0.025316013023257256, -0.34068137407302856, -0.2211291342973709, 0.38395199179649353, -0.3060538172721863, 0.08232942968606949, -0.1754467487335205, 0.04278825223445892, -0.03406541049480438, -0.24324075877666473, -0.042656127363443375, 0.18749338388442993, 0.21665847301483154, 0.042248353362083435, 0.23043592274188995, 0.0012255571782588959, 0.1125812828540802, 0.13914158940315247, -0.10558630526065826, -0.10437044501304626, 0.5303356647491455, 0.27584004402160645, -0.06240950524806976, -0.19226285815238953, 0.23816801607608795, 0.07783679664134979, 0.10142309963703156, 0.6716704368591309, 0.03906027227640152, -0.05559539049863815, -0.06814546138048172, 0.0057654231786727905, 0.3251340687274933, -0.1148088201880455, -0.21514946222305298, -0.3067004680633545, 0.08728203922510147, 0.07104206085205078, -0.2673068046569824, -0.21797055006027222, 0.1069883331656456, 0.10014531016349792, -0.2690121829509735, -0.41902437806129456, -0.36483168601989746, -0.19585199654102325, 0.07973796874284744, -0.2726512551307678, 0.3813120126724243, 0.009816615842282772, -0.15502551198005676, 0.13104580342769623, 0.1450662612915039, 0.3880736231803894, -0.07437723129987717, 0.347133994102478, 0.27463987469673157, -0.016507018357515335, -0.30216899514198303, 0.2685924470424652, -0.136091411113739, -0.0050390176475048065, -0.22477178275585175, -0.16654254496097565, 0.20244866609573364, -0.26308172941207886, 0.0775960311293602, -0.04299274459481239, 0.4014994204044342, -0.0002394840121269226, 0.37281325459480286, 0.32136309146881104, 0.40310484170913696, 0.10458315908908844, -0.019626177847385406, -0.07396245747804642, -0.0863814577460289, 0.258507639169693, 0.19309937953948975, -0.22396455705165863, -0.2521964907646179, -0.06776197999715805, -0.020372584462165833, -0.22098663449287415, -0.24022755026817322, 0.33338066935539246, 0.19345812499523163, -0.4989130198955536, -0.023181702941656113, 0.5008065700531006, 0.0060212500393390656, -0.16556993126869202, 0.10502782464027405, -0.1418714076280594, -0.12604521214962006, -0.30947235226631165, 0.07590886950492859, 0.2528243064880371, -0.16018125414848328, 0.2812947630882263, -0.07687214016914368, 0.49812597036361694, 0.04750313609838486, 0.30355268716812134, 0.3707237243652344, 0.096248097717762, -0.1441243588924408, -0.20200631022453308, -0.3485706150531769, 0.1855834424495697, 0.0507267490029335, -0.1315109133720398, -0.030971579253673553, -0.11490444093942642, -0.27944108843803406, -0.07502098381519318, 0.050973307341337204, -0.25767451524734497, -0.05764269828796387, 0.02539883553981781, 0.3534553349018097, 0.09787649661302567, -0.2471977025270462, 0.0006358996033668518, -0.24852745234966278, -0.1676691323518753, 0.06729397177696228, 0.39564600586891174, -0.1806042641401291, -0.3170182406902313, -0.5069208145141602, 0.16817256808280945, 0.43953606486320496, 0.18799075484275818, 0.16617606580257416, -0.24754562973976135, -0.032942190766334534, 0.2556147575378418, -0.34265583753585815, 0.27378490567207336, 0.11720743775367737, -0.305671751499176, -0.23019136488437653, 0.08945822715759277, -0.3044486939907074, -0.21890729665756226, 0.08569763600826263, -0.14614686369895935, -0.06445053219795227, 0.2756136953830719, 0.26867708563804626, -0.17650912702083588, -0.012937309220433235, -0.2883862555027008, -0.25624534487724304, 0.2462271749973297, 0.17889517545700073, 0.37698835134506226, -0.037900738418102264, 0.15235093235969543, 0.538781464099884, 0.06697943806648254, -0.38386017084121704, -0.03002522699534893, -0.156999871134758, 0.5144122242927551, -0.46003031730651855, 0.38163986802101135, -0.21484914422035217, -0.11701158434152603, 0.17491978406906128, -0.38291752338409424, 0.3466678559780121, 0.15802478790283203, 0.05973019450902939, -0.08232412487268448, 0.00481763482093811, 0.3204120099544525, -0.044858139008283615, -0.07983560860157013, -0.1922534704208374, -0.2519010901451111, -0.15173843502998352, -0.23149123787879944, -0.2959873676300049, -0.18502183258533478, -0.040659449994564056, 0.22060538828372955, 0.36955732107162476, -0.021372562274336815, -0.14678031206130981, -0.08496762812137604, 0.3680126368999481, 0.1280679851770401, 0.16879601776599884, -0.036535318940877914, -0.20773816108703613, 0.1696757972240448, -0.18458296358585358, 0.2581161558628082, 0.27523720264434814, 0.40712791681289673, 0.02619856223464012, 0.2872464954853058, 0.2546694576740265, -0.10023227334022522, -0.10496433079242706, -0.2070123255252838, 0.2988642752170563, 0.30754148960113525, 0.23850303888320923, 0.29118162393569946, 0.09369339048862457, -0.27139928936958313, 0.09760597348213196, -0.060411788523197174, 0.17113259434700012, 0.39449170231819153, -0.13852442800998688, -0.08198687434196472, -0.1640886515378952, 0.14725808799266815, 0.1415473222732544, 0.36072492599487305, -0.052627794444561005, 0.017791762948036194, 0.23726016283035278, -0.4145682156085968, -0.28494346141815186, -0.08580277115106583, 0.05542195960879326, 0.14266939461231232, 0.06415874511003494, 0.4149327576160431, 0.05140446126461029, -0.0188014954328537, 0.0447479709982872, -0.12116052955389023, 0.15478205680847168, -0.2346334159374237, -0.04626794904470444, 0.08961528539657593, 0.09873834997415543, 0.0901554748415947, 0.025140583515167236, 0.07456117868423462, 0.1143125370144844, -0.14357632398605347, 0.3518413007259369, 0.13265903294086456, 0.42570608854293823, -0.23266732692718506, -0.3643445670604706, -0.2572852075099945, -0.2122786045074463, 0.03864738345146179, 0.3008294105529785, 0.07326662540435791, -0.08552314341068268, 0.2842051386833191, 0.5591573119163513, -0.0492175817489624, -0.10149649530649185, -0.063578300178051, -0.29799944162368774, -0.2619438171386719, 0.013049319386482239, -0.047071781009435654, 0.021043960005044937, -0.01213170401751995, 0.12732505798339844, -0.07290872931480408, -0.1648215502500534, -0.791871964931488, -0.007138855755329132, 0.06195426732301712, 0.2949560582637787, 0.12773045897483826, -0.13784661889076233, -0.3270549774169922, -0.001270536333322525, 0.01725994236767292, -0.40893107652664185, -0.023082325235009193, 0.333400696516037, -0.24032647907733917, 0.020099468529224396, -0.03319774568080902, -0.19501295685768127, -0.12176128476858139, -0.0371471643447876, 0.035057950764894485, 0.17747166752815247, 0.037284571677446365, -0.12759655714035034, 0.11328987032175064, 0.32145223021507263, 0.42720887064933777, -0.1533857136964798, -0.03843604773283005, 0.1742308884859085, 0.4179564118385315, -0.2875511050224304, -0.3364487290382385, -0.11314113438129425, -0.04294079542160034, 0.09721885621547699, 0.02003619447350502, -0.33385950326919556, -0.3707353472709656, -0.13501015305519104, -0.15471312403678894, 0.16762450337409973, 0.21132662892341614, 0.08064785599708557, 0.13571622967720032, 0.08123931288719177, -0.1643407940864563, -0.11949416995048523, -0.051210664212703705, 0.16528776288032532, 0.6772114634513855, -0.2268790602684021, -0.1361590325832367, 0.09984970092773438, 0.8651848435401917, 0.14755195379257202, -0.45604637265205383, 0.036251723766326904, 0.3171021640300751, 0.025822825729846954, -0.03095582127571106, -0.291069895029068, 0.18813374638557434, -0.24658456444740295, -0.2843208312988281, 0.2827780842781067, 0.10243097692728043, -0.29742860794067383, -0.3031907081604004, -0.20264632999897003, 0.506892204284668, -0.12234582006931305, 0.07363028824329376, 0.11307651549577713, 0.0693340077996254, -0.004519810900092125, -0.06540961563587189, -0.2938372790813446, -0.12536656856536865, 0.24079278111457825, -0.007610015571117401, 0.09480364620685577, -0.037457797676324844, 0.44763636589050293, 0.003257923061028123, -0.5484259128570557, 0.3487659692764282, 0.05293440446257591, -0.5662438869476318, 0.14763320982456207, -0.07294204831123352, -0.039310112595558167, -0.04653788357973099, -0.3226349353790283, 0.07470069080591202, 0.036167632788419724, -0.14837059378623962, -0.06120404973626137, -0.25290030241012573, 0.09232476353645325, -0.22582730650901794, -0.47378313541412354, 0.5314682126045227, 0.33735042810440063, -0.38272061944007874, 0.09321527183055878, 0.22271139919757843, 0.31544360518455505, -0.20287923514842987, -0.01469384878873825, -0.2976093292236328, -0.2209089994430542, 0.07640571147203445, -0.07465633004903793, -0.026433972641825676, 0.22570650279521942, -0.051201701164245605, -0.027302740141749382, 0.1271093636751175, -0.10163205116987228, 0.18122626841068268, 0.07489539682865143, 0.008387213572859764, -0.02671448513865471, 0.15502208471298218, 0.11128585040569305, -0.07816717028617859, 0.19248150289058685, 0.3351227343082428, -0.41935256123542786, -0.5140181183815002, 0.11239588260650635, 0.15022076666355133, 0.1733611822128296, 0.09620565921068192, 0.0997607633471489, -0.010332837700843811, 0.40235763788223267, 0.28301751613616943, 0.16237877309322357, 0.26863452792167664, 0.13129714131355286, 0.4590049982070923, 0.06577195227146149, -0.14670036733150482, 0.3369627296924591, -0.1843062937259674, 0.029224611818790436, 0.3064761459827423, -0.09339839220046997, -0.026042615994811058, 0.17510423064231873, 0.45825260877609253, 0.9547624588012695, 0.11388440430164337, 0.07662218809127808, 0.2924111783504486, -0.12001800537109375, 0.3372717797756195, -0.2470250129699707, 0.05289619788527489, -0.31194964051246643, -0.37743079662323, -0.03756645694375038, -0.08929362893104553, -0.4033791720867157, -0.27996718883514404, -0.20358093082904816, 0.11634592711925507, -0.34473085403442383, -0.22656014561653137, 0.3043632507324219, 0.32432472705841064, 0.07537992298603058, 0.021569836884737015, 0.287003755569458, 0.06190452724695206, -0.20935119688510895, 0.31468191742897034, -0.0779997855424881, -0.23030854761600494, -0.23614338040351868, 0.10131241381168365, -0.187404602766037, 0.26554423570632935, -0.057189591228961945, -0.006722116842865944, 0.010055450722575188, -0.3199191093444824, 0.07066012918949127, 0.06933721899986267, 0.7521310448646545, 0.11961214244365692, -0.37206533551216125, 0.148244708776474, -0.17030683159828186, 0.13719098269939423, 0.1343541443347931, -0.16840770840644836, 0.36602407693862915, -0.14322543144226074, -0.16689930856227875, -0.0825682207942009, 0.2166527807712555, -0.2843283414840698, 0.16792216897010803, -0.173920139670372, -0.1123281717300415, -0.07144229859113693, 0.4080495536327362, 0.32134342193603516, 0.07981818914413452, -0.20158815383911133, 0.09625990688800812, 0.2228284776210785, -0.2782500684261322, 0.08793088793754578, 0.13651295006275177, -0.23687361180782318, -0.00305919349193573, 0.021061338484287262, -0.04544374346733093, -0.07609303295612335, 0.20994733273983002, 0.18963415920734406, -0.12245554476976395, 0.20275983214378357, -0.5506261587142944, 0.5572553277015686, -0.4292539358139038, 0.11070036888122559, -0.13716785609722137, -0.022318467497825623, 0.03415050730109215, 0.614263117313385, 0.46294867992401123, 0.17757567763328552, -0.4233265221118927, 0.1225862205028534, -0.42704424262046814, 0.10487539321184158, -0.18003784120082855, 0.3510757386684418, -0.2739526927471161, -0.00676131434738636, 0.006709648296236992, 0.38980719447135925, -0.2270161658525467, -0.026360461488366127, 0.12607695162296295, -0.07171663641929626, -0.02165127918124199, -0.4830068051815033, 0.11179284006357193, 0.04559263586997986, -0.2318335324525833, -0.10764293372631073, 0.08288025110960007, -0.042953141033649445, -0.46715980768203735, 0.15198089182376862, 0.1879013180732727, 0.05123550444841385, -0.2913530170917511, -0.34454625844955444, -0.09238672256469727, -0.1816118210554123, -0.0012826751917600632, -0.035988446325063705, 0.07129773497581482, 0.37686648964881897, 0.16384737193584442, 0.03225523978471756, -0.2014034390449524, -0.058261938393116, 0.38416942954063416, 0.11524109542369843, 0.06221017614006996, -0.08224429190158844, 0.39128321409225464, -0.023472528904676437, 0.02685277909040451, 0.0503184050321579, 0.027712710201740265, 0.08725352585315704, 0.425647109746933, 0.0977625846862793, 0.15738634765148163, -0.4112390875816345, 0.28078633546829224, 0.1774875968694687, 0.11828821897506714, -0.13777074217796326, 0.18009963631629944, 0.10431303083896637, -0.224992036819458, -0.23325394093990326, -0.19721508026123047, 0.12463855743408203, -0.1126864105463028, 0.16435353457927704, -0.11488209664821625, 0.3551517426967621, 0.3632742762565613, -0.009765916503965855, -0.24801523983478546, -0.3071395754814148, 0.04222584888339043, 0.05488782376050949, -0.1848628669977188, 0.27694782614707947, 0.4547150433063507, 0.06889287382364273, 0.10064513981342316, -0.44657522439956665, -0.09266727417707443, 0.08696957677602768, -0.3843233585357666, -0.08820292353630066, 0.34193098545074463, 0.19567814469337463, -0.15060898661613464, -0.20592045783996582, 0.15797919034957886, -0.11409712582826614, -0.13242556154727936, 0.36024948954582214, -0.24125291407108307, -0.14182516932487488, -0.054071806371212006, 0.3252495527267456, 0.02627585642039776, 0.0775642842054367, 0.05787578970193863, -0.10660601407289505, 0.06521810591220856, -0.006445923820137978, 0.10747767239809036, 0.06735964119434357, -0.01296483725309372, 0.12857083976268768, -0.11017973721027374, 0.2070382535457611, -0.6235056519508362, 0.35154885053634644, 0.3406658172607422, 0.04048712179064751, -0.2386520653963089, 0.1311081200838089, -0.06537596881389618, -0.26793164014816284, 0.33320534229278564, -0.07992002367973328, 0.20084738731384277, -0.3434513807296753, 0.08213335275650024, 0.28266215324401855, -0.054147496819496155, 0.06991564482450485, 0.20671403408050537, 0.4017025828361511, -0.08867113292217255, -0.05176244676113129, -0.05980346351861954, -0.17698337137699127, 0.3869963586330414, -0.4427589178085327, 0.16624006628990173, -0.09579436480998993, 0.9571192860603333, 0.3474618196487427, -0.24923890829086304, -0.24940219521522522, -0.14705103635787964, -0.34628552198410034, 0.1076839417219162, 0.17763426899909973, -0.03834070265293121, -0.0039489297196269035, -0.06391055136919022, 0.0028831809759140015, 0.05237804725766182, 0.15503710508346558, 0.22364646196365356, 0.5317481160163879, -0.35456860065460205, 0.299388587474823, -0.22832512855529785, 0.10329064726829529, 0.2751069962978363, 0.13227303326129913, -0.127249613404274, -0.01768401637673378, -0.009352385997772217, -0.04927118867635727, 0.026652716100215912, -0.32910609245300293, 0.06418018043041229, -0.14669649302959442, -0.2652188837528229, -0.21753938496112823, -0.04903087019920349, 0.03659139573574066, 0.04084458202123642, -0.06513194739818573, 0.032620396465063095, -0.12107520550489426, -0.10738211870193481, -0.052252963185310364, -0.3160066604614258, 0.348541796207428, 0.3667259216308594, 0.08986833691596985, -0.02784176729619503, 0.6427887678146362, 0.15500786900520325, -0.1114484965801239, 0.2014165222644806, 0.14983777701854706, -0.11773532629013062, -0.2359752655029297, 0.022950423881411552, 0.30525529384613037, -0.11438214033842087, -0.3438956141471863, -0.49479174613952637, 0.6355993747711182, 0.17985564470291138, -0.5693952441215515, -0.17015691101551056, -0.24646635353565216, -0.17225560545921326, -0.11563274264335632, 0.3047388195991516, 0.3367745876312256, 0.10700065642595291, -0.017186086624860764, -0.36785197257995605, 0.06726876646280289, 0.4848315119743347, -0.4563571512699127, -0.07332669198513031, -0.11960713565349579, 0.1667352020740509, 0.2647434175014496, -0.19481265544891357, -0.9240175485610962, -0.1374061107635498, 0.14410820603370667, -0.01407289132475853, -0.17872872948646545, 0.298997163772583, 0.24600133299827576, 0.2937111258506775, -0.2554592192173004, 0.5315841436386108, -0.049537889659404755, -0.09967632591724396, -0.4590686857700348, -0.21204854547977448 ]
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
[ 0.03105197474360466, -0.2103957235813141, 0.11031089723110199, 0.24948230385780334, 0.28502875566482544, -0.33688661456108093, 0.2980252504348755, 0.14864566922187805, 0.09188990294933319, 0.16699329018592834, -0.30923908948898315, 0.3939817249774933, -0.48465636372566223, 0.25464630126953125, 0.0468795970082283, -0.08954024314880371, -0.06799928098917007, 0.05894642695784569, 0.10282014310359955, -0.17925673723220825, -0.11254642903804779, 0.1559169441461563, 0.11106309294700623, -0.10564196109771729, -0.19486533105373383, 0.3722075819969177, -0.06696510314941406, 0.11051197350025177, -0.2348370999097824, -0.05581817403435707, 0.23994289338588715, -0.09313477575778961, 0.19716298580169678, 0.7407883405685425, -0.00012301238893996924, -0.29678821563720703, 0.10562047362327576, -0.013817953877151012, -0.39566442370414734, -0.22906653583049774, 0.39341676235198975, -0.3064042925834656, 0.14753983914852142, -0.18821850419044495, -0.023235976696014404, -0.04682978242635727, -0.2829231023788452, -0.11125530302524567, 0.13092002272605896, 0.24891841411590576, 0.03568290174007416, 0.12069830298423767, -0.04003624618053436, 0.10717988759279251, 0.1736428290605545, -0.19217033684253693, -0.025240492075681686, 0.5683524012565613, 0.3351752460002899, -0.019421137869358063, -0.22494250535964966, 0.1777404099702835, 0.01149292103946209, 0.10073009133338928, 0.661795973777771, 0.04060619696974754, -0.050766654312610626, -0.0463242381811142, 0.04265300929546356, 0.3301611840724945, -0.11374533921480179, -0.1908949315547943, -0.25507664680480957, 0.12239804118871689, 0.07248985022306442, -0.24157071113586426, -0.23092666268348694, 0.16072112321853638, 0.10856536030769348, -0.24158066511154175, -0.3910423517227173, -0.3491133749485016, -0.14091533422470093, 0.13794507086277008, -0.2982495427131653, 0.3092898726463318, -0.06806907057762146, -0.12711897492408752, 0.15347877144813538, 0.0853552371263504, 0.34188732504844666, -0.0955546572804451, 0.35434016585350037, 0.19703879952430725, 0.08634331077337265, -0.18238544464111328, 0.22257307171821594, 0.04210386052727699, -0.062443435192108154, -0.3093153238296509, -0.12535421550273895, 0.20691800117492676, -0.2814028859138489, 0.011797267943620682, -0.027306731790304184, 0.3203352689743042, 0.040419403463602066, 0.35977381467819214, 0.28627586364746094, 0.37072044610977173, 0.07752560079097748, -0.04536345601081848, -0.08907102048397064, -0.15487778186798096, 0.24097369611263275, 0.19222837686538696, -0.2800595164299011, -0.3287021517753601, -0.033770088106393814, -0.05419699102640152, -0.21769770979881287, -0.10590723901987076, 0.34869498014450073, 0.13636153936386108, -0.49230244755744934, -0.06439290940761566, 0.6397177577018738, 0.046870630234479904, -0.18193814158439636, 0.07984868437051773, -0.14149247109889984, -0.09508395195007324, -0.3073860704898834, 0.06758337467908859, 0.21250422298908234, -0.19385336339473724, 0.27882277965545654, 0.03200087696313858, 0.5187743306159973, 0.12528753280639648, 0.16171284019947052, 0.4335992932319641, 0.16154703497886658, -0.19240939617156982, -0.2416660487651825, -0.2961934506893158, 0.22628909349441528, 0.026455819606781006, -0.11240114271640778, 0.08938474953174591, -0.12872382998466492, -0.3263464570045471, -0.1520967185497284, 0.023959849029779434, -0.28712815046310425, -0.019190020859241486, -0.010954417288303375, 0.27803778648376465, 0.10987874120473862, -0.2487732321023941, -0.03219764679670334, -0.3951016068458557, -0.12442648410797119, 0.09461551159620285, 0.4095954895019531, -0.20389004051685333, -0.40325531363487244, -0.42737653851509094, 0.10449196398258209, 0.4528105556964874, 0.33462104201316833, 0.09447555243968964, -0.22268010675907135, -0.20602339506149292, 0.182228684425354, -0.35699212551116943, 0.24966296553611755, 0.1920621693134308, -0.27465391159057617, -0.22991600632667542, 0.13340473175048828, -0.25309333205223083, -0.21538230776786804, 0.025944694876670837, -0.1508408635854721, -0.06364277750253677, 0.3645898699760437, 0.11044356226921082, -0.18229268491268158, -0.02314547449350357, -0.3769245743751526, -0.1707116812467575, 0.24087752401828766, 0.14524507522583008, 0.3112166225910187, 0.04421839118003845, 0.13978596031665802, 0.545229434967041, 0.10949307680130005, -0.37678149342536926, 0.03864999860525131, -0.20870767533779144, 0.4394534230232239, -0.38880056142807007, 0.36291447281837463, -0.13229471445083618, -0.04362037777900696, 0.2286229431629181, -0.32526838779449463, 0.4498295783996582, 0.15172049403190613, 0.13739268481731415, -0.11641093343496323, -0.11465416103601456, 0.2857360541820526, -0.0415586419403553, -0.09858444333076477, -0.175385981798172, -0.28409701585769653, -0.22045296430587769, -0.147940993309021, -0.2546970546245575, -0.23914557695388794, -0.05173944681882858, 0.1921086609363556, 0.32608386874198914, -0.06349757313728333, -0.16392400860786438, -0.038680702447891235, 0.3862399756908417, 0.1850813329219818, 0.21768535673618317, -0.07578517496585846, -0.2285490334033966, 0.11393851041793823, -0.20823408663272858, 0.14065778255462646, 0.1960555911064148, 0.3851838707923889, 0.031623486429452896, 0.40312522649765015, 0.21448063850402832, -0.11146409064531326, -0.09128458797931671, -0.12201724201440811, 0.2812281847000122, 0.324666291475296, 0.2669116258621216, 0.2754969000816345, 0.12106325477361679, -0.24934867024421692, 0.10694250464439392, -0.04537442699074745, 0.10155414044857025, 0.35155966877937317, -0.16831736266613007, 0.06461989879608154, -0.16824544966220856, 0.1615699678659439, 0.29566341638565063, 0.2568989396095276, -0.059767477214336395, -0.0450265035033226, 0.21921998262405396, -0.46931278705596924, -0.27227386832237244, 0.011140124872326851, 0.07040084898471832, 0.08245272934436798, 0.07317212224006653, 0.4600592255592346, 0.14472335577011108, -0.01683230698108673, 0.02773122489452362, -0.08601688593626022, 0.14787472784519196, -0.11266074329614639, -0.06314841657876968, 0.12503033876419067, 0.06865347176790237, 0.07814255356788635, 0.10393920540809631, 0.05523441731929779, 0.1424945890903473, -0.18085737526416779, 0.33856457471847534, -0.006266901269555092, 0.4345984160900116, -0.249354749917984, -0.3888390064239502, -0.22619426250457764, -0.24832506477832794, -0.034891463816165924, 0.32378798723220825, -0.01339777372777462, -0.13628649711608887, 0.19925041496753693, 0.40451356768608093, -0.08564773201942444, -0.05584080517292023, -0.11347314715385437, -0.26379668712615967, -0.36608439683914185, -0.0012674015015363693, -0.06414400041103363, 0.06512763351202011, -0.08705278486013412, 0.03587926924228668, -0.15460710227489471, -0.1930072158575058, -0.7795090675354004, 0.06021347641944885, 0.03927228972315788, 0.30001893639564514, 0.20597079396247864, -0.1349009871482849, -0.31092119216918945, 0.044446591287851334, 0.10181473940610886, -0.37420210242271423, -0.026990581303834915, 0.417681485414505, -0.23497293889522552, 0.07670997828245163, -0.03322906792163849, -0.22106102108955383, -0.12886755168437958, -0.02256774716079235, -0.05517931282520294, 0.12109814584255219, 0.0652790516614914, -0.20432519912719727, 0.13932447135448456, 0.3219204246997833, 0.41521957516670227, -0.09877263754606247, -0.06713970005512238, 0.22492268681526184, 0.42780059576034546, -0.25694626569747925, -0.301351934671402, -0.18492011725902557, -0.0005777478218078613, 0.017254244536161423, 0.0003764871507883072, -0.3237825036048889, -0.36015433073043823, -0.11460591852664948, -0.15125621855258942, 0.20219986140727997, 0.16915762424468994, 0.06157908961176872, 0.224295973777771, 0.06467749178409576, -0.21677854657173157, -0.16279278695583344, -0.07023169845342636, 0.24304130673408508, 0.6987829208374023, -0.23153802752494812, -0.09995263069868088, 0.0897039920091629, 0.752469003200531, 0.13057518005371094, -0.423186719417572, 0.06254187226295471, 0.26073333621025085, 0.06086135655641556, -0.11394082754850388, -0.26714572310447693, 0.15565060079097748, -0.3145396411418915, -0.27640247344970703, 0.3156626224517822, 0.11460062116384506, -0.28042182326316833, -0.26113060116767883, -0.2004973441362381, 0.49796172976493835, -0.17563274502754211, 0.06963612884283066, 0.03468531370162964, -0.04997820407152176, 0.02282186597585678, -0.1170809417963028, -0.27406227588653564, -0.11454538255929947, 0.2060009241104126, 0.02208690345287323, 0.0904163047671318, 0.03668069466948509, 0.45275968313217163, 0.015573451295495033, -0.7020947933197021, 0.4291396141052246, 0.09122688323259354, -0.6415270566940308, 0.20079074800014496, -0.15091948211193085, -0.08306723833084106, -0.05323367193341255, -0.3633190393447876, 0.11488113552331924, 0.0365847684442997, -0.06340625137090683, -0.024901432916522026, -0.28089118003845215, 0.019936252385377884, -0.26069605350494385, -0.3109963834285736, 0.6108319163322449, 0.30939918756484985, -0.3783528208732605, 0.05132340267300606, 0.23827819526195526, 0.4477608799934387, -0.1608956903219223, 0.03582214191555977, -0.3568879961967468, -0.21575215458869934, 0.09272745251655579, -0.11704526096582413, -0.03972261771559715, 0.2884737551212311, -0.12854893505573273, -0.07872042804956436, 0.24400566518306732, -0.08774784207344055, 0.1997191160917282, 0.049700599163770676, 0.10960305482149124, -0.09483009576797485, 0.17061875760555267, 0.05049663782119751, 0.06471322476863861, 0.24659648537635803, 0.41444963216781616, -0.5077788233757019, -0.47134923934936523, 0.21681636571884155, 0.16121608018875122, 0.17137649655342102, 0.1093549132347107, 0.10491938143968582, -0.13601121306419373, 0.43549782037734985, 0.27228671312332153, 0.18568487465381622, 0.23380330204963684, 0.039173658937215805, 0.36824843287467957, 0.11878341436386108, -0.19488810002803802, 0.3924102485179901, -0.170674130320549, 0.014420703053474426, 0.2270651012659073, 0.06649242341518402, -0.012271774932742119, 0.12193980813026428, 0.4338003396987915, 0.9371415972709656, 0.14221984148025513, 0.09444331377744675, 0.2991996705532074, -0.13560274243354797, 0.3273821175098419, -0.045953571796417236, 0.02885357476770878, -0.3419894278049469, -0.2665644884109497, 0.043637558817863464, -0.04958975315093994, -0.44529539346694946, -0.16226740181446075, -0.19859670102596283, 0.14733000099658966, -0.3593901991844177, -0.25683680176734924, 0.2827572226524353, 0.2808370292186737, 0.10884825885295868, 0.030932091176509857, 0.2770262658596039, 0.08055364340543747, -0.17511048913002014, 0.26924458146095276, -0.08487468957901001, -0.1843993365764618, -0.26210451126098633, 0.13207873702049255, -0.18024921417236328, 0.37191352248191833, -0.10059570521116257, -0.044855475425720215, -0.060901377350091934, -0.3492479920387268, 0.0737575888633728, 0.0327632874250412, 0.7054666876792908, 0.09231684356927872, -0.3070915639400482, 0.08027985692024231, -0.062137700617313385, 0.17744171619415283, 0.19616656005382538, -0.14955508708953857, 0.35956045985221863, -0.13380126655101776, -0.1622401773929596, -0.16799667477607727, 0.14697028696537018, -0.3749425411224365, 0.22124546766281128, -0.16459882259368896, -0.045548535883426666, -0.11824847757816315, 0.40338510274887085, 0.33453625440597534, 0.1593393236398697, -0.2553209662437439, 0.07714979350566864, 0.20327964425086975, -0.2602023780345917, 0.045820098370313644, 0.11606155335903168, -0.2364221066236496, 0.0402362160384655, -0.009904071688652039, 0.021524876356124878, -0.0013384893536567688, 0.23494352400302887, 0.17420583963394165, -0.11456694453954697, 0.23590180277824402, -0.6396331787109375, 0.5488191246986389, -0.3684130609035492, 0.1757143884897232, -0.1360732764005661, -0.055178992450237274, 0.12224508821964264, 0.5733035206794739, 0.4357832074165344, 0.21459415555000305, -0.3986991345882416, 0.18546873331069946, -0.3889482319355011, 0.08772922307252884, -0.18641024827957153, 0.3603321313858032, -0.17637723684310913, 0.0029720868915319443, -0.11693356931209564, 0.23120135068893433, -0.20673413574695587, 0.03993794322013855, 0.19829165935516357, -0.03068380430340767, 0.009558885358273983, -0.4418536424636841, 0.02748611569404602, 0.03583340346813202, -0.23903554677963257, -0.0639573484659195, 0.11508999019861221, -0.02289605140686035, -0.4367915391921997, 0.1588520109653473, 0.2320312261581421, 0.009140817448496819, -0.32059186697006226, -0.38937908411026, -0.1304398477077484, -0.1896282583475113, 0.06611350178718567, 0.027370154857635498, 0.0649438351392746, 0.21625937521457672, 0.2079448103904724, 0.020306043326854706, -0.16019943356513977, -0.04887603968381882, 0.41021987795829773, 0.16648009419441223, 0.009113239124417305, -0.034340761601924896, 0.42611533403396606, -0.02806224673986435, 0.19161036610603333, 0.024117454886436462, -0.025749370455741882, 0.16406989097595215, 0.4158223867416382, 0.13150396943092346, 0.15301206707954407, -0.39151960611343384, 0.31367525458335876, 0.2520207464694977, 0.20041455328464508, -0.20123547315597534, 0.2113136351108551, 0.08531980961561203, -0.24712762236595154, -0.1780570149421692, -0.20906847715377808, 0.050173066556453705, -0.1371205747127533, 0.17602597177028656, -0.11124125123023987, 0.36048269271850586, 0.3827468752861023, -0.061240192502737045, -0.23618574440479279, -0.26754826307296753, 0.05084863677620888, 0.044831614941358566, -0.2170989215373993, 0.31279855966567993, 0.5844203233718872, 0.04400472342967987, 0.05221609026193619, -0.3825650215148926, -0.11984279006719589, 0.008465098217129707, -0.2900795638561249, -0.10884629189968109, 0.3588649034500122, 0.20460045337677002, -0.10920138657093048, -0.09902498126029968, 0.15128180384635925, -0.0717029795050621, -0.12134341895580292, 0.39471155405044556, -0.40146973729133606, -0.16225993633270264, -0.03691476956009865, 0.3568076491355896, -0.02662954106926918, 0.06826359033584595, 0.10880716145038605, -0.21613872051239014, 0.07733315229415894, -0.015025246888399124, -0.01827465370297432, 0.09222211688756943, 0.07474467158317566, 0.10134054720401764, -0.09447591006755829, 0.1506623923778534, -0.7600218653678894, 0.3500680923461914, 0.2666798233985901, 0.07755827903747559, -0.1919763684272766, 0.1510920524597168, -0.11532817780971527, -0.23615196347236633, 0.3830995559692383, -0.016696017235517502, 0.25774872303009033, -0.3252435624599457, -0.041428934782743454, 0.4066010117530823, -0.02356446161866188, 0.043433502316474915, 0.1191042810678482, 0.4557110667228699, -0.13713541626930237, -0.08016903698444366, -0.08495017886161804, -0.17106102406978607, 0.334772527217865, -0.3921130895614624, 0.2219248265028, 0.057305265218019485, 0.8864078521728516, 0.3305407762527466, -0.20438870787620544, -0.27269217371940613, -0.09083288162946701, -0.29800689220428467, 0.04675421491265297, 0.1034170389175415, -0.08217354118824005, 0.027585923671722412, -0.027783703058958054, 0.006058171391487122, 0.12412020564079285, 0.1421758532524109, 0.26756611466407776, 0.29824957251548767, -0.3599909842014313, 0.2530735731124878, -0.2842276692390442, 0.15552237629890442, 0.1820707768201828, 0.06622336804866791, -0.23234081268310547, -0.0077626658603549, 0.0711766928434372, -0.10512158274650574, 0.057173073291778564, -0.23002056777477264, 0.07925764471292496, -0.21019110083580017, -0.234105184674263, -0.2609882652759552, 0.06782655417919159, -0.014428498223423958, 0.004517506808042526, -0.08634749054908752, 0.02315201237797737, -0.11450622975826263, -0.09457816183567047, 0.021241137757897377, -0.2878338098526001, 0.3728131353855133, 0.3003571629524231, 0.12165670096874237, 0.05715637654066086, 0.5850411653518677, 0.12948980927467346, -0.17101655900478363, 0.26426154375076294, 0.13617144525051117, -0.10343565791845322, -0.18146252632141113, -0.043379973620176315, 0.2637508809566498, -0.050897277891635895, -0.2635940611362457, -0.5351039171218872, 0.550689160823822, 0.11400303244590759, -0.5117972493171692, -0.2289525866508484, -0.16178415715694427, -0.18558961153030396, -0.10991455614566803, 0.24824468791484833, 0.3811723589897156, 0.13884267210960388, -0.02862614206969738, -0.4515787363052368, 0.00406445749104023, 0.465069979429245, -0.4754205346107483, -0.12336903810501099, -0.1212315633893013, 0.18406851589679718, 0.29907095432281494, -0.12919802963733673, -0.8295280337333679, -0.1543295979499817, 0.09583029896020889, -0.038087889552116394, -0.13884219527244568, 0.2332545816898346, 0.11724242568016052, 0.2540249228477478, -0.29439660906791687, 0.5520853400230408, 0.04639887809753418, -0.16173940896987915, -0.4791439175605774, -0.1919277459383011 ]
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 :)
[ -0.09875033795833588, 0.2994914650917053, -0.1906658262014389, 0.08413086831569672, -0.09124993532896042, 0.10564102977514267, 0.4507472515106201, -0.14683814346790314, -0.21770641207695007, 0.004500426352024078, -0.20793591439723969, -0.17496585845947266, -0.15123486518859863, 0.27440688014030457, 0.07923612743616104, -0.36933478713035583, -0.001966707408428192, 0.033536963164806366, 0.0868884027004242, -0.34479355812072754, -0.10768920928239822, 0.17830190062522888, -0.049756813794374466, 0.19358950853347778, -0.00017957063391804695, 0.09124097973108292, -0.22928372025489807, -0.15842890739440918, -0.1509992629289627, -0.02232060395181179, 0.13173751533031464, 0.24340614676475525, 0.07313701510429382, 0.38489121198654175, -0.00010290565842296928, -0.21261009573936462, 0.12683197855949402, -0.12616373598575592, 0.029008451849222183, 0.032224707305431366, 0.02200361341238022, -0.2267683744430542, -0.3399676978588104, 0.037781842052936554, -0.1280568689107895, -0.2536057233810425, -0.18752998113632202, -0.5613652467727661, -0.1343904435634613, 0.5067586302757263, 0.24277807772159576, -0.17855897545814514, 0.13484226167201996, 0.014460690319538116, 0.1603565514087677, 0.2705265283584595, -0.32320085167884827, -0.1657284051179886, 0.6291850805282593, 0.14990399777889252, 0.034464359283447266, 0.299895703792572, 0.22463494539260864, -0.37106797099113464, -0.2069711983203888, -0.13152313232421875, -0.05224614590406418, -0.3681877553462982, 0.027923583984375, 0.35658785700798035, 0.6327370405197144, -0.08151998370885849, -0.5868775248527527, -0.09462009370326996, 0.24371466040611267, -0.07554932683706284, -0.10097461193799973, -0.13449391722679138, 0.2280341386795044, 0.19478893280029297, 0.3044319748878479, -0.40359747409820557, -0.06974363327026367, 0.0990145206451416, 0.08143571019172668, 0.009418077766895294, -0.05463429540395737, -0.017630569636821747, 0.11922620236873627, -0.16042016446590424, 0.010838668793439865, 0.16477179527282715, -0.060169246047735214, 0.36071068048477173, -0.21490664780139923, -0.3206721246242523, 0.06903428584337234, -0.05932850390672684, 0.3171025812625885, 0.07181338965892792, -0.608530580997467, 0.03394018113613129, -0.4754469394683838, -0.13653287291526794, 0.23075494170188904, 0.1410786360502243, 0.11318837851285934, -0.2693191468715668, 0.1695142388343811, -0.34914854168891907, 0.12838324904441833, -0.00893649086356163, -0.006788323167711496, -0.025769703090190887, -0.28193041682243347, 0.06205940619111061, -0.09378920495510101, -0.3450583219528198, 0.00953446514904499, 0.028470683842897415, 0.19382166862487793, -0.2102639526128769, -0.10387542098760605, 0.19578397274017334, 0.0845952033996582, 0.31471413373947144, -0.18962429463863373, 0.19519789516925812, -0.1819043904542923, -0.6139768958091736, -0.15016724169254303, 0.04668860137462616, -0.2589320242404938, 0.033653706312179565, 0.1179800033569336, -0.3058912456035614, 0.04503721743822098, 0.10600259155035019, -0.03552224487066269, 0.24374499917030334, 0.18960943818092346, -0.330634206533432, -0.14126470685005188, -0.2343374639749527, -0.2384476661682129, 0.08291614800691605, 0.006255749613046646, 0.29035305976867676, -0.27429625391960144, -0.132715106010437, 0.0041040703654289246, -0.0777202919125557, -0.549354612827301, 0.25537705421447754, 0.12873254716396332, -0.2849130630493164, -0.05030667036771774, 0.7471095323562622, -0.1838758885860443, -0.2807391285896301, 0.06724202632904053, -0.05907435715198517, -0.09857086837291718, -0.06438565999269485, 0.035185277462005615, 0.33346879482269287, -0.5780576467514038, -0.14078916609287262, -0.21537350118160248, -0.3041473925113678, 0.26338276267051697, 0.24018315970897675, -0.013686489313840866, 0.12782764434814453, -0.27479541301727295, -0.02918938174843788, 0.054756857454776764, -0.05104214698076248, -0.007746802642941475, 0.19158364832401276, -0.17146112024784088, 0.25183263421058655, 0.005483254790306091, 0.07924217730760574, -0.01534961722791195, -0.07730845361948013, 0.29996687173843384, 0.3543703556060791, -0.2617481052875519, -0.07039570063352585, -0.036199793219566345, -0.4247276782989502, 0.16641637682914734, 0.1873418539762497, 0.07433468103408813, -0.0027771443128585815, 0.1444615125656128, -0.23539042472839355, 0.13004983961582184, -0.06374673545360565, 0.223867729306221, 0.2670321464538574, 0.13562460243701935, 0.52492356300354, 0.06076273322105408, 0.04892569035291672, -0.15978315472602844, -0.015663962811231613, 0.10479092597961426, 0.4020426273345947, 0.31529003381729126, -0.1573580801486969, -0.025230739265680313, -0.12310608476400375, -0.24866123497486115, -0.36729195713996887, 0.24655038118362427, 0.2638358175754547, 0.019912991672754288, 0.33392003178596497, -0.357746422290802, 0.2866944372653961, -0.22801707684993744, -0.25668105483055115, -0.1881212592124939, 0.2682744860649109, -0.08029606938362122, 0.02375076897442341, -0.1958562582731247, 0.35900285840034485, -0.10002193599939346, -0.0650896430015564, -0.10964334011077881, -0.1666446030139923, 0.21401216089725494, 0.1449785828590393, 0.5165335536003113, 0.2998847961425781, 0.2574204206466675, -0.24572421610355377, -0.1387699991464615, 0.18689005076885223, 0.06296168267726898, -0.20906659960746765, 0.14302563667297363, 0.2379070222377777, 0.18251925706863403, 0.051116518676280975, 0.06885787844657898, -0.024304836988449097, 0.3960515856742859, 0.21396595239639282, -0.07813738286495209, -0.1446685492992401, 0.30037227272987366, -0.06700195372104645, 0.28173166513442993, -0.07894673198461533, -0.39791518449783325, -0.021499350666999817, 0.3684978783130646, 0.022610630840063095, 0.11486069113016129, 0.1395164430141449, -0.1527702510356903, -0.10070927441120148, 0.21678513288497925, -0.48139747977256775, 0.12765121459960938, 0.26268669962882996, 0.22171497344970703, -0.16540634632110596, 0.25981947779655457, 0.03460785746574402, 0.04588094353675842, -0.25480198860168457, -0.10957492142915726, -0.039888568222522736, 0.1586693376302719, 0.07674157619476318, -0.34561803936958313, -0.4028259515762329, -0.0025885868817567825, 0.08199048042297363, 0.049554575234651566, -0.19892188906669617, 0.07890412211418152, -0.05405363813042641, 0.23634327948093414, -0.3461318910121918, -0.2608318626880646, -0.09982599318027496, 0.24856789410114288, -0.33776822686195374, -0.006427828222513199, 0.042945168912410736, -0.1900702714920044, 0.32391369342803955, -0.03564951941370964, -0.00315234437584877, -0.24567316472530365, 0.00588980270549655, -0.4746783971786499, 0.29337117075920105, -0.053815145045518875, -0.31592875719070435, 0.38174858689308167, -0.07545958459377289, 0.1476750671863556, -0.33012962341308594, -0.5107157230377197, 0.1697302758693695, -0.12999437749385834, 0.575386106967926, 0.12542343139648438, 0.02576528489589691, 0.013467654585838318, -0.12582723796367645, 0.02396179921925068, -0.05465292930603027, -0.1641533523797989, -0.11448454856872559, -0.08873169124126434, 0.42236241698265076, 0.0386972539126873, -0.45314255356788635, -0.42036664485931396, -0.1723112016916275, 0.2529330849647522, -0.16396479308605194, -0.0030090510845184326, 0.22998592257499695, -0.1378037929534912, 0.13826453685760498, 0.009906891733407974, 0.13187910616397858, -0.26472756266593933, 0.027845803648233414, 0.20042890310287476, -0.3106869161128998, -0.2774639129638672, 0.061187949031591415, -0.09069793671369553, 0.0852125734090805, -0.24416659772396088, -0.3625800609588623, 0.2580147683620453, -0.09587226808071136, 0.27385061979293823, -0.19152070581912994, 0.07490339875221252, 0.1914641261100769, 0.16966593265533447, -0.06745731085538864, 0.11042968183755875, 0.007872123271226883, -0.13218480348587036, 0.010557470843195915, 0.0869903415441513, -0.10921663790941238, -0.0853852927684784, -0.011631377041339874, 0.2945597767829895, -0.011589999310672283, 0.06750217825174332, 0.2806149125099182, 0.026751503348350525, 0.23155353963375092, -0.3706285059452057, -0.4429999589920044, 0.35534635186195374, 0.040036000311374664, 0.04863486438989639, 0.41959822177886963, 0.21355536580085754, -0.2142554670572281, -0.20989453792572021, 0.05257746949791908, 0.18030855059623718, -0.07134474813938141, 0.040761854499578476, -0.10820119082927704, 0.16658490896224976, -0.24493514001369476, -0.2167084515094757, -0.17305870354175568, -0.10115177929401398, 0.25474393367767334, 0.3276863694190979, 0.093620665371418, -0.07407154887914658, -0.22788836061954498, 0.23817841708660126, -0.300315260887146, -0.044358447194099426, 0.08408119529485703, 0.25461193919181824, -0.2716078460216522, 0.33530646562576294, 0.030030101537704468, 0.15427611768245697, 0.21860197186470032, -0.6904956102371216, -0.4939383268356323, 0.13585558533668518, 0.3060668706893921, 0.1556713879108429, 0.1767359972000122, -0.1837189644575119, -0.22654761373996735, 0.30841830372810364, 0.2455105483531952, -0.16246208548545837, 0.029782280325889587, 0.19417054951190948, 0.42701590061187744, -0.0171998031437397, -0.046921733766794205, 0.0955195277929306, -0.16586410999298096, -0.09738194942474365, 0.39466896653175354, -0.02931247651576996, 0.017172763124108315, 0.08575226366519928, 0.137074276804924, -0.0335642471909523, 0.2016945779323578, 0.2797214388847351, -0.059396203607320786, -0.16069501638412476, 0.15213672816753387, 0.0868341326713562, -0.04038210213184357, -0.48504725098609924, 0.07861019670963287, 0.19921687245368958, -0.020880240947008133, -0.5251501202583313, 0.0876171886920929, -0.23867137730121613, 0.4695502817630768, 0.16023927927017212, 0.2618004381656647, 0.32919129729270935, -0.2775520384311676, -0.21814961731433868, 0.030388478189706802, 0.2831471264362335, 0.4071330428123474, -0.03894490376114845, -0.43501928448677063, -0.776552677154541, -0.0014821141958236694, -0.09129558503627777, 0.016643479466438293, 0.41284841299057007, 0.43938320875167847, -0.13270777463912964, 0.24752633273601532, 0.076760433614254, 0.9190419912338257, 0.23129485547542572, 0.041549693793058395, 0.07719247043132782, -0.08021663874387741, 0.1929486244916916, -0.3251569867134094, -0.13492462038993835, 0.04967120662331581, -0.23054398596286774, 0.005983946844935417, 0.26198527216911316, 0.34716683626174927, 0.2499637007713318, -0.23360739648342133, 0.10049983859062195, 0.09623244404792786, -0.19594307243824005, 0.29610949754714966, 0.3082045912742615, 0.10677415877580643, -0.4312557280063629, 0.03384609892964363, 0.23043572902679443, 0.043792806565761566, 0.20384739339351654, -0.07254531979560852, -0.05668361485004425, 0.10696379840373993, 0.1314014494419098, -0.45485883951187134, -0.07609505951404572, -0.5070959329605103, 0.12641814351081848, -0.4729512631893158, 0.08941704034805298, 0.2029142826795578, -0.08584138751029968, 0.42142343521118164, 0.47171857953071594, 0.040915802121162415, 0.022925127297639847, 0.47623881697654724, 0.06283987313508987, -0.08770649880170822, -0.23555314540863037, 0.2755262553691864, -0.050539638847112656, -0.2636359632015228, 0.32308340072631836, -0.20167532563209534, -0.13621468842029572, -0.38254696130752563, -0.1163196712732315, -0.10659092664718628, -0.04572846740484238, -0.17525967955589294, 0.41036543250083923, 0.17232657968997955, -0.19534622132778168, 0.14802269637584686, 0.3497508764266968, -0.2356005609035492, -0.19416096806526184, 0.10732913017272949, -0.09747309982776642, -0.3136826455593109, -0.22666007280349731, 0.24906349182128906, -0.6465165615081787, 0.29450729489326477, 0.4389953315258026, -0.15036293864250183, -0.23534393310546875, 0.08640588074922562, 0.44002726674079895, -0.4701535403728485, -0.38490724563598633, 0.0059770564548671246, -0.14805962145328522, -0.042697399854660034, 0.2869289219379425, 0.06660977005958557, -0.2061111330986023, 0.1507674604654312, -0.09458740800619125, -0.36929062008857727, -0.028940849006175995, 0.1487242430448532, 0.12537972629070282, 0.27110984921455383, -0.2238439917564392, 0.03578522801399231, 0.07909592986106873, -0.3972316384315491, -0.10924950987100601, -0.10658740252256393, 0.0030478201806545258, 0.24276475608348846, 0.05568774417042732, 0.014138064347207546, 0.03401078283786774, 0.19847622513771057, -0.34704339504241943, -0.07309698313474655, -0.18337324261665344, -0.1146322637796402, 0.07452747970819473, 0.13493061065673828, 0.1228223368525505, -0.077947698533535, -0.035056471824645996, -0.10303807258605957, -0.14566455781459808, 0.11736613512039185, 0.07653297483921051, -0.07971978187561035, 0.25643283128738403, -0.09329646825790405, 0.011540405452251434, -0.16124552488327026, 0.0922037810087204, 0.2833353281021118, 0.04205460846424103, 0.28873109817504883, 0.16807055473327637, 0.18502826988697052, -0.13068319857120514, 0.08269279450178146, 0.2942376434803009, -0.004716634750366211, 0.3498418927192688, -0.1388331651687622, 0.037948597222566605, 0.28711998462677, 0.30442380905151367, -0.07962283492088318, 0.02395400032401085, -0.10387270152568817, -0.038740258663892746, 0.3412598669528961, 0.28051409125328064, -0.3252316117286682, 0.16337814927101135, 0.10603345930576324, -0.14815394580364227, 0.19855047762393951, 0.2169756293296814, 0.32979583740234375, 0.29217708110809326, 0.4217357635498047, -0.15993690490722656, -0.13561317324638367, 0.010873289778828621, 0.4031975567340851, 0.3240828514099121, -0.037429079413414, -0.1339365839958191, 0.09704168140888214, 0.264242559671402, 0.010727807879447937, 0.274061918258667, -0.13993841409683228, 0.32852232456207275, 0.19472768902778625, 0.2168823778629303, -0.012067921459674835, -0.2332325428724289, 0.4450658857822418, 0.3071078658103943, 0.28177204728126526, 0.22357061505317688, 0.10950889438390732, -0.14016768336296082, -0.1730029135942459, -0.22349724173545837, -0.04635661840438843, 0.0801677256822586, -0.2020120918750763, -0.001011483371257782, 0.4020218551158905, -0.3433103561401367, -0.18070578575134277, -0.22531436383724213, -0.053367067128419876, -0.0010698288679122925, -0.2768160402774811, 0.112814761698246, 0.1313040852546692, -0.3100360333919525, 0.06996661424636841, 0.02971084788441658, 0.026648633182048798, -0.06110558658838272, 0.2780030071735382, 0.010399229824543, 0.17701563239097595, 0.047720108181238174, 0.4197746217250824, -0.13973702490329742, 0.08330446481704712, 0.016874466091394424, 0.31802231073379517, -0.20493702590465546, -0.01801310107111931, 0.08799592405557632, 0.32789117097854614, 0.06277631223201752, 0.2459595799446106, 0.320583313703537, 0.12034431099891663, -0.1926158219575882, -0.015043161809444427, 0.17180296778678894, -0.19165131449699402, -0.41362860798835754, 0.012699954211711884, -0.03840883821249008, 0.08607941120862961, -0.37970414757728577, -0.031050771474838257, -0.12854592502117157, -0.13488523662090302, 0.3256210684776306, -0.2585844397544861, -0.10838155448436737, -0.07899247854948044, 0.13846108317375183, 0.16617566347122192, -0.2029038965702057, 0.4855582118034363, 0.23405641317367554, -0.16686522960662842, -0.21744246780872345, -0.4381391108036041, 0.051265373826026917, 0.08015064895153046, 0.1744614839553833, -0.1245618611574173, 0.06258368492126465, 0.3831964433193207, 0.11679412424564362, -0.042772114276885986, -0.06421244889497757, 0.16320037841796875, 0.3232917785644531, -0.34794703125953674, 0.0629393607378006, 0.2859339714050293, 0.4084406793117523, -0.030207963660359383, -0.018503978848457336, -0.07329057157039642, 0.41259074211120605, 0.1505436897277832, -0.13348513841629028, -0.2462119311094284, 0.13039067387580872, 0.12726637721061707, 0.3557191491127014, 0.22650910913944244, 0.5275260210037231, -0.22152623534202576, -0.037970881909132004, -0.2685335576534271, -0.22323279082775116, -0.06925410777330399, 0.17556259036064148, 0.02854483760893345, 0.06362191587686539, -0.11477160453796387, -0.31324541568756104, -0.24616079032421112, 0.203011617064476, 0.17591363191604614, 0.2730839252471924, -0.5236432552337646, 0.21069705486297607, 0.09567048400640488, -0.37458232045173645, -0.18295717239379883, 0.14599262177944183, 0.041923414915800095, -0.036862365901470184, -0.09746697545051575, -0.05345077067613602, 0.0436975359916687, -0.1767207235097885, -0.23924516141414642, -0.1110527440905571, 0.13255849480628967, 0.31383413076400757, 0.1159164160490036, -0.36679232120513916, 0.17310664057731628, 0.2874091863632202, 0.22819122672080994, 0.28344497084617615, -0.02660796046257019, -0.020219523459672928, 0.03794587776064873, -0.2808718681335449, 0.3004860281944275, 0.1155647337436676, 0.09652014821767807, 0.1939515769481659, -0.22160011529922485 ]
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`
[ -0.20377331972122192, 0.10567396879196167, -0.03363513946533203, 0.06406787782907486, -0.10075315833091736, -0.1691393107175827, 0.13653059303760529, 0.10918207466602325, -0.3450912833213806, 0.1280626356601715, -0.16435293853282928, 0.5187444686889648, 0.18093863129615784, 0.08213922381401062, 0.16695699095726013, 0.07651332020759583, 0.06449422985315323, 0.46056950092315674, 0.04686892777681351, 0.15668262541294098, -0.2131081521511078, 0.4624912738800049, -0.4079589247703552, 0.1537039875984192, -0.23863652348518372, -0.16774117946624756, -0.20612671971321106, 0.27909159660339355, -0.5138829946517944, -0.27585065364837646, 0.297965943813324, 0.14172382652759552, -0.008438993245363235, 0.18113583326339722, -0.00011696519504766911, -0.23673871159553528, 0.12079568952322006, -0.051851581782102585, -0.3485306203365326, -0.3542289137840271, -0.008578561246395111, -0.4436558187007904, 0.15956906974315643, -0.34196075797080994, 0.14092859625816345, 0.3921468257904053, 0.3511866331100464, 0.37066900730133057, 0.3492595851421356, 0.3874807357788086, 0.15486444532871246, 0.3137633800506592, 0.21581004559993744, -0.054475974291563034, 0.16369754076004028, -0.2806618809700012, -0.2976788580417633, 0.32439976930618286, 0.5822188258171082, -0.16857078671455383, 0.0035515576601028442, 0.1059035211801529, -0.09550733119249344, 0.05590522289276123, 0.12202002853155136, 0.1638885736465454, -0.25351929664611816, -0.42911747097969055, 0.2994453012943268, 0.12129329144954681, 0.0607747808098793, -0.2001691460609436, -0.01815180666744709, 0.0014596953988075256, 0.2585047483444214, -0.3444949984550476, 0.21898558735847473, 0.3701314330101013, -0.34308353066444397, 0.07960154861211777, 0.0347071997821331, -0.032762810587882996, -0.217060849070549, 0.230853870511055, 0.056651949882507324, 0.3647039234638214, -0.06477495282888412, 0.010259903967380524, 0.2563156485557556, 0.07807134091854095, 0.22043833136558533, 0.2324579954147339, -0.19939710199832916, -0.01257722731679678, -0.020614508539438248, -0.19665183126926422, -0.15251672267913818, 0.3759249150753021, 0.23538938164710999, -0.3535800576210022, 0.196884885430336, -0.06267080456018448, 0.19996103644371033, 0.19693246483802795, 0.02321106567978859, -0.012032496742904186, 0.35832950472831726, -0.04169832170009613, 0.39903324842453003, 0.4018419086933136, -0.02484768256545067, 0.0016648806631565094, 0.04790092259645462, -0.177496999502182, -0.2790900468826294, -0.12938418984413147, 0.06468921899795532, -0.189718559384346, -0.2175167053937912, -0.016829531639814377, -0.15009905397891998, 0.27821674942970276, 0.03317305073142052, 0.19462904334068298, -0.06817463040351868, -0.23874244093894958, 0.10872025787830353, -0.022551842033863068, -0.3674684762954712, -0.17430125176906586, 0.009949443861842155, -0.0019467584788799286, -0.1876400262117386, 0.10968397557735443, 0.2619292140007019, 0.18368858098983765, 0.4778350293636322, -0.15345947444438934, 0.09960543364286423, -0.11781322956085205, 0.052396245300769806, -0.2653234302997589, -0.10134170204401016, 0.3144099712371826, 0.3155171573162079, -0.22006909549236298, -0.2717234492301941, -0.5757358074188232, 0.04511953145265579, 0.1976000964641571, -0.03203395754098892, -0.012889672070741653, -0.12502753734588623, 0.10841798782348633, -0.2844674587249756, -0.07026495039463043, -0.3243393003940582, 0.054984934628009796, -0.21550452709197998, -0.25893157720565796, 0.04896059259772301, -0.1539314091205597, -0.245650514960289, -0.0019285448361188173, -0.058479346334934235, 0.18696627020835876, -0.0323915034532547, -0.09881709516048431, 0.16054767370224, 0.00841245986521244, -0.0025032609701156616, 0.2557321786880493, -0.14600391685962677, -0.08713565021753311, 0.06068771332502365, 0.19247467815876007, 0.38259610533714294, -0.3142409026622772, -0.7157806158065796, 0.12331622838973999, -0.07620730251073837, -0.2864677309989929, -0.09852322190999985, 0.181364044547081, 0.4203035235404968, -0.18550926446914673, 0.07004640251398087, 0.4979860186576843, -0.17399761080741882, 0.1083311215043068, -0.2697705924510956, -0.17382073402404785, -0.12889905273914337, 0.05015195533633232, 0.11421224474906921, -0.017191380262374878, -0.03910631686449051, 0.7044433951377869, 0.16681408882141113, -0.09890884160995483, -0.20148733258247375, 0.05992184951901436, 0.3261891007423401, -0.3010638356208801, -0.0882209837436676, -0.14634931087493896, -0.5062057971954346, 0.01004686951637268, -0.3460908532142639, 0.21325284242630005, 0.0968346819281578, -0.21113018691539764, -0.05918348208069801, -0.06821369379758835, -0.01218407228589058, -0.17810165882110596, 0.10006794333457947, 0.020486939698457718, 0.048870284110307693, -0.04860570281744003, 0.09019947052001953, 0.39791059494018555, -0.19552581012248993, 0.4248863756656647, -0.20517784357070923, 0.18992064893245697, -0.34831029176712036, -0.36242300271987915, 0.06926039606332779, 0.31324702501296997, 0.0034317485988140106, -0.1705968677997589, -0.10006601363420486, 0.2572607696056366, 0.08105379343032837, -0.07469089329242706, 0.10122917592525482, -0.36668211221694946, 0.3960619270801544, 0.06066770479083061, -0.04035735875368118, -0.06082839146256447, 0.012146678753197193, -0.03536980599164963, -0.02880086749792099, 0.2245766669511795, -0.18602298200130463, -0.04142371565103531, 0.17305608093738556, 0.20941871404647827, 0.039265647530555725, -0.11988596618175507, -0.20445342361927032, 0.1320449560880661, 0.32316601276397705, 0.2368137091398239, 0.04117957130074501, 0.010915195569396019, -0.36299949884414673, 0.21057859063148499, 0.4938598871231079, 0.027466241270303726, 0.24809953570365906, 0.01987292990088463, 0.04852696508169174, 0.0026454664766788483, 0.34254059195518494, 0.08806069195270538, 0.3751702308654785, 0.11413420736789703, -0.11013118922710419, -0.003826858475804329, -0.18535751104354858, -0.30785664916038513, 0.1314210295677185, 0.0830453634262085, 0.1299072653055191, -0.16973936557769775, 0.3030274212360382, 0.014845994301140308, -0.36467915773391724, -0.1110706627368927, -0.07071467489004135, 0.23503398895263672, 0.10903173685073853, 0.3059641122817993, -0.14340360462665558, -0.6106245517730713, -0.2032153457403183, -0.11518280953168869, -0.1861463189125061, -0.18630245327949524, -0.054775431752204895, 0.39907020330429077, 0.3937791585922241, 0.13219667971134186, 0.3591201603412628, 0.20930393040180206, -0.11759829521179199, 0.18299829959869385, 0.1719692349433899, -0.04389527067542076, -0.42944610118865967, 0.03136976808309555, 0.3862445652484894, 0.29799884557724, 0.10854163765907288, -0.23654919862747192, -0.24920272827148438, 0.015471708960831165, -0.4298044443130493, 0.26033616065979004, -0.3362565040588379, 0.2794796824455261, 0.09976346790790558, 0.24782228469848633, -0.5951438546180725, -0.10310759395360947, 0.16499000787734985, -0.11449187994003296, -0.12290894240140915, -0.09292814135551453, -0.03688323125243187, -0.4055462181568146, -0.26714786887168884, -0.20519253611564636, -0.4359728693962097, -0.2693944275379181, 0.3358285427093506, 0.01575186476111412, 0.21157878637313843, 0.3739762306213379, 0.09294816851615906, 0.31005221605300903, 0.04953598231077194, -0.2053631693124771, 0.12083633244037628, 0.03579314425587654, 0.0885268971323967, -0.2505459785461426, -0.4985419809818268, 0.09977518022060394, 0.05173453688621521, 0.037645094096660614, -0.24499185383319855, -0.41610968112945557, -0.5225145816802979, 0.12612798810005188, 0.1882447749376297, -0.0012071207165718079, -0.07886308431625366, 0.20694276690483093, -0.19732968509197235, 0.1095580905675888, -0.01307712122797966, -0.24611112475395203, -0.12782594561576843, 0.0830916091799736, 0.3247511386871338, -0.15073630213737488, 0.10587995499372482, 0.14839689433574677, 0.19341638684272766, 0.31456393003463745, -0.07250475883483887, 0.11436943709850311, 0.12013456970453262, 0.27393531799316406, -0.21050935983657837, -0.3405758738517761, 0.3424801826477051, 0.010134965181350708, -0.009619051590561867, 0.3454561233520508, -0.05972875654697418, -0.5129147171974182, -0.18136189877986908, 0.10224531590938568, -0.25864046812057495, -0.35660701990127563, 0.10800474882125854, 0.43654659390449524, -0.0429634153842926, 0.2192896157503128, 0.176939457654953, -0.0064318375661969185, -0.26712295413017273, -0.39389076828956604, 0.2829705476760864, -0.23305228352546692, 0.10045421123504639, 0.18662957847118378, -0.45402929186820984, -0.6995073556900024, 0.1291135847568512, 0.020111992955207825, 0.29739686846733093, -0.008754976093769073, 0.2319486141204834, 0.1322501301765442, -0.028030648827552795, 0.24217189848423004, 0.08653083443641663, -0.1293119341135025, 0.05553670600056648, 0.02404181659221649, -0.25207918882369995, 0.14389944076538086, -0.3590138852596283, 0.5358404517173767, 0.8802664875984192, 0.37709736824035645, -0.16434848308563232, 0.05778127163648605, 0.5292019844055176, -0.032791536301374435, 0.1895885169506073, 0.10803986340761185, -0.30443376302719116, -0.3213100731372833, -0.14403459429740906, 0.05696239322423935, -0.048293691128492355, -0.0072516086511313915, 0.09713459759950638, -0.10260340571403503, 0.2945187985897064, -0.13049399852752686, -0.1803799420595169, 0.49008429050445557, -0.04314805567264557, 0.06898368895053864, 0.034631866961717606, -0.20050598680973053, -0.14851167798042297, 0.2358812689781189, 0.2782011926174164, -0.21470855176448822, -0.2510850429534912, 0.19106850028038025, -0.2993552088737488, 0.12639901041984558, 0.5574198961257935, -0.012704958207905293, -0.08767616003751755, -0.1413710117340088, -0.010357685387134552, 0.37668415904045105, 0.008204463869333267, 0.34758469462394714, 0.13797669112682343, -0.08299775421619415, 0.15213747322559357, 0.2743895649909973, -0.010234429500997066, 0.060648828744888306, -0.12669657170772552, -0.13302627205848694, -0.26653560996055603, 0.43179649114608765, 0.0313241109251976, 0.8313065767288208, -0.19805578887462616, 0.18259061872959137, 0.21659579873085022, 0.03458178788423538, 0.700577974319458, 0.09818445146083832, 0.29737430810928345, -0.1605040431022644, -0.26129215955734253, -0.02691712975502014, 0.03417275473475456, 0.2771380543708801, 0.003568492829799652, -0.6976840496063232, 0.3543376326560974, 0.5069888234138489, -0.06665970385074615, 0.1408870369195938, 0.20467042922973633, -0.20412833988666534, -0.3309398293495178, -0.6929692029953003, 0.1499154269695282, -0.24720902740955353, 0.21208223700523376, -0.23182713985443115, -0.01660650037229061, -0.22074241936206818, -0.055851973593235016, -0.39808905124664307, 0.09784277528524399, -0.49289393424987793, -0.14910347759723663, 0.30979645252227783, -0.2716407775878906, -0.015284748747944832, 0.318087100982666, 0.2980523109436035, 0.21503791213035583, -0.04117577522993088, 0.05580143630504608, -0.009489171206951141, -0.4288063049316406, 0.069615438580513, 0.12072674930095673, 0.07078708708286285, -0.2315792739391327, -0.011699400842189789, -0.15616866946220398, 0.19199419021606445, -0.14730089902877808, 0.3778708577156067, 0.26534610986709595, 0.11379937827587128, 0.15317006409168243, -0.11881119012832642, -0.04288323223590851, -0.18436691164970398, -0.2605475187301636, 0.1095840334892273, -0.10283772647380829, -0.1080937311053276, 0.17686614394187927, 0.17040139436721802, -0.16747748851776123, -0.018127765506505966, 0.24444758892059326, -0.10914572328329086, 0.05315779149532318, 0.21790295839309692, 0.2029287964105606, -0.04807587340474129, -0.21867182850837708, 0.1641601324081421, -0.4710911810398102, -0.2075665295124054, 0.019297659397125244, 0.004306101240217686, -0.17338921129703522, -0.056421104818582535, 0.6419909596443176, 0.29731160402297974, 0.03289363533258438, -0.1205991804599762, -0.36713284254074097, -0.22014771401882172, 0.20411597192287445, 0.2204362004995346, -0.1217895895242691, 0.3346557915210724, 0.3340521454811096, 0.28580614924430847, 0.0965176597237587, -0.2278682440519333, 0.2888352870941162, -0.37428563833236694, 0.1768331676721573, 0.03506293147802353, -0.2467694878578186, 0.1611359417438507, -0.13581226766109467, -0.05651426315307617, 0.547860860824585, 0.06339478492736816, -0.2198670208454132, -0.13317617774009705, 0.06871268153190613, 0.11896972358226776, -0.10746381431818008, -0.31719645857810974, -0.06813296675682068, -0.09852121770381927, -0.04035230726003647, 0.07765203714370728, -0.0682106763124466, 0.10933534801006317, -0.06905272603034973, 0.26677948236465454, 0.2946944832801819, 0.2482583075761795, 0.07181014120578766, -0.05436810851097107, 0.21100717782974243, 0.12927581369876862, 0.2501763701438904, 0.2140113115310669, 0.03418012708425522, 0.14100198447704315, -0.00031037628650665283, -0.07852867245674133, 0.15759477019309998, 0.266745001077652, -0.4023352265357971, 0.10853855311870575, -0.06370002031326294, -0.04671737924218178, 0.38915032148361206, 0.05957898497581482, -0.1627657562494278, 0.25530385971069336, 0.07318656891584396, -0.14528286457061768, 0.0629551038146019, 0.5381839275360107, 0.16475962102413177, 0.07420110702514648, 0.48672521114349365, 0.11547712981700897, -0.048144444823265076, 0.1442146748304367, 0.10595778375864029, 0.23524419963359833, 0.2533341348171234, 0.011607512831687927, -0.08355844020843506, 0.013890925794839859, 0.11324334144592285, 0.08847473561763763, 0.05775381997227669, 0.1947414129972458, 0.008706476539373398, -0.00255652517080307, 0.18441610038280487, -0.13208413124084473, 0.09944799542427063, 0.2976967692375183, -0.03657403588294983, -0.21635495126247406, 0.04410649836063385, 0.2792079448699951, 0.062182217836380005, 0.13555821776390076, 0.8394277095794678, -0.535256028175354, -0.25631389021873474, -0.11296945065259933, 0.13962063193321228, -0.36160191893577576, -0.006912203039973974, -0.3097972273826599, 0.15970873832702637, -0.13192445039749146, -0.018767990171909332, -0.15464213490486145, -0.04683637619018555, 0.30263882875442505, 0.2967252731323242, -0.1764540821313858, -0.11941330134868622, -0.21040698885917664, 0.09632765501737595, -0.1580304205417633, -0.04321964830160141, 0.5650871992111206, -0.4615534543991089, 0.15587055683135986, -0.11572045087814331, 0.5128489136695862, 0.1959538757801056, 0.23445622622966766, -0.17214342951774597, -0.16993466019630432, -0.040043652057647705, 0.021636465564370155, -0.09688606858253479, -0.01911655068397522, 0.4978407621383667, 0.2487410008907318, 0.1886564940214157, 0.030080711469054222, -0.11569540202617645, 0.19563598930835724, 0.15852324664592743, -0.2628212869167328, 0.05865900591015816, -0.10240822285413742, -0.1074177622795105, -0.14664116501808167, -0.2744993567466736, -0.043056122958660126, -0.3473239839076996, 0.5429143905639648, -0.2994093596935272, -0.1069813072681427, 0.0449775867164135, -0.09464637190103531, 0.0484454482793808, -0.004604628309607506, 0.5350642204284668, 0.403215616941452, 0.02547585219144821, -0.21520183980464935, -0.4277659058570862, -0.38895684480667114, 0.2161230444908142, -0.2563636600971222, 0.011096818372607231, 0.059635311365127563, 0.18237672746181488, -0.10124049335718155, 0.41397738456726074, -0.07893024384975433, 0.04184630885720253, -0.12649881839752197, -0.06532084196805954, -0.47243618965148926, -0.1160053163766861, 0.28529053926467896, -0.2760682702064514, -0.06966478377580643, -0.22366377711296082, 0.15214930474758148, -0.2058599591255188, -0.07175612449645996, -0.007133167237043381, -0.3104133605957031, -0.2900140881538391, -0.10448214411735535, 0.07206257432699203, 0.17017734050750732, 0.10107715427875519, 0.019608356058597565, -0.32109400629997253, -0.04657530412077904, -0.2240411788225174, -0.271959125995636, 0.07557492703199387, 0.2256832867860794, 0.2975493371486664, -0.05231298506259918, 0.3451107442378998, -0.16596199572086334, 0.5020350813865662, 0.10082212090492249, 0.03670166805386543, -0.012219477444887161, 0.0015544518828392029, -0.12299051135778427, 0.08524467796087265, -0.14766019582748413, 0.45889905095100403, -0.2785676419734955, 0.17531099915504456, -0.4899505376815796, -0.3166499733924866, 0.3832274377346039, -0.4803461730480194, -0.09981128573417664, 0.0017299461178481579, 0.20958417654037476, 0.13941442966461182, -0.30026501417160034, -0.39764609932899475, -0.02579391747713089, -0.15764199197292328, -0.17507284879684448, -0.4597727954387665, 0.13977783918380737, -0.20215071737766266, 0.05126635357737541, -0.0826166495680809, 0.4182423949241638, 0.07982709258794785, -0.25584203004837036, 0.015247397124767303, -0.15567579865455627 ]
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
[ -0.16229428350925446, 0.28865036368370056, -0.016385259106755257, 0.08907576650381088, -0.09513925015926361, -0.13527382910251617, 0.24156032502651215, 0.04239333048462868, -0.4444727599620819, 0.05577059090137482, -0.3425390422344208, 0.5006805658340454, 0.19029931724071503, 0.05172187089920044, 0.23175173997879028, 0.12949393689632416, 0.05272127315402031, 0.45624175667762756, -0.005170382559299469, 0.15157285332679749, -0.1371893286705017, 0.4322417974472046, -0.3422397971153259, 0.19298404455184937, -0.2837298512458801, -0.20665043592453003, -0.3387557864189148, 0.33662930130958557, -0.3838270604610443, -0.36793971061706543, 0.30076923966407776, 0.11652305722236633, -0.010650593787431717, 0.24211174249649048, -0.00011587276821956038, -0.26339447498321533, 0.07170158624649048, 0.05097352713346481, -0.32355111837387085, -0.18543969094753265, 0.1579289734363556, -0.2767643332481384, 0.3587702810764313, -0.36968278884887695, 0.23197349905967712, 0.2630351483821869, 0.19682085514068604, 0.41317909955978394, 0.378421425819397, 0.25508973002433777, 0.15432633459568024, 0.36232462525367737, 0.15149959921836853, -0.049903079867362976, 0.3280055522918701, -0.240153506398201, -0.24223308265209198, 0.5541877746582031, 0.5191040635108948, -0.17601542174816132, -0.0643187090754509, -0.023654695600271225, -0.15008221566677094, 0.10818663239479065, 0.09065616130828857, 0.2510840594768524, -0.30428019165992737, -0.35956698656082153, 0.14779290556907654, 0.07521237432956696, 0.01005680114030838, -0.25509873032569885, 0.08777572959661484, -0.08877255767583847, 0.2548277676105499, -0.4677804410457611, 0.22016599774360657, 0.29525285959243774, -0.3333255648612976, 0.047399066388607025, 0.06930747628211975, -0.09359443932771683, -0.15576386451721191, 0.1831720769405365, 0.11445970833301544, 0.3580797016620636, -0.0566365122795105, 0.01170644536614418, 0.25798267126083374, 0.11046642810106277, 0.2554870843887329, 0.26135581731796265, -0.12736041843891144, -0.12093034386634827, -0.048167597502470016, -0.10802482068538666, -0.11227396875619888, 0.353577584028244, 0.17000266909599304, -0.2848018407821655, 0.18919840455055237, -0.08284571766853333, 0.180411696434021, 0.1894938349723816, -0.025866255164146423, 0.011831950396299362, 0.17808930575847626, -0.227344810962677, 0.4025893807411194, 0.26510074734687805, -0.1325012743473053, -0.019606098532676697, 0.06968577951192856, -0.08283194899559021, -0.18123900890350342, -0.2002732753753662, 0.11146120727062225, -0.17330406606197357, -0.21198901534080505, -0.06052526831626892, -0.21012082695960999, 0.10313369333744049, 0.030671827495098114, 0.22637057304382324, -0.04862555116415024, -0.3248879611492157, -0.008808787912130356, 0.009723534807562828, -0.3067668378353119, -0.1658761203289032, -0.02070273831486702, -0.015813101083040237, -0.3158724009990692, 0.2927926778793335, 0.25611111521720886, 0.01416999101638794, 0.4369167387485504, -0.0667463019490242, 0.11256466805934906, -0.06157010793685913, -0.03654201328754425, -0.25050827860832214, -0.1376703977584839, 0.19454526901245117, 0.28504160046577454, -0.13704144954681396, -0.274625688791275, -0.5182225108146667, -0.029475823044776917, 0.13417834043502808, -0.09137134999036789, -0.10038185119628906, -0.14188069105148315, 0.11342830955982208, -0.28249549865722656, -0.0845005214214325, -0.24448567628860474, -0.011239003390073776, -0.27759864926338196, -0.32704615592956543, 0.06488142162561417, -0.11334045231342316, -0.11958074569702148, -0.17336557805538177, -0.0513911210000515, 0.22777405381202698, -0.1550368368625641, -0.060792289674282074, 0.12664705514907837, 0.08688856661319733, 0.05646301433444023, 0.36917534470558167, -0.22387832403182983, 0.05881846696138382, -0.030297335237264633, 0.27381205558776855, 0.4754830300807953, -0.13133665919303894, -0.5511143803596497, 0.02224794775247574, -0.09273926168680191, -0.2200123369693756, -0.17768478393554688, 0.1853412687778473, 0.43592777848243713, -0.1713544875383377, 0.12053315341472626, 0.47766441106796265, -0.12799139320850372, 0.08927568793296814, -0.25258973240852356, -0.2019869089126587, 0.021061614155769348, 0.0683601051568985, 0.10308273136615753, -0.04792409390211105, 0.010847516357898712, 0.6572158336639404, 0.15238052606582642, -0.1279432773590088, -0.2924891412258148, 0.012023558840155602, 0.2637319266796112, -0.15930792689323425, -0.08614962548017502, -0.10207740217447281, -0.49761849641799927, -0.01891493611037731, -0.38266482949256897, 0.22172364592552185, 0.11234728246927261, -0.126118466258049, -0.019689809530973434, -0.02919624000787735, 0.013085011392831802, -0.10262607038021088, 0.11551791429519653, 0.09121492505073547, 0.03127623721957207, 0.03552217409014702, 0.011038664728403091, 0.3079177439212799, -0.2108154445886612, 0.40704068541526794, -0.16428785026073456, 0.3394855260848999, -0.37261924147605896, -0.45434367656707764, 0.03883136063814163, 0.3175657093524933, -0.005407072603702545, -0.18586820363998413, -0.0847700759768486, 0.30383995175361633, -0.03310529887676239, -0.0015199631452560425, -0.0419730544090271, -0.21679885685443878, 0.41755738854408264, -0.03388039767742157, 0.040236301720142365, -0.15333695709705353, 0.011028864420950413, -0.11775416135787964, 0.12128815799951553, 0.15278573334217072, -0.27263888716697693, -0.0467425212264061, 0.11932708323001862, 0.08081527799367905, 0.0008184816688299179, -0.12340371310710907, -0.12160488963127136, 0.20139914751052856, 0.25048884749412537, 0.23211517930030823, 0.05561453476548195, 0.003987550735473633, -0.3448365330696106, 0.16886049509048462, 0.6388943791389465, 0.009611144661903381, 0.1933998167514801, 0.00266338512301445, -0.05535616725683212, -0.09187646955251694, 0.36545637249946594, -0.1382659524679184, 0.4473469853401184, 0.08535989373922348, -0.11062639951705933, 0.046115897595882416, -0.20016111433506012, -0.3595981001853943, 0.08434490114450455, 0.1504209339618683, 0.3393532633781433, -0.16112473607063293, 0.32813283801078796, 0.18927931785583496, -0.4165051579475403, -0.1966778188943863, -0.11323972791433334, 0.16031938791275024, 0.046376269310712814, 0.31053587794303894, -0.22582700848579407, -0.636313796043396, -0.2138478010892868, -0.2246876358985901, -0.1736542284488678, -0.21213839948177338, -0.06543605029582977, 0.41710567474365234, 0.37957391142845154, 0.10483014583587646, 0.390836626291275, 0.2919412851333618, -0.1407936066389084, 0.1514676809310913, 0.07684770226478577, -0.11653365194797516, -0.4451969861984253, 0.03559618815779686, 0.440340518951416, 0.30272260308265686, 0.1866854727268219, -0.18262280523777008, -0.14730453491210938, 0.041650332510471344, -0.47399434447288513, 0.25526338815689087, -0.28294235467910767, 0.250566691160202, 0.005636274814605713, 0.2146396040916443, -0.5915750861167908, -0.2966306507587433, 0.17837530374526978, -0.19174475967884064, -0.06382707506418228, -0.07141566276550293, -0.09075494855642319, -0.36869990825653076, -0.23214147984981537, -0.19074279069900513, -0.3892240822315216, -0.251996248960495, 0.30632466077804565, 0.04709013178944588, 0.1246849074959755, 0.3451833128929138, -0.05784675106406212, 0.4203663170337677, -0.006661331746727228, -0.04496560990810394, 0.0683775469660759, 0.032240767031908035, 0.06294314563274384, -0.2506498694419861, -0.3901831805706024, 0.08632995188236237, 0.08140083402395248, 0.07147687673568726, -0.32992255687713623, -0.4478316903114319, -0.4516489505767822, 0.10525127500295639, 0.2513579726219177, 0.009099237620830536, -0.09579459577798843, 0.20816729962825775, -0.1809079349040985, 0.05214749276638031, 0.12245678156614304, -0.35239288210868835, -0.2072189748287201, 0.15541231632232666, 0.3015429377555847, -0.15515083074569702, 0.14002060890197754, -0.03658069670200348, 0.2472066581249237, 0.12030628323554993, -0.16269686818122864, 0.2434939742088318, 0.0470380038022995, 0.2754899859428406, -0.2692047357559204, -0.3264658451080322, 0.3509717881679535, -0.13879045844078064, -0.07192058861255646, 0.35272228717803955, -0.17399722337722778, -0.543537974357605, -0.17621366679668427, 0.08835037052631378, -0.3520292341709137, -0.35120058059692383, 0.14087727665901184, 0.4890918731689453, -0.12838980555534363, 0.16154582798480988, 0.10040249675512314, 0.13662998378276825, -0.3128577470779419, -0.28019729256629944, 0.14688032865524292, -0.31283360719680786, 0.015016519464552402, 0.0840366780757904, -0.5422759652137756, -0.6152955293655396, 0.09041263908147812, 0.02460484765470028, 0.16473181545734406, 0.010636936873197556, 0.16953417658805847, 0.1697925478219986, -0.07759535312652588, 0.18881714344024658, 0.239374041557312, -0.21169348061084747, 0.10137900710105896, 0.10286170244216919, -0.24931302666664124, 0.004477288573980331, -0.4466741681098938, 0.4395610988140106, 0.8527019023895264, 0.21628931164741516, -0.16088272631168365, 0.1470588594675064, 0.49671170115470886, 0.08200763911008835, 0.1896102875471115, 0.13085351884365082, -0.27469202876091003, -0.3481546640396118, -0.20279721915721893, 0.02792271226644516, -0.20913085341453552, 0.03567733243107796, -0.04792370647192001, -0.05737582594156265, 0.33631300926208496, -0.05814089626073837, -0.11765992641448975, 0.5659452676773071, -0.05071742832660675, 0.1383465975522995, -0.0365656279027462, -0.16751006245613098, -0.05538645386695862, 0.20562489330768585, 0.31446439027786255, -0.14917083084583282, -0.11714661121368408, 0.25812581181526184, -0.17916470766067505, 0.1546858251094818, 0.5725120306015015, 0.03370543196797371, -0.01572665572166443, -0.25497832894325256, -0.027280151844024658, 0.3534432649612427, 0.03846055269241333, 0.3919437825679779, 0.02410222217440605, -0.09691464155912399, 0.22327588498592377, 0.36032718420028687, 0.006348045542836189, 0.005739316344261169, -0.05591969192028046, -0.13157901167869568, -0.16340459883213043, 0.5269417762756348, 0.09646141529083252, 0.9199798107147217, -0.13518962264060974, 0.2328004240989685, 0.39558547735214233, 0.005457688122987747, 0.8213492035865784, 0.036442771553993225, 0.29021501541137695, -0.2200111299753189, -0.10888098925352097, 0.038619667291641235, -0.027458976954221725, 0.14550791680812836, -0.15644913911819458, -0.6769137978553772, 0.24205626547336578, 0.45106345415115356, 0.08140227198600769, 0.05928388983011246, 0.22673889994621277, -0.040526434779167175, -0.2793492078781128, -0.6700121760368347, 0.16516238451004028, -0.23910804092884064, 0.25718143582344055, -0.3153548538684845, 0.006179021671414375, -0.12149988114833832, -0.01944376714527607, -0.41948968172073364, 0.1596042513847351, -0.3940163552761078, -0.10973115265369415, 0.25963613390922546, -0.24909450113773346, -0.003667149692773819, 0.19052889943122864, 0.323668897151947, 0.18673259019851685, -0.0359213650226593, -0.02719733491539955, -0.07232923060655594, -0.34440359473228455, 0.1388995498418808, 0.07803389430046082, 0.13850179314613342, -0.23670151829719543, -0.086870476603508, 0.028653942048549652, 0.24055737257003784, -0.17888350784778595, 0.2510043978691101, 0.26750999689102173, 0.09652285277843475, 0.03877583518624306, 0.021772412583231926, 0.049311600625514984, -0.12440276145935059, -0.25639116764068604, 0.08975756168365479, -0.10483016073703766, -0.04636775329709053, 0.25974172353744507, 0.16537608206272125, -0.15297949314117432, -0.09329453110694885, 0.16011497378349304, -0.12814617156982422, 0.09165763854980469, 0.18983621895313263, 0.2613019049167633, -0.07720856368541718, -0.17520993947982788, 0.045474305748939514, -0.5221548676490784, -0.3000226616859436, -0.07522928714752197, 0.02601168490946293, -0.05449220538139343, 0.07419305294752121, 0.6213568449020386, 0.3452058434486389, 0.1389111578464508, -0.09827123582363129, -0.30082517862319946, -0.1382514238357544, 0.22200584411621094, 0.31040114164352417, -0.10538050532341003, 0.43308311700820923, 0.4210222363471985, 0.2333647608757019, 0.170184925198555, -0.23408553004264832, 0.3082627058029175, -0.34560006856918335, 0.1343725323677063, -0.04723048955202103, -0.2782139182090759, 0.1894894689321518, -0.06866849958896637, -0.05093331262469292, 0.5490370988845825, 0.13024351000785828, -0.19484561681747437, -0.0554991252720356, 0.06485946476459503, 0.08955325931310654, -0.10635159909725189, -0.355571985244751, -0.08047802746295929, 0.07174202799797058, 0.00746369082480669, 0.11996142566204071, -0.13209018111228943, 0.15138033032417297, -0.06826154887676239, 0.37282606959342957, 0.24534690380096436, 0.13897669315338135, 0.11990594863891602, 0.018168345093727112, 0.23148804903030396, 0.16493651270866394, 0.26410728693008423, 0.11129942536354065, -0.008376307785511017, 0.10891096293926239, 0.01216726005077362, -0.04408532381057739, 0.0655946359038353, 0.1964225471019745, -0.38780468702316284, 0.07796412706375122, -0.23517318069934845, -0.037391506135463715, 0.48941290378570557, 0.13260655105113983, -0.18248435854911804, 0.2481166124343872, 0.06972344219684601, -0.11593437194824219, 0.022276239469647408, 0.4485892355442047, 0.18295283615589142, -0.019731197506189346, 0.4950128495693207, 0.1573648303747177, -0.003334715962409973, 0.12629817426204681, 0.04812316596508026, 0.2687561810016632, 0.36889469623565674, 0.0004860414192080498, -0.06775552034378052, -0.06217379868030548, 0.03523818403482437, 0.17615734040737152, 0.060218654572963715, 0.15215088427066803, -0.06740161776542664, -0.12788814306259155, 0.15209059417247772, -0.289011150598526, 0.06382182240486145, 0.3066234290599823, -0.05420027673244476, -0.2557840645313263, 0.11049554497003555, 0.3738560676574707, -0.02909044921398163, 0.08440802246332169, 0.8333595395088196, -0.35441017150878906, -0.3209327161312103, -0.1020396426320076, 0.2567328214645386, -0.345008909702301, 0.09901490807533264, -0.27242404222488403, 0.10394691675901413, -0.08362768590450287, 0.08783477544784546, -0.17857876420021057, -0.13201038539409637, 0.3258436918258667, 0.2735811173915863, -0.2538757920265198, -0.15999843180179596, -0.16739705204963684, 0.05449536442756653, -0.1816900074481964, 0.07494211196899414, 0.5572719573974609, -0.33135414123535156, 0.09935422241687775, -0.23304104804992676, 0.528035581111908, 0.289602667093277, 0.2048736810684204, -0.19099849462509155, -0.1523382067680359, -0.040742866694927216, 0.04505090415477753, -0.014735380187630653, -0.06881915777921677, 0.4951958954334259, 0.10088448226451874, 0.18965791165828705, 0.049777351319789886, -0.13975872099399567, 0.28492471575737, 0.13173896074295044, -0.21514549851417542, 0.037306033074855804, -0.10045898705720901, 0.0037534888833761215, -0.17565684020519257, -0.3053843677043915, -0.050677500665187836, -0.31464946269989014, 0.5148182511329651, -0.14559148252010345, -0.11733540147542953, -0.04240352660417557, -0.13935600221157074, 0.05286145955324173, 0.09525205194950104, 0.6466183066368103, 0.5057283639907837, 0.15907415747642517, -0.19428105652332306, -0.4506126046180725, -0.41047388315200806, 0.2689817547798157, -0.30991464853286743, 0.11366626620292664, 0.05045691505074501, 0.1924663633108139, -0.16570335626602173, 0.4790000915527344, -0.026662632822990417, 0.11698726564645767, -0.24810901284217834, -0.15506228804588318, -0.5336962342262268, -0.015051908791065216, 0.11213739216327667, -0.3295289874076843, -0.1642407327890396, -0.043305810540914536, 0.21823453903198242, -0.3421745002269745, -0.07783357799053192, -0.06188158690929413, -0.30674275755882263, -0.2668483257293701, -0.07545420527458191, 0.25476449728012085, 0.10352201759815216, 0.13317328691482544, 0.07666158676147461, -0.3111726939678192, 0.02930312603712082, -0.38363757729530334, -0.1451123058795929, 0.04734295606613159, 0.17323514819145203, 0.19779308140277863, 0.0003887750208377838, 0.36838898062705994, -0.19359779357910156, 0.504595160484314, 0.09660807251930237, 0.06460712105035782, -0.002981509082019329, -0.09048981219530106, -0.07207770645618439, 0.08612068742513657, -0.0878099799156189, 0.42042770981788635, -0.23913711309432983, 0.2577548027038574, -0.547700047492981, -0.26309603452682495, 0.4129125475883484, -0.4527483284473419, -0.04186302050948143, -0.00435636006295681, 0.19462785124778748, 0.11911725997924805, -0.19335907697677612, -0.4065576493740082, -0.09077239036560059, -0.10226772725582123, -0.12823066115379333, -0.44656145572662354, 0.18074040114879608, -0.12650536000728607, -0.009618472307920456, -0.11334371566772461, 0.2060469388961792, 0.044330112636089325, -0.275480717420578, -0.0735137015581131, -0.18137381970882416 ]
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 :)
[ -0.20812611281871796, 0.08529740571975708, -0.08818325400352478, 0.16447675228118896, -0.023115716874599457, -0.10333463549613953, 0.3287059962749481, 0.08260271698236465, -0.3126956820487976, 0.21142813563346863, -0.39940762519836426, 0.33623337745666504, 0.2900557816028595, 0.16409508883953094, 0.15483728051185608, 0.164514422416687, 0.09830836951732635, 0.37275904417037964, -0.06697463989257812, 0.14713607728481293, -0.22851459681987762, 0.4517594575881958, -0.3685060143470764, 0.06003374978899956, -0.27340254187583923, -0.20931918919086456, -0.23765408992767334, 0.273295134305954, -0.4236902594566345, -0.2180948704481125, 0.2879370152950287, 0.18390336632728577, -0.11174572259187698, 0.2599034309387207, -0.00011433531471993774, -0.24122940003871918, -0.020322293043136597, -0.06533646583557129, -0.3073402941226959, -0.16067226231098175, -0.08061999082565308, -0.23325759172439575, 0.2200169563293457, -0.4098558723926544, 0.1319078803062439, 0.41221514344215393, 0.2227526307106018, 0.286211758852005, 0.2939855456352234, 0.29529327154159546, 0.18044836819171906, 0.4040350914001465, 0.12714536488056183, -0.17765109241008759, 0.32549697160720825, -0.2722628116607666, -0.24315336346626282, 0.410020112991333, 0.627405047416687, -0.08831330388784409, -0.06562700122594833, 0.13427631556987762, -0.12808924913406372, 0.12494772672653198, 0.08908423036336899, 0.24771210551261902, -0.2894337773323059, -0.4838736951351166, 0.34367427229881287, 0.11621173471212387, 0.15104791522026062, -0.12315120548009872, 0.058925919234752655, 0.05899590253829956, 0.1713838130235672, -0.29327523708343506, 0.26158592104911804, 0.3350571095943451, -0.22270549833774567, 0.13434647023677826, 0.006289578974246979, 0.08987294882535934, -0.1818217933177948, 0.13155537843704224, -0.027878839522600174, 0.16018429398536682, -0.14802968502044678, -0.09576212614774704, 0.22563056647777557, 0.011602772399783134, 0.2465774267911911, 0.09906674921512604, -0.16215816140174866, -0.09483564645051956, -0.1089966893196106, -0.24236761033535004, -0.2443491518497467, 0.24999386072158813, 0.23728668689727783, -0.24156509339809418, 0.3187462091445923, -0.029161054641008377, 0.2450110912322998, 0.1847766935825348, -0.031999439001083374, -0.14745502173900604, 0.25958141684532166, -0.1022157296538353, 0.4240211844444275, 0.41627928614616394, -0.1882748007774353, -0.09184844046831131, -0.02441922202706337, -0.17939020693302155, -0.22493113577365875, -0.09151172637939453, 0.04073415696620941, -0.24053452908992767, -0.2889053225517273, 0.07782167941331863, -0.1368495523929596, 0.2789806127548218, 0.06914238631725311, 0.2426397204399109, -0.001409911666996777, -0.4101417362689972, 0.03285718709230423, -0.010167215019464493, -0.2695293724536896, -0.22026358544826508, -0.030173445120453835, -0.09110906720161438, -0.1644534468650818, 0.26104265451431274, 0.24966920912265778, 0.136254221200943, 0.5002055764198303, -0.09450667351484299, 0.23768068850040436, -0.2036779224872589, 0.03617642819881439, -0.42523881793022156, -0.09555710107088089, 0.3575984239578247, 0.3498214781284332, -0.203954815864563, -0.22149424254894257, -0.5597000122070312, 0.10842228680849075, 0.07068213075399399, -0.03556116670370102, -0.022771235555410385, -0.037245579063892365, 0.1511412113904953, -0.20522908866405487, 0.02430899254977703, -0.31279516220092773, 0.04790033400058746, -0.28003260493278503, -0.1421736478805542, 0.01360168308019638, -0.03400634229183197, -0.2153872549533844, -0.03196972608566284, -0.016684822738170624, 0.2605353593826294, -0.10019350796937943, 0.00802689790725708, 0.20574261248111725, -0.04610013589262962, -0.040377043187618256, 0.31282347440719604, -0.15650902688503265, -0.009523682296276093, -0.010019519366323948, 0.2053345888853073, 0.3080560863018036, -0.2581486999988556, -0.7119432687759399, 0.05824534222483635, 0.005247361958026886, -0.26256269216537476, -0.13132771849632263, 0.23769718408584595, 0.5601800680160522, -0.1987905353307724, 0.11458456516265869, 0.4635341167449951, -0.18903322517871857, 0.09863816946744919, -0.15008224546909332, -0.1512318104505539, -0.1275019496679306, 0.12364053726196289, 0.20790448784828186, -0.006194032728672028, 0.07666581869125366, 0.5572872161865234, 0.16624769568443298, -0.1265098601579666, -0.3530396819114685, 0.07210428267717361, 0.27843502163887024, -0.3214608430862427, -0.09803193807601929, -0.10858356207609177, -0.5646517872810364, 0.06257221847772598, -0.28644707798957825, 0.08351267129182816, 0.13649031519889832, -0.23094579577445984, -0.0580952987074852, -0.0626564472913742, -0.07724547386169434, -0.03654152527451515, 0.1337592601776123, -0.09502677619457245, -0.033458929508924484, 0.020008746534585953, 0.11398935317993164, 0.3453357517719269, -0.2902800738811493, 0.5099153518676758, 0.008631348609924316, 0.32485464215278625, -0.27559128403663635, -0.30001163482666016, 0.1337188333272934, 0.09635904431343079, 0.08106764405965805, -0.1677875816822052, -0.14541804790496826, 0.29110416769981384, 0.19663825631141663, 0.031059160828590393, 0.14562588930130005, -0.27314692735671997, 0.4598880112171173, 0.058973077684640884, -0.035778917372226715, -0.13375794887542725, 0.08443674445152283, -0.19134965538978577, -0.04629848897457123, 0.2320573627948761, -0.24288012087345123, -0.04818817228078842, 0.11958301067352295, 0.1858396679162979, 0.04129054397344589, -0.13352656364440918, -0.15656763315200806, 0.09857840836048126, 0.23880954086780548, 0.20704035460948944, -0.0013627633452415466, -0.04337596893310547, -0.34334397315979004, 0.19563522934913635, 0.46204400062561035, -0.005481511354446411, 0.20487985014915466, -0.06295571476221085, -0.06925413757562637, 0.0003483295440673828, 0.42172321677207947, 0.04121627286076546, 0.4216998815536499, 0.08651918917894363, -0.090688556432724, 0.0007988698780536652, -0.03490293025970459, -0.3570866584777832, 0.16904473304748535, 0.14286863803863525, 0.2106134444475174, -0.1635541021823883, 0.22982488572597504, -0.036555614322423935, -0.3995290696620941, -0.10759542882442474, -0.08823565393686295, 0.13155025243759155, 0.13491052389144897, 0.3200078308582306, -0.11699086427688599, -0.5118411779403687, -0.07171622663736343, -0.1881691813468933, -0.09109161794185638, -0.2565024793148041, -0.047091953456401825, 0.506132960319519, 0.3253307044506073, 0.013789152726531029, 0.3485431671142578, 0.34163233637809753, -0.002122672274708748, 0.3129968047142029, 0.11849890649318695, -0.04346625134348869, -0.28942447900772095, 0.07920357584953308, 0.42273062467575073, 0.24148361384868622, 0.21288280189037323, -0.30292388796806335, -0.15141017735004425, -0.14659321308135986, -0.28685131669044495, 0.29411792755126953, -0.19982564449310303, 0.23458154499530792, -0.02369757741689682, 0.1992698311805725, -0.5368194580078125, -0.12628576159477234, 0.1590675413608551, -0.010400773026049137, -0.08413989841938019, -0.05235573649406433, -0.01666244864463806, -0.4458804428577423, -0.20343223214149475, -0.1740431785583496, -0.3917514979839325, -0.24149511754512787, 0.18787041306495667, -0.11721916496753693, 0.1827543079853058, 0.2395993322134018, 0.065528005361557, 0.40171733498573303, -0.03590880334377289, -0.24859121441841125, 0.018446635454893112, -0.1290372908115387, 0.2076355516910553, -0.32397735118865967, -0.5115103721618652, 0.10641662776470184, 0.18794181942939758, 0.0032007023692131042, -0.26909077167510986, -0.43559908866882324, -0.3778632879257202, 0.049906257539987564, 0.119134321808815, 0.023840852081775665, -0.2214813530445099, 0.005934268236160278, -0.26215100288391113, 0.04465475678443909, -0.010199185460805893, -0.2739880084991455, -0.11105111986398697, 0.07669292390346527, 0.30780330300331116, -0.19712606072425842, 0.11028680205345154, 0.03152870759367943, 0.25294387340545654, 0.3045100271701813, -0.11898021399974823, 0.18053938448429108, 0.1001780778169632, 0.217914417386055, -0.2815532982349396, -0.44954508543014526, 0.3379516899585724, -0.1619504988193512, -0.06860882043838501, 0.35916173458099365, -0.09100829809904099, -0.5044240951538086, -0.17165996134281158, 0.072667196393013, -0.26782113313674927, -0.42276081442832947, 0.07086481899023056, 0.44430238008499146, -0.14769354462623596, 0.11258593201637268, 0.21931472420692444, 0.019851818680763245, -0.2291976511478424, -0.3125329911708832, 0.33609849214553833, -0.28678032755851746, 0.001498315017670393, 0.15573740005493164, -0.41713234782218933, -0.6907585859298706, 0.16188699007034302, -0.0503639280796051, 0.1151643842458725, -0.11067374050617218, 0.29069948196411133, 0.10505314916372299, 0.03925466164946556, 0.3141642212867737, 0.11675354838371277, -0.15859441459178925, 0.04357834532856941, -0.07495133578777313, -0.19444707036018372, 0.03445034846663475, -0.46410566568374634, 0.5842931270599365, 0.791657030582428, 0.39100512862205505, -0.04272051155567169, 0.14425702393054962, 0.5085097551345825, -0.03295799344778061, 0.15193405747413635, 0.08288567513227463, -0.14068740606307983, -0.3031001091003418, -0.1240249052643776, 0.05947873368859291, -0.027399515733122826, -0.060650307685136795, 0.033765941858291626, -0.20056183636188507, 0.30888551473617554, -0.07161420583724976, -0.0267878919839859, 0.5213572382926941, -0.019524237141013145, 0.08624415844678879, 0.10341477394104004, -0.10577695071697235, -0.11154292523860931, 0.4208713471889496, 0.30464044213294983, -0.20926691591739655, -0.2674808204174042, 0.2845478951931, -0.49008119106292725, 0.09635571390390396, 0.5848199725151062, -0.010691358707845211, -0.031013168394565582, -0.1739792674779892, 0.0050660669803619385, 0.3778679072856903, 0.05310952290892601, 0.40148672461509705, 0.17706425487995148, -0.02953927032649517, 0.17760561406612396, 0.257417768239975, -0.09426917135715485, 0.01887008547782898, 0.014035731554031372, -0.1412169188261032, -0.23207984864711761, 0.44888657331466675, 0.07712198048830032, 0.8232882618904114, -0.1511876881122589, 0.06643524020910263, 0.17773103713989258, 0.005393411964178085, 0.5558375716209412, -0.015879733487963676, 0.22724460065364838, -0.1396792232990265, -0.4450140595436096, -0.022555772215127945, -0.10461308807134628, 0.28905585408210754, -0.026632964611053467, -0.7058284878730774, 0.2149587869644165, 0.2657940983772278, 0.15262751281261444, 0.02041568234562874, 0.05576954782009125, -0.14507755637168884, -0.24285975098609924, -0.5719939470291138, 0.22100290656089783, -0.17519983649253845, 0.2350521981716156, -0.28548693656921387, 0.03061573952436447, -0.25519639253616333, 0.043435677886009216, -0.45206600427627563, 0.06690526008605957, -0.4822937846183777, 0.013023794628679752, 0.14497774839401245, -0.25090157985687256, 0.04381660744547844, 0.3447955846786499, 0.38753020763397217, 0.17426258325576782, -0.04343307018280029, -0.01864882931113243, -0.03557267785072327, -0.30341026186943054, -0.0630999282002449, -0.01694113202393055, 0.005117092281579971, -0.2645796239376068, -0.06806783378124237, -0.07059996575117111, 0.30341193079948425, -0.14938153326511383, 0.3579692840576172, 0.1762150526046753, 0.12822237610816956, -0.06678523123264313, -0.05379936099052429, -0.11600913107395172, -0.1748107224702835, -0.29133135080337524, 0.1321408748626709, -0.1205013245344162, -0.005070164799690247, 0.29780736565589905, 0.18357326090335846, -0.12276741117238998, -0.035556141287088394, 0.2913229167461395, -0.2046390175819397, 0.10926903784275055, 0.21874366700649261, 0.25724437832832336, -0.12845025956630707, -0.2801276445388794, 0.30033043026924133, -0.4932372570037842, -0.17429225146770477, 0.11607924103736877, -0.17401978373527527, -0.12756580114364624, -0.13076527416706085, 0.6403630971908569, 0.41495847702026367, -0.010212138295173645, -0.06361081451177597, -0.36319300532341003, -0.20702557265758514, 0.20271044969558716, 0.25006306171417236, 0.02594696916639805, 0.22087985277175903, 0.19487769901752472, 0.21906661987304688, 0.06693298369646072, -0.270333468914032, 0.373213529586792, -0.4460946023464203, 0.05523914471268654, -0.03002113476395607, -0.24697262048721313, 0.23904503881931305, -0.15189622342586517, -0.03373747318983078, 0.502168595790863, 0.06566665321588516, -0.26115095615386963, -0.06453022360801697, 0.06806580722332001, 0.00853058323264122, -0.12818123400211334, -0.2609414756298065, 0.004728136584162712, -0.03035365790128708, -0.06656614691019058, 0.17197631299495697, -0.058868810534477234, 0.2700216770172119, -0.12606224417686462, 0.33104389905929565, 0.20877617597579956, 0.15549926459789276, 0.0846952572464943, 0.013366863131523132, 0.20232270658016205, 0.12882864475250244, 0.12136552482843399, 0.1580013632774353, -0.07078966498374939, 0.25705859065055847, 0.02487761527299881, -0.07617715001106262, -0.007841411978006363, 0.33350980281829834, -0.33863329887390137, 0.15942130982875824, 0.0026581971906125546, -0.16646437346935272, 0.3518216907978058, -0.0122822942212224, -0.12231290340423584, 0.10254073888063431, 0.13919846713542938, -0.10381705313920975, -0.05152583494782448, 0.5680357813835144, 0.20062626898288727, 0.030121412128210068, 0.5001777410507202, 0.08568538725376129, -0.04003031924366951, 0.2109011858701706, 0.044015124440193176, 0.3199205994606018, 0.477145254611969, 0.07485055923461914, -0.11321045458316803, 0.025738749653100967, 0.23483610153198242, 0.11907745152711868, 0.05644068866968155, 0.16241036355495453, -0.04504361003637314, -0.050260014832019806, 0.11100438237190247, -0.17791956663131714, 0.08935988694429398, 0.49929049611091614, -0.10340704023838043, -0.22849802672863007, 0.042944394052028656, 0.10159806907176971, 0.027173761278390884, 0.12443164736032486, 0.738997757434845, -0.48798197507858276, -0.26755407452583313, -0.19579683244228363, 0.1438283771276474, -0.4215511977672577, 0.08674022555351257, -0.31011244654655457, 0.06441142410039902, -0.09468194842338562, -0.0019607506692409515, -0.16727225482463837, -0.05734232813119888, 0.3959067165851593, 0.2621211111545563, -0.19068606197834015, 0.03612062335014343, -0.12258759886026382, 0.10783541947603226, -0.13998591899871826, -0.012021541595458984, 0.4424281716346741, -0.30636030435562134, 0.2628045082092285, 0.014340197667479515, 0.6072162389755249, 0.18633612990379333, 0.20774537324905396, -0.10181604325771332, -0.2245771735906601, 0.06042516231536865, -0.08757299929857254, 0.012050177901983261, -0.055068016052246094, 0.4431183338165283, 0.08779536187648773, 0.26832982897758484, 0.08101987838745117, -0.14020629227161407, 0.22144773602485657, 0.1270507127046585, -0.18602310121059418, 0.034433431923389435, -0.29817187786102295, -0.0920553207397461, -0.14600978791713715, -0.29239657521247864, -0.17628207802772522, -0.3040122985839844, 0.4630802571773529, -0.23748037219047546, -0.15882517397403717, 0.09287767112255096, 0.026172105222940445, 0.07179025560617447, 0.00036934204399585724, 0.5908597111701965, 0.3205854296684265, -0.0013034939765930176, -0.2988016605377197, -0.5059593915939331, -0.3325135409832001, 0.1911904215812683, -0.16133569180965424, 0.13938629627227783, 0.08501014113426208, 0.1867828071117401, -0.02311478555202484, 0.45860767364501953, 0.1640261709690094, -0.08083482086658478, -0.2100592404603958, -0.06281013786792755, -0.5517159700393677, -0.13538366556167603, 0.26514410972595215, -0.26066315174102783, -0.16066502034664154, -0.16178691387176514, 0.28976377844810486, -0.1578645408153534, 0.008336901664733887, -0.04479454457759857, -0.27315443754196167, -0.33197638392448425, -0.24862372875213623, 0.15242043137550354, 0.12959279119968414, 0.12297084927558899, 0.007584661245346069, -0.33314165472984314, 0.12094786763191223, -0.2889213562011719, -0.10682777315378189, 0.011426374316215515, 0.21625448763370514, 0.250301331281662, 0.02853419817984104, 0.26356276869773865, -0.09871037304401398, 0.4951654374599457, 0.1588352471590042, 0.02855980396270752, -0.16628402471542358, -0.08736930787563324, -0.07266160845756531, 0.059975218027830124, -0.04567209258675575, 0.40819433331489563, -0.32333117723464966, 0.2676665186882019, -0.4084075689315796, -0.2796306312084198, 0.392464816570282, -0.37216269969940186, -0.07345762848854065, -0.01743394136428833, 0.2117195427417755, 0.13282813131809235, -0.13608139753341675, -0.32911065220832825, -0.07494746148586273, -0.11623641848564148, -0.2900269031524658, -0.45486074686050415, 0.2603672742843628, -0.07690216600894928, 0.06409212946891785, -0.10680924355983734, 0.34088853001594543, 0.1294020712375641, -0.2829594314098358, -0.16729773581027985, -0.21795813739299774 ]
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) ```
[ -0.2108590304851532, 0.31305992603302, -0.049380332231521606, 0.03972186520695686, -0.08726876974105835, -0.18556517362594604, 0.2631244361400604, 0.15955516695976257, -0.44758912920951843, 0.10873610526323318, -0.3607417941093445, 0.5732998847961426, 0.19588486850261688, 0.08197712898254395, 0.17324364185333252, 0.19502784311771393, -0.008884638547897339, 0.3963029682636261, -0.057467564940452576, 0.16443295776844025, -0.22316746413707733, 0.3445679247379303, -0.3933257460594177, 0.1032145544886589, -0.22195595502853394, -0.19702430069446564, -0.15914304554462433, 0.29760801792144775, -0.39194461703300476, -0.2167634516954422, 0.27606332302093506, 0.1715928465127945, -0.059993281960487366, 0.17929421365261078, -0.00011473386257421225, -0.20161773264408112, -0.030263230204582214, -0.07803517580032349, -0.3729619085788727, -0.23220182955265045, 0.03527386486530304, -0.33603227138519287, 0.3040671646595001, -0.4480282664299011, 0.13219532370567322, 0.4575706422328949, 0.2886238396167755, 0.35761699080467224, 0.29694077372550964, 0.2759973108768463, 0.1642184555530548, 0.3264271020889282, 0.13450878858566284, 0.005115261767059565, 0.23249945044517517, -0.2531128525733948, -0.2693357467651367, 0.42635324597358704, 0.5949572324752808, -0.26348480582237244, -0.13266947865486145, 0.10558639466762543, -0.1691100001335144, 0.15007247030735016, 0.11090759932994843, 0.22836950421333313, -0.344156950712204, -0.4259018003940582, 0.22296041250228882, 0.060683850198984146, 0.02652881294488907, -0.26247644424438477, 0.054827980697155, 0.005878575146198273, 0.27631425857543945, -0.3067043721675873, 0.15786680579185486, 0.35337457060813904, -0.3750433027744293, 0.11998167634010315, 0.02427566424012184, 0.033813994377851486, -0.19935107231140137, 0.1858348846435547, 0.05883972346782684, 0.343229204416275, -0.02436065301299095, -0.013873949646949768, 0.21716883778572083, 0.08212460577487946, 0.3446105122566223, 0.21561802923679352, -0.03924686834216118, -0.0030005155131220818, -0.1363031566143036, -0.20195722579956055, -0.14123904705047607, 0.23442456126213074, 0.23353010416030884, -0.3234664797782898, 0.28888267278671265, 0.006387138739228249, 0.2911266088485718, 0.19484877586364746, -0.05898584797978401, -0.11289198696613312, 0.21247822046279907, -0.13682159781455994, 0.39696359634399414, 0.3778129816055298, -0.10544446110725403, 0.002824101597070694, -0.044340867549180984, -0.1690581887960434, -0.15089067816734314, -0.10000819712877274, 0.12246547639369965, -0.17493733763694763, -0.2805670201778412, 0.023782603442668915, -0.22518011927604675, 0.29884639382362366, 0.04263242706656456, 0.24172699451446533, -0.06156925484538078, -0.37630218267440796, 0.06695341318845749, -0.02502242848277092, -0.34266865253448486, -0.2041155844926834, -0.0023026647977530956, -0.019037291407585144, -0.16797615587711334, 0.09720903635025024, 0.30007511377334595, 0.1617031842470169, 0.4099462628364563, -0.11885754019021988, 0.1719551980495453, -0.10903345793485641, 0.03236810863018036, -0.2970982789993286, -0.09300362318754196, 0.27639320492744446, 0.38804447650909424, -0.24145792424678802, -0.2290213257074356, -0.5571017265319824, 0.045669712126255035, 0.16808468103408813, -0.012525200843811035, -0.008397631347179413, -0.15670877695083618, 0.1424696296453476, -0.2702634930610657, -0.04810602217912674, -0.30732473731040955, 0.010949742048978806, -0.2560478746891022, -0.18128329515457153, 0.03165686875581741, -0.07667209953069687, -0.17649976909160614, -0.07915295660495758, 0.0010234005749225616, 0.1980469524860382, -0.16310574114322662, -0.03464559465646744, 0.12622477114200592, 0.03138294070959091, 0.04735119640827179, 0.23515060544013977, -0.1826513558626175, 0.054192714393138885, 0.028838926926255226, 0.22627674043178558, 0.4111819863319397, -0.20428423583507538, -0.6995334029197693, -0.007170290220528841, -0.08847233653068542, -0.2417573630809784, -0.13286659121513367, 0.24760931730270386, 0.44656020402908325, -0.1435137689113617, 0.21046529710292816, 0.5739884376525879, -0.16754485666751862, 0.1404290497303009, -0.22272250056266785, -0.17239852249622345, -0.06703240424394608, 0.18465368449687958, 0.14324384927749634, -0.10967373102903366, 0.011932075023651123, 0.637118935585022, 0.23053854703903198, -0.1181582510471344, -0.232115238904953, -0.004396114498376846, 0.2663671672344208, -0.307822585105896, -0.10001197457313538, -0.05420035496354103, -0.5955290794372559, 0.030195817351341248, -0.27251386642456055, 0.19959959387779236, -0.01254892349243164, -0.19998201727867126, -0.03226301074028015, 0.002767818048596382, -0.021475505083799362, -0.041503820568323135, 0.14903368055820465, -0.10393498837947845, 0.03776175156235695, 0.01641112193465233, 0.13712643086910248, 0.3391304612159729, -0.2325786054134369, 0.4468236565589905, -0.003963744267821312, 0.2574148178100586, -0.2883979082107544, -0.3253239393234253, 0.15580134093761444, 0.24388766288757324, 0.03058543987572193, -0.1248806044459343, -0.06957944482564926, 0.2165549397468567, 0.10290727764368057, -0.08505699038505554, 0.06705203652381897, -0.3381919860839844, 0.3775499761104584, 0.08620890974998474, -0.05293312668800354, -0.07910649478435516, 0.07668915390968323, -0.0767320841550827, 0.018533185124397278, 0.22335022687911987, -0.31230801343917847, -0.07031091302633286, 0.10400648415088654, 0.21962209045886993, 0.06820762902498245, -0.13731427490711212, -0.1277414709329605, 0.21088209748268127, 0.22927658259868622, 0.244489386677742, 0.015720881521701813, 0.038441531360149384, -0.3653057813644409, 0.27148914337158203, 0.4710398018360138, 0.009044945240020752, 0.2805316746234894, -0.027613885700702667, -0.06783580034971237, -0.017648398876190186, 0.3719351887702942, 0.023189764469861984, 0.32941877841949463, 0.11665686219930649, -0.11657072603702545, 0.01113898679614067, -0.10210777819156647, -0.29719579219818115, 0.11165915429592133, 0.1278635412454605, 0.3178746700286865, -0.10101091116666794, 0.3305141031742096, -0.034372154623270035, -0.3430921733379364, -0.20008860528469086, -0.09329666942358017, 0.2421555519104004, 0.12846019864082336, 0.25913605093955994, -0.13541017472743988, -0.6098312735557556, -0.05507991462945938, -0.16726814210414886, -0.090360626578331, -0.22800885140895844, -0.023899409919977188, 0.45616281032562256, 0.276153564453125, 0.07358154654502869, 0.2938823103904724, 0.21407930552959442, 0.0367489755153656, 0.08619954437017441, 0.15598152577877045, -0.11963637173175812, -0.31850939989089966, 0.0718802958726883, 0.39522674679756165, 0.24352465569972992, 0.08686231076717377, -0.24151507019996643, -0.18731874227523804, -0.09849104285240173, -0.3265635371208191, 0.2032727152109146, -0.23567236959934235, 0.30599111318588257, -0.017747044563293457, 0.2388731837272644, -0.6264257431030273, -0.21012380719184875, 0.10280880331993103, -0.12800614535808563, -0.11295357346534729, -0.03887312859296799, -0.005858607590198517, -0.3610164523124695, -0.24185693264007568, -0.14751112461090088, -0.5018797516822815, -0.24258753657341003, 0.2574930191040039, -0.0037882179021835327, 0.19984561204910278, 0.2799195349216461, 0.07124128937721252, 0.3175303339958191, 0.09950865805149078, -0.2388117015361786, 0.06216571480035782, 0.02206607535481453, 0.23596256971359253, -0.24667274951934814, -0.5061877965927124, 0.06595106422901154, 0.05682873725891113, 0.032461293041706085, -0.2706143260002136, -0.4002765417098999, -0.465181440114975, 0.14091917872428894, 0.23564434051513672, -0.04908043146133423, -0.13650193810462952, 0.13784277439117432, -0.18993277847766876, 0.07168592512607574, 0.01500818133354187, -0.2303987443447113, -0.17598643898963928, 0.13017922639846802, 0.2832840383052826, -0.25433680415153503, 0.04595976695418358, 0.15705351531505585, 0.22293870151042938, 0.11176830530166626, -0.083133265376091, 0.15639086067676544, 0.0955338254570961, 0.1974157691001892, -0.18021884560585022, -0.42029011249542236, 0.33444952964782715, -0.11916840076446533, -0.09323754161596298, 0.34363552927970886, -0.15984144806861877, -0.604213535785675, -0.12987607717514038, 0.06717176735401154, -0.38668757677078247, -0.46388745307922363, 0.16150718927383423, 0.5007524490356445, -0.1398274004459381, 0.2503450810909271, 0.18707329034805298, 0.008182933554053307, -0.34842953085899353, -0.3713095784187317, 0.2292095273733139, -0.26665523648262024, 0.04256106913089752, 0.20368073880672455, -0.42458483576774597, -0.6710546612739563, 0.18562442064285278, 0.07730090618133545, 0.12946704030036926, 0.03373346105217934, 0.16176483035087585, 0.14359444379806519, 0.01068422943353653, 0.2023099660873413, 0.23272587358951569, -0.043824389576911926, 0.0860024243593216, 0.05781089514493942, -0.284172922372818, 0.02707858756184578, -0.39344486594200134, 0.5534502863883972, 0.82149738073349, 0.21728666126728058, -0.12233014404773712, 0.161693274974823, 0.512523353099823, -0.02566528506577015, 0.16802293062210083, 0.13292090594768524, -0.22827960550785065, -0.37134337425231934, -0.14193890988826752, 0.10894770175218582, -0.05841958522796631, 0.07994873821735382, 0.09687226265668869, -0.06940171867609024, 0.2451850324869156, -0.07987070083618164, -0.05815878510475159, 0.5733697414398193, -0.04309689998626709, 0.17280890047550201, 0.07421660423278809, -0.18566349148750305, -0.13516639173030853, 0.305774986743927, 0.2549997568130493, -0.048399049788713455, -0.2348625659942627, 0.2255319058895111, -0.31759583950042725, 0.16910533607006073, 0.5100562572479248, -0.028746042400598526, -0.008707329630851746, -0.2693237364292145, 0.009692050516605377, 0.3701845109462738, 0.0820201113820076, 0.256671279668808, 0.07617249339818954, -0.08871469646692276, 0.2006741166114807, 0.2374025583267212, -0.12386049330234528, 0.02468188852071762, -0.15152502059936523, -0.19545288383960724, -0.20539426803588867, 0.4353124499320984, 0.09283722192049026, 0.7878344655036926, -0.23235715925693512, 0.16138100624084473, 0.27821919322013855, 0.02919618785381317, 0.6408751010894775, -0.03151513263583183, 0.33673012256622314, -0.18474218249320984, -0.2963073253631592, 0.053685300052165985, -0.02115555852651596, 0.1532127559185028, 0.014196395874023438, -0.6053224802017212, 0.23526014387607574, 0.3834209442138672, 0.02982737310230732, 0.08889052271842957, 0.13620389997959137, -0.11280248314142227, -0.32899385690689087, -0.6385902166366577, 0.17903026938438416, -0.31387943029403687, 0.19821517169475555, -0.30542829632759094, 0.003389583434909582, -0.2572369873523712, 0.05614405870437622, -0.36994582414627075, 0.15100428462028503, -0.5142830610275269, -0.048113904893398285, 0.2063174694776535, -0.30066657066345215, -0.10240314900875092, 0.300640344619751, 0.17832377552986145, 0.20467081665992737, 0.06039530038833618, 0.06155680865049362, -0.058590881526470184, -0.30445829033851624, 0.08371365070343018, 0.03717843443155289, 0.07622479647397995, -0.2088422179222107, -0.08287840336561203, -0.07315363734960556, 0.27861690521240234, -0.23776978254318237, 0.3863203525543213, 0.2796516418457031, 0.18206268548965454, 0.020238172262907028, -0.06534850597381592, -0.037681981921195984, -0.1724327802658081, -0.28082188963890076, 0.10838428139686584, -0.10882590711116791, -0.06522068381309509, 0.28962716460227966, 0.1406569480895996, -0.17361317574977875, -0.0035924874246120453, 0.2703544497489929, -0.1790098249912262, 0.1063862144947052, 0.20714975893497467, 0.2732166349887848, -0.11991051584482193, -0.21892857551574707, 0.19247543811798096, -0.5636967420578003, -0.17945496737957, 0.02946116402745247, -0.07871925085783005, -0.026361390948295593, -0.08183688670396805, 0.631582498550415, 0.3025014400482178, 0.02565683051943779, -0.07841385900974274, -0.4058826267719269, -0.18444348871707916, 0.2054409384727478, 0.1786111444234848, -0.1519635021686554, 0.37673142552375793, 0.3331974148750305, 0.193341925740242, 0.04874858260154724, -0.23890846967697144, 0.2942793071269989, -0.40905633568763733, 0.16265268623828888, -0.057454802095890045, -0.3167046904563904, 0.1807573288679123, -0.16345810890197754, -0.006830867379903793, 0.5193843245506287, 0.12761372327804565, -0.237626314163208, -0.08323725312948227, 0.06300914287567139, 0.02087610960006714, -0.10651836544275284, -0.3156396448612213, -0.07924634218215942, -0.03797822818160057, -0.030868735164403915, 0.09227944910526276, -0.11090949177742004, 0.2121518850326538, -0.045085735619068146, 0.2560538947582245, 0.2237636148929596, 0.18538793921470642, 0.03360719978809357, -0.02662503719329834, 0.2571994364261627, 0.103535957634449, 0.208764910697937, 0.17264524102210999, 0.038589365780353546, 0.22618544101715088, -0.04137689620256424, -0.16636626422405243, 0.07988855242729187, 0.3123323917388916, -0.36842775344848633, 0.06631982326507568, -0.12071607261896133, -0.08334916085004807, 0.3914593756198883, 0.0945458635687828, -0.22146737575531006, 0.2422187626361847, 0.12057972699403763, -0.2034115493297577, 0.05944051593542099, 0.5620882511138916, 0.1985337734222412, 0.11759072542190552, 0.5118353962898254, 0.003311287611722946, -0.0001255124807357788, 0.14947068691253662, 0.1118350401520729, 0.29032543301582336, 0.40054959058761597, -0.05944668874144554, -0.1089259535074234, -0.040777262300252914, 0.1046377420425415, 0.12247244268655777, 0.104865163564682, 0.1897202730178833, -0.003135383129119873, -0.08938490599393845, 0.09117859601974487, -0.21281325817108154, 0.07399358600378036, 0.3456506133079529, -0.0707034021615982, -0.2760521471500397, 0.061199285089969635, 0.22912639379501343, 0.0624624565243721, 0.07917775958776474, 0.7251800298690796, -0.34647998213768005, -0.2281430959701538, -0.02409774251282215, 0.16770753264427185, -0.39005348086357117, 0.1112295538187027, -0.2578761577606201, 0.10365492850542068, -0.07326438277959824, 0.06669114530086517, -0.15407678484916687, -0.13330844044685364, 0.3501731753349304, 0.2939985692501068, -0.1731889396905899, -0.024732721969485283, -0.10646162182092667, 0.1602122187614441, -0.10827084630727768, -0.023033855482935905, 0.3853769898414612, -0.34098494052886963, 0.19518744945526123, -0.184858500957489, 0.5654686093330383, 0.17510512471199036, 0.18691007792949677, -0.2589702606201172, -0.19222979247570038, -0.04654514417052269, -0.009189987555146217, -0.12912294268608093, -0.10915717482566833, 0.4811045229434967, 0.1795612871646881, 0.2597264051437378, 0.07102728635072708, -0.13908211886882782, 0.2577846646308899, 0.19828620553016663, -0.2809126675128937, 0.14965423941612244, -0.2773096561431885, 0.04808777570724487, -0.17036616802215576, -0.25108009576797485, -0.10992150753736496, -0.26709216833114624, 0.41964513063430786, -0.26302942633628845, -0.08880849182605743, 0.0830579325556755, -0.06837192177772522, 0.06529147922992706, 0.036942094564437866, 0.5547948479652405, 0.33888235688209534, 0.017961259931325912, -0.25608697533607483, -0.40430766344070435, -0.3963000774383545, 0.22965607047080994, -0.2289820909500122, -0.07774485647678375, 0.08669283241033554, 0.2964426577091217, -0.050080738961696625, 0.49343276023864746, 0.014594167470932007, 0.044201672077178955, -0.18333584070205688, -0.009463541209697723, -0.5380403995513916, -0.06271453201770782, 0.21355772018432617, -0.33248093724250793, -0.07470738142728806, -0.2555917799472809, 0.22630353271961212, -0.3445778489112854, 0.011229854077100754, -0.022472016513347626, -0.370181679725647, -0.24126386642456055, -0.08208002150058746, 0.17296934127807617, 0.07968718558549881, 0.06808533519506454, 0.01229977235198021, -0.23334704339504242, 0.03424596041440964, -0.2586531341075897, -0.20849581062793732, 0.040847696363925934, 0.20121587812900543, 0.22902452945709229, 0.08602142333984375, 0.36334502696990967, -0.09533359110355377, 0.5152747631072998, 0.09685724973678589, 0.04516313597559929, -0.04169933870434761, -0.024246975779533386, -0.12852951884269714, 0.13830913603305817, -0.22996149957180023, 0.44079622626304626, -0.3113741874694824, 0.2767087519168854, -0.4855022430419922, -0.3400691747665405, 0.36335206031799316, -0.5322354435920715, -0.1876487284898758, 0.04002277925610542, 0.22818812727928162, 0.08981306105852127, -0.17658741772174835, -0.29772669076919556, -0.02463051676750183, -0.11291200667619705, -0.23598924279212952, -0.41634872555732727, 0.23685230314731598, -0.1849517822265625, 0.04134880378842354, -0.09677921235561371, 0.28482678532600403, 0.03882337361574173, -0.244221493601799, -0.21315842866897583, -0.21907676756381989 ]
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 ?
[ -0.06062939018011093, -0.43357160687446594, -0.05662921816110611, -0.015485309064388275, -0.16559098660945892, 0.04524776339530945, 0.18558672070503235, 0.18404962122440338, 0.20415988564491272, -0.14907832443714142, -0.24861812591552734, 0.2844125032424927, -0.1800074279308319, 0.47155213356018066, 0.3802827000617981, -0.4318718910217285, 0.07262524217367172, -0.14835751056671143, -0.3279542624950409, 0.3465222120285034, 0.043432027101516724, 0.12348020076751709, -0.18072369694709778, 0.029902007430791855, -0.35131070017814636, -0.16163277626037598, -0.3850715458393097, 0.07486452162265778, 0.014117991551756859, 0.031469352543354034, 0.06988038122653961, 0.3947067856788635, -0.04164344444870949, 0.18248435854911804, -0.00011498481762828305, -0.10626955330371857, 0.11025393009185791, -0.1087462455034256, 0.09373389184474945, -0.03407789766788483, 0.13381794095039368, -0.3790649175643921, 0.058088354766368866, -0.16541621088981628, 0.20267626643180847, 0.14574658870697021, 0.1934349536895752, 0.05130394548177719, 0.3670564889907837, -0.2185332477092743, 0.08402933180332184, 0.26443198323249817, -0.13484972715377808, 0.3218310475349426, -0.0651739165186882, 0.02047070860862732, 0.044885843992233276, 0.23994137346744537, 0.6195873022079468, -0.47857508063316345, -0.027776941657066345, 0.22522087395191193, -0.011841628700494766, -0.012316416949033737, 0.12501248717308044, -0.06387162208557129, -0.2911246716976166, -0.2389596402645111, -0.3664330244064331, 0.24455338716506958, -0.04346802085638046, -0.21901509165763855, -0.22686205804347992, -0.5487809777259827, -0.025223106145858765, -0.01662977784872055, -0.326879620552063, 0.06366682797670364, -0.16956283152103424, -0.05313921347260475, -0.1652791053056717, -0.0777100995182991, -0.06440579891204834, 0.1307315230369568, 0.29616230726242065, 0.5442551374435425, 0.0387100987136364, 0.053566258400678635, 0.2903417646884918, 0.036765966564416885, -0.11274214088916779, -0.2751785218715668, 0.3076981008052826, 0.07872704416513443, -0.32095393538475037, -0.34821999073028564, 0.0695447325706482, -0.31259289383888245, 0.32180315256118774, 0.08210974186658859, 0.13770149648189545, 0.06355192512273788, -0.13061709702014923, 0.21757471561431885, 0.38325244188308716, 0.008566847071051598, -0.0750824585556984, -0.16860586404800415, 0.007640670984983444, -0.14267084002494812, 0.1661435067653656, 0.1513281613588333, 0.11848486959934235, -0.004000247921794653, -0.4497514069080353, -0.06211620569229126, -0.0710141658782959, 0.05218597874045372, -0.12578356266021729, -0.5080181360244751, -0.11419260501861572, -0.09674488008022308, -0.11398350447416306, -0.17604142427444458, -0.26717716455459595, 0.17692387104034424, 0.003627914935350418, 0.27180927991867065, -0.1638910472393036, 0.03908424824476242, -0.03798029571771622, 0.01682923547923565, -0.4351593852043152, -0.08235833048820496, 0.2814764380455017, 0.08760927617549896, 0.05325385928153992, 0.16166581213474274, -0.04158489406108856, 0.06549911201000214, 0.332254022359848, 0.028684310615062714, 0.13818120956420898, -0.04580060765147209, 0.17860060930252075, -0.26701080799102783, -0.010549895465373993, 0.15709362924098969, -0.3184623420238495, -0.12240960448980331, -0.15247127413749695, -0.20893260836601257, 0.05303369462490082, 0.010931023396551609, -0.17490319907665253, -0.28919920325279236, -0.4556143581867218, 0.7248806357383728, 0.015414074063301086, -0.08319571614265442, -0.19418630003929138, 0.13860708475112915, -0.1514989137649536, -0.09328716993331909, 0.22360345721244812, 0.11550799012184143, -0.07817286252975464, -0.27755212783813477, 0.019101375713944435, -0.08028481900691986, 0.012738101184368134, 0.304551899433136, -0.06466139107942581, 0.2548964023590088, 0.06150007247924805, 0.2096283882856369, 0.30092281103134155, -0.39963531494140625, -0.0024379361420869827, 0.34459465742111206, -0.05548939108848572, 0.3846750855445862, 0.10268242657184601, 0.12903325259685516, 0.00221403781324625, -0.19203858077526093, 0.2654207944869995, 0.9147472977638245, -0.38112103939056396, 0.06711127609014511, -0.19821713864803314, -0.3537558317184448, 0.5127946138381958, 0.2841840982437134, 0.011132195591926575, -0.3445623815059662, -0.13308793306350708, -0.0049951025284826756, 0.07843337953090668, 0.0030071884393692017, 0.03737063705921173, -0.02651938796043396, -0.3109757602214813, 0.07546744495630264, -0.19495409727096558, -0.115241140127182, -0.4247463643550873, -0.1003587543964386, 0.04577895998954773, 0.3210233747959137, 0.26057329773902893, -0.15339195728302002, 0.22681349515914917, -0.3086334466934204, -0.15567214787006378, -0.2565402388572693, 0.13309957087039948, -0.026752278208732605, -0.014473576098680496, -0.20608532428741455, -0.09645234048366547, 0.32367751002311707, 0.0577915757894516, -0.07589920610189438, -0.4959026575088501, 0.24824412167072296, -0.077762171626091, 0.020057471469044685, 0.07954811304807663, 0.6802391409873962, -0.18799707293510437, -0.07588966190814972, 0.1854725182056427, 0.2653414309024811, -0.3162192702293396, 0.25385794043540955, 0.22005969285964966, 0.33087337017059326, 0.2140362411737442, -0.0697961375117302, 0.10537343472242355, -0.0637868195772171, -0.04165664687752724, -0.22341936826705933, -0.19158995151519775, 0.43757516145706177, -0.17595253884792328, 0.463652104139328, -0.03217608481645584, 0.02915244735777378, -0.22143886983394623, -0.01690283790230751, -0.23426449298858643, 0.38474154472351074, 0.31314295530319214, -0.20771293342113495, 0.33080944418907166, -0.046972163021564484, -0.44120243191719055, 0.2577657401561737, 0.24474388360977173, 0.06818315386772156, 0.09573803842067719, -0.009551957249641418, -0.045534223318099976, -0.3349413573741913, -0.060299456119537354, 0.34204697608947754, 0.5666442513465881, 0.26995715498924255, 0.19837501645088196, 0.06839533895254135, -0.2664468288421631, -0.11241086572408676, 0.0692981630563736, 0.16281472146511078, 0.012806608341634274, 0.2764888405799866, 0.03590695559978485, 0.07019908726215363, -0.01635851338505745, -0.09048351645469666, 0.13983921706676483, -0.18486753106117249, -0.15955151617527008, 0.013510974124073982, -0.23286262154579163, -0.661297082901001, -0.2655385434627533, 0.05956576392054558, -0.3062054514884949, -0.319520503282547, 0.12936243414878845, 0.017231788486242294, -0.14742322266101837, 0.17125503718852997, 0.23219692707061768, 0.607446551322937, -0.39070576429367065, -0.1505921334028244, -0.06880813837051392, -0.3318815529346466, -0.1426524519920349, 0.07787388563156128, 0.549889862537384, 0.3654056191444397, 0.43545612692832947, 0.17507439851760864, 0.0016453061252832413, 0.04031530022621155, -0.4197556972503662, 0.09459864348173141, -0.3344561457633972, 0.16005419194698334, 0.06284233927726746, -0.1625741869211197, 0.14567077159881592, -0.42311927676200867, -0.08997100591659546, 0.024390386417508125, -0.09786457568407059, 0.017521221190690994, -0.0396757647395134, 0.027871476486325264, -0.22343939542770386, -0.3024548888206482, -0.5716314911842346, -0.4520237147808075, 0.3431037664413452, -0.27985459566116333, 0.14475837349891663, 0.0694185420870781, -0.22567537426948547, 0.14136293530464172, -0.17480261623859406, 0.07033857703208923, -0.11356350034475327, -0.10608261823654175, 0.042191363871097565, -0.34368401765823364, -0.2414880245923996, -0.38526827096939087, -0.026993216946721077, 0.1701684594154358, -0.0015302281826734543, -0.06010874733328819, -0.2366362065076828, 0.07347780466079712, -0.21928173303604126, 0.11132682859897614, 0.29573196172714233, 0.1376895308494568, -0.16666603088378906, 0.13377836346626282, -0.04997451975941658, -0.18687541782855988, 0.20780950784683228, 0.32819244265556335, 0.3863658905029297, 0.15270429849624634, 0.0017188973724842072, 0.2775653302669525, 0.6705092191696167, 0.3482150733470917, 0.04389109089970589, 0.09031020104885101, 0.25889748334884644, 0.06046076864004135, 0.1286640763282776, -0.18067094683647156, 0.26686447858810425, -0.20002683997154236, 0.5133737325668335, 0.1512202024459839, 0.1806647777557373, -0.10541103035211563, -0.2131507396697998, 0.12126965820789337, -0.3510887920856476, -0.1996564269065857, 0.3020295798778534, -0.31928250193595886, 0.021582331508398056, -0.05449039489030838, -0.3992007374763489, -0.2533678114414215, -0.04955428093671799, 0.017427291721105576, 0.3520650267601013, -0.05508100241422653, 0.19885984063148499, -0.4190732538700104, -0.09170428663492203, -0.46005338430404663, 0.19697128236293793, 0.08974097669124603, 0.3336322605609894, -0.06856122612953186, -0.11177203059196472, -0.08517734706401825, 0.1311119943857193, 0.5710868835449219, -0.5576232671737671, -0.2217063158750534, 0.17702358961105347, 0.1506168097257614, -0.08356249332427979, -0.12241517007350922, -0.38056671619415283, -0.026714732870459557, 0.46636343002319336, 0.42542821168899536, -0.4082850515842438, -0.35964280366897583, 0.2669551968574524, -0.16865240037441254, -0.020689569413661957, -0.13011807203292847, -0.20354948937892914, -0.21533101797103882, -0.17324474453926086, 0.026581868529319763, 0.22452165186405182, 0.16896408796310425, -0.1544281542301178, 0.15072380006313324, -0.046022042632102966, 0.23721414804458618, 0.21118459105491638, -0.012833455577492714, -0.05038946494460106, 0.26388460397720337, 0.0650329738855362, 0.07246600836515427, -0.03061072528362274, -0.2702557444572449, 0.10697230696678162, -0.13292673230171204, 0.06653112173080444, 0.2735390067100525, -0.07887067645788193, 0.7340445518493652, 0.15070095658302307, 0.3863653838634491, 0.17446628212928772, -0.1042366474866867, 0.020268280059099197, -0.2474546730518341, 0.21436722576618195, 0.20877262949943542, 0.24122527241706848, -0.1877690851688385, -0.19011326134204865, 0.5577906370162964, 0.33030593395233154, -0.015693001449108124, 0.41446152329444885, -0.4715427756309509, -0.2955447733402252, 0.3173159956932068, 0.037461526691913605, 0.7747876644134521, -0.0559433251619339, 0.3117319345474243, 0.08896906673908234, -0.4208648204803467, 0.5706254839897156, -0.43463313579559326, 0.15796628594398499, -0.09969194233417511, -0.18400435149669647, 0.030260616913437843, -0.10041100531816483, -0.026499107480049133, 0.2533981204032898, -0.38214847445487976, 0.3142700791358948, 0.4203941226005554, -0.25699344277381897, -0.03189202398061752, 0.5367090106010437, -0.05316828936338425, -0.3928298056125641, -0.05631323903799057, 0.04181140661239624, -0.32738596200942993, 0.08863338828086853, -0.12786144018173218, -0.3133983016014099, 0.23929113149642944, 0.1586398184299469, -0.3130236864089966, 0.10464895516633987, -0.06307268142700195, 0.14268921315670013, 0.22421951591968536, -0.035015303641557693, 0.24764713644981384, -0.13931328058242798, 0.0429934561252594, -0.1307624876499176, -0.06671702861785889, 0.03300987556576729, -0.10206114500761032, 0.07684885710477829, 0.09859734773635864, -0.16361364722251892, 0.5141545534133911, -0.0033410824835300446, 0.008826300501823425, -0.29763033986091614, 0.2612650990486145, -0.3895629644393921, -0.276950865983963, -0.024676434695720673, 0.18370632827281952, 0.09997154027223587, -0.29114770889282227, 0.6143852472305298, 0.0873115211725235, 0.0357685349881649, 0.04025207459926605, 0.1385815590620041, -0.23400579392910004, 0.5710294842720032, -0.24325956404209137, -0.01429014652967453, 0.08371679484844208, -0.013701401650905609, -0.11303932219743729, -0.4529513716697693, -0.0207376591861248, 0.38573089241981506, 0.07245340943336487, -0.022204995155334473, 0.06766734272241592, 0.1305583119392395, -0.06083385646343231, -0.017283156514167786, -0.2517119348049164, -0.29927992820739746, 0.10248804837465286, 0.24880841374397278, 0.24665340781211853, 0.15516209602355957, -0.041943833231925964, -0.24314746260643005, -0.13975267112255096, 0.5461680889129639, -0.03090217337012291, 0.26203492283821106, -0.051449257880449295, 0.3808257579803467, 0.3815823495388031, 0.5050756931304932, -0.2507596015930176, 0.07350698858499527, 0.06648766249418259, 0.1924353688955307, 0.44197627902030945, -0.26780688762664795, -0.032334066927433014, 0.24625711143016815, -0.1350875347852707, 0.10860054194927216, -0.32450005412101746, -0.15105046331882477, -0.20546089112758636, 0.10554122179746628, 0.19726675748825073, 0.02915121428668499, 0.047509338706731796, -0.033540308475494385, -0.13116846978664398, -0.10678388178348541, 0.021674226969480515, -0.37187138199806213, -0.12796550989151, 0.23094749450683594, -0.1180993914604187, 0.0367705412209034, 0.07956629991531372, 0.13500702381134033, 0.004463784396648407, -0.09252685308456421, 0.31753668189048767, 0.09242818504571915, 0.06634555011987686, -0.1891518533229828, -0.18550202250480652, 0.04238702729344368, 0.23217496275901794, 0.1891823709011078, 0.0023199990391731262, 0.08651171624660492, -0.17059358954429626, -0.1667363941669464, 0.21267817914485931, 0.2336457073688507, 0.01032559759914875, -0.09136780351400375, 0.1436552107334137, 0.09549880772829056, -0.0544309988617897, -0.005216378252953291, 0.22590729594230652, 0.05995369330048561, -0.007213480770587921, 0.06645393371582031, 0.19826093316078186, 0.39348655939102173, -0.18172939121723175, 0.045478835701942444, 0.035260286182165146, -0.39731937646865845, 0.1519799381494522, -0.040100280195474625, 0.00020148977637290955, 0.00528678297996521, 0.06006809324026108, 0.0377880223095417, 0.12491361796855927, -0.32155483961105347, 0.32893696427345276, 0.262973815202713, 0.025910526514053345, 0.035576287657022476, 0.17994658648967743, -0.028497472405433655, -0.15546952188014984, 0.27964523434638977, 0.07556164264678955, 0.43249401450157166, 0.0586092472076416, 0.8388553857803345, -0.2928153872489929, -0.513799250125885, 0.11348996311426163, 0.424418568611145, -0.002087157219648361, -0.19419698417186737, 0.054136864840984344, 0.11889065057039261, -0.3490162789821625, 0.07172706723213196, -0.06716296076774597, -0.1238756775856018, 0.5954532027244568, -0.0036203712224960327, 0.1662234663963318, -0.22186864912509918, -0.17409703135490417, 0.28185030817985535, 0.2393641471862793, -0.11452360451221466, 0.11490707099437714, 0.08922533690929413, -0.12978611886501312, -0.15155154466629028, 0.5569117069244385, 0.25876203179359436, 0.05257067829370499, -0.4229050278663635, 0.014233014546334743, 0.09556449949741364, 0.10881701111793518, 0.07581400871276855, 0.08942347019910812, 0.45313453674316406, -0.1393473595380783, 0.27781942486763, -0.011475536040961742, -0.011696679517626762, -0.154131218791008, 0.033524323254823685, -0.4807209372520447, -0.3438222110271454, 0.39191505312919617, 0.1465260088443756, -0.16903084516525269, 0.10381101816892624, 0.26207026839256287, -0.7270404696464539, -0.2905240058898926, 0.41981521248817444, -0.028902582824230194, -0.03651423752307892, -0.20693422853946686, 0.06619679182767868, 0.18844348192214966, 0.5728141665458679, 0.27638599276542664, 0.32753732800483704, -0.38855424523353577, -0.11769816279411316, -0.5355229377746582, -0.17271271347999573, 0.1998899132013321, -0.04911524057388306, -0.186818465590477, 0.14797520637512207, -0.02143321931362152, 0.3246809244155884, -0.26161473989486694, 0.10892422497272491, 0.16829948127269745, 0.07349081337451935, -0.5057327151298523, 0.11960236728191376, 0.20687857270240784, 0.04922626540064812, 0.19890852272510529, -0.057058461010456085, -0.08498969674110413, 0.13170494139194489, -0.07632112503051758, -0.0857621505856514, 0.005865886807441711, 0.2451675534248352, 0.16403938829898834, -0.026462681591510773, 0.08232583105564117, 0.4630805552005768, -0.01387227512896061, -0.19936423003673553, -0.019803274422883987, 0.04457425698637962, -0.4071342647075653, -0.07029031962156296, -0.09318269789218903, 0.1952255666255951, -0.39534685015678406, 0.29372358322143555, 0.008449711836874485, -0.11666037887334824, 0.0764622613787651, 0.02875891514122486, -0.15167874097824097, -0.049831509590148926, 0.08785512298345566, 0.1956939697265625, 0.29837992787361145, 0.740128219127655, -0.0052440352737903595, 0.12361189723014832, -0.17261390388011932, -0.06902056932449341, 0.027409732341766357, -0.053223028779029846, -0.23431609570980072, -0.10208185017108917, 0.06130241975188255, 0.16821064054965973, 0.21723411977291107, -0.6970353126525879, -0.3775850236415863, 0.03720245510339737, -0.0556478425860405, 0.013423949480056763, 0.7778046131134033, 0.13142307102680206, -0.16813664138317108, -0.16086889803409576, -0.3078276515007019, 0.011136813089251518, -0.297132670879364, 0.04810846969485283, -0.45130521059036255 ]
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.
[ -0.07244556397199631, -0.467218816280365, -0.040159933269023895, -0.055031321942806244, -0.304560124874115, 0.0747440978884697, 0.1744338721036911, 0.06729500740766525, 0.2653241455554962, -0.1193469762802124, -0.2597564160823822, 0.3054881989955902, -0.16949790716171265, 0.3780532777309418, 0.35094723105430603, -0.4499828517436981, -0.021549955010414124, -0.14634406566619873, -0.2921192944049835, 0.324207067489624, 0.05431655794382095, 0.11644261330366135, -0.15435390174388885, 0.06620006263256073, -0.2752176523208618, -0.17669454216957092, -0.34399914741516113, 0.12340910732746124, 0.08483219146728516, 0.07976438105106354, 0.13584452867507935, 0.39544519782066345, -0.018593134358525276, 0.12798629701137543, -0.000111412737169303, -0.09476036578416824, 0.09925207495689392, -0.09712206572294235, 0.1526375710964203, 0.03999316692352295, 0.1779927909374237, -0.3429226577281952, 0.0634026750922203, -0.14008671045303345, 0.13159680366516113, 0.24719160795211792, 0.19724278151988983, 0.08876565098762512, 0.4500899910926819, -0.3533526659011841, 0.10489170253276825, 0.3131389617919922, -0.2438289225101471, 0.30103304982185364, -0.08568572252988815, -0.1528027057647705, 0.011506261304020882, 0.28885069489479065, 0.6218158006668091, -0.5280038714408875, -0.015143148601055145, 0.10488273203372955, -0.04628778249025345, 0.04999203234910965, 0.13104325532913208, -0.0666382759809494, -0.28239738941192627, -0.25352394580841064, -0.3627510964870453, 0.2775130569934845, -0.12239693850278854, -0.06288975477218628, -0.192586749792099, -0.5656943321228027, -0.003744387999176979, 0.05627430975437164, -0.3290136158466339, 0.08781717717647552, -0.2160564512014389, 0.015035653486847878, -0.12497919052839279, -0.1268686205148697, -0.007691536098718643, 0.11148081719875336, 0.30812910199165344, 0.5256880521774292, 0.07436874508857727, 0.13465401530265808, 0.35248714685440063, 0.10151573270559311, -0.14089272916316986, -0.26514488458633423, 0.29077792167663574, 0.13547110557556152, -0.3288780450820923, -0.37709754705429077, 0.15168583393096924, -0.2695583701133728, 0.28027063608169556, 0.09671274572610855, 0.14681369066238403, 0.05229128897190094, -0.08265077322721481, 0.26438283920288086, 0.38912948966026306, 0.012631913647055626, -0.09676509350538254, -0.11974359303712845, 0.02395952120423317, -0.10591056942939758, 0.19155767560005188, 0.13948854804039001, 0.14466069638729095, 0.009004228748381138, -0.4349183440208435, -0.11233647912740707, -0.19096265733242035, 0.09654764831066132, -0.18051843345165253, -0.5517171621322632, -0.08570057153701782, -0.10183704644441605, -0.020415570586919785, -0.12436326593160629, -0.2886071503162384, 0.1530826836824417, -0.07474653422832489, 0.15041980147361755, -0.18044017255306244, 0.02078830450773239, -0.07014726847410202, 0.06636682152748108, -0.462980717420578, -0.03496813774108887, 0.2592128813266754, 0.10559295117855072, 0.07144852727651596, 0.1174030676484108, -0.020503327250480652, 0.16793939471244812, 0.42610931396484375, 0.030805256217718124, 0.028726674616336823, -0.03104657307267189, 0.14761847257614136, -0.3091065287590027, -0.031393419951200485, 0.23605014383792877, -0.3518315851688385, -0.09727918356657028, -0.14454936981201172, -0.17756307125091553, 0.13741560280323029, 0.04549838975071907, -0.2158234715461731, -0.26927056908607483, -0.441557377576828, 0.7260332107543945, -0.013283692300319672, -0.02393115684390068, -0.17176997661590576, 0.259479820728302, -0.20922605693340302, -0.11802931129932404, 0.18940530717372894, 0.09559951722621918, -0.021369582042098045, -0.31907832622528076, -0.10603727400302887, -0.06091995909810066, -0.015615783631801605, 0.3213063180446625, -0.151783749461174, 0.1992667317390442, 0.1068788468837738, 0.17462219297885895, 0.2618972957134247, -0.4294564425945282, -0.05689644813537598, 0.3527347445487976, -0.061471495777368546, 0.38411834836006165, 0.05087510496377945, 0.17250452935695648, -0.009615043178200722, -0.21683649718761444, 0.24459490180015564, 0.983712911605835, -0.404123991727829, 0.0024759043008089066, -0.185939759016037, -0.4244101345539093, 0.5637450218200684, 0.3482625186443329, 0.0648256316781044, -0.29352274537086487, -0.18867358565330505, 0.03436806797981262, 0.046717651188373566, 0.0923098772764206, 0.07490967959165573, -0.08023171126842499, -0.2644890546798706, 0.010359544306993484, -0.22413411736488342, -0.1698399782180786, -0.42165297269821167, -0.00873611494898796, -0.036963459104299545, 0.3787110149860382, 0.37277835607528687, -0.14035218954086304, 0.237596333026886, -0.30643683671951294, -0.11149434745311737, -0.31171655654907227, 0.14455872774124146, -0.06806737184524536, 0.012566164135932922, -0.23941916227340698, -0.06527689844369888, 0.2053755819797516, 0.09740221500396729, -0.10502316802740097, -0.4230077266693115, 0.29147908091545105, -0.04073402285575867, 0.021507807075977325, -0.013076692819595337, 0.614617109298706, -0.2161233127117157, -0.116585873067379, 0.20798203349113464, 0.291594535112381, -0.31383368372917175, 0.2109994739294052, 0.30932095646858215, 0.3545565605163574, 0.1958429217338562, -0.06140464171767235, 0.09926991164684296, -0.012211006134748459, -0.08991379290819168, -0.17108172178268433, -0.08095481246709824, 0.4797084927558899, -0.20128488540649414, 0.3575448989868164, -0.02191038988530636, 0.03489571064710617, -0.20772439241409302, -0.05692202225327492, -0.1906934380531311, 0.3881007730960846, 0.3106745183467865, -0.22799547016620636, 0.3347799479961395, -0.0767851322889328, -0.4003092050552368, 0.2434409260749817, 0.21397249400615692, 0.013479337096214294, 0.09026619046926498, -0.023068010807037354, 0.0312410369515419, -0.3101979196071625, -0.06050332635641098, 0.3375436067581177, 0.5514167547225952, 0.2894958555698395, 0.17737038433551788, 0.016729244962334633, -0.21758225560188293, -0.07269231975078583, 0.04261384159326553, 0.1653767228126526, -0.06318723410367966, 0.14198440313339233, -0.041099268943071365, 0.09890101104974747, 0.02665790170431137, -0.10134835541248322, 0.12281974405050278, -0.14453928172588348, -0.04268159717321396, -0.037505075335502625, -0.15475626289844513, -0.5833339095115662, -0.19117631018161774, 0.052576228976249695, -0.35153090953826904, -0.2960585355758667, 0.2087559700012207, 0.05685410648584366, -0.17757241427898407, 0.16798126697540283, 0.19975121319293976, 0.6398568749427795, -0.39235299825668335, -0.1751716583967209, -0.05248171091079712, -0.29990026354789734, -0.11523335427045822, 0.1373639702796936, 0.43582555651664734, 0.30238285660743713, 0.49836045503616333, 0.23070229589939117, -0.03288009762763977, 0.12016383558511734, -0.3862852454185486, 0.11135131120681763, -0.3567821979522705, 0.2921696603298187, 0.09976066648960114, -0.10114516317844391, 0.0970672070980072, -0.3661125898361206, -0.07395435869693756, -0.09430330991744995, -0.06166940927505493, -0.04039396345615387, 0.008098820224404335, 0.026680387556552887, -0.2271045446395874, -0.2285715937614441, -0.5969739556312561, -0.49180614948272705, 0.2876458168029785, -0.2523242235183716, 0.09606648981571198, 0.043231528252363205, -0.2876448929309845, 0.20857645571231842, -0.1413346379995346, 0.03815591707825661, -0.18621958792209625, -0.07839317619800568, -0.02086154744029045, -0.3604215979576111, -0.17928363382816315, -0.422829806804657, -0.012583755888044834, 0.1930728256702423, -0.07771419733762741, -0.06733407080173492, -0.2891920506954193, 0.032975926995277405, -0.10656090080738068, 0.08936822414398193, 0.23315834999084473, 0.14413048326969147, -0.20324985682964325, 0.07360047847032547, -0.02718188613653183, -0.1418774425983429, 0.2734167277812958, 0.39811575412750244, 0.36558014154434204, 0.14558780193328857, -0.02538781240582466, 0.21974122524261475, 0.7495047450065613, 0.28150755167007446, 0.08701758086681366, 0.10910997539758682, 0.29363590478897095, -0.03916487470269203, 0.09498448669910431, -0.17914023995399475, 0.25715699791908264, -0.11236843466758728, 0.4973568916320801, 0.1744980365037918, 0.15857043862342834, -0.20294521749019623, -0.27151721715927124, 0.019823230803012848, -0.2972184121608734, -0.21401159465312958, 0.2666808068752289, -0.2744174003601074, -0.002879694104194641, -0.10535693168640137, -0.4169650077819824, -0.23437228798866272, -0.08540157228708267, -0.02018485963344574, 0.28638947010040283, 0.04233963042497635, 0.17477644979953766, -0.37227681279182434, -0.024910947307944298, -0.4501193165779114, 0.13502947986125946, 0.053518146276474, 0.3062891662120819, -0.12003666162490845, -0.10522732138633728, -0.099813312292099, 0.15174566209316254, 0.5571808218955994, -0.5467590689659119, -0.15157878398895264, 0.08499222993850708, 0.08082776516675949, -0.13787327706813812, -0.07683352380990982, -0.34837594628334045, -0.013061403296887875, 0.4500737488269806, 0.3170126676559448, -0.43303248286247253, -0.36202460527420044, 0.22607110440731049, -0.1992088109254837, -0.07860724627971649, -0.15291933715343475, -0.20009441673755646, -0.18559390306472778, -0.12708865106105804, 0.014105066657066345, 0.26386529207229614, 0.2031843066215515, -0.09490790218114853, 0.18410980701446533, -0.02212667278945446, 0.26804283261299133, 0.22520719468593597, 0.0337846465408802, -0.08329089730978012, 0.2344714105129242, 0.14794911444187164, 0.07315370440483093, -0.027881748974323273, -0.3805592954158783, 0.07879495620727539, -0.08251582086086273, 0.11951687932014465, 0.30150672793388367, 0.006075269542634487, 0.6325334310531616, 0.08441071212291718, 0.3359811007976532, 0.2805604040622711, -0.10613861680030823, 0.08464629203081131, -0.2193402200937271, 0.13918522000312805, 0.15871590375900269, 0.25185441970825195, -0.09183425456285477, -0.23387929797172546, 0.5279425978660583, 0.28383785486221313, -0.050370652228593826, 0.42784854769706726, -0.48821645975112915, -0.2839679718017578, 0.2870752811431885, 0.09380435943603516, 0.7551211714744568, -0.021680839359760284, 0.3053111433982849, 0.11571769416332245, -0.3540657162666321, 0.6180205345153809, -0.3570942282676697, 0.10530238598585129, -0.07974395155906677, -0.19833479821681976, 0.03314223885536194, -0.05398918315768242, -0.12255165725946426, 0.26841914653778076, -0.34920966625213623, 0.335441529750824, 0.4309313893318176, -0.2644554376602173, -0.045098140835762024, 0.5335498452186584, -0.02422489784657955, -0.3266540467739105, -0.020113375037908554, 0.07147333025932312, -0.30048394203186035, 0.06917276978492737, -0.14425423741340637, -0.24988864362239838, 0.21163970232009888, 0.17763064801692963, -0.25000855326652527, 0.05973845720291138, -0.03575151786208153, 0.1542450487613678, 0.21908296644687653, 0.02060939371585846, 0.2647174596786499, -0.2386094629764557, -0.07471178472042084, -0.16828171908855438, -0.1355242282152176, 0.0720594972372055, -0.19097557663917542, 0.04476761445403099, 0.11568400263786316, -0.09888505935668945, 0.5525389313697815, -0.07500897347927094, -0.10550448298454285, -0.2602429986000061, 0.25004419684410095, -0.4098776578903198, -0.21422843635082245, 0.020000256597995758, 0.1741640418767929, 0.1515708863735199, -0.35633909702301025, 0.5705783367156982, 0.06015283986926079, 0.05220331624150276, 0.052962057292461395, 0.16527235507965088, -0.2551437020301819, 0.4723822772502899, -0.24417735636234283, 0.006947994232177734, 0.05782755836844444, 0.009747333824634552, -0.102012999355793, -0.3523838520050049, -0.06964657455682755, 0.31335726380348206, 0.09305255115032196, -0.042641185224056244, 0.03194166719913483, 0.13272316753864288, -0.11251083016395569, 0.019939225167036057, -0.1706300973892212, -0.36915016174316406, 0.14089323580265045, 0.24945934116840363, 0.2711485028266907, 0.18250614404678345, -0.11808463931083679, -0.20866364240646362, -0.10900400578975677, 0.5103971362113953, 0.0016739293932914734, 0.2612941861152649, -0.13015809655189514, 0.3655264973640442, 0.42578956484794617, 0.4940814971923828, -0.2902446389198303, 0.03717299550771713, 0.037440225481987, 0.09641530364751816, 0.48279711604118347, -0.2636583149433136, 0.02178003638982773, 0.2817063629627228, -0.13145862519741058, 0.058572232723236084, -0.34479525685310364, -0.1682952344417572, -0.22006326913833618, 0.11335426568984985, 0.10334259271621704, -0.020654119551181793, 0.07225695997476578, -0.04305354505777359, -0.14204126596450806, -0.06888562440872192, 0.045750442892313004, -0.49925723671913147, -0.08053195476531982, 0.2202770709991455, -0.1371796429157257, -0.011609427630901337, 0.02034958079457283, 0.23941396176815033, -0.008112229406833649, -0.12586088478565216, 0.29131388664245605, 0.1330929398536682, 0.12393414229154587, -0.1924992948770523, -0.16767145693302155, 0.019878968596458435, 0.21255391836166382, 0.17324021458625793, -0.056935764849185944, 0.19535769522190094, -0.17024734616279602, -0.14223472774028778, 0.24023392796516418, 0.12741324305534363, 0.04383229836821556, -0.17915286123752594, 0.17719945311546326, 0.1416485458612442, -0.09756141155958176, 0.06519679725170135, 0.22627584636211395, 0.08455795794725418, 0.12077274918556213, 0.09019751846790314, 0.1219424307346344, 0.4357948303222656, -0.2544054388999939, -0.017903020605444908, 0.1419234275817871, -0.351421594619751, 0.1340951770544052, -0.04542107507586479, -0.07775237411260605, 0.0030962564051151276, 0.11714660376310349, -0.07418832182884216, 0.12427617609500885, -0.3563074469566345, 0.3523235321044922, 0.27313128113746643, -0.01798110082745552, 0.06044319272041321, 0.1776667684316635, 0.005221795290708542, -0.13915102183818817, 0.3545141816139221, 0.10290132462978363, 0.40711650252342224, 0.03838269039988518, 0.9214759469032288, -0.28261417150497437, -0.4910769760608673, 0.074270099401474, 0.4407668709754944, 0.013904735445976257, -0.20886003971099854, 0.06338049471378326, 0.0767066553235054, -0.349092036485672, 0.09109146893024445, 0.02814275398850441, -0.1617656946182251, 0.6003193855285645, -0.0926634818315506, 0.19700390100479126, -0.12050983309745789, -0.18638822436332703, 0.30416521430015564, 0.2193455994129181, -0.0848199725151062, 0.03866316005587578, 0.08103355765342712, -0.09659154713153839, -0.2082844078540802, 0.539487898349762, 0.1378851681947708, -0.03067024052143097, -0.4813683331012726, 0.02654212713241577, 0.1303127110004425, 0.16194486618041992, -0.042261797934770584, 0.1696631759405136, 0.42150866985321045, -0.1417672038078308, 0.29074108600616455, 0.0038774237036705017, -0.09474688023328781, -0.18484871089458466, 0.07130374759435654, -0.4407298266887665, -0.35878968238830566, 0.38067737221717834, 0.14892904460430145, -0.1816977560520172, 0.12254293262958527, 0.2738814353942871, -0.6083030104637146, -0.22651799023151398, 0.28876474499702454, -0.16612592339515686, -0.08401194959878922, -0.22621002793312073, 0.08732544630765915, 0.05002979189157486, 0.6215313673019409, 0.2386113405227661, 0.3424452245235443, -0.31244197487831116, -0.11013562232255936, -0.5452611446380615, -0.12234259396791458, 0.15473434329032898, -0.12682385742664337, -0.16537782549858093, 0.17525804042816162, -0.025400705635547638, 0.3293079137802124, -0.26487916707992554, 0.18954189121723175, 0.1383688747882843, 0.08010149747133255, -0.45095768570899963, 0.1427934616804123, 0.08015468716621399, 0.0685940533876419, 0.27189093828201294, -0.1098073422908783, -0.17113739252090454, 0.08312532305717468, -0.06507846713066101, -0.06606283038854599, -0.0689278393983841, 0.2686031460762024, 0.16191941499710083, 0.03138861805200577, 0.07718020677566528, 0.4355566203594208, 0.004979666322469711, -0.20629256963729858, 0.0015182457864284515, 0.0014965813606977463, -0.3515438437461853, -0.19137030839920044, -0.07274559140205383, 0.1683378964662552, -0.3356752097606659, 0.2971011996269226, -0.029855817556381226, -0.11843480169773102, 0.1227511614561081, 0.06191503256559372, -0.1948867291212082, -0.09245669096708298, 0.10430537909269333, 0.15145081281661987, 0.21517954766750336, 0.7912341356277466, -0.033916689455509186, 0.05124349892139435, -0.10069805383682251, -0.010347947478294373, 0.04032726585865021, -0.09624392539262772, -0.22648505866527557, -0.07546033710241318, 0.03443174064159393, 0.0869658961892128, 0.10204719007015228, -0.6406650543212891, -0.37819695472717285, 0.03704700991511345, -0.04062342643737793, -0.0003268904983997345, 0.7422401905059814, 0.09145697951316833, -0.1913265585899353, -0.0755397230386734, -0.32336387038230896, 0.008202241733670235, -0.29208534955978394, 0.1692119836807251, -0.3407365679740906 ]
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.
[ -0.09163219481706619, -0.4211600720882416, -0.04513248801231384, -0.08009868860244751, -0.2404671162366867, 0.059044964611530304, 0.1587897539138794, 0.05346059799194336, 0.30978649854660034, -0.15817712247371674, -0.23331528902053833, 0.3175208568572998, -0.1927580088376999, 0.35245880484580994, 0.42011570930480957, -0.4721607267856598, 0.027789317071437836, -0.20416474342346191, -0.3079902231693268, 0.35350745916366577, 0.0078004151582717896, 0.11438114196062088, -0.17745058238506317, 0.08323551714420319, -0.29803040623664856, -0.16597478091716766, -0.35576871037483215, 0.13368433713912964, 0.09495039284229279, 0.019253242760896683, 0.17773222923278809, 0.3762141764163971, -0.0357888862490654, 0.0648593083024025, -0.00011105541489087045, -0.11337066441774368, 0.21172308921813965, -0.08244490623474121, 0.09594912827014923, -0.04149681329727173, 0.14164206385612488, -0.38433295488357544, 0.0004229896003380418, -0.0986679345369339, 0.18146944046020508, 0.16644404828548431, 0.19007359445095062, 0.049618855118751526, 0.3385031223297119, -0.29403936862945557, 0.1220092922449112, 0.3003107011318207, -0.16042426228523254, 0.25943702459335327, -0.07904580980539322, -0.07974834740161896, 0.04912269115447998, 0.22365301847457886, 0.6150230169296265, -0.5491244196891785, -0.0011966824531555176, 0.1282312422990799, -0.0015522688627243042, -0.024032112210989, 0.1823764443397522, -0.04576140269637108, -0.3587004244327545, -0.24357569217681885, -0.3750145435333252, 0.23614278435707092, -0.11858702450990677, -0.15378594398498535, -0.2185506522655487, -0.5235364437103271, 0.02771851420402527, 0.016797097399830818, -0.2812066078186035, 0.02997349575161934, -0.16756169497966766, -0.04844861850142479, -0.16560259461402893, -0.12260453402996063, 0.013107195496559143, 0.15340232849121094, 0.3461077809333801, 0.5681386590003967, 0.06152692064642906, 0.16792187094688416, 0.40035420656204224, 0.0016654245555400848, -0.15272551774978638, -0.3198026120662689, 0.29201003909111023, 0.09851133823394775, -0.38790521025657654, -0.3371363580226898, 0.1489792913198471, -0.18864706158638, 0.24584227800369263, 0.15510164201259613, 0.10689498484134674, 0.08710213005542755, -0.11532724648714066, 0.27482134103775024, 0.3207060694694519, 0.0658683329820633, -0.06847185641527176, -0.11937612295150757, 0.027747048065066338, -0.08491230756044388, 0.21637919545173645, 0.12413544207811356, 0.2167028784751892, -0.02785976231098175, -0.4467775523662567, -0.14667266607284546, -0.17592653632164001, 0.12826527655124664, -0.17640042304992676, -0.524179995059967, -0.05227186158299446, -0.10353738069534302, 0.08943959325551987, -0.048955611884593964, -0.28327587246894836, 0.10214965045452118, -0.07685297727584839, 0.20829130709171295, -0.06274810433387756, 0.07493423670530319, -0.07943785935640335, 0.12596245110034943, -0.4375801086425781, -0.01497543416917324, 0.30172044038772583, 0.11206872761249542, 0.11711080372333527, 0.15991513431072235, -0.02833031862974167, 0.10989698767662048, 0.3795184791088104, -0.016750697046518326, 0.050976626574993134, -0.007164042443037033, 0.1398804932832718, -0.32601621747016907, -0.0488298200070858, 0.1761285662651062, -0.34295204281806946, -0.14655029773712158, -0.13508251309394836, -0.20721761882305145, 0.14267075061798096, 0.05246469005942345, -0.19705796241760254, -0.27266278862953186, -0.469783753156662, 0.7273021936416626, -0.06460843235254288, -0.043151721358299255, -0.13948577642440796, 0.27761369943618774, -0.2422705888748169, -0.09580087661743164, 0.19651944935321808, 0.09004785120487213, -0.008326750248670578, -0.3287340998649597, -0.09379439800977707, -0.1102161854505539, -0.03254228085279465, 0.26917797327041626, -0.17343416810035706, 0.2014375925064087, 0.048721395432949066, 0.20125935971736908, 0.2725919187068939, -0.467446893453598, -0.0660242959856987, 0.3148563504219055, -0.021598059684038162, 0.37245580554008484, 0.047157011926174164, 0.2215045690536499, -0.009119553491473198, -0.19109781086444855, 0.27087265253067017, 0.9994992613792419, -0.39096599817276, -0.0052987802773714066, -0.14486417174339294, -0.36753806471824646, 0.5181840658187866, 0.34996935725212097, 0.06550046056509018, -0.3326672911643982, -0.14244595170021057, -0.023778382688760757, 0.08096598088741302, 0.0770779699087143, 0.08081132173538208, -0.023395702242851257, -0.3496008515357971, 0.02572062239050865, -0.21601687371730804, -0.13175401091575623, -0.39827898144721985, -0.013193009421229362, -0.011377250775694847, 0.3362920582294464, 0.2629062533378601, -0.11397732049226761, 0.23644927144050598, -0.26847612857818604, -0.1313517987728119, -0.3461121916770935, 0.1502704918384552, -0.06126667559146881, 0.051079463213682175, -0.2666863799095154, -0.036756664514541626, 0.3218362331390381, 0.1138492226600647, -0.11466069519519806, -0.5048949718475342, 0.2798417806625366, -0.08512654155492783, 0.032010518014431, -0.015921875834465027, 0.6123453378677368, -0.13566911220550537, -0.09030590206384659, 0.18309861421585083, 0.3067482113838196, -0.32861700654029846, 0.2714981436729431, 0.3248351514339447, 0.2580251395702362, 0.1833747923374176, -0.02368352562189102, 0.11495939642190933, -0.03480041027069092, -0.03136957809329033, -0.20113030076026917, -0.10553111881017685, 0.4779738187789917, -0.17939849197864532, 0.40344473719596863, -0.027947911992669106, 0.02752644009888172, -0.19954337179660797, -0.036002691835165024, -0.22701188921928406, 0.367566853761673, 0.3015568256378174, -0.19315442442893982, 0.3859136998653412, -0.09095410257577896, -0.3488398790359497, 0.17814302444458008, 0.20084065198898315, 0.013798952102661133, 0.09303046762943268, -0.016065359115600586, -0.0603153333067894, -0.3105744421482086, -0.06443570554256439, 0.4159681797027588, 0.6467360854148865, 0.2788432538509369, 0.1465170532464981, 0.05391903221607208, -0.23704062402248383, -0.07217651605606079, 0.02709732949733734, 0.17971578240394592, -0.022057941183447838, 0.13218240439891815, 0.028972607105970383, 0.053669046610593796, -0.0197980348020792, -0.14596885442733765, 0.1391361653804779, -0.18290135264396667, -0.026788270100951195, 0.0350450798869133, -0.19006139039993286, -0.5519410967826843, -0.2311941236257553, 0.04605114087462425, -0.375975638628006, -0.3040531575679779, 0.12783879041671753, 0.025094876065850258, -0.19186356663703918, 0.1835845559835434, 0.2177857607603073, 0.5667122602462769, -0.4283854067325592, -0.25661733746528625, -0.10000119358301163, -0.25852108001708984, -0.10058089345693588, 0.12403920292854309, 0.41809219121932983, 0.30088603496551514, 0.4967927932739258, 0.28931277990341187, -0.014469848945736885, 0.020839190110564232, -0.3807588815689087, 0.04763735085725784, -0.288138210773468, 0.2660779356956482, 0.10908018052577972, -0.23446260392665863, 0.1544487029314041, -0.4112611711025238, -0.02723073586821556, -0.06777597963809967, -0.05139286071062088, -0.049117833375930786, -0.015199120156466961, 0.0207084771245718, -0.17686575651168823, -0.22578808665275574, -0.6239805221557617, -0.4909670948982239, 0.28933680057525635, -0.28130975365638733, 0.0613841712474823, 0.016535308212041855, -0.26013949513435364, 0.19389955699443817, -0.15522347390651703, 0.03759928047657013, -0.17455731332302094, -0.10408574342727661, 0.04657413065433502, -0.3616589605808258, -0.16067299246788025, -0.41546830534935, -0.11438252776861191, 0.23246780037879944, -0.08507470786571503, -0.0425742045044899, -0.23878808319568634, 0.048094674944877625, -0.12757475674152374, 0.09415815770626068, 0.2228919416666031, 0.1751595139503479, -0.22287091612815857, 0.09907601773738861, -0.036777425557374954, -0.26462864875793457, 0.1948915719985962, 0.33807286620140076, 0.42725706100463867, 0.17781081795692444, 0.01658076047897339, 0.17279289662837982, 0.7245492935180664, 0.25919094681739807, 0.05576366186141968, 0.15051709115505219, 0.3070625364780426, 0.03356906771659851, 0.14654755592346191, -0.23375555872917175, 0.24164286255836487, -0.16472789645195007, 0.5056428909301758, 0.09843314439058304, 0.1719014048576355, -0.2043481171131134, -0.28388190269470215, -0.012782398611307144, -0.2579216957092285, -0.18797719478607178, 0.2683839499950409, -0.22223517298698425, 0.020278356969356537, -0.055379871279001236, -0.3724345564842224, -0.25288158655166626, -0.04279504343867302, 0.017948660999536514, 0.31589123606681824, -0.030992522835731506, 0.15221890807151794, -0.34515342116355896, -0.0411997027695179, -0.4830809235572815, 0.15105335414409637, 0.08815167844295502, 0.27570006251335144, -0.08164797723293304, -0.0731469988822937, -0.1406986564397812, 0.10453666746616364, 0.5943602323532104, -0.5221483111381531, -0.16565899550914764, 0.09090529382228851, 0.08042608946561813, -0.09018901735544205, -0.08775423467159271, -0.2720241844654083, -0.04363056644797325, 0.4483170211315155, 0.36958035826683044, -0.3692058026790619, -0.3136216998100281, 0.2565195858478546, -0.18059846758842468, -0.03166762739419937, -0.19901442527770996, -0.18238741159439087, -0.18923181295394897, -0.14206206798553467, 0.03215467929840088, 0.1843511015176773, 0.2113642543554306, -0.10405097156763077, 0.17735259234905243, -0.05196564644575119, 0.2529606819152832, 0.20239824056625366, 0.01307758316397667, -0.040300264954566956, 0.22695446014404297, 0.15439800918102264, 0.031860142946243286, -0.12105235457420349, -0.29360511898994446, 0.09633069485425949, -0.06996095180511475, 0.11004846543073654, 0.271496057510376, -0.00305559067055583, 0.6472855806350708, 0.05717632919549942, 0.31057435274124146, 0.2889767289161682, -0.12562023103237152, 0.05900801718235016, -0.23012974858283997, 0.14177188277244568, 0.22950918972492218, 0.25832289457321167, -0.13930945098400116, -0.19662606716156006, 0.5935574173927307, 0.3162984549999237, -0.07326257973909378, 0.38970157504081726, -0.46200162172317505, -0.2712932527065277, 0.2523413896560669, 0.027918927371501923, 0.6917436122894287, -0.025947771966457367, 0.2671622931957245, 0.06503260135650635, -0.384538471698761, 0.6288003325462341, -0.3909117579460144, 0.1575741320848465, -0.07772613316774368, -0.23540401458740234, 0.0326017290353775, -0.09457425773143768, -0.06971939653158188, 0.26830774545669556, -0.3717164695262909, 0.34969374537467957, 0.4773365259170532, -0.20441070199012756, -0.05118029564619064, 0.5518015027046204, -0.029192546382546425, -0.400279700756073, -0.10723662376403809, 0.08906093239784241, -0.28121334314346313, 0.10603442788124084, -0.16336649656295776, -0.27637630701065063, 0.24864256381988525, 0.11350046098232269, -0.27409273386001587, 0.11336158961057663, -0.03516446799039841, 0.1855698674917221, 0.2648400366306305, 0.016313955187797546, 0.25120601058006287, -0.17889165878295898, -0.0666220486164093, -0.1549469530582428, -0.0886358767747879, 0.1312912553548813, -0.14661234617233276, 0.05025830119848251, 0.030554739758372307, -0.2045278698205948, 0.5105484127998352, -0.07615140080451965, -0.10789109766483307, -0.24378013610839844, 0.24476522207260132, -0.43671631813049316, -0.1542508751153946, 0.11010661721229553, 0.16822589933872223, 0.2132159322500229, -0.31623944640159607, 0.6001381874084473, 0.08641384541988373, -0.009584203362464905, 0.055802345275878906, 0.1773013174533844, -0.23682695627212524, 0.4851786196231842, -0.25789588689804077, -0.004610419273376465, 0.052440494298934937, 0.010878428816795349, -0.1622539460659027, -0.37275445461273193, -0.03353308141231537, 0.2981380224227905, 0.08636647462844849, -0.016164803877472878, -0.0029818490147590637, 0.1021042987704277, -0.07975640147924423, -0.003008240833878517, -0.15544848144054413, -0.3322067856788635, 0.1165371909737587, 0.14724498987197876, 0.19470559060573578, 0.16701370477676392, -0.053953059017658234, -0.24001404643058777, -0.18599310517311096, 0.46923357248306274, 0.04062017798423767, 0.2234199494123459, -0.13615617156028748, 0.3405207097530365, 0.42906028032302856, 0.5344385504722595, -0.28160953521728516, 0.07217328995466232, 0.032790035009384155, 0.08643410354852676, 0.4831205904483795, -0.2349259853363037, 0.07466655224561691, 0.2786791920661926, -0.11924923956394196, 0.0491994246840477, -0.34023600816726685, -0.1727285087108612, -0.2547360062599182, 0.09835311770439148, 0.11223381012678146, -0.0018388675525784492, 0.0054129986092448235, -0.06127312406897545, -0.07910172641277313, -0.08158965408802032, 0.016571901738643646, -0.3894309103488922, -0.06932204961776733, 0.15564624965190887, -0.11541067063808441, 0.015997931361198425, 0.059237927198410034, 0.20302146673202515, 0.009794354438781738, -0.0524289608001709, 0.2911641597747803, 0.12709856033325195, 0.11118440330028534, -0.1909799873828888, -0.17631687223911285, 0.004691518843173981, 0.21382761001586914, 0.14815178513526917, -0.02907838672399521, 0.1604178547859192, -0.23436830937862396, -0.1400955319404602, 0.2342926561832428, 0.1486867070198059, 0.020620571449398994, -0.12043088674545288, 0.1056559756398201, 0.14062277972698212, -0.12489896267652512, -0.02853161096572876, 0.17814530432224274, 0.11084794253110886, 0.1366320550441742, 0.11059849709272385, 0.15870606899261475, 0.44991493225097656, -0.21001620590686798, 0.025973491370677948, 0.11526631563901901, -0.3685958981513977, 0.12917537987232208, -0.11547407507896423, -0.040456511080265045, 0.005308408290147781, 0.1474834382534027, -0.0111294686794281, 0.13959282636642456, -0.32164716720581055, 0.30607354640960693, 0.30969566106796265, -0.023712676018476486, 0.05864635109901428, 0.23911476135253906, 0.01480795443058014, -0.13310979306697845, 0.30467915534973145, 0.01857598125934601, 0.3192398250102997, 0.042772985994815826, 0.8754275441169739, -0.272227019071579, -0.45505988597869873, 0.10648228973150253, 0.43658554553985596, 0.03211100772023201, -0.21552914381027222, 0.15521663427352905, 0.09118608385324478, -0.40597593784332275, 0.1209784597158432, 0.05810922011733055, -0.1637459695339203, 0.657473087310791, -0.11649444699287415, 0.16466790437698364, -0.15811486542224884, -0.14994096755981445, 0.3523966372013092, 0.19673556089401245, -0.09543836116790771, 0.010715967044234276, 0.11992718279361725, -0.11069288849830627, -0.16985253989696503, 0.5139833688735962, 0.2285723239183426, 0.034792959690093994, -0.4124753475189209, 0.03735283389687538, 0.107136070728302, 0.14290352165699005, -0.07321160286664963, 0.138036847114563, 0.44498148560523987, -0.18126782774925232, 0.2542766332626343, 0.036787159740924835, -0.07882071286439896, -0.13540196418762207, 0.043383754789829254, -0.38847362995147705, -0.39733782410621643, 0.4565662145614624, 0.10967668145895004, -0.11010195314884186, 0.04526514559984207, 0.3206997513771057, -0.6264177560806274, -0.2374812215566635, 0.3614809215068817, -0.1362902969121933, -0.0500934012234211, -0.24483497440814972, 0.08715891093015671, 0.08407621085643768, 0.6122646927833557, 0.21460853517055511, 0.3534941077232361, -0.4017142355442047, -0.14038477838039398, -0.5742546916007996, -0.21603280305862427, 0.17194154858589172, -0.07661580294370651, -0.2584016025066376, 0.14676685631275177, -0.06859225779771805, 0.28950589895248413, -0.21401123702526093, 0.1178646832704544, 0.17773322761058807, 0.02705768123269081, -0.430539071559906, 0.11578315496444702, 0.058791302144527435, 0.004421898163855076, 0.31976118683815, -0.11052559316158295, -0.1742173284292221, 0.08242564648389816, -0.05696408078074455, -0.09089021384716034, -0.06792822480201721, 0.23781514167785645, 0.16779708862304688, 0.09648948162794113, 0.08136258274316788, 0.4660528302192688, -0.04932480677962303, -0.2016535997390747, -0.008784513920545578, 0.012513291090726852, -0.37709811329841614, -0.1661680042743683, -0.07718321681022644, 0.1801460236310959, -0.33483511209487915, 0.25497040152549744, -0.016634495928883553, -0.08482318371534348, 0.11921893060207367, 0.01841846853494644, -0.1893487125635147, -0.04945801943540573, 0.07420556992292404, 0.195068821310997, 0.22280582785606384, 0.7319666147232056, -0.01056225597858429, 0.08575521409511566, -0.09298598021268845, -0.03421401232481003, 0.07086607813835144, -0.052111849188804626, -0.20773279666900635, -0.13613852858543396, 0.02423723042011261, 0.09727302193641663, 0.13568156957626343, -0.6967558860778809, -0.33582890033721924, 0.014331405982375145, -0.022863823920488358, 0.03177652135491371, 0.7183234095573425, 0.06665454804897308, -0.14486512541770935, -0.11753140389919281, -0.2986988425254822, 0.053424231708049774, -0.2820817530155182, 0.18920615315437317, -0.34770846366882324 ]
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).
[ -0.07760576903820038, -0.4301373362541199, -0.05128105729818344, -0.007647901773452759, -0.24152879416942596, 0.11484136432409286, 0.1743364930152893, 0.09230956435203552, 0.28521808981895447, -0.11796191334724426, -0.24373745918273926, 0.27350884675979614, -0.19464346766471863, 0.41696271300315857, 0.39819997549057007, -0.45603516697883606, 0.007997095584869385, -0.21966683864593506, -0.2606225907802582, 0.37996965646743774, 0.04182820022106171, 0.09555917233228683, -0.14734555780887604, 0.06278294324874878, -0.2902207374572754, -0.1371721774339676, -0.34358692169189453, 0.08050669729709625, 0.10446936637163162, 0.0340154692530632, 0.19900573790073395, 0.37377867102622986, 0.00899704359471798, 0.10891927778720856, -0.00011252853437326849, -0.09093979001045227, 0.16050821542739868, -0.0712214782834053, 0.1302853673696518, -0.023051872849464417, 0.12360096722841263, -0.35964643955230713, 0.06394293904304504, -0.10716303437948227, 0.18151751160621643, 0.18675681948661804, 0.20179066061973572, 0.08393155038356781, 0.3181023597717285, -0.3155003786087036, 0.0999816283583641, 0.29038479924201965, -0.18832486867904663, 0.280779629945755, -0.05406240373849869, -0.0642036646604538, 0.03068152628839016, 0.22856448590755463, 0.6673146486282349, -0.5304582118988037, 0.04461165517568588, 0.06839239597320557, -0.02713405340909958, 0.04156407713890076, 0.11788807809352875, -0.08518808335065842, -0.30709508061408997, -0.239467591047287, -0.37350428104400635, 0.28026336431503296, -0.07192579656839371, -0.10643924027681351, -0.23320834338665009, -0.5333685874938965, 0.01573336496949196, 0.07696226239204407, -0.2952456772327423, 0.05960957705974579, -0.1627965122461319, -0.019170589745044708, -0.12913011014461517, -0.12187917530536652, -0.001456335186958313, 0.1736183613538742, 0.3086414337158203, 0.5555930137634277, 0.06791813671588898, 0.16606809198856354, 0.3354538083076477, 0.06970074027776718, -0.09848681837320328, -0.2893276810646057, 0.2836745083332062, 0.06580854207277298, -0.4028254747390747, -0.35675525665283203, 0.13010713458061218, -0.26543211936950684, 0.28047773241996765, 0.13964121043682098, 0.16255822777748108, 0.09564383327960968, -0.1668533980846405, 0.2518714666366577, 0.33976078033447266, -0.02247934229671955, -0.12287168949842453, -0.13603001832962036, 0.017504721879959106, -0.10887739807367325, 0.16654373705387115, 0.13609996438026428, 0.17564214766025543, -0.0186630729585886, -0.4268651306629181, -0.11984290927648544, -0.14647668600082397, 0.0982045829296112, -0.18795841932296753, -0.6031383275985718, -0.09075364470481873, -0.10779067128896713, 0.03100643679499626, -0.15765944123268127, -0.2877180278301239, 0.092652827501297, -0.10346348583698273, 0.24397854506969452, -0.13179226219654083, 0.00854057352989912, -0.048577066510915756, 0.06551392376422882, -0.423928439617157, -0.04849141836166382, 0.29413989186286926, 0.09935413300991058, 0.04831109195947647, 0.16498593986034393, -0.03427750617265701, 0.12868380546569824, 0.42059457302093506, 0.029048018157482147, 0.014121629297733307, -0.0367378331720829, 0.17303933203220367, -0.3192979693412781, -0.011867158114910126, 0.2226768583059311, -0.37173980474472046, -0.11837560683488846, -0.16984781622886658, -0.2173645943403244, 0.09760744869709015, 0.03343378007411957, -0.21831941604614258, -0.3184994161128998, -0.4913887679576874, 0.7129409909248352, -0.044741190969944, -0.04646164923906326, -0.16467785835266113, 0.2666395902633667, -0.2322615534067154, -0.09799232333898544, 0.16326361894607544, 0.08497167378664017, 0.013150323182344437, -0.2949587106704712, -0.09312959760427475, -0.10383567214012146, -0.06481081992387772, 0.30172601342201233, -0.16893525421619415, 0.21227431297302246, 0.06585872173309326, 0.13697415590286255, 0.2494572103023529, -0.4536503851413727, -0.048898711800575256, 0.33244845271110535, -0.02496878057718277, 0.35705429315567017, 0.07683297246694565, 0.2123105525970459, -0.0030691931024193764, -0.2040671855211258, 0.19492629170417786, 1.017795205116272, -0.3952714800834656, -0.021733975037932396, -0.11022846400737762, -0.39061519503593445, 0.5425779819488525, 0.36107009649276733, 0.06735080480575562, -0.3310126066207886, -0.13180634379386902, -0.008706456050276756, 0.09087163209915161, 0.07975569367408752, 0.04832107573747635, -0.0628935918211937, -0.34346044063568115, 0.04504135251045227, -0.24580256640911102, -0.19732028245925903, -0.4192027747631073, -0.020278949290513992, 0.016957003623247147, 0.29230332374572754, 0.3340206742286682, -0.10002608597278595, 0.27135199308395386, -0.3039945662021637, -0.1537623554468155, -0.31332990527153015, 0.11975884437561035, -0.005619511008262634, 0.03799665346741676, -0.24834029376506805, -0.030674505978822708, 0.31918051838874817, 0.09926745295524597, -0.13511954247951508, -0.4337626099586487, 0.33981865644454956, -0.047643229365348816, 0.04676056280732155, -0.04639905318617821, 0.6206656098365784, -0.19886362552642822, -0.09158611297607422, 0.1701807677745819, 0.24870342016220093, -0.3234960436820984, 0.2560337483882904, 0.32414016127586365, 0.35003429651260376, 0.19196629524230957, -0.08836960047483444, 0.08618822693824768, -0.012060839682817459, -0.05773837864398956, -0.19454273581504822, -0.08406423777341843, 0.4658960700035095, -0.18547146022319794, 0.41023844480514526, -0.05356256663799286, 0.014754710718989372, -0.19410167634487152, -0.05506932735443115, -0.18381235003471375, 0.4107012152671814, 0.3079740107059479, -0.1972614824771881, 0.3365919291973114, -0.11086127161979675, -0.36304888129234314, 0.1967424750328064, 0.18514330685138702, 0.056777406483888626, 0.07615286111831665, -0.03248913586139679, -0.0191662535071373, -0.29046300053596497, -0.11108554899692535, 0.40457239747047424, 0.632398784160614, 0.28050968050956726, 0.11706024408340454, -0.021244825795292854, -0.25007444620132446, -0.10282351821660995, 0.011356614530086517, 0.20606467127799988, -0.016843674704432487, 0.15028434991836548, 0.04689302667975426, 0.07360973954200745, 0.043971844017505646, -0.14593511819839478, 0.1934909075498581, -0.1794009506702423, -0.04956897348165512, 0.024611497297883034, -0.1928529292345047, -0.5652398467063904, -0.21911503374576569, 0.0437881201505661, -0.3279030919075012, -0.3013741374015808, 0.15034149587154388, 0.05334058031439781, -0.176954984664917, 0.15528973937034607, 0.174662783741951, 0.5224665999412537, -0.44439470767974854, -0.2338436096906662, -0.06413104385137558, -0.28284752368927, -0.08073657006025314, 0.11607673764228821, 0.49301397800445557, 0.2858504056930542, 0.45890358090400696, 0.2643522024154663, -0.04553256928920746, 0.0037953639402985573, -0.3939478397369385, 0.029011324048042297, -0.30691754817962646, 0.2791736125946045, 0.0830259770154953, -0.21713970601558685, 0.15055492520332336, -0.39499631524086, -0.06443783640861511, -0.04882184416055679, -0.07792936265468597, -0.028692929074168205, -0.025047870352864265, 0.0897224098443985, -0.17804251611232758, -0.2269406020641327, -0.6084065437316895, -0.4707542657852173, 0.2826777398586273, -0.28210893273353577, 0.07274039834737778, 0.038764555007219315, -0.27793413400650024, 0.15081928670406342, -0.15593431890010834, 0.028028899803757668, -0.15452151000499725, -0.11299090832471848, -0.01863935962319374, -0.3668549358844757, -0.15962430834770203, -0.41231536865234375, -0.07378329336643219, 0.22635772824287415, -0.07291725277900696, -0.0498352125287056, -0.2201995700597763, 0.06068301200866699, -0.10324518382549286, 0.0564827024936676, 0.17424608767032623, 0.1718524992465973, -0.22455747425556183, 0.11065275967121124, -0.04123270884156227, -0.20303326845169067, 0.20850899815559387, 0.36925259232521057, 0.3985558748245239, 0.17247524857521057, -0.04534270614385605, 0.15867747366428375, 0.6988726258277893, 0.22570152580738068, 0.07469471544027328, 0.1957734227180481, 0.300357460975647, -0.008464980870485306, 0.14122040569782257, -0.233286514878273, 0.21620109677314758, -0.13694700598716736, 0.49532997608184814, 0.11535324156284332, 0.1564604938030243, -0.17848846316337585, -0.2667303681373596, 0.05356660857796669, -0.2868148386478424, -0.18034493923187256, 0.23250894248485565, -0.2959993779659271, 0.0018163472414016724, -0.06489185988903046, -0.46274006366729736, -0.28313708305358887, -0.04907607659697533, -0.028592389076948166, 0.3023619055747986, -0.01832866668701172, 0.16230705380439758, -0.3587011694908142, -0.02396807074546814, -0.359493225812912, 0.16688692569732666, 0.09561147540807724, 0.25671476125717163, -0.08165299892425537, -0.045284949243068695, -0.07541713118553162, 0.13249599933624268, 0.594111442565918, -0.5726907253265381, -0.11468027532100677, 0.11226196587085724, 0.06943963468074799, -0.10702905058860779, -0.11246588826179504, -0.25311192870140076, -0.022778909653425217, 0.41089943051338196, 0.3724684417247772, -0.3512176275253296, -0.3359483778476715, 0.27706092596054077, -0.1780259758234024, -0.06660392880439758, -0.1773485690355301, -0.1805456429719925, -0.19922927021980286, -0.11174333840608597, -0.015011996030807495, 0.22503185272216797, 0.23601852357387543, -0.09397950023412704, 0.1691284477710724, -0.11902868747711182, 0.23430871963500977, 0.17777666449546814, -0.020016271620988846, -0.08604373782873154, 0.21967238187789917, 0.19012929499149323, 0.04374058172106743, -0.12046508491039276, -0.2854847311973572, 0.0902773067355156, -0.08531789481639862, 0.04106679558753967, 0.32314562797546387, -0.01695924811065197, 0.6543562412261963, 0.0035463571548461914, 0.35406139492988586, 0.29701897501945496, -0.15183007717132568, 0.035120029002428055, -0.22525565326213837, 0.1592191606760025, 0.1932837814092636, 0.2645816504955292, -0.1670910269021988, -0.2223948836326599, 0.5615092515945435, 0.3239040970802307, -0.04171903058886528, 0.3951147198677063, -0.5345005989074707, -0.27463921904563904, 0.27251511812210083, 0.029302049428224564, 0.6709376573562622, -0.08567734062671661, 0.2815999984741211, 0.012769563123583794, -0.4185329079627991, 0.6093618273735046, -0.356319785118103, 0.14552359282970428, -0.07994046807289124, -0.2315620630979538, 0.021089831367135048, -0.07193601876497269, -0.09989524632692337, 0.2743527889251709, -0.34961992502212524, 0.3014124631881714, 0.44390183687210083, -0.18051117658615112, -0.03987343609333038, 0.5531432628631592, -0.015312885865569115, -0.357075572013855, -0.03860893100500107, 0.06670358777046204, -0.2900908589363098, 0.07109349220991135, -0.1547674834728241, -0.25153255462646484, 0.24413558840751648, 0.1927369385957718, -0.3443204164505005, 0.06983192265033722, 0.018976649269461632, 0.18553580343723297, 0.22446368634700775, -0.00212039053440094, 0.296608030796051, -0.16361579298973083, -0.031091906130313873, -0.1591515690088272, -0.11198890209197998, 0.1209486573934555, -0.18845388293266296, 0.004920526873320341, 0.04263247922062874, -0.1961912214756012, 0.5082728266716003, -0.05026669800281525, -0.17390188574790955, -0.2432452142238617, 0.2771672308444977, -0.4623684883117676, -0.23436494171619415, 0.12246377766132355, 0.19409795105457306, 0.218156099319458, -0.31469637155532837, 0.5614801645278931, 0.127997025847435, -0.011074503883719444, 0.044554583728313446, 0.134901762008667, -0.20838184654712677, 0.47530749440193176, -0.2853441834449768, -0.02586296945810318, 0.06688466668128967, -0.004448145627975464, -0.09673929959535599, -0.3717647194862366, -0.04464828968048096, 0.317708283662796, 0.16569329798221588, -0.028410455211997032, 0.07400874048471451, 0.13821393251419067, -0.0919293612241745, -0.007477550767362118, -0.16431035101413727, -0.2981984615325928, 0.17177724838256836, 0.1826208084821701, 0.2225337028503418, 0.18226583302021027, -0.07159329205751419, -0.23599836230278015, -0.122200608253479, 0.4769054651260376, -0.008985381573438644, 0.24603387713432312, -0.13987493515014648, 0.3845987617969513, 0.4304969310760498, 0.5406715273857117, -0.25706756114959717, 0.12530672550201416, 0.04334171861410141, 0.08136966824531555, 0.5019499659538269, -0.2255435287952423, 0.03800460323691368, 0.27418333292007446, -0.1421244889497757, 0.04171761870384216, -0.2888588309288025, -0.16853713989257812, -0.21639379858970642, 0.11692989617586136, 0.1269606649875641, 0.02414373867213726, 0.03174247592687607, -0.05489098280668259, -0.1292056441307068, -0.05813824385404587, -0.010787772946059704, -0.37812095880508423, -0.04661107808351517, 0.13476987183094025, -0.062149014323949814, 0.04813924431800842, 0.045553840696811676, 0.24881413578987122, -0.005483634769916534, -0.04891292750835419, 0.3026910424232483, 0.1137172058224678, 0.13134883344173431, -0.19243137538433075, -0.22584998607635498, -0.02276790142059326, 0.27811670303344727, 0.16005298495292664, -0.08049098402261734, 0.15094342827796936, -0.21033044159412384, -0.1457175314426422, 0.1973719596862793, 0.10546581447124481, 0.06881042569875717, -0.08835802972316742, 0.1466383934020996, 0.10758335143327713, -0.0755477249622345, 0.013402312994003296, 0.1942145824432373, 0.07465550303459167, 0.07250766456127167, 0.10837576538324356, 0.1365595906972885, 0.4735901355743408, -0.2260930985212326, 0.003198942169547081, 0.12325045466423035, -0.3594176769256592, 0.15418648719787598, -0.08749888837337494, -0.06692662090063095, -0.03102230653166771, 0.1314779371023178, -0.059005070477724075, 0.13142891228199005, -0.3177986741065979, 0.31435197591781616, 0.2902788519859314, -0.016615577042102814, -0.023868408054113388, 0.24449306726455688, 0.03816736862063408, -0.14794549345970154, 0.2975652813911438, 0.04241354390978813, 0.3625638484954834, 0.03182324767112732, 0.8982052206993103, -0.28706395626068115, -0.4490219056606293, 0.08506786078214645, 0.4442313611507416, 0.09313870966434479, -0.21840958297252655, 0.12272955477237701, 0.10584568232297897, -0.3336717188358307, 0.12666091322898865, 0.027430633082985878, -0.1679341197013855, 0.6969411373138428, -0.08871020376682281, 0.13946983218193054, -0.15565209090709686, -0.16006094217300415, 0.37912964820861816, 0.15306229889392853, -0.08720095455646515, 0.005124622955918312, 0.11803263425827026, -0.11495539546012878, -0.1713058054447174, 0.5400869250297546, 0.19663335382938385, -0.016986824572086334, -0.44110897183418274, -0.002918851561844349, 0.09282286465167999, 0.18584536015987396, -0.003108755685389042, 0.11664828658103943, 0.42755234241485596, -0.17765967547893524, 0.2757040858268738, 0.016501324251294136, -0.042451731860637665, -0.12721548974514008, 0.04194376617670059, -0.39963504672050476, -0.44053223729133606, 0.43349674344062805, 0.1266334503889084, -0.10133137553930283, 0.12239136546850204, 0.34221863746643066, -0.625470757484436, -0.21663399040699005, 0.32839295268058777, -0.13515399396419525, -0.09978494048118591, -0.21278639137744904, 0.08299478888511658, 0.05206345394253731, 0.6268002986907959, 0.26694661378860474, 0.35594016313552856, -0.367862343788147, -0.10077333450317383, -0.5913139581680298, -0.1538389027118683, 0.15314844250679016, -0.07130032777786255, -0.22856467962265015, 0.15664465725421906, -0.04573541134595871, 0.27950361371040344, -0.16623638570308685, 0.15578556060791016, 0.22006641328334808, 0.0632854551076889, -0.4629290997982025, 0.05816925689578056, 0.15672540664672852, 0.0008714217692613602, 0.26979321241378784, -0.09562158584594727, -0.18361014127731323, 0.10721481591463089, -0.08659321069717407, -0.13303770124912262, -0.08308319747447968, 0.2785082459449768, 0.1443416178226471, 0.1257953643798828, 0.060521967709064484, 0.48311856389045715, -0.013349970802664757, -0.19535678625106812, 0.00032370537519454956, -0.017219023779034615, -0.36596015095710754, -0.133477121591568, -0.11219032108783722, 0.16162464022636414, -0.40695860981941223, 0.3112413287162781, 0.020267846062779427, -0.11749889701604843, 0.07889258116483688, 0.08205994963645935, -0.19863159954547882, -0.1077202707529068, 0.0925135388970375, 0.1740230917930603, 0.22926753759384155, 0.7064047455787659, -0.009655654430389404, 0.08157234638929367, -0.08162674307823181, -0.024642569944262505, 0.013805702328681946, 0.001977735199034214, -0.23659107089042664, -0.09722098708152771, 0.035987578332424164, 0.04707097262144089, 0.14540337026119232, -0.6879175305366516, -0.3390010893344879, 0.07124495506286621, 0.008100252598524094, 0.060267698019742966, 0.7683095335960388, 0.13407115638256073, -0.14332085847854614, -0.10919389128684998, -0.306508332490921, 0.028837325051426888, -0.27584418654441833, 0.18561233580112457, -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.
[ -0.08985172212123871, -0.3969329297542572, -0.031665366142988205, -0.037351638078689575, -0.2264488786458969, 0.08738168329000473, 0.196954607963562, 0.06602276861667633, 0.33273035287857056, -0.13962094485759735, -0.22614936530590057, 0.3318239152431488, -0.2025051862001419, 0.36451777815818787, 0.4015558958053589, -0.5019195079803467, -0.009031489491462708, -0.173209547996521, -0.3404366970062256, 0.34348368644714355, 0.01590494066476822, 0.08975976705551147, -0.1299690455198288, 0.053992584347724915, -0.2659047842025757, -0.14809374511241913, -0.3480379581451416, 0.10398368537425995, 0.09025157988071442, 0.00021010730415582657, 0.26544636487960815, 0.3834899067878723, -0.004184707999229431, 0.11468689143657684, -0.00011099899711553007, -0.10409414768218994, 0.17799219489097595, -0.061887532472610474, 0.13401706516742706, -0.022732838988304138, 0.20591771602630615, -0.374109148979187, 0.041294753551483154, -0.14299587905406952, 0.09510346502065659, 0.18651211261749268, 0.14864693582057953, 0.06131582707166672, 0.4389665126800537, -0.32718124985694885, 0.1137162297964096, 0.27153220772743225, -0.21278294920921326, 0.21606014668941498, -0.10353533178567886, -0.08171606063842773, 0.03618605434894562, 0.294840008020401, 0.6407874822616577, -0.525549590587616, -0.026187747716903687, 0.09548534452915192, -0.02736218273639679, -0.026917368173599243, 0.16530895233154297, -0.05579408258199692, -0.3531099855899811, -0.2346816062927246, -0.3933756947517395, 0.26181381940841675, -0.19633162021636963, -0.1110704317688942, -0.22825053334236145, -0.5546967387199402, 0.04327018931508064, 0.05650652199983597, -0.32799801230430603, 0.04984400421380997, -0.16306138038635254, 0.01506592333316803, -0.16031597554683685, -0.1137707456946373, -0.03513527661561966, 0.15094447135925293, 0.3342221677303314, 0.5391837358474731, 0.07172629982233047, 0.18103159964084625, 0.37623268365859985, 0.06643140316009521, -0.15596938133239746, -0.25584715604782104, 0.26040059328079224, 0.1465936303138733, -0.30014485120773315, -0.3815108835697174, 0.10290298610925674, -0.20207494497299194, 0.26482439041137695, 0.16467323899269104, 0.09347937256097794, 0.1269809901714325, -0.11747140437364578, 0.2820117175579071, 0.3259006440639496, -0.0005844989791512489, -0.05843723565340042, -0.08932635933160782, 0.03309009224176407, -0.08139177411794662, 0.16326531767845154, 0.12141101807355881, 0.1739644706249237, 0.06176479905843735, -0.4309346079826355, -0.12424077838659286, -0.20832158625125885, 0.14269258081912994, -0.12683850526809692, -0.5166333913803101, -0.06742709875106812, -0.08390922099351883, 0.06903015822172165, -0.08924680203199387, -0.3010077476501465, 0.11446747928857803, -0.022935112938284874, 0.21466878056526184, -0.06969944387674332, 0.014183173887431622, -0.05495689436793327, 0.11815536022186279, -0.4216558337211609, -0.016975730657577515, 0.3122232258319855, 0.09955611824989319, 0.09712179005146027, 0.1739337146282196, -0.07775558531284332, 0.12298354506492615, 0.42270323634147644, 0.02999717742204666, 0.02881624549627304, -0.03779754415154457, 0.14484578371047974, -0.32560503482818604, -0.00640280544757843, 0.18319693207740784, -0.36248600482940674, -0.15413984656333923, -0.09662603586912155, -0.16510260105133057, 0.12379004806280136, 0.05943754315376282, -0.225196972489357, -0.278982937335968, -0.37818217277526855, 0.7292916178703308, -0.0695025846362114, -0.0614316388964653, -0.17215341329574585, 0.29769450426101685, -0.25143396854400635, -0.07137693464756012, 0.18124361336231232, 0.07104992866516113, -0.032820604741573334, -0.38990500569343567, -0.10039550811052322, -0.10813398659229279, -0.07206135243177414, 0.2687499225139618, -0.20015178620815277, 0.21656769514083862, 0.04004433751106262, 0.16750960052013397, 0.25886863470077515, -0.40507903695106506, -0.10159724950790405, 0.28701913356781006, -0.037553705275058746, 0.3858807384967804, 0.0768330916762352, 0.24547545611858368, 0.08919303864240646, -0.17639228701591492, 0.2526644170284271, 1.007195234298706, -0.3976956903934479, -0.06931684166193008, -0.10972339659929276, -0.3863373100757599, 0.5040086507797241, 0.3668290972709656, 0.10830037295818329, -0.23684129118919373, -0.1491307020187378, -0.05921328812837601, 0.06534458696842194, 0.04500487446784973, 0.10629874467849731, -0.08196783810853958, -0.34277451038360596, 0.0306718572974205, -0.21699601411819458, -0.20223864912986755, -0.36724191904067993, -0.0022578760981559753, -0.0650901198387146, 0.3331246078014374, 0.3727763295173645, -0.1935800313949585, 0.18135881423950195, -0.2651146948337555, -0.07853227853775024, -0.2893359065055847, 0.13996614515781403, -0.07206805050373077, 0.05796957388520241, -0.2934381663799286, -0.04770263284444809, 0.3240184485912323, 0.06630030274391174, -0.09036258608102798, -0.4725525379180908, 0.31648772954940796, -0.06607367843389511, 0.041039589792490005, -0.06410311907529831, 0.6213017702102661, -0.1682785153388977, -0.1320490837097168, 0.2150786817073822, 0.22514674067497253, -0.3285653293132782, 0.2118608057498932, 0.3367438316345215, 0.24053168296813965, 0.17107175290584564, -0.054370369762182236, 0.10397013276815414, 0.019159464165568352, -0.012394202873110771, -0.2296966016292572, -0.10859379917383194, 0.5694274306297302, -0.06802934408187866, 0.4326314330101013, -0.02600593864917755, 0.04270171374082565, -0.18314510583877563, -0.11107334494590759, -0.24281013011932373, 0.38110747933387756, 0.2990497648715973, -0.21203340590000153, 0.35235998034477234, -0.0513424351811409, -0.3553396761417389, 0.13639584183692932, 0.13285274803638458, 0.0007288865745067596, 0.0813232958316803, 0.018113721162080765, -0.06393145769834518, -0.24718385934829712, -0.097596175968647, 0.41166672110557556, 0.5995680689811707, 0.2978718876838684, 0.14136414229869843, 0.07985891401767731, -0.2595907747745514, -0.07501399517059326, 0.03891145437955856, 0.19106222689151764, -0.07945381850004196, 0.15421119332313538, 0.010218745097517967, 0.057113002985715866, -0.015598870813846588, -0.09665244817733765, 0.12611661851406097, -0.2244982272386551, -0.02574707567691803, 0.04047059267759323, -0.15040066838264465, -0.5861160159111023, -0.23188872635364532, 0.005161036271601915, -0.31230658292770386, -0.264523983001709, 0.1403566598892212, 0.03932912275195122, -0.17840608954429626, 0.15653905272483826, 0.18521946668624878, 0.6711623072624207, -0.4588552415370941, -0.2282126098871231, -0.05708080902695656, -0.2689540684223175, -0.11032205075025558, 0.12571704387664795, 0.42610594630241394, 0.27763330936431885, 0.5188313722610474, 0.30895715951919556, -0.03489592671394348, 0.06895482540130615, -0.3962318003177643, 0.05782635509967804, -0.2660856246948242, 0.3105848729610443, 0.11344976723194122, -0.22008086740970612, 0.17031775414943695, -0.37428954243659973, 0.0022700424306094646, -0.10265406221151352, -0.03408002853393555, -0.07405854016542435, -0.0439092256128788, 0.08226968348026276, -0.2027607560157776, -0.19468384981155396, -0.6382986903190613, -0.47390538454055786, 0.3052135407924652, -0.21334078907966614, 0.06977495551109314, 0.019458502531051636, -0.25893715023994446, 0.17852020263671875, -0.13915321230888367, -0.013687675818800926, -0.19458447396755219, -0.12896260619163513, 0.01676156371831894, -0.3375963270664215, -0.1641012728214264, -0.40518635511398315, -0.20086050033569336, 0.22619260847568512, -0.09001278877258301, -0.06369993835687637, -0.2524302005767822, 0.07718165218830109, -0.12796750664710999, 0.11129318177700043, 0.2199832648038864, 0.1511763036251068, -0.2682066857814789, 0.09364662319421768, -0.029154960066080093, -0.20805367827415466, 0.25715190172195435, 0.29497888684272766, 0.37840813398361206, 0.1561039388179779, 0.038204602897167206, 0.13332581520080566, 0.7063828110694885, 0.244694322347641, 0.031950805336236954, 0.1457706242799759, 0.30789634585380554, 0.017253555357456207, 0.1577180027961731, -0.22609084844589233, 0.19193778932094574, -0.1703614592552185, 0.44894129037857056, 0.10616723448038101, 0.19418981671333313, -0.20405080914497375, -0.2649410665035248, 0.05006055906414986, -0.22742325067520142, -0.23034411668777466, 0.26681944727897644, -0.18870264291763306, 0.01172114908695221, -0.0904022604227066, -0.36917439103126526, -0.24698743224143982, -0.022600457072257996, 0.01634928770363331, 0.2658746540546417, -0.023195773363113403, 0.12559154629707336, -0.4113297164440155, -0.03634204342961311, -0.43079227209091187, 0.1822599619626999, 0.07882989197969437, 0.26000621914863586, -0.06049288064241409, -0.07115326076745987, -0.08950991928577423, 0.10068007558584213, 0.5856121182441711, -0.5367428660392761, -0.17423446476459503, 0.11988365650177002, 0.07640215754508972, -0.14059610664844513, -0.06939823180437088, -0.27343234419822693, -0.044390156865119934, 0.45053955912590027, 0.3297434449195862, -0.35634875297546387, -0.2867952883243561, 0.29957589507102966, -0.17227555811405182, -0.09658411145210266, -0.20668786764144897, -0.15361671149730682, -0.1994931697845459, -0.13871708512306213, 0.05471811443567276, 0.18321537971496582, 0.15969343483448029, -0.10423214733600616, 0.12230968475341797, -0.11102566123008728, 0.2156314253807068, 0.19780366122722626, 0.009142642840743065, -0.0450814850628376, 0.21428577601909637, 0.17898498475551605, 0.032369524240493774, -0.14142334461212158, -0.34136611223220825, 0.1296176016330719, -0.06286951899528503, 0.129246324300766, 0.23772463202476501, -0.015323157422244549, 0.6900310516357422, 0.042731717228889465, 0.3528766334056854, 0.31852808594703674, -0.10639488697052002, 0.009463511407375336, -0.3053719699382782, 0.130976140499115, 0.16529572010040283, 0.24443557858467102, -0.11793570220470428, -0.22836287319660187, 0.5872114896774292, 0.3297061622142792, -0.08609949797391891, 0.3757953941822052, -0.4737694263458252, -0.3104903995990753, 0.32061368227005005, 0.04829271882772446, 0.7352595925331116, -0.020149655640125275, 0.2903289496898651, 0.06005868315696716, -0.380578875541687, 0.6396002173423767, -0.3495901823043823, 0.1265997290611267, -0.1399349421262741, -0.20601844787597656, 0.00896688736975193, -0.08151331543922424, -0.06173834949731827, 0.2745792865753174, -0.3608980178833008, 0.3065803349018097, 0.37564271688461304, -0.22479894757270813, -0.026938384398818016, 0.5579148530960083, -0.02848581038415432, -0.40201908349990845, -0.07268825173377991, 0.10170149803161621, -0.2750025987625122, 0.06954959034919739, -0.16664573550224304, -0.23241929709911346, 0.2085062861442566, 0.135392963886261, -0.2649180293083191, 0.10901264101266861, -0.021995829418301582, 0.15653371810913086, 0.2781093120574951, 0.021803833544254303, 0.27406981587409973, -0.2461930215358734, -0.08067634701728821, -0.10604271292686462, -0.08823463320732117, 0.13408753275871277, -0.22273996472358704, 0.059132736176252365, 0.07656439393758774, -0.16619165241718292, 0.529438853263855, -0.029330454766750336, -0.10595725476741791, -0.20204520225524902, 0.23767605423927307, -0.428960919380188, -0.16126343607902527, 0.10507818311452866, 0.19494932889938354, 0.12262531369924545, -0.35108932852745056, 0.5494362115859985, 0.09299609065055847, -0.030742866918444633, 0.05949225276708603, 0.160847470164299, -0.285043329000473, 0.46693888306617737, -0.20441322028636932, 0.014505095779895782, 0.06675855070352554, -0.03496868163347244, -0.0981704369187355, -0.3632012903690338, -0.03582313656806946, 0.2880770266056061, 0.14274710416793823, -0.045032914727926254, 0.008256874978542328, 0.17375093698501587, -0.06492220610380173, -0.000905974768102169, -0.10859966278076172, -0.367808997631073, 0.13817241787910461, 0.15505200624465942, 0.28181618452072144, 0.20770156383514404, -0.1084919422864914, -0.2485351711511612, -0.22846619784832, 0.45401978492736816, 0.08564331382513046, 0.24932020902633667, -0.13096801936626434, 0.36532971262931824, 0.40376007556915283, 0.555581271648407, -0.28828108310699463, 0.047473788261413574, 0.030073771253228188, 0.09261932969093323, 0.48831918835639954, -0.24070215225219727, 0.0004641436389647424, 0.271899551153183, -0.1266116350889206, 0.06990136206150055, -0.3354588449001312, -0.18014976382255554, -0.2789696156978607, 0.1032845601439476, 0.1215905100107193, 0.004378753714263439, 0.01943439617753029, -0.025179700925946236, -0.09327095746994019, -0.07621203362941742, -0.032401952892541885, -0.42004576325416565, -0.05962035059928894, 0.13473327457904816, -0.07899188995361328, 0.04574663192033768, -0.004300221800804138, 0.21682439744472504, -0.0051980093121528625, -0.04506361484527588, 0.27724528312683105, 0.08747966587543488, 0.14421889185905457, -0.17655183374881744, -0.142299085855484, 0.006133183836936951, 0.3090972304344177, 0.10410726070404053, -0.028175629675388336, 0.1752343475818634, -0.2041565477848053, -0.07347287237644196, 0.26419705152511597, 0.19137436151504517, 0.02827773615717888, -0.06573785841464996, 0.12197957932949066, 0.15415149927139282, -0.13383066654205322, 0.03742099180817604, 0.24652691185474396, 0.052805643528699875, 0.12569384276866913, 0.13203656673431396, 0.15263524651527405, 0.4616323709487915, -0.22779320180416107, -0.009581963531672955, 0.13401474058628082, -0.3001980781555176, 0.16639268398284912, -0.1424228399991989, -0.0036706216633319855, -0.018843088299036026, 0.14484070241451263, -0.030460692942142487, 0.17718975245952606, -0.3630979657173157, 0.29315540194511414, 0.3141861855983734, -0.01202450692653656, 0.04972722753882408, 0.21571552753448486, 0.011386286467313766, -0.1301044374704361, 0.30930817127227783, 0.09537190198898315, 0.33699968457221985, 0.003113105893135071, 0.8959373831748962, -0.3085148334503174, -0.4987756013870239, 0.08208086341619492, 0.4228001832962036, 0.03482485190033913, -0.19934433698654175, 0.08230765163898468, 0.1272721290588379, -0.36981821060180664, 0.07597455382347107, 0.07843572646379471, -0.18659235537052155, 0.6474287509918213, -0.11679467558860779, 0.16359634697437286, -0.1702311635017395, -0.20496293902397156, 0.3429817855358124, 0.15966227650642395, -0.1373148262500763, 0.027703264728188515, 0.11107563227415085, -0.06848166882991791, -0.162620410323143, 0.5162652134895325, 0.1743650883436203, 0.01610345020890236, -0.44874659180641174, 0.041226696223020554, 0.0500359982252121, 0.15217286348342896, -0.08710524439811707, 0.1192086935043335, 0.4416902959346771, -0.16612105071544647, 0.24096046388149261, 0.038404934108257294, -0.10613935440778732, -0.15496879816055298, 0.014972109347581863, -0.3957519233226776, -0.4279155433177948, 0.4407232403755188, 0.12134898453950882, -0.15100565552711487, 0.018220771104097366, 0.25174224376678467, -0.6123269200325012, -0.22717365622520447, 0.3545632064342499, -0.1733732670545578, -0.0767514556646347, -0.22562576830387115, 0.08382952958345413, 0.03794964775443077, 0.6115517616271973, 0.21737883985042572, 0.35958364605903625, -0.38429316878318787, -0.1715811789035797, -0.5617932081222534, -0.18688622117042542, 0.14032000303268433, -0.10462857782840729, -0.24724486470222473, 0.16251887381076813, -0.07911864668130875, 0.296741247177124, -0.22634278237819672, 0.12328320741653442, 0.22285419702529907, -0.024598680436611176, -0.4108704924583435, 0.12129530310630798, 0.12449541687965393, 0.021148012951016426, 0.31346920132637024, -0.08089778572320938, -0.21032868325710297, 0.07324421405792236, -0.07079584896564484, -0.03564128652215004, -0.08734466880559921, 0.26318782567977905, 0.1474705934524536, 0.1126747578382492, 0.0629216656088829, 0.5086536407470703, -0.0262728463858366, -0.16295559704303741, 0.02011765167117119, -0.04937830939888954, -0.38858604431152344, -0.1232491210103035, -0.08058828115463257, 0.21543635427951813, -0.356221467256546, 0.2585195004940033, 0.004246872384101152, -0.15498501062393188, 0.14422056078910828, 0.03260624781250954, -0.2562313973903656, -0.058259010314941406, -0.01238970085978508, 0.21576139330863953, 0.21925483644008636, 0.7426397800445557, 0.0279543437063694, 0.08706794679164886, -0.039152711629867554, -0.009825295768678188, 0.0879296138882637, -0.040988899767398834, -0.22142860293388367, -0.16324160993099213, 0.010918691754341125, 0.015441585332155228, 0.09817143529653549, -0.664664626121521, -0.36807185411453247, 0.04237039387226105, -0.00760621577501297, 0.09818348288536072, 0.7144815325737, 0.08214789628982544, -0.14683596789836884, -0.14546498656272888, -0.2757914364337921, 0.034837059676647186, -0.245181605219841, 0.1932796835899353, -0.39100319147109985 ]
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.
[ -0.029288088902831078, -0.3940451741218567, -0.061029281467199326, -0.025661781430244446, -0.2608686685562134, 0.08258078247308731, 0.18105274438858032, 0.04695507138967514, 0.2868426442146301, -0.1594534069299698, -0.23304535448551178, 0.3090498447418213, -0.20747075974941254, 0.38534146547317505, 0.37004518508911133, -0.44683244824409485, -0.007145114243030548, -0.20719191431999207, -0.2622194290161133, 0.3925706148147583, 0.03699757903814316, 0.11199010163545609, -0.14147131145000458, 0.10980463773012161, -0.30104702711105347, -0.13675542175769806, -0.3321911692619324, 0.10305800288915634, 0.11648961901664734, 0.01812893897294998, 0.21229058504104614, 0.3617035448551178, -0.04755072668194771, 0.1438959836959839, -0.00011497208470245823, -0.044561855494976044, 0.14054232835769653, -0.07454654574394226, 0.12798750400543213, -0.038532182574272156, 0.1727735996246338, -0.38791531324386597, 0.06748510897159576, -0.13466739654541016, 0.21631953120231628, 0.15920688211917877, 0.1809328943490982, 0.06926152110099792, 0.40569382905960083, -0.32851293683052063, 0.08759269118309021, 0.28492650389671326, -0.19006860256195068, 0.34371376037597656, -0.10634823888540268, -0.1040673702955246, 0.03765280544757843, 0.2589143216609955, 0.586479902267456, -0.5919791460037231, 0.053165215998888016, 0.08587002754211426, 0.012978112325072289, -0.03543013706803322, 0.15835490822792053, -0.028579317033290863, -0.3110813796520233, -0.22444447875022888, -0.41868290305137634, 0.1887262761592865, -0.10443753749132156, -0.15518945455551147, -0.2215287685394287, -0.5562809109687805, -0.017073914408683777, -0.0037985313683748245, -0.3029254674911499, -0.002131476067006588, -0.17310236394405365, -0.03949776664376259, -0.14767667651176453, -0.10621414333581924, -0.05107507482171059, 0.15497082471847534, 0.3348459005355835, 0.5709912180900574, 0.1295989602804184, 0.17109742760658264, 0.3198666572570801, 0.05387607216835022, -0.13317297399044037, -0.2922164499759674, 0.30810126662254333, 0.1642889529466629, -0.36470866203308105, -0.3597911298274994, 0.19177573919296265, -0.25898075103759766, 0.23242563009262085, 0.09343147277832031, 0.12703199684619904, 0.08136390149593353, -0.1447870135307312, 0.26923462748527527, 0.33163169026374817, 0.036866866052150726, -0.10527094453573227, -0.12733322381973267, 0.028762737289071083, -0.15994568169116974, 0.23521608114242554, 0.096702441573143, 0.18593524396419525, -0.022179249674081802, -0.39544251561164856, -0.16337111592292786, -0.13463500142097473, 0.14888417720794678, -0.19198034703731537, -0.5355586409568787, -0.16265323758125305, -0.13995644450187683, 0.04749830812215805, -0.13046738505363464, -0.31422993540763855, 0.07659865915775299, -0.030082136392593384, 0.1895587146282196, -0.14268620312213898, 0.12083224207162857, -0.052640534937381744, 0.07124754786491394, -0.44809192419052124, -0.046076878905296326, 0.2776446044445038, 0.10952538251876831, 0.08483003824949265, 0.1338437795639038, 0.014841295778751373, 0.12547995150089264, 0.4019385874271393, 0.03929182514548302, -0.019937366247177124, -0.03314119204878807, 0.16819153726100922, -0.3317154347896576, -0.006262622773647308, 0.17574232816696167, -0.3530604839324951, -0.1490301489830017, -0.15961137413978577, -0.1868329644203186, 0.12111593037843704, 0.024497399106621742, -0.22084029018878937, -0.320146381855011, -0.4782353341579437, 0.6768113970756531, -0.02763400226831436, -0.06473112851381302, -0.16472402215003967, 0.27578598260879517, -0.2691701650619507, -0.12027681618928909, 0.17447832226753235, 0.04582887887954712, 0.0011192001402378082, -0.3560561239719391, -0.06673183292150497, -0.07920421659946442, -0.07760482281446457, 0.25847122073173523, -0.1712941974401474, 0.2623891234397888, 0.09088423103094101, 0.1857634335756302, 0.34088096022605896, -0.4731498658657074, -0.022683417424559593, 0.2970294952392578, -0.05563241243362427, 0.41089126467704773, 0.07046730816364288, 0.2279609888792038, -0.08120601624250412, -0.2089109718799591, 0.2719021439552307, 1.0307587385177612, -0.3764412999153137, -0.02222232148051262, -0.09570206701755524, -0.3904014229774475, 0.5535184741020203, 0.34943294525146484, 0.07714053988456726, -0.3079705238342285, -0.1021806076169014, -0.009633601643145084, 0.06673574447631836, 0.08410999178886414, 0.07473351806402206, -0.08781086653470993, -0.3393521010875702, 0.11391237378120422, -0.21026675403118134, -0.1518992781639099, -0.3530142307281494, 0.033923424780368805, 0.024167057126760483, 0.31401926279067993, 0.25735369324684143, -0.1301783323287964, 0.2086988389492035, -0.28740155696868896, -0.14865362644195557, -0.2948305606842041, 0.10739141702651978, -0.07125721871852875, 0.00028781965374946594, -0.2379792034626007, -0.03177569806575775, 0.3077721893787384, 0.13992391526699066, -0.13622434437274933, -0.4772080183029175, 0.29331836104393005, -0.06322217732667923, 0.05785415694117546, -0.056329090148210526, 0.6482879519462585, -0.16215437650680542, -0.06727927178144455, 0.19220829010009766, 0.32548788189888, -0.3306891620159149, 0.2363167107105255, 0.3137441873550415, 0.3130967319011688, 0.16561448574066162, 0.006151299923658371, 0.1179901510477066, 0.011639727279543877, -0.001240383367985487, -0.21448662877082825, -0.09475650638341904, 0.5063755512237549, -0.19477367401123047, 0.4248299300670624, -0.023951157927513123, -0.002447165548801422, -0.20531035959720612, -0.015217259526252747, -0.17521768808364868, 0.41539132595062256, 0.29060012102127075, -0.23537112772464752, 0.29709145426750183, -0.16056767106056213, -0.37394168972969055, 0.17211991548538208, 0.17921608686447144, 0.018867194652557373, 0.03440475091338158, -0.04453393071889877, 0.0077328160405159, -0.2754381597042084, -0.10600511729717255, 0.31868237257003784, 0.657733142375946, 0.2544577419757843, 0.1376100331544876, 0.06570452451705933, -0.2021634876728058, -0.05955902859568596, 0.048522621393203735, 0.21359135210514069, -0.04889417812228203, 0.15909138321876526, 0.036246854811906815, 0.08083906024694443, 0.062116049230098724, -0.16772831976413727, 0.1392093449831009, -0.2044977992773056, -0.11621741950511932, 0.054170407354831696, -0.18401747941970825, -0.6338428258895874, -0.18930378556251526, -0.009177061729133129, -0.33165132999420166, -0.2957558333873749, 0.13720130920410156, 0.05582486838102341, -0.11464425921440125, 0.13877373933792114, 0.1970454901456833, 0.5815644264221191, -0.43081381916999817, -0.2768559753894806, -0.03663846105337143, -0.31371769309043884, -0.10173463076353073, 0.1042480319738388, 0.4355733096599579, 0.2959771454334259, 0.5036947131156921, 0.2990516424179077, -0.06889364868402481, 0.0986972227692604, -0.4021720886230469, 0.06541772186756134, -0.35339850187301636, 0.2585299611091614, 0.13582871854305267, -0.2217332273721695, 0.14582960307598114, -0.3897812068462372, -0.05303334444761276, -0.025669991970062256, -0.07492633908987045, -0.05133765935897827, 0.054068006575107574, 0.06931324303150177, -0.19855494797229767, -0.19036796689033508, -0.6149934530258179, -0.44191789627075195, 0.3062751293182373, -0.22799837589263916, 0.08272532373666763, 0.06216789409518242, -0.2821543216705322, 0.23458461463451385, -0.08945947885513306, 0.025363514199852943, -0.1702771782875061, -0.05233986675739288, 0.010179545730352402, -0.3848069906234741, -0.1447841376066208, -0.3393521010875702, -0.039895057678222656, 0.22139781713485718, -0.06605792045593262, -0.029407277703285217, -0.275092214345932, 0.08280210942029953, -0.08778975158929825, 0.07219231128692627, 0.26224303245544434, 0.16565276682376862, -0.2060343325138092, 0.13177178800106049, -0.08111491054296494, -0.27193501591682434, 0.21034687757492065, 0.33023664355278015, 0.37373843789100647, 0.18449366092681885, 0.03631783649325371, 0.18707406520843506, 0.7639586329460144, 0.24814295768737793, 0.0630788654088974, 0.14840328693389893, 0.2611876130104065, -0.012645263224840164, 0.16147339344024658, -0.1882694661617279, 0.1840166300535202, -0.1733154058456421, 0.5277929902076721, 0.08122701942920685, 0.1695801317691803, -0.14921659231185913, -0.3321092426776886, 0.01802697777748108, -0.3142361044883728, -0.18452826142311096, 0.19016119837760925, -0.2771725356578827, 0.01715637743473053, -0.0910215675830841, -0.4362218976020813, -0.28573817014694214, -0.037379682064056396, -0.017000479623675346, 0.3356783390045166, -0.0067284852266311646, 0.17077067494392395, -0.33541008830070496, -0.04341872036457062, -0.42174914479255676, 0.1442059576511383, 0.058303818106651306, 0.23728248476982117, -0.08073723316192627, -0.07213886082172394, -0.10103416442871094, 0.11182694882154465, 0.5621090531349182, -0.5473767518997192, -0.10901273787021637, 0.0950554609298706, 0.07976454496383667, -0.12696273624897003, -0.11654843389987946, -0.2918790578842163, 0.02043408341705799, 0.42763039469718933, 0.37687990069389343, -0.37812376022338867, -0.29430466890335083, 0.2408275604248047, -0.253957599401474, -0.0721559077501297, -0.1933041512966156, -0.2405911237001419, -0.2112833559513092, -0.12438013404607773, -0.005925334990024567, 0.2536102533340454, 0.23441138863563538, -0.12050190567970276, 0.18363387882709503, -0.08123081177473068, 0.2604242265224457, 0.24369797110557556, 0.010381605476140976, -0.11040753871202469, 0.2441927194595337, 0.17174628376960754, 0.038996510207653046, -0.12102335691452026, -0.35914620757102966, 0.09985445439815521, -0.021673571318387985, 0.0665009617805481, 0.2553040385246277, -0.0018159598112106323, 0.6958091259002686, 0.046072423458099365, 0.3278181254863739, 0.27770739793777466, -0.14805655181407928, -0.00294533371925354, -0.22390860319137573, 0.18585385382175446, 0.23404698073863983, 0.24906158447265625, -0.14151260256767273, -0.23254254460334778, 0.5836414098739624, 0.30703142285346985, -0.08686482906341553, 0.43852701783180237, -0.49418801069259644, -0.33374685049057007, 0.21709124743938446, 0.02576620876789093, 0.7134889960289001, -0.022532179951667786, 0.3297560513019562, 0.11808131635189056, -0.41327381134033203, 0.5890750885009766, -0.33622241020202637, 0.12408437579870224, -0.054837848991155624, -0.17275115847587585, -0.011037005111575127, -0.07591135799884796, -0.10295803099870682, 0.2057555615901947, -0.3488755226135254, 0.3369857966899872, 0.403509259223938, -0.23023736476898193, -0.047393523156642914, 0.5434771776199341, -0.01447100006043911, -0.3405782878398895, -0.07095815986394882, 0.051304690539836884, -0.2821178436279297, 0.06948010623455048, -0.2126551866531372, -0.25520122051239014, 0.21050375699996948, 0.15928402543067932, -0.27442002296447754, 0.06679576635360718, 0.008197695016860962, 0.2008352279663086, 0.22477677464485168, 0.0355929359793663, 0.26400211453437805, -0.17304955422878265, -0.1037311851978302, -0.1767415553331375, -0.06488702446222305, 0.10578872263431549, -0.2313130795955658, 0.03633556514978409, 0.0953330397605896, -0.1507091224193573, 0.4938497543334961, -0.07441737502813339, -0.1055268719792366, -0.2258998453617096, 0.26180723309516907, -0.44568145275115967, -0.2548680305480957, 0.13174155354499817, 0.16105252504348755, 0.18171535432338715, -0.32188621163368225, 0.594508171081543, 0.08666715025901794, 0.05346443131566048, 0.018005715683102608, 0.17048421502113342, -0.23972740769386292, 0.46453961730003357, -0.32235217094421387, 0.0002985820174217224, 0.09918470680713654, 0.00014899671077728271, -0.1518298089504242, -0.4189448356628418, -0.053586021065711975, 0.2748265862464905, 0.12449447810649872, 0.055341511964797974, -0.01699470728635788, 0.11429896205663681, -0.03799336776137352, -0.0555114783346653, -0.11189214885234833, -0.2893211245536804, 0.1342899203300476, 0.2440171241760254, 0.1765127182006836, 0.1978479027748108, -0.060454923659563065, -0.2944514751434326, -0.13006703555583954, 0.49924999475479126, -0.016898449510335922, 0.2320835292339325, -0.15137694776058197, 0.37086302042007446, 0.4106280207633972, 0.49037304520606995, -0.24309992790222168, 0.057044617831707, 0.09853257983922958, 0.10004319995641708, 0.4676119089126587, -0.22680512070655823, 0.047387488186359406, 0.2821199893951416, -0.14611601829528809, 0.031245801597833633, -0.3519592583179474, -0.1358805149793625, -0.24969330430030823, 0.12456847727298737, 0.12881559133529663, -0.0284756850451231, 0.028416700661182404, -0.11128637194633484, -0.10461029410362244, -0.057755425572395325, 0.04444064572453499, -0.36330124735832214, -0.09757275879383087, 0.1614966243505478, -0.09412811696529388, 0.08671195060014725, 0.057051338255405426, 0.20702838897705078, 0.06947362422943115, -0.07310886681079865, 0.33872056007385254, 0.12007106840610504, 0.08686429262161255, -0.16815894842147827, -0.14083035290241241, 0.01643984392285347, 0.29556000232696533, 0.180315762758255, -0.06893271952867508, 0.1422436386346817, -0.20938445627689362, -0.19972947239875793, 0.207700714468956, 0.20243296027183533, 0.03528792783617973, -0.13023819029331207, 0.10691758990287781, 0.09648489952087402, -0.06001542881131172, -0.01290728710591793, 0.23015180230140686, 0.08779247850179672, 0.07244819402694702, 0.11680930107831955, 0.11368687450885773, 0.4800723195075989, -0.28227096796035767, 0.032748743891716, 0.09447218477725983, -0.3594552278518677, 0.11638280749320984, -0.08243821561336517, -0.08555788546800613, -0.014490868896245956, 0.12477198243141174, 0.00008924305438995361, 0.11933603882789612, -0.33420151472091675, 0.3240496516227722, 0.26727110147476196, 0.015397857874631882, 0.01500190794467926, 0.1387517899274826, 0.04229075089097023, -0.17318110167980194, 0.2679608464241028, 0.03501954674720764, 0.3751722574234009, 0.06041339784860611, 0.8366438746452332, -0.2134696990251541, -0.39151203632354736, 0.15799500048160553, 0.42227044701576233, 0.08762919157743454, -0.21886204183101654, 0.13186512887477875, 0.09811235219240189, -0.37548473477363586, 0.06979610025882721, 0.0569925457239151, -0.1820814460515976, 0.6210576295852661, -0.08598870038986206, 0.16826140880584717, -0.11792021989822388, -0.18515899777412415, 0.3580675721168518, 0.21875646710395813, -0.0734490305185318, -0.03959990292787552, 0.12822821736335754, -0.061942100524902344, -0.19880662858486176, 0.47432956099510193, 0.2133026421070099, -0.0024094320833683014, -0.48168647289276123, 0.03337985277175903, 0.10145311057567596, 0.17900331318378448, 0.0025961026549339294, 0.1369650661945343, 0.4460619390010834, -0.15026038885116577, 0.20349398255348206, -0.007044810801744461, -0.03401816636323929, -0.13445675373077393, 0.0641835480928421, -0.4029550850391388, -0.4237152338027954, 0.4620606303215027, 0.1102583110332489, -0.14952698349952698, 0.11755184829235077, 0.37620967626571655, -0.6295855045318604, -0.26949018239974976, 0.33437567949295044, -0.05823199078440666, -0.08502033352851868, -0.19609683752059937, 0.06549707800149918, 0.07512514293193817, 0.6126105189323425, 0.20807671546936035, 0.4082014858722687, -0.3459770679473877, -0.09671205282211304, -0.5682133436203003, -0.20369422435760498, 0.17174018919467926, -0.15298008918762207, -0.22418370842933655, 0.1183871328830719, -0.06933384388685226, 0.27628377079963684, -0.1890554279088974, 0.1030978187918663, 0.24528977274894714, 0.049819692969322205, -0.4366210699081421, 0.08494247496128082, 0.19068224728107452, 0.0404028445482254, 0.2738911807537079, -0.06394287198781967, -0.22662179172039032, 0.04732493311166763, -0.06836362183094025, -0.09690745174884796, -0.11419853568077087, 0.26815515756607056, 0.1839274764060974, 0.09595856070518494, 0.057073406875133514, 0.5114501714706421, 0.010969426482915878, -0.15749332308769226, 0.0035766661167144775, 0.04371960833668709, -0.36338210105895996, -0.13569670915603638, -0.09348143637180328, 0.13298873603343964, -0.39160358905792236, 0.32264938950538635, 0.0057074506767094135, -0.07805059105157852, 0.09187640249729156, 0.0003340801631566137, -0.12278414517641068, -0.11219880729913712, 0.04377824068069458, 0.2214047908782959, 0.18611274659633636, 0.7520402073860168, -0.012685023248195648, 0.13772812485694885, -0.09494926035404205, -0.008783180266618729, 0.027268782258033752, -0.06434722989797592, -0.254292756319046, -0.09406182914972305, 0.0674457997083664, 0.06809153407812119, 0.1432076394557953, -0.6984319090843201, -0.3541291356086731, 0.05046321079134941, 0.01615240052342415, 0.015999507158994675, 0.7800260186195374, 0.09473781287670135, -0.13276270031929016, -0.08945003151893616, -0.36224567890167236, 0.0005289893597364426, -0.22626979649066925, 0.16598626971244812, -0.3741370439529419 ]
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!
[ -0.04102962464094162, -0.3800145387649536, -0.05613580346107483, -0.05533301830291748, -0.22696126997470856, 0.038144730031490326, 0.19116359949111938, 0.028541235253214836, 0.31403058767318726, -0.16932375729084015, -0.2113303542137146, 0.3226812183856964, -0.21177791059017181, 0.3609279990196228, 0.41081103682518005, -0.4081554114818573, 0.049611590802669525, -0.23854923248291016, -0.30185526609420776, 0.38168203830718994, 0.029433421790599823, 0.11054189503192902, -0.18156084418296814, 0.11165952682495117, -0.27534911036491394, -0.14701829850673676, -0.3376019597053528, 0.10576724261045456, 0.12588635087013245, 0.0001229001209139824, 0.2541155219078064, 0.4050154387950897, -0.0645340085029602, 0.16401571035385132, -0.00011387035192456096, -0.028973150998353958, 0.23698236048221588, -0.06454761326313019, 0.11490742862224579, -0.07505476474761963, 0.20829960703849792, -0.42716363072395325, 0.05915150046348572, -0.16756829619407654, 0.1729607880115509, 0.15671946108341217, 0.1583680659532547, 0.04002629965543747, 0.39363449811935425, -0.2916637063026428, 0.10325049608945847, 0.2875029742717743, -0.15446846187114716, 0.27613866329193115, -0.12511226534843445, -0.047687649726867676, 0.036738425493240356, 0.2582833766937256, 0.5602096319198608, -0.6138644218444824, 0.003775171935558319, 0.1158638596534729, 0.0011098384857177734, 0.004889128729701042, 0.15319645404815674, -0.0207950621843338, -0.35217273235321045, -0.2176881730556488, -0.403907835483551, 0.18628636002540588, -0.1451551616191864, -0.20172220468521118, -0.23753756284713745, -0.5448267459869385, 0.00467009749263525, 0.006173262372612953, -0.26987841725349426, -0.018358614295721054, -0.20179776847362518, -0.06324602663516998, -0.12326838821172714, -0.08969501405954361, -0.03366662561893463, 0.15525320172309875, 0.31824880838394165, 0.5842540860176086, 0.09058930724859238, 0.16732612252235413, 0.36797162890434265, 0.04139312356710434, -0.15085740387439728, -0.2804940640926361, 0.3275359869003296, 0.14582251012325287, -0.36985042691230774, -0.303060919046402, 0.20416340231895447, -0.23676475882530212, 0.22982433438301086, 0.10457167774438858, 0.09047483652830124, 0.11722014099359512, -0.14798322319984436, 0.26190412044525146, 0.30457472801208496, 0.04612195864319801, -0.09614870697259903, -0.09544776380062103, 0.03467622399330139, -0.08594397455453873, 0.19273574650287628, 0.1121780201792717, 0.1854979693889618, -0.012385403737425804, -0.3950411379337311, -0.13390576839447021, -0.12934401631355286, 0.16474467515945435, -0.21528448164463043, -0.4849030375480652, -0.10857385396957397, -0.11952342092990875, 0.11771436780691147, -0.10458620637655258, -0.3091358244419098, 0.09875422716140747, -0.03895314037799835, 0.2612786889076233, -0.1374974101781845, 0.13061656057834625, -0.07104397565126419, 0.08409787714481354, -0.4412364065647125, -0.07710470259189606, 0.3059290945529938, 0.15183445811271667, 0.08070219308137894, 0.1679019182920456, -0.0014318525791168213, 0.12360233068466187, 0.389382004737854, 0.015577927231788635, 0.0013970881700515747, -0.0009414032101631165, 0.13964273035526276, -0.3161430358886719, 0.020423350855708122, 0.1589083969593048, -0.34745216369628906, -0.19029992818832397, -0.11934822052717209, -0.16854071617126465, 0.14057381451129913, 0.03215441107749939, -0.20059601962566376, -0.23849985003471375, -0.49818840622901917, 0.7015707492828369, -0.05721471458673477, -0.0934588611125946, -0.175914466381073, 0.30375146865844727, -0.24646390974521637, -0.08007052540779114, 0.21196673810482025, 0.057640403509140015, 0.011351071298122406, -0.3442132771015167, -0.10541735589504242, -0.0986817479133606, -0.02154744416475296, 0.25007933378219604, -0.16200295090675354, 0.27927154302597046, 0.05273354798555374, 0.20901422202587128, 0.279838889837265, -0.48212292790412903, -0.06264183670282364, 0.2596435844898224, -0.044113803654909134, 0.42299899458885193, 0.05807698890566826, 0.19613584876060486, -0.02601388469338417, -0.1701488345861435, 0.29014694690704346, 0.9980716109275818, -0.34383922815322876, -0.019431522116065025, -0.10914511233568192, -0.359876424074173, 0.5201607346534729, 0.3251276910305023, 0.10994377732276917, -0.2907424867153168, -0.13588455319404602, -0.015212766826152802, 0.12174706161022186, 0.07488220185041428, 0.08309371024370193, -0.0717742070555687, -0.3480933904647827, 0.06251135468482971, -0.21467044949531555, -0.16926828026771545, -0.34028953313827515, 0.023376867175102234, 0.05374082177877426, 0.33713722229003906, 0.211589515209198, -0.14518499374389648, 0.19584709405899048, -0.255342960357666, -0.14328446984291077, -0.30226704478263855, 0.14626285433769226, -0.06994940340518951, 0.013663891702890396, -0.24087928235530853, -0.031897496432065964, 0.3690546154975891, 0.11622273921966553, -0.12677861750125885, -0.4973999857902527, 0.3005133867263794, -0.07422433793544769, 0.06661362200975418, -0.03259079158306122, 0.6585509777069092, -0.15063244104385376, -0.0851476788520813, 0.1811935007572174, 0.28125569224357605, -0.2942325174808502, 0.20129701495170593, 0.3036305606365204, 0.19760726392269135, 0.15173554420471191, 0.004687663167715073, 0.07329729199409485, 0.052242256700992584, 0.024796411395072937, -0.23360514640808105, -0.09858066588640213, 0.5100671052932739, -0.1613478660583496, 0.452299028635025, -0.028002260252833366, 0.025157766416668892, -0.1460137516260147, -0.02968289516866207, -0.20386555790901184, 0.38019752502441406, 0.28941023349761963, -0.18660050630569458, 0.3278832733631134, -0.16236703097820282, -0.3643941581249237, 0.15504823625087738, 0.13120117783546448, 0.012626104056835175, 0.054630741477012634, -0.043492428958415985, -0.013248145580291748, -0.27911409735679626, -0.10775341093540192, 0.3511252999305725, 0.6806016564369202, 0.23463593423366547, 0.19186411798000336, 0.06999503076076508, -0.2855544984340668, -0.07115587592124939, 0.0013609528541564941, 0.22168385982513428, -0.04060807079076767, 0.16736525297164917, 0.06225525960326195, 0.042443666607141495, 0.02196458727121353, -0.152826726436615, 0.1556808203458786, -0.18520916998386383, -0.093662790954113, 0.10740455240011215, -0.22168302536010742, -0.593887984752655, -0.17749641835689545, 0.0022361506707966328, -0.28781720995903015, -0.3085006773471832, 0.10392729938030243, 0.05332016199827194, -0.14587950706481934, 0.16206085681915283, 0.20494990050792694, 0.5456448197364807, -0.42458266019821167, -0.27564889192581177, -0.04172290861606598, -0.31425341963768005, -0.11091042309999466, 0.10897095501422882, 0.45676755905151367, 0.33519771695137024, 0.48909059166908264, 0.3100678324699402, -0.0816788449883461, 0.04969361424446106, -0.40164679288864136, 0.047573596239089966, -0.32548993825912476, 0.26705124974250793, 0.11735353618860245, -0.2622724175453186, 0.16213956475257874, -0.37220484018325806, -0.0118264677003026, -0.08734836429357529, -0.08597197383642197, -0.021902648732066154, -0.019531672820448875, 0.05914955586194992, -0.21000738441944122, -0.15615111589431763, -0.6618931293487549, -0.45373156666755676, 0.3190036416053772, -0.2383420616388321, 0.08222013711929321, 0.059315554797649384, -0.295492947101593, 0.222370445728302, -0.06498803198337555, 0.016361389309167862, -0.1773642748594284, -0.048845671117305756, 0.029317837208509445, -0.3534867763519287, -0.17807669937610626, -0.37236499786376953, -0.13820210099220276, 0.2802262604236603, -0.06061965972185135, -0.00712372362613678, -0.25036323070526123, 0.09986145049333572, -0.0790095180273056, 0.1057049036026001, 0.24656324088573456, 0.17851471900939941, -0.2391587495803833, 0.13448840379714966, -0.060159504413604736, -0.2647288143634796, 0.24683836102485657, 0.28179123997688293, 0.3843090236186981, 0.18339258432388306, 0.05993513762950897, 0.1708054095506668, 0.6879076361656189, 0.22980087995529175, 0.06567364931106567, 0.1472759246826172, 0.25494474172592163, 0.019310712814331055, 0.16987480223178864, -0.2620755732059479, 0.20821648836135864, -0.21381089091300964, 0.4608769416809082, 0.045081254094839096, 0.1656011939048767, -0.2366361767053604, -0.29566115140914917, 0.02418256551027298, -0.3138537108898163, -0.2150166630744934, 0.19742123782634735, -0.2264554798603058, 0.0024595558643341064, -0.07362702488899231, -0.379216730594635, -0.2847122848033905, -0.03315741568803787, 0.004512320272624493, 0.3330303430557251, -0.0960419550538063, 0.11804568022489548, -0.3570634126663208, -0.0310059804469347, -0.46974027156829834, 0.14145812392234802, 0.008326739072799683, 0.25842157006263733, -0.05097510665655136, -0.07873968780040741, -0.08971892297267914, 0.07503249496221542, 0.6008839011192322, -0.4771212637424469, -0.12468117475509644, 0.15265104174613953, 0.11117656528949738, -0.13729418814182281, -0.1263008564710617, -0.29125574231147766, 0.046896789222955704, 0.46849650144577026, 0.37232542037963867, -0.3875475227832794, -0.2634357810020447, 0.22910921275615692, -0.21606238186359406, -0.04498942941427231, -0.1908903568983078, -0.2358568012714386, -0.2560034692287445, -0.09917148947715759, 0.0442420095205307, 0.21777860820293427, 0.2320205718278885, -0.12347625941038132, 0.21100057661533356, -0.10897153615951538, 0.26126062870025635, 0.22021985054016113, 0.004205815494060516, -0.027149219065904617, 0.20108652114868164, 0.17098407447338104, 0.03842533752322197, -0.1398620754480362, -0.3714185953140259, 0.13281747698783875, 0.004647500813007355, 0.07882440090179443, 0.24406200647354126, -0.007657070644199848, 0.6334802508354187, 0.07391240447759628, 0.2964061498641968, 0.2896258234977722, -0.14868685603141785, 0.017745528370141983, -0.2544417679309845, 0.19163630902767181, 0.24283285439014435, 0.26264265179634094, -0.14039891958236694, -0.2564387619495392, 0.5900333523750305, 0.30693069100379944, -0.09948912262916565, 0.4368514120578766, -0.49968451261520386, -0.3505488932132721, 0.2613230347633362, 0.022101111710071564, 0.6753665804862976, 0.005988597869873047, 0.31732290983200073, 0.16142034530639648, -0.37519150972366333, 0.5796291828155518, -0.3777523636817932, 0.16100859642028809, -0.07037866860628128, -0.1492103487253189, 0.004808923229575157, -0.08049367368221283, -0.03711894899606705, 0.22904330492019653, -0.3471164405345917, 0.33152300119400024, 0.44647926092147827, -0.20433306694030762, -0.06859056651592255, 0.5156770944595337, -0.023617537692189217, -0.37427833676338196, -0.12388628721237183, 0.07793339341878891, -0.2607240080833435, 0.09877165406942368, -0.15557417273521423, -0.24808023869991302, 0.193296879529953, 0.14641070365905762, -0.28790104389190674, 0.1049903854727745, 0.030854009091854095, 0.23830074071884155, 0.2569228708744049, 0.009897217154502869, 0.21698251366615295, -0.194374218583107, -0.054441049695014954, -0.1598697453737259, -0.06007956713438034, 0.12273116409778595, -0.20815375447273254, 0.08337394893169403, 0.08152663707733154, -0.2225058376789093, 0.5000624656677246, -0.05308302864432335, -0.1086621880531311, -0.24004215002059937, 0.26080310344696045, -0.4327811002731323, -0.18399743735790253, 0.1675368994474411, 0.16328491270542145, 0.23204074800014496, -0.3219020962715149, 0.6011402010917664, 0.11713387072086334, -0.009043615311384201, 0.03524141013622284, 0.18379531800746918, -0.2748798131942749, 0.5102388858795166, -0.35661590099334717, 0.011908546090126038, 0.08584189414978027, 0.01222093403339386, -0.14734867215156555, -0.37615886330604553, -0.03224923461675644, 0.29934805631637573, 0.08523876219987869, 0.031986914575099945, -0.01090395450592041, 0.10516078025102615, -0.0341632217168808, -0.0448414646089077, -0.09733837097883224, -0.32608121633529663, 0.12282869964838028, 0.1837870478630066, 0.17996639013290405, 0.17217344045639038, -0.08546741306781769, -0.27731460332870483, -0.18056967854499817, 0.4803909659385681, 0.023020224645733833, 0.20212793350219727, -0.15253102779388428, 0.3534679412841797, 0.41968730092048645, 0.5000770688056946, -0.2531452178955078, 0.055512458086013794, 0.07240822911262512, 0.07157933712005615, 0.4242958724498749, -0.24605277180671692, 0.06022543087601662, 0.2450602799654007, -0.12159861624240875, 0.04352039843797684, -0.3518276512622833, -0.14629203081130981, -0.2637699544429779, 0.108505479991436, 0.10938438028097153, 0.0003897491842508316, 0.05136275663971901, -0.09670047461986542, -0.09724637866020203, -0.06615409255027771, -0.012980936095118523, -0.3419173061847687, -0.07169054448604584, 0.12081153690814972, -0.06952276825904846, 0.04318970441818237, 0.07352231442928314, 0.18629994988441467, 0.039545293897390366, -0.07555213570594788, 0.31191930174827576, 0.09127123653888702, 0.1235487088561058, -0.16061893105506897, -0.13676072657108307, 0.020297743380069733, 0.2227650284767151, 0.15596720576286316, -0.04287651181221008, 0.11895954608917236, -0.2020348459482193, -0.1555565893650055, 0.2734251618385315, 0.22079382836818695, 0.02745632827281952, -0.13336586952209473, 0.13120368123054504, 0.11210887134075165, -0.11988216638565063, -0.025684526190161705, 0.23964142799377441, 0.08068189024925232, 0.10649418830871582, 0.12201157957315445, 0.11496567726135254, 0.46063417196273804, -0.2895886301994324, 0.04949895292520523, 0.11247304826974869, -0.3702648878097534, 0.11843657493591309, -0.1373874396085739, -0.010501008480787277, 0.026158351451158524, 0.14512374997138977, -0.004540286958217621, 0.11540011316537857, -0.24399949610233307, 0.32384398579597473, 0.3000832796096802, -0.010189592838287354, 0.03903508931398392, 0.1394422948360443, 0.021483201533555984, -0.17022664844989777, 0.22123786807060242, -0.012152723968029022, 0.3521032929420471, 0.0319325216114521, 0.8429731726646423, -0.30105873942375183, -0.3967202305793762, 0.12651823461055756, 0.40005436539649963, 0.06075844168663025, -0.23886547982692719, 0.13217173516750336, 0.10595058649778366, -0.3579234182834625, 0.060700055211782455, 0.0641697570681572, -0.15428316593170166, 0.6550420522689819, -0.14000152051448822, 0.16470623016357422, -0.1756533533334732, -0.21198037266731262, 0.3319481611251831, 0.20278312265872955, -0.07796873897314072, -0.011605443432927132, 0.12744788825511932, -0.060983046889305115, -0.19573882222175598, 0.4920666217803955, 0.2548990547657013, 0.05156734585762024, -0.46568816900253296, 0.007283506449311972, 0.09706340730190277, 0.13736309111118317, -0.0935613363981247, 0.1530829221010208, 0.4537775218486786, -0.1514500081539154, 0.20709161460399628, 0.004273159429430962, -0.05843540281057358, -0.09915964305400848, 0.08272354304790497, -0.3991204798221588, -0.40854743123054504, 0.5045661330223083, 0.06087705120444298, -0.11606736481189728, 0.08093591034412384, 0.3370765447616577, -0.6575289964675903, -0.260475218296051, 0.34948861598968506, -0.08489373326301575, -0.03731125593185425, -0.22380787134170532, 0.07407914847135544, 0.07273796200752258, 0.6086328029632568, 0.16365867853164673, 0.3771527111530304, -0.39702364802360535, -0.10719912499189377, -0.6464778184890747, -0.21320584416389465, 0.0997607558965683, -0.1670617014169693, -0.2602815628051758, 0.13774150609970093, -0.038988031446933746, 0.2751610279083252, -0.2075420767068863, 0.05163240432739258, 0.25321558117866516, 0.026921281591057777, -0.45630544424057007, 0.07905134558677673, 0.1370767205953598, 0.0201097559183836, 0.28374919295310974, -0.07789784669876099, -0.22131921350955963, 0.06585610657930374, -0.056704044342041016, -0.05049337446689606, -0.08613930642604828, 0.27794787287712097, 0.17560851573944092, 0.03599392622709274, 0.04806513339281082, 0.5123879313468933, -0.022415578365325928, -0.17422173917293549, -0.02602824941277504, 0.02857363596558571, -0.3725864589214325, -0.11181623488664627, -0.12267301976680756, 0.21131892502307892, -0.37687671184539795, 0.27004748582839966, 0.01695701666176319, -0.10174858570098877, 0.12268924713134766, -0.009532656520605087, -0.17501015961170197, -0.021082734689116478, 0.04411768168210983, 0.2265997976064682, 0.21104493737220764, 0.7624980211257935, -0.00015050917863845825, 0.1709391474723816, -0.07620479166507721, -0.06651207059621811, 0.07553748041391373, -0.11749524623155594, -0.26458320021629333, -0.15456560254096985, 0.0648573786020279, 0.04752340540289879, 0.11066454648971558, -0.7056499123573303, -0.3468591272830963, 0.016169268637895584, -0.019177090376615524, 0.05395834520459175, 0.7361038327217102, 0.07266171276569366, -0.14398175477981567, -0.1310943216085434, -0.3231854736804962, 0.02054028958082199, -0.23995469510555267, 0.18913601338863373, -0.3676549792289734 ]
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).
[ -0.030960330739617348, -0.4186744689941406, -0.03965791314840317, -0.04687012732028961, -0.22470249235630035, 0.09376334398984909, 0.16783766448497772, 0.037585847079753876, 0.2795366048812866, -0.16709871590137482, -0.22088007628917694, 0.3342224955558777, -0.21486341953277588, 0.3291553556919098, 0.4196423292160034, -0.4405980110168457, 0.02405460923910141, -0.21915334463119507, -0.2579808831214905, 0.37987810373306274, 0.065569207072258, 0.10199691355228424, -0.1636253297328949, 0.05979228764772415, -0.26799413561820984, -0.16470766067504883, -0.3614473044872284, 0.12241487950086594, 0.13698215782642365, -0.010515410453081131, 0.21925154328346252, 0.3853780925273895, -0.02397031895816326, 0.11684758216142654, -0.00011339464981574565, -0.07660691440105438, 0.1594887226819992, -0.12002448737621307, 0.09109000861644745, 0.006583958864212036, 0.15462055802345276, -0.4277091920375824, 0.07293383032083511, -0.15843729674816132, 0.14493289589881897, 0.15658533573150635, 0.14506758749485016, -0.0069006867706775665, 0.4177415370941162, -0.255648672580719, 0.09768406301736832, 0.3037808835506439, -0.1650868058204651, 0.2707972824573517, -0.14200884103775024, -0.06699219346046448, 0.027930142357945442, 0.2208583503961563, 0.6148960590362549, -0.5899091362953186, 0.0041680708527565, 0.07856716215610504, -0.014349911361932755, 0.007405588403344154, 0.15548290312290192, -0.029883679002523422, -0.3357184827327728, -0.26306936144828796, -0.39329853653907776, 0.21611791849136353, -0.13005810976028442, -0.17696866393089294, -0.25059211254119873, -0.5384417772293091, 0.017066888511180878, 0.03541639447212219, -0.2903541922569275, 0.031573329120874405, -0.2046491801738739, -0.05459118261933327, -0.1233537420630455, -0.10292129218578339, -0.006522908806800842, 0.15645073354244232, 0.3066308796405792, 0.571914553642273, 0.12777642905712128, 0.17128872871398926, 0.3525916635990143, 0.02389596775174141, -0.14905191957950592, -0.2876003086566925, 0.29404017329216003, 0.16211548447608948, -0.4089536666870117, -0.350779265165329, 0.13092613220214844, -0.24935418367385864, 0.2844473719596863, 0.08810119330883026, 0.11836829036474228, 0.05591227114200592, -0.13708598911762238, 0.28888919949531555, 0.3011118173599243, 0.03557530790567398, -0.07668820768594742, -0.1267349272966385, 0.021894332021474838, -0.13308824598789215, 0.18895933032035828, 0.12626680731773376, 0.18178308010101318, -0.008661901578307152, -0.37493807077407837, -0.13579264283180237, -0.1604587286710739, 0.1738724410533905, -0.16109979152679443, -0.5166131258010864, -0.11007818579673767, -0.09324751049280167, 0.07216590642929077, -0.09246727079153061, -0.29434117674827576, 0.07609951496124268, -0.04885469749569893, 0.2247583419084549, -0.08821694552898407, 0.0521588958799839, -0.06363727897405624, 0.08743637800216675, -0.427127480506897, -0.031100252643227577, 0.31112298369407654, 0.0953134298324585, 0.07913778722286224, 0.168320894241333, -0.009065136313438416, 0.10786354541778564, 0.42135104537010193, 0.020736411213874817, -0.005292035639286041, -0.025720451027154922, 0.20524586737155914, -0.2978519797325134, 0.013126444071531296, 0.1394488662481308, -0.34433114528656006, -0.12336943298578262, -0.1408035159111023, -0.16177307069301605, 0.10516715794801712, 0.026167768985033035, -0.22706972062587738, -0.26348572969436646, -0.478713721036911, 0.7096784114837646, -0.01631738245487213, -0.04073471203446388, -0.1800367534160614, 0.28762441873550415, -0.26460060477256775, -0.11316057294607162, 0.22921299934387207, 0.10030773282051086, -0.043902475386857986, -0.2998162806034088, -0.07132072746753693, -0.08353069424629211, -0.08975840359926224, 0.2788395583629608, -0.1998571902513504, 0.22274690866470337, 0.06801249086856842, 0.1548837423324585, 0.3027404546737671, -0.4629935622215271, -0.036578841507434845, 0.26916882395744324, -0.035776667296886444, 0.38466617465019226, 0.0656006932258606, 0.22492584586143494, 0.019382182508707047, -0.20013092458248138, 0.26398831605911255, 1.0421268939971924, -0.4086361527442932, 0.0009426698088645935, -0.09630396962165833, -0.35686030983924866, 0.5225275158882141, 0.3448227047920227, 0.05289360508322716, -0.2893567979335785, -0.1473955512046814, -0.005420439410954714, 0.07239223271608353, 0.058151524513959885, 0.09516756236553192, -0.036987483501434326, -0.3416345417499542, 0.069547139108181, -0.22003597021102905, -0.15002982318401337, -0.44301098585128784, 0.019452519714832306, 0.04457852244377136, 0.343417763710022, 0.2352166473865509, -0.13884490728378296, 0.22767773270606995, -0.29061394929885864, -0.13999690115451813, -0.3124927580356598, 0.1317206770181656, -0.050935596227645874, 0.031070612370967865, -0.2207598239183426, -0.035360001027584076, 0.3621315360069275, 0.14933723211288452, -0.12701550126075745, -0.48026585578918457, 0.327831506729126, -0.07594688981771469, 0.03664093837141991, -0.027286112308502197, 0.6227851510047913, -0.15069270133972168, -0.07301953434944153, 0.1906040608882904, 0.26689162850379944, -0.28597161173820496, 0.2114320993423462, 0.3612278699874878, 0.27097657322883606, 0.17123503983020782, -0.034092649817466736, 0.09529763460159302, 0.023809412494301796, -0.0051742177456617355, -0.22024020552635193, -0.08570583909749985, 0.523391604423523, -0.17810945212841034, 0.4187559187412262, -0.04507885128259659, 0.03660005331039429, -0.135530486702919, -0.01814746856689453, -0.19881552457809448, 0.41928932070732117, 0.2726353406906128, -0.17454449832439423, 0.3594204783439636, -0.1392040103673935, -0.3428288698196411, 0.16008315980434418, 0.15589238703250885, 0.038506489247083664, 0.09078461676836014, -0.0031082406640052795, -0.04850035905838013, -0.2614932954311371, -0.10891598463058472, 0.3499148190021515, 0.6467791199684143, 0.2586943507194519, 0.13757577538490295, 0.013220415450632572, -0.25058871507644653, -0.06948138028383255, 0.03184850513935089, 0.16933591663837433, -0.028486156836152077, 0.14386525750160217, 0.06159069389104843, 0.04908040165901184, 0.029332436621189117, -0.17154915630817413, 0.18374310433864594, -0.16221751272678375, -0.07556957751512527, 0.07238435000181198, -0.16648511588573456, -0.5669125318527222, -0.2283155769109726, 0.02773299813270569, -0.3815072476863861, -0.2889532148838043, 0.1047787070274353, 0.05141313001513481, -0.12023964524269104, 0.18662790954113007, 0.20177417993545532, 0.5722042918205261, -0.3877086043357849, -0.29302436113357544, -0.029145535081624985, -0.3389906585216522, -0.11456451565027237, 0.11243927478790283, 0.4557243883609772, 0.2603592276573181, 0.4842230975627899, 0.3259252607822418, -0.03397265449166298, 0.0463758260011673, -0.39842456579208374, 0.017644278705120087, -0.3012903332710266, 0.2676166594028473, 0.07377395778894424, -0.18586845695972443, 0.15339384973049164, -0.40689191222190857, -0.005612120032310486, -0.021065721288323402, -0.07087572664022446, -0.05289333313703537, 0.004416038282215595, 0.09541960060596466, -0.18347781896591187, -0.19025975465774536, -0.6412384510040283, -0.43016278743743896, 0.2673931121826172, -0.26016369462013245, 0.08659752458333969, 0.014041760936379433, -0.2555754482746124, 0.17967817187309265, -0.10786710679531097, -0.0026078885421156883, -0.16347719728946686, -0.038953542709350586, 0.04661323130130768, -0.35634130239486694, -0.17569808661937714, -0.3697788119316101, -0.1142725795507431, 0.24605382978916168, -0.06947679817676544, -0.024847764521837234, -0.2876040041446686, 0.09452758729457855, -0.11205729842185974, 0.05305677652359009, 0.20907112956047058, 0.19394834339618683, -0.22071023285388947, 0.13459806144237518, -0.07369859516620636, -0.29403573274612427, 0.23027512431144714, 0.27270710468292236, 0.38758090138435364, 0.1529558002948761, 0.07122115045785904, 0.17559175193309784, 0.7356224656105042, 0.25420889258384705, 0.018226033076643944, 0.15264949202537537, 0.25192463397979736, -0.008517477661371231, 0.17827630043029785, -0.24685533344745636, 0.17666278779506683, -0.15473127365112305, 0.48757994174957275, 0.09875655174255371, 0.150802344083786, -0.24501647055149078, -0.3062400817871094, 0.05892613157629967, -0.2685535252094269, -0.2171599119901657, 0.24243071675300598, -0.22875815629959106, 0.014071222394704819, -0.08966025710105896, -0.4162385165691376, -0.2906377613544464, -0.053883858025074005, -0.022309914231300354, 0.3469157814979553, -0.07768469303846359, 0.157317653298378, -0.34134039282798767, -0.046045079827308655, -0.4354320168495178, 0.14782822132110596, 0.07338887453079224, 0.28789249062538147, -0.06971289962530136, -0.024439603090286255, -0.11077514290809631, 0.10292797535657883, 0.6152392029762268, -0.5185810327529907, -0.12251298129558563, 0.10406072437763214, 0.08391129970550537, -0.1375814527273178, -0.10161551833152771, -0.2601819336414337, 0.014867657795548439, 0.4297337532043457, 0.3357178568840027, -0.3773830831050873, -0.2934283912181854, 0.17445799708366394, -0.20981255173683167, -0.0844765156507492, -0.16246354579925537, -0.2532844543457031, -0.22844791412353516, -0.12505650520324707, 0.041776541620492935, 0.22972021996974945, 0.2200998067855835, -0.06440145522356033, 0.16535009443759918, -0.10738156735897064, 0.20170798897743225, 0.21851877868175507, 0.023033052682876587, -0.040012627840042114, 0.2556941509246826, 0.1977343112230301, 0.039545901119709015, -0.12689794600009918, -0.3682379722595215, 0.06594017893075943, -0.05863915756344795, 0.06726761162281036, 0.25432831048965454, -0.03158362954854965, 0.6494699120521545, 0.04232773929834366, 0.339296817779541, 0.2956755757331848, -0.171039417386055, 0.00383719801902771, -0.227918803691864, 0.16946972906589508, 0.22283972799777985, 0.22637011110782623, -0.15907901525497437, -0.26187992095947266, 0.562938392162323, 0.29968640208244324, -0.07139334082603455, 0.451677531003952, -0.4898034334182739, -0.3304513394832611, 0.2624640464782715, 0.0708131194114685, 0.6956494450569153, -0.0626806989312172, 0.29476526379585266, 0.07930435985326767, -0.36408501863479614, 0.6168705224990845, -0.3644068241119385, 0.1506238430738449, -0.1062217652797699, -0.19690929353237152, 0.012675641104578972, -0.09358709305524826, -0.0731005147099495, 0.2685149312019348, -0.38254672288894653, 0.3077969253063202, 0.42872554063796997, -0.20226317644119263, -0.025360627099871635, 0.5634380578994751, 0.05623336881399155, -0.3985060155391693, -0.15839296579360962, 0.06879620999097824, -0.3081548810005188, 0.0543656125664711, -0.1665422022342682, -0.2580891251564026, 0.22570592164993286, 0.15583230555057526, -0.3013504147529602, 0.06678961217403412, -0.03523048013448715, 0.2177829146385193, 0.21286426484584808, 0.015265248715877533, 0.26790186762809753, -0.11605891585350037, -0.037850163877010345, -0.15350617468357086, -0.08397115767002106, 0.13552787899971008, -0.19333329796791077, 0.0378146730363369, 0.01826782338321209, -0.1867406666278839, 0.4988887310028076, -0.06247975677251816, -0.11852211505174637, -0.21425199508666992, 0.28020739555358887, -0.4461520314216614, -0.19157014787197113, 0.14757505059242249, 0.1756839156150818, 0.1982739418745041, -0.3049491345882416, 0.5695933699607849, 0.10945115983486176, 0.011156491935253143, 0.0341043621301651, 0.2101193368434906, -0.27115949988365173, 0.5002744793891907, -0.2811298668384552, 0.009963221848011017, 0.10484032332897186, 0.02505221962928772, -0.11265336722135544, -0.38574928045272827, -0.030786365270614624, 0.31705957651138306, 0.13859906792640686, 0.008726745843887329, 0.028529450297355652, 0.11237525194883347, -0.06501572579145432, -0.02706974744796753, -0.14490534365177155, -0.35135406255722046, 0.131876140832901, 0.18006372451782227, 0.18757539987564087, 0.2329903244972229, -0.028973333537578583, -0.2715376913547516, -0.16603974997997284, 0.44810593128204346, 0.03526157885789871, 0.20998844504356384, -0.1294512003660202, 0.34659919142723083, 0.40914109349250793, 0.5483714938163757, -0.2595946192741394, 0.07455826550722122, 0.055365972220897675, 0.11097997426986694, 0.459245502948761, -0.2511613667011261, 0.076079823076725, 0.2771490514278412, -0.12474247813224792, 0.03135416656732559, -0.3204188942909241, -0.1535581350326538, -0.24974076449871063, 0.1109316274523735, 0.10442157834768295, 0.0042015002109110355, 0.004017968662083149, -0.10565099865198135, -0.11311247944831848, -0.08425325900316238, -0.00029867514967918396, -0.3816279470920563, -0.05959092080593109, 0.12420295178890228, -0.03806604817509651, 0.0650605633854866, 0.036878831684589386, 0.18779055774211884, 0.023436088114976883, -0.10253900289535522, 0.34939655661582947, 0.11723890900611877, 0.10813349485397339, -0.1778067648410797, -0.16322247684001923, 0.023343484848737717, 0.2428969442844391, 0.15198221802711487, -0.06412412226200104, 0.12868019938468933, -0.24157720804214478, -0.12066946178674698, 0.236313134431839, 0.16919121146202087, 0.017447862774133682, -0.1027127355337143, 0.13327622413635254, 0.11498117446899414, -0.09520798921585083, -0.03510065749287605, 0.27450206875801086, 0.07827584445476532, 0.134038507938385, 0.14364203810691833, 0.09875935316085815, 0.4728016257286072, -0.23880432546138763, 0.009598851203918457, 0.14305885136127472, -0.35201001167297363, 0.12352623045444489, -0.1303715705871582, -0.03644517436623573, -0.011555921286344528, 0.10323746502399445, -0.024381965398788452, 0.141191303730011, -0.2700303792953491, 0.28077515959739685, 0.30520468950271606, 0.000443827360868454, -0.03391337767243385, 0.2237342894077301, 0.05606946721673012, -0.14373500645160675, 0.2464371919631958, 0.04788047820329666, 0.32649004459381104, 0.021484695374965668, 0.8898254632949829, -0.25932177901268005, -0.4453703463077545, 0.10914941877126694, 0.4276833236217499, 0.04372840002179146, -0.20927919447422028, 0.1428680717945099, 0.10250184684991837, -0.3528323173522949, 0.07324333488941193, 0.0668150931596756, -0.21987570822238922, 0.650164008140564, -0.10448980331420898, 0.21174612641334534, -0.13323836028575897, -0.16710615158081055, 0.3940386176109314, 0.17511966824531555, -0.06507784128189087, 0.0157078355550766, 0.14367759227752686, -0.12336216866970062, -0.17897970974445343, 0.43864330649375916, 0.21286191046237946, 0.027808908373117447, -0.46195387840270996, 0.0628170520067215, 0.09056828916072845, 0.14693021774291992, -0.03955453634262085, 0.10659772902727127, 0.43498724699020386, -0.1517581194639206, 0.23860199749469757, 0.013740068301558495, -0.05349789932370186, -0.19473107159137726, 0.07981115579605103, -0.42988622188568115, -0.378812700510025, 0.4788786768913269, 0.06758509576320648, -0.11324054002761841, 0.05943562462925911, 0.33388450741767883, -0.6175366044044495, -0.2591392397880554, 0.3267904818058014, -0.09543348848819733, -0.06627240777015686, -0.1992466002702713, 0.07281956076622009, 0.06321091949939728, 0.6397793292999268, 0.17726364731788635, 0.36059918999671936, -0.3750837743282318, -0.12910102307796478, -0.6346372961997986, -0.15605154633522034, 0.11392465233802795, -0.17616349458694458, -0.25313717126846313, 0.1707487851381302, -0.04700475186109543, 0.3227485418319702, -0.20954696834087372, 0.09758217632770538, 0.2715473473072052, 0.027108391746878624, -0.4724886417388916, 0.0400509275496006, 0.09500536322593689, 0.04796040430665016, 0.28179454803466797, -0.07571332156658173, -0.19103200733661652, 0.06382099539041519, -0.06736119091510773, -0.08385732024908066, -0.11887842416763306, 0.2494056522846222, 0.17716890573501587, 0.06206147000193596, 0.0344526432454586, 0.5532029867172241, 0.010818343609571457, -0.1616266816854477, 0.031095368787646294, 0.028595712035894394, -0.3770715296268463, -0.1337224245071411, -0.08216693997383118, 0.20839737355709076, -0.39294594526290894, 0.2529560923576355, -0.007655319292098284, -0.025235764682292938, 0.11938372254371643, 0.011496465653181076, -0.2017444372177124, -0.02404690906405449, 0.08408021181821823, 0.24017967283725739, 0.20282059907913208, 0.7438056468963623, -0.01443392038345337, 0.17315666377544403, -0.08187657594680786, -0.03745286911725998, 0.06602832674980164, -0.13763946294784546, -0.24784840643405914, -0.13568419218063354, 0.03555174171924591, 0.06208350509405136, 0.12089323252439499, -0.676857054233551, -0.3293685019016266, 0.045108404010534286, -0.0008667614310979843, 0.004000037908554077, 0.7409059405326843, 0.06504139304161072, -0.11601370573043823, -0.11704592406749725, -0.3568640649318695, 0.033418286591768265, -0.25153300166130066, 0.18174003064632416, -0.37132352590560913 ]
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`.
[ -0.08343673497438431, -0.37729135155677795, -0.0540122427046299, -0.08327743411064148, -0.2122468650341034, 0.016944333910942078, 0.13349920511245728, 0.06816373765468597, 0.31245875358581543, -0.0252598375082016, -0.23778603971004486, 0.3599814176559448, -0.15870383381843567, 0.3497789800167084, 0.3823589086532593, -0.4798983037471771, -0.0307442769408226, -0.13304588198661804, -0.368801474571228, 0.3823729157447815, 0.04090229421854019, 0.09933474659919739, -0.12265481799840927, 0.0074624791741371155, -0.2362724095582962, -0.15221793949604034, -0.34329670667648315, 0.09290042519569397, 0.18456219136714935, -0.055031318217515945, 0.19634635746479034, 0.44995513558387756, -0.017280213534832, 0.052321407943964005, -0.00011072367487940937, -0.08731884509325027, 0.13175973296165466, -0.12002041190862656, 0.049835313111543655, 0.03783103823661804, 0.21656614542007446, -0.4117274880409241, -0.06971552222967148, -0.18803787231445312, 0.10876727849245071, 0.21512407064437866, 0.18305854499340057, 0.024411894381046295, 0.41552120447158813, -0.32108616828918457, 0.1293700933456421, 0.32684794068336487, -0.22198718786239624, 0.23453129827976227, -0.09464385360479355, -0.15330490469932556, 0.09412934631109238, 0.3127889931201935, 0.4916841983795166, -0.5477129817008972, -0.06117916852235794, 0.14151130616664886, 0.0077215563505887985, 0.059617672115564346, 0.10257229208946228, -0.051516398787498474, -0.3251841366291046, -0.20651036500930786, -0.3548204004764557, 0.29408255219459534, -0.18114712834358215, -0.14917248487472534, -0.19877654314041138, -0.5856704711914062, 0.06005871295928955, -0.010826671496033669, -0.2997104227542877, 0.02046133577823639, -0.14967502653598785, -0.01279304176568985, -0.12954580783843994, -0.09831427037715912, 0.07478511333465576, 0.09907751530408859, 0.32749229669570923, 0.5603529214859009, 0.09386050701141357, 0.18363526463508606, 0.39189448952674866, 0.07738843560218811, -0.21224841475486755, -0.246026873588562, 0.2851274013519287, 0.09630567580461502, -0.28267213702201843, -0.3462059199810028, 0.1630009114742279, -0.26397940516471863, 0.23295828700065613, 0.05783981829881668, 0.1677578091621399, 0.11276699602603912, -0.13788184523582458, 0.2494640052318573, 0.28214171528816223, 0.05985120311379433, 0.010651610791683197, -0.11803437769412994, 0.01498420536518097, -0.05733748525381088, 0.1585502028465271, 0.09276801347732544, 0.21617823839187622, -0.04497029632329941, -0.40501436591148376, -0.08261941373348236, -0.2754175066947937, 0.1954956352710724, -0.20601993799209595, -0.5297425389289856, -0.018626544624567032, -0.11480327695608139, 0.023904889822006226, -0.10028742998838425, -0.3187437355518341, 0.16014358401298523, -0.056539006531238556, 0.1746998280286789, -0.1174469143152237, 0.07287582010030746, -0.08563584089279175, 0.08851851522922516, -0.49863627552986145, 0.02519451640546322, 0.25986430048942566, 0.08004923164844513, 0.16220256686210632, 0.16258688271045685, -0.026162348687648773, 0.1506979614496231, 0.39101892709732056, 0.059601109474897385, 0.0018358007073402405, -0.058130573481321335, 0.159010648727417, -0.2949698865413666, -0.042152438312768936, 0.1823381781578064, -0.42464199662208557, -0.15480473637580872, -0.1469176709651947, -0.07604150474071503, 0.16089080274105072, 0.04554303362965584, -0.24469687044620514, -0.2981192469596863, -0.4465663731098175, 0.7699316143989563, -0.009562753140926361, -0.1129106730222702, -0.10955851525068283, 0.20794659852981567, -0.30656173825263977, -0.0800890251994133, 0.2194678783416748, 0.032092414796352386, -0.01905878446996212, -0.40339744091033936, -0.04226965457201004, -0.037179749459028244, -0.10395515710115433, 0.2644660174846649, -0.21608860790729523, 0.18137189745903015, 0.07405565679073334, 0.17394699156284332, 0.2999929189682007, -0.4679257273674011, -0.08489702641963959, 0.23441095650196075, -0.036087457090616226, 0.3289673626422882, 0.08498428016901016, 0.12413075566291809, 0.05196559429168701, -0.17159166932106018, 0.17609819769859314, 0.9709120392799377, -0.3548182249069214, 0.024371890351176262, -0.1063789427280426, -0.3863513469696045, 0.5253020524978638, 0.2912898063659668, 0.07079442590475082, -0.2695537209510803, -0.12689140439033508, 0.004664952866733074, 0.09227924048900604, 0.01856328919529915, 0.08767051994800568, -0.04912332445383072, -0.3308199942111969, 0.00919756293296814, -0.2163790911436081, -0.15521496534347534, -0.3733971118927002, 0.05382620543241501, -0.061897244304418564, 0.3879176378250122, 0.39867132902145386, -0.1182982549071312, 0.2000267207622528, -0.3578914999961853, -0.033523861318826675, -0.28234654664993286, 0.1489844024181366, -0.013988636434078217, 0.04843536391854286, -0.220911905169487, -0.049511704593896866, 0.23565392196178436, 0.135026752948761, -0.1555178165435791, -0.40747788548469543, 0.30450940132141113, 0.020174410194158554, -0.0015047937631607056, -0.05420171096920967, 0.5833069086074829, -0.1648831069469452, -0.10873427987098694, 0.2087189257144928, 0.3047448396682739, -0.3626171946525574, 0.29410797357559204, 0.31045353412628174, 0.19204077124595642, 0.15806689858436584, 0.019926097244024277, 0.1577107310295105, -0.014536786824464798, -0.06476853042840958, -0.16084536910057068, -0.041563235223293304, 0.641684353351593, -0.04830150678753853, 0.34820738434791565, -0.029241543263196945, -0.031101044267416, -0.17479848861694336, -0.1673583984375, -0.25965818762779236, 0.3309371769428253, 0.2930154800415039, -0.27841484546661377, 0.3624514043331146, -0.12749820947647095, -0.4090743660926819, 0.15859650075435638, 0.19831007719039917, 0.05338207259774208, 0.12208328396081924, 0.011676155030727386, -0.05402829498052597, -0.30141010880470276, -0.10330075025558472, 0.34477925300598145, 0.6520266532897949, 0.28751140832901, 0.168321892619133, 0.02160434052348137, -0.2816527485847473, -0.0925324335694313, 0.0023983940482139587, 0.13403041660785675, -0.02989664115011692, 0.18200238049030304, 0.10243838280439377, 0.031474776566028595, -0.08563199639320374, -0.06431353092193604, 0.1644388884305954, -0.21713696420192719, -0.03231696039438248, 0.03986576944589615, -0.1786455363035202, -0.5128587484359741, -0.29099470376968384, 0.035596031695604324, -0.3008247911930084, -0.30237075686454773, 0.180471733212471, 0.0911056250333786, -0.22132958471775055, 0.21164461970329285, 0.23787373304367065, 0.6204198598861694, -0.39749449491500854, -0.1298801749944687, -0.10952049493789673, -0.346415251493454, -0.10274597257375717, 0.12991203367710114, 0.3555278778076172, 0.2745726406574249, 0.4790083169937134, 0.2562704086303711, -0.058410752564668655, 0.1019943505525589, -0.40407466888427734, 0.026286358013749123, -0.31978166103363037, 0.25025177001953125, 0.11853522062301636, -0.18916471302509308, 0.15250933170318604, -0.3981904983520508, 0.011865782551467419, -0.14020149409770966, -0.029884323477745056, -0.059307198971509933, 0.011886273510754108, 0.1486315131187439, -0.20273490250110626, -0.2027556300163269, -0.6083055734634399, -0.42396363615989685, 0.2456139773130417, -0.21168050169944763, 0.04542766883969307, 0.051857348531484604, -0.2142132669687271, 0.14821317791938782, -0.16540053486824036, 0.015459399670362473, -0.1877848356962204, -0.10263331979513168, -0.006060328334569931, -0.3906427025794983, -0.10505229234695435, -0.5042768120765686, -0.12487228959798813, 0.2693501114845276, -0.10490936785936356, -0.10202071070671082, -0.32057926058769226, 0.05709948390722275, -0.09570686519145966, 0.01693490333855152, 0.2777612507343292, 0.10743658244609833, -0.22773580253124237, 0.03465273976325989, -0.06954748928546906, -0.21417123079299927, 0.21847206354141235, 0.41094449162483215, 0.3875404894351959, 0.11120001971721649, 0.003796897828578949, 0.14250244200229645, 0.7288059592247009, 0.20422549545764923, -0.006215959787368774, 0.24172458052635193, 0.2230886071920395, 0.059005945920944214, 0.1169361025094986, -0.14880691468715668, 0.21649548411369324, -0.13906502723693848, 0.5181334018707275, 0.13408225774765015, 0.14490383863449097, -0.23636023700237274, -0.26169705390930176, 0.001547597348690033, -0.20625880360603333, -0.24861380457878113, 0.2672934830188751, -0.2329823076725006, -0.03897733986377716, -0.04308027774095535, -0.35451382398605347, -0.32712942361831665, -0.04302404820919037, 0.021407146006822586, 0.16934068500995636, -0.009691014885902405, 0.13491588830947876, -0.44665461778640747, 0.007309667766094208, -0.4875246286392212, 0.17486964166164398, 0.03552931174635887, 0.20631295442581177, -0.014624364674091339, -0.09552451968193054, -0.13732275366783142, 0.10531467944383621, 0.49268484115600586, -0.5116745829582214, -0.0789996087551117, 0.08137321472167969, 0.06042468547821045, -0.1340131163597107, -0.1311509609222412, -0.3227555453777313, -0.07703298330307007, 0.44155260920524597, 0.37745487689971924, -0.42304694652557373, -0.3025462031364441, 0.22415241599082947, -0.16462092101573944, -0.09548047184944153, -0.17159664630889893, -0.21112145483493805, -0.1963445544242859, -0.15463081002235413, 0.07413271814584732, 0.14330701529979706, 0.26712173223495483, -0.1253158301115036, 0.19032739102840424, -0.10813811421394348, 0.18458861112594604, 0.17987221479415894, 0.015827741473913193, -0.04419182613492012, 0.35392293334007263, 0.11021867394447327, 0.07063794136047363, -0.1312553584575653, -0.331958532333374, 0.10548806190490723, -0.1351999193429947, 0.20508211851119995, 0.22697043418884277, 0.09067465364933014, 0.5530034899711609, 0.11436168849468231, 0.35768502950668335, 0.30341091752052307, -0.06347424536943436, 0.12543818354606628, -0.31384581327438354, 0.20288746058940887, 0.17584112286567688, 0.255044549703598, -0.10684063285589218, -0.2142479568719864, 0.5952865481376648, 0.28747767210006714, -0.08879175782203674, 0.3823321461677551, -0.5090104341506958, -0.3641585409641266, 0.2897244095802307, 0.08487685024738312, 0.656855583190918, -0.01266615092754364, 0.3372664153575897, 0.05917869508266449, -0.304044246673584, 0.619907796382904, -0.275953471660614, 0.14735527336597443, -0.13500432670116425, -0.24565301835536957, 0.03667672723531723, -0.04864157363772392, -0.08241482824087143, 0.2677815854549408, -0.35127562284469604, 0.3676716387271881, 0.4438672661781311, -0.1998153030872345, -0.12936098873615265, 0.5204607844352722, -0.08692405372858047, -0.3249291777610779, -0.11418372392654419, 0.09938503801822662, -0.2334737628698349, 0.11657506227493286, -0.1201619803905487, -0.25054484605789185, 0.22407999634742737, 0.13571172952651978, -0.22700893878936768, 0.04554489254951477, -0.0658809021115303, 0.13693293929100037, 0.26868557929992676, 0.030229032039642334, 0.19252631068229675, -0.2200113832950592, -0.025525979697704315, -0.10796047747135162, -0.10625419020652771, 0.12955453991889954, -0.1636422872543335, 0.07055357843637466, 0.13920392096042633, -0.1475970298051834, 0.519082248210907, 0.012029826641082764, -0.11477293074131012, -0.19632449746131897, 0.26937335729599, -0.3825450539588928, -0.17794661223888397, 0.024098139256238937, 0.09247929602861404, 0.10347151011228561, -0.3090542256832123, 0.5627153515815735, 0.12510044872760773, -0.0283498615026474, 0.06166983023285866, 0.10210748761892319, -0.22989562153816223, 0.5123823285102844, -0.18565776944160461, -0.03955163061618805, 0.05415235832333565, 0.09490837901830673, -0.162653386592865, -0.3041132986545563, -0.04040642827749252, 0.24128533899784088, 0.1346210241317749, -0.014453219249844551, 0.014386706054210663, 0.16485227644443512, -0.17030954360961914, 0.04330530762672424, -0.07534758746623993, -0.34250274300575256, 0.15653546154499054, 0.2247285395860672, 0.21894203126430511, 0.2962503433227539, -0.11109837889671326, -0.185403510928154, -0.12255635112524033, 0.4910305142402649, 0.09181749820709229, 0.2689322829246521, -0.11989662051200867, 0.32257211208343506, 0.35474756360054016, 0.5900279879570007, -0.3123571574687958, 0.07766600698232651, 0.06387640535831451, 0.1348428875207901, 0.5099009275436401, -0.23741289973258972, 0.008969585411250591, 0.3305620849132538, -0.11035560816526413, 0.07590006291866302, -0.32956793904304504, -0.17685595154762268, -0.2944236099720001, 0.08765851706266403, 0.07303470373153687, 0.01262526586651802, 0.02389490231871605, -0.03426353633403778, -0.045068588107824326, -0.1670873761177063, 0.005586780607700348, -0.41685113310813904, -0.10029646754264832, 0.20000974833965302, -0.06102355197072029, 0.06438520550727844, 0.1422736495733261, 0.22846247255802155, 0.009681642055511475, -0.1569228619337082, 0.2438974529504776, 0.16898371279239655, 0.10425110906362534, -0.12324473261833191, -0.1688486933708191, -0.0028999894857406616, 0.21621885895729065, 0.08830443024635315, -0.12145934253931046, 0.19059349596500397, -0.20523498952388763, -0.09061136096715927, 0.3178473114967346, 0.18736661970615387, 0.04893610253930092, -0.1805138736963272, 0.1527685821056366, 0.16645275056362152, -0.11254394799470901, 0.00039309359272010624, 0.15967324376106262, 0.11689897626638412, 0.15863487124443054, 0.14670749008655548, 0.14573132991790771, 0.38194525241851807, -0.1910199075937271, -0.041007544845342636, 0.08764426410198212, -0.32143473625183105, 0.11080721020698547, -0.060433123260736465, -0.020613621920347214, -0.03212401643395424, 0.18058408796787262, -0.0632687360048294, 0.20519569516181946, -0.4029552936553955, 0.2911439836025238, 0.33818021416664124, -0.008856095373630524, 0.030189277604222298, 0.12258760631084442, 0.011403318494558334, -0.13385038077831268, 0.3777519166469574, 0.05748779699206352, 0.3323083221912384, 0.05188056826591492, 0.9268810749053955, -0.26295799016952515, -0.4481849670410156, 0.06358546018600464, 0.531395673751831, 0.029138321056962013, -0.3214201331138611, 0.06138929724693298, 0.06478064507246017, -0.3745325803756714, 0.16029533743858337, 0.011612344533205032, -0.1540061980485916, 0.6641727685928345, -0.15340936183929443, 0.19332930445671082, -0.16817286610603333, -0.22600221633911133, 0.28352779150009155, 0.1881144493818283, -0.1497504562139511, 0.044861018657684326, 0.07804491370916367, -0.06418494880199432, -0.12437944114208221, 0.5293444395065308, 0.17358122766017914, -0.013322815299034119, -0.4142281711101532, -0.0032766656950116158, 0.14429627358913422, 0.16323961317539215, -0.12115471810102463, 0.12707525491714478, 0.4776846468448639, -0.17132806777954102, 0.2919730842113495, 0.0513656809926033, -0.1416742503643036, -0.16505253314971924, 0.0537252277135849, -0.4065130352973938, -0.4111208915710449, 0.4177228808403015, 0.13737979531288147, -0.04147929698228836, 0.06708236783742905, 0.28556138277053833, -0.5498070120811462, -0.22149813175201416, 0.25499197840690613, -0.17315132915973663, -0.07097940146923065, -0.22309298813343048, 0.09759379178285599, 0.044758934527635574, 0.6479761004447937, 0.2121247947216034, 0.252043753862381, -0.35490432381629944, -0.09994027018547058, -0.5510979890823364, -0.13988235592842102, 0.1789744794368744, -0.10730129480361938, -0.27738454937934875, 0.16404403746128082, -0.03695596009492874, 0.27869564294815063, -0.18398259580135345, 0.18127024173736572, 0.21143768727779388, 0.009155172854661942, -0.4085698425769806, 0.1516011357307434, 0.018774323165416718, 0.041605498641729355, 0.31257539987564087, -0.11525028198957443, -0.15051111578941345, 0.021374668926000595, -0.02663564123213291, -0.04524409770965576, -0.08165740966796875, 0.24790722131729126, 0.18512989580631256, 0.1656785011291504, 0.013480477035045624, 0.4840942621231079, -0.07819250226020813, -0.12652689218521118, -0.02078770473599434, -0.0746103897690773, -0.4193073511123657, -0.20675155520439148, -0.09964479506015778, 0.21812179684638977, -0.279577374458313, 0.3174080550670624, -0.009172753430902958, -0.11152945458889008, 0.15208293497562408, 0.06763426214456558, -0.10436652600765228, -0.05300787836313248, 0.11525581032037735, 0.09159116446971893, 0.19210441410541534, 0.7487508654594421, 0.06614496558904648, 0.0984087884426117, -0.08251773566007614, -0.11334868520498276, 0.0889340415596962, -0.06935074925422668, -0.18787801265716553, -0.1619156152009964, 0.012331299483776093, 0.10883492976427078, 0.0035943370312452316, -0.6421046257019043, -0.30900102853775024, 0.0285690538585186, 0.04078555852174759, 0.03615572676062584, 0.7093765139579773, 0.05208500474691391, -0.16649790108203888, -0.12391172349452972, -0.34871944785118103, 0.08365851640701294, -0.33575519919395447, 0.06608189642429352, -0.37589794397354126 ]
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.
[ -0.08342069387435913, -0.3881210684776306, -0.05889216065406799, -0.057947948575019836, -0.2545417547225952, 0.06940446048974991, 0.19493108987808228, 0.05140962824225426, 0.3405598998069763, -0.14826445281505585, -0.2024366557598114, 0.3717307448387146, -0.23944707214832306, 0.34754085540771484, 0.3872973918914795, -0.4599965512752533, 0.024137496948242188, -0.26757046580314636, -0.26987913250923157, 0.372872531414032, 0.06227091699838638, 0.08812969923019409, -0.18115802109241486, 0.08036620914936066, -0.253978431224823, -0.16344726085662842, -0.34938856959342957, 0.10462121665477753, 0.15728846192359924, -0.013106939382851124, 0.21203763782978058, 0.38126319646835327, -0.04049615561962128, 0.13177934288978577, -0.00011317195458104834, -0.07457607984542847, 0.21052531898021698, -0.07623793929815292, 0.08656489849090576, -0.06524446606636047, 0.22962728142738342, -0.4317649304866791, 0.036712341010570526, -0.14333145320415497, 0.14059683680534363, 0.15172182023525238, 0.16424933075904846, -0.007431387901306152, 0.3844814896583557, -0.27362796664237976, 0.1117820143699646, 0.2874106466770172, -0.15955950319766998, 0.26453638076782227, -0.10869453102350235, -0.04743973910808563, 0.04901283234357834, 0.26195451617240906, 0.6061145067214966, -0.6340843439102173, -0.014124855399131775, 0.10082323849201202, -0.04980574548244476, 0.03194368630647659, 0.1234845295548439, -0.08227281272411346, -0.38277190923690796, -0.23529961705207825, -0.39225131273269653, 0.24548795819282532, -0.17746689915657043, -0.19453588128089905, -0.22744829952716827, -0.5580011010169983, 0.006252024322748184, 0.041994303464889526, -0.3048388659954071, 0.05022188276052475, -0.19981540739536285, -0.06731575727462769, -0.13756799697875977, -0.0884879007935524, 0.011829979717731476, 0.17321613430976868, 0.3156491816043854, 0.5342774391174316, 0.10947057604789734, 0.17463421821594238, 0.34226953983306885, 0.042830877006053925, -0.1386733502149582, -0.2869671583175659, 0.34202083945274353, 0.1264592707157135, -0.3712506592273712, -0.3075862526893616, 0.1824488788843155, -0.26114827394485474, 0.25415700674057007, 0.11465080082416534, 0.11636269837617874, 0.0927191823720932, -0.12476164847612381, 0.2851978838443756, 0.3020440638065338, 0.015004457905888557, -0.08135757595300674, -0.08082705736160278, 0.0235547237098217, -0.13282251358032227, 0.22979426383972168, 0.12041135877370834, 0.16450080275535583, -0.030592776834964752, -0.4087803363800049, -0.06945730000734329, -0.16123726963996887, 0.19764044880867004, -0.2383095771074295, -0.4846024811267853, -0.14205893874168396, -0.08315498381853104, 0.07923724502325058, -0.09192671626806259, -0.2979579269886017, 0.10018016397953033, -0.0730477049946785, 0.2468690127134323, -0.10002531111240387, 0.07183870673179626, -0.0563451312482357, 0.09835048019886017, -0.4394397735595703, -0.06991691142320633, 0.32330963015556335, 0.1422458440065384, 0.1164548471570015, 0.1751137524843216, -0.016917668282985687, 0.1301395744085312, 0.44203123450279236, -0.001940365880727768, 0.02095779776573181, -0.014059159904718399, 0.19523228704929352, -0.3251470923423767, 0.004884026944637299, 0.22755245864391327, -0.37265825271606445, -0.14289042353630066, -0.12809118628501892, -0.15609100461006165, 0.11269738525152206, 0.03565201163291931, -0.23599132895469666, -0.2588002383708954, -0.4986611306667328, 0.6829741597175598, -0.029003135859966278, -0.04187577962875366, -0.15797123312950134, 0.2522135376930237, -0.24844467639923096, -0.12514978647232056, 0.21692107617855072, 0.06579447537660599, -0.0024429522454738617, -0.31193840503692627, -0.12970860302448273, -0.08177660405635834, -0.03710111230611801, 0.22530728578567505, -0.160078763961792, 0.2730770409107208, 0.0468321219086647, 0.20835669338703156, 0.2641535997390747, -0.5000428557395935, -0.04386594146490097, 0.2669200003147125, -0.04217652231454849, 0.4183538258075714, 0.04843825474381447, 0.21280845999717712, 0.034804292023181915, -0.16171272099018097, 0.2985629439353943, 1.0328640937805176, -0.38605228066444397, 0.004557004198431969, -0.09491436183452606, -0.36517012119293213, 0.523317277431488, 0.33920562267303467, 0.0907391905784607, -0.31951484084129333, -0.1159665659070015, -0.05300162360072136, 0.10058605670928955, 0.0858922004699707, 0.07345293462276459, -0.07229658216238022, -0.36503931879997253, 0.05978412181138992, -0.2487354576587677, -0.19382643699645996, -0.3421441316604614, 0.03035864233970642, 0.11190621554851532, 0.2797664403915405, 0.24866735935211182, -0.13200034201145172, 0.22297078371047974, -0.25494569540023804, -0.1298157125711441, -0.3390694558620453, 0.141840398311615, -0.059791356325149536, 0.03747944161295891, -0.25106966495513916, 0.0015004873275756836, 0.34114083647727966, 0.1274174451828003, -0.1505715250968933, -0.47374796867370605, 0.28402599692344666, -0.055938661098480225, 0.039368610829114914, -0.022019052878022194, 0.6338948607444763, -0.1646694540977478, -0.0881076231598854, 0.1917855441570282, 0.28774401545524597, -0.26854750514030457, 0.2172980159521103, 0.31826168298721313, 0.1692703366279602, 0.14662028849124908, -0.02283206582069397, 0.09416678547859192, 0.062102653086185455, -0.02409166470170021, -0.21376866102218628, -0.08747979253530502, 0.536893367767334, -0.15059490501880646, 0.439378559589386, -0.06569953262805939, 0.06121225655078888, -0.1688978224992752, -0.01765916496515274, -0.18787500262260437, 0.4131902754306793, 0.3052521049976349, -0.1949424296617508, 0.3491774797439575, -0.18313732743263245, -0.3764328956604004, 0.15709848701953888, 0.08929839730262756, 0.029830005019903183, 0.09827025234699249, -0.043594785034656525, -0.05164376646280289, -0.2501388192176819, -0.06704864650964737, 0.3425806164741516, 0.6638853549957275, 0.26290982961654663, 0.18624214828014374, 0.044392190873622894, -0.24098923802375793, -0.061129119247198105, 0.008607722818851471, 0.1747158169746399, -0.08582823723554611, 0.16375651955604553, 0.05545458942651749, 0.033264581114053726, 0.049514539539813995, -0.13218633830547333, 0.15407155454158783, -0.18271619081497192, -0.07683469355106354, 0.09348670393228531, -0.20606431365013123, -0.5439789295196533, -0.1966906040906906, 0.022410670295357704, -0.2867784798145294, -0.2818972170352936, 0.09111475944519043, 0.09463920444250107, -0.1386154294013977, 0.15519262850284576, 0.213479146361351, 0.5341398119926453, -0.3972027003765106, -0.29688242077827454, -0.021574970334768295, -0.34398654103279114, -0.13003480434417725, 0.1255968064069748, 0.43872177600860596, 0.33325695991516113, 0.5198891758918762, 0.3033435642719269, -0.08945310860872269, 0.047933146357536316, -0.3372720777988434, 0.010944869369268417, -0.2858090400695801, 0.285250723361969, 0.13329178094863892, -0.24219299852848053, 0.1681118756532669, -0.3485582172870636, -0.025232821702957153, -0.07369424402713776, -0.08946512639522552, -0.03885030001401901, -0.0026784262154251337, 0.07000210881233215, -0.17970307171344757, -0.17550259828567505, -0.6369869112968445, -0.43897488713264465, 0.29911163449287415, -0.2640557587146759, 0.09133274853229523, -0.016800738871097565, -0.2485501617193222, 0.18541932106018066, -0.1101645976305008, 0.0057690260000526905, -0.17779715359210968, -0.09301428496837616, 0.026734184473752975, -0.3839552402496338, -0.18263915181159973, -0.38308316469192505, -0.1138099804520607, 0.21733051538467407, -0.019879233092069626, -0.023588884621858597, -0.2534290552139282, 0.10554011911153793, -0.06678193807601929, 0.06014016270637512, 0.2171061635017395, 0.15147016942501068, -0.23825182020664215, 0.13472546637058258, -0.058886781334877014, -0.23463891446590424, 0.2361760139465332, 0.2706865072250366, 0.4174257814884186, 0.16836750507354736, 0.04018959030508995, 0.20723821222782135, 0.6981124877929688, 0.2486543357372284, 0.028429972007870674, 0.14254799485206604, 0.2371979057788849, -0.005118515342473984, 0.188836932182312, -0.2463115155696869, 0.21866951882839203, -0.216200053691864, 0.48305827379226685, 0.028078727424144745, 0.1475924849510193, -0.21976663172245026, -0.2751489281654358, 0.0053660087287425995, -0.2993876039981842, -0.19954906404018402, 0.22123710811138153, -0.27302905917167664, 0.011601664125919342, -0.11817152798175812, -0.34711340069770813, -0.27148744463920593, -0.03676827624440193, 0.01355339027941227, 0.30513429641723633, -0.08174071460962296, 0.15701259672641754, -0.3513631522655487, -0.049009110778570175, -0.4496108293533325, 0.15219715237617493, 0.041846055537462234, 0.2807386517524719, -0.06174740195274353, -0.060966841876506805, -0.10398003458976746, 0.09561648964881897, 0.5847180485725403, -0.5166595578193665, -0.0955328494310379, 0.1287485510110855, 0.12548871338367462, -0.13120079040527344, -0.11914443969726562, -0.2858884632587433, 0.07362529635429382, 0.4597931206226349, 0.34093767404556274, -0.3444136381149292, -0.2527545690536499, 0.219550222158432, -0.1727745682001114, -0.05759384483098984, -0.18763361871242523, -0.23618847131729126, -0.2354423701763153, -0.08206910640001297, 0.055361997336149216, 0.22892631590366364, 0.2281399816274643, -0.11155793070793152, 0.19489000737667084, -0.12732549011707306, 0.24054816365242004, 0.21754425764083862, -0.0025133593007922173, -0.024932891130447388, 0.1810740828514099, 0.17871515452861786, 0.044977836310863495, -0.14879077672958374, -0.3022290766239166, 0.09064795076847076, -0.05727584287524223, 0.11080880463123322, 0.22470298409461975, -0.020743336528539658, 0.629221498966217, 0.04769931733608246, 0.31856662034988403, 0.2778649628162384, -0.19443348050117493, -0.002981148660182953, -0.2579376697540283, 0.20183704793453217, 0.22035077214241028, 0.26212093234062195, -0.11640968173742294, -0.2639237940311432, 0.5500637888908386, 0.3211117386817932, -0.0659315288066864, 0.44204992055892944, -0.5858772397041321, -0.31232112646102905, 0.24330048263072968, 0.06730468571186066, 0.6210485100746155, -0.01113668829202652, 0.2869623601436615, 0.08895482122898102, -0.3830382227897644, 0.5410966873168945, -0.3828495740890503, 0.1639019399881363, -0.08226895332336426, -0.1536538153886795, 0.01521679200232029, -0.09603606164455414, -0.040717847645282745, 0.25265881419181824, -0.32958027720451355, 0.3504188358783722, 0.4120699167251587, -0.20682260394096375, -0.06753432005643845, 0.5238791108131409, -0.013963611796498299, -0.35267695784568787, -0.04201924055814743, 0.08234072476625443, -0.2836204767227173, 0.1062176525592804, -0.14440613985061646, -0.24390070140361786, 0.21701034903526306, 0.15858152508735657, -0.3457740545272827, 0.08260709792375565, -0.035912834107875824, 0.22727586328983307, 0.21372637152671814, -0.03624611347913742, 0.20517532527446747, -0.1772371083498001, -0.07493124902248383, -0.1512720286846161, -0.03667150437831879, 0.12637118995189667, -0.14632153511047363, 0.04019317030906677, 0.0808696299791336, -0.2181077003479004, 0.4567464292049408, -0.04643137380480766, -0.1222226470708847, -0.19975224137306213, 0.28687262535095215, -0.42228835821151733, -0.1803591549396515, 0.15953995287418365, 0.1727864146232605, 0.252602756023407, -0.31159907579421997, 0.5563429594039917, 0.1483863890171051, -0.01022356003522873, 0.03905479609966278, 0.18057703971862793, -0.26042017340660095, 0.5079049468040466, -0.36686787009239197, 0.014650382101535797, 0.08976055681705475, 0.020912379026412964, -0.13653337955474854, -0.36404064297676086, -0.048243485391139984, 0.3149685859680176, 0.11398130655288696, 0.00437699630856514, 0.06601301580667496, 0.12075292319059372, 0.013668261468410492, -0.04037732630968094, -0.16214576363563538, -0.34312987327575684, 0.11653603613376617, 0.1604323536157608, 0.1777603030204773, 0.18723280727863312, -0.04996239393949509, -0.2939014434814453, -0.16337966918945312, 0.4305415153503418, 0.044078417122364044, 0.2024313360452652, -0.13496211171150208, 0.35182130336761475, 0.3814525008201599, 0.49481624364852905, -0.25800997018814087, 0.07705625146627426, 0.09505408257246017, 0.09604034572839737, 0.44298055768013, -0.21446135640144348, 0.05208288133144379, 0.25334668159484863, -0.11877486109733582, 0.0298747755587101, -0.32196879386901855, -0.15790030360221863, -0.23781806230545044, 0.09774885326623917, 0.07003472000360489, 0.01756487973034382, 0.04143724963068962, -0.10240134596824646, -0.08863097429275513, -0.07675354927778244, -0.01117253303527832, -0.32483306527137756, -0.06518663465976715, 0.1471289098262787, -0.10908211767673492, 0.04781116917729378, 0.07037888467311859, 0.19740544259548187, -0.002258516848087311, -0.08733029663562775, 0.2813519537448883, 0.12359531968832016, 0.18367765843868256, -0.16559207439422607, -0.1435876190662384, -0.025252990424633026, 0.21271595358848572, 0.14612016081809998, -0.07139419764280319, 0.16139699518680573, -0.2196010947227478, -0.15503615140914917, 0.25354963541030884, 0.22577472031116486, 0.02459389716386795, -0.10969046503305435, 0.147075355052948, 0.11975523084402084, -0.12432171404361725, -0.0021177558228373528, 0.20850613713264465, 0.09803561121225357, 0.11718396842479706, 0.1034604012966156, 0.10285216569900513, 0.46599674224853516, -0.3051430583000183, 0.03137544170022011, 0.10533096641302109, -0.3483791947364807, 0.11580254137516022, -0.08272500336170197, -0.0022730231285095215, 0.0024448297917842865, 0.10375705361366272, -0.0023502446711063385, 0.14066556096076965, -0.23288805782794952, 0.3303033411502838, 0.3737262487411499, 0.0639888197183609, 0.007040075957775116, 0.1710142195224762, 0.0021042227745056152, -0.11631642282009125, 0.23376409709453583, 0.020215503871440887, 0.35285112261772156, 0.03638788312673569, 0.8677213788032532, -0.31640490889549255, -0.44532158970832825, 0.10563317686319351, 0.39686667919158936, 0.05611910670995712, -0.23129118978977203, 0.1447218805551529, 0.10362931340932846, -0.3610672354698181, 0.0887259840965271, 0.06926026195287704, -0.17860956490039825, 0.668048620223999, -0.12029631435871124, 0.16749344766139984, -0.1809268295764923, -0.22236594557762146, 0.3528715968132019, 0.1628759801387787, -0.10737734287977219, -0.03565685451030731, 0.1072143018245697, -0.10292467474937439, -0.199831023812294, 0.4940883219242096, 0.24655316770076752, 0.048753246665000916, -0.5133512020111084, 0.011973072774708271, 0.06946771591901779, 0.15743617713451385, -0.07946465909481049, 0.10217936336994171, 0.4472751021385193, -0.17720068991184235, 0.2422112673521042, 0.015447285026311874, -0.057762470096349716, -0.10079395771026611, 0.09602126479148865, -0.4450973570346832, -0.3861229419708252, 0.4605049788951874, 0.07121430337429047, -0.10437770932912827, 0.0653752014040947, 0.3186512291431427, -0.6325953006744385, -0.26431936025619507, 0.35300230979919434, -0.08635836839675903, -0.05638347566127777, -0.20915158092975616, 0.08163756877183914, 0.042112432420253754, 0.6012605428695679, 0.1831812858581543, 0.3498595654964447, -0.3703472316265106, -0.11620760709047318, -0.6050469875335693, -0.19771051406860352, 0.051925208419561386, -0.15896077454090118, -0.2622525990009308, 0.14968301355838776, -0.06800580769777298, 0.326211154460907, -0.23237250745296478, 0.09375449270009995, 0.263223797082901, 0.10629783570766449, -0.45955148339271545, 0.056498199701309204, 0.1048322468996048, 0.030725767835974693, 0.3046696186065674, -0.10057763755321503, -0.22762726247310638, 0.046522561460733414, -0.055756255984306335, -0.06158817559480667, -0.14107653498649597, 0.2911347448825836, 0.18577246367931366, 0.08580746501684189, 0.025451794266700745, 0.5246702432632446, -0.0034383945167064667, -0.14572028815746307, -0.025782551616430283, 0.054995931684970856, -0.3764566481113434, -0.12250711768865585, -0.0945226401090622, 0.1602577418088913, -0.3817417621612549, 0.272300660610199, 0.011709699407219887, -0.04529961571097374, 0.11852079629898071, 0.0029892444144934416, -0.19630929827690125, -0.030839137732982635, 0.06354299932718277, 0.23153410851955414, 0.17898130416870117, 0.7314297556877136, -0.025483064353466034, 0.1557295322418213, -0.07545021176338196, -0.035311587154865265, 0.05345030128955841, -0.13339698314666748, -0.2366471290588379, -0.2042839229106903, 0.06664758920669556, 0.05969949811697006, 0.1490631103515625, -0.6794450879096985, -0.33675849437713623, 0.03388065844774246, -0.0028483830392360687, 0.060768578201532364, 0.7086071968078613, 0.07913337647914886, -0.1290806531906128, -0.11463317275047302, -0.3271321952342987, -0.00030950456857681274, -0.25829946994781494, 0.2029401659965515, -0.37882471084594727 ]
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.
[ -0.005270712077617645, -0.4564434885978699, -0.04090085253119469, -0.033153802156448364, -0.24214433133602142, 0.0899883434176445, 0.22626745700836182, 0.03379298001527786, 0.3455828130245209, -0.18324880301952362, -0.1746247559785843, 0.27823755145072937, -0.21712930500507355, 0.3147799074649811, 0.3954232633113861, -0.4474583864212036, 0.05297614261507988, -0.268787145614624, -0.2423509955406189, 0.37680453062057495, 0.023829326033592224, 0.1376442164182663, -0.165048286318779, 0.10277897119522095, -0.3490171432495117, -0.14932627975940704, -0.29104581475257874, 0.10549008846282959, 0.14364416897296906, 0.0494338721036911, 0.24418769776821136, 0.3417590260505676, -0.031972579658031464, 0.1957811713218689, -0.00011775486927945167, -0.02630411460995674, 0.2295907735824585, -0.05339072644710541, 0.13802748918533325, -0.05143082141876221, 0.13814207911491394, -0.37766194343566895, 0.06772827357053757, -0.12424972653388977, 0.19982796907424927, 0.26353877782821655, 0.18150970339775085, 0.130659282207489, 0.3444681763648987, -0.3399111032485962, 0.05876673758029938, 0.27942800521850586, -0.20459185540676117, 0.3041354715824127, -0.08195006102323532, 0.047511495649814606, 0.0006847735494375229, 0.24062517285346985, 0.6316858530044556, -0.6122636795043945, 0.009547203779220581, 0.11814531683921814, 0.020631732419133186, -0.07559317350387573, 0.21316513419151306, -0.07582761347293854, -0.3081117272377014, -0.20631641149520874, -0.401638001203537, 0.19827395677566528, -0.13543862104415894, -0.12603074312210083, -0.17508959770202637, -0.5726632475852966, -0.09653259068727493, 0.07819853723049164, -0.2844803035259247, 0.015646645799279213, -0.1918289214372635, -0.048152048140764236, -0.13264106214046478, -0.12773174047470093, -0.03177602216601372, 0.13176490366458893, 0.3510734736919403, 0.6222617626190186, 0.1479904055595398, 0.15523606538772583, 0.38221311569213867, 0.1009344831109047, -0.08788469433784485, -0.2404811978340149, 0.313900351524353, 0.1148008406162262, -0.39394763112068176, -0.3623136281967163, 0.2042005956172943, -0.2805655598640442, 0.22935116291046143, 0.06943339109420776, 0.16647127270698547, 0.08447271585464478, -0.1230989471077919, 0.26120567321777344, 0.31291449069976807, 0.004651205614209175, -0.13495278358459473, -0.09768775850534439, 0.0522313192486763, -0.07296288758516312, 0.23544877767562866, 0.10745201259851456, 0.1639312356710434, -0.01664864271879196, -0.43312445282936096, -0.08164000511169434, -0.13500460982322693, 0.16052022576332092, -0.12132032215595245, -0.5605202317237854, -0.03214416280388832, -0.14209012687206268, 0.1316174864768982, -0.09903281182050705, -0.305805504322052, 0.13915681838989258, -0.08490349352359772, 0.2681118845939636, -0.09449831396341324, 0.13290256261825562, -0.03960466757416725, 0.048502255231142044, -0.44950002431869507, -0.06491013616323471, 0.32811102271080017, 0.2135879248380661, 0.052876412868499756, 0.13712136447429657, -0.001447327435016632, 0.07443195581436157, 0.35362860560417175, 0.054831210523843765, -0.010001584887504578, 0.04295624420046806, 0.07790768146514893, -0.30953723192214966, -0.027146007865667343, 0.21304695308208466, -0.3466987609863281, -0.135694682598114, -0.16489878296852112, -0.2356416881084442, 0.1343439370393753, -0.014896339736878872, -0.18980468809604645, -0.2510468661785126, -0.5263704657554626, 0.659267246723175, -0.030561067163944244, -0.057073093950748444, -0.12782695889472961, 0.32526862621307373, -0.24029703438282013, -0.07828085869550705, 0.12975382804870605, 0.045841217041015625, 0.03465936705470085, -0.37421631813049316, -0.09624490141868591, -0.08568520843982697, -0.04035698622465134, 0.30118250846862793, -0.18863238394260406, 0.2752721905708313, 0.09550352394580841, 0.13118724524974823, 0.21981319785118103, -0.48723840713500977, -0.03061848320066929, 0.29354098439216614, -0.02629059925675392, 0.37811198830604553, 0.07428215444087982, 0.24824485182762146, -0.04977637156844139, -0.18438968062400818, 0.25284382700920105, 0.9806959629058838, -0.3445678651332855, 0.007813209667801857, -0.12597227096557617, -0.3848101496696472, 0.593819797039032, 0.3565863072872162, 0.1247875839471817, -0.28234657645225525, -0.13511595129966736, 0.08289138972759247, 0.09046825021505356, 0.11593087017536163, 0.07905609905719757, -0.09205912053585052, -0.3260628283023834, 0.023914657533168793, -0.16235792636871338, -0.11649957299232483, -0.4359433054924011, 0.010673858225345612, -0.029216019436717033, 0.35977673530578613, 0.21428415179252625, -0.10928668826818466, 0.2714816927909851, -0.2791427969932556, -0.12076298892498016, -0.3025941848754883, 0.07069007307291031, -0.1252288818359375, -0.0336623452603817, -0.28421685099601746, -0.014253232628107071, 0.28782302141189575, 0.17413687705993652, -0.09134804457426071, -0.4625680446624756, 0.3106144964694977, -0.1135096624493599, 0.07947522401809692, -0.07837453484535217, 0.6710865497589111, -0.13304346799850464, -0.05170928314328194, 0.18674391508102417, 0.2545601725578308, -0.30537280440330505, 0.2602539658546448, 0.3174130320549011, 0.3043522834777832, 0.19623766839504242, -0.05067839100956917, 0.08027397096157074, 0.07100621610879898, -0.006333452183753252, -0.2542903423309326, -0.168838769197464, 0.45542824268341064, -0.17983192205429077, 0.46422210335731506, -0.07356216758489609, -0.006369562819600105, -0.20803198218345642, -0.04436568543314934, -0.2130458801984787, 0.44610467553138733, 0.29147422313690186, -0.16969943046569824, 0.21452738344669342, -0.10940782725811005, -0.35217347741127014, 0.14893800020217896, 0.13902010023593903, 0.03116176277399063, 0.03513643145561218, 0.012584585696458817, 0.006687812507152557, -0.2945304214954376, -0.11442479491233826, 0.2900373935699463, 0.5758762955665588, 0.218854159116745, 0.13711366057395935, 0.01795647293329239, -0.20882201194763184, -0.09033512324094772, 0.013694152235984802, 0.2752791941165924, 0.01692286692559719, 0.19390946626663208, 0.015642074868083, 0.06455541402101517, 0.04245805740356445, -0.20732952654361725, 0.15761087834835052, -0.20261123776435852, -0.0598207488656044, 0.03191634267568588, -0.18620602786540985, -0.5582317113876343, -0.25548040866851807, -0.010924034751951694, -0.3671207129955292, -0.2729373276233673, 0.15407200157642365, 0.0434349849820137, -0.16207273304462433, 0.1376049518585205, 0.2474062144756317, 0.6375616192817688, -0.4645482301712036, -0.23034031689167023, -0.14063900709152222, -0.3725029528141022, -0.07144743204116821, 0.07383935153484344, 0.4597983956336975, 0.2861062288284302, 0.38312235474586487, 0.33520054817199707, -0.07496833056211472, 0.10937522351741791, -0.4088929295539856, 0.01039048284292221, -0.28102684020996094, 0.24833275377750397, 0.08054860681295395, -0.21447692811489105, 0.17429618537425995, -0.39816129207611084, -0.06309565901756287, -0.02732512727379799, -0.06673935800790787, -0.004362684674561024, -0.0064717913046479225, 0.08041532337665558, -0.19725613296031952, -0.16866174340248108, -0.6268006563186646, -0.4757333993911743, 0.32473599910736084, -0.2877430319786072, 0.06258346885442734, 0.09646835178136826, -0.27044060826301575, 0.12282240390777588, -0.07755089551210403, -0.006085807457566261, -0.12584716081619263, -0.042075514793395996, 0.027647659182548523, -0.3368200957775116, -0.14505460858345032, -0.3748606741428375, -0.09668285399675369, 0.22548219561576843, -0.07548362761735916, -0.014169111847877502, -0.14891380071640015, 0.10138890147209167, -0.16277535259723663, 0.1472623199224472, 0.13594429194927216, 0.1742737591266632, -0.2007133513689041, 0.16263291239738464, -0.018707377836108208, -0.32956135272979736, 0.2292134165763855, 0.37306100130081177, 0.45728522539138794, 0.17798158526420593, 0.054322533309459686, 0.16411493718624115, 0.7202291488647461, 0.2120826542377472, 0.11296965181827545, 0.1207050234079361, 0.339720755815506, -0.09869544208049774, 0.16566073894500732, -0.23940826952457428, 0.2550593912601471, -0.17682713270187378, 0.47674137353897095, 0.049709927290678024, 0.2018958032131195, -0.20101477205753326, -0.281904935836792, -0.030092597007751465, -0.29803550243377686, -0.21822118759155273, 0.2677379250526428, -0.2083997130393982, 0.040124133229255676, -0.07382592558860779, -0.40090104937553406, -0.2804727256298065, -0.06336963176727295, 0.056344639509916306, 0.4238353371620178, -0.05789089947938919, 0.17249959707260132, -0.33453816175460815, -0.06981953233480453, -0.40455499291419983, 0.05911160632967949, 0.0287389624863863, 0.23816978931427002, -0.09146816283464432, -0.06206570565700531, -0.0657404214143753, 0.16588729619979858, 0.6495797038078308, -0.47983986139297485, -0.10412196815013885, 0.08365978300571442, 0.08293940126895905, -0.043583016842603683, -0.1442505419254303, -0.24003960192203522, -0.045977845788002014, 0.43813803791999817, 0.4039085805416107, -0.3940846025943756, -0.2811402976512909, 0.1908789575099945, -0.25683602690696716, -0.12578126788139343, -0.23165160417556763, -0.21141904592514038, -0.17475789785385132, -0.1496141254901886, 0.051150932908058167, 0.264576256275177, 0.21153360605239868, -0.1040395125746727, 0.13656336069107056, -0.01972043327987194, 0.23842531442642212, 0.21176308393478394, -0.014663837850093842, -0.0997283011674881, 0.2485458254814148, 0.12894083559513092, 0.03632749244570732, -0.16067296266555786, -0.34276020526885986, 0.07137365639209747, 0.04164756461977959, 0.1356404423713684, 0.28387314081192017, -0.00841048639267683, 0.6613308787345886, 0.07133970409631729, 0.38200536370277405, 0.295071005821228, -0.03617727756500244, -0.006972946226596832, -0.27927374839782715, 0.13812753558158875, 0.18902543187141418, 0.2439100444316864, -0.19070518016815186, -0.17326833307743073, 0.5538422465324402, 0.2548624277114868, -0.10853821784257889, 0.38894686102867126, -0.5794233083724976, -0.28470340371131897, 0.28209590911865234, 0.06647150218486786, 0.7568140029907227, -0.022754795849323273, 0.29468196630477905, 0.05107128620147705, -0.46020108461380005, 0.5372833609580994, -0.4241015911102295, 0.1517772525548935, -0.048981014639139175, -0.11968086659908295, -0.010851796716451645, -0.10564872622489929, -0.10179810971021652, 0.30793994665145874, -0.33810174465179443, 0.30630993843078613, 0.49197500944137573, -0.23215794563293457, -0.07935892790555954, 0.5174073576927185, 0.09096956253051758, -0.3882366120815277, 0.00893926341086626, 0.008923366665840149, -0.25484174489974976, 0.1692943274974823, -0.19708052277565002, -0.2969512343406677, 0.19138270616531372, 0.1274462193250656, -0.2990952134132385, 0.08042872697114944, 0.08365587890148163, 0.2613987922668457, 0.32197779417037964, 0.001385912299156189, 0.30562031269073486, -0.21544677019119263, -0.08355700969696045, -0.18756043910980225, -0.10171757638454437, 0.16740213334560394, -0.22281646728515625, 0.07013489305973053, 0.04552353546023369, -0.18126130104064941, 0.5122028589248657, -0.07783924788236618, -0.12757037580013275, -0.275188684463501, 0.25064870715141296, -0.45339494943618774, -0.2337384968996048, 0.1493915319442749, 0.14604678750038147, 0.2355833649635315, -0.3322303891181946, 0.5851309299468994, 0.08234508335590363, -0.005585554987192154, 0.01226550992578268, 0.18602252006530762, -0.20392747223377228, 0.5055809617042542, -0.3174498975276947, 0.04855109751224518, 0.08217541873455048, -0.043954454362392426, -0.1198650524020195, -0.40590372681617737, -0.09712445735931396, 0.3543564975261688, 0.09507563710212708, 0.025455109775066376, -0.05688740685582161, 0.13590461015701294, -0.1269838511943817, -0.05918329581618309, -0.1386878788471222, -0.39263230562210083, 0.10445506870746613, 0.2105829417705536, 0.201763316988945, 0.16465380787849426, -0.07526347041130066, -0.25805097818374634, -0.21936975419521332, 0.5150886178016663, -0.02673773095011711, 0.20104709267616272, -0.16419154405593872, 0.375638484954834, 0.42429500818252563, 0.5418851375579834, -0.2086775302886963, 0.06487932801246643, 0.06177882105112076, 0.08104370534420013, 0.37772032618522644, -0.2797585427761078, 0.04444727674126625, 0.22707773745059967, -0.18765462934970856, 0.006583523005247116, -0.27863770723342896, -0.10919923335313797, -0.2729395031929016, 0.13855046033859253, 0.147890105843544, 0.035220738500356674, -0.003477866295725107, -0.08391064405441284, -0.045493949204683304, -0.018514038994908333, -0.061204250901937485, -0.38703739643096924, -0.02604971081018448, 0.18790309131145477, -0.04360799118876457, 0.05576164275407791, 0.017927583307027817, 0.2448146641254425, 0.04288823530077934, -0.06855219602584839, 0.36694198846817017, 0.09157950431108475, 0.15586097538471222, -0.14030107855796814, -0.2409391552209854, 0.020427793264389038, 0.2335359752178192, 0.139428973197937, -0.009182199835777283, 0.19073864817619324, -0.23085370659828186, -0.13711145520210266, 0.1696343868970871, 0.12699580192565918, 0.07510606944561005, -0.10606863349676132, 0.09852870553731918, 0.07580418139696121, -0.11297906190156937, -0.023444492369890213, 0.21591240167617798, 0.09666826575994492, 0.04326470568776131, 0.11526884138584137, 0.14180317521095276, 0.49411487579345703, -0.2662070393562317, 0.08396876603364944, 0.1088305115699768, -0.3563464283943176, 0.13451825082302094, -0.1554356962442398, -0.020224642008543015, 0.029280882328748703, 0.1533055156469345, -0.01826348900794983, 0.11684288829565048, -0.35832130908966064, 0.3812861144542694, 0.2211315631866455, -0.003973167389631271, 0.03103131428360939, 0.20100942254066467, 0.050393808633089066, -0.2398848980665207, 0.20653469860553741, 0.027465254068374634, 0.28734877705574036, 0.06632684916257858, 0.8376128673553467, -0.2126876711845398, -0.3841235935688019, 0.14744503796100616, 0.4088926911354065, 0.046499524265527725, -0.20879638195037842, 0.16752348840236664, 0.13624891638755798, -0.3738655149936676, 0.06661009788513184, 0.04919600859284401, -0.14546628296375275, 0.6858696937561035, -0.10418245196342468, 0.11888362467288971, -0.07461725920438766, -0.21313542127609253, 0.384389728307724, 0.16913066804409027, -0.0546514056622982, -0.06694069504737854, 0.13790003955364227, -0.11123915016651154, -0.1763106733560562, 0.49055102467536926, 0.20801791548728943, 0.009057771414518356, -0.45847073197364807, 0.020228872075676918, 0.1295265257358551, 0.13868214190006256, -0.05805091932415962, 0.17673495411872864, 0.47014448046684265, -0.07212406396865845, 0.15682095289230347, -0.03370370715856552, -0.010842632502317429, -0.12366794049739838, 0.04220827668905258, -0.41784194111824036, -0.3098796010017395, 0.5140645503997803, 0.09148631989955902, -0.17002195119857788, 0.09728050231933594, 0.3429412841796875, -0.6166440844535828, -0.2568531334400177, 0.29943743348121643, -0.11552081257104874, -0.09028030931949615, -0.23393946886062622, 0.0465732216835022, 0.1041128933429718, 0.6218817830085754, 0.14806923270225525, 0.3906702995300293, -0.3981969654560089, -0.10698513686656952, -0.5736942887306213, -0.23161479830741882, 0.13995277881622314, -0.05170750617980957, -0.17874516546726227, 0.09441905468702316, -0.06492910534143448, 0.22461533546447754, -0.256145179271698, -0.019427163526415825, 0.21327170729637146, 0.09001754224300385, -0.4094822406768799, 0.16877318918704987, 0.18962767720222473, 0.068280890583992, 0.3064979016780853, -0.03220444172620773, -0.15833808481693268, 0.20729203522205353, -0.1309206485748291, -0.13873668015003204, -0.08773781359195709, 0.24361661076545715, 0.18980833888053894, 0.046843431890010834, 0.021143078804016113, 0.4844205677509308, -0.004028007388114929, -0.2120710015296936, 0.010847926139831543, -0.0026888842694461346, -0.3256473243236542, -0.17872539162635803, -0.07713405787944794, 0.20804090797901154, -0.36006876826286316, 0.2944648563861847, -0.009957587346434593, -0.1363239586353302, 0.02387673407793045, -0.008793317712843418, -0.2192276418209076, -0.025523681193590164, 0.03367118537425995, 0.23170852661132812, 0.24931330978870392, 0.6675169467926025, 0.01665070280432701, 0.09321210533380508, -0.08999074250459671, -0.05048945173621178, 0.026074931025505066, -0.01502169854938984, -0.24134978652000427, -0.14494483172893524, 0.03302671015262604, 0.01468677818775177, 0.18479904532432556, -0.7851413488388062, -0.39457643032073975, -0.021979182958602905, 0.03141116723418236, 0.0848199725151062, 0.701366126537323, 0.12302403151988983, -0.1750701665878296, -0.09732478857040405, -0.34183692932128906, 0.020174965262413025, -0.22876331210136414, 0.17516592144966125, -0.36070096492767334 ]
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.
[ -0.06561257690191269, -0.3723561465740204, -0.06819301843643188, -0.049267709255218506, -0.2852722406387329, 0.13376593589782715, 0.21648038923740387, 0.12065184116363525, 0.3021675944328308, -0.1421176940202713, -0.1687704473733902, 0.3203653395175934, -0.16379880905151367, 0.3486941456794739, 0.4185852110385895, -0.41218042373657227, 0.031688444316387177, -0.21076568961143494, -0.21456116437911987, 0.4246906638145447, 0.0625796988606453, 0.08338528871536255, -0.16562975943088531, 0.10172660648822784, -0.29676884412765503, -0.1877509206533432, -0.33034709095954895, 0.12289676070213318, 0.1802750527858734, -0.006745514459908009, 0.21674300730228424, 0.3386645019054413, -0.005247997120022774, 0.07626133412122726, -0.00011275351425865665, -0.07976371049880981, 0.21804775297641754, -0.05337457358837128, 0.10312984883785248, -0.013962417840957642, 0.19306248426437378, -0.36390167474746704, 0.04667441546916962, -0.10680350661277771, 0.168757826089859, 0.1935688704252243, 0.17223404347896576, -0.07640406489372253, 0.3412652909755707, -0.2545016407966614, 0.10234159231185913, 0.3156692087650299, -0.23991090059280396, 0.29766952991485596, -0.11138313263654709, -0.08681921660900116, 0.011975457891821861, 0.24389182031154633, 0.671851396560669, -0.6273881196975708, -0.06007193773984909, 0.057241737842559814, -0.03238601237535477, 0.04977741837501526, 0.09267555177211761, -0.13015508651733398, -0.3021056056022644, -0.21089527010917664, -0.33772188425064087, 0.25109922885894775, -0.15973111987113953, -0.13706070184707642, -0.22889843583106995, -0.5552747249603271, -0.013059410266578197, 0.051291950047016144, -0.3080897033214569, 0.031833216547966, -0.17259512841701508, -0.006247218698263168, -0.10114537924528122, -0.12405448406934738, 0.057553812861442566, 0.08909878879785538, 0.33570656180381775, 0.5584868788719177, 0.12079384922981262, 0.1740795075893402, 0.36989516019821167, 0.0731532946228981, -0.08478496223688126, -0.25911930203437805, 0.27419862151145935, 0.10016472637653351, -0.4158090353012085, -0.31035926938056946, 0.1395321786403656, -0.2330121099948883, 0.3023061752319336, 0.09700149297714233, 0.13515186309814453, 0.11098425835371017, -0.12362208217382431, 0.2736952304840088, 0.3402844965457916, 0.03425619378685951, -0.07769947499036789, -0.16922658681869507, 0.010786432772874832, -0.14165517687797546, 0.18027429282665253, 0.11355341970920563, 0.20747898519039154, 0.02580343186855316, -0.43926802277565, -0.13691231608390808, -0.22848019003868103, 0.163488507270813, -0.25956225395202637, -0.5688910484313965, -0.11211642622947693, -0.09568820148706436, 0.06208207085728645, -0.1388760209083557, -0.26140207052230835, 0.13918064534664154, -0.09216346591711044, 0.26953399181365967, -0.06885512173175812, 0.006288326345384121, -0.05109159275889397, 0.08796742558479309, -0.45554232597351074, -0.010129228234291077, 0.33236196637153625, 0.10064949095249176, 0.08620400726795197, 0.1457323431968689, -0.019783131778240204, 0.062231436371803284, 0.38701415061950684, 0.011976772919297218, -0.004459694027900696, -0.012135099619626999, 0.18673944473266602, -0.3140554130077362, 0.02921934798359871, 0.18315307796001434, -0.3368549346923828, -0.08135093003511429, -0.14553308486938477, -0.14867334067821503, 0.1425742655992508, 0.03213296830654144, -0.22536857426166534, -0.28603845834732056, -0.5482082962989807, 0.6673356294631958, -0.011244222521781921, -0.06299498677253723, -0.16312560439109802, 0.3023531436920166, -0.3023782968521118, -0.09308987855911255, 0.17095297574996948, 0.013165690004825592, 0.021748218685388565, -0.27748388051986694, -0.16175632178783417, -0.07043299078941345, -0.10190135985612869, 0.3153511583805084, -0.1921916902065277, 0.30844223499298096, 0.04201127588748932, 0.1713876873254776, 0.2346702516078949, -0.48465967178344727, -0.032916322350502014, 0.3313533663749695, -0.07133089005947113, 0.401894748210907, 0.050246935337781906, 0.22615402936935425, 0.055643998086452484, -0.17825017869472504, 0.23286056518554688, 0.9699898958206177, -0.3880669176578522, 0.009526556357741356, -0.08592958003282547, -0.3920598328113556, 0.5433835983276367, 0.3569908142089844, 0.059595491737127304, -0.24156567454338074, -0.14711663126945496, -0.012277972884476185, 0.08712254464626312, 0.10982911288738251, 0.09951728582382202, -0.07516493648290634, -0.42072439193725586, 0.023374471813440323, -0.2236105054616928, -0.19864052534103394, -0.43184375762939453, 0.018178671598434448, 0.021138910204172134, 0.22960224747657776, 0.3293682634830475, -0.13701951503753662, 0.24782595038414001, -0.279499351978302, -0.046550583094358444, -0.37557777762413025, 0.1250409185886383, -0.051529660820961, 0.008913092315196991, -0.20238712430000305, -0.04816081374883652, 0.3567744493484497, 0.17522487044334412, -0.12952224910259247, -0.3912130892276764, 0.3071179986000061, -0.052944689989089966, 0.03812495991587639, -0.040555454790592194, 0.6141847968101501, -0.21523290872573853, -0.057151585817337036, 0.1891898810863495, 0.24514198303222656, -0.27274125814437866, 0.28847983479499817, 0.35366785526275635, 0.27539190649986267, 0.19267572462558746, -0.03591683506965637, 0.054176971316337585, 0.06028128042817116, -0.032949987798929214, -0.2539370059967041, -0.05172360688447952, 0.49388670921325684, -0.12789620459079742, 0.46705737709999084, -0.06947565078735352, 0.053518541157245636, -0.1416846364736557, -0.03650067746639252, -0.1959550529718399, 0.44116976857185364, 0.29005083441734314, -0.21228128671646118, 0.3862041234970093, -0.10933716595172882, -0.3189101815223694, 0.17448747158050537, 0.13803300261497498, 0.0512457899749279, 0.06468277424573898, -0.009072475135326385, -0.03657042235136032, -0.272716224193573, -0.0716029703617096, 0.3707674443721771, 0.6577780246734619, 0.23860488831996918, 0.17744719982147217, 0.014817635528743267, -0.22931432723999023, -0.0808015987277031, -0.031207077205181122, 0.19398143887519836, 0.026519775390625, 0.09888973087072372, 0.013474501669406891, 0.022317294031381607, 0.048615291714668274, -0.140078604221344, 0.15098978579044342, -0.16983996331691742, -0.009785115718841553, 0.04137495905160904, -0.22050823271274567, -0.5566984415054321, -0.21707725524902344, 0.029173608869314194, -0.3283434808254242, -0.2532225251197815, 0.10271672159433365, 0.12225048243999481, -0.13139253854751587, 0.1512339860200882, 0.23217689990997314, 0.6160979270935059, -0.4460151791572571, -0.2688652276992798, -0.08353511989116669, -0.3008219003677368, -0.11323411017656326, 0.12900160253047943, 0.45876020193099976, 0.301447331905365, 0.3852967321872711, 0.3329293727874756, -0.009069617837667465, -0.021664714440703392, -0.34239283204078674, -0.03896350413560867, -0.2504136562347412, 0.25898978114128113, 0.03648291528224945, -0.24169911444187164, 0.14477303624153137, -0.38711997866630554, -0.0762285441160202, -0.076802097260952, -0.07060395181179047, -0.08701084554195404, 0.031285360455513, 0.1460472047328949, -0.18527090549468994, -0.15247860550880432, -0.5852789282798767, -0.47434401512145996, 0.29749834537506104, -0.29475653171539307, 0.05387391522526741, -0.10809855908155441, -0.31336092948913574, 0.19335013628005981, -0.12611320614814758, -0.019432134926319122, -0.13752521574497223, -0.09073202311992645, -0.05549386516213417, -0.33968839049339294, -0.15745006501674652, -0.39757785201072693, -0.0830496996641159, 0.2783389091491699, -0.04936185106635094, -0.042478594928979874, -0.15597312152385712, 0.13113842904567719, -0.15483488142490387, 0.06006859242916107, 0.09263874590396881, 0.17417500913143158, -0.23634229600429535, 0.12928353250026703, -0.028270836919546127, -0.26209717988967896, 0.23682940006256104, 0.28308534622192383, 0.3746201694011688, 0.17052826285362244, 0.022374117746949196, 0.238032728433609, 0.7015316486358643, 0.23946097493171692, 0.05487408488988876, 0.16642330586910248, 0.28170809149742126, -0.023641463369131088, 0.11926232278347015, -0.2199317067861557, 0.15194281935691833, -0.1884620487689972, 0.4781736135482788, 0.10397278517484665, 0.15760928392410278, -0.25659656524658203, -0.27070897817611694, 0.030139897018671036, -0.2531300187110901, -0.17043861746788025, 0.21320009231567383, -0.3159545958042145, 0.03342357650399208, -0.06611943244934082, -0.39819636940956116, -0.22175011038780212, -0.01891930401325226, -0.014532038941979408, 0.28800368309020996, -0.04299434274435043, 0.1606459766626358, -0.3352032005786896, -0.031863465905189514, -0.408568799495697, 0.11785230040550232, 0.09698818624019623, 0.33181166648864746, -0.08659662306308746, -0.045031748712062836, -0.08494347333908081, 0.11186127364635468, 0.5522233247756958, -0.4672243297100067, -0.09274457395076752, 0.10849891602993011, 0.06989067047834396, -0.09652112424373627, -0.08832485973834991, -0.2206001579761505, 0.045805923640728, 0.42899057269096375, 0.3607071042060852, -0.2990397810935974, -0.28864553570747375, 0.15508930385112762, -0.21957731246948242, -0.04865565150976181, -0.18476983904838562, -0.21510964632034302, -0.1798703968524933, -0.13402646780014038, 0.06568881869316101, 0.23183833062648773, 0.2185508757829666, -0.08076180517673492, 0.15293242037296295, -0.11033450067043304, 0.15788644552230835, 0.1860540509223938, -0.043711405247449875, -0.09453729540109634, 0.15242967009544373, 0.20288051664829254, 0.013336161151528358, -0.1537550985813141, -0.30702757835388184, 0.02104792185127735, 0.013778546825051308, 0.13222724199295044, 0.24650096893310547, -0.017967522144317627, 0.6386029720306396, -0.04344119131565094, 0.3078649342060089, 0.3434125781059265, -0.15040098130702972, -0.06010924279689789, -0.28928837180137634, 0.16420391201972961, 0.18684609234333038, 0.2574945390224457, -0.12326844781637192, -0.187174990773201, 0.5725806355476379, 0.33587220311164856, -0.06958144903182983, 0.4539729654788971, -0.5722923874855042, -0.31260859966278076, 0.24356551468372345, 0.05923977866768837, 0.5774691104888916, -0.08398005366325378, 0.2616599202156067, -0.0199696384370327, -0.41017359495162964, 0.5315449237823486, -0.37374430894851685, 0.1534079760313034, -0.05188099294900894, -0.24385567009449005, 0.020601386204361916, -0.08406281471252441, -0.007088042795658112, 0.3127877712249756, -0.3229181170463562, 0.3246552050113678, 0.48921066522598267, -0.22829663753509521, -0.0012176278978586197, 0.5655282139778137, 0.025864582508802414, -0.37018075585365295, 0.01694345660507679, 0.06858357042074203, -0.2902955412864685, 0.08192048966884613, -0.15320712327957153, -0.21072693169116974, 0.22035029530525208, 0.17975349724292755, -0.352019727230072, 0.058837808668613434, 0.026805616915225983, 0.21444062888622284, 0.20536550879478455, 0.03536912798881531, 0.2878018915653229, -0.17419615387916565, -0.060969769954681396, -0.15786191821098328, -0.08689059317111969, 0.12871664762496948, -0.19305962324142456, 0.0364055410027504, 0.09449713677167892, -0.20959554612636566, 0.45646899938583374, -0.060330793261528015, -0.14519859850406647, -0.22870060801506042, 0.27023714780807495, -0.5092924237251282, -0.14957068860530853, 0.1981714516878128, 0.147306427359581, 0.28635287284851074, -0.29994630813598633, 0.5590230822563171, 0.10002879798412323, -0.015114353969693184, 0.04153593257069588, 0.19969823956489563, -0.23633220791816711, 0.5029985308647156, -0.3028043806552887, -0.015231341123580933, 0.06073188781738281, 0.05679795891046524, -0.1458495557308197, -0.3981848657131195, -0.07032734155654907, 0.2884272634983063, 0.14057910442352295, -0.004567045718431473, 0.0554143562912941, 0.09924254566431046, 0.050652071833610535, 0.019722655415534973, -0.20058375597000122, -0.324292927980423, 0.13359788060188293, 0.14717687666416168, 0.20381349325180054, 0.22745782136917114, -0.0952349454164505, -0.3026587665081024, -0.19763077795505524, 0.45944756269454956, -0.029067862778902054, 0.17826800048351288, -0.12853685021400452, 0.34765633940696716, 0.40746334195137024, 0.5269021987915039, -0.2459169179201126, 0.11764289438724518, 0.04560831934213638, 0.1269548535346985, 0.5051091909408569, -0.21541807055473328, -0.0042850300669670105, 0.3125104010105133, -0.13643276691436768, 0.06286744773387909, -0.2663431763648987, -0.16179120540618896, -0.19228850305080414, 0.10798419266939163, 0.12295631319284439, 0.025694001466035843, 0.02970491349697113, -0.11089315265417099, -0.09539107978343964, -0.040935371071100235, -0.055806245654821396, -0.3989924490451813, -0.06124714016914368, 0.09844258427619934, -0.06748118996620178, 0.07659310102462769, 0.024723634123802185, 0.2167934626340866, -0.03503827005624771, -0.1232563853263855, 0.24483482539653778, 0.11760400980710983, 0.17357222735881805, -0.15246367454528809, -0.1608661264181137, -0.036695994436740875, 0.22497503459453583, 0.1443028301000595, -0.05042234808206558, 0.19050449132919312, -0.23372295498847961, -0.07541358470916748, 0.24634647369384766, 0.10478648543357849, 0.09123042970895767, -0.08847752213478088, 0.17272531986236572, 0.10370082408189774, -0.08446706831455231, -0.020609991624951363, 0.23100976645946503, 0.05959821864962578, 0.11645430326461792, 0.13184748589992523, 0.1963878571987152, 0.46792519092559814, -0.18571580946445465, -0.0031028352677822113, 0.11268300563097, -0.38476091623306274, 0.12622153759002686, -0.1879337579011917, -0.07245918363332748, -0.028666596859693527, 0.13434644043445587, -0.017931785434484482, 0.1367465853691101, -0.2621413469314575, 0.3380008637905121, 0.3889702558517456, 0.0716743916273117, -0.007467169314622879, 0.22169539332389832, 0.06501398980617523, -0.15316633880138397, 0.25695449113845825, 0.05426362156867981, 0.3002220690250397, 0.012245923280715942, 0.8360199928283691, -0.324025422334671, -0.4433022141456604, 0.11316011101007462, 0.36561208963394165, 0.051835596561431885, -0.21638917922973633, 0.23777315020561218, 0.13523787260055542, -0.33631283044815063, 0.0876116156578064, 0.06379147619009018, -0.1809615045785904, 0.7053232192993164, -0.1334047168493271, 0.14903201162815094, -0.08333371579647064, -0.2459767758846283, 0.36664891242980957, 0.1151495948433876, -0.07119186222553253, -0.00048232264816761017, 0.05708689987659454, -0.11510632932186127, -0.17619889974594116, 0.53257817029953, 0.18094301223754883, 0.018203895539045334, -0.44259965419769287, 0.01007156539708376, 0.014929790049791336, 0.17563772201538086, -0.07289613783359528, 0.1774269938468933, 0.4411538541316986, -0.1820056140422821, 0.24563755095005035, 0.005940483883023262, -0.04617801681160927, -0.1398257613182068, 0.046848829835653305, -0.4108136296272278, -0.4485563635826111, 0.4289397895336151, 0.07663736492395401, -0.12908506393432617, 0.08759172260761261, 0.36286261677742004, -0.5635618567466736, -0.25134044885635376, 0.31865552067756653, -0.20191657543182373, -0.1316317617893219, -0.20967750251293182, 0.0823572501540184, 0.049965281039476395, 0.6422715783119202, 0.20109823346138, 0.39153286814689636, -0.3583631217479706, -0.17929266393184662, -0.5960274338722229, -0.17861217260360718, 0.08641770482063293, -0.08518502861261368, -0.22825589776039124, 0.1549057960510254, -0.11009537428617477, 0.33917397260665894, -0.17584379017353058, 0.07582423090934753, 0.2411535084247589, 0.00003824010491371155, -0.414818674325943, 0.09212320297956467, 0.12537765502929688, -0.017391210421919823, 0.28541383147239685, -0.10768580436706543, -0.14416517317295074, 0.1217658519744873, -0.07111340016126633, -0.07582584768533707, -0.11642725765705109, 0.2526834011077881, 0.1565539836883545, 0.1055905669927597, 0.01008404977619648, 0.4512917101383209, 0.01785183697938919, -0.1485949009656906, -0.051521334797143936, 0.03749128058552742, -0.3527217209339142, -0.18147292733192444, -0.10226456820964813, 0.14909955859184265, -0.4346385598182678, 0.29576852917671204, 0.02136603742837906, -0.06180842965841293, 0.1016729325056076, 0.005676994565874338, -0.25479844212532043, -0.058563075959682465, 0.031436383724212646, 0.1994289755821228, 0.2543521523475647, 0.7323935627937317, -0.00232667475938797, 0.1040797159075737, -0.05692112073302269, 0.03094632364809513, -0.02811811864376068, -0.08862250298261642, -0.25010132789611816, -0.13463179767131805, 0.05298812687397003, 0.07657415419816971, 0.0841163694858551, -0.6738630533218384, -0.40565288066864014, 0.053689200431108475, 0.022415632382035255, 0.05789883807301521, 0.7152572274208069, 0.02755647711455822, -0.1257437765598297, -0.1139771044254303, -0.3888225853443146, 0.008579626679420471, -0.22798483073711395, 0.21922320127487183, -0.40185993909835815 ]
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.
[ -0.002134092152118683, -0.39335891604423523, -0.04490877687931061, 0.011329784989356995, -0.22390787303447723, 0.021761052310466766, 0.19789783656597137, 0.027536004781723022, 0.41430220007896423, -0.19782714545726776, -0.23707209527492523, 0.39821067452430725, -0.19453246891498566, 0.302828848361969, 0.3515961468219757, -0.4019326865673065, -0.022123195230960846, -0.24198278784751892, -0.2949601113796234, 0.35135418176651, 0.060441792011260986, 0.19776856899261475, -0.21249975264072418, 0.06919260323047638, -0.21596956253051758, -0.17891055345535278, -0.3551909029483795, 0.15981325507164001, 0.0868036150932312, 0.008535115979611874, 0.2525821328163147, 0.42071613669395447, -0.047911107540130615, 0.1720024198293686, -0.00011538730177562684, -0.04062813147902489, 0.22244948148727417, -0.12262426316738129, 0.05334298685193062, -0.0625763088464737, 0.1727963387966156, -0.41804441809654236, 0.07642097771167755, -0.17792534828186035, 0.168304443359375, 0.1507711559534073, 0.12851466238498688, 0.018031105399131775, 0.39563190937042236, -0.26081591844558716, 0.10167448222637177, 0.3133843243122101, -0.15474574267864227, 0.3200138211250305, -0.13475903868675232, -0.021238267421722412, 0.025054844096302986, 0.25627315044403076, 0.6043381690979004, -0.547825813293457, -0.039860256016254425, 0.09073145687580109, 0.020382096990942955, -0.08250826597213745, 0.12837648391723633, -0.027154281735420227, -0.3430268168449402, -0.2517932951450348, -0.3898296058177948, 0.220865398645401, -0.0587255135178566, -0.18009454011917114, -0.22417059540748596, -0.5440153479576111, -0.05742250755429268, 0.0014682058244943619, -0.2000161111354828, -0.09796569496393204, -0.20692285895347595, -0.07884536683559418, -0.08055788278579712, -0.10031530261039734, -0.05526137351989746, 0.1227860078215599, 0.28622689843177795, 0.5722646713256836, 0.09168432652950287, 0.16959592700004578, 0.4112890958786011, -0.00377008318901062, -0.1038934737443924, -0.27631962299346924, 0.2889995872974396, 0.10878873616456985, -0.4652678370475769, -0.35177066922187805, 0.2077229917049408, -0.27356263995170593, 0.2420552670955658, 0.11815419048070908, 0.1267671287059784, 0.14092016220092773, -0.17259527742862701, 0.23502519726753235, 0.31019917130470276, 0.16164404153823853, -0.08544173091650009, -0.14018547534942627, -0.0408443957567215, -0.10049774497747421, 0.2605676054954529, 0.09310336410999298, 0.13976170122623444, -0.024156570434570312, -0.43835675716400146, -0.1129164919257164, -0.1902087926864624, 0.17506974935531616, -0.23685747385025024, -0.46511077880859375, -0.1156223937869072, -0.13347432017326355, 0.16438165307044983, -0.0893157348036766, -0.32635942101478577, 0.09385561943054199, -0.06360591948032379, 0.22851410508155823, -0.16430963575839996, 0.09540645778179169, -0.05092413350939751, 0.07936447858810425, -0.4200189709663391, -0.10349610447883606, 0.3041563630104065, 0.20076040923595428, 0.08877842128276825, 0.10183322429656982, 0.041355617344379425, 0.16144338250160217, 0.3386405408382416, -0.018589990213513374, 0.07116767764091492, -0.00675012543797493, 0.1700589954853058, -0.24186667799949646, 0.023634018376469612, 0.08572132885456085, -0.327629029750824, -0.14327853918075562, -0.1162455603480339, -0.17121268808841705, 0.09555090963840485, 0.001571817323565483, -0.2470017969608307, -0.21890217065811157, -0.5296895503997803, 0.6929859519004822, 0.009013980627059937, -0.07522498071193695, -0.17376235127449036, 0.21700376272201538, -0.2249675691127777, -0.11346260458230972, 0.18577587604522705, 0.13063430786132812, -0.019459588453173637, -0.2366185039281845, -0.08683812618255615, -0.11565864086151123, 0.007798895239830017, 0.25118836760520935, -0.1722692847251892, 0.311110258102417, 0.0009649060666561127, 0.25110113620758057, 0.26969873905181885, -0.4574740529060364, -0.08002284914255142, 0.29960525035858154, -0.07061275094747543, 0.49156177043914795, 0.06509648263454437, 0.2020913064479828, 0.01994803547859192, -0.20112177729606628, 0.27416419982910156, 0.9725066423416138, -0.3398754298686981, -0.045466549694538116, -0.13535133004188538, -0.3082437217235565, 0.5843701958656311, 0.2936512231826782, 0.0460602231323719, -0.23651698231697083, -0.1625548005104065, 0.025136027485132217, 0.13693943619728088, 0.11461334675550461, 0.10209725052118301, -0.043395839631557465, -0.3110347390174866, 0.06801388412714005, -0.16123905777931213, -0.13331198692321777, -0.4370240569114685, 0.03781503438949585, 0.04610864818096161, 0.34421592950820923, 0.1861216127872467, -0.04249907284975052, 0.2013804018497467, -0.24291640520095825, -0.20070357620716095, -0.3231874108314514, 0.12822404503822327, -0.034077659249305725, -0.006764721125364304, -0.3257838785648346, -0.04058927297592163, 0.39110177755355835, 0.11225387454032898, -0.14964769780635834, -0.5206971764564514, 0.2977626323699951, -0.08318758755922318, 0.07015018165111542, 0.00023986026644706726, 0.6380751729011536, -0.09240628033876419, -0.1353624016046524, 0.18318232893943787, 0.3113526403903961, -0.247738316655159, 0.14653019607067108, 0.3683391809463501, 0.17878328263759613, 0.17248530685901642, -0.053787339478731155, 0.04117811471223831, 0.014778591692447662, 0.06409095227718353, -0.2540859282016754, -0.04174523800611496, 0.4979396462440491, -0.20704549551010132, 0.4949684143066406, -0.05947215482592583, 0.09615933895111084, -0.07772010564804077, 0.003605946898460388, -0.21134963631629944, 0.41704657673835754, 0.2573397755622864, -0.16382722556591034, 0.31977397203445435, -0.1749093234539032, -0.389285683631897, 0.2181686908006668, 0.14777934551239014, 0.0507119856774807, -0.0239395871758461, -0.004027266055345535, -0.004173837602138519, -0.2835710644721985, -0.08406413346529007, 0.2592909634113312, 0.6727411150932312, 0.21467217803001404, 0.17579340934753418, -0.0035536978393793106, -0.21288852393627167, -0.07563171535730362, 0.01703471690416336, 0.21315500140190125, -0.018384750932455063, 0.1383446455001831, 0.06322658807039261, 0.055012837052345276, 0.02725990116596222, -0.16887202858924866, 0.16363954544067383, -0.15025180578231812, -0.11828188598155975, 0.11910352855920792, -0.2400861531496048, -0.6449166536331177, -0.1902637928724289, -0.005381044000387192, -0.33743420243263245, -0.3379825949668884, 0.06197679787874222, 0.09609253704547882, -0.09279252588748932, 0.17458873987197876, 0.33908799290657043, 0.5370780825614929, -0.42260855436325073, -0.2787001132965088, -0.07020194083452225, -0.31300246715545654, -0.22870567440986633, 0.10595953464508057, 0.47611004114151, 0.3147261440753937, 0.4362727403640747, 0.33345887064933777, -0.07831181585788727, -0.01110665313899517, -0.40816831588745117, 0.07382825762033463, -0.3157109022140503, 0.2591186761856079, 0.10779214650392532, -0.16198177635669708, 0.1569957733154297, -0.3664079010486603, -0.05445985123515129, 0.02567255310714245, -0.10825203359127045, 0.00573563389480114, -0.02132081426680088, 0.06178305298089981, -0.2249404340982437, -0.20574063062667847, -0.6049152612686157, -0.3986760079860687, 0.33175164461135864, -0.26845747232437134, 0.15955230593681335, 0.11135069280862808, -0.2901904284954071, 0.21595753729343414, -0.06657834351062775, -0.03549857437610626, -0.19768497347831726, -0.0266132615506649, 0.04863820970058441, -0.33357229828834534, -0.22930164635181427, -0.33347997069358826, -0.0837489664554596, 0.2851191461086273, -0.06061399728059769, -0.01924038678407669, -0.21448425948619843, 0.13713258504867554, -0.09409275650978088, 0.1131659746170044, 0.18199409544467926, 0.262392520904541, -0.22190825641155243, 0.151618093252182, -0.13297538459300995, -0.33784160017967224, 0.28923559188842773, 0.3065319359302521, 0.4667800962924957, 0.2628144323825836, 0.05958127602934837, 0.19852983951568604, 0.7363940477371216, 0.2467275857925415, 0.0307711623609066, 0.16282278299331665, 0.24470634758472443, -0.008579332381486893, 0.12864334881305695, -0.29556161165237427, 0.24368076026439667, -0.20570030808448792, 0.489105224609375, 0.046842485666275024, 0.12142494320869446, -0.2784973382949829, -0.3198149800300598, 0.09729421138763428, -0.29605522751808167, -0.24099452793598175, 0.22886088490486145, -0.2286773920059204, -0.029668159782886505, -0.0462922677397728, -0.37965264916419983, -0.30280715227127075, -0.03775858134031296, 0.024953655898571014, 0.36506718397140503, -0.08353621512651443, 0.1451680064201355, -0.3331809639930725, -0.05883096531033516, -0.47370094060897827, 0.11649078875780106, 0.02723412774503231, 0.30089789628982544, -0.07650651037693024, -0.10251370072364807, -0.09614039957523346, 0.09850404411554337, 0.7110233902931213, -0.4442327916622162, -0.1452292501926422, 0.14357194304466248, 0.09769856929779053, -0.09330227971076965, -0.12175722420215607, -0.286529541015625, 0.11032100766897202, 0.5207927823066711, 0.3619166314601898, -0.45472973585128784, -0.26354849338531494, 0.14256997406482697, -0.21809375286102295, -0.04638509079813957, -0.18992525339126587, -0.2505958378314972, -0.267026424407959, -0.09710871428251266, 0.042247943580150604, 0.22509314119815826, 0.2796405553817749, -0.11629276722669601, 0.2169063538312912, -0.07204927504062653, 0.17009702324867249, 0.21983088552951813, 0.0008597709238529205, -0.026834923774003983, 0.2566705346107483, 0.1444438099861145, 0.08391493558883667, -0.09641501307487488, -0.3558114767074585, 0.1244937926530838, -0.028675835579633713, 0.13755030930042267, 0.3451254367828369, -0.06954695284366608, 0.5981544256210327, 0.15671080350875854, 0.300218790769577, 0.22123482823371887, -0.13670648634433746, -0.013893641531467438, -0.1888938546180725, 0.23657336831092834, 0.21810726821422577, 0.20831653475761414, -0.14313693344593048, -0.24706344306468964, 0.4849056303501129, 0.25417226552963257, -0.13027405738830566, 0.476606547832489, -0.46295660734176636, -0.3232702612876892, 0.23429572582244873, 0.08237838000059128, 0.7584080100059509, -0.025766268372535706, 0.24822090566158295, 0.2218545377254486, -0.3872930407524109, 0.5620597004890442, -0.2961157560348511, 0.15221048891544342, -0.060211144387722015, -0.15607114136219025, -0.00347258523106575, -0.10052992403507233, -0.06659532338380814, 0.19200065732002258, -0.3005223870277405, 0.3170953691005707, 0.47413086891174316, -0.25450611114501953, -0.02542540617287159, 0.5520507097244263, 0.07409633696079254, -0.4449000060558319, -0.06325917690992355, 0.05981096625328064, -0.27754122018814087, 0.1279604285955429, -0.1231735423207283, -0.32199662923812866, 0.19604867696762085, 0.14368201792240143, -0.3340367078781128, 0.08418518304824829, 0.0025326660834252834, 0.2179051786661148, 0.3024013042449951, -0.018724113702774048, 0.24353772401809692, -0.15538766980171204, -0.06911833584308624, -0.2024182826280594, -0.08242576569318771, 0.12338769435882568, -0.15316855907440186, 0.11678958684206009, 0.09096910059452057, -0.23128847777843475, 0.5451806783676147, -0.09264493733644485, -0.07114842534065247, -0.24564054608345032, 0.22469687461853027, -0.40141743421554565, -0.20007510483264923, 0.20054173469543457, 0.16138680279254913, 0.1955033242702484, -0.2907208204269409, 0.5905134081840515, 0.12673096358776093, -0.003993481397628784, 0.01494215615093708, 0.22409652173519135, -0.23790514469146729, 0.4939917027950287, -0.33471783995628357, 0.03277841955423355, 0.10144218802452087, 0.04158319905400276, -0.10776830464601517, -0.43384358286857605, -0.0012157373130321503, 0.33424779772758484, 0.07582005858421326, 0.051432013511657715, -0.08145014941692352, 0.0753398984670639, -0.04821406304836273, -0.09353338181972504, -0.1852855235338211, -0.343398779630661, 0.08340653777122498, 0.2625080943107605, 0.14033231139183044, 0.22585955262184143, -0.062351763248443604, -0.30122557282447815, -0.18288251757621765, 0.4282068610191345, -0.09056003391742706, 0.17181049287319183, -0.1609405279159546, 0.31565916538238525, 0.41203197836875916, 0.46285828948020935, -0.24442191421985626, 0.016906801611185074, 0.022539852187037468, 0.09322279691696167, 0.3734447658061981, -0.2316666543483734, 0.07709383964538574, 0.24294085800647736, -0.13898323476314545, 0.04263536259531975, -0.34411028027534485, -0.12457103282213211, -0.26963165402412415, 0.11243967711925507, 0.13527867197990417, -0.04004848748445511, 0.004309933632612228, -0.13928300142288208, -0.1367930918931961, -0.06827154010534286, -0.03434345871210098, -0.40052008628845215, -0.10314901173114777, 0.08782876282930374, -0.04722622409462929, 0.06675795465707779, 0.06176084280014038, 0.16811507940292358, 0.05270761623978615, -0.08132509887218475, 0.3830561935901642, 0.09156361222267151, 0.07138912379741669, -0.18555381894111633, -0.16798439621925354, 0.05471840500831604, 0.27849191427230835, 0.1424303948879242, -0.04896392673254013, 0.06975305825471878, -0.2131606638431549, -0.1310153305530548, 0.21939092874526978, 0.2282593697309494, 0.05144962668418884, -0.19707223773002625, 0.1387762725353241, 0.08696110546588898, -0.07299178838729858, -0.02421017922461033, 0.2587135136127472, 0.061155978590250015, 0.058103013783693314, 0.15578004717826843, 0.13392274081707, 0.434079110622406, -0.18399573862552643, 0.04718610644340515, 0.168609157204628, -0.37249696254730225, 0.053001344203948975, -0.1925286501646042, -0.12392839789390564, 0.04347006604075432, 0.1779736578464508, -0.02071366086602211, 0.11280078440904617, -0.18534862995147705, 0.2707878351211548, 0.3122725486755371, 0.009716758504509926, 0.047666821628808975, 0.13717655837535858, -0.05930173769593239, -0.18108198046684265, 0.20019416511058807, 0.057559527456760406, 0.344394713640213, 0.018697895109653473, 0.8508671522140503, -0.27619698643684387, -0.3828795254230499, 0.17217466235160828, 0.40964171290397644, 0.03172902390360832, -0.18231992423534393, 0.08489298075437546, 0.07741566747426987, -0.34510067105293274, 0.01853763684630394, 0.03847333788871765, -0.17038504779338837, 0.6866495609283447, -0.13125543296337128, 0.11843216419219971, -0.22235529124736786, -0.2760984003543854, 0.3105562627315521, 0.17144030332565308, -0.05652574449777603, 0.06068401038646698, 0.13091380894184113, -0.09606015682220459, -0.16307991743087769, 0.4281635582447052, 0.20537415146827698, 0.07103633880615234, -0.4505877196788788, -0.05537906661629677, 0.10227978229522705, 0.11845840513706207, -0.08575428277254105, 0.1549680233001709, 0.44376876950263977, -0.12202142179012299, 0.2705669105052948, -0.023743653669953346, -0.02321956679224968, -0.13216739892959595, 0.16326655447483063, -0.3313332200050354, -0.3735797107219696, 0.4598417580127716, 0.0828370600938797, -0.11625508964061737, 0.096751868724823, 0.29102209210395813, -0.6527736783027649, -0.24476014077663422, 0.25609290599823, -0.061829231679439545, -0.0006644940003752708, -0.1802777647972107, 0.06906038522720337, 0.08660949766635895, 0.6433354616165161, 0.1645336002111435, 0.30600106716156006, -0.4014545679092407, -0.12584923207759857, -0.6316568851470947, -0.21747547388076782, 0.17937438189983368, -0.1384008228778839, -0.17036277055740356, 0.18058349192142487, -0.007043749094009399, 0.24776320159435272, -0.26092439889907837, -0.006971346214413643, 0.32927176356315613, 0.042966850101947784, -0.5192553400993347, 0.039342351257801056, 0.16072025895118713, 0.11304482817649841, 0.26529285311698914, -0.020683681592345238, -0.20508523285388947, 0.10758026689291, -0.06017396226525307, -0.0842016264796257, -0.11822482943534851, 0.23890498280525208, 0.2081911563873291, 0.005505956709384918, 0.06280198693275452, 0.4919825494289398, 0.00655258446931839, -0.1701352894306183, -0.0049099065363407135, 0.09330549091100693, -0.3385924696922302, -0.1612526774406433, -0.10824744403362274, 0.22235333919525146, -0.34201085567474365, 0.3595678508281708, -0.017937785014510155, -0.08411504328250885, 0.13189372420310974, 0.025731749832630157, -0.2636851668357849, 0.02853337861597538, -0.008091215044260025, 0.27323272824287415, 0.20005379617214203, 0.7247750163078308, 0.03976457193493843, 0.17260219156742096, -0.16154934465885162, -0.04141373932361603, 0.10445643961429596, -0.17149558663368225, -0.28233763575553894, -0.1106865257024765, 0.0700385794043541, 0.031754955649375916, 0.13340839743614197, -0.7072269320487976, -0.3839820623397827, 0.006639160681515932, -0.032671090215444565, 0.024425268173217773, 0.6573563814163208, 0.045901108533144, -0.10253019630908966, -0.13903044164180756, -0.32310807704925537, 0.09402622282505035, -0.23133821785449982, 0.18395085632801056, -0.3546333312988281 ]
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?
[ -0.089605912566185, -0.37051939964294434, -0.04412396624684334, -0.06419703364372253, -0.22869302332401276, 0.02895078808069229, 0.26768994331359863, 0.023359674960374832, 0.39509493112564087, -0.10933448374271393, -0.17059293389320374, 0.32033535838127136, -0.2015971690416336, 0.37548592686653137, 0.39844679832458496, -0.4378180503845215, -0.002576962113380432, -0.25555214285850525, -0.32734400033950806, 0.386793851852417, 0.0030521005392074585, 0.1184615045785904, -0.14524981379508972, 0.056555770337581635, -0.28478386998176575, -0.1290682703256607, -0.3320092260837555, 0.1075512170791626, 0.09705879539251328, -0.011513208039104939, 0.2389511913061142, 0.3638802170753479, -0.0029583927243947983, 0.1261327862739563, -0.00011568747140699998, -0.0413927361369133, 0.22455735504627228, -0.08851228654384613, 0.06489241123199463, -0.04569900035858154, 0.17900121212005615, -0.3761778771877289, 0.0587877556681633, -0.14258654415607452, 0.20260292291641235, 0.17821501195430756, 0.17752519249916077, 0.04718811810016632, 0.3459145426750183, -0.325434148311615, 0.08241245150566101, 0.32058289647102356, -0.20827698707580566, 0.26400527358055115, -0.11156594008207321, -0.05021269619464874, 0.07479416579008102, 0.3185981512069702, 0.6831016540527344, -0.5693524479866028, -0.07522604614496231, 0.10948421061038971, 0.01886456273496151, -0.06389616429805756, 0.15395689010620117, -0.08444515615701675, -0.30900242924690247, -0.24644428491592407, -0.4362654387950897, 0.2125432789325714, -0.11863141506910324, -0.1288323998451233, -0.2360723912715912, -0.6045428514480591, -0.029127713292837143, 0.07235994935035706, -0.2734473645687103, -0.009662054479122162, -0.21529704332351685, -0.049840640276670456, -0.17386050522327423, -0.1056186780333519, -0.031104769557714462, 0.14210346341133118, 0.3534389138221741, 0.5484317541122437, 0.07717563211917877, 0.21021884679794312, 0.41617828607559204, 0.0839209035038948, -0.11613135039806366, -0.27954208850860596, 0.3173259496688843, 0.12179888784885406, -0.362846702337265, -0.34425222873687744, 0.17837625741958618, -0.2727919816970825, 0.2653511166572571, 0.14319467544555664, 0.13283757865428925, 0.05661197006702423, -0.1277903914451599, 0.2421325445175171, 0.3584884703159332, 0.08159485459327698, -0.09061955660581589, -0.11054693162441254, 0.011070534586906433, -0.104486383497715, 0.22195321321487427, 0.10372351109981537, 0.1453361213207245, -0.015399766154587269, -0.47164249420166016, -0.08467715978622437, -0.2308618575334549, 0.11420813202857971, -0.18811652064323425, -0.5452074408531189, -0.0772758275270462, -0.14436036348342896, 0.11034156382083893, -0.0899566188454628, -0.2681082487106323, 0.11734405905008316, -0.05604647845029831, 0.2535909414291382, -0.13841092586517334, 0.0487503707408905, -0.04957445710897446, 0.06132807955145836, -0.45848968625068665, -0.09332135319709778, 0.2870364189147949, 0.1720787137746811, 0.12903091311454773, 0.12023808807134628, -0.019157595932483673, 0.14410105347633362, 0.33735769987106323, 0.028595805168151855, 0.013049967586994171, 0.010992974042892456, 0.18654197454452515, -0.30493977665901184, 0.011405512690544128, 0.23146317899227142, -0.34904929995536804, -0.10899230092763901, -0.20152640342712402, -0.191081702709198, 0.16736048460006714, 0.01393076404929161, -0.21046435832977295, -0.26141491532325745, -0.5597923994064331, 0.6944102644920349, -0.03270464390516281, -0.10934729874134064, -0.1405429244041443, 0.25908997654914856, -0.16008715331554413, -0.08462845534086227, 0.15515024960041046, 0.048239149153232574, -0.04864697903394699, -0.36557188630104065, -0.07868274301290512, -0.1202850490808487, 0.05071509629487991, 0.2679360508918762, -0.22832441329956055, 0.2701062560081482, 0.01878887228667736, 0.19740863144397736, 0.2476557493209839, -0.47684672474861145, -0.03297758102416992, 0.3645944595336914, -0.04434584081172943, 0.4032135009765625, 0.10897742956876755, 0.20473596453666687, 0.042717427015304565, -0.20092684030532837, 0.24048028886318207, 1.0242564678192139, -0.3771039843559265, 0.008823161944746971, -0.12261933088302612, -0.394635945558548, 0.541912853717804, 0.3336596190929413, 0.08739255368709564, -0.26467570662498474, -0.11319314688444138, 0.02093157172203064, 0.14002995193004608, 0.06807200610637665, 0.09149998426437378, -0.06338704377412796, -0.3728696405887604, 0.0450839102268219, -0.21324503421783447, -0.0941624864935875, -0.36404529213905334, 0.02004189044237137, -0.07089449465274811, 0.3300633728504181, 0.2768281102180481, -0.1019618809223175, 0.29011070728302, -0.25997963547706604, -0.135525643825531, -0.29892638325691223, 0.08099536597728729, -0.09198111295700073, 0.019625477492809296, -0.3017862141132355, -0.04352949559688568, 0.3114902675151825, 0.09314753115177155, -0.11146549880504608, -0.5201099514961243, 0.2931293249130249, -0.03599860519170761, 0.06156433746218681, -0.05629589036107063, 0.5926676392555237, -0.1611739993095398, -0.12305725365877151, 0.1540277600288391, 0.30346107482910156, -0.26907262206077576, 0.22715787589550018, 0.28728780150413513, 0.18800130486488342, 0.16485334932804108, -0.016521990299224854, 0.13337045907974243, 0.005181150510907173, -0.012539949268102646, -0.22247466444969177, -0.07112563401460648, 0.4943295121192932, -0.17993195354938507, 0.48739567399024963, -0.044665250927209854, 0.015739629045128822, -0.18632057309150696, -0.047534674406051636, -0.23269318044185638, 0.4075046479701996, 0.31928834319114685, -0.17949745059013367, 0.3725854158401489, -0.08623794466257095, -0.35132768750190735, 0.17856436967849731, 0.09627662599086761, 0.00019782409071922302, 0.021255694329738617, -0.012915138155221939, -0.009134769439697266, -0.3338388502597809, -0.11009958386421204, 0.3355424404144287, 0.6644570827484131, 0.24936822056770325, 0.17727503180503845, -0.02865043841302395, -0.19699712097644806, -0.09434285759925842, 0.03606759011745453, 0.21235433220863342, -0.024822061881422997, 0.15700215101242065, -0.010826410725712776, 0.03673912212252617, -0.001910082995891571, -0.14514292776584625, 0.14484195411205292, -0.20404064655303955, -0.0847463607788086, 0.08304194360971451, -0.20195826888084412, -0.6090875864028931, -0.24757292866706848, -0.017396800220012665, -0.32901883125305176, -0.2885229289531708, 0.10992706567049026, 0.08244891464710236, -0.15863145887851715, 0.14156243205070496, 0.22592276334762573, 0.6497545838356018, -0.4458458721637726, -0.2383548468351364, -0.0980750024318695, -0.2736438512802124, -0.117884062230587, 0.0897928923368454, 0.43891486525535583, 0.2987906038761139, 0.4678769111633301, 0.33576488494873047, -0.07294581085443497, 0.09495680779218674, -0.3517071604728699, 0.03553077206015587, -0.2930786609649658, 0.30988240242004395, 0.13900691270828247, -0.1824033409357071, 0.19600293040275574, -0.3770199120044708, -0.06796406954526901, -0.04905340075492859, -0.09277120977640152, -0.06871572136878967, -0.026185188442468643, 0.07875314354896545, -0.19790859520435333, -0.14304092526435852, -0.6475343108177185, -0.4276149272918701, 0.26240986585617065, -0.2563132047653198, 0.05381118506193161, 0.056188713759183884, -0.27896520495414734, 0.16724108159542084, -0.11408741027116776, -0.030198896303772926, -0.16675274074077606, -0.11565938591957092, -0.022315409034490585, -0.34140515327453613, -0.16686159372329712, -0.43680325150489807, -0.08814220875501633, 0.20312467217445374, -0.01946449466049671, -0.019430149346590042, -0.1868307739496231, 0.15789034962654114, -0.12860511243343353, 0.1338551938533783, 0.22555002570152283, 0.14002405107021332, -0.24906912446022034, 0.1449832171201706, -0.06924641877412796, -0.2159963995218277, 0.22665101289749146, 0.31023120880126953, 0.467271089553833, 0.20428210496902466, 0.021375082433223724, 0.1477930098772049, 0.7359363436698914, 0.2251024693250656, 0.06165266036987305, 0.15805065631866455, 0.2971978187561035, -0.01152196153998375, 0.10848712921142578, -0.20904192328453064, 0.2638307213783264, -0.1977590024471283, 0.48094671964645386, 0.08564785122871399, 0.1445295810699463, -0.18779566884040833, -0.2808536887168884, 0.008959051221609116, -0.2754811942577362, -0.22052472829818726, 0.24758362770080566, -0.27491945028305054, 0.008521296083927155, -0.07700526714324951, -0.36289533972740173, -0.28628259897232056, -0.039261214435100555, -0.007199836894869804, 0.3148205876350403, 0.03027878701686859, 0.16325710713863373, -0.3991277515888214, -0.0698733702301979, -0.4792785048484802, 0.12462051212787628, 0.05004148557782173, 0.2723785936832428, -0.07456652075052261, -0.1068066880106926, -0.0809648185968399, 0.13911302387714386, 0.6718359589576721, -0.4680737555027008, -0.12010304629802704, 0.04675668850541115, 0.08207770437002182, -0.059772148728370667, -0.12703664600849152, -0.2370416522026062, 0.020004617050290108, 0.4522354304790497, 0.35549548268318176, -0.41076329350471497, -0.2772553265094757, 0.22630614042282104, -0.16873975098133087, -0.08126749098300934, -0.22157131135463715, -0.21372485160827637, -0.17879274487495422, -0.12051863968372345, 0.022924095392227173, 0.19871008396148682, 0.2567111849784851, -0.17587724328041077, 0.1802804172039032, -0.09640815109014511, 0.2749904990196228, 0.18857160210609436, -0.004854738712310791, -0.061227165162563324, 0.19523654878139496, 0.1555696725845337, 0.07697056978940964, -0.113992840051651, -0.2888616919517517, 0.11771614104509354, -0.09440399706363678, 0.09064487367868423, 0.27966850996017456, -0.02780107781291008, 0.689308226108551, 0.08675223588943481, 0.33508390188217163, 0.26983949542045593, -0.08435344696044922, -0.015831395983695984, -0.2550775706768036, 0.17849378287792206, 0.1570008099079132, 0.2542005777359009, -0.15993042290210724, -0.26166704297065735, 0.5813794732093811, 0.30217254161834717, -0.07488243281841278, 0.4123922884464264, -0.5658517479896545, -0.318134605884552, 0.32106220722198486, 0.008235007524490356, 0.6767164468765259, -0.011694639921188354, 0.2966558635234833, 0.09642036259174347, -0.29173049330711365, 0.5918289422988892, -0.30680662393569946, 0.13567262887954712, -0.08629504591226578, -0.1570109874010086, -0.0007165279239416122, -0.1022060215473175, -0.06825951486825943, 0.2856675386428833, -0.26939907670021057, 0.34069037437438965, 0.45153963565826416, -0.25932610034942627, -0.062463678419589996, 0.49701982736587524, 0.0402609184384346, -0.3902255594730377, -0.053388237953186035, 0.03806687891483307, -0.2636737823486328, 0.15802627801895142, -0.1339806616306305, -0.2760329842567444, 0.2424040138721466, 0.09275698661804199, -0.2599031627178192, 0.10794831067323685, 0.00268064858391881, 0.21478167176246643, 0.31649458408355713, 0.024473756551742554, 0.24765875935554504, -0.18959127366542816, -0.15023545920848846, -0.17084519565105438, -0.06086115539073944, 0.1315786987543106, -0.14995667338371277, 0.07499956339597702, 0.102250836789608, -0.18703822791576385, 0.5334511399269104, -0.060373108834028244, -0.08038945496082306, -0.2575185298919678, 0.22078363597393036, -0.45324862003326416, -0.15675629675388336, 0.15572863817214966, 0.17293328046798706, 0.1708429902791977, -0.3983144462108612, 0.5807892680168152, 0.13632537424564362, 0.01606684923171997, -0.0007992703467607498, 0.16640503704547882, -0.22039063274860382, 0.5007694363594055, -0.3107818067073822, 0.009038269519805908, 0.07070198655128479, 0.006287731230258942, -0.12925174832344055, -0.41762956976890564, -0.04808884114027023, 0.29771342873573303, 0.0980951339006424, 0.050419896841049194, -0.009111844003200531, 0.1802223026752472, -0.11553867161273956, -0.03041183203458786, -0.11281539499759674, -0.29369795322418213, 0.13766293227672577, 0.18977539241313934, 0.2009892612695694, 0.19345170259475708, -0.080145925283432, -0.32933810353279114, -0.19233445823192596, 0.4907073974609375, -0.03517528250813484, 0.20113804936408997, -0.08926887065172195, 0.35346731543540955, 0.3922426402568817, 0.45730820298194885, -0.23124614357948303, 0.034198909997940063, 0.07259204983711243, 0.09041831642389297, 0.42385488748550415, -0.21112433075904846, 0.054924800992012024, 0.29967108368873596, -0.15968696773052216, 0.03289360553026199, -0.29686519503593445, -0.12285886704921722, -0.24775631725788116, 0.11770706623792648, 0.12449154257774353, -0.019673800095915794, 0.0004341951571404934, -0.10363554954528809, -0.059410009533166885, -0.056695543229579926, 0.01615842431783676, -0.3881985545158386, -0.07074013352394104, 0.0800229012966156, -0.05661024525761604, 0.0591755136847496, 0.018896304070949554, 0.30536022782325745, 0.03774084895849228, -0.04799099266529083, 0.3157254755496979, 0.12572357058525085, 0.08717876672744751, -0.15319716930389404, -0.1950373500585556, 0.01724487543106079, 0.30177685618400574, 0.17275962233543396, -0.03927139192819595, 0.16133424639701843, -0.16719713807106018, -0.15903237462043762, 0.22609040141105652, 0.22319786250591278, 0.040098268538713455, -0.1603206843137741, 0.11157452315092087, 0.08390440046787262, -0.09768814593553543, 0.01281345821917057, 0.1506921797990799, 0.11321600526571274, 0.061028238385915756, 0.1405305117368698, 0.1759590059518814, 0.45911073684692383, -0.25382113456726074, -0.0028148703277111053, 0.14726129174232483, -0.241718128323555, 0.09182550013065338, -0.07453352212905884, -0.04242352768778801, 0.022125471383333206, 0.17669157683849335, 0.016910109668970108, 0.12707734107971191, -0.3183577060699463, 0.2580792307853699, 0.3101721704006195, 0.005789011716842651, 0.014650922268629074, 0.09429790824651718, -0.06749280542135239, -0.18665936589241028, 0.3170960545539856, 0.00961320847272873, 0.3464251458644867, 0.04529720917344093, 0.8753758072853088, -0.333763062953949, -0.4275134801864624, 0.15399391949176788, 0.4489407539367676, 0.05686268210411072, -0.22211354970932007, 0.037358127534389496, 0.1282808780670166, -0.3681759238243103, 0.11881013214588165, 0.016469597816467285, -0.1489974409341812, 0.6261696815490723, -0.11506859958171844, 0.14155372977256775, -0.19208960235118866, -0.25172507762908936, 0.3088245391845703, 0.1510942578315735, -0.108719602227211, 0.003981549292802811, 0.08623930811882019, -0.03195403516292572, -0.08976559340953827, 0.4948278069496155, 0.24426431953907013, 0.044602397829294205, -0.4217642843723297, -0.05275574326515198, 0.06980518996715546, 0.15835505723953247, -0.09271027147769928, 0.14168429374694824, 0.45747143030166626, -0.1465977430343628, 0.2598172724246979, -0.01205808948725462, -0.028776679188013077, -0.10949544608592987, 0.10569151490926743, -0.40160274505615234, -0.36630117893218994, 0.45456865429878235, 0.09633444249629974, -0.10901318490505219, 0.05089858919382095, 0.2974779009819031, -0.6373802423477173, -0.23752887547016144, 0.30419301986694336, -0.12987762689590454, -0.07157358527183533, -0.18063612282276154, 0.06053189933300018, 0.05796731263399124, 0.6411347985267639, 0.24478359520435333, 0.35875189304351807, -0.41919270157814026, -0.12837077677249908, -0.5469263195991516, -0.25038862228393555, 0.14156486093997955, -0.07870977371931076, -0.26345714926719666, 0.1379135400056839, -0.0405026450753212, 0.25394564867019653, -0.2701578736305237, 0.10107435286045074, 0.23565635085105896, 0.09575535356998444, -0.43658798933029175, 0.08838328719139099, 0.13122275471687317, 0.10667721927165985, 0.25308340787887573, -0.095354363322258, -0.17096666991710663, 0.10993611812591553, -0.12099019438028336, -0.08666327595710754, -0.08444841206073761, 0.3115096390247345, 0.22627808153629303, 0.048147864639759064, 0.0510241873562336, 0.5658936500549316, 0.008747503161430359, -0.1783832162618637, 0.01966799423098564, 0.009437719359993935, -0.34356987476348877, -0.1645628809928894, -0.08115561306476593, 0.15427017211914062, -0.3199593722820282, 0.341757595539093, -0.01973627507686615, -0.11823641508817673, 0.10428188741207123, 0.0029659494757652283, -0.22956092655658722, -0.05096949636936188, 0.005004383623600006, 0.18253229558467865, 0.25139597058296204, 0.7283111810684204, 0.028254516422748566, 0.09515015035867691, -0.07251247763633728, 0.00252331979572773, 0.03102754056453705, -0.10433545708656311, -0.21575942635536194, -0.17576856911182404, 0.044877730309963226, 0.013303913176059723, 0.12388886511325836, -0.7102764844894409, -0.3642635643482208, 0.019617607817053795, 0.013225267641246319, 0.08402018249034882, 0.7129794359207153, 0.07473532855510712, -0.14252953231334686, -0.1384013146162033, -0.3290698826313019, 0.051101960241794586, -0.23991230130195618, 0.19118718802928925, -0.3383631706237793 ]
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!
[ -0.09565509855747223, -0.43038490414619446, -0.04629374295473099, -0.06745029985904694, -0.25789403915405273, 0.00994720309972763, 0.20224888622760773, 0.0762408971786499, 0.29039692878723145, -0.11092473566532135, -0.18591398000717163, 0.29462260007858276, -0.23009233176708221, 0.3789541721343994, 0.41002357006073, -0.45200076699256897, 0.024441152811050415, -0.17720666527748108, -0.29455074667930603, 0.3627450466156006, 0.015423532575368881, 0.14288915693759918, -0.15470293164253235, 0.08106376975774765, -0.3128843605518341, -0.1692613661289215, -0.34849420189857483, 0.1008625477552414, 0.13701575994491577, -0.02462046965956688, 0.2090308964252472, 0.3778529763221741, 0.03400932997465134, 0.1013946384191513, -0.00011142913717776537, -0.09728643298149109, 0.20507881045341492, -0.09062542021274567, 0.10297657549381256, -0.05900318920612335, 0.16401755809783936, -0.36792224645614624, -0.0027356501668691635, -0.1269502341747284, 0.15940651297569275, 0.22511357069015503, 0.20238101482391357, 0.08568350970745087, 0.39888113737106323, -0.30972468852996826, 0.11813074350357056, 0.3286946713924408, -0.18170085549354553, 0.2379731386899948, -0.08357029408216476, -0.06251056492328644, 0.060989417135715485, 0.2345285415649414, 0.6071350574493408, -0.5887547135353088, -0.06404925137758255, 0.14051803946495056, 0.038545072078704834, 0.021421082317829132, 0.14441192150115967, -0.03873608261346817, -0.29622358083724976, -0.2327665090560913, -0.436235636472702, 0.2806541621685028, -0.16127413511276245, -0.13062629103660583, -0.26345258951187134, -0.5633004903793335, -0.0008530186023563147, 0.0014622323215007782, -0.330098420381546, 0.0447884127497673, -0.21435779333114624, -0.03346632048487663, -0.1524394005537033, -0.060267433524131775, -0.040601130574941635, 0.11201082170009613, 0.3631240725517273, 0.5616533756256104, 0.08171787858009338, 0.13045671582221985, 0.37812310457229614, 0.058417245745658875, -0.19902943074703217, -0.25620630383491516, 0.283736914396286, 0.11374115943908691, -0.3466759920120239, -0.3419148325920105, 0.17506253719329834, -0.1930898129940033, 0.32280901074409485, 0.10108862072229385, 0.09349600970745087, 0.09113761782646179, -0.10029951483011246, 0.29205912351608276, 0.3476564884185791, 0.12236909568309784, -0.05660543590784073, -0.12424623966217041, 0.03975958377122879, -0.06336762011051178, 0.20475304126739502, 0.1504272222518921, 0.1986272633075714, -0.007035126909613609, -0.4239499568939209, -0.11359044164419174, -0.1826791614294052, 0.1420108526945114, -0.21429969370365143, -0.5234177112579346, -0.02995612658560276, -0.13044431805610657, 0.08226405084133148, -0.05673923343420029, -0.2853320837020874, 0.09736412763595581, -0.08805631101131439, 0.1967231184244156, -0.12846842408180237, 0.07483383268117905, -0.08253393322229385, 0.07176254689693451, -0.4690384864807129, -0.04859989881515503, 0.30152595043182373, 0.10744746029376984, 0.13133534789085388, 0.13816618919372559, -0.037743523716926575, 0.13288018107414246, 0.33942851424217224, 0.03031250089406967, -0.0003314018249511719, -0.03616652265191078, 0.1575368493795395, -0.33455827832221985, -0.04376903548836708, 0.15608330070972443, -0.4061703383922577, -0.15703800320625305, -0.1264195740222931, -0.21524950861930847, 0.133074089884758, 0.054886993020772934, -0.2345520704984665, -0.25436773896217346, -0.4572625458240509, 0.7054925560951233, -0.04623884707689285, -0.03340630233287811, -0.13866668939590454, 0.2801205515861511, -0.2386481612920761, -0.07164382934570312, 0.15914331376552582, 0.09545499086380005, 0.0009736567735671997, -0.3780563175678253, -0.03246134892106056, -0.10533332824707031, -0.06803711503744125, 0.31936413049697876, -0.17716695368289948, 0.18924599885940552, 0.05113614350557327, 0.20222042500972748, 0.24152293801307678, -0.49662354588508606, -0.03344199061393738, 0.30526047945022583, -0.03650673106312752, 0.35196825861930847, 0.07787430286407471, 0.1823500245809555, 0.09818743914365768, -0.1924564093351364, 0.26532721519470215, 1.0035319328308105, -0.38286933302879333, 0.02205587737262249, -0.12745912373065948, -0.4010966122150421, 0.485562801361084, 0.32760196924209595, 0.023421039804816246, -0.31363844871520996, -0.13007831573486328, -0.025605181232094765, 0.09931136667728424, 0.06988118588924408, 0.076529361307621, -0.0697745680809021, -0.31873446702957153, 0.013977974653244019, -0.15631629526615143, -0.13527324795722961, -0.42362886667251587, 0.03415093570947647, -0.06946314871311188, 0.3329486846923828, 0.32506078481674194, -0.13765057921409607, 0.21762609481811523, -0.2634049654006958, -0.10048367083072662, -0.3129211664199829, 0.14184829592704773, -0.10447484254837036, 0.05392620339989662, -0.2799592614173889, -0.0749286562204361, 0.3310255706310272, 0.08667045831680298, -0.12241573631763458, -0.46758055686950684, 0.24261586368083954, -0.10650403052568436, 0.01754605397582054, 0.0072994716465473175, 0.6171311140060425, -0.14959514141082764, -0.07000717520713806, 0.2050333321094513, 0.28201788663864136, -0.2821090817451477, 0.2547474801540375, 0.2886514663696289, 0.22653207182884216, 0.18389946222305298, 0.01291370764374733, 0.10476886481046677, -0.018239326775074005, -0.03309943526983261, -0.1843896210193634, -0.1276172399520874, 0.5337749123573303, -0.11340467631816864, 0.42374786734580994, 0.010280590504407883, -0.0016263797879219055, -0.18574842810630798, -0.08734390139579773, -0.2569582760334015, 0.3532719314098358, 0.32422077655792236, -0.21554210782051086, 0.3575747609138489, -0.09278329461812973, -0.3873850405216217, 0.1576668620109558, 0.18133816123008728, 0.022451937198638916, 0.08836881071329117, 0.02248459868133068, -0.04399360716342926, -0.28928107023239136, -0.0826866552233696, 0.3609902560710907, 0.6006553769111633, 0.3052535951137543, 0.1733636111021042, 0.02706008218228817, -0.30429166555404663, -0.11205678433179855, 0.01522761583328247, 0.18314236402511597, -0.030197111889719963, 0.15285545587539673, 0.04158672317862511, 0.05599968880414963, -0.04895707964897156, -0.16611436009407043, 0.09252741187810898, -0.17580817639827728, -0.0818902924656868, 0.039575956761837006, -0.1671431064605713, -0.5780366659164429, -0.2460935115814209, 0.04412256181240082, -0.3262541890144348, -0.2550859749317169, 0.15884125232696533, 0.029434477910399437, -0.16992655396461487, 0.1729968786239624, 0.2292766571044922, 0.6286777257919312, -0.43886107206344604, -0.18500059843063354, -0.10761904716491699, -0.3222368061542511, -0.15262538194656372, 0.12096057832241058, 0.4160206913948059, 0.29050713777542114, 0.4932442307472229, 0.28309208154678345, -0.02600112557411194, 0.10474321991205215, -0.37736454606056213, 0.0544515997171402, -0.2754265069961548, 0.28514426946640015, 0.11039245128631592, -0.2209576517343521, 0.15785640478134155, -0.39220544695854187, -0.05006009340286255, -0.12310837209224701, -0.09991288930177689, -0.08157984912395477, -0.026580937206745148, 0.042091358453035355, -0.19186259806156158, -0.1675722897052765, -0.6334234476089478, -0.46810728311538696, 0.34189003705978394, -0.21221736073493958, 0.10471151024103165, 0.05770028755068779, -0.260682612657547, 0.16616946458816528, -0.15556229650974274, 0.02795892022550106, -0.16115033626556396, -0.08060229569673538, 0.02716047689318657, -0.3650359511375427, -0.16097742319107056, -0.42607778310775757, -0.13156574964523315, 0.2117251604795456, -0.08573517948389053, -0.01816057786345482, -0.25038695335388184, 0.07165047526359558, -0.13314282894134521, 0.11693156510591507, 0.26750215888023376, 0.1917031705379486, -0.23755095899105072, 0.09199521690607071, -0.030880069360136986, -0.23764263093471527, 0.24059653282165527, 0.3334428071975708, 0.39254167675971985, 0.14052046835422516, 0.030540477484464645, 0.19585338234901428, 0.7353606820106506, 0.25642988085746765, 0.023826051503419876, 0.18662703037261963, 0.262111634016037, 0.02463659644126892, 0.1601390242576599, -0.16981855034828186, 0.25950542092323303, -0.17571207880973816, 0.4492540955543518, 0.12856148183345795, 0.18453162908554077, -0.2333071380853653, -0.24578745663166046, -0.01603533700108528, -0.26773956418037415, -0.20620620250701904, 0.24365229904651642, -0.2004663497209549, -0.005440384149551392, -0.09856343269348145, -0.3882952928543091, -0.2598908245563507, -0.032904163002967834, 0.01779084838926792, 0.28791072964668274, 0.002848885953426361, 0.16430744528770447, -0.3194672763347626, -0.05563710257411003, -0.4845142960548401, 0.14956317842006683, 0.0722145363688469, 0.2689267098903656, -0.06599932909011841, -0.10260532796382904, -0.11336347460746765, 0.08600842952728271, 0.6084622740745544, -0.46098771691322327, -0.12899242341518402, 0.029675118625164032, 0.09475579857826233, -0.12481449544429779, -0.10218705236911774, -0.26836177706718445, -0.06343230605125427, 0.46247395873069763, 0.37933453917503357, -0.3683632016181946, -0.30867621302604675, 0.25277942419052124, -0.21676290035247803, -0.055353112518787384, -0.23541662096977234, -0.21979960799217224, -0.1814945936203003, -0.1528942734003067, 0.09308324754238129, 0.22387365996837616, 0.17905890941619873, -0.10410721600055695, 0.12050007283687592, -0.07052141427993774, 0.20699217915534973, 0.23102611303329468, 0.03242017328739166, -0.02489125356078148, 0.20212386548519135, 0.1299932450056076, 0.07291962206363678, -0.12812314927577972, -0.3435955345630646, 0.052055779844522476, -0.07871638238430023, 0.1674790382385254, 0.2720282971858978, 0.038946542888879776, 0.6369619965553284, 0.09319525957107544, 0.34834831953048706, 0.24644875526428223, -0.10002975910902023, 0.028285130858421326, -0.26440945267677307, 0.11702169477939606, 0.19253741204738617, 0.24081192910671234, -0.1332411915063858, -0.1824222058057785, 0.614434003829956, 0.32102546095848083, -0.07534254342317581, 0.391568660736084, -0.4802115559577942, -0.29888367652893066, 0.3047894239425659, 0.04767713323235512, 0.6780330538749695, -0.008658498525619507, 0.2885777950286865, 0.10952302813529968, -0.35490506887435913, 0.566443920135498, -0.3462432622909546, 0.15785478055477142, -0.1039699837565422, -0.19105485081672668, 0.007184026762843132, -0.08794987201690674, -0.0487314909696579, 0.26887238025665283, -0.346708744764328, 0.34354397654533386, 0.4566826820373535, -0.2356591522693634, -0.0808056890964508, 0.5609844923019409, -0.04154323786497116, -0.40955644845962524, -0.08284197747707367, 0.07587035745382309, -0.28342705965042114, 0.1413346827030182, -0.17460066080093384, -0.2745565176010132, 0.25787800550460815, 0.1064663827419281, -0.27389299869537354, 0.09171746671199799, -0.050028443336486816, 0.17283344268798828, 0.27562379837036133, 0.03175937384366989, 0.1917441040277481, -0.20281726121902466, -0.06302553415298462, -0.19644108414649963, -0.07966642826795578, 0.1545015126466751, -0.14524778723716736, 0.1055343821644783, 0.08784648030996323, -0.20984166860580444, 0.5108199119567871, -0.053067002445459366, -0.10169781744480133, -0.2674711346626282, 0.2604370415210724, -0.404293417930603, -0.17751352488994598, 0.1069343164563179, 0.13696087896823883, 0.20796982944011688, -0.2924808859825134, 0.562729001045227, 0.09027546644210815, -0.005863044410943985, 0.0553344264626503, 0.15226761996746063, -0.27232444286346436, 0.5453824996948242, -0.2550191581249237, 0.01121600717306137, 0.051902733743190765, 0.011895991861820221, -0.14888283610343933, -0.36376792192459106, -0.08748200535774231, 0.2741847634315491, 0.09738309681415558, -0.0001752413809299469, -0.013708025217056274, 0.14800365269184113, -0.11333446204662323, -0.02041766792535782, -0.12666328251361847, -0.3535959720611572, 0.1792508363723755, 0.21609561145305634, 0.255737841129303, 0.2160109281539917, -0.07465711236000061, -0.2411779910326004, -0.1948138028383255, 0.481928288936615, 0.027669668197631836, 0.2639501988887787, -0.13278573751449585, 0.3500426709651947, 0.41728389263153076, 0.5425980091094971, -0.2901419401168823, 0.019488610327243805, 0.0454469695687294, 0.11256442964076996, 0.45981836318969727, -0.24196511507034302, 0.07676403969526291, 0.28588104248046875, -0.13011102378368378, 0.07015644013881683, -0.36212149262428284, -0.17070725560188293, -0.26530221104621887, 0.10153092443943024, 0.10833417624235153, 0.047189466655254364, 0.003176143392920494, -0.019687756896018982, -0.07955944538116455, -0.09573731571435928, 0.01848379150032997, -0.3760214149951935, -0.09828223288059235, 0.15372036397457123, -0.06953905522823334, 0.04239383339881897, 0.12043493986129761, 0.20509055256843567, 0.04669894278049469, -0.06932926177978516, 0.32367491722106934, 0.12583911418914795, 0.07482606917619705, -0.152202770113945, -0.16026020050048828, 0.007074616849422455, 0.228685662150383, 0.13647708296775818, -0.06157446652650833, 0.16175296902656555, -0.2716624438762665, -0.13608995079994202, 0.24081090092658997, 0.14597174525260925, 0.07060440629720688, -0.1579345017671585, 0.11631929874420166, 0.14452940225601196, -0.11241625249385834, 0.0010166900465264916, 0.20374658703804016, 0.05654647946357727, 0.1279764175415039, 0.15175622701644897, 0.16456280648708344, 0.45382368564605713, -0.2142961472272873, 0.036483701318502426, 0.1422029286623001, -0.37099385261535645, 0.1153038889169693, -0.13379791378974915, -0.025103040039539337, 0.024295121431350708, 0.1567724496126175, -0.009453639388084412, 0.21674735844135284, -0.34670162200927734, 0.286684513092041, 0.31906288862228394, -0.03136871010065079, 0.055136896669864655, 0.19727474451065063, -0.0020191464573144913, -0.16688492894172668, 0.27943533658981323, 0.0712997242808342, 0.3188144564628601, 0.05534609407186508, 0.8951666951179504, -0.24943052232265472, -0.4379865825176239, 0.0884794294834137, 0.3954373002052307, 0.024619698524475098, -0.22906865179538727, 0.10322864353656769, 0.08457624167203903, -0.3302842974662781, 0.09990286827087402, 0.04174529388546944, -0.1267833560705185, 0.6448746919631958, -0.08233323693275452, 0.14766822755336761, -0.15032018721103668, -0.23305943608283997, 0.3292797803878784, 0.17817792296409607, -0.11302310228347778, 0.01350913755595684, 0.09902162849903107, -0.10855188965797424, -0.1989586353302002, 0.5276767611503601, 0.24399083852767944, 0.055927641689777374, -0.4059012234210968, 0.007024959195405245, 0.11586366593837738, 0.1438957005739212, -0.0914529487490654, 0.10737704485654831, 0.45036906003952026, -0.14932742714881897, 0.2122393548488617, 0.03401569277048111, -0.06872900575399399, -0.09852886199951172, 0.03311299532651901, -0.4560564458370209, -0.3647690713405609, 0.45689019560813904, 0.10242491215467453, -0.08989409357309341, 0.042917583137750626, 0.30867794156074524, -0.6284581422805786, -0.18853935599327087, 0.2987588047981262, -0.13009721040725708, -0.07465946674346924, -0.226114422082901, 0.08871965855360031, 0.07111990451812744, 0.5859489440917969, 0.21885667741298676, 0.3504166305065155, -0.40086090564727783, -0.09598863869905472, -0.6091051697731018, -0.20071932673454285, 0.11933498084545135, -0.08799003809690475, -0.2137719988822937, 0.1967511624097824, -0.10631450265645981, 0.2746657431125641, -0.1899932473897934, 0.11173813045024872, 0.1823192536830902, 0.019365869462490082, -0.410337895154953, 0.15257962048053741, 0.09875050187110901, 0.011352420784533024, 0.30674174427986145, -0.09572204947471619, -0.17017151415348053, 0.08457707613706589, -0.07569790631532669, -0.08605210483074188, -0.051421307027339935, 0.2711256146430969, 0.177827849984169, 0.04539426043629646, 0.05552592873573303, 0.46028751134872437, -0.04975072667002678, -0.1465172916650772, -0.02097504958510399, -0.02138887718319893, -0.39996081590652466, -0.17749813199043274, -0.08906073868274689, 0.17003121972084045, -0.3348320424556732, 0.2429094910621643, -0.045284342020750046, -0.10608506947755814, 0.12069018185138702, 0.01244242861866951, -0.13162121176719666, -0.03240446001291275, 0.03830353915691376, 0.1873740255832672, 0.22846576571464539, 0.7360840439796448, 0.037493448704481125, 0.08689642697572708, -0.09599662572145462, -0.05491313338279724, 0.05435870587825775, -0.0933588370680809, -0.23394675552845, -0.17539487779140472, 0.07947839796543121, 0.09167444705963135, 0.07787445187568665, -0.6713752746582031, -0.35351088643074036, 0.010525266639888287, 0.010181214660406113, 0.07833001017570496, 0.7260480523109436, 0.0725812315940857, -0.15513916313648224, -0.15334570407867432, -0.35617515444755554, 0.06008966267108917, -0.2716900408267975, 0.18187421560287476, -0.4164087772369385 ]
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.
[ -0.09692946076393127, -0.4622396230697632, -0.0760454311966896, 0.017250370234251022, -0.3047633171081543, 0.05286409705877304, 0.22632108628749847, 0.0648951306939125, 0.27689695358276367, -0.11790058016777039, -0.17251448333263397, 0.2042178213596344, -0.25846895575523376, 0.4444766640663147, 0.4080318808555603, -0.41051334142684937, 0.01412723958492279, -0.2733161449432373, -0.25001782178878784, 0.39677631855010986, 0.05055207014083862, 0.08481955528259277, -0.17099374532699585, 0.06342427432537079, -0.2406720668077469, -0.1341973841190338, -0.33941224217414856, 0.08957769721746445, 0.13409462571144104, 0.027393341064453125, 0.14990709722042084, 0.3057301640510559, 0.027069568634033203, 0.14519339799880981, -0.00011703865311574191, -0.10317887365818024, 0.22462084889411926, -0.054976049810647964, 0.1232357919216156, -0.038753315806388855, 0.11764193326234818, -0.4690667986869812, 0.0013235188089311123, -0.0907580778002739, 0.2729416787624359, 0.1990501880645752, 0.24085494875907898, 0.05113822966814041, 0.30715295672416687, -0.3036268651485443, 0.08462028950452805, 0.2716985046863556, -0.23924672603607178, 0.3131469190120697, -0.04093202203512192, -0.03336239606142044, 0.06543538719415665, 0.2598797082901001, 0.7037147283554077, -0.5388643741607666, 0.0035533681511878967, 0.11596065759658813, 0.028103845193982124, -0.02197091281414032, 0.13279148936271667, -0.07223215699195862, -0.2183319330215454, -0.2454034984111786, -0.3490258753299713, 0.20940148830413818, -0.05419587343931198, -0.08673951774835587, -0.21434032917022705, -0.5842260718345642, -0.01856834813952446, 0.05227081477642059, -0.31707563996315, 0.1362728774547577, -0.20855456590652466, -0.07745707035064697, -0.19316914677619934, -0.10364755243062973, -0.01954682543873787, 0.10086800158023834, 0.37002772092819214, 0.5515717267990112, 0.05214407667517662, 0.16430193185806274, 0.34780994057655334, 0.022836746647953987, -0.10691865533590317, -0.2457597851753235, 0.2928490936756134, 0.12592414021492004, -0.3565714359283447, -0.31134793162345886, 0.18628336489200592, -0.2371802031993866, 0.2858080267906189, 0.0631810873746872, 0.138044074177742, 0.03951591998338699, -0.10145283490419388, 0.20296594500541687, 0.38573843240737915, 0.0993378609418869, -0.08045104891061783, -0.14554136991500854, 0.02665870450437069, -0.16947418451309204, 0.19106584787368774, 0.15759554505348206, 0.1533164232969284, -0.08121535927057266, -0.4698123335838318, -0.13959723711013794, -0.24632911384105682, 0.12415027618408203, -0.18019165098667145, -0.5548356175422668, -0.0691334456205368, -0.20435214042663574, 0.02727869153022766, -0.12810075283050537, -0.28909409046173096, 0.13472476601600647, -0.08762967586517334, 0.214901864528656, -0.13607242703437805, 0.0736110731959343, -0.04291800782084465, 0.06124958395957947, -0.4702683687210083, -0.10851015150547028, 0.2724396586418152, 0.18824900686740875, 0.07662239670753479, 0.11885008960962296, -0.04511091113090515, 0.11725416779518127, 0.3278566598892212, 0.015721550211310387, -0.005909577012062073, 0.018595241010189056, 0.21519234776496887, -0.30850884318351746, -0.09372919797897339, 0.26277410984039307, -0.38236239552497864, -0.049495212733745575, -0.22280102968215942, -0.17501145601272583, 0.15861161053180695, 0.013053161092102528, -0.13337019085884094, -0.24674835801124573, -0.5358600616455078, 0.697175920009613, 0.02937692403793335, -0.061253421008586884, -0.15396425127983093, 0.2474062740802765, -0.2110527604818344, -0.1469140499830246, 0.08802923560142517, 0.06152331084012985, -0.07157882302999496, -0.3473765552043915, -0.03984358534216881, -0.08777216076850891, 0.016239985823631287, 0.3379491865634918, -0.19569717347621918, 0.2697061002254486, 0.07623136043548584, 0.12101085484027863, 0.2956903874874115, -0.476119726896286, -0.04213965684175491, 0.390820175409317, -0.029752977192401886, 0.35634106397628784, 0.12997449934482574, 0.24874745309352875, -0.019198745489120483, -0.1718997359275818, 0.271645188331604, 0.9539661407470703, -0.40128350257873535, 0.03243052214384079, -0.1284985989332199, -0.42581912875175476, 0.543549656867981, 0.2887674570083618, 0.09507393836975098, -0.29588890075683594, -0.08115897327661514, 0.05353764444589615, 0.1472449004650116, 0.11203370988368988, 0.12126529216766357, -0.06787087023258209, -0.39558130502700806, 0.023653171956539154, -0.19152000546455383, -0.10706014931201935, -0.4190426766872406, -0.00041070953011512756, -0.09667230397462845, 0.28670555353164673, 0.34507930278778076, -0.07108791172504425, 0.30363717675209045, -0.31333690881729126, -0.08020229637622833, -0.3391849398612976, 0.07497363537549973, -0.11293651163578033, 0.008267991244792938, -0.3027854859828949, -0.06041332334280014, 0.3214498460292816, 0.13272058963775635, -0.15883849561214447, -0.46373850107192993, 0.21984809637069702, -0.06655080616474152, 0.07297089695930481, -0.00666128471493721, 0.6013066172599792, -0.2415640652179718, -0.09500335901975632, 0.14528265595436096, 0.2679252028465271, -0.299379825592041, 0.23697508871555328, 0.27708864212036133, 0.26271796226501465, 0.21460813283920288, 0.0410454235970974, 0.18638215959072113, -0.10739432275295258, -0.06513874232769012, -0.21716970205307007, -0.07372831553220749, 0.46423768997192383, -0.13604456186294556, 0.4431438744068146, -0.04130062460899353, -0.0566461980342865, -0.23851515352725983, -0.04269199073314667, -0.20221060514450073, 0.36900603771209717, 0.346292644739151, -0.18958206474781036, 0.29735490679740906, -0.10137459635734558, -0.3327430486679077, 0.15534891188144684, 0.11895051598548889, 0.09320920705795288, 0.05557187274098396, -0.012828655540943146, -0.03845524787902832, -0.2931990623474121, -0.14500270783901215, 0.3609068989753723, 0.6997568607330322, 0.21742130815982819, 0.18000757694244385, 0.027522051706910133, -0.15481916069984436, -0.1363481879234314, 0.01572834700345993, 0.17999452352523804, -0.04321042820811272, 0.17419418692588806, 0.020082473754882812, 0.04398118332028389, 0.05878257006406784, -0.13080242276191711, 0.1434115320444107, -0.20149387419223785, -0.047282423824071884, 0.007521247491240501, -0.2275293916463852, -0.6512364149093628, -0.31002578139305115, 0.024343740195035934, -0.32361945509910583, -0.24214555323123932, 0.07698924094438553, -0.0048471540212631226, -0.14679472148418427, 0.09195514023303986, 0.1915389895439148, 0.6404305100440979, -0.4318293333053589, -0.1751062422990799, -0.024588823318481445, -0.25083744525909424, -0.15639027953147888, 0.09369319677352905, 0.46999046206474304, 0.24401961266994476, 0.44433027505874634, 0.28168702125549316, -0.06684494763612747, 0.1097661629319191, -0.36882132291793823, 0.022475695237517357, -0.23586614429950714, 0.2514667510986328, 0.13616536557674408, -0.2515857219696045, 0.19181431829929352, -0.33062925934791565, -0.0595003105700016, -0.10711772739887238, -0.0802377462387085, -0.14460302889347076, -0.017944375053048134, 0.06091843172907829, -0.2195328027009964, -0.16286656260490417, -0.6563556790351868, -0.4274236559867859, 0.2951757311820984, -0.2354220300912857, 0.09205666184425354, 0.043100208044052124, -0.27662020921707153, 0.16108903288841248, -0.11539875715970993, 0.005506050772964954, -0.11043908447027206, -0.12340137362480164, -0.06430646777153015, -0.3830776810646057, -0.1486847698688507, -0.3794017434120178, -0.032762687653303146, 0.21915201842784882, -0.013218598440289497, -0.06567730754613876, -0.13860058784484863, 0.09972245246171951, -0.19404825568199158, 0.07510499656200409, 0.28883856534957886, 0.07945150882005692, -0.319480299949646, 0.13742370903491974, -0.041642311960458755, -0.24689237773418427, 0.1678505539894104, 0.3495776653289795, 0.4843297600746155, 0.18409183621406555, 0.022530749440193176, 0.16180536150932312, 0.7138544321060181, 0.23926421999931335, 0.0766744539141655, 0.12391601502895355, 0.2778066396713257, -0.027644697576761246, 0.1562185138463974, -0.1513795107603073, 0.34679901599884033, -0.13601285219192505, 0.4500523805618286, 0.12394806742668152, 0.21397534012794495, -0.1946689486503601, -0.26240062713623047, -0.008469192311167717, -0.26845115423202515, -0.1703488528728485, 0.2559317946434021, -0.25556179881095886, 0.04936530441045761, -0.06720227003097534, -0.47130462527275085, -0.26852676272392273, -0.04070999100804329, -0.04501011222600937, 0.3134015202522278, -0.07623101025819778, 0.19009241461753845, -0.3363175690174103, -0.1390690803527832, -0.45425480604171753, 0.14053946733474731, 0.0639503002166748, 0.31060701608657837, -0.04748229682445526, -0.06474945694208145, -0.08007955551147461, 0.16260182857513428, 0.632561206817627, -0.48288804292678833, -0.14991649985313416, 0.01430671289563179, 0.10305802524089813, -0.08692990243434906, -0.1750025898218155, -0.24498915672302246, 0.004539084155112505, 0.479754239320755, 0.3700738549232483, -0.3253914415836334, -0.2802450656890869, 0.22733959555625916, -0.20513926446437836, -0.05097851902246475, -0.2206481695175171, -0.22033728659152985, -0.12599384784698486, -0.0952446386218071, 0.07316489517688751, 0.17764103412628174, 0.24161872267723083, -0.15876393020153046, 0.18788014352321625, -0.020658457651734352, 0.25508973002433777, 0.19459588825702667, -0.011927492916584015, -0.09128223359584808, 0.25237223505973816, 0.11457347869873047, -0.0004393830895423889, -0.0798569768667221, -0.2460368275642395, 0.07207027077674866, -0.10971249639987946, 0.190028578042984, 0.3117481768131256, -0.054916299879550934, 0.6821686029434204, 0.07290752232074738, 0.3269188702106476, 0.2302166223526001, -0.09951400756835938, 0.04348275437951088, -0.1924857199192047, 0.13781525194644928, 0.17686715722084045, 0.22169776260852814, -0.18922297656536102, -0.2239363044500351, 0.528230607509613, 0.36822310090065, -0.021432846784591675, 0.3635031282901764, -0.6384041905403137, -0.29502901434898376, 0.24399606883525848, 0.03697985038161278, 0.6201876997947693, 0.0012556388974189758, 0.33115464448928833, -0.028299134224653244, -0.3404572010040283, 0.5145268440246582, -0.34301185607910156, 0.17477640509605408, -0.06714172661304474, -0.1301722675561905, 0.022789793089032173, -0.10523968935012817, -0.034400612115859985, 0.31344133615493774, -0.27450644969940186, 0.3359679579734802, 0.456587016582489, -0.252425879240036, -0.12036937475204468, 0.5019968152046204, -0.0001567639410495758, -0.35710829496383667, -0.02680341899394989, 0.025746513158082962, -0.24538461863994598, 0.13652129471302032, -0.2202281355857849, -0.2273554801940918, 0.22728466987609863, 0.15298545360565186, -0.3146008253097534, 0.043000757694244385, -0.022224413231015205, 0.1911669373512268, 0.2674507200717926, 0.009535141289234161, 0.28806072473526, -0.20601823925971985, -0.11121505498886108, -0.15486127138137817, -0.06647570431232452, 0.09129486232995987, -0.11013207584619522, 0.09081569314002991, 0.08554096519947052, -0.18792080879211426, 0.4651317000389099, -0.0855831652879715, -0.04987715184688568, -0.23923173546791077, 0.24239274859428406, -0.3915676474571228, -0.22097721695899963, 0.20412571728229523, 0.19909922778606415, 0.23439092934131622, -0.3496168255805969, 0.5967292189598083, 0.1162034422159195, 0.06519228219985962, 0.0013032052665948868, 0.17796741425991058, -0.1788136065006256, 0.5910233855247498, -0.314308762550354, -0.025047242641448975, 0.04015878587961197, -0.06175950914621353, -0.03130720555782318, -0.43369677662849426, -0.07607503980398178, 0.29465779662132263, 0.14731964468955994, 0.04011417180299759, 0.0717901960015297, 0.06344659626483917, -0.02486107125878334, 0.006633184850215912, -0.08219344168901443, -0.27645164728164673, 0.21565306186676025, 0.21482691168785095, 0.1768282800912857, 0.24704399704933167, -0.06320678442716599, -0.2921832203865051, -0.11573338508605957, 0.46784716844558716, -0.030524883419275284, 0.21433034539222717, -0.14435702562332153, 0.42935043573379517, 0.48022541403770447, 0.530646562576294, -0.20600423216819763, 0.08634337782859802, 0.11954668909311295, 0.09168408066034317, 0.41203543543815613, -0.11529964953660965, 0.016234416514635086, 0.3485611081123352, -0.20053066313266754, 0.06389868259429932, -0.34794434905052185, -0.11550619453191757, -0.23809635639190674, 0.12806586921215057, 0.13986657559871674, 0.02418089471757412, 0.04037763178348541, -0.10755089670419693, -0.05919549986720085, -0.025316566228866577, 0.01723877340555191, -0.3293403387069702, -0.06795446574687958, 0.10453064739704132, -0.02551407739520073, 0.01560242846608162, 0.09028692543506622, 0.3568156063556671, 0.07297879457473755, -0.09809212386608124, 0.3494691550731659, 0.13369181752204895, 0.12953081727027893, -0.15184414386749268, -0.21585924923419952, -0.04204421490430832, 0.25913774967193604, 0.17298948764801025, -0.14142590761184692, 0.12014445662498474, -0.23301860690116882, -0.12351971864700317, 0.2052648663520813, 0.23218905925750732, 0.07684158533811569, -0.122011199593544, 0.13692417740821838, 0.06485418975353241, -0.04432760179042816, -0.004679646342992783, 0.18947122991085052, 0.09242015331983566, 0.021239016205072403, 0.1604587584733963, 0.18984167277812958, 0.491852343082428, -0.2660159468650818, 0.02571577951312065, 0.08390630781650543, -0.3478933572769165, 0.10506007075309753, -0.15809708833694458, -0.051642633974552155, 0.019299719482660294, 0.19063270092010498, -0.05377303436398506, 0.13644802570343018, -0.29196619987487793, 0.3588845729827881, 0.25594577193260193, 0.05868406966328621, -0.013341095298528671, 0.13964232802391052, -0.039308078587055206, -0.20437990128993988, 0.33291295170783997, 0.001500256359577179, 0.3294956684112549, 0.0227276012301445, 0.8239642381668091, -0.22451286017894745, -0.4310363233089447, 0.1677999198436737, 0.4035542905330658, 0.02263050153851509, -0.24922962486743927, 0.12385421991348267, 0.10800222307443619, -0.3104625344276428, 0.15777724981307983, 0.009082529693841934, -0.11470045149326324, 0.610750675201416, -0.0843510627746582, 0.16867589950561523, -0.1848052740097046, -0.23177820444107056, 0.36075055599212646, 0.06150561198592186, -0.11912199854850769, 0.031436216086149216, 0.01417490839958191, -0.08980607986450195, -0.13866446912288666, 0.5110017657279968, 0.19786587357521057, 0.081528440117836, -0.36255207657814026, -0.08625247329473495, 0.12035311758518219, 0.23069064319133759, -0.13932983577251434, 0.1242808923125267, 0.5147860050201416, -0.18943671882152557, 0.22483006119728088, -0.03141586855053902, -0.027002189308404922, -0.07785665988922119, 0.05451780557632446, -0.46020689606666565, -0.439922958612442, 0.43461498618125916, 0.04109242931008339, -0.11985763907432556, 0.05708657205104828, 0.3916843831539154, -0.6262600421905518, -0.21965065598487854, 0.3360733687877655, -0.1508597582578659, -0.08783331513404846, -0.24786406755447388, 0.0547993928194046, 0.061759378761053085, 0.6547306776046753, 0.2686643898487091, 0.36167648434638977, -0.3674938678741455, -0.13172833621501923, -0.5774894952774048, -0.18853691220283508, 0.09687363356351852, -0.026180535554885864, -0.2875889539718628, 0.23825030028820038, -0.11827684193849564, 0.21939203143119812, -0.1281169354915619, 0.10893572866916656, 0.17497077584266663, 0.07805293798446655, -0.42623019218444824, 0.11324359476566315, 0.0924319177865982, 0.08165434002876282, 0.2756972908973694, -0.0810507982969284, -0.145060196518898, 0.03609946370124817, -0.12800279259681702, -0.14890071749687195, -0.09801290929317474, 0.325968861579895, 0.22520777583122253, 0.09227751195430756, 0.03513903170824051, 0.5028657913208008, 0.028904594480991364, -0.19265466928482056, -0.07568523287773132, 0.004324430599808693, -0.35453489422798157, -0.16355344653129578, -0.058457862585783005, 0.1632135659456253, -0.4496609568595886, 0.3605155050754547, 0.04926588013768196, -0.11139626055955887, 0.09741541743278503, 0.021948814392089844, -0.09885645657777786, -0.09097623080015182, 0.08899564296007156, 0.15876539051532745, 0.28199058771133423, 0.7396944165229797, 0.018032066524028778, 0.0463823527097702, -0.09544427692890167, 0.010541828349232674, -0.01951901614665985, -0.07139810919761658, -0.21920379996299744, -0.20084403455257416, 0.07530634105205536, -0.03850242495536804, 0.08397188782691956, -0.6680576205253601, -0.4091363251209259, 0.031875986605882645, 0.04847077280282974, 0.072431281208992, 0.7247989177703857, 0.054000239819288254, -0.18400411307811737, -0.11854550242424011, -0.3672237694263458, 0.031231006607413292, -0.27500683069229126, 0.1617080718278885, -0.3264085054397583 ]
https://github.com/huggingface/datasets/issues/217
Multi-task dataset mixing
> @zphang is having different batch sizes per task actually helpful? Would be interesting to know as it's not something I've come across as a technique used by any MTL papers. I think having different batch sizes per task is particularly helpful in some scenarios where each task has different amount of data. For example, the problem I'm currently facing is one task has tens of thousands of samples while one task has a couple hundreds. I think in this case different batch size could help. But if using the same batch size is a lot simpler to implement, I guess it makes sense to go with that.
It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly.
108
Multi-task dataset mixing It seems like many of the best performing models on the GLUE benchmark make some use of multitask learning (simultaneous training on multiple tasks). The [T5 paper](https://arxiv.org/pdf/1910.10683.pdf) highlights multiple ways of mixing the tasks together during finetuning: - **Examples-proportional mixing** - sample from tasks proportionally to their dataset size - **Equal mixing** - sample uniformly from each task - **Temperature-scaled mixing** - The generalized approach used by multilingual BERT which uses a temperature T, where the mixing rate of each task is raised to the power 1/T and renormalized. When T=1 this is equivalent to equal mixing, and becomes closer to equal mixing with increasing T. Following this discussion https://github.com/huggingface/transformers/issues/4340 in [transformers](https://github.com/huggingface/transformers), @enzoampil suggested that the `nlp` library might be a better place for this functionality. Some method for combining datasets could be implemented ,e.g. ``` dataset = nlp.load_multitask(['squad','imdb','cnn_dm'], temperature=2.0, ...) ``` We would need a few additions: - Method of identifying the tasks - how can we support adding a string to each task as an identifier: e.g. 'summarisation: '? - Method of combining the metrics - a standard approach is to use the specific metric for each task and add them together for a combined score. It would be great to support common use cases such as pretraining on the GLUE benchmark before fine-tuning on each GLUE task in turn. I'm willing to write bits/most of this I just need some guidance on the interface and other library details so I can integrate it properly. > @zphang is having different batch sizes per task actually helpful? Would be interesting to know as it's not something I've come across as a technique used by any MTL papers. I think having different batch sizes per task is particularly helpful in some scenarios where each task has different amount of data. For example, the problem I'm currently facing is one task has tens of thousands of samples while one task has a couple hundreds. I think in this case different batch size could help. But if using the same batch size is a lot simpler to implement, I guess it makes sense to go with that.
[ -0.15286675095558167, -0.4851095676422119, -0.08828960359096527, 0.04388965666294098, -0.30230510234832764, 0.02592063695192337, 0.26860570907592773, 0.09124504774808884, 0.27223020792007446, -0.07411901652812958, -0.16744045913219452, 0.12118718028068542, -0.2572723925113678, 0.44997718930244446, 0.3634524643421173, -0.4062099754810333, 0.030634447932243347, -0.2730120122432709, -0.22879037261009216, 0.4066848158836365, 0.036920905113220215, 0.051403194665908813, -0.13749776780605316, 0.048646025359630585, -0.2384946346282959, -0.1524750143289566, -0.31967318058013916, 0.055579692125320435, 0.19754554331302643, 0.0446510948240757, 0.11467425525188446, 0.3015516400337219, 0.047315604984760284, 0.15700912475585938, -0.00011640706361504272, -0.08212161064147949, 0.22991494834423065, -0.020130040124058723, 0.07947728037834167, -0.019034460186958313, 0.13090094923973083, -0.5154120326042175, -0.03750036656856537, -0.10425849258899689, 0.27347856760025024, 0.21355989575386047, 0.2659730613231659, 0.03253559768199921, 0.25440889596939087, -0.2947164475917816, 0.09862411767244339, 0.32211872935295105, -0.2970360517501831, 0.2996929883956909, -0.010349847376346588, -0.014742083847522736, 0.05328204482793808, 0.26293087005615234, 0.6999955177307129, -0.49931326508522034, 0.0012699216604232788, 0.16061906516551971, 0.027350855991244316, 0.02736656367778778, 0.1041140928864479, -0.05754975974559784, -0.14754286408424377, -0.2259558141231537, -0.3079743981361389, 0.24689793586730957, -0.08192551881074905, -0.06607357412576675, -0.23695698380470276, -0.6513186097145081, -0.007324228063225746, 0.04993279278278351, -0.350864440202713, 0.186945840716362, -0.18846456706523895, -0.09649688005447388, -0.15846090018749237, -0.05361082777380943, -0.007652949541807175, 0.06648024171590805, 0.37570899724960327, 0.505257248878479, 0.013429434970021248, 0.1736924797296524, 0.3467084467411041, 0.03121255710721016, -0.10651373118162155, -0.24471548199653625, 0.2853563129901886, 0.10458505898714066, -0.370259553194046, -0.2722153663635254, 0.1617216169834137, -0.2643088102340698, 0.2712612748146057, 0.023664042353630066, 0.1709330826997757, 0.021655306220054626, -0.0688675120472908, 0.17013107240200043, 0.40976235270500183, 0.1125168651342392, -0.09519147127866745, -0.1455383598804474, 0.04109416529536247, -0.1776212751865387, 0.18310831487178802, 0.20799770951271057, 0.14859290421009064, -0.13576693832874298, -0.46563732624053955, -0.08216294646263123, -0.25629374384880066, 0.149321511387825, -0.20054562389850616, -0.5421443581581116, -0.04683934897184372, -0.23184806108474731, -0.00907492358237505, -0.11373697966337204, -0.29090937972068787, 0.14404302835464478, -0.11999651044607162, 0.22714763879776, -0.16093432903289795, 0.07435385882854462, -0.05789913237094879, 0.004685141146183014, -0.44629794359207153, -0.1151442751288414, 0.2692190110683441, 0.16773712635040283, 0.08348635584115982, 0.0706988051533699, -0.06737266480922699, 0.1179807186126709, 0.28367072343826294, -0.004396373406052589, -0.032980822026729584, 0.0396023765206337, 0.22117272019386292, -0.2768903374671936, -0.1352517157793045, 0.3050745725631714, -0.3997878432273865, -0.022525280714035034, -0.23434147238731384, -0.19428640604019165, 0.18709267675876617, 0.020744837820529938, -0.1286899745464325, -0.22317004203796387, -0.5973215103149414, 0.6830006837844849, 0.06751823425292969, -0.046523068100214005, -0.15667149424552917, 0.17821726202964783, -0.19402453303337097, -0.17894062399864197, 0.011592505499720573, 0.08312460780143738, -0.06165241450071335, -0.3461354076862335, 0.0005171513184905052, -0.11840665340423584, 0.01542605459690094, 0.3973841071128845, -0.1924615055322647, 0.28655773401260376, 0.09445212036371231, 0.12006457149982452, 0.2664092183113098, -0.4741748869419098, -0.049382515251636505, 0.45079997181892395, -0.019094228744506836, 0.35060596466064453, 0.1672840714454651, 0.26022273302078247, 0.0020201322622597218, -0.17475318908691406, 0.23725655674934387, 0.897196888923645, -0.4060541093349457, 0.08358100056648254, -0.13938406109809875, -0.42281439900398254, 0.5240345597267151, 0.3101251423358917, 0.10595034062862396, -0.3104816973209381, -0.06986040621995926, 0.04622741416096687, 0.13682276010513306, 0.09810622781515121, 0.12557418644428253, -0.057762593030929565, -0.4325707256793976, -0.0024137496948242188, -0.19192519783973694, -0.06833679974079132, -0.49648332595825195, 0.0032335594296455383, -0.15541516244411469, 0.2653364837169647, 0.373635470867157, -0.058589108288288116, 0.34154415130615234, -0.3247836232185364, -0.0682685375213623, -0.3196665942668915, 0.07444004714488983, -0.095672607421875, 0.03928646072745323, -0.3197908401489258, -0.06206240504980087, 0.3237813413143158, 0.12032066285610199, -0.14445088803768158, -0.42558753490448, 0.1923220306634903, -0.037724703550338745, 0.08093027770519257, 0.016697179526090622, 0.5754867196083069, -0.2911287546157837, -0.12075145542621613, 0.10930070281028748, 0.24802979826927185, -0.28013497591018677, 0.218608558177948, 0.21097564697265625, 0.2668554484844208, 0.2549041509628296, 0.12753558158874512, 0.2021847516298294, -0.1201392412185669, -0.09720626473426819, -0.2197658121585846, -0.03213990479707718, 0.4492030143737793, -0.09584534168243408, 0.4679342210292816, -0.04142171889543533, -0.1082647442817688, -0.22867216169834137, -0.04206334054470062, -0.20951564610004425, 0.35186707973480225, 0.3388044238090515, -0.1527581363916397, 0.3212238848209381, -0.087537482380867, -0.32107263803482056, 0.15290342271327972, 0.0971042811870575, 0.0753893256187439, 0.05987535044550896, -0.015522070229053497, -0.04942823201417923, -0.29519227147102356, -0.16740113496780396, 0.3495880365371704, 0.7306103110313416, 0.22298802435398102, 0.199818953871727, -0.0016516447067260742, -0.16649731993675232, -0.16180893778800964, -0.007260546088218689, 0.19259782135486603, -0.019021527841687202, 0.2063293755054474, 0.026077844202518463, 0.010703608393669128, 0.047556810081005096, -0.14258532226085663, 0.12812857329845428, -0.19433893263339996, -0.039021871984004974, 0.001227010041475296, -0.21815864741802216, -0.6630979180335999, -0.34492602944374084, 0.03966861218214035, -0.2795931100845337, -0.22961638867855072, 0.09637437760829926, 0.006200697273015976, -0.20221716165542603, 0.09744243323802948, 0.254486620426178, 0.6557924747467041, -0.4205453097820282, -0.1220015361905098, -0.011811692267656326, -0.2465246468782425, -0.18224331736564636, 0.08044181764125824, 0.4787087142467499, 0.1971724033355713, 0.4226318597793579, 0.25549378991127014, -0.07075680047273636, 0.08091455698013306, -0.3518063724040985, 0.025702081620693207, -0.2472161501646042, 0.24243246018886566, 0.11621172726154327, -0.23505277931690216, 0.19698333740234375, -0.33835020661354065, -0.03175120800733566, -0.11529534310102463, -0.07352644205093384, -0.1828647404909134, -0.022766556590795517, 0.04582684114575386, -0.21434494853019714, -0.14710214734077454, -0.6761797070503235, -0.423672616481781, 0.3288804292678833, -0.2589181661605835, 0.08312936127185822, 0.06865055114030838, -0.2425619661808014, 0.11010250449180603, -0.12082317471504211, 0.02017778903245926, -0.10749750584363937, -0.1358654946088791, -0.10516944527626038, -0.3834722936153412, -0.16408587992191315, -0.44173675775527954, -0.0516645610332489, 0.2295835167169571, 0.027340183034539223, -0.10455960780382156, -0.13729692995548248, 0.09618252515792847, -0.22930790483951569, 0.07182342559099197, 0.30790358781814575, 0.05160655081272125, -0.3560315668582916, 0.13332213461399078, -0.012046035379171371, -0.21837742626667023, 0.1295343041419983, 0.3778134286403656, 0.4913676977157593, 0.14920908212661743, 0.024773813784122467, 0.124551922082901, 0.7539142370223999, 0.20179390907287598, 0.05993613600730896, 0.10636746883392334, 0.24882110953330994, -0.016944710165262222, 0.1358460783958435, -0.11611085385084152, 0.39773058891296387, -0.13664868474006653, 0.42631685733795166, 0.17098723351955414, 0.19975197315216064, -0.2030813843011856, -0.1982661336660385, -0.04472185671329498, -0.2616236209869385, -0.1858825981616974, 0.28222328424453735, -0.2242661416530609, 0.052601393312215805, -0.08051186800003052, -0.43985608220100403, -0.28615260124206543, -0.030420005321502686, -0.0606733113527298, 0.27411603927612305, -0.09199515730142593, 0.1816534548997879, -0.3203604519367218, -0.15978267788887024, -0.4608957767486572, 0.1459830105304718, 0.08875874429941177, 0.3017204701900482, -0.012456808239221573, -0.08777286857366562, -0.07358933985233307, 0.20257049798965454, 0.6373814344406128, -0.448762983083725, -0.1601756364107132, -0.005617394112050533, 0.060697779059410095, -0.1331735998392105, -0.21247753500938416, -0.25151216983795166, -0.0027016708627343178, 0.46255674958229065, 0.3910315930843353, -0.35132133960723877, -0.29208096861839294, 0.19622765481472015, -0.16269074380397797, -0.05591331794857979, -0.23227892816066742, -0.24113264679908752, -0.10254239290952682, -0.10813840478658676, 0.11773519963026047, 0.15848436951637268, 0.23454245924949646, -0.16833293437957764, 0.16883863508701324, -0.03416452556848526, 0.21754100918769836, 0.17560702562332153, 0.0023399870842695236, -0.08446931838989258, 0.28032708168029785, 0.11757074296474457, -0.018795277923345566, -0.052353084087371826, -0.18998877704143524, 0.08903008699417114, -0.16537903249263763, 0.1914764791727066, 0.32191795110702515, -0.03718516603112221, 0.6493061780929565, 0.09719540178775787, 0.33552613854408264, 0.19747206568717957, -0.057175107300281525, 0.10656420886516571, -0.20644065737724304, 0.16744470596313477, 0.1632019430398941, 0.19573616981506348, -0.2039257436990738, -0.25002917647361755, 0.5089166164398193, 0.4079136550426483, -0.00573991984128952, 0.33328256011009216, -0.6443974375724792, -0.25987666845321655, 0.23563896119594574, 0.007103554904460907, 0.5431056618690491, 0.005898706614971161, 0.3782307207584381, -0.04366758465766907, -0.3120470643043518, 0.480791300535202, -0.329947829246521, 0.16691383719444275, -0.04985711723566055, -0.14279991388320923, 0.07683635503053665, -0.09651876986026764, -0.05870380252599716, 0.3432524800300598, -0.2126275897026062, 0.3445494771003723, 0.4961307644844055, -0.2248227298259735, -0.17989489436149597, 0.47908806800842285, -0.012490028515458107, -0.37976929545402527, 0.0026413705199956894, 0.029900487512350082, -0.2305547446012497, 0.1625853329896927, -0.21588605642318726, -0.20611076056957245, 0.2247900366783142, 0.13790561258792877, -0.3079603314399719, -0.02638571709394455, -0.056044258177280426, 0.19251877069473267, 0.2610135078430176, -0.017162799835205078, 0.25520721077919006, -0.24698877334594727, -0.1174718588590622, -0.17436350882053375, -0.05939333140850067, 0.0534650981426239, -0.0917482003569603, 0.11164909601211548, 0.07159196585416794, -0.196152463555336, 0.46401092410087585, -0.07571911811828613, -0.021194279193878174, -0.22428280115127563, 0.2308792769908905, -0.3635616898536682, -0.22265037894248962, 0.20926348865032196, 0.14877155423164368, 0.23195645213127136, -0.31442898511886597, 0.5743623971939087, 0.08713354170322418, 0.07795141637325287, 0.010979234240949154, 0.15689313411712646, -0.12796767055988312, 0.632997453212738, -0.2941191792488098, -0.07013058662414551, -0.006362538784742355, -0.08990371972322464, 0.01585724577307701, -0.44756096601486206, -0.06190013140439987, 0.27504709362983704, 0.13799363374710083, 0.028833702206611633, 0.09607603400945663, 0.04632272198796272, -0.07321404665708542, 0.024421604350209236, -0.05657602846622467, -0.26283934712409973, 0.21956144273281097, 0.2410237193107605, 0.18731345236301422, 0.27172982692718506, -0.0618717260658741, -0.2633691430091858, -0.10183098167181015, 0.46383529901504517, -0.013535238802433014, 0.24274982511997223, -0.09981803596019745, 0.4236372113227844, 0.4712507724761963, 0.5466607809066772, -0.20703789591789246, 0.07370539009571075, 0.13231545686721802, 0.09527561068534851, 0.39160600304603577, -0.05915419012308121, 0.04130518063902855, 0.40140238404273987, -0.18922612071037292, 0.08920939266681671, -0.34877458214759827, -0.12751220166683197, -0.22972330451011658, 0.1274731308221817, 0.15703818202018738, 0.03752823919057846, 0.05876712128520012, -0.08815532922744751, -0.013388663530349731, -0.044957131147384644, 0.06431731581687927, -0.2907070815563202, -0.07052233815193176, 0.08662541955709457, 0.019152451306581497, 0.014970462769269943, 0.12828858196735382, 0.4145223796367645, 0.07437916845083237, -0.08982579410076141, 0.3278343081474304, 0.1382105052471161, 0.1572459191083908, -0.12971729040145874, -0.21711261570453644, -0.08144602924585342, 0.271614134311676, 0.1716535985469818, -0.1672322154045105, 0.09358740597963333, -0.21850991249084473, -0.08033723384141922, 0.1846635341644287, 0.24251602590084076, 0.07674828171730042, -0.12868699431419373, 0.17664793133735657, 0.07307992875576019, -0.023402612656354904, -0.0032811930868774652, 0.13657668232917786, 0.09947994351387024, 0.039155375212430954, 0.2060372680425644, 0.2117108404636383, 0.5047973990440369, -0.23348025977611542, 0.01652560383081436, 0.06762964278459549, -0.3135484457015991, 0.09969207644462585, -0.13878469169139862, -0.016195103526115417, 0.028218304738402367, 0.2262643724679947, -0.08678199350833893, 0.15465793013572693, -0.30596089363098145, 0.32733529806137085, 0.2416602075099945, 0.03566158562898636, -0.023880433291196823, 0.12478074431419373, -0.03723393380641937, -0.20662887394428253, 0.3375938832759857, -0.029000192880630493, 0.31513136625289917, 0.009087376296520233, 0.7983553409576416, -0.2380809187889099, -0.3960985839366913, 0.15436698496341705, 0.4071566164493561, 0.008456403389573097, -0.2650139033794403, 0.10717974603176117, 0.13453203439712524, -0.2611269950866699, 0.20869114995002747, -0.047895628958940506, -0.0980665534734726, 0.5933933258056641, -0.06555300951004028, 0.17357707023620605, -0.16121459007263184, -0.26063650846481323, 0.34757816791534424, 0.008399970829486847, -0.12483534216880798, 0.05821116641163826, -0.029602721333503723, -0.09513767063617706, -0.10379637777805328, 0.5061762928962708, 0.18733710050582886, 0.09599249064922333, -0.34532904624938965, -0.12983378767967224, 0.14326107501983643, 0.2453797608613968, -0.18993660807609558, 0.11646673828363419, 0.4996979534626007, -0.19961269199848175, 0.2697572112083435, -0.025633160024881363, -0.051957614719867706, -0.04092942550778389, 0.044215790927410126, -0.48897045850753784, -0.41631606221199036, 0.4167274534702301, 0.03329203277826309, -0.09751677513122559, 0.05186869949102402, 0.38108333945274353, -0.6517300009727478, -0.19317051768302917, 0.30628684163093567, -0.1815161406993866, -0.08955494314432144, -0.24177534878253937, 0.05955817550420761, 0.039202868938446045, 0.7016072273254395, 0.2924351990222931, 0.32732290029525757, -0.3832719027996063, -0.08002852648496628, -0.571548342704773, -0.1767008900642395, 0.018201462924480438, 0.000060731545090675354, -0.2943786084651947, 0.26411956548690796, -0.12518629431724548, 0.22162367403507233, -0.0940609872341156, 0.08814497292041779, 0.14160360395908356, 0.07060161978006363, -0.4327945411205292, 0.09113295376300812, 0.050961028784513474, 0.08235752582550049, 0.30425775051116943, -0.07882454246282578, -0.13369040191173553, 0.05160808935761452, -0.13522720336914062, -0.17467184364795685, -0.08309835195541382, 0.31040307879447937, 0.21442070603370667, 0.10870148241519928, 0.01488940604031086, 0.4607813358306885, 0.05151563882827759, -0.18575263023376465, -0.12474758923053741, -0.022356001660227776, -0.3448331952095032, -0.1339934766292572, -0.05004814639687538, 0.17086800932884216, -0.4427906274795532, 0.3610150218009949, 0.04574093595147133, -0.12132153660058975, 0.11556059122085571, 0.034114062786102295, -0.050387606024742126, -0.06718190014362335, 0.12175998836755753, 0.12575607001781464, 0.279092401266098, 0.6992523670196533, 0.02689368464052677, 0.020200582221150398, -0.08986503630876541, -0.024383937940001488, -0.03306381404399872, -0.05258099362254143, -0.20456336438655853, -0.21896520256996155, 0.095838762819767, -0.060808442533016205, 0.1070447787642479, -0.6410905718803406, -0.38680341839790344, 0.05448320880532265, 0.06030602380633354, 0.058844875544309616, 0.6827636361122131, 0.07146307826042175, -0.2238549441099167, -0.11899231374263763, -0.37256860733032227, 0.041019268333911896, -0.29618901014328003, 0.1502608358860016, -0.3555028438568115 ]
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.
[ -0.10336709767580032, -0.43319427967071533, -0.02786410227417946, -0.014036096632480621, -0.2707732319831848, 0.08221616595983505, 0.15497680008411407, 0.07477396726608276, 0.3294964134693146, -0.09032988548278809, -0.16610202193260193, 0.26055219769477844, -0.1972145438194275, 0.4019600450992584, 0.3530553877353668, -0.464892715215683, -0.05361878126859665, -0.20055332779884338, -0.3664042055606842, 0.33402031660079956, 0.015646427869796753, 0.08674705028533936, -0.1794654130935669, 0.019072607159614563, -0.2719663977622986, -0.11051621288061142, -0.39493507146835327, 0.1083938479423523, 0.11043483018875122, -0.013923369348049164, 0.1876085251569748, 0.344235360622406, 0.06541892141103745, 0.05401723459362984, -0.00011230397649342194, -0.12137141078710556, 0.21756750345230103, -0.09191820025444031, 0.13066019117832184, -0.000581890344619751, 0.07844532281160355, -0.47050097584724426, -0.007732077036052942, -0.1358966827392578, 0.20232844352722168, 0.17046016454696655, 0.29044148325920105, 0.00826919823884964, 0.39299511909484863, -0.330892413854599, 0.09726569056510925, 0.25307148694992065, -0.15323781967163086, 0.26218825578689575, -0.1310252845287323, -0.02023126929998398, 0.10178595036268234, 0.2531665563583374, 0.6352720260620117, -0.534798264503479, -0.036832697689533234, 0.10696162283420563, 0.025654474273324013, 0.03074496239423752, 0.1634942889213562, -0.08338319510221481, -0.319130539894104, -0.33498701453208923, -0.4021603465080261, 0.23699527978897095, -0.14883258938789368, -0.08392999321222305, -0.24520662426948547, -0.5854561924934387, 0.0016671568155288696, 0.016477355733513832, -0.31532639265060425, 0.11600280553102493, -0.20372825860977173, -0.0343724749982357, -0.25125688314437866, -0.14696136116981506, -0.0434427373111248, 0.14494892954826355, 0.3343847692012787, 0.5547493696212769, 0.08374859392642975, 0.16105499863624573, 0.3695610463619232, 0.057474397122859955, -0.12632228434085846, -0.31179752945899963, 0.29279690980911255, 0.10796751827001572, -0.36916103959083557, -0.3550264537334442, 0.13920462131500244, -0.20298609137535095, 0.2949889004230499, 0.10299038887023926, 0.025053909048438072, 0.0836867168545723, -0.07621697336435318, 0.25609761476516724, 0.371712327003479, 0.09638257324695587, -0.08300160616636276, -0.06298963725566864, 0.002955302596092224, -0.06848469376564026, 0.143419548869133, 0.17668959498405457, 0.1238449364900589, 0.003416826482862234, -0.4794355034828186, -0.12967386841773987, -0.258689820766449, 0.07781024277210236, -0.15980178117752075, -0.47638750076293945, -0.06893062591552734, -0.11731699109077454, 0.029383741319179535, -0.08602423220872879, -0.2983452081680298, 0.11831432580947876, -0.09277249127626419, 0.23073945939540863, -0.0812850296497345, 0.03701058402657509, -0.04833853244781494, 0.07137130200862885, -0.4358130693435669, -0.058910809457302094, 0.2667330503463745, 0.12061843276023865, 0.10626675933599472, 0.16640640795230865, 0.013974253088235855, 0.15166302025318146, 0.3776472508907318, 0.016560018062591553, 0.005949541926383972, -0.06492958962917328, 0.21329852938652039, -0.3031078577041626, 0.0006738565862178802, 0.1707688421010971, -0.36183294653892517, -0.14570215344429016, -0.19035518169403076, -0.16295641660690308, 0.0947001725435257, 0.017278699204325676, -0.1942499876022339, -0.26328516006469727, -0.44688764214515686, 0.7332510352134705, 0.026847876608371735, -0.017139391973614693, -0.1802855134010315, 0.26064807176589966, -0.2235586792230606, -0.06396756321191788, 0.2086058259010315, 0.12475068867206573, -0.01690937951207161, -0.3514384627342224, -0.046797093003988266, -0.014810100197792053, 0.034520648419857025, 0.36155638098716736, -0.22166237235069275, 0.12773405015468597, 0.06172608584165573, 0.1461106836795807, 0.2180098593235016, -0.4322619140148163, -0.09815126657485962, 0.4060342013835907, -0.01648067682981491, 0.3736610412597656, 0.1182851642370224, 0.19480188190937042, 0.12559908628463745, -0.1956963837146759, 0.2478085160255432, 1.0400981903076172, -0.4225592613220215, 0.022193072363734245, -0.1649436354637146, -0.3881130814552307, 0.5052050352096558, 0.36283397674560547, 0.09461911022663116, -0.3206784129142761, -0.1631169617176056, -0.02661627158522606, 0.11877448856830597, 0.041435401886701584, 0.10451119393110275, -0.009234648197889328, -0.4052465260028839, -0.022486738860607147, -0.22144317626953125, -0.07816123217344284, -0.33089834451675415, 0.015293266624212265, -0.08741414546966553, 0.32746365666389465, 0.34056544303894043, -0.1058153510093689, 0.2896713316440582, -0.2984875738620758, -0.13058176636695862, -0.33861082792282104, 0.11447054147720337, -0.03951864689588547, 0.07346740365028381, -0.2365601360797882, -0.06718042492866516, 0.35986921191215515, 0.08904337137937546, -0.1385277658700943, -0.5096867680549622, 0.3090386390686035, -0.057400934398174286, 0.06364452093839645, -0.025107532739639282, 0.5965563654899597, -0.19375944137573242, -0.12890784442424774, 0.144881010055542, 0.22485199570655823, -0.3346216678619385, 0.1949298083782196, 0.33768734335899353, 0.1734292060136795, 0.20787860453128815, -0.08267894387245178, 0.17195725440979004, -0.029911600053310394, -0.09459737688302994, -0.1404891014099121, -0.03039642423391342, 0.500379741191864, -0.11950579285621643, 0.44350945949554443, -0.018835991621017456, 0.015261845663189888, -0.2615099847316742, -0.09255734086036682, -0.21653792262077332, 0.35943979024887085, 0.33664095401763916, -0.19295254349708557, 0.36073407530784607, -0.05211370438337326, -0.35946547985076904, 0.14874233305454254, 0.20372150838375092, 0.08420239388942719, 0.046032264828681946, 0.020731162279844284, -0.10165785998106003, -0.2789604663848877, -0.10182248055934906, 0.461579829454422, 0.5798784494400024, 0.27897465229034424, 0.19102543592453003, 0.0029391925781965256, -0.2511984705924988, -0.0801590234041214, 0.03535642474889755, 0.2158656269311905, -0.05848614126443863, 0.140546053647995, 0.024648314341902733, 0.009310554713010788, 0.01067931205034256, -0.11853775382041931, 0.14683032035827637, -0.18777890503406525, -0.01490294560790062, 0.06336440145969391, -0.19838862121105194, -0.616385817527771, -0.3052637577056885, -0.015071871690452099, -0.3515801727771759, -0.2869986593723297, 0.13720376789569855, -0.003792528063058853, -0.138868510723114, 0.16575412452220917, 0.12699683010578156, 0.5480082631111145, -0.42024415731430054, -0.17030328512191772, -0.02358926087617874, -0.23749884963035583, -0.12512165307998657, 0.10783182084560394, 0.47286322712898254, 0.2611252963542938, 0.43125152587890625, 0.28128984570503235, -0.03624339774250984, 0.12175818532705307, -0.34444907307624817, 0.04084836691617966, -0.2936261296272278, 0.2540331780910492, 0.011807881295681, -0.20614619553089142, 0.13463850319385529, -0.4294755458831787, -0.060518667101860046, -0.01937848888337612, -0.053349561989307404, -0.028743047267198563, -0.02822124771773815, 0.03259540721774101, -0.17166337370872498, -0.20089346170425415, -0.6588301658630371, -0.45629560947418213, 0.28352153301239014, -0.20028935372829437, 0.09491465985774994, 0.02516014315187931, -0.23395322263240814, 0.08987870067358017, -0.16715110838413239, -0.0019509382545948029, -0.12760667502880096, -0.12098366022109985, 0.08369611948728561, -0.3582075834274292, -0.16718408465385437, -0.47265341877937317, -0.1242687925696373, 0.24180172383785248, -0.07113957405090332, -0.07343463599681854, -0.18778802454471588, 0.07646550983190536, -0.19489236176013947, 0.07409609109163284, 0.26539111137390137, 0.10548775643110275, -0.2660580277442932, 0.10515215992927551, 0.002967294305562973, -0.13249753415584564, 0.2176908701658249, 0.345712274312973, 0.4796205461025238, 0.15022984147071838, 0.11335104703903198, 0.19090749323368073, 0.6792360544204712, 0.32960841059684753, 0.007964439690113068, 0.21351279318332672, 0.3103100657463074, 0.024073563516139984, 0.19754567742347717, -0.21085034310817719, 0.34230101108551025, -0.11590605974197388, 0.454321026802063, 0.10699529945850372, 0.151146799325943, -0.24436338245868683, -0.24342335760593414, 0.024328678846359253, -0.20640727877616882, -0.17803643643856049, 0.2593960165977478, -0.22649115324020386, 0.020161356776952744, -0.09770283102989197, -0.36491405963897705, -0.23557128012180328, -0.05439337342977524, -0.031282514333724976, 0.2283371239900589, 0.011614084243774414, 0.17223304510116577, -0.3531283736228943, -0.10165946185588837, -0.4612160325050354, 0.20466774702072144, 0.05774303153157234, 0.301775187253952, -0.01998644694685936, -0.08469295501708984, -0.1292436569929123, 0.12709130346775055, 0.5680522918701172, -0.4728291928768158, -0.2209477573633194, -0.025382880121469498, 0.09025241434574127, -0.09211970865726471, -0.10572792589664459, -0.21971237659454346, -0.025454366579651833, 0.41992703080177307, 0.3062419295310974, -0.36397260427474976, -0.27792614698410034, 0.2682923674583435, -0.1646314561367035, -0.07220476865768433, -0.14969685673713684, -0.21267953515052795, -0.13832350075244904, -0.17088139057159424, 0.0775844007730484, 0.12965193390846252, 0.19336852431297302, -0.13969531655311584, 0.16703511774539948, -0.10889354348182678, 0.2603113055229187, 0.1859169900417328, 0.05234674736857414, -0.027580726891756058, 0.19997909665107727, 0.17190344631671906, 0.05351324379444122, -0.10656812787055969, -0.23571938276290894, 0.1216953694820404, -0.032500628381967545, 0.07200022041797638, 0.29760974645614624, 0.03904136270284653, 0.6938973069190979, 0.042601972818374634, 0.367055743932724, 0.22503027319908142, -0.06475947052240372, 0.08747007697820663, -0.21818453073501587, 0.1336088329553604, 0.22997234761714935, 0.20788517594337463, -0.16687704622745514, -0.22111359238624573, 0.5431409478187561, 0.2985454499721527, -0.00961148738861084, 0.3315664231777191, -0.5390052199363708, -0.2511659264564514, 0.29675745964050293, 0.040164604783058167, 0.6300408840179443, -0.007101081311702728, 0.2938927710056305, -0.03298947587609291, -0.3371446132659912, 0.5710722804069519, -0.330727756023407, 0.19199493527412415, -0.14019986987113953, -0.24169905483722687, -0.00279105082154274, -0.0857037603855133, -0.07936576753854752, 0.2763812243938446, -0.33719825744628906, 0.36917930841445923, 0.4814932346343994, -0.2480582892894745, -0.1045876145362854, 0.5050842761993408, -0.05645810812711716, -0.3726613223552704, -0.11116635799407959, 0.06115485727787018, -0.27586644887924194, 0.09509329497814178, -0.11868716776371002, -0.3008018732070923, 0.27083778381347656, 0.12162388861179352, -0.24231112003326416, 0.16438499093055725, -0.050431910902261734, 0.1845231056213379, 0.25303924083709717, 0.008703038096427917, 0.27064093947410583, -0.1624799370765686, -0.08966235816478729, -0.11695253849029541, -0.07923764735460281, 0.11518245935440063, -0.14977088570594788, 0.012844467535614967, 0.11990493535995483, -0.2024591565132141, 0.4892711639404297, -0.009573008865118027, -0.13599589467048645, -0.21550917625427246, 0.3113110661506653, -0.38764089345932007, -0.17851831018924713, 0.10874960571527481, 0.18113838136196136, 0.22955557703971863, -0.35972678661346436, 0.5780019760131836, 0.1318092793226242, 0.027300912886857986, 0.03078719787299633, 0.14182829856872559, -0.2512468993663788, 0.5130019783973694, -0.27598828077316284, -0.038096215575933456, 0.04208821803331375, 0.007354706525802612, -0.08804501593112946, -0.35103803873062134, -0.03188447654247284, 0.283353328704834, 0.12923085689544678, 0.021556593477725983, 0.06551934033632278, 0.17462104558944702, -0.07396022230386734, -0.06540042906999588, -0.13226768374443054, -0.3194485902786255, 0.15650279819965363, 0.07593652606010437, 0.19611230492591858, 0.1613200157880783, -0.07628889381885529, -0.21275311708450317, -0.21801771223545074, 0.5112912654876709, 0.021172910928726196, 0.25781238079071045, -0.18163955211639404, 0.30822059512138367, 0.3887070417404175, 0.5697489976882935, -0.2556871175765991, 0.11726082861423492, 0.055237285792827606, 0.06974649429321289, 0.46103349328041077, -0.2023731768131256, 0.04363477602601051, 0.2892819941043854, -0.14841970801353455, 0.06235472857952118, -0.3408699035644531, -0.14281439781188965, -0.23465700447559357, 0.09880520403385162, 0.15117910504341125, -0.019357137382030487, 0.009457303211092949, -0.06798816472291946, -0.09655873477458954, -0.11638102680444717, 0.013011522591114044, -0.4335089921951294, -0.068189337849617, 0.11059035360813141, -0.03718041256070137, -0.045462995767593384, 0.07459377497434616, 0.23765985667705536, 0.001246139407157898, -0.020071543753147125, 0.3473857641220093, 0.13781443238258362, 0.13269758224487305, -0.16679096221923828, -0.27622732520103455, -0.01261458545923233, 0.2735847234725952, 0.1634385883808136, -0.07122205942869186, 0.10352066904306412, -0.24374069273471832, -0.08736574649810791, 0.26286137104034424, 0.20636211335659027, 0.0771169364452362, -0.17260372638702393, 0.15144741535186768, 0.11271845549345016, -0.07894737273454666, -0.0010420656763017178, 0.20892126858234406, 0.07174256443977356, 0.08645141124725342, 0.17023196816444397, 0.15795131027698517, 0.4282063841819763, -0.24424384534358978, -0.007555047050118446, 0.10386931896209717, -0.34369808435440063, 0.14428013563156128, -0.07468444108963013, -0.01689911261200905, -0.05834301933646202, 0.11644040048122406, -0.042675282806158066, 0.22477833926677704, -0.2821715474128723, 0.250227153301239, 0.30582141876220703, 0.00996827520430088, -0.042123544961214066, 0.20314428210258484, 0.0018033422529697418, -0.22454063594341278, 0.33291447162628174, 0.0851997509598732, 0.3855224847793579, 0.048954181373119354, 0.8560094237327576, -0.24788807332515717, -0.4657871425151825, 0.037295930087566376, 0.4859018623828888, 0.04150879383087158, -0.27084872126579285, 0.14337950944900513, 0.09598004072904587, -0.4118724465370178, 0.16141974925994873, 0.07498539984226227, -0.18442364037036896, 0.6622331142425537, -0.05848798155784607, 0.16207745671272278, -0.19061189889907837, -0.17011427879333496, 0.36753860116004944, 0.14315113425254822, -0.13632890582084656, 0.07419116050004959, 0.1387922167778015, -0.10969686508178711, -0.14287377893924713, 0.5203854441642761, 0.2531953752040863, 0.03317749872803688, -0.4345836639404297, -0.022072333842515945, 0.16021208465099335, 0.16314104199409485, -0.10989037156105042, 0.1235361397266388, 0.40921056270599365, -0.1710437536239624, 0.2681717872619629, 0.013457048684358597, -0.0643920972943306, -0.17000040411949158, -0.0010199770331382751, -0.4380497634410858, -0.38811764121055603, 0.4196394085884094, 0.11488635092973709, -0.05141694098711014, 0.019021131098270416, 0.2460399568080902, -0.687714159488678, -0.2315683513879776, 0.3309122920036316, -0.1353486329317093, -0.10280797630548477, -0.2420894056558609, 0.07230069488286972, -0.015390504151582718, 0.6274807453155518, 0.29352179169654846, 0.33659446239471436, -0.4087728261947632, -0.15738613903522491, -0.5566297173500061, -0.1426028609275818, 0.10983379185199738, -0.1562354564666748, -0.2828674912452698, 0.23817487061023712, -0.09585996717214584, 0.23567375540733337, -0.21898658573627472, 0.13313177227973938, 0.21304944157600403, 0.09838514029979706, -0.4491214454174042, 0.12216770648956299, 0.10001176595687866, 0.04428112506866455, 0.2792525291442871, -0.09820935130119324, -0.1715533584356308, 0.053720925003290176, -0.10322336852550507, -0.008149050176143646, -0.016879189759492874, 0.24552473425865173, 0.0991356372833252, 0.12277588248252869, 0.0672542005777359, 0.5235679149627686, -0.033524513244628906, -0.2191493958234787, 0.005622748285531998, -0.0402335561811924, -0.41220858693122864, -0.13063281774520874, -0.07771554589271545, 0.2438381016254425, -0.34682708978652954, 0.26382020115852356, -0.04028177261352539, -0.12499444931745529, 0.12921610474586487, 0.031963687390089035, -0.12255164980888367, -0.05809997022151947, 0.05785603076219559, 0.21735650300979614, 0.21858006715774536, 0.712125837802887, 0.03826141357421875, 0.10666997730731964, -0.028916187584400177, -0.024187881499528885, 0.0697459951043129, -0.08190585672855377, -0.22436916828155518, -0.2049376666545868, 0.02550646662712097, 0.06479815393686295, 0.07541317492723465, -0.6441550254821777, -0.3265171945095062, 0.011607188731431961, 0.004460446536540985, 0.04752181097865105, 0.7026143074035645, 0.11499851942062378, -0.14029523730278015, -0.133981391787529, -0.2580876052379608, -0.0005197040736675262, -0.32983556389808655, 0.20005002617835999, -0.30358052253723145 ]
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).
[ -0.0643157884478569, -0.46592262387275696, -0.08466394245624542, -0.03106280416250229, -0.2937632203102112, 0.014267697930335999, 0.27487197518348694, -0.0015316680073738098, 0.32739582657814026, -0.08819639682769775, -0.14839480817317963, 0.1096205785870552, -0.166634663939476, 0.4398263096809387, 0.3444587290287018, -0.3664792776107788, 0.03876933455467224, -0.2969401478767395, -0.27085453271865845, 0.3853917121887207, 0.03808407485485077, 0.041931621730327606, -0.13354316353797913, 0.08822565525770187, -0.28080931305885315, -0.2636697292327881, -0.38970765471458435, 0.03443445637822151, 0.1399758756160736, 0.051349662244319916, 0.23408935964107513, 0.307551771402359, 0.048616886138916016, 0.27919378876686096, -0.00011687174264807254, -0.08377620577812195, 0.2308100014925003, -0.01913651078939438, 0.11680890619754791, -0.0018155276775360107, 0.15380114316940308, -0.4783225357532501, 0.005464307498186827, -0.11083638668060303, 0.2739708125591278, 0.15680430829524994, 0.25918132066726685, 0.10646338760852814, 0.28866541385650635, -0.3447949290275574, 0.09772253036499023, 0.2746815085411072, -0.21083860099315643, 0.22634679079055786, -0.11846130341291428, -0.01419290155172348, 0.07735439389944077, 0.2606106102466583, 0.7055602073669434, -0.5410453081130981, -0.0727652832865715, 0.07944448292255402, 0.04223468899726868, 0.011969641782343388, 0.1090795248746872, -0.17149625718593597, -0.2628779411315918, -0.297519713640213, -0.32956618070602417, 0.22085723280906677, -0.08587924391031265, -0.09882543236017227, -0.26016151905059814, -0.6264215111732483, -0.03642530366778374, 0.03155381232500076, -0.3546121120452881, 0.16949914395809174, -0.2045310139656067, -0.1291305273771286, -0.234602689743042, -0.10024101287126541, -0.026420921087265015, 0.07539796084165573, 0.34746411442756653, 0.5395692586898804, 0.0419699102640152, 0.12721998989582062, 0.39268121123313904, -0.03826463222503662, -0.08897200226783752, -0.27377161383628845, 0.2567116618156433, 0.13433712720870972, -0.39920154213905334, -0.29389309883117676, 0.1546253263950348, -0.26450997591018677, 0.31528714299201965, 0.08219506591558456, 0.08174525946378708, 0.051612235605716705, -0.14183208346366882, 0.19767314195632935, 0.4014679193496704, 0.1253281980752945, -0.09977950900793076, -0.11220314353704453, 0.04722511023283005, -0.1615312397480011, 0.21224674582481384, 0.18241596221923828, 0.14265777170658112, -0.13860581815242767, -0.5396937131881714, -0.16719037294387817, -0.27625465393066406, 0.1414400339126587, -0.18073806166648865, -0.574489176273346, -0.02044694870710373, -0.15083536505699158, 0.020509587600827217, -0.09436807781457901, -0.27212539315223694, 0.09609212726354599, -0.11833249777555466, 0.26906847953796387, -0.16005444526672363, 0.0736631453037262, -0.05052114650607109, 0.04711295664310455, -0.47076326608657837, -0.1046871617436409, 0.2876485586166382, 0.17828719317913055, 0.09426694363355637, 0.18709905445575714, -0.03195817023515701, 0.08593122661113739, 0.30259835720062256, -0.06191828101873398, -0.05103866010904312, 0.039355069398880005, 0.24605825543403625, -0.26802659034729004, -0.087934210896492, 0.2486291378736496, -0.3791605830192566, -0.05903386324644089, -0.23170897364616394, -0.15932096540927887, 0.15850937366485596, -0.0022806283086538315, -0.1443961262702942, -0.2504304051399231, -0.5782618522644043, 0.6736925840377808, 0.06236782297492027, -0.05668482556939125, -0.16031387448310852, 0.2204204499721527, -0.14365476369857788, -0.18698127567768097, 0.0599556565284729, 0.018210165202617645, -0.032834481447935104, -0.3329956531524658, 0.03664219379425049, -0.03524135798215866, 0.003587670624256134, 0.4364960789680481, -0.1851608157157898, 0.28957483172416687, 0.08100872486829758, 0.12182895839214325, 0.2664524018764496, -0.4424835443496704, -0.011515280231833458, 0.41733318567276, 0.05032651871442795, 0.3307779133319855, 0.12340635061264038, 0.23189181089401245, 0.04421749711036682, -0.2073780745267868, 0.2358897477388382, 0.97413569688797, -0.37452682852745056, 0.09668097645044327, -0.05804971978068352, -0.46580228209495544, 0.4411907196044922, 0.2258761078119278, 0.08976149559020996, -0.2925753593444824, -0.07390519231557846, -0.054094135761260986, 0.057376813143491745, 0.08964527398347855, 0.09594698995351791, 0.047746121883392334, -0.4209632873535156, 0.0964864194393158, -0.18344923853874207, -0.050520528107881546, -0.44608891010284424, -0.010563749819993973, -0.08277759701013565, 0.19561369717121124, 0.3746560215950012, -0.09083861112594604, 0.3362386226654053, -0.3263072669506073, -0.09256401658058167, -0.34133198857307434, 0.06611593812704086, -0.12174712121486664, 0.09767484664916992, -0.30773478746414185, 0.0007403865456581116, 0.33542773127555847, 0.09459519386291504, -0.10449420660734177, -0.4976728558540344, 0.17593219876289368, -0.036461055278778076, -0.008630004711449146, 0.05814778804779053, 0.5408447980880737, -0.1912010908126831, -0.08414341509342194, 0.1043061912059784, 0.28971901535987854, -0.2685345411300659, 0.2316524237394333, 0.31108570098876953, 0.26073765754699707, 0.21135921776294708, 0.118392214179039, 0.2055724859237671, -0.10977359116077423, -0.09943695366382599, -0.22493010759353638, -0.10347156971693039, 0.46075230836868286, -0.13459336757659912, 0.5010107755661011, -0.02969755232334137, -0.07168645411729813, -0.18867169320583344, -0.017303455621004105, -0.1613268405199051, 0.45697250962257385, 0.3395251929759979, -0.17075707018375397, 0.3268982768058777, -0.09833280742168427, -0.3560428321361542, 0.05383462458848953, 0.022897612303495407, 0.05386388674378395, 0.08113538473844528, -0.0080033540725708, -0.030827999114990234, -0.29537585377693176, -0.1262146532535553, 0.3653463125228882, 0.6546183824539185, 0.22161778807640076, 0.2182420790195465, -0.06149466335773468, -0.22549623250961304, -0.15221518278121948, -0.025874458253383636, 0.2021447867155075, 0.006639918778091669, 0.16049888730049133, 0.026298027485609055, -0.005736664868891239, 0.08471231907606125, -0.1053382009267807, 0.11840752512216568, -0.22522202134132385, -0.08519266545772552, 0.005405334755778313, -0.22219176590442657, -0.6311858296394348, -0.3055891692638397, 0.015079880133271217, -0.30497050285339355, -0.23138725757598877, 0.03909820318222046, 0.07374662160873413, -0.11424049735069275, 0.08188289403915405, 0.21158628165721893, 0.637373685836792, -0.43466487526893616, -0.10998263210058212, -0.03147666156291962, -0.24741141498088837, -0.14098459482192993, 0.0675954818725586, 0.4684332013130188, 0.2097017914056778, 0.4605121910572052, 0.3219369053840637, -0.10924215614795685, 0.11420060694217682, -0.32258519530296326, 0.026337826624512672, -0.2005293071269989, 0.242381289601326, 0.04370402172207832, -0.15862421691417694, 0.176504448056221, -0.33668437600135803, -0.07314237952232361, -0.06108460575342178, -0.05285245180130005, -0.1735837459564209, -0.07025719434022903, 0.05648849159479141, -0.14883579313755035, -0.10039744526147842, -0.6727861166000366, -0.39646661281585693, 0.3388640284538269, -0.25696566700935364, 0.06550872325897217, 0.02542121335864067, -0.19426095485687256, 0.07539035379886627, -0.14416266977787018, -0.017736678943037987, -0.07496719062328339, -0.12288616597652435, -0.02914408966898918, -0.4028762876987457, -0.1595148891210556, -0.37320849299430847, -0.02167428284883499, 0.19834546744823456, 0.0170576311647892, -0.06768269091844559, -0.09868606925010681, 0.02911701239645481, -0.1789454072713852, 0.1095881536602974, 0.34813860058784485, 0.083240807056427, -0.3999960124492645, 0.15140101313591003, -0.055636562407016754, -0.24908709526062012, 0.14666146039962769, 0.2859276831150055, 0.4913572371006012, 0.21992048621177673, 0.008941875770688057, 0.15517903864383698, 0.7662396430969238, 0.2730218172073364, 0.02362293377518654, 0.0985933467745781, 0.28743621706962585, 0.02550457790493965, 0.12757091224193573, -0.11887826025485992, 0.3374464809894562, -0.17567282915115356, 0.4366201162338257, 0.06794396042823792, 0.20402318239212036, -0.16058145463466644, -0.23380237817764282, -0.0395277664065361, -0.32576462626457214, -0.17143847048282623, 0.25072628259658813, -0.2517654299736023, 0.07811987400054932, -0.1435822695493698, -0.43167445063591003, -0.29271435737609863, -0.016961131244897842, -0.0676080733537674, 0.3017239272594452, -0.081898532807827, 0.20029884576797485, -0.33506572246551514, -0.16427873075008392, -0.38935917615890503, 0.17401209473609924, -0.002374277450144291, 0.299874484539032, -0.037796445190906525, -0.014264572411775589, -0.07898098230361938, 0.17510312795639038, 0.5862996578216553, -0.4357353448867798, -0.14293211698532104, -0.014258298091590405, 0.1116923838853836, -0.1409677416086197, -0.19628870487213135, -0.25150614976882935, 0.053072646260261536, 0.5074841380119324, 0.37952733039855957, -0.3611524701118469, -0.31687700748443604, 0.19735835492610931, -0.19273711740970612, -0.0932251363992691, -0.22613464295864105, -0.21406161785125732, -0.16825458407402039, -0.0791085958480835, 0.14117708802223206, 0.13951720297336578, 0.15919187664985657, -0.2341109663248062, 0.14947552978992462, -0.07838371396064758, 0.28221195936203003, 0.17640581727027893, -0.018854636698961258, -0.07879453897476196, 0.24801918864250183, 0.10285070538520813, 0.017769701778888702, -0.08682434260845184, -0.16271844506263733, 0.03729332238435745, -0.131301611661911, 0.1557673066854477, 0.29673823714256287, 0.013812731020152569, 0.6570740938186646, 0.10724851489067078, 0.4018983542919159, 0.20059719681739807, -0.11568357050418854, 0.05698535218834877, -0.19889219105243683, 0.12703147530555725, 0.23611553013324738, 0.24626266956329346, -0.22283609211444855, -0.24729734659194946, 0.5878698229789734, 0.3816869556903839, -0.054454490542411804, 0.40503570437431335, -0.6244817972183228, -0.26055192947387695, 0.231800839304924, 0.024833209812641144, 0.599092423915863, -0.013116806745529175, 0.29962143301963806, -0.07944035530090332, -0.35769355297088623, 0.459354043006897, -0.3115878105163574, 0.1631142646074295, -0.07900004088878632, -0.08968059718608856, 0.07022058218717575, -0.13893498480319977, 0.018397986888885498, 0.2944927215576172, -0.235673725605011, 0.37945178151130676, 0.4121044874191284, -0.2506578862667084, -0.1553017944097519, 0.4758174419403076, -0.034751228988170624, -0.3831048309803009, 0.05561334267258644, 0.024736668914556503, -0.2695632576942444, 0.17046722769737244, -0.21481552720069885, -0.200601726770401, 0.23190727829933167, 0.16014274954795837, -0.35234177112579346, 0.023304615169763565, -0.016125215217471123, 0.22780883312225342, 0.2794572114944458, -0.049164604395627975, 0.2548297047615051, -0.16728553175926208, -0.10953828692436218, -0.12556107342243195, -0.03871750086545944, 0.10054507851600647, -0.11218119412660599, 0.03601551055908203, 0.03847880661487579, -0.17122697830200195, 0.43994638323783875, -0.05404619872570038, -0.08606556057929993, -0.2332594096660614, 0.22317694127559662, -0.2976759970188141, -0.2963246703147888, 0.1337805688381195, 0.12846246361732483, 0.15110750496387482, -0.31660398840904236, 0.5542105436325073, 0.16135236620903015, 0.0931546688079834, 0.007157033309340477, 0.16346269845962524, -0.12557080388069153, 0.6486729979515076, -0.3296874761581421, -0.010702870786190033, 0.03142017126083374, -0.09077974408864975, 0.006326589733362198, -0.4751659631729126, -0.05922219157218933, 0.2901131510734558, 0.11815185844898224, 0.04533172398805618, 0.09262455254793167, 0.0931677594780922, 0.026724770665168762, -0.010842042975127697, -0.10102419555187225, -0.2596004009246826, 0.17278236150741577, 0.16765490174293518, 0.18941771984100342, 0.19720377027988434, -0.05612855777144432, -0.27054280042648315, -0.10723855346441269, 0.4969039559364319, -0.01926862820982933, 0.21876312792301178, -0.11857306212186813, 0.3374301791191101, 0.4945826828479767, 0.5996097922325134, -0.1953844577074051, 0.09189862757921219, 0.1511169970035553, 0.09730382263660431, 0.4130692183971405, -0.085114024579525, 0.01588013395667076, 0.35309118032455444, -0.20409706234931946, 0.037824276834726334, -0.2977036237716675, -0.1182650551199913, -0.1918211579322815, 0.11325032263994217, 0.1212540715932846, 0.046959877014160156, 0.028699621558189392, -0.14802198112010956, 0.009298676624894142, -0.05196170136332512, 0.06106187030673027, -0.25682711601257324, -0.07784226536750793, 0.11179018765687943, 0.0003042444586753845, 0.08506157249212265, 0.14502769708633423, 0.37398841977119446, 0.04396069794893265, -0.06393073499202728, 0.3479340076446533, 0.10780222713947296, 0.17816533148288727, -0.15824896097183228, -0.21071095764636993, -0.05424991995096207, 0.2973994016647339, 0.14296922087669373, -0.1791614294052124, 0.08633353561162949, -0.1817280501127243, -0.12257352471351624, 0.14351685345172882, 0.2863256335258484, 0.08999732881784439, -0.11031541973352432, 0.21582332253456116, 0.061415016651153564, 0.006081629544496536, 0.002305425703525543, 0.17192593216896057, 0.1008213683962822, -0.004941221326589584, 0.1974208801984787, 0.17235052585601807, 0.43909674882888794, -0.2788703441619873, 0.029636621475219727, 0.09111108630895615, -0.34259915351867676, 0.09730203449726105, -0.1335470825433731, -0.0017445646226406097, -0.021511781960725784, 0.2189338356256485, 0.0036812033504247665, 0.14514659345149994, -0.3186529278755188, 0.34929797053337097, 0.2836127281188965, 0.041865069419145584, -0.03126334771513939, 0.1697319746017456, -0.13565263152122498, -0.23886196315288544, 0.2818862199783325, -0.007455453276634216, 0.2914676368236542, 0.006257347762584686, 0.8371669054031372, -0.27607426047325134, -0.4076002240180969, 0.11187384277582169, 0.40588104724884033, -0.008414365351200104, -0.24195267260074615, 0.05891671031713486, 0.1356607973575592, -0.21899615228176117, 0.18768543004989624, 0.005946461111307144, -0.15572084486484528, 0.6309082508087158, -0.0868590772151947, 0.15201355516910553, -0.13946127891540527, -0.1847788691520691, 0.3204897940158844, 0.06872246414422989, -0.04573506861925125, 0.10238135606050491, -0.0006037354469299316, -0.08211371302604675, -0.13777108490467072, 0.5545239448547363, 0.27132341265678406, 0.05069413036108017, -0.3948327600955963, -0.05802081897854805, 0.12167657911777496, 0.18451707065105438, -0.09898988157510757, 0.12961964309215546, 0.4572368562221527, -0.21894459426403046, 0.21530205011367798, -0.018487699329853058, -0.033718667924404144, -0.06840468943119049, 0.03277232125401497, -0.46991270780563354, -0.35822170972824097, 0.4419861137866974, 0.03924168273806572, -0.15338796377182007, 0.09869474172592163, 0.3518230617046356, -0.6634207963943481, -0.2220752239227295, 0.39307525753974915, -0.15513268113136292, -0.11853007227182388, -0.19786436855793, 0.05703792721033096, 0.12376956641674042, 0.650640606880188, 0.30440041422843933, 0.315585196018219, -0.33722466230392456, -0.15077969431877136, -0.5069020390510559, -0.21034398674964905, -0.015344548970460892, -0.04891415685415268, -0.31534719467163086, 0.21218296885490417, -0.13832110166549683, 0.16271944344043732, -0.1201680451631546, 0.09354358166456223, 0.22755911946296692, 0.18092060089111328, -0.460758239030838, 0.12454557418823242, 0.09379181265830994, 0.10597744584083557, 0.2719220519065857, -0.08295974135398865, -0.15479181706905365, 0.08311773091554642, -0.1303596794605255, -0.0644197016954422, -0.005010437220335007, 0.37014827132225037, 0.1331891119480133, 0.08900219202041626, 0.06753277778625488, 0.5516717433929443, 0.08134457468986511, -0.16196118295192719, -0.06577123701572418, 0.034855637699365616, -0.3513203561306, -0.1360032856464386, -0.08042474091053009, 0.1183290183544159, -0.4727519750595093, 0.30819860100746155, 0.06809171289205551, -0.1501397043466568, 0.051804423332214355, 0.016798460856080055, -0.035492897033691406, -0.10962076485157013, 0.10881707817316055, 0.15477821230888367, 0.34218019247055054, 0.717380702495575, 0.013488717377185822, 0.051124148070812225, -0.11207713931798935, 0.016379019245505333, -0.041447192430496216, -0.10039984434843063, -0.20996205508708954, -0.24219325184822083, 0.09355348348617554, -0.043549247086048126, 0.12603627145290375, -0.5950586199760437, -0.34497857093811035, 0.003527292050421238, 0.05697786062955856, 0.09147070348262787, 0.6922088265419006, 0.08021368086338043, -0.19983115792274475, -0.12055380642414093, -0.3673868775367737, 0.051314473152160645, -0.29429328441619873, 0.0956127792596817, -0.31406641006469727 ]
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.
[ -0.033307574689388275, -0.5136857032775879, -0.02777966670691967, 0.06664924323558807, -0.25032734870910645, 0.04263012856245041, 0.1753961741924286, 0.060455355793237686, 0.20672665536403656, -0.1286959946155548, -0.2596333622932434, 0.17233319580554962, -0.16145704686641693, 0.5526630878448486, 0.3689039647579193, -0.42828068137168884, -0.033644311130046844, -0.22553157806396484, -0.3311319649219513, 0.3586280345916748, 0.16255369782447815, 0.12674082815647125, -0.1310737431049347, 0.011625438928604126, -0.3748592734336853, -0.05233381688594818, -0.3290127217769623, -0.014158792793750763, 0.07128327339887619, 0.1565447300672531, 0.14056597650051117, 0.343787282705307, -0.045731376856565475, 0.3042815923690796, -0.00012068864452885464, -0.0008572041988372803, 0.061606451869010925, -0.12618549168109894, 0.11212606728076935, -0.03864641487598419, 0.2874492406845093, -0.3399031162261963, 0.053131043910980225, -0.17077600955963135, 0.13239771127700806, 0.13732117414474487, 0.23634573817253113, 0.07320478558540344, 0.4259689450263977, -0.22620758414268494, 0.042793527245521545, 0.46675702929496765, -0.28901153802871704, 0.29696953296661377, 0.003143012523651123, -0.037487998604774475, -0.0336274728178978, 0.21709588170051575, 0.7237206697463989, -0.404282808303833, 0.06341977417469025, 0.04022970795631409, -0.013890817761421204, -0.011367347091436386, 0.21762865781784058, -0.007717118598520756, -0.3298623263835907, -0.2981940805912018, -0.3776232600212097, 0.249018132686615, 0.0038080736994743347, -0.18128225207328796, -0.22731944918632507, -0.5786259174346924, -0.052297983318567276, -0.008305491879582405, -0.3058779537677765, 0.02888064831495285, -0.28044208884239197, -0.07985904812812805, -0.15180154144763947, -0.13654440641403198, -0.11704954504966736, 0.18049582839012146, 0.3751619756221771, 0.46168068051338196, 0.06983566284179688, 0.17935660481452942, 0.25237786769866943, 0.13546818494796753, -0.05440989509224892, -0.30060815811157227, 0.29202115535736084, 0.18487879633903503, -0.35682132840156555, -0.3436088562011719, 0.07527995854616165, -0.3887392282485962, 0.35672807693481445, 0.12050751596689224, 0.16586101055145264, 0.010490525513887405, -0.14146848022937775, 0.11952164024114609, 0.3928675651550293, -0.09637761116027832, -0.1670006811618805, -0.11347857862710953, -0.025100111961364746, -0.25130605697631836, 0.07018402218818665, 0.09315343201160431, 0.1699230968952179, -0.10998507589101791, -0.36253732442855835, -0.12453190237283707, -0.1382128745317459, 0.09857957065105438, -0.032947439700365067, -0.5224338173866272, -0.21595120429992676, -0.1834096610546112, -0.03497084230184555, -0.08714625984430313, -0.284088671207428, 0.0498402938246727, -0.0025754645466804504, 0.21522320806980133, -0.2706935405731201, 0.04158273711800575, -0.04056726396083832, 0.02727458067238331, -0.3746243119239807, -0.04783228039741516, 0.29492121934890747, 0.05317297205328941, 0.1146203801035881, 0.16285616159439087, 0.19290168583393097, 0.12154622375965118, 0.4489949941635132, -0.028123173862695694, -0.010045245289802551, -0.023936744779348373, 0.14568474888801575, -0.21127967536449432, -0.019119448959827423, 0.33989524841308594, -0.3035700023174286, -0.06994908303022385, -0.1554131805896759, -0.19051958620548248, 0.09611622989177704, -0.05094454810023308, -0.2012614905834198, -0.3853227198123932, -0.4683864414691925, 0.6835903525352478, 0.0639658123254776, -0.0959390252828598, -0.20830407738685608, 0.20528583228588104, -0.21099448204040527, -0.0864718109369278, 0.1614258587360382, 0.06823398917913437, -0.01320141926407814, -0.37244030833244324, 0.041984524577856064, -0.10054001212120056, 0.008564971387386322, 0.30295640230178833, -0.15408705174922943, 0.22199967503547668, 0.08476705104112625, 0.11213968694210052, 0.3399461805820465, -0.479181170463562, -0.009698720648884773, 0.3526689410209656, -0.04269430786371231, 0.29657697677612305, 0.1470525711774826, 0.27371567487716675, -0.07260718941688538, -0.15075428783893585, 0.12967126071453094, 0.9934558272361755, -0.3805125653743744, -0.017214447259902954, -0.0795985758304596, -0.3645017445087433, 0.48016631603240967, 0.35858508944511414, 0.19760920107364655, -0.36574697494506836, -0.18771952390670776, -0.005163594149053097, 0.0852503553032875, 0.006785046309232712, 0.08330795913934708, 0.024008266627788544, -0.23442037403583527, 0.10306419432163239, -0.19651690125465393, -0.23357394337654114, -0.4084452688694, 0.041435323655605316, -0.05092465132474899, 0.37805959582328796, 0.35174524784088135, -0.05343758687376976, 0.2586511969566345, -0.2782157063484192, -0.12434379756450653, -0.3101751208305359, 0.034565918147563934, -0.023269139230251312, 0.0006010346114635468, -0.26450568437576294, -0.09321190416812897, 0.260556161403656, 0.03617440164089203, -0.0949815958738327, -0.4310052692890167, 0.4275183379650116, 0.02262904867529869, -0.03463855758309364, -0.03160158544778824, 0.5317655205726624, -0.24605217576026917, -0.11966552585363388, 0.10470794141292572, 0.21626326441764832, -0.20524659752845764, 0.15738517045974731, 0.22795149683952332, 0.3987584710121155, 0.23988604545593262, -0.08323948830366135, 0.05731647461652756, -0.05171791836619377, -0.06435868889093399, -0.21896958351135254, -0.004287399351596832, 0.40685945749282837, -0.2748064398765564, 0.40373435616493225, -0.007485587149858475, -0.02760295756161213, -0.2262641042470932, 0.010464470833539963, -0.12387334555387497, 0.4250919222831726, 0.34055137634277344, -0.24968089163303375, 0.3180716037750244, -0.13490408658981323, -0.347717821598053, 0.23370882868766785, 0.12683451175689697, 0.019252199679613113, 0.00021502375602722168, -0.04031014442443848, 0.015348073095083237, -0.2949424386024475, -0.13681767880916595, 0.3106762766838074, 0.6342358589172363, 0.19913358986377716, 0.13627763092517853, -0.06178681552410126, -0.18401041626930237, -0.061751168221235275, 0.07712703943252563, 0.1738758236169815, -0.07647082209587097, 0.21138346195220947, -0.013369981199502945, 0.10678858309984207, 0.041692160069942474, -0.16831189393997192, 0.25196754932403564, -0.17216463387012482, -0.16441883146762848, 0.048586875200271606, -0.17014376819133759, -0.6914837956428528, -0.11620110273361206, 0.08361642807722092, -0.29779037833213806, -0.25566723942756653, 0.15189403295516968, 0.130579873919487, -0.09642691910266876, 0.17519433796405792, 0.16043365001678467, 0.49070101976394653, -0.3046950697898865, -0.3377012014389038, -0.0032319799065589905, -0.2629941403865814, -0.07772117853164673, 0.04284592345356941, 0.4357629120349884, 0.4183099567890167, 0.49025580286979675, 0.19848904013633728, -0.07153211534023285, -0.04202023148536682, -0.38981831073760986, 0.15624688565731049, -0.3618420362472534, 0.36245065927505493, 0.06319830566644669, -0.12011861801147461, 0.07015429437160492, -0.3186580538749695, -0.014650960452854633, -0.0883055105805397, -0.14835074543952942, -0.0626271665096283, -0.09027517586946487, 0.07648736238479614, -0.1306277960538864, -0.16166287660598755, -0.6078256964683533, -0.39203643798828125, 0.1981445550918579, -0.18852905929088593, 0.07596120238304138, 0.15446597337722778, -0.2840275764465332, 0.24272455275058746, -0.1273290067911148, 0.08162131905555725, -0.132224902510643, -0.14825524389743805, -0.016928914934396744, -0.40536534786224365, -0.22016756236553192, -0.4019721746444702, -0.01956525817513466, 0.16767029464244843, 0.052456971257925034, -0.0996360257267952, -0.39100709557533264, 0.0016656983643770218, 0.0276724211871624, 0.058360401540994644, 0.2569243311882019, 0.11975869536399841, -0.28187525272369385, 0.1681641787290573, -0.14121797680854797, -0.18737377226352692, 0.2724555730819702, 0.3120083510875702, 0.36048623919487, 0.17845940589904785, -0.0021588318049907684, 0.12795649468898773, 0.9331138134002686, 0.2947481870651245, 0.004239201545715332, 0.18400788307189941, 0.13027022778987885, 0.01915886253118515, 0.0827958956360817, -0.19149282574653625, 0.20076070725917816, -0.11971068382263184, 0.4675973653793335, 0.10902044922113419, 0.08792636543512344, -0.016809163615107536, -0.2927389144897461, 0.09595124423503876, -0.3363724946975708, -0.2571498155593872, 0.20428669452667236, -0.34242308139801025, 0.10670784115791321, -0.1269156038761139, -0.40862467885017395, -0.37843674421310425, -0.07600965350866318, -0.09919141232967377, 0.2852097153663635, 0.10526390373706818, 0.1705242097377777, -0.38959264755249023, -0.050177156925201416, -0.4277200698852539, 0.14170679450035095, 0.12198345363140106, 0.2932141423225403, -0.08066634088754654, -0.0009058080613613129, -0.061951398849487305, 0.1394701898097992, 0.5762676000595093, -0.48940569162368774, -0.09264400601387024, 0.1321839541196823, -0.09517456591129303, -0.1648751050233841, -0.02935931831598282, -0.3096029460430145, 0.05311177670955658, 0.3403048813343048, 0.34198248386383057, -0.4822687804698944, -0.4431769847869873, 0.2855309844017029, -0.07070653140544891, -0.06597499549388885, -0.16693171858787537, -0.16285689175128937, -0.23329249024391174, -0.048672597855329514, -0.08596397191286087, 0.3517230451107025, 0.24646826088428497, -0.14004094898700714, 0.2022763192653656, -0.0320151150226593, 0.2508801221847534, 0.20811602473258972, -0.0859517753124237, -0.07969799637794495, 0.19456399977207184, 0.2547481060028076, 0.03491140902042389, -0.10541506111621857, -0.30441224575042725, 0.14299854636192322, -0.14487148821353912, -0.11757621169090271, 0.2817603051662445, 0.06203855946660042, 0.6452473402023315, 0.0720042958855629, 0.34857362508773804, 0.34181270003318787, -0.14397569000720978, 0.0882176011800766, -0.10354958474636078, 0.20663754642009735, 0.22724993526935577, 0.25939416885375977, -0.11130636185407639, -0.2703574001789093, 0.545784056186676, 0.3118358850479126, 0.018796496093273163, 0.4668085277080536, -0.566470205783844, -0.26695170998573303, 0.2087348848581314, 0.043551236391067505, 0.7242981791496277, -0.06499601900577545, 0.4172830879688263, 0.14827373623847961, -0.3961779475212097, 0.5780623555183411, -0.215780571103096, 0.04198579862713814, -0.14468619227409363, -0.3205392062664032, 0.024116365239024162, -0.13038353621959686, -0.20336803793907166, 0.22062033414840698, -0.2274605631828308, 0.2735256254673004, 0.370996356010437, -0.21724531054496765, -0.03466954082250595, 0.48724400997161865, -0.08320412784814835, -0.3102732300758362, -0.007831737399101257, -0.005010316148400307, -0.27206099033355713, 0.1534225195646286, -0.15633094310760498, -0.27250993251800537, 0.18027876317501068, 0.22555722296237946, -0.36308759450912476, -0.03686496615409851, -0.04107169061899185, 0.17539943754673004, 0.27989450097084045, -0.06461715698242188, 0.22811898589134216, -0.058995965868234634, -0.044314779341220856, -0.18439990282058716, -0.13139384984970093, 0.11050482094287872, -0.25417792797088623, 0.03881377726793289, 0.13085919618606567, -0.08861708641052246, 0.47457197308540344, -0.10375446081161499, -0.14080700278282166, -0.21365085244178772, 0.2007814347743988, -0.453219473361969, -0.2897721529006958, -0.02821923792362213, 0.046036411076784134, 0.1446157693862915, -0.312000572681427, 0.5833741426467896, 0.02064930647611618, 0.0554836131632328, -0.03234788402915001, 0.20651651918888092, -0.1736728399991989, 0.4253310561180115, -0.3541755974292755, -0.043095387518405914, 0.05548543483018875, 0.0381491482257843, -0.04922188073396683, -0.3953079879283905, 0.07578082382678986, 0.4156874120235443, 0.16333788633346558, 0.031431667506694794, 0.1543506383895874, 0.10619043558835983, -0.18556252121925354, 0.11073176562786102, -0.07481802254915237, -0.19344593584537506, 0.12716929614543915, 0.3633471131324768, 0.16826599836349487, 0.12084878981113434, -0.0636971965432167, -0.2312946617603302, -0.18166883289813995, 0.5214850306510925, -0.014888845384120941, 0.22884076833724976, -0.15737080574035645, 0.4722117781639099, 0.42730841040611267, 0.47415345907211304, -0.19213160872459412, 0.11956621706485748, 0.11680854111909866, 0.08993293344974518, 0.5097169876098633, -0.24610280990600586, 0.04991716891527176, 0.26596879959106445, -0.21029984951019287, -0.0020564980804920197, -0.3129919469356537, -0.08803519606590271, -0.26381611824035645, 0.1546061933040619, 0.1326218843460083, 0.017633747309446335, 0.08132201433181763, -0.10233192890882492, -0.20836412906646729, -0.10318643599748611, 0.13714444637298584, -0.2286568284034729, -0.07985039055347443, 0.2186506688594818, 0.004845131188631058, -0.015345089137554169, -0.068305104970932, 0.32523876428604126, 0.06706158816814423, -0.003959596157073975, 0.3423418700695038, 0.15840312838554382, 0.16730660200119019, -0.1784554421901703, -0.16622069478034973, -0.02740786224603653, 0.32666876912117004, 0.09608419239521027, -0.06478893756866455, 0.058879733085632324, -0.1197936087846756, -0.18889734148979187, 0.18732768297195435, 0.252206027507782, 0.030034566298127174, -0.027139021083712578, 0.21464210748672485, 0.032238349318504333, -0.03687313199043274, 0.0016330728540197015, 0.13352444767951965, 0.005256228148937225, 0.08753183484077454, 0.12357059866189957, 0.009909719228744507, 0.4867095947265625, -0.1714266985654831, -0.04055001586675644, 0.0002595053520053625, -0.25957179069519043, 0.16239137947559357, 0.033877670764923096, -0.054903820157051086, -0.018305223435163498, 0.0721062496304512, -0.03674392029643059, 0.12226544320583344, -0.33030378818511963, 0.3012450635433197, 0.27697938680648804, -0.08719229698181152, -0.015070218592882156, 0.23038509488105774, -0.08356519043445587, -0.1979319304227829, 0.31108635663986206, 0.0794963538646698, 0.4178653061389923, -0.009903520345687866, 0.9180453419685364, -0.3006817102432251, -0.46471014618873596, 0.0114987101405859, 0.521208643913269, 0.1481102854013443, -0.25235995650291443, 0.06891939043998718, 0.08991686254739761, -0.2977737486362457, 0.09773918986320496, -0.00985107570886612, -0.18153205513954163, 0.3974483013153076, -0.010155841708183289, 0.19723674654960632, -0.13021765649318695, -0.1488286554813385, 0.34921994805336, 0.17988654971122742, -0.006385877728462219, 0.08323796093463898, 0.03972797095775604, -0.10503023862838745, -0.17178542912006378, 0.512056291103363, 0.08268113434314728, -0.07022848725318909, -0.5144467353820801, 0.02289751172065735, 0.15500828623771667, 0.1871427595615387, 0.12166750431060791, 0.2046428918838501, 0.4158378839492798, -0.19626441597938538, 0.36525964736938477, -0.08221453428268433, -0.03379257768392563, -0.1805470585823059, 0.060573041439056396, -0.42835405468940735, -0.46482202410697937, 0.46828126907348633, 0.05242529138922691, -0.1495819091796875, 0.16356703639030457, 0.24570178985595703, -0.6194168329238892, -0.27177971601486206, 0.3253089487552643, -0.17062537372112274, -0.09467184543609619, -0.14008155465126038, 0.035228483378887177, 0.015107783488929272, 0.7085172533988953, 0.24064163863658905, 0.44719892740249634, -0.315765917301178, -0.06770211458206177, -0.5137102603912354, -0.13571268320083618, 0.18594777584075928, -0.07559124380350113, -0.18541374802589417, 0.055791862308979034, 0.04734198376536369, 0.31720685958862305, -0.08298158645629883, 0.16424773633480072, 0.18251392245292664, 0.19075411558151245, -0.5117713212966919, -0.04429372027516365, -0.02375224232673645, 0.23374833166599274, 0.24095776677131653, -0.1344553381204605, -0.21129365265369415, 0.11336624622344971, -0.11925771832466125, -0.1077665239572525, -0.12228408455848694, 0.3494461178779602, 0.057619608938694, 0.10408392548561096, 0.07577160745859146, 0.5753726959228516, 0.06265414506196976, -0.19473980367183685, 0.04764499515295029, -0.05116250365972519, -0.2982564866542816, -0.09186168760061264, -0.07321731746196747, 0.11370620131492615, -0.3598741590976715, 0.3518761694431305, 0.013054776936769485, -0.02684737928211689, 0.11974893510341644, -0.05373900383710861, -0.07982286810874939, -0.13715922832489014, 0.06085861474275589, 0.19985537230968475, 0.25871768593788147, 0.7949317097663879, -0.15307685732841492, 0.13903599977493286, -0.136479914188385, 0.054477520287036896, 0.10467610508203506, -0.12959527969360352, -0.21673181653022766, -0.11664855480194092, 0.056867823004722595, 0.04489794000983238, 0.1433515101671219, -0.5357406735420227, -0.2840758264064789, 0.12216448783874512, 0.005257561802864075, 0.01748190075159073, 0.7596602439880371, 0.2824134826660156, -0.17513993382453918, -0.05176655203104019, -0.18941739201545715, 0.008700868114829063, -0.22647400200366974, 0.09785190224647522, -0.2878146767616272 ]
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, -0.44010576605796814, -0.09260924160480499, 0.38070476055145264, -0.059492915868759155, -0.13712067902088165, -0.35067734122276306, 0.09061785042285919, 0.023502390831708908, 0.23529517650604248, -0.25593215227127075, 0.16841155290603638, -0.021551262587308884, 0.0022065434604883194, 0.09740171581506729, -0.3178660273551941, 0.03742338716983795, -0.034754619002342224, 0.16254767775535583, -0.20558613538742065, -0.15831172466278076, 0.3564315736293793, -0.06628049910068512, 0.2676071226596832, 0.05491020530462265, 0.2283007800579071, 0.1845669150352478, 0.20209896564483643, -0.41247719526290894, -0.17358002066612244, 0.14888527989387512, 0.005113102495670319, 0.23019932210445404, 0.2997414171695709, -0.00011345054372213781, -0.22685754299163818, 0.1634780466556549, -0.09964220225811005, 0.10824410617351532, -0.5389814972877502, 0.06842269003391266, -0.28030285239219666, -0.11710111051797867, -0.050361551344394684, -0.07701852172613144, -0.2652798295021057, -0.30860161781311035, -0.15958648920059204, 0.30006319284439087, 0.1558416783809662, 0.13660411536693573, -0.20554155111312866, -0.062215328216552734, -0.10612590610980988, 0.2618381977081299, -0.10174711048603058, 0.11459247022867203, 0.5414437055587769, 0.2648921608924866, 0.003537917509675026, 0.1269058734178543, 0.1456470936536789, 0.257468044757843, 0.06553101539611816, 0.3002891540527344, 0.21130600571632385, 0.31865453720092773, -0.11210070550441742, -0.19202959537506104, 0.2805585265159607, 0.12960241734981537, -0.13421276211738586, -0.3024217486381531, 0.10026524215936661, 0.07765419036149979, -0.5770214796066284, -0.2875575125217438, -0.1081860288977623, 0.15919633209705353, -0.21189618110656738, -0.19137609004974365, -0.043242041021585464, -0.15449902415275574, 0.10635067522525787, 0.07333448529243469, -0.011296898126602173, -0.05365904048085213, 0.1988222599029541, 0.18406569957733154, -0.003397800028324127, -0.5394580364227295, 0.21521182358264923, -0.31420913338661194, 0.2942136824131012, -0.3065517246723175, -0.09514324367046356, 0.1869504153728485, 0.40665122866630554, 0.10325301438570023, 0.3489795923233032, 0.11953452974557877, -0.2580716609954834, -0.2762497663497925, 0.08729451149702072, -0.0699591338634491, 0.4045150578022003, 0.1960495412349701, 0.33812788128852844, -0.028442755341529846, 0.06927013397216797, 0.19545549154281616, -0.0628727450966835, 0.24693496525287628, -0.477272629737854, -0.06356947124004364, 0.47708311676979065, -0.007767818868160248, -0.5834073424339294, -0.4587995707988739, -0.028733234852552414, -0.33794182538986206, -0.20797036588191986, 0.15752938389778137, 0.10549388825893402, -0.20728549361228943, 0.32331016659736633, -0.03419395536184311, 0.2794032096862793, -0.3262602984905243, -0.11367299407720566, -0.15017735958099365, 0.2062860131263733, -0.20226049423217773, 0.22359485924243927, 0.15393972396850586, 0.1776576191186905, 0.2562238574028015, 0.21195411682128906, 0.2691628336906433, 0.05984320491552353, 0.206268310546875, -0.23722785711288452, -0.1175270602107048, -0.14814883470535278, -0.034479040652513504, 0.07741092145442963, 0.18840716779232025, -0.27248072624206543, -0.12267474830150604, -0.0750536099076271, -0.3015037775039673, -0.008742023259401321, 0.15920142829418182, 0.13326062262058258, 0.1258605420589447, 0.016422824934124947, -0.21359172463417053, 0.3548072576522827, -0.15039575099945068, -0.2694520652294159, 0.20135003328323364, -0.07212784141302109, -0.2583310306072235, -0.12636497616767883, 0.229312926530838, -0.3264336585998535, -0.1708659529685974, 0.044310376048088074, 0.2678411304950714, -0.08543580770492554, 0.0589672327041626, 0.3666200339794159, 0.001542154117487371, 0.19801342487335205, -0.15210579335689545, -0.5923177599906921, 0.6651851534843445, -0.7773891091346741, -0.06579917669296265, -0.0005512686911970377, -0.11361725628376007, -0.15772925317287445, 0.13908840715885162, 0.029332460835576057, -0.29635360836982727, -0.20137320458889008, 0.34006622433662415, -0.031104527413845062, -0.15010927617549896, -0.3194095492362976, -0.09263476729393005, 0.3367890417575836, 0.4528050422668457, -0.1628921777009964, -0.04249311611056328, -0.11781185120344162, -0.08548697084188461, -0.04174085706472397, 0.40881645679473877, 0.06966608017683029, -0.1271551549434662, 0.021893665194511414, -0.18314959108829498, 0.16435667872428894, 0.22998657822608948, -0.29093894362449646, 0.25405171513557434, 0.10766575485467911, -0.9638234376907349, -0.22757470607757568, 0.4533107876777649, -0.2547580599784851, -0.6550406217575073, -0.20376752316951752, 0.30906230211257935, -0.267272412776947, 0.09786076843738556, -0.16166384518146515, 0.36613017320632935, 0.06151813641190529, -0.12694668769836426, 0.0966961458325386, 0.06017623841762543, -0.0041922759264707565, -0.3270718455314636, -0.13261853158473969, 0.10517538338899612, 0.07881283015012741, 0.2529539167881012, 0.45298293232917786, 0.45500147342681885, 0.2661784291267395, 0.14760726690292358, 0.23407462239265442, 0.22817136347293854, 0.11697328090667725, 0.31659629940986633, 0.2823440134525299, 0.16588659584522247, 0.06816911697387695, -0.06692637503147125, 0.45915114879608154, 0.3160407245159149, -0.1627158522605896, -0.1957533061504364, 0.4131929874420166, -0.48462292551994324, 0.12427695840597153, 0.02790401130914688, 0.012188298627734184, -0.20597581565380096, 0.12720182538032532, -0.40661925077438354, 0.004663527011871338, 0.23972061276435852, -0.10887444019317627, 0.13831034302711487, -0.26548025012016296, 0.07673659920692444, -0.06142597645521164, 0.19538286328315735, 0.021580249071121216, -0.036684490740299225, -0.13535982370376587, -0.0005483701825141907, -0.04335637390613556, -0.09061428904533386, -0.20829978585243225, 0.18548327684402466, 0.17604230344295502, 0.2232331484556198, 0.12233252823352814, -0.36821696162223816, -0.1841251105070114, -0.05356618016958237, -0.05732809007167816, 0.01605123095214367, 0.023728959262371063, 0.11102567613124847, -0.04540029168128967, 0.13835087418556213, -0.2840757369995117, -0.2592443823814392, 0.10575349628925323, -0.3508680760860443, 0.10057410597801208, -0.09940055012702942, -0.11004475504159927, -0.17763330042362213, -0.5282851457595825, -0.13439330458641052, -0.34777477383613586, -0.02615341544151306, 0.07551563531160355, 0.29761871695518494, -0.2034304291009903, 0.21936559677124023, 0.1803237348794937, -0.10814427584409714, 0.32815560698509216, 0.10603396594524384, -0.21061009168624878, -0.2679211497306824, 0.17807482182979584, -0.22155316174030304, 0.00794026255607605, 0.24635574221611023, -0.12963180243968964, -0.0070066433399915695, 0.15435971319675446, -0.25496116280555725, -0.09858129918575287, -0.05393761396408081, 0.22666512429714203, 0.06242278963327408, -0.28835684061050415, -0.20405727624893188, 0.3045308291912079, 0.23586663603782654, -0.02450849488377571, -0.26168835163116455, 0.03409318998456001, -0.19851455092430115, 0.20603370666503906, 0.05213009938597679, -0.1808527708053589, 0.06330019235610962, -0.08067408204078674, 0.48391538858413696, 0.2462226003408432, -0.07296387851238251, 0.2791990339756012, -0.31181883811950684, 0.25193145871162415, 0.31410759687423706, 0.37752383947372437, -0.3375564217567444, -0.2846829891204834, 0.3150167763233185, -0.41623079776763916, -0.3131350874900818, -0.09678226709365845, 0.06253829598426819, -0.01308748871088028, -0.08516383916139603, 0.06592926383018494, -0.34703490138053894, 0.038814082741737366, -0.04131891578435898, -0.12303416430950165, 0.18541406095027924, 0.24729016423225403, -0.26667720079421997, -0.05914100259542465, -0.10685434937477112, 0.08153663575649261, -0.013448402285575867, 0.10353471338748932, 0.539858877658844, -0.2782626748085022, 0.117824025452137, 0.1310109794139862, 0.34448790550231934, 0.3714722990989685, 0.434383362531662, 0.054828941822052, 0.16002950072288513, 0.029156070202589035, 0.08967453241348267, -0.17208315432071686, 0.30430662631988525, 0.07725410908460617, -0.1583978682756424, 0.07002720236778259, 0.15498226881027222, 0.4240894615650177, -0.286934494972229, -0.032308634370565414, -0.10816021263599396, 0.10679011046886444, -0.08260096609592438, -0.2662002444267273, -0.039601199328899384, -0.08259433507919312, 0.04221716523170471, -0.1264973282814026, -0.10205937922000885, 0.3080264925956726, 0.036289945244789124, 0.20763379335403442, 0.1973811239004135, -0.3730541467666626, -0.08608078956604004, -0.3050489127635956, 0.47342896461486816, 0.09159312397241592, -0.003033999353647232, 0.12427844107151031, 0.09895651787519455, -0.017055027186870575, -0.1943625658750534, 0.4446818232536316, 0.036086879670619965, 0.1932583898305893, 0.2945857346057892, -0.19206935167312622, -0.6826237440109253, -0.17620927095413208, -0.311749130487442, 0.27777206897735596, 0.16995884478092194, 0.08621401339769363, -0.17443932592868805, 0.05788734555244446, 0.19328731298446655, 0.0027839879039674997, -0.1897599697113037, 0.027196433395147324, -0.3189568519592285, 0.015282370150089264, 0.32539811730384827, -0.21736928820610046, -0.1152779683470726, 0.5167665481567383, -0.08533593267202377, 0.3249630033969879, 0.09882186353206635, -0.25806885957717896, 0.4550524950027466, -0.0021251123398542404, 0.24309459328651428, 0.1178697943687439, -0.18184949457645416, 0.15938431024551392, -0.13895103335380554, -0.5745752453804016, 0.33481115102767944, 0.09035544097423553, -0.5968520045280457, -0.10880055278539658, 0.008119766600430012, 0.482616126537323, 0.030374452471733093, -0.05051330104470253, -0.08396968990564346, -0.44186127185821533, -0.11237742006778717, -0.5094912648200989, -0.13879278302192688, 0.27017706632614136, -0.006576824467629194, -0.11409132182598114, -0.3621308505535126, 0.5230097770690918, 0.3907356858253479, -0.3316461443901062, 0.35048040747642517, -0.2329196184873581, -0.20785191655158997, 0.2235560566186905, 0.07861392945051193, 0.775036633014679, 0.2355675995349884, -0.03162864223122597, 0.12106257677078247, -0.25697922706604004, 0.5335091352462769, -0.6500673294067383, 0.07334858179092407, -0.21515586972236633, -0.2684263288974762, -0.0399848148226738, 0.06009354814887047, 0.048211779445409775, 0.18227218091487885, 0.08014319837093353, 0.5681716799736023, -0.060079678893089294, -0.1525992602109909, -0.030920585617423058, 0.253612220287323, 0.09966520965099335, -0.12943436205387115, -0.1757190078496933, 0.16745387017726898, -0.13291697204113007, 0.00017710775136947632, -0.2708410322666168, 0.11507498472929001, -0.5599819421768188, 0.06690462678670883, -0.14745119214057922, -0.13702517747879028, -0.18371473252773285, 0.17790457606315613, 0.13588501513004303, -0.422540545463562, -0.3310343325138092, 0.5675361156463623, 0.7250475883483887, 0.06679413467645645, -0.2974841892719269, 0.2634715139865875, -0.17508473992347717, 0.11545809358358383, 0.03641796484589577, -0.09256507456302643, 0.05644341930747032, -0.028397109359502792, -0.02011878788471222, 0.3248104751110077, 0.34698501229286194, -0.002152740955352783, -0.17017726600170135, 0.009425781667232513, -0.14369578659534454, 0.0441509485244751, -0.059258632361888885, 0.24112549424171448, 0.3410099744796753, 0.07543078064918518, 0.0871412456035614, 0.22685506939888, -0.5656342506408691, -0.052645422518253326, 0.2004537433385849, -0.19524170458316803, 0.1426815539598465, 0.2905636429786682, 0.03757728263735771, -0.2935364842414856, -0.22153693437576294, 0.087562695145607, 0.07894840836524963, -0.005834855139255524, -0.04641561210155487, 0.016472788527607918, -0.06334849447011948, 0.018273118883371353, 0.28729677200317383, 0.15044651925563812, 0.03996008634567261, 0.5065686702728271, 0.31823664903640747, -0.1310807168483734, 0.25226715207099915, -0.4538985788822174, -0.2241072654724121, 0.19057047367095947, 0.08486626297235489, -0.0875307023525238, -0.08937273919582367, -0.08392098546028137, 0.13785170018672943, 0.5422527194023132, -0.2892587184906006, -0.2375672161579132, -0.04240000993013382, 0.14428775012493134, -0.16340003907680511, -0.1392480432987213, -0.10749558359384537, 0.14073379337787628, 0.06438962370157242, 0.04496776685118675, -0.18612557649612427, -0.17945623397827148, -0.2494579404592514, 0.13138766586780548, -0.1495247781276703, 0.1688951998949051, -0.06284715980291367, 0.0750800371170044, 0.184679314494133, -0.0817183405160904, -0.05887413024902344, 0.3747633099555969, 0.37351560592651367, 0.3379557132720947, -0.034392427653074265, 0.10218104720115662, -0.056325025856494904, -0.01782960817217827, 0.21329344809055328, 0.2078014612197876, 0.07900924235582352, -0.008537612855434418, 0.14385941624641418, -0.016868434846401215, -0.2875359356403351, -0.08986053615808487, -0.14779503643512726, -0.22689494490623474, 0.14476278424263, -0.01856979727745056, 0.20657122135162354, -0.2133338451385498, 0.37987709045410156, 0.13268625736236572, -0.07608423382043839, 0.18973122537136078, -0.18290534615516663, 0.142856165766716, -0.1198011115193367, 0.4061797559261322, 0.1726001352071762, 0.25368455052375793, -0.21503499150276184, 0.258207231760025, -0.03328472375869751, 0.15419086813926697, -0.011892654001712799, -0.008982258848845959, -0.027472414076328278, -0.3529618978500366, 0.21023724973201752, -0.45130017399787903, 0.13748317956924438, -0.07608513534069061, 0.34516963362693787, -0.14173296093940735, 0.0866357758641243, 0.09466511011123657, 0.11317452043294907, 0.20618659257888794, 0.057095009833574295, 0.38195013999938965, 0.05865370109677315, 0.16471195220947266, -0.335029274225235, 0.25033360719680786, 0.03729299455881119, -0.056034669280052185, 0.06057799607515335, 0.13486745953559875, 0.040382590144872665, -0.1946026086807251, -0.1217292845249176, 0.10459183156490326, 0.12302770465612411, -0.3937950134277344, -0.3598281443119049, -0.2577280104160309, -0.1975317746400833, -0.056885093450546265, -0.004423275589942932, 0.09338743984699249, -0.14682918787002563, 0.0563250407576561, -0.05959995090961456, -0.29237163066864014, -0.39882606267929077, 0.15627151727676392, 0.3441053032875061, -0.07433145493268967, 0.10791286826133728, 0.20615988969802856, 0.31274670362472534, 0.31578972935676575, 0.3679596781730652, 0.13274119794368744, 0.08966721594333649, -0.5088682770729065, 0.05510064586997032, -0.1820521205663681, -0.07493680715560913, 0.036594685167074203, 0.2758980691432953, 0.19292783737182617, -0.17557163536548615, 0.3166452646255493, 0.1541426181793213, -0.10532339662313461, 0.3391841650009155, -0.14225688576698303, 0.05937499925494194, -0.4119409918785095, 0.5525149703025818, 0.03649605065584183, 0.20788806676864624, -0.20443935692310333, 0.05917360261082649, -0.4736991822719574, -0.18732239305973053, -0.5115628242492676, 0.06559547781944275, 0.061522603034973145, -0.3025529980659485, 0.0779670923948288, 0.0898432731628418, 0.02408474311232567, 0.08686594665050507, 0.42110979557037354, -0.337667852640152, -0.19173681735992432, -0.3841714859008789, -0.36229798197746277, -0.0484888069331646, 0.11754702031612396, -0.24215459823608398, 0.287210077047348, -0.06774679571390152, 0.5412331223487854, 0.15837721526622772, 0.48342883586883545, 0.053993307054042816, 0.17747852206230164, -0.2007937878370285, -0.012257155030965805, -0.09126216173171997, -0.1045587807893753, -0.10259967297315598, 0.2286399006843567, -0.21448929607868195, -0.1760862171649933, 0.03683061897754669, -0.15790441632270813, -0.26630884408950806, 0.24678465723991394, 0.691580057144165, 0.3145759701728821, 0.3558570146560669, 0.46494171023368835, -0.01913050189614296, 0.03927036374807358, -0.05851421132683754, -0.0018537133000791073, -0.05582123622298241, 0.15834486484527588, -0.38414108753204346, 0.3125041127204895, -0.3025667369365692, -0.22766633331775665, 0.10168598592281342, 0.6001899242401123, 0.0094134621322155, -0.20100079476833344, -0.3931237757205963, -0.07098311185836792, 0.01847902685403824, -0.14986251294612885, -0.4123910963535309, 0.3378525674343109, -0.015125848352909088, 0.44271761178970337, -0.009619802236557007, -0.13172677159309387, 0.4783710837364197, -0.36559662222862244, 0.06721772253513336, 0.06162618100643158, 0.38429126143455505, 0.2618565261363983, -0.07773260772228241, -0.19687703251838684, -0.03381256014108658, 0.3542482852935791, 0.19263824820518494, -0.0012842845171689987, 0.16451631486415863, -0.18422242999076843, -0.10394720733165741, -0.15975956618785858, -0.019801974296569824, 0.06336133182048798, 0.041653603315353394, 0.10203620791435242, -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 :)
[ 0.12398985028266907, -0.4979516863822937, -0.08898454159498215, 0.3463147282600403, -0.08415494859218597, -0.22060829401016235, -0.37294113636016846, 0.08035792410373688, 0.03604584187269211, 0.22104912996292114, -0.21847869455814362, 0.23340590298175812, -0.011109267361462116, -0.0014878809452056885, 0.07686223834753036, -0.28436294198036194, -0.011588193476200104, -0.03263472020626068, 0.08533450961112976, -0.24821986258029938, -0.10775785148143768, 0.4054567217826843, -0.1544710248708725, 0.2524571418762207, 0.0654294490814209, 0.31669625639915466, 0.17007912695407867, 0.2143232226371765, -0.432865709066391, -0.18212689459323883, 0.1652716100215912, 0.050362877547740936, 0.1496959924697876, 0.23267795145511627, -0.00011426618584664539, -0.2925373613834381, 0.1849438101053238, -0.1414392739534378, 0.1738622784614563, -0.6005355715751648, 0.058325670659542084, -0.26914966106414795, -0.08119745552539825, -0.06456467509269714, -0.11102306842803955, -0.19320020079612732, -0.2885153889656067, -0.12736788392066956, 0.37436389923095703, 0.23226219415664673, 0.12809447944164276, -0.20203982293605804, -0.06131395697593689, -0.05112773925065994, 0.2469228208065033, -0.08830800652503967, 0.08550673723220825, 0.5601398348808289, 0.2660447955131531, -0.037467509508132935, 0.08142390102148056, 0.15097472071647644, 0.23171231150627136, 0.009744605049490929, 0.3903578817844391, 0.1786264032125473, 0.2437993288040161, -0.09382613003253937, -0.2231394350528717, 0.3274887502193451, 0.18336084485054016, -0.15424370765686035, -0.23367217183113098, 0.15244010090827942, 0.02727609872817993, -0.5276973247528076, -0.2502237558364868, -0.09636014699935913, 0.2021206021308899, -0.15361133217811584, -0.24245628714561462, -0.030754176899790764, -0.21640445291996002, 0.08503448963165283, 0.0001808665692806244, -0.007078118622303009, -0.0669672042131424, 0.16879890859127045, 0.20319344103336334, 0.011866047978401184, -0.5337000489234924, 0.2507757544517517, -0.3072173595428467, 0.34942129254341125, -0.37095728516578674, -0.16411639750003815, 0.1327212005853653, 0.4145554304122925, 0.10024333745241165, 0.40116071701049805, 0.22469398379325867, -0.14048725366592407, -0.2672233581542969, 0.10991980135440826, -0.03308255225419998, 0.4469318687915802, 0.22990906238555908, 0.31486496329307556, -0.027802959084510803, 0.0251830592751503, 0.14823849499225616, -0.02264481410384178, 0.23133672773838043, -0.4569520354270935, -0.038008276373147964, 0.39786893129348755, -0.027401171624660492, -0.6806557774543762, -0.4113559126853943, 0.051422327756881714, -0.3216814398765564, -0.14020563662052155, 0.11257842928171158, 0.09805742651224136, -0.19216439127922058, 0.3302401006221771, -0.007844313979148865, 0.23737473785877228, -0.33151912689208984, -0.14027734100818634, -0.07846415042877197, 0.2287808507680893, -0.25133201479911804, 0.21520091593265533, 0.18882833421230316, 0.21518711745738983, 0.24487322568893433, 0.20560823380947113, 0.30322444438934326, 0.029208816587924957, 0.2829257845878601, -0.24809260666370392, -0.02712104469537735, -0.14943145215511322, -0.07167099416255951, 0.06048443913459778, 0.11274956166744232, -0.2844976484775543, -0.09193716943264008, -0.05520528554916382, -0.2921653687953949, 0.009701360017061234, 0.14610490202903748, 0.13343559205532074, 0.07843891531229019, 0.012796679511666298, -0.11202722042798996, 0.3773157596588135, -0.20596805214881897, -0.266198992729187, 0.21629923582077026, -0.14620406925678253, -0.28263214230537415, -0.1256779283285141, 0.21286293864250183, -0.30434638261795044, -0.1561819463968277, 0.08477066457271576, 0.252817302942276, -0.05326258763670921, 0.014622434973716736, 0.4060991406440735, 0.0686083734035492, 0.1522846221923828, -0.09097429364919662, -0.5238792896270752, 0.6502180695533752, -0.7507557272911072, -0.09616754949092865, 0.03446264564990997, -0.08055101335048676, -0.10891734063625336, 0.0822356790304184, 0.021923156455159187, -0.31465744972229004, -0.14471255242824554, 0.2815755009651184, 0.010718575678765774, -0.08700942993164062, -0.3107841908931732, -0.1504412591457367, 0.39708665013313293, 0.4027329087257385, -0.07194815576076508, -0.08274514973163605, -0.1770651638507843, -0.1273576319217682, -0.01692304015159607, 0.4076850116252899, 0.03755701333284378, -0.12502336502075195, 0.11389748752117157, -0.19675952196121216, 0.14640352129936218, 0.2724829614162445, -0.3126000463962555, 0.3493416905403137, 0.07731414586305618, -0.8894877433776855, -0.1724054217338562, 0.4622158408164978, -0.2256922870874405, -0.6651107668876648, -0.215151846408844, 0.31093209981918335, -0.3100336790084839, 0.10217571258544922, -0.22574597597122192, 0.3516812324523926, 0.1139783263206482, -0.1536424160003662, 0.20693835616111755, 0.06197982281446457, 0.0069307852536439896, -0.319097638130188, -0.16207371652126312, 0.10615421831607819, 0.058505259454250336, 0.25599825382232666, 0.4742416441440582, 0.4499499499797821, 0.2268802970647812, 0.1688266396522522, 0.2143983542919159, 0.2538841962814331, 0.03497163951396942, 0.2582990527153015, 0.2095763087272644, 0.19598032534122467, -0.018469657748937607, -0.07744132727384567, 0.5199008584022522, 0.30376124382019043, -0.1541360318660736, -0.19438409805297852, 0.4662749767303467, -0.490393728017807, 0.0574420765042305, 0.030122719705104828, 0.0035352911800146103, -0.2344088852405548, 0.048947256058454514, -0.42692822217941284, 0.008210279047489166, 0.33786410093307495, -0.07221493124961853, 0.1508098691701889, -0.21438349783420563, 0.07765595614910126, -0.05872390419244766, 0.16696760058403015, -0.023158831521868706, -0.07007049769163132, -0.09916801005601883, 0.04828318953514099, -0.05247817933559418, -0.08713283389806747, -0.14648213982582092, 0.1367141604423523, 0.14246432483196259, 0.17149512469768524, 0.2192104160785675, -0.411747545003891, -0.18401682376861572, -0.007912367582321167, -0.1532539576292038, 0.01218420173972845, 0.015409678220748901, 0.06150931119918823, -0.07383880019187927, 0.051839955151081085, -0.30490612983703613, -0.30222752690315247, 0.06790249794721603, -0.3260825276374817, 0.0888369232416153, -0.10181143134832382, -0.14128339290618896, -0.17434781789779663, -0.4596985876560211, -0.1507117599248886, -0.3067990243434906, -0.037475667893886566, 0.07321762293577194, 0.2997561991214752, -0.29844796657562256, 0.21187353134155273, 0.20868787169456482, -0.0963938757777214, 0.3506270945072174, 0.09900619089603424, -0.29849109053611755, -0.21889367699623108, 0.14851553738117218, -0.22972330451011658, 0.010996153578162193, 0.20694665610790253, -0.15003857016563416, -0.0032068798318505287, 0.14732468128204346, -0.2885560691356659, -0.09995420277118683, -0.02525819092988968, 0.16488490998744965, 0.05572685971856117, -0.3303431272506714, -0.2132261097431183, 0.35807982087135315, 0.1570579558610916, -0.0432528555393219, -0.250110387802124, 0.09181634336709976, -0.12921284139156342, 0.13602469861507416, -0.005322178825736046, -0.20498058199882507, 0.07565270364284515, -0.016803644597530365, 0.4890824556350708, 0.23215307295322418, -0.02597884088754654, 0.2851446270942688, -0.354024738073349, 0.21252381801605225, 0.2565453052520752, 0.39148786664009094, -0.3267376720905304, -0.2725765109062195, 0.39423611760139465, -0.4358576238155365, -0.37219110131263733, -0.1029462218284607, 0.10516533255577087, 0.02943887561559677, -0.04442455992102623, 0.05712844431400299, -0.3295972943305969, 0.05410837382078171, -0.09084348380565643, -0.17520207166671753, 0.13301944732666016, 0.224800243973732, -0.26698797941207886, -0.04010269045829773, -0.12790319323539734, 0.09605056047439575, -0.0023123174905776978, 0.13142825663089752, 0.5086548924446106, -0.24810606241226196, 0.07726961374282837, 0.1404632031917572, 0.2756560444831848, 0.40292391180992126, 0.4460241198539734, 0.049148768186569214, 0.16979750990867615, 0.02271243929862976, 0.08366765081882477, -0.13092896342277527, 0.28202977776527405, 0.042150162160396576, -0.12101099640130997, 0.10032165050506592, 0.18470141291618347, 0.33166617155075073, -0.34148117899894714, -0.08046337962150574, -0.09187647700309753, 0.09235559403896332, -0.06381132453680038, -0.24431443214416504, -0.034373775124549866, -0.05484950542449951, 0.05709163099527359, -0.09766252338886261, -0.13494375348091125, 0.3102976977825165, 0.022706955671310425, 0.21679842472076416, 0.1691882163286209, -0.37488988041877747, -0.028955137357115746, -0.3219936788082123, 0.4225342273712158, 0.06536135077476501, -0.03763532638549805, 0.1324227899312973, 0.06528028845787048, 0.00621332973241806, -0.2297491878271103, 0.3863006830215454, 0.06365104764699936, 0.21256199479103088, 0.26525598764419556, -0.22254759073257446, -0.6708530783653259, -0.18083666265010834, -0.3159851133823395, 0.28959229588508606, 0.1978680044412613, 0.02305518463253975, -0.15612193942070007, 0.08311497420072556, 0.18468429148197174, -0.028113244101405144, -0.18701854348182678, 0.04569631069898605, -0.2581061124801636, 0.009999211877584457, 0.36460068821907043, -0.22924280166625977, -0.1427791267633438, 0.4782615602016449, -0.06073053926229477, 0.3421839475631714, 0.14983788132667542, -0.21545132994651794, 0.4536169767379761, 0.047774624079465866, 0.27631422877311707, 0.026590503752231598, -0.20807871222496033, 0.16249088943004608, -0.09441934525966644, -0.6795375943183899, 0.33890706300735474, 0.1927538961172104, -0.5905706882476807, -0.16465136408805847, -0.037636782974004745, 0.47643154859542847, 0.05458669364452362, -0.011679282411932945, -0.09295854717493057, -0.48114213347435, -0.0621471107006073, -0.5296534299850464, -0.12390255182981491, 0.2654699385166168, -0.008956381119787693, -0.11165179312229156, -0.3016607165336609, 0.48601317405700684, 0.3428497612476349, -0.3132480978965759, 0.3125554025173187, -0.19474510848522186, -0.20292159914970398, 0.2572427988052368, 0.1027376800775528, 0.8113808631896973, 0.21375523507595062, 0.0021938756108283997, 0.07617828249931335, -0.35134565830230713, 0.4707513749599457, -0.7156763672828674, 0.04499911516904831, -0.26291918754577637, -0.2532733976840973, -0.041826069355010986, 0.046717915683984756, 0.04493224620819092, 0.21422642469406128, 0.06681868433952332, 0.5717447400093079, 0.06466993689537048, -0.1611304134130478, -0.0035069938749074936, 0.27374058961868286, 0.13994716107845306, -0.05350300297141075, -0.14959964156150818, 0.14553578197956085, -0.12895610928535461, -0.027124818414449692, -0.24744370579719543, 0.0909351333975792, -0.5433633327484131, 0.07716075330972672, -0.1685914546251297, -0.12107547372579575, -0.1406957358121872, 0.11651520431041718, 0.13030724227428436, -0.39353418350219727, -0.25777560472488403, 0.6006635427474976, 0.7157094478607178, 0.12420657277107239, -0.37131816148757935, 0.2642662227153778, -0.2064560055732727, 0.15178190171718597, 0.10992161184549332, -0.05415476858615875, 0.045877523720264435, -0.057032763957977295, -0.06457622349262238, 0.276641845703125, 0.35905739665031433, -0.010357465595006943, -0.2205694317817688, -0.0072341859340667725, -0.1653808355331421, 0.07030605524778366, -0.03975100442767143, 0.20711562037467957, 0.3018413782119751, 0.10604226589202881, 0.08863203227519989, 0.2199646234512329, -0.5469434261322021, -0.029552176594734192, 0.24314643442630768, -0.2311176210641861, 0.16570517420768738, 0.3313015103340149, 0.008900599554181099, -0.2933350205421448, -0.25438016653060913, 0.07475247979164124, 0.08233662694692612, -0.005110431462526321, -0.03546134755015373, -0.020485907793045044, -0.041176121681928635, 0.020232368260622025, 0.25103893876075745, 0.13301165401935577, 0.021783920004963875, 0.434613436460495, 0.30642181634902954, -0.11115405708551407, 0.23509378731250763, -0.4188060760498047, -0.19548580050468445, 0.13808919489383698, 0.0805044174194336, -0.11472414433956146, -0.031742505729198456, -0.15798786282539368, 0.17435529828071594, 0.5180923938751221, -0.2837792634963989, -0.18762150406837463, -0.02894975244998932, 0.15590384602546692, -0.13453315198421478, -0.20898255705833435, -0.20269297063350677, 0.0788305252790451, 0.04642462730407715, 0.06440538167953491, -0.18087099492549896, -0.16448324918746948, -0.23848401010036469, 0.14039763808250427, -0.1358509361743927, 0.13904498517513275, -0.09388555586338043, 0.0531131848692894, 0.1827079951763153, -0.0889543890953064, -0.07194386422634125, 0.34918099641799927, 0.3699827790260315, 0.4574727714061737, -0.10201866924762726, 0.08419951796531677, -0.11546285450458527, -0.03458692133426666, 0.21736623346805573, 0.22923855483531952, 0.08090734481811523, -0.02448383718729019, 0.1727311611175537, -0.03619274124503136, -0.24741241335868835, -0.05410424619913101, -0.20524130761623383, -0.27699869871139526, 0.20450440049171448, -0.0031601451337337494, 0.229315847158432, -0.25947126746177673, 0.4298003911972046, 0.16129784286022186, -0.09349481761455536, 0.14037218689918518, -0.18729963898658752, 0.13596060872077942, -0.11948462575674057, 0.3579462468624115, 0.2065812647342682, 0.2541394531726837, -0.2403791844844818, 0.223332017660141, -0.016074348241090775, 0.08376480638980865, 0.06476977467536926, -0.031833112239837646, -0.02862314134836197, -0.32644128799438477, 0.18215462565422058, -0.42984622716903687, 0.15969720482826233, -0.10262607038021088, 0.26413482427597046, -0.1583990901708603, 0.06268270313739777, 0.20385761559009552, 0.11040225625038147, 0.22919906675815582, 0.02511720359325409, 0.42615145444869995, 0.04126839339733124, 0.24798648059368134, -0.3098810911178589, 0.2689865231513977, 0.07964637875556946, -0.031282611191272736, 0.10673385113477707, 0.1408994197845459, 0.11347515881061554, -0.23334597051143646, -0.1448434293270111, 0.06972706317901611, 0.12309956550598145, -0.3971679210662842, -0.3614509701728821, -0.2509462237358093, -0.26933151483535767, -0.017956912517547607, -0.05178672447800636, 0.12138805538415909, -0.10855117440223694, 0.10018302500247955, -0.044841229915618896, -0.24225075542926788, -0.37579551339149475, 0.07377775758504868, 0.44615018367767334, -0.10802555084228516, 0.10562580078840256, 0.1896156370639801, 0.3804336190223694, 0.2985629439353943, 0.33002007007598877, 0.12830069661140442, 0.0885031670331955, -0.5871762633323669, 0.1041392832994461, -0.1879848837852478, -0.0834156721830368, 0.07672026753425598, 0.3212108314037323, 0.17876584827899933, -0.16076809167861938, 0.3263954818248749, 0.12612119317054749, -0.06357879936695099, 0.3487561047077179, -0.11649841070175171, 0.05914187431335449, -0.39986035227775574, 0.4350467920303345, 0.11445913463830948, 0.211585134267807, -0.25026389956474304, 0.01833103597164154, -0.5301827788352966, -0.18016104400157928, -0.5192384719848633, 0.1562500298023224, 0.057314030826091766, -0.33192959427833557, 0.07285404205322266, 0.12624770402908325, 0.07426103204488754, 0.05576527118682861, 0.3771965503692627, -0.3097309172153473, -0.2862943410873413, -0.41255640983581543, -0.32417938113212585, 0.023713327944278717, 0.07984786480665207, -0.20767951011657715, 0.22943317890167236, -0.04843645542860031, 0.5633388161659241, 0.1081700474023819, 0.4685989320278168, 0.05568799376487732, 0.1747456192970276, -0.24061770737171173, 0.055369362235069275, -0.0938052386045456, -0.13297021389007568, -0.14306612312793732, 0.24368232488632202, -0.17661771178245544, -0.18919794261455536, 0.03014836460351944, -0.12438850104808807, -0.22878040373325348, 0.21405865252017975, 0.7476032972335815, 0.2672616243362427, 0.3172944486141205, 0.4162449240684509, -0.002823207527399063, 0.03634314984083176, -0.007560711354017258, 0.06603831797838211, 0.007561356760561466, 0.2057744264602661, -0.30258432030677795, 0.2575332224369049, -0.21722067892551422, -0.2541283667087555, 0.15533418953418732, 0.6273364424705505, -0.04024696722626686, -0.20091114938259125, -0.35345590114593506, -0.1567518264055252, 0.05964159220457077, -0.16473793983459473, -0.518647313117981, 0.39035049080848694, -0.04144945740699768, 0.40933364629745483, -0.0025366097688674927, -0.08588693290948868, 0.4886842668056488, -0.33689677715301514, 0.11393262445926666, 0.11974432319402695, 0.3727896213531494, 0.2416967749595642, -0.11075747758150101, -0.23718181252479553, -0.051548443734645844, 0.2927353084087372, 0.16268792748451233, 0.0006722100079059601, 0.20673133432865143, -0.15648120641708374, -0.10198025405406952, -0.17159123718738556, 0.06903013586997986, 0.11228369921445847, 0.060577183961868286, 0.062373917549848557, -0.1644407957792282 ]
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
[ -0.13186068832874298, 0.14030981063842773, 0.05940939113497734, 0.4210398197174072, -0.10276152193546295, 0.06391056627035141, -0.09939047694206238, 0.24033156037330627, -0.16995178163051605, 0.16742539405822754, -0.002576854545623064, 0.2693632245063782, 0.10592739284038544, -0.02921498566865921, -0.06128937005996704, 0.11617531627416611, -0.07574360817670822, 0.23838576674461365, 0.06643003225326538, 0.009823672473430634, -0.029608845710754395, 0.4090341627597809, -0.290790855884552, -0.08376243710517883, -0.2635086476802826, -0.20447701215744019, -0.007695583626627922, 0.2664903700351715, -0.028038201853632927, -0.2925416827201843, 0.27207711338996887, 0.011068349704146385, 0.13318169116973877, 0.1919163018465042, -0.00012480426812544465, 0.02284415066242218, 0.5506137609481812, -0.25053170323371887, -0.21130108833312988, -0.2491968423128128, -0.4140856862068176, -0.28161290287971497, 0.039414264261722565, -0.3059539496898651, -0.07530711591243744, 0.13373978435993195, 0.16358286142349243, -0.052976563572883606, 0.24514071643352509, 0.3204346299171448, 0.11070700734853745, 0.07963929325342178, -0.2686450183391571, 0.1345806121826172, 0.12662167847156525, 0.08180851489305496, -0.03699054569005966, -0.07447116076946259, -0.18289318680763245, -0.17931108176708221, -0.0033599063754081726, 0.4617772102355957, -0.2776550054550171, 0.13988222181797028, 0.19181618094444275, 0.17948710918426514, 0.348262757062912, -0.13864782452583313, 0.15919563174247742, 0.3170469105243683, 0.42594918608665466, 0.15770426392555237, -0.20480979979038239, -0.42427897453308105, -0.0219615176320076, 0.10003815591335297, 0.4799433648586273, 0.32488197088241577, -0.23730987310409546, 0.03319871425628662, -0.433061808347702, -0.07635169476270676, -0.03681832551956177, 0.009392645210027695, 0.15958695113658905, 0.32183772325515747, 0.08606128394603729, 0.07757256180047989, 0.09996238350868225, 0.013011597096920013, 0.3436041474342346, -0.27978405356407166, -0.21615895628929138, 0.04208419844508171, -0.10203404724597931, -0.19778043031692505, -0.18273749947547913, 0.06244150549173355, 0.15654721856117249, 0.07547703385353088, 0.37092092633247375, -0.08625371009111404, 0.17984464764595032, 0.11776173114776611, 0.3715384006500244, -0.013677000999450684, -0.028967447578907013, 0.5226830840110779, 0.10686895251274109, -0.009327450767159462, 0.1176343560218811, 0.04502790421247482, 0.29349997639656067, -0.1499851495027542, -0.19445769488811493, 0.12557056546211243, 0.16165246069431305, -0.6070359349250793, -0.4450760781764984, 0.016515780240297318, -0.5715518593788147, -0.20781360566616058, -0.055248793214559555, 0.24616670608520508, -0.36231160163879395, 0.2653512954711914, -0.18779277801513672, -0.03759102523326874, -0.32581865787506104, -0.01764378882944584, -0.18110673129558563, 0.1677558720111847, -0.1367722898721695, 0.25721195340156555, 0.14054302871227264, -0.18570293486118317, 0.41987064480781555, -0.17853707075119019, -0.030709147453308105, -0.15490424633026123, 0.2606305181980133, -0.1932949423789978, 0.07086895406246185, 0.22536788880825043, 0.0033082999289035797, 0.16629427671432495, -0.09200568497180939, -0.2952803075313568, -0.35808518528938293, 0.15888839960098267, -0.2516799569129944, -0.5216352343559265, 0.10345858335494995, 0.1205809935927391, -0.47296392917633057, -0.036487333476543427, -0.036902524530887604, 0.2329140454530716, 0.4576258659362793, -0.33405205607414246, -0.02695993334054947, -0.1752258688211441, -0.2176084816455841, -0.04367905110120773, -0.2373843491077423, 0.43923449516296387, 0.14354448020458221, -0.23715516924858093, 0.13986878097057343, 0.06780452281236649, 0.6044769883155823, 0.3637809157371521, -0.0025907307863235474, 0.07247652113437653, -0.1579892486333847, 0.33445459604263306, 0.13933111727237701, -0.3792102336883545, -0.22019106149673462, 0.47593674063682556, 0.07704773545265198, 0.27838781476020813, 0.26170918345451355, 0.1201210543513298, 0.19283688068389893, -0.14579878747463226, 0.13192394375801086, 0.3967822790145874, -0.01601727120578289, 0.06469926238059998, -0.4757278859615326, -0.3340061604976654, 0.45108896493911743, 0.10481983423233032, 0.10609757155179977, 0.016658365726470947, -0.06624103337526321, 0.453386127948761, 0.2774915099143982, -0.11796833574771881, -0.03857593238353729, 0.178915336728096, -0.2975044250488281, -0.0030195489525794983, -0.008994619362056255, 0.045939862728118896, -0.43105632066726685, -0.06189686805009842, -0.16155469417572021, 0.2573896646499634, 0.021547138690948486, -0.3396948575973511, -0.12602588534355164, -0.3853745758533478, -0.012610990554094315, -0.03815138712525368, -0.024999532848596573, 0.32791680097579956, 0.38916337490081787, 0.020113889127969742, 0.04731035605072975, 0.5165187120437622, -0.07561039179563522, 0.19800005853176117, -0.7434919476509094, 0.24971456825733185, -0.0970013365149498, -0.22091659903526306, 0.48767125606536865, 0.29407188296318054, 0.1883361041545868, -0.20826861262321472, -0.12820076942443848, 0.4203099310398102, -0.09913244843482971, 0.41020524501800537, -0.3028305768966675, 0.009452924132347107, 0.28660088777542114, -0.21590326726436615, -0.0950908288359642, 0.3127250373363495, -0.014362134039402008, -0.02967958152294159, -0.08529884368181229, 0.4560617208480835, -0.35522958636283875, 0.19494345784187317, 0.06018040329217911, -0.07444212585687637, 0.11264720559120178, -0.21182698011398315, -0.2256215363740921, -0.46470898389816284, 0.5550782680511475, -0.011282149702310562, -0.33899542689323425, 0.09569765627384186, -0.22145242989063263, 0.09830480068922043, 0.471059113740921, 0.09217573702335358, -0.13324370980262756, -0.10313469916582108, 0.10757986456155777, -0.13031743466854095, 0.2442014217376709, 0.2132987231016159, 0.4979696273803711, 0.12661860883235931, -0.025470968335866928, 0.05887509137392044, -0.14599889516830444, -0.3809732496738434, 0.13225601613521576, -0.017310798168182373, 0.1433498114347458, 0.2880011200904846, 0.09366247802972794, -0.040374506264925, -0.2727144658565521, -0.04973514750599861, 0.1679784059524536, 0.24612028896808624, -0.2921508848667145, -0.04091402515769005, -0.3865091800689697, -0.47475746273994446, -0.2583891451358795, -0.1584344208240509, -0.21669253706932068, -0.33608385920524597, 0.16131684184074402, 0.24778471887111664, -0.1326110064983368, 0.06925041228532791, -0.16835512220859528, 0.06610099971294403, -0.31220901012420654, -0.07321520894765854, -0.06265108287334442, -0.28624194860458374, -0.24312901496887207, -0.04386451840400696, 0.19816158711910248, 0.37293606996536255, 0.20360085368156433, -0.4253942668437958, -0.22502323985099792, 0.14127753674983978, -0.2914535105228424, 0.14501729607582092, -0.026525668799877167, -0.12445494532585144, 0.08017541468143463, -0.2173847109079361, 0.2646990418434143, -0.24175013601779938, 0.0449194461107254, -0.10440611094236374, -0.33106258511543274, -0.10031034797430038, 0.12558411061763763, 0.004528342746198177, -0.21038497984409332, -0.4195679724216461, -0.25907838344573975, -0.34470659494400024, -0.10632041096687317, 0.28229814767837524, 0.3047817349433899, 0.11865761876106262, -0.15986120700836182, -0.11309616267681122, -0.05772746726870537, 0.21963723003864288, -0.3405010998249054, -0.09820809960365295, 0.23921023309230804, -0.08729145675897598, -0.11389191448688507, 0.011300381273031235, -0.05697101354598999, -0.03342808410525322, 0.15223176777362823, -0.5655202865600586, -0.20712752640247345, -0.1663607358932495, -0.10165074467658997, -0.2521023750305176, -0.2789222300052643, 0.5203965902328491, -0.05680610239505768, 0.035450875759124756, 0.06587979942560196, -0.45105597376823425, 0.0322563573718071, 0.16768893599510193, 0.5530168414115906, -0.18503111600875854, 0.2017785608768463, -0.059397418051958084, 0.4893559217453003, 0.5270171761512756, 0.0005122683942317963, -0.1543310284614563, 0.15243995189666748, 0.14646774530410767, 0.07625628262758255, -0.03768445551395416, 0.3673478066921234, 0.05485774576663971, -0.22224962711334229, 0.4481472969055176, -0.0057780202478170395, -0.22578765451908112, -0.18706944584846497, -0.10959067940711975, -0.26541081070899963, -0.2833309471607208, 0.13767114281654358, -0.1600017100572586, 0.13109618425369263, 0.2893316447734833, 0.3210524916648865, -0.07106977701187134, -0.5795766711235046, -0.07721082866191864, 0.39042043685913086, 0.15024107694625854, 0.1212044358253479, -0.39578667283058167, 0.06771234422922134, -0.5080149173736572, 0.3432265520095825, 0.24453184008598328, 0.28298795223236084, -0.15224531292915344, -0.07150012999773026, 0.10070975124835968, -0.26286590099334717, 0.2946960926055908, -0.450811505317688, -0.10308392345905304, -0.056670330464839935, -0.03291384130716324, -0.3269294798374176, -0.19002771377563477, -0.16598448157310486, 0.4601231515407562, 0.47317570447921753, 0.465853214263916, -0.3707205057144165, -0.08754470944404602, 0.2038038671016693, -0.12906353175640106, -0.13388808071613312, -0.17332911491394043, -0.32800331711769104, -0.4676930606365204, -0.16358797252178192, -0.1462998390197754, -0.12559810280799866, 0.33448296785354614, -0.013892686925828457, -0.04778723046183586, 0.08378373831510544, 0.09194675832986832, 0.3580939769744873, 0.34259915351867676, 0.07846525311470032, 0.11075650155544281, 0.10656103491783142, 0.21385885775089264, 0.5311782360076904, 0.12926630675792694, 0.5807284116744995, 0.1455211490392685, -0.06633424758911133, -0.019350875169038773, -0.24416658282279968, 0.14895093441009521, 0.4908931851387024, 0.1351771205663681, 0.053816210478544235, 0.037483520805835724, -0.08207686245441437, -0.2781609296798706, 0.07523505389690399, -0.007691921666264534, -0.05447133257985115, -0.4696718752384186, -0.33327388763427734, 0.2758778929710388, 0.29490888118743896, -0.14097779989242554, -0.1551348716020584, -0.24002976715564728, -0.3468589782714844, 0.2835216522216797, 0.24371957778930664, 1.0874791145324707, 0.2417486160993576, -0.02210623398423195, -0.0018390249460935593, -0.15207722783088684, 0.9227550625801086, -0.276262491941452, 0.4379432797431946, -0.3907809257507324, -0.2073115110397339, -0.042194053530693054, -0.16487109661102295, 0.1221630796790123, 0.022072769701480865, -0.38290151953697205, 0.3094184696674347, 0.3011414408683777, 0.020166737958788872, -0.11545133590698242, 0.3871205151081085, -0.21533185243606567, -0.14622919261455536, -0.285542756319046, -0.02860609069466591, -0.3070119023323059, 0.4314887821674347, -0.02508636936545372, -0.05846847593784332, -0.21924766898155212, -0.21921728551387787, -0.3474550247192383, 0.2333538830280304, 0.03732386976480484, -0.0337875597178936, 0.26438504457473755, -0.2253141552209854, -0.05423031002283096, 0.05196887254714966, 0.3515733480453491, -0.08515338599681854, -0.10732707381248474, -0.0710335522890091, -0.18470579385757446, -0.14279161393642426, 0.12849223613739014, 0.014371451921761036, 0.23452402651309967, -0.08470719307661057, 0.06245040148496628, 0.06788527965545654, -0.09753068536520004, -0.3518664836883545, -0.11455514281988144, -0.06113250553607941, -0.20195893943309784, -0.15386666357517242, -0.04333508759737015, 0.13102227449417114, -0.08529549837112427, -0.03876178339123726, 0.0029937177896499634, -0.1441469043493271, -0.22105348110198975, 0.2700084149837494, -0.06912082433700562, -0.38187360763549805, -0.15728235244750977, 0.3526187837123871, 0.44284093379974365, 0.036589257419109344, 0.3667510747909546, -0.152848020195961, -0.07765865325927734, -0.1503259837627411, -0.15030476450920105, -0.3644888401031494, -0.20347857475280762, 0.23492255806922913, -0.23662817478179932, -0.12594090402126312, 0.14240820705890656, 0.4553375244140625, 0.18074476718902588, 0.33318740129470825, -0.06118413060903549, -0.510886549949646, -0.008190404623746872, 0.17526480555534363, 0.032078810036182404, 0.05828070640563965, -0.058064572513103485, 0.5259189605712891, -0.2515856623649597, 0.21031665802001953, -0.1909005045890808, -0.14888900518417358, -0.44389593601226807, 0.3342974781990051, 0.1320309340953827, -0.043771445751190186, -0.00957229919731617, 0.00005443580448627472, -0.16391296684741974, 0.3139306902885437, -0.06522475928068161, -0.005955733358860016, -0.06145970895886421, 0.20649993419647217, 0.1785380244255066, -0.1087699830532074, -0.174637109041214, 0.1248009204864502, 0.07677891105413437, 0.03751734644174576, 0.03463679552078247, 0.2473377287387848, 0.04405125230550766, 0.2818426787853241, -0.3356453478336334, 0.11022813618183136, -0.07830432802438736, 0.22765110433101654, -0.2985169589519501, 0.3604103922843933, -0.05913567915558815, 0.30647528171539307, -0.10010340064764023, 0.015594683587551117, 0.15056994557380676, -0.036064304411411285, 0.0484817698597908, 0.09732745587825775, 0.07677573710680008, -0.17980913817882538, 0.1473214328289032, 0.18416355550289154, -0.08916452527046204, 0.6363036036491394, -0.07548834383487701, -0.19107255339622498, 0.10431263595819473, 0.04541569575667381, -0.08597458899021149, 0.0015469093341380358, 0.2222021222114563, 0.025047065690159798, -0.05612965673208237, 0.5138844847679138, 0.08978398144245148, -0.13231056928634644, -0.137436643242836, 0.11166287213563919, 0.33755630254745483, -0.0757070928812027, 0.10394451022148132, 0.19666683673858643, -0.20539212226867676, 0.09556091576814651, 0.3046140670776367, -0.2812727093696594, 0.15195713937282562, -0.000051431357860565186, 0.027243182063102722, 0.4766053259372711, 0.16119647026062012, 0.3706953227519989, -0.02431722730398178, -0.11332204937934875, 0.0299596656113863, 0.6413366794586182, 0.28227782249450684, 0.37167811393737793, 0.2842690348625183, 0.387371689081192, -0.2766481935977936, 0.02878662385046482, -0.36788809299468994, -0.17549064755439758, -0.06385762989521027, -0.14051616191864014, -0.46062153577804565, -0.04844362661242485, -0.17240488529205322, 0.20627908408641815, -0.14023509621620178, 0.12604156136512756, 0.4599732756614685, -0.07544796168804169, -0.20757706463336945, -0.37508833408355713, 0.12007813155651093, 0.028892848640680313, 0.16698630154132843, -0.2368585765361786, 0.31650424003601074, -0.07666333019733429, -0.0866452008485794, 0.32038334012031555, 0.5079480409622192, 0.4745161533355713, 0.30628952383995056, -0.05767250433564186, -0.20901255309581757, 0.0815262645483017, 0.11034242808818817, 0.2296566367149353, 0.7636833190917969, 0.2975425124168396, -0.2571280896663666, 0.3881787657737732, -0.06476611644029617, -0.07674150913953781, 0.4682409465312958, 0.07057179510593414, -0.08752591907978058, -0.022166043519973755, 0.4095025062561035, -0.0731789767742157, 0.11578703671693802, 0.021555233746767044, 0.13730394840240479, -0.2554468810558319, -0.27098923921585083, 0.06821246445178986, -0.27079296112060547, 0.30509594082832336, 0.013167710974812508, 0.022203564643859863, -0.30443084239959717, 0.6118448972702026, 0.3897586166858673, -0.12752357125282288, -0.20482808351516724, -0.07230465114116669, -0.43062710762023926, 0.34660300612449646, -0.42263078689575195, -0.23992247879505157, -0.29968753457069397, 0.5030856728553772, -0.22747161984443665, 0.40453994274139404, -0.2877247929573059, 0.37956419587135315, -0.06517479568719864, -0.031675487756729126, -0.22983890771865845, -0.06720118224620819, -0.12271597981452942, -0.29663899540901184, 0.02004878968000412, -0.2617557644844055, 0.2225911021232605, 0.07699591666460037, -0.13086313009262085, 0.0015138424932956696, -0.01643592119216919, 0.016017727553844452, 0.40780171751976013, -0.012907572090625763, 0.5090906023979187, 0.39042145013809204, -0.14212578535079956, -0.15031196177005768, -0.005523864179849625, -0.3197630047798157, -0.3111400902271271, 0.41492000222206116, 0.2139553278684616, 0.3597555160522461, -0.174125537276268, 0.6665087342262268, -0.06747272610664368, 0.3192364275455475, 0.05682755261659622, 0.02263082005083561, -0.25127550959587097, -0.057016998529434204, -0.2459755539894104, 0.03582847863435745, 0.06754429638385773, 0.45341363549232483, 0.15364189445972443, -0.027757912874221802, -0.2425970435142517, -0.3491297662258148, 0.47973042726516724, -0.3962055742740631, 0.029295947402715683, -0.006679593585431576, 0.1837034374475479, 0.24500307440757751, -0.22634100914001465, -0.7466574907302856, -0.07005758583545685, 0.25677353143692017, -0.2124132513999939, 0.05072428658604622, -0.02079637348651886, -0.05016607418656349, -0.2243598997592926, -0.12880520522594452, -0.2339140772819519, 0.21499401330947876, -0.41724684834480286, 0.286338746547699, -0.1293872594833374 ]
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
[ -0.21565434336662292, 0.12347611784934998, 0.027687493711709976, 0.48136138916015625, -0.011550076305866241, 0.09883285313844681, -0.13197048008441925, 0.3685871958732605, -0.051696889102458954, 0.18515688180923462, 0.07529840618371964, 0.1368841975927353, 0.08542821556329727, -0.13447432219982147, -0.032238900661468506, 0.025298824533820152, -0.09251757711172104, 0.19295063614845276, 0.038546644151210785, -0.09255413711071014, -0.1750839650630951, 0.4164971113204956, -0.3340628147125244, 0.007659561932086945, -0.2115810513496399, -0.10277733951807022, 0.08266931772232056, 0.4248708486557007, 0.005414196290075779, -0.3878881335258484, 0.017097117379307747, -0.01549144834280014, 0.07717113941907883, 0.20357677340507507, -0.0001237727265106514, 0.05251894146203995, 0.6025201082229614, -0.26091182231903076, -0.15160660445690155, -0.24732501804828644, -0.39834678173065186, -0.3468354344367981, -0.15325507521629333, -0.3098127543926239, 0.04340548813343048, -0.13741998374462128, 0.27026909589767456, -0.21010427176952362, 0.3015902638435364, 0.3589797019958496, 0.10676811635494232, 0.1728033870458603, -0.0673220157623291, 0.13837473094463348, 0.278253972530365, 0.06059841811656952, -0.05707038193941116, -0.07430325448513031, -0.030549347400665283, -0.2105666995048523, -0.09983684867620468, 0.3154500722885132, -0.2957831025123596, -0.004714282229542732, 0.17416709661483765, 0.0982152670621872, 0.3729245960712433, -0.20413394272327423, 0.17276903986930847, 0.29544177651405334, 0.5296697020530701, 0.13004115223884583, -0.21195344626903534, -0.4479368031024933, -0.15491218864917755, 0.11767435073852539, 0.4065711796283722, 0.4266493618488312, -0.20510539412498474, -0.0010265074670314789, -0.26339811086654663, -0.034661389887332916, -0.19299185276031494, 0.06627212464809418, 0.2431643307209015, 0.24165329337120056, 0.08146435022354126, 0.06608115136623383, -0.02012054994702339, 0.07188792526721954, 0.09668942540884018, -0.2602767050266266, -0.14188729226589203, 0.09542854130268097, 0.029838938266038895, -0.07551892846822739, -0.1188725158572197, -0.07541238516569138, 0.20608875155448914, 0.20481674373149872, 0.36635321378707886, 0.019850444048643112, 0.2098536342382431, 0.13504046201705933, 0.33779874444007874, -0.07288758456707001, 0.048855364322662354, 0.4268062710762024, 0.20002338290214539, -0.03538358956575394, 0.08910275995731354, 0.01330018974840641, 0.22526928782463074, -0.2736580967903137, -0.15773549675941467, 0.17779546976089478, 0.24600408971309662, -0.49686187505722046, -0.4554826021194458, 0.14754261076450348, -0.42383873462677, -0.2733585834503174, -0.25034916400909424, 0.26493072509765625, -0.30463919043540955, 0.3659850060939789, -0.13079878687858582, 0.08853168785572052, -0.38467368483543396, -0.10326562821865082, -0.13025537133216858, 0.16457189619541168, -0.1479642242193222, 0.20909681916236877, 0.1741015762090683, -0.1450197994709015, 0.36790335178375244, -0.17715896666049957, -0.06438577175140381, -0.2138010561466217, 0.36703091859817505, -0.17966467142105103, 0.04696182161569595, 0.25024035573005676, 0.10622438788414001, 0.17181415855884552, -0.11795464158058167, -0.2918159067630768, -0.285082072019577, 0.09585841745138168, -0.27959397435188293, -0.393039733171463, 0.11864625662565231, 0.14444109797477722, -0.3303492069244385, -0.04221513122320175, 0.0714513435959816, 0.04952285438776016, 0.41477566957473755, -0.4803284704685211, -0.09707238525152206, -0.22827772796154022, -0.15888912975788116, -0.1285473108291626, -0.2621917724609375, 0.6329419612884521, 0.13807927072048187, -0.10230564326047897, 0.2826194167137146, -0.029114600270986557, 0.5553823113441467, 0.3593013286590576, 0.06276628375053406, 0.1281600445508957, -0.19398050010204315, 0.47451382875442505, 0.11766644567251205, -0.41009315848350525, -0.1508031040430069, 0.469144344329834, -0.04133346676826477, 0.3392702639102936, 0.43024924397468567, 0.10246627032756805, 0.2844761610031128, -0.17926861345767975, -0.042457446455955505, 0.3146088123321533, 0.06323046237230301, -0.028442656621336937, -0.45213353633880615, -0.14667300879955292, 0.436776340007782, 0.10489121079444885, 0.18474258482456207, 0.01015954464673996, 0.1415489912033081, 0.4389921724796295, 0.3131595253944397, -0.10660916566848755, 0.10542171448469162, 0.2982519268989563, -0.4476800560951233, 0.1695100963115692, 0.10063304007053375, -0.1074298843741417, -0.3479068875312805, -0.07296061515808105, -0.3559933304786682, 0.14380577206611633, 0.0009279176592826843, -0.22381561994552612, -0.17331531643867493, -0.3992129862308502, -0.07715778052806854, -0.003456883132457733, -0.02226208709180355, 0.3663998246192932, 0.28739380836486816, 0.154518261551857, -0.059055425226688385, 0.36677902936935425, -0.033696845173835754, 0.18653325736522675, -0.8968737721443176, 0.2386501580476761, -0.15186844766139984, -0.2217145562171936, 0.4185647666454315, 0.3158281743526459, 0.19127628207206726, -0.19332465529441833, -0.0772937685251236, 0.38104304671287537, -0.10586760938167572, 0.4759340286254883, -0.2341872751712799, 0.07466614246368408, 0.18912166357040405, -0.31206458806991577, -0.020262254402041435, 0.405687540769577, 0.09887292236089706, -0.1194395050406456, -0.03732912987470627, 0.422737717628479, -0.34207528829574585, 0.19422794878482819, 0.1411232054233551, -0.06570770591497421, 0.06375399231910706, -0.26771435141563416, -0.24543547630310059, -0.52154541015625, 0.6221478581428528, -0.06730543076992035, -0.26908358931541443, 0.04679863154888153, -0.2562190592288971, 0.025931358337402344, 0.5729615092277527, 0.0020754151046276093, -0.08300849050283432, -0.06289932131767273, 0.06719455122947693, -0.06301282346248627, -0.018116867169737816, 0.1845977008342743, 0.45713484287261963, 0.09094727039337158, -0.05557703599333763, 0.06711187213659286, -0.1318192183971405, -0.4584287405014038, 0.1281435191631317, -0.04832863807678223, -0.027972206473350525, 0.5724952816963196, -0.011887378990650177, 0.06637779623270035, -0.3477863073348999, -0.02124866098165512, 0.1145634725689888, 0.13852645456790924, -0.3325786590576172, -0.0929332971572876, -0.32997098565101624, -0.4057203233242035, -0.3795689046382904, -0.07616891711950302, -0.1871165782213211, -0.3854970335960388, 0.1835039258003235, 0.21754324436187744, -0.2500670254230499, 0.02811361849308014, -0.1977851539850235, 0.005206704139709473, -0.34196656942367554, 0.04885482043027878, -0.13127970695495605, -0.19375471770763397, -0.21058034896850586, -0.06514140963554382, 0.2704784572124481, 0.38981252908706665, 0.2202957719564438, -0.5454668402671814, -0.2100881040096283, -0.04477045685052872, -0.20753580331802368, 0.020225614309310913, -0.05990787595510483, -0.17760729789733887, 0.09295939654111862, -0.15598003566265106, 0.5214743614196777, -0.16104164719581604, 0.14776454865932465, -0.12871910631656647, -0.27084916830062866, 0.12490451335906982, 0.22166506946086884, -0.14447900652885437, -0.22062841057777405, -0.5866003036499023, -0.17907948791980743, -0.4568946957588196, -0.09897778928279877, 0.2818935215473175, 0.29488685727119446, 0.04235469549894333, -0.16215060651302338, -0.10634422302246094, -0.21439433097839355, 0.356164813041687, -0.16313152015209198, 0.05585532262921333, 0.32341793179512024, -0.1839478611946106, -0.27022504806518555, -0.021707266569137573, 0.1376059502363205, -0.05147732421755791, 0.14158521592617035, -0.46213555335998535, -0.059683170169591904, -0.18343858420848846, -0.15344777703285217, -0.28296756744384766, -0.11173214018344879, 0.4699302911758423, -0.03412950783967972, -0.004535004496574402, -0.056615859270095825, -0.2683960199356079, 0.017017774283885956, 0.054796285927295685, 0.300724059343338, -0.1281159222126007, 0.2338898479938507, -0.09957724064588547, 0.41609013080596924, 0.5284243822097778, 0.1846238523721695, -0.06775498390197754, 0.07574257999658585, 0.27277037501335144, 0.16633430123329163, -0.016635319218039513, 0.33610039949417114, -0.02377237379550934, -0.09101869910955429, 0.4331532120704651, -0.0023463983088731766, -0.09685909003019333, -0.14553609490394592, -0.24594686925411224, -0.34901532530784607, -0.19343861937522888, 0.14629052579402924, -0.1666952669620514, 0.14870798587799072, 0.22088949382305145, 0.3066439926624298, -0.03872033208608627, -0.5155571699142456, 0.024541765451431274, 0.45322561264038086, 0.32034769654273987, 0.18500789999961853, -0.4796754717826843, 0.09125615656375885, -0.4247417151927948, 0.29793304204940796, 0.14495402574539185, 0.3121426999568939, -0.18283198773860931, -0.14990225434303284, 0.10623293370008469, -0.2515309751033783, 0.2325572371482849, -0.4161829650402069, 0.019843805581331253, 0.059046562761068344, -0.0065513551235198975, -0.3043077290058136, -0.1795516163110733, -0.2614225447177887, 0.4053453505039215, 0.3865751028060913, 0.43123432993888855, -0.3831046223640442, -0.16495168209075928, 0.09735054522752762, 0.005135940853506327, -0.10864178836345673, -0.08253543823957443, -0.29944607615470886, -0.3675912618637085, -0.22960831224918365, -0.29456833004951477, -0.08849793672561646, 0.26921510696411133, -0.09233544021844864, -0.02708631195127964, 0.03468576818704605, 0.10009204596281052, 0.3609338402748108, 0.27080318331718445, 0.08658649772405624, -0.10598383843898773, 0.07422317564487457, 0.15665021538734436, 0.5734142065048218, 0.11814229935407639, 0.5895763635635376, 0.06899486482143402, -0.18111127614974976, 0.024747120216488838, -0.4304133653640747, 0.19162777066230774, 0.46582555770874023, 0.081842340528965, 0.044300612062215805, 0.12616181373596191, -0.024200141429901123, -0.4446394741535187, -0.05971920117735863, 0.16697165369987488, -0.08109410107135773, -0.40408653020858765, -0.42042431235313416, 0.27812016010284424, 0.20545518398284912, -0.1348201483488083, -0.1648535281419754, -0.13094422221183777, -0.2597302496433258, 0.3258017897605896, 0.13428542017936707, 1.0393658876419067, 0.2653292715549469, 0.07107443362474442, -0.06675074994564056, -0.19366438686847687, 0.7017489671707153, -0.23345686495304108, 0.4610561728477478, -0.45273011922836304, -0.18189243972301483, -0.09579324722290039, -0.18577441573143005, 0.2801097333431244, 0.13668379187583923, -0.21743783354759216, 0.3387278914451599, 0.20434297621250153, -0.02720646932721138, -0.10105759650468826, 0.47604691982269287, -0.2647891044616699, -0.03463727980852127, -0.10286451131105423, -0.05299068242311478, -0.18343351781368256, 0.456830233335495, -0.0012250952422618866, -0.1371757835149765, -0.1618817299604416, -0.27150124311447144, -0.3544192910194397, 0.28696689009666443, 0.11578189581632614, -0.07829161733388901, 0.18359063565731049, -0.1999584138393402, 0.05459507927298546, 0.20607027411460876, 0.2277209460735321, -0.08420178294181824, -0.03865048289299011, -0.09670551121234894, -0.1803741753101349, -0.1994977593421936, 0.1871034801006317, 0.12525026500225067, 0.3472328186035156, -0.11429058015346527, 0.062304623425006866, 0.1867659091949463, -0.1503848284482956, -0.27849090099334717, -0.3265236020088196, 0.020360734313726425, -0.2225656658411026, -0.1865178346633911, -0.14518161118030548, 0.015972092747688293, -0.16698449850082397, -0.0472404845058918, -0.0031716078519821167, -0.0846782922744751, -0.0963406041264534, 0.19067338109016418, -0.1776462197303772, -0.4502430558204651, -0.1255148947238922, 0.2783941328525543, 0.36409792304039, -0.030169572681188583, 0.43103885650634766, -0.0935879498720169, -0.12448450177907944, -0.19571992754936218, -0.1521313637495041, -0.23456509411334991, -0.4015829861164093, 0.17225393652915955, -0.20613950490951538, -0.10457724332809448, 0.17576433718204498, 0.24396362900733948, 0.08783169090747833, 0.1960269808769226, -0.12787160277366638, -0.6653318405151367, 0.12647387385368347, 0.1522238850593567, 0.218843474984169, -0.019053328782320023, -0.04890122264623642, 0.32298335433006287, -0.23107977211475372, 0.1628679484128952, -0.18060219287872314, -0.07511378079652786, -0.30882522463798523, 0.3002735674381256, 0.16600315272808075, 0.04077087715268135, -0.10946865379810333, 0.029191341251134872, -0.135566845536232, 0.2487284392118454, -0.17746086418628693, 0.03926803171634674, -0.14119011163711548, 0.2227548360824585, 0.15719327330589294, -0.05965302884578705, -0.10897029936313629, 0.019748196005821228, 0.09330268204212189, -0.021798525005578995, 0.033921077847480774, 0.29959672689437866, -0.0386262983083725, 0.34890031814575195, -0.3989139795303345, 0.14669688045978546, -0.09037596732378006, 0.2649862468242645, -0.3462001085281372, 0.38974738121032715, -0.20149533450603485, 0.24060280621051788, -0.03934531658887863, 0.0229201540350914, -0.09879704564809799, -0.03177809715270996, 0.12119582295417786, 0.06413545459508896, 0.31527379155158997, -0.15873874723911285, 0.21573194861412048, 0.2088232785463333, -0.03734010457992554, 0.7320029139518738, -0.12561552226543427, -0.11623473465442657, 0.062017686665058136, 0.05478934198617935, -0.11067084223031998, -0.017984287813305855, 0.18334759771823883, -0.08880674839019775, -0.08237115293741226, 0.36358341574668884, 0.19489237666130066, -0.2530840039253235, -0.1885806769132614, 0.10765566676855087, 0.41206881403923035, -0.08131382614374161, 0.0942225307226181, 0.23868374526500702, -0.011601142585277557, 0.017933592200279236, 0.19761022925376892, -0.1533752828836441, 0.0799495130777359, 0.1149350255727768, 0.09460396319627762, 0.523320734500885, 0.1414513736963272, 0.4351023733615875, -0.23179784417152405, -0.1278584599494934, 0.018450116738677025, 0.6700650453567505, 0.19276079535484314, 0.48585203289985657, 0.25121283531188965, 0.3217865824699402, -0.29630574584007263, 0.15058808028697968, -0.26697838306427, -0.17418262362480164, -0.03776218742132187, -0.1280360072851181, -0.4697096347808838, -0.12692967057228088, -0.20043063163757324, 0.06337586045265198, -0.07095864415168762, 0.2730647623538971, 0.3169322609901428, -0.05591939389705658, -0.19864140450954437, -0.29472509026527405, 0.17275679111480713, -0.15478473901748657, 0.1828123778104782, -0.3227279484272003, 0.27617940306663513, 0.10261663049459457, -0.032872408628463745, 0.3759278357028961, 0.5440003871917725, 0.6268026232719421, 0.4050438702106476, -0.10989484190940857, -0.12854225933551788, -0.05433844402432442, 0.047856591641902924, 0.18827101588249207, 0.8659350872039795, 0.21049317717552185, -0.173279270529747, 0.37000465393066406, -0.03926464170217514, -0.08896410465240479, 0.3375842869281769, 0.16805602610111237, -0.04063234478235245, 0.047046732157468796, 0.3882298171520233, -0.1983952671289444, -0.0021037310361862183, -0.0224141925573349, 0.034169699996709824, -0.4142058193683624, -0.2665117383003235, 0.17878387868404388, -0.14471019804477692, 0.2802107632160187, -0.15230447053909302, 0.019083548337221146, -0.17731128633022308, 0.5529252290725708, 0.47746336460113525, -0.17004427313804626, -0.09665463119745255, -0.022314898669719696, -0.5047860145568848, 0.5072606801986694, -0.41914352774620056, -0.3187641203403473, -0.2432267963886261, 0.3742888867855072, -0.19792011380195618, 0.26287841796875, -0.09300824999809265, 0.4435423016548157, -0.13743431866168976, 0.10049092024564743, -0.2129075676202774, -0.019861925393342972, 0.014264997094869614, -0.15694934129714966, -0.09848317503929138, -0.24548834562301636, 0.25242987275123596, 0.04092523828148842, -0.07209350168704987, -0.01097242534160614, 0.1572078913450241, -0.0024539101868867874, 0.274394690990448, -0.026171691715717316, 0.4778502583503723, 0.44526028633117676, -0.2909839451313019, -0.2139376699924469, -0.11031825840473175, -0.23036065697669983, -0.2815762460231781, 0.4010995924472809, 0.22582034766674042, 0.3603361248970032, -0.20492468774318695, 0.5031101703643799, -0.09256672859191895, 0.2626860439777374, -0.15710869431495667, 0.14505548775196075, -0.2188476175069809, 0.03149867057800293, -0.23324859142303467, -0.01633145473897457, 0.15726915001869202, 0.45763304829597473, 0.18879970908164978, -0.12539446353912354, -0.23119822144508362, -0.373332142829895, 0.41276806592941284, -0.3910871744155884, -0.07988372445106506, -0.017234738916158676, 0.19397437572479248, 0.12862254679203033, -0.041532017290592194, -0.6443723440170288, -0.0595877505838871, 0.2283196747303009, -0.11385009437799454, 0.029822751879692078, -0.06642253696918488, -0.15205009281635284, -0.3520394563674927, -0.034354664385318756, -0.2507988512516022, 0.27073606848716736, -0.41924959421157837, 0.40055108070373535, -0.10350938886404037 ]
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.
[ -0.15917661786079407, 0.061384305357933044, 0.06137007102370262, 0.3985082507133484, -0.09111974388360977, 0.0681394562125206, -0.1541566103696823, 0.3743869364261627, -0.05398889631032944, 0.1291232705116272, 0.09490998089313507, 0.16967818140983582, 0.050891757011413574, -0.07023631036281586, -0.02994246408343315, -0.031388700008392334, -0.04626775532960892, 0.20410287380218506, 0.04388732090592384, 0.001029621809720993, -0.019637158140540123, 0.44040247797966003, -0.30174973607063293, -0.022966966032981873, -0.2469935566186905, -0.16833093762397766, -0.04212997481226921, 0.32819753885269165, 0.0039550624787807465, -0.3751760721206665, 0.24395911395549774, -0.07514706254005432, 0.13529175519943237, 0.19081103801727295, -0.00012187100219307467, -0.013404510915279388, 0.5785959959030151, -0.2156340628862381, -0.20625445246696472, -0.25492215156555176, -0.4047054648399353, -0.3622797131538391, -0.08360154926776886, -0.23551985621452332, -0.02168666198849678, -0.055038001388311386, 0.17205533385276794, -0.08597995340824127, 0.2725290060043335, 0.2854219675064087, 0.11910401284694672, 0.15230633318424225, -0.1194787472486496, 0.03129195421934128, 0.1314687430858612, 0.11148032546043396, -0.03822559118270874, -0.09602130949497223, -0.0876149833202362, -0.2862391769886017, -0.10378170758485794, 0.43996766209602356, -0.19715236127376556, 0.15519720315933228, 0.2258736789226532, 0.11542969197034836, 0.4008738100528717, -0.11456099152565002, 0.03758153319358826, 0.3904353976249695, 0.5094952583312988, 0.16033314168453217, -0.20868922770023346, -0.5066426992416382, -0.050456468015909195, 0.19382116198539734, 0.41565215587615967, 0.3648850917816162, -0.23909999430179596, 0.008471596986055374, -0.4111096262931824, -0.07393138110637665, -0.0905098170042038, 0.10173960030078888, 0.2668370306491852, 0.19856902956962585, 0.11837191134691238, 0.006740126758813858, 0.12967170774936676, -0.041098929941654205, 0.11463387310504913, -0.24428820610046387, -0.13999293744564056, 0.112827830016613, -0.1838788092136383, -0.1930742710828781, -0.1448516845703125, -0.017406605184078217, 0.2120879590511322, 0.2155565768480301, 0.351828396320343, -0.054053351283073425, 0.22565490007400513, 0.1119692251086235, 0.42831626534461975, 0.04790709167718887, -0.010735899209976196, 0.5196449756622314, 0.08321750164031982, 0.0033717905171215534, 0.11515559256076813, 0.059673406183719635, 0.265980988740921, -0.2190958708524704, -0.14518535137176514, 0.1199166476726532, 0.15946324169635773, -0.5667430758476257, -0.4398612678050995, 0.10815741866827011, -0.5615882277488708, -0.24343468248844147, -0.09248389303684235, 0.17566204071044922, -0.3614131212234497, 0.3320038616657257, -0.15387743711471558, 0.04779893159866333, -0.3616383373737335, -0.05944567546248436, -0.2066727578639984, 0.17067909240722656, -0.19749875366687775, 0.20532439649105072, 0.20146071910858154, -0.15544496476650238, 0.40264222025871277, -0.10782983154058456, -0.05867202579975128, -0.19689896702766418, 0.4011359214782715, -0.10814116895198822, 0.020010434091091156, 0.1572449654340744, -0.029180467128753662, 0.16998447477817535, -0.0008436329662799835, -0.216933012008667, -0.38861730694770813, 0.07954563200473785, -0.22302427887916565, -0.5368489027023315, 0.10469014197587967, 0.1530231237411499, -0.44328850507736206, -0.06129131466150284, 0.09627173840999603, 0.21096181869506836, 0.36438441276550293, -0.33245113492012024, -0.09742449969053268, -0.22020064294338226, -0.12947271764278412, -0.05456848070025444, -0.20552809536457062, 0.4950895309448242, 0.14020079374313354, -0.266052782535553, 0.13045363128185272, -0.018282391130924225, 0.5335451364517212, 0.3493090569972992, 0.034334439784288406, 0.09644124656915665, -0.1925896555185318, 0.48758596181869507, 0.13571110367774963, -0.35429847240448, -0.16301201283931732, 0.5454862713813782, 0.028683774173259735, 0.28070491552352905, 0.25120460987091064, 0.11916983872652054, 0.1942966878414154, -0.09376268088817596, 0.01077117770910263, 0.408086895942688, -0.021602002903819084, 0.03519278019666672, -0.4514918923377991, -0.3096883296966553, 0.4148997664451599, 0.20669451355934143, 0.11799587309360504, -0.04226302355527878, -0.0019413307309150696, 0.4964992105960846, 0.30462905764579773, -0.13594847917556763, 0.05396605283021927, 0.29197072982788086, -0.3540269434452057, 0.1288684904575348, 0.038535453379154205, -0.11504384875297546, -0.3697326183319092, -0.023021630942821503, -0.28698253631591797, 0.2818140387535095, 0.10698975622653961, -0.2584027945995331, -0.1996178925037384, -0.4155747592449188, -0.025226790457963943, -0.02996254339814186, 0.010714028030633926, 0.2815219461917877, 0.39745810627937317, 0.05119476094841957, 0.06851144134998322, 0.428454726934433, -0.06218875199556351, 0.17390741407871246, -0.796796441078186, 0.23201800882816315, -0.11876515299081802, -0.21117733418941498, 0.5127648711204529, 0.3413342833518982, 0.22369951009750366, -0.16931097209453583, -0.09573248028755188, 0.4365609586238861, -0.10373702645301819, 0.40195852518081665, -0.17406710982322693, 0.010552763938903809, 0.20788681507110596, -0.29545900225639343, 0.018217746168375015, 0.3747367858886719, -0.0284036286175251, -0.06868936866521835, -0.13359332084655762, 0.4167672395706177, -0.33828726410865784, 0.16545343399047852, 0.06876222044229507, 0.01500893197953701, 0.07816560566425323, -0.2387959510087967, -0.19139666855335236, -0.45546427369117737, 0.6415620446205139, -0.0951366126537323, -0.32226407527923584, 0.042292818427085876, -0.20106329023838043, 0.03303185850381851, 0.43965309858322144, 0.021865837275981903, -0.09082555025815964, -0.0264102965593338, 0.02838919684290886, -0.1608659327030182, 0.0957215428352356, 0.2750535011291504, 0.5138395428657532, 0.12376817315816879, 0.023313958197832108, 0.015095815993845463, -0.1677716225385666, -0.33145835995674133, 0.12296850234270096, -0.0629400908946991, 0.027277668938040733, 0.36706891655921936, -0.0033676116727292538, -0.031910091638565063, -0.3108045756816864, -0.012369511649012566, 0.15956242382526398, 0.24881474673748016, -0.32820871472358704, -0.13419055938720703, -0.35838791728019714, -0.40440231561660767, -0.3453233242034912, -0.1350657194852829, -0.28269660472869873, -0.36944037675857544, 0.20206354558467865, 0.19996565580368042, -0.1682591289281845, 0.13886970281600952, -0.1668083816766739, 0.04333499073982239, -0.3156355023384094, -0.05282643437385559, -0.01851584017276764, -0.21042899787425995, -0.27342867851257324, -0.02624230459332466, 0.22404198348522186, 0.42671772837638855, 0.24252188205718994, -0.39223507046699524, -0.15776924788951874, 0.04673943668603897, -0.2580878734588623, 0.06250384449958801, -0.05618477612733841, -0.15136504173278809, 0.1768348515033722, -0.1386253833770752, 0.37565261125564575, -0.22181446850299835, 0.10164041817188263, -0.1899556964635849, -0.3374048173427582, 0.027316994965076447, 0.17911186814308167, 0.03542134910821915, -0.19235356152057648, -0.5341885685920715, -0.26857152581214905, -0.36351504921913147, -0.06236624717712402, 0.28557682037353516, 0.24182623624801636, 0.03172805905342102, -0.1204400360584259, -0.13270418345928192, -0.1704920083284378, 0.2503783702850342, -0.30170783400535583, -0.05889040231704712, 0.24042047560214996, -0.09786741435527802, -0.16751530766487122, -0.03147032484412193, 0.0028387531638145447, -0.05066491290926933, 0.12287375330924988, -0.5468325018882751, -0.10958747565746307, -0.20560337603092194, -0.07945811003446579, -0.3269421458244324, -0.2575342655181885, 0.5263427495956421, -0.07347503304481506, -0.024189021438360214, 0.013537459075450897, -0.3449536859989166, 0.06836865842342377, 0.1019066646695137, 0.5138316750526428, -0.16110703349113464, 0.18403863906860352, -0.014320679008960724, 0.5651073455810547, 0.5774591565132141, 0.07665694504976273, -0.11549916863441467, 0.17679192125797272, 0.09244534373283386, 0.17404824495315552, 0.009827882051467896, 0.37702691555023193, 0.022528842091560364, -0.1074550524353981, 0.4157789349555969, 0.08360455185174942, -0.1998012810945511, -0.11438945680856705, -0.1578519493341446, -0.31042420864105225, -0.1871032863855362, 0.16156162321567535, -0.2299748659133911, 0.20063230395317078, 0.20180656015872955, 0.23883286118507385, -0.18510441482067108, -0.5294701457023621, -0.012654252350330353, 0.3591078519821167, 0.2272566258907318, 0.17686012387275696, -0.4701034724712372, 0.14273430407047272, -0.4302378296852112, 0.2958553731441498, 0.21121390163898468, 0.2849108576774597, -0.21085688471794128, -0.15540382266044617, 0.08103788644075394, -0.19880712032318115, 0.29555171728134155, -0.4056425988674164, -0.015020042657852173, -0.04621124267578125, -0.03189551830291748, -0.342242032289505, -0.2017230987548828, -0.13815273344516754, 0.3731078505516052, 0.5086655616760254, 0.4213860332965851, -0.4272281527519226, -0.12459447234869003, 0.16328121721744537, -0.05940474942326546, -0.13091081380844116, -0.16943767666816711, -0.25253307819366455, -0.33338984847068787, -0.17163142561912537, -0.14729195833206177, -0.1254299134016037, 0.30011674761772156, -0.02195614017546177, -0.002084653824567795, 0.1447324901819229, 0.12521997094154358, 0.3355754613876343, 0.2571823298931122, 0.16852501034736633, -0.03306189551949501, 0.10766860842704773, 0.2788693904876709, 0.540444016456604, 0.08358269184827805, 0.5131438374519348, 0.10595917701721191, -0.058910682797431946, -0.025092247873544693, -0.3047671616077423, 0.07861298322677612, 0.43017813563346863, 0.17479118704795837, 0.0852496474981308, 0.14247438311576843, -0.027434512972831726, -0.46311166882514954, -0.025743912905454636, 0.11071310937404633, -0.07656627148389816, -0.37168172001838684, -0.3909304440021515, 0.25883975625038147, 0.26937922835350037, -0.1429109275341034, -0.14606890082359314, -0.19557951390743256, -0.2553572654724121, 0.23188114166259766, 0.1594078540802002, 0.9950311779975891, 0.26009055972099304, -0.026021379977464676, -0.027609582990407944, -0.22299738228321075, 0.8679138422012329, -0.25553590059280396, 0.4108973741531372, -0.42821571230888367, -0.2512686848640442, -0.06234835088253021, -0.20891474187374115, 0.1523512750864029, 0.1181773692369461, -0.29969677329063416, 0.33157145977020264, 0.2687565088272095, -0.01166745275259018, -0.10332239419221878, 0.4452238082885742, -0.24535900354385376, -0.07309824228286743, -0.17734670639038086, 0.01480168104171753, -0.288422167301178, 0.5008419752120972, -0.01772240735590458, -0.12158550322055817, -0.21846343576908112, -0.21833479404449463, -0.4154765009880066, 0.25962021946907043, 0.18067587912082672, 0.005110269412398338, 0.2780006527900696, -0.1839195191860199, -0.0008592866361141205, 0.091573067009449, 0.3283592462539673, -0.1827334463596344, -0.17481862008571625, 0.006974555552005768, -0.2024371325969696, -0.13968569040298462, 0.19291821122169495, 0.07448725402355194, 0.33971625566482544, -0.12549349665641785, 0.0013905540108680725, 0.05642957240343094, -0.12228953838348389, -0.3300777077674866, -0.3035987317562103, -0.1133045107126236, -0.20215778052806854, -0.18769235908985138, -0.08816827833652496, 0.07827645540237427, -0.10347446799278259, -0.06149611994624138, 0.04047757014632225, -0.040233857929706573, -0.21182696521282196, 0.23607870936393738, -0.09049469232559204, -0.3304225504398346, -0.11680580675601959, 0.39779967069625854, 0.3670256733894348, -0.03411338850855827, 0.44290435314178467, -0.1041085422039032, -0.09311012923717499, -0.20934069156646729, -0.12919022142887115, -0.31093570590019226, -0.34310370683670044, 0.16432727873325348, -0.268461674451828, -0.08667010068893433, 0.194707453250885, 0.29301851987838745, 0.23840950429439545, 0.3419274091720581, -0.001604028046131134, -0.5291138887405396, 0.016075069084763527, 0.13710251450538635, -0.024978507310152054, 0.07527955621480942, -0.1436307728290558, 0.42183154821395874, -0.20619772374629974, 0.2404181957244873, -0.21923071146011353, -0.08254688233137131, -0.4039372205734253, 0.32245495915412903, 0.11327122896909714, -0.10176185518503189, -0.03889915347099304, 0.015869105234742165, -0.11925475299358368, 0.21726515889167786, -0.14230133593082428, 0.0013298988342285156, -0.11548508703708649, 0.20137105882167816, 0.147395059466362, -0.08197376877069473, -0.10277969390153885, 0.114006906747818, 0.07142955809831619, 0.02901693433523178, -0.0539865717291832, 0.2069932222366333, -0.022113539278507233, 0.32321470975875854, -0.3484603762626648, 0.04232325404882431, -0.06938486546278, 0.23659777641296387, -0.3490239977836609, 0.29365915060043335, -0.09470129758119583, 0.21986159682273865, -0.08393579721450806, -0.00015819072723388672, -0.03363386169075966, 0.013329416513442993, 0.08372709155082703, 0.07243642210960388, 0.18849554657936096, -0.17304065823554993, 0.1039828360080719, 0.22139424085617065, 0.021713346242904663, 0.5436677932739258, -0.07953482121229172, -0.10975439101457596, 0.07915399968624115, 0.10070711374282837, -0.12908628582954407, -0.03161070495843887, 0.2547113597393036, -0.030462689697742462, -0.12914377450942993, 0.35050174593925476, 0.14029525220394135, -0.16056475043296814, -0.2005099505186081, 0.10796525329351425, 0.3961574137210846, -0.13299141824245453, 0.12379775941371918, 0.29219895601272583, -0.16179753839969635, -0.015710052102804184, 0.23938491940498352, -0.2638282775878906, 0.13483266532421112, 0.010987766087055206, 0.04418575391173363, 0.4613368809223175, 0.09753677248954773, 0.39205455780029297, -0.14451506733894348, -0.13713553547859192, -0.05386102572083473, 0.7125874757766724, 0.3032740354537964, 0.40754762291908264, 0.23052367568016052, 0.475411981344223, -0.3398900032043457, -0.003783765248954296, -0.35419604182243347, -0.18464218080043793, -0.037353575229644775, -0.10190211236476898, -0.5155592560768127, -0.10836469382047653, -0.2322472631931305, 0.12218470871448517, -0.08818584680557251, 0.14326107501983643, 0.4093429148197174, -0.0715935081243515, -0.18918578326702118, -0.4787328243255615, 0.05307842046022415, 0.014809217303991318, 0.17545342445373535, -0.3287905156612396, 0.15796281397342682, 0.036844149231910706, -0.1626705676317215, 0.39498797059059143, 0.5192059874534607, 0.5169357657432556, 0.30216145515441895, -0.14195631444454193, -0.21230356395244598, 0.054576385766267776, 0.151615709066391, 0.23794233798980713, 0.768156886100769, 0.2752503454685211, -0.2857544720172882, 0.3937153220176697, -0.05130111426115036, -0.08288401365280151, 0.40311741828918457, 0.05604337155818939, -0.06899301707744598, -0.02687562070786953, 0.44262176752090454, -0.13705362379550934, 0.11488338559865952, -0.004703536629676819, 0.11616162210702896, -0.29875025153160095, -0.3102468252182007, 0.10357506573200226, -0.17181900143623352, 0.2889469265937805, -0.07533550262451172, 0.03425173833966255, -0.31278958916664124, 0.5959123969078064, 0.43539929389953613, -0.212471604347229, -0.21041031181812286, -0.11609205603599548, -0.4998016059398651, 0.34542739391326904, -0.46122199296951294, -0.2328447550535202, -0.1896783858537674, 0.47722530364990234, -0.2099061906337738, 0.29365766048431396, -0.1712389439344406, 0.46486949920654297, -0.1285218894481659, 0.0775582417845726, -0.26981472969055176, -0.05188273265957832, -0.0817444771528244, -0.19109514355659485, 0.021983973681926727, -0.22614023089408875, 0.19919097423553467, 0.10465008020401001, -0.08411657810211182, 0.03213798627257347, 0.05748949199914932, 0.026997938752174377, 0.42762643098831177, -0.08215092122554779, 0.5023901462554932, 0.42900338768959045, -0.14516234397888184, -0.21732522547245026, -0.08768832683563232, -0.28880035877227783, -0.29746630787849426, 0.37864717841148376, 0.1362174153327942, 0.4217302203178406, -0.1751050353050232, 0.6471750736236572, -0.12871085107326508, 0.3200819194316864, 0.01959068328142166, 0.011846438981592655, -0.32712477445602417, -0.03769771009683609, -0.214523583650589, 0.026058319956064224, 0.15193985402584076, 0.471005916595459, 0.17632994055747986, -0.017446991056203842, -0.18155960738658905, -0.30553677678108215, 0.42345255613327026, -0.31965968012809753, 0.03429244086146355, -0.02669631689786911, 0.18106959760189056, 0.20320919156074524, -0.11101242154836655, -0.6931034922599792, -0.0768067017197609, 0.2931540310382843, -0.11016219109296799, 0.08574353158473969, -0.035791508853435516, -0.07137183845043182, -0.2900952994823456, -0.022308053448796272, -0.22079938650131226, 0.23416650295257568, -0.3639991879463196, 0.384461909532547, -0.08561214059591293 ]
https://github.com/huggingface/datasets/issues/215
NonMatchingSplitsSizesError when loading blog_authorship_corpus
I'm getting this same issue when loading the `imdb` corpus via `dataset = load_dataset("imdb")`. When I try `ignore_verifications=True`, no examples are read into the `train` portion of the dataset.
Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on.
29
NonMatchingSplitsSizesError when loading blog_authorship_corpus Getting this error when i run `nlp.load_dataset('blog_authorship_corpus')`. ``` raise NonMatchingSplitsSizesError(str(bad_splits)) nlp.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=610252351, num_examples=532812, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='train', num_bytes=616473500, num_examples=536323, dataset_name='blog_authorship_corpus')}, {'expected': SplitInfo(name='validation', num_bytes=37500394, num_examples=31277, dataset_name='blog_authorship_corpus'), 'recorded': SplitInfo(name='validation', num_bytes=30786661, num_examples=27766, dataset_name='blog_authorship_corpus')}] ``` Upon checking it seems like there is a disparity between the information in `datasets/blog_authorship_corpus/dataset_infos.json` and what was downloaded. Although I can get away with this by passing `ignore_verifications=True` in `load_dataset`, I'm thinking doing so might give problems later on. I'm getting this same issue when loading the `imdb` corpus via `dataset = load_dataset("imdb")`. When I try `ignore_verifications=True`, no examples are read into the `train` portion of the dataset.
[ -0.1060912013053894, 0.17192918062210083, 0.08614695072174072, 0.37152886390686035, -0.07061881572008133, 0.19752010703086853, -0.08851537108421326, 0.3589627742767334, -0.07059349119663239, 0.13236775994300842, 0.010963218286633492, 0.1926116794347763, 0.019733257591724396, -0.13543306291103363, -0.06135275214910507, 0.001089625060558319, -0.049684636294841766, 0.151227205991745, 0.08074234426021576, -0.000766657292842865, -0.01908314786851406, 0.43668192625045776, -0.3507993221282959, -0.06492380797863007, -0.27976858615875244, -0.17144758999347687, 0.005060212686657906, 0.30681169033050537, 0.06779033690690994, -0.33830156922340393, 0.24676106870174408, -0.08211196959018707, 0.1046873927116394, 0.25874534249305725, -0.00012322825205046684, 0.03466298431158066, 0.6713199019432068, -0.23092038929462433, -0.19163353741168976, -0.34996122121810913, -0.33645808696746826, -0.2255391627550125, -0.0071844905614852905, -0.2159276008605957, 0.0023932382464408875, -0.07125576585531235, 0.19345584511756897, -0.15920564532279968, 0.17167355120182037, 0.3034624457359314, 0.09692041575908661, 0.1682632714509964, -0.17914779484272003, 0.05318162962794304, 0.22526121139526367, 0.18425463140010834, -0.08278965950012207, -0.08363880217075348, -0.09633514285087585, -0.13919712603092194, -0.13139402866363525, 0.4392056465148926, -0.243143692612648, 0.1521119922399521, 0.27531176805496216, 0.16114851832389832, 0.2986014783382416, -0.20217843353748322, 0.018696190789341927, 0.37583062052726746, 0.4876858592033386, 0.10426037013530731, -0.3134579658508301, -0.47657036781311035, -0.02254026010632515, 0.18319030106067657, 0.36662253737449646, 0.3496968448162079, -0.24039380252361298, 0.03685608133673668, -0.42643657326698303, -0.029454322531819344, -0.13254500925540924, 0.08995271474123001, 0.21369346976280212, 0.10086237639188766, 0.15537028014659882, -0.0032709985971450806, 0.04466111958026886, -0.09545892477035522, 0.1590355783700943, -0.23730173707008362, -0.11320334672927856, 0.04265632480382919, -0.13779090344905853, -0.13386563956737518, -0.09080664068460464, 0.015388961881399155, 0.1906350553035736, 0.20482176542282104, 0.3135945200920105, 0.05322418361902237, 0.2250058650970459, 0.21918317675590515, 0.47183409333229065, -0.04583967849612236, -0.08414595574140549, 0.5593453645706177, 0.24461223185062408, 0.030025452375411987, 0.060235749930143356, -0.010037921369075775, 0.2779306471347809, -0.12376952171325684, -0.08744187653064728, 0.025136426091194153, 0.25217562913894653, -0.5562329292297363, -0.4696899950504303, 0.10254070162773132, -0.4754391312599182, -0.11085563153028488, -0.049945469945669174, 0.2599767744541168, -0.3506721258163452, 0.31802502274513245, -0.11212270706892014, 0.11180038750171661, -0.33322635293006897, -0.0451674610376358, -0.1908116191625595, 0.1568988561630249, -0.10932102799415588, 0.18961702287197113, 0.3031001687049866, -0.09979420900344849, 0.46441444754600525, -0.1483692228794098, 0.05276651680469513, -0.16524440050125122, 0.4266863763332367, -0.176682248711586, 0.12088873237371445, 0.22699227929115295, 0.07362121343612671, 0.23083259165287018, -0.022406894713640213, -0.14349472522735596, -0.3749014735221863, 0.14235332608222961, -0.1875194013118744, -0.5286055207252502, 0.11082946509122849, 0.125046044588089, -0.399397611618042, -0.05795259028673172, -0.010794363915920258, 0.19038967788219452, 0.3466760516166687, -0.38587918877601624, -0.0868326798081398, -0.16370455920696259, -0.07959919422864914, 0.004998224787414074, -0.1589997410774231, 0.46130794286727905, 0.18528062105178833, -0.2022431641817093, 0.15319886803627014, 0.05393008887767792, 0.5437243580818176, 0.3525787889957428, 0.02455816976726055, 0.022253315895795822, -0.14071279764175415, 0.45799392461776733, 0.18906007707118988, -0.4663136601448059, -0.12786981463432312, 0.5851624011993408, 0.0991259217262268, 0.3170833885669708, 0.22275221347808838, 0.14682313799858093, 0.23375362157821655, -0.06082319840788841, 0.040691111236810684, 0.400925874710083, -0.08982719480991364, 0.07596190273761749, -0.3674880862236023, -0.3140704035758972, 0.40807420015335083, 0.17905962467193604, 0.1312665194272995, 0.010799884796142578, -0.051287077367305756, 0.4866936206817627, 0.26025259494781494, -0.08486109972000122, 0.09025527536869049, 0.270938515663147, -0.4059145748615265, 0.13252195715904236, 0.0775730088353157, -0.09696778655052185, -0.36799970269203186, 0.0032126829028129578, -0.18765036761760712, 0.2800235152244568, 0.05730115622282028, -0.32654303312301636, -0.15178191661834717, -0.42282459139823914, -0.007933657616376877, 0.024887505918741226, -0.017824368551373482, 0.2539554834365845, 0.32474076747894287, 0.09333169460296631, -0.019458234310150146, 0.37591367959976196, -0.026518842205405235, 0.2407076358795166, -0.8501718044281006, 0.13954225182533264, -0.176375150680542, -0.2839401662349701, 0.4763646721839905, 0.276856392621994, 0.18114018440246582, -0.14808836579322815, -0.07467490434646606, 0.4480014145374298, -0.0722123309969902, 0.46613049507141113, -0.3489922285079956, 0.022523581981658936, 0.31674832105636597, -0.2829039394855499, -0.019446251913905144, 0.35974612832069397, 0.012247209437191486, -0.06112988293170929, -0.1798420250415802, 0.3726075291633606, -0.3008643686771393, 0.2264280915260315, 0.10579578578472137, 0.009542830288410187, 0.12567394971847534, -0.22969752550125122, -0.2462267428636551, -0.5783404111862183, 0.5604589581489563, -0.10399386286735535, -0.3391011953353882, -0.001096842810511589, -0.2969704270362854, -0.01972806453704834, 0.42888882756233215, 0.11643408238887787, -0.06571271270513535, -0.03648044168949127, -0.00525684654712677, -0.15044254064559937, 0.19341441988945007, 0.1804913431406021, 0.42518150806427, 0.12823976576328278, 0.08470874279737473, 0.012771890498697758, -0.17027883231639862, -0.3577145040035248, 0.1167982891201973, 0.05171636492013931, 0.003451473545283079, 0.38919001817703247, 0.0633619874715805, 0.032328858971595764, -0.34547922015190125, -0.027957897633314133, 0.20833459496498108, 0.10395724326372147, -0.38145801424980164, -0.12965868413448334, -0.4263538122177124, -0.39681047201156616, -0.2394590675830841, -0.17718367278575897, -0.12151668965816498, -0.376502126455307, 0.2047390192747116, 0.33553194999694824, -0.20929209887981415, 0.11408486217260361, -0.2720203697681427, -0.012221306562423706, -0.33466166257858276, -0.06907996535301208, -0.0707184225320816, -0.2245490998029709, -0.2937888503074646, -0.03721298649907112, 0.1956603080034256, 0.45404261350631714, 0.2249227613210678, -0.3926709294319153, -0.16009247303009033, 0.054461658000946045, -0.26246970891952515, 0.0824066773056984, -0.08520620316267014, -0.07174213975667953, 0.11912095546722412, -0.16228099167346954, 0.37041449546813965, -0.2849782407283783, 0.02957264706492424, -0.16127032041549683, -0.2947843372821808, 0.036483246833086014, 0.153392031788826, 0.03398400545120239, -0.25572696328163147, -0.5162034630775452, -0.21861325204372406, -0.3656136691570282, -0.10801637172698975, 0.26958948373794556, 0.18670199811458588, 0.0947457104921341, -0.15174269676208496, -0.09986476600170135, -0.03366941213607788, 0.30374419689178467, -0.2530960738658905, -0.11181238293647766, 0.3357352316379547, -0.22152790427207947, -0.13774646818637848, 0.054698433727025986, -0.05541529506444931, -0.05785643681883812, 0.12819057703018188, -0.5356529355049133, -0.11282755434513092, -0.22534403204917908, -0.05317845940589905, -0.21877068281173706, -0.30812549591064453, 0.4669431746006012, -0.07963355630636215, -0.024783868342638016, -0.03770994395017624, -0.3831615447998047, 0.07807578146457672, 0.06121492385864258, 0.40866583585739136, -0.11687021702528, 0.27626001834869385, -0.001049097627401352, 0.4330562353134155, 0.5189312100410461, 0.0885111540555954, 0.003336705267429352, 0.13722161948680878, 0.12602093815803528, 0.10288116335868835, 0.028162097558379173, 0.303975373506546, -0.020976334810256958, -0.12858650088310242, 0.35867953300476074, 0.0653780847787857, -0.2829357981681824, -0.1473606377840042, -0.11749204993247986, -0.3157091736793518, -0.26602551341056824, 0.1234094575047493, -0.34433069825172424, 0.12941250205039978, 0.19345565140247345, 0.3407934308052063, -0.08182297646999359, -0.5395622849464417, -0.018391627818346024, 0.28625598549842834, 0.2730642557144165, 0.13522294163703918, -0.4073798656463623, 0.15566496551036835, -0.4685330390930176, 0.2718776762485504, 0.18034248054027557, 0.3475733697414398, -0.1694795936346054, -0.15780410170555115, 0.05548808351159096, -0.20148704946041107, 0.339417964220047, -0.4746239483356476, 0.013200923800468445, -0.0041796425357460976, -0.07114433497190475, -0.4237631857395172, -0.189687579870224, -0.18537120521068573, 0.40302515029907227, 0.4932022988796234, 0.41495800018310547, -0.37820670008659363, -0.07719447463750839, 0.24144402146339417, -0.08462471514940262, -0.05414053797721863, -0.15137629210948944, -0.21289490163326263, -0.3216528296470642, -0.27340924739837646, -0.15363115072250366, 0.005820173770189285, 0.2583274841308594, -0.10156774520874023, -0.1375637948513031, 0.09534488618373871, 0.14404428005218506, 0.41291892528533936, 0.2945849299430847, 0.1835898458957672, 0.018144451081752777, 0.14950424432754517, 0.35043197870254517, 0.4505654573440552, 0.17933370172977448, 0.4881417453289032, 0.06032627075910568, -0.15531286597251892, 0.06298692524433136, -0.3202924430370331, 0.04378961771726608, 0.4678903818130493, 0.19955551624298096, 0.021901868283748627, 0.1291140466928482, -0.14375878870487213, -0.394074410200119, -0.00532188918441534, 0.06702914088964462, -0.17417608201503754, -0.33612456917762756, -0.36778974533081055, 0.2681983411312103, 0.34013622999191284, -0.11092524975538254, -0.16600102186203003, -0.15098713338375092, -0.3498978614807129, 0.3399809002876282, 0.17212873697280884, 0.9456440210342407, 0.2882981598377228, -0.01568662002682686, -0.0824362263083458, -0.17538723349571228, 0.8163899183273315, -0.236005961894989, 0.43048322200775146, -0.3811832666397095, -0.24644877016544342, -0.08875585347414017, -0.17461098730564117, 0.1384231150150299, 0.1136191338300705, -0.3597881495952606, 0.32473045587539673, 0.1303011178970337, 0.034613996744155884, -0.1917153298854828, 0.4030136466026306, -0.28494516015052795, -0.19131125509738922, -0.15553028881549835, 0.023692864924669266, -0.16818127036094666, 0.38430389761924744, 0.055374160408973694, -0.07974905520677567, -0.2306404411792755, -0.21798528730869293, -0.4340413808822632, 0.23839318752288818, 0.1523972898721695, 0.041992783546447754, 0.2618045210838318, -0.211929589509964, -0.02144467644393444, 0.12297064065933228, 0.33395975828170776, -0.17732937633991241, -0.17432652413845062, -0.0057398416101932526, -0.10337153822183609, -0.18696154654026031, 0.14108417928218842, 0.05936410278081894, 0.3844880163669586, -0.09873633831739426, -0.011704005300998688, 0.0815645158290863, -0.15946197509765625, -0.26500940322875977, -0.17423103749752045, -0.1326061338186264, -0.1864425241947174, -0.18964163959026337, -0.16944310069084167, 0.013096466660499573, -0.06749565899372101, -0.08927694708108902, 0.013088760897517204, -0.13440144062042236, -0.15610651671886444, 0.2143878936767578, -0.03449402377009392, -0.3317170739173889, -0.10979071259498596, 0.4217984080314636, 0.3353780210018158, -0.0034990832209587097, 0.4691973924636841, -0.06988655775785446, -0.11542753130197525, -0.15430672466754913, -0.07997772842645645, -0.14900021255016327, -0.3325285315513611, 0.209935262799263, -0.22150230407714844, -0.08268897235393524, 0.2382514476776123, 0.2852216362953186, 0.13365080952644348, 0.2915157675743103, -0.12268856167793274, -0.471201092004776, -0.00956140086054802, 0.2567778527736664, -0.019193951040506363, 0.11481359601020813, -0.1499737948179245, 0.3957260549068451, -0.1862943321466446, 0.15356986224651337, -0.19016431272029877, -0.07312840223312378, -0.48063111305236816, 0.35391807556152344, 0.09942642599344254, -0.007975932210683823, -0.0888342335820198, -0.027649054303765297, -0.16565971076488495, 0.28192412853240967, -0.0191587433218956, 0.0067089758813381195, -0.13281415402889252, 0.19781990349292755, 0.09791306406259537, -0.10110466182231903, -0.1399448961019516, 0.11612597107887268, 0.020546656101942062, -0.0048479437828063965, 0.03788452595472336, 0.26702621579170227, -0.008993938565254211, 0.38938796520233154, -0.36771664023399353, 0.0906316339969635, -0.0589551106095314, 0.18782053887844086, -0.3241511881351471, 0.2810356616973877, -0.12034697830677032, 0.21074873208999634, -0.044932786375284195, 0.04436337947845459, 0.03362684324383736, -0.028521068394184113, 0.03280778229236603, 0.04359268397092819, 0.1490325927734375, -0.16278497874736786, 0.140609472990036, 0.33817797899246216, -0.02393568679690361, 0.5712108612060547, -0.06357824057340622, -0.1615837961435318, 0.030100101605057716, 0.0879780650138855, -0.08462013304233551, -0.0033468804322183132, 0.20258064568042755, -0.060471802949905396, -0.06835417449474335, 0.3554293215274811, 0.18379716575145721, -0.10070470720529556, -0.1846417635679245, 0.11385100334882736, 0.4215511083602905, -0.17783145606517792, 0.19242455065250397, 0.1978377252817154, -0.2146584391593933, -0.01578349992632866, 0.16785669326782227, -0.20996692776679993, 0.1965465545654297, 0.04577769339084625, 0.020209401845932007, 0.4743586480617523, 0.15783251821994781, 0.3953703045845032, -0.07260806113481522, -0.21184641122817993, -0.06808555126190186, 0.6018927097320557, 0.3082670569419861, 0.4337640404701233, 0.22709372639656067, 0.3257054090499878, -0.2783255875110626, -0.00010927417315542698, -0.3504144549369812, -0.183759868144989, -0.011417478322982788, -0.1646636426448822, -0.47189947962760925, -0.0938858762383461, -0.23944692313671112, 0.09697262942790985, -0.15479224920272827, 0.15204116702079773, 0.39516106247901917, -0.04441917687654495, -0.20598699152469635, -0.5548534989356995, 0.1378760039806366, -0.13042554259300232, 0.19311949610710144, -0.26797130703926086, 0.18461112678050995, 0.05767602100968361, -0.11533723771572113, 0.43101680278778076, 0.5478075742721558, 0.5472566485404968, 0.32145917415618896, -0.17351791262626648, -0.11064151674509048, 0.03344130888581276, 0.032779306173324585, 0.19032762944698334, 0.733572781085968, 0.3221980333328247, -0.27110177278518677, 0.3745807409286499, -0.059075117111206055, -0.05481623113155365, 0.41394397616386414, 0.12746432423591614, 0.015450948849320412, -0.06342163681983948, 0.5102359056472778, -0.1708884984254837, 0.10250923782587051, 0.006851306185126305, 0.02524959295988083, -0.2746659219264984, -0.3511103689670563, 0.2094680368900299, -0.3275744616985321, 0.2222515046596527, -0.030705541372299194, 0.018956240266561508, -0.25469717383384705, 0.5723084807395935, 0.4384695589542389, -0.14093264937400818, -0.21884925663471222, -0.010653063654899597, -0.4845343232154846, 0.35602089762687683, -0.3736163079738617, -0.16402584314346313, -0.13282263278961182, 0.5531368851661682, -0.2862464487552643, 0.23929312825202942, -0.2779918909072876, 0.38563624024391174, -0.17468668520450592, 0.16580131649971008, -0.25717684626579285, -0.013567924499511719, -0.10855506360530853, -0.22183682024478912, 0.006634432822465897, -0.24003097414970398, 0.14324951171875, 0.042736832052469254, -0.12452109158039093, 0.03154309093952179, -0.024334553629159927, 0.06714528799057007, 0.4972288906574249, -0.0039021968841552734, 0.47871333360671997, 0.4656328856945038, -0.1569477766752243, -0.23094283044338226, -0.16266821324825287, -0.3780435025691986, -0.34398841857910156, 0.3681851625442505, 0.17288753390312195, 0.4330157935619354, -0.17476186156272888, 0.5583014488220215, -0.04618224874138832, 0.27481594681739807, -0.04320555180311203, -0.0047761425375938416, -0.3202861547470093, 0.012487802654504776, -0.2711293697357178, 0.025067638605833054, 0.12636426091194153, 0.3703688085079193, 0.1730210781097412, -0.02074553444981575, -0.1875908523797989, -0.4143213629722595, 0.430174857378006, -0.35417822003364563, 0.05320895090699196, -0.13932479918003082, 0.15484215319156647, 0.2172609567642212, -0.14222191274166107, -0.7538331151008606, -0.08064009994268417, 0.2726384699344635, -0.16095545887947083, 0.05061447247862816, -0.013103971257805824, -0.014782052487134933, -0.27094611525535583, -0.1554519087076187, -0.25527331233024597, 0.24762418866157532, -0.37345200777053833, 0.30728623270988464, -0.09805581718683243 ]
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.
[ -0.17601029574871063, 0.1923229992389679, 0.06580723822116852, 0.415630578994751, 0.012082453817129135, 0.19369161128997803, -0.060682814568281174, 0.3878972828388214, -0.05917363613843918, 0.13248199224472046, -0.07451555132865906, 0.1716667264699936, 0.00934138149023056, -0.13655923306941986, -0.07436807453632355, 0.00014028698205947876, -0.061030514538288116, 0.14836759865283966, 0.07987777143716812, 0.035928815603256226, -0.06956406682729721, 0.4301077127456665, -0.3462488055229187, -0.08884064853191376, -0.23781460523605347, -0.11275551468133926, 0.02941947616636753, 0.3264774680137634, 0.04625571146607399, -0.3628297746181488, 0.2542961537837982, -0.06689697504043579, 0.12759649753570557, 0.30131080746650696, -0.00012377803795970976, 0.01873653382062912, 0.6639291048049927, -0.24313701689243317, -0.19303658604621887, -0.38231271505355835, -0.2983901798725128, -0.27950337529182434, -0.031189940869808197, -0.23511448502540588, 0.00004652142524719238, -0.045577239245176315, 0.1376456320285797, -0.17305010557174683, 0.22491493821144104, 0.3440399467945099, 0.0974617749452591, 0.18349336087703705, -0.09255968779325485, 0.11123840510845184, 0.21578535437583923, 0.12078313529491425, -0.048925578594207764, -0.014452159404754639, -0.017779268324375153, -0.15842965245246887, -0.03852691501379013, 0.3862314820289612, -0.3328869342803955, 0.11739915609359741, 0.25707224011421204, 0.14909526705741882, 0.3444841206073761, -0.19640594720840454, 0.03148144111037254, 0.38532519340515137, 0.45523858070373535, 0.020421288907527924, -0.3434726297855377, -0.4385927617549896, -0.0033315056934952736, 0.09549721330404282, 0.385768860578537, 0.3656495213508606, -0.2501451373100281, 0.04120176285505295, -0.34488174319267273, 0.020254844799637794, -0.12321119010448456, 0.10563604533672333, 0.1657620221376419, 0.08337247371673584, 0.15274578332901, -0.013326473534107208, 0.040776655077934265, -0.03611783683300018, 0.14461666345596313, -0.2656342089176178, -0.1059991717338562, 0.09341879934072495, -0.16324205696582794, -0.12530268728733063, -0.18923768401145935, -0.011101484298706055, 0.16719725728034973, 0.1999482661485672, 0.29252347350120544, 0.010139748454093933, 0.19105952978134155, 0.18825817108154297, 0.469216525554657, -0.05503105744719505, -0.10778697580099106, 0.5959746241569519, 0.21419714391231537, 0.08869078755378723, -0.005386956036090851, -0.031112052500247955, 0.2485344111919403, -0.2011575847864151, 0.02705814316868782, 0.12978479266166687, 0.24780960381031036, -0.5492243766784668, -0.5105422735214233, 0.178507998585701, -0.47248125076293945, -0.1294342279434204, -0.03337736055254936, 0.2145405113697052, -0.39336279034614563, 0.32460302114486694, -0.06497441232204437, 0.12263622879981995, -0.3191176950931549, -0.0042855278588831425, -0.15956468880176544, 0.1274050921201706, -0.13459908962249756, 0.16218197345733643, 0.2891721725463867, -0.17284761369228363, 0.4029974639415741, -0.1103069856762886, 0.06361111253499985, -0.21592643857002258, 0.3070898950099945, -0.2068900465965271, 0.065682053565979, 0.2388724982738495, 0.047185469418764114, 0.2478986233472824, 0.0007344074547290802, -0.20997636020183563, -0.3573243319988251, 0.13655634224414825, -0.2047910988330841, -0.4814677834510803, 0.07134076952934265, 0.12073985487222672, -0.3954429626464844, -0.04784272611141205, -0.07426493614912033, 0.036178454756736755, 0.3651830554008484, -0.4267297089099884, -0.09594767540693283, -0.22851373255252838, -0.07225896418094635, -0.035934239625930786, -0.10529256612062454, 0.5175377726554871, 0.15426748991012573, -0.16895271837711334, 0.1386125534772873, 0.05945858731865883, 0.5408574342727661, 0.391432523727417, 0.05832092463970184, 0.025509662926197052, -0.19143261015415192, 0.37669724225997925, 0.20354309678077698, -0.5307165384292603, -0.2367611825466156, 0.5350231528282166, 0.13700176775455475, 0.36929187178611755, 0.22536331415176392, 0.14809520542621613, 0.26581189036369324, -0.06201304867863655, 0.016567617654800415, 0.4239158630371094, -0.04632773995399475, 0.07779435813426971, -0.37284496426582336, -0.29308465123176575, 0.3921940326690674, 0.14195109903812408, 0.1503073275089264, 0.04352620244026184, -0.02133478969335556, 0.44595351815223694, 0.3237895369529724, -0.0458550788462162, 0.09353777021169662, 0.2675280272960663, -0.28064805269241333, 0.25819653272628784, 0.049429990351200104, -0.22007182240486145, -0.3241133987903595, 0.027645796537399292, -0.1992626041173935, 0.299434632062912, -0.09630544483661652, -0.2697671949863434, -0.17187514901161194, -0.4570290148258209, -0.042080823332071304, -0.00748959556221962, -0.006650581955909729, 0.20860198140144348, 0.28892526030540466, 0.1307036578655243, -0.06835608929395676, 0.4250880181789398, -0.09385355561971664, 0.2411915510892868, -0.8024852871894836, 0.24098919332027435, -0.19558899104595184, -0.31735971570014954, 0.4423464238643646, 0.2632880210876465, 0.19873633980751038, -0.14953134953975677, -0.14108476042747498, 0.4424898028373718, -0.031132789328694344, 0.4205712080001831, -0.3902108371257782, 0.07099071890115738, 0.32512369751930237, -0.3283196687698364, -0.013647114858031273, 0.37926772236824036, 0.06837370246648788, -0.09130765497684479, -0.09971357136964798, 0.3737080693244934, -0.3763323724269867, 0.21936804056167603, 0.1308305412530899, 0.029671097174286842, 0.13359251618385315, -0.21173621714115143, -0.2728499174118042, -0.529153048992157, 0.5458309650421143, -0.06431084871292114, -0.3128311336040497, 0.004712603986263275, -0.23172332346439362, -0.06337665766477585, 0.41675665974617004, 0.039075907319784164, -0.07154937833547592, -0.016353987157344818, -0.005610503256320953, -0.09737547487020493, 0.0756484791636467, 0.20833925902843475, 0.44202858209609985, 0.09718981385231018, 0.013755030930042267, 0.07849985361099243, -0.20874279737472534, -0.3179134726524353, 0.16330751776695251, 0.01875659078359604, 0.0033224241342395544, 0.4489210247993469, 0.013933427631855011, 0.0849306508898735, -0.2789892554283142, -0.0302160382270813, 0.11369865387678146, 0.18893904983997345, -0.42710933089256287, -0.09499017149209976, -0.4230707287788391, -0.38129299879074097, -0.3252098858356476, -0.19172726571559906, -0.15214784443378448, -0.3741565942764282, 0.10221859812736511, 0.33567163348197937, -0.09335123002529144, 0.15202751755714417, -0.27478036284446716, -0.07710932195186615, -0.28563809394836426, -0.16717529296875, -0.04007302224636078, -0.25538623332977295, -0.222213476896286, -0.06030028685927391, 0.2851209044456482, 0.41625288128852844, 0.21915368735790253, -0.4314367473125458, -0.1876332312822342, -0.003962108865380287, -0.21282917261123657, 0.047138720750808716, -0.09212160110473633, -0.010913737118244171, 0.12462525069713593, -0.09668281674385071, 0.35306113958358765, -0.3084551692008972, 0.13665345311164856, -0.18659070134162903, -0.28652623295783997, 0.08701904863119125, 0.16549672186374664, 0.005901121534407139, -0.16749265789985657, -0.4755880534648895, -0.2604392468929291, -0.35200002789497375, -0.15064652264118195, 0.2697538137435913, 0.1826927661895752, 0.09125352650880814, -0.12194868177175522, -0.10851744562387466, -0.03932376578450203, 0.3502366244792938, -0.214968740940094, -0.12616385519504547, 0.2993685305118561, -0.1570669412612915, -0.17148832976818085, 0.006665978580713272, -0.019033823162317276, -0.028024282306432724, 0.12993308901786804, -0.5250759720802307, -0.12667787075042725, -0.2553532123565674, -0.01944783329963684, -0.2235790491104126, -0.21697384119033813, 0.4017179310321808, -0.019970007240772247, 0.007223976776003838, -0.007094044238328934, -0.3523035943508148, 0.027413323521614075, 0.053191713988780975, 0.38160037994384766, -0.13469213247299194, 0.30944913625717163, -0.03444170206785202, 0.4264928102493286, 0.5960376262664795, 0.0360533632338047, 0.02764013409614563, 0.07134276628494263, 0.14365211129188538, 0.07474596798419952, -0.10545020550489426, 0.26571565866470337, -0.07296621799468994, -0.07890313863754272, 0.2505122423171997, 0.008917722851037979, -0.22383230924606323, -0.17834460735321045, -0.12964656949043274, -0.45140305161476135, -0.27320659160614014, 0.10059580206871033, -0.3199053704738617, 0.17668426036834717, 0.13752348721027374, 0.2814924418926239, -0.055178675800561905, -0.5646655559539795, 0.029398571699857712, 0.41505634784698486, 0.21458125114440918, 0.11504055559635162, -0.33853164315223694, 0.1420566290616989, -0.3411736488342285, 0.26438724994659424, 0.20610526204109192, 0.37408316135406494, -0.1111692488193512, -0.11376402527093887, 0.03909432142972946, -0.19258281588554382, 0.3788115382194519, -0.4746643006801605, 0.05551697686314583, -0.00408543087542057, -0.07041024416685104, -0.4118961989879608, -0.21917130053043365, -0.18260174989700317, 0.3691921532154083, 0.4657864570617676, 0.4691724479198456, -0.35209381580352783, -0.061966490000486374, 0.2556946873664856, 0.021391037851572037, -0.1011052131652832, -0.174580916762352, -0.22858764231204987, -0.3355112373828888, -0.2618897557258606, -0.17329686880111694, -0.04387642443180084, 0.31121692061424255, -0.07534471899271011, -0.08241952210664749, 0.04225067049264908, 0.16651961207389832, 0.425326406955719, 0.23124046623706818, 0.20768505334854126, -0.03240104764699936, 0.22409969568252563, 0.25176823139190674, 0.40947458148002625, 0.2128596007823944, 0.5298258066177368, -0.0027628839015960693, -0.19281616806983948, -0.07655413448810577, -0.30114465951919556, 0.005492262542247772, 0.43741244077682495, 0.2044098973274231, -0.030449964106082916, 0.12637130916118622, -0.13738712668418884, -0.42967453598976135, -0.05294419452548027, 0.1266050636768341, -0.1013169214129448, -0.38544777035713196, -0.49731114506721497, 0.29982176423072815, 0.3504994809627533, -0.11287234723567963, -0.14339759945869446, -0.09955067932605743, -0.3347935676574707, 0.34836745262145996, 0.1721331775188446, 0.9992426037788391, 0.22708752751350403, -0.02504497393965721, -0.06836072355508804, -0.2009786069393158, 0.7844555974006653, -0.21298271417617798, 0.3598969578742981, -0.4350261390209198, -0.2963215112686157, -0.10601989924907684, -0.18125872313976288, 0.19104528427124023, 0.10323550552129745, -0.33262690901756287, 0.3883803188800812, 0.10925772786140442, -0.01733897253870964, -0.20386344194412231, 0.4057892858982086, -0.28847673535346985, -0.1145377904176712, -0.15640376508235931, 0.020807690918445587, -0.1877886950969696, 0.43178772926330566, 0.08245304971933365, -0.05726049095392227, -0.2624441087245941, -0.16835300624370575, -0.5494994521141052, 0.22166454792022705, 0.14272379875183105, 0.02637326717376709, 0.21180008351802826, -0.22485177218914032, -0.013590821996331215, 0.13423489034175873, 0.32788556814193726, -0.11514613032341003, -0.10681673884391785, -0.0016026571393013, -0.022615842521190643, -0.15573088824748993, 0.20076419413089752, 0.057710736989974976, 0.4340367317199707, -0.10155399143695831, -0.061763565987348557, 0.10912123322486877, -0.07918648421764374, -0.2611497640609741, -0.1338934302330017, -0.13959364593029022, -0.18021485209465027, -0.21085694432258606, -0.2031717300415039, 0.014056168496608734, -0.029257072135806084, -0.09731970727443695, 0.017006656154990196, -0.059214264154434204, -0.19780589640140533, 0.25401172041893005, -0.09184752404689789, -0.3513319194316864, -0.09548133611679077, 0.4479711949825287, 0.2841583788394928, -0.04742593690752983, 0.4655079245567322, -0.044768765568733215, -0.11638534069061279, -0.142870232462883, -0.07991980016231537, -0.22824016213417053, -0.3608661890029907, 0.1705559492111206, -0.25035595893859863, 0.0032198280096054077, 0.21465757489204407, 0.27662038803100586, 0.23138077557086945, 0.2816075384616852, -0.09972803294658661, -0.5418651700019836, -0.005885621532797813, 0.16270053386688232, 0.09235391020774841, 0.1533018797636032, -0.19041401147842407, 0.3851987421512604, -0.2032112181186676, 0.0885583907365799, -0.18876288831233978, 0.0016788651701062918, -0.4216451048851013, 0.3190789818763733, 0.0986839234828949, -0.010334260761737823, -0.10648006200790405, -0.05362049490213394, -0.10954704880714417, 0.23286375403404236, -0.008623586967587471, -0.0021029282361268997, -0.1735309362411499, 0.19913391768932343, 0.07818430662155151, -0.1462133526802063, -0.12777991592884064, 0.02929745241999626, 0.07194782793521881, -0.039905861020088196, -0.03288396820425987, 0.3182600736618042, 0.016946248710155487, 0.43039947748184204, -0.2967921495437622, 0.007057413458824158, -0.05222383886575699, 0.25376251339912415, -0.37408027052879333, 0.32040339708328247, -0.06483311206102371, 0.23397673666477203, -0.0029911978635936975, 0.07607707381248474, 0.005250210408121347, -0.01093979924917221, -0.03330609202384949, 0.04425101727247238, 0.26031824946403503, -0.18230938911437988, 0.15160910785198212, 0.3269649147987366, 0.0630289614200592, 0.66217440366745, -0.0977916419506073, -0.14357538521289825, 0.10021911561489105, 0.08938557654619217, -0.12129510194063187, 0.0176846906542778, 0.2556929588317871, -0.06330009549856186, -0.08007554709911346, 0.3751468062400818, 0.16038791835308075, -0.11737298965454102, -0.2403324693441391, 0.09514571726322174, 0.5035054087638855, -0.13529187440872192, 0.18721407651901245, 0.2672545909881592, -0.05208103731274605, -0.03511747345328331, 0.11959374696016312, -0.21357037127017975, 0.11305534839630127, 0.08282115310430527, -0.016397520899772644, 0.4081759452819824, 0.05979757383465767, 0.3810717761516571, -0.1107976958155632, -0.19281329214572906, -0.016884218901395798, 0.5765299201011658, 0.19954770803451538, 0.4437112510204315, 0.23778961598873138, 0.34543317556381226, -0.33760300278663635, 0.025902289897203445, -0.3117019236087799, -0.1353357583284378, -0.013288278132677078, -0.17353425920009613, -0.44935426115989685, -0.13270974159240723, -0.22633275389671326, 0.11731503903865814, -0.13277365267276764, 0.16289067268371582, 0.32127314805984497, 0.033501870930194855, -0.20877082645893097, -0.512785017490387, 0.06974249333143234, -0.11926928907632828, 0.22779160737991333, -0.2851017415523529, 0.11935082077980042, 0.12183379381895065, -0.06800270080566406, 0.38052862882614136, 0.5955421328544617, 0.6343860626220703, 0.3927338123321533, -0.19025832414627075, -0.0560276098549366, 0.0363975428044796, 0.05950076878070831, 0.20654883980751038, 0.7163812518119812, 0.2461318075656891, -0.220201775431633, 0.3402116894721985, -0.033732496201992035, -0.06545168161392212, 0.37195467948913574, 0.07000861316919327, 0.0255957692861557, -0.08627039194107056, 0.5236676335334778, -0.19162189960479736, 0.11693494766950607, -0.05106053501367569, 0.0679503083229065, -0.32060736417770386, -0.323677659034729, 0.26137760281562805, -0.26411721110343933, 0.24652719497680664, -0.0030775396153330803, 0.015305664390325546, -0.2047778218984604, 0.5182005167007446, 0.5190638303756714, -0.09691687673330307, -0.24339550733566284, -0.01123107224702835, -0.5534390807151794, 0.3506774604320526, -0.3848564028739929, -0.21972526609897614, -0.14044570922851562, 0.5029476284980774, -0.2913248836994171, 0.271891713142395, -0.16698040068149567, 0.4217168986797333, -0.1999897062778473, 0.23926503956317902, -0.26129385828971863, -0.04820116236805916, -0.08566737174987793, -0.24912545084953308, -0.041208043694496155, -0.23633703589439392, 0.12473779916763306, 0.0006039002910256386, -0.09424837678670883, 0.030080754309892654, -0.026399530470371246, 0.06136685982346535, 0.49497443437576294, 0.024868406355381012, 0.4822665750980377, 0.5218864679336548, -0.1439880132675171, -0.2633267641067505, -0.17555569112300873, -0.33674079179763794, -0.3082938492298126, 0.43964216113090515, 0.1312471628189087, 0.37442272901535034, -0.1879602074623108, 0.4798943102359772, -0.13751064240932465, 0.33619341254234314, -0.025599170476198196, -0.1167680025100708, -0.3755322992801666, -0.05126340687274933, -0.22220423817634583, -0.03113829344511032, 0.09889931231737137, 0.42117664217948914, 0.15882252156734467, 0.08846063911914825, -0.17987768352031708, -0.4090624451637268, 0.4370060861110687, -0.327399343252182, 0.043092552572488785, -0.11276359856128693, 0.12388109415769577, 0.14992263913154602, -0.03915152698755264, -0.7592726349830627, -0.035973746329545975, 0.3005138337612152, -0.177036315202713, -0.02731669321656227, -0.013707553967833519, -0.019451463595032692, -0.22175200283527374, -0.11064097285270691, -0.30953407287597656, 0.22752606868743896, -0.3596225380897522, 0.31355851888656616, -0.08145659416913986 ]
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')
[ -0.05952122062444687, 0.15316832065582275, 0.0704789012670517, 0.38789206743240356, -0.015147943049669266, 0.1419079601764679, -0.09238578379154205, 0.32008153200149536, -0.04935669153928757, 0.008110485970973969, -0.08665413409471512, 0.15768729150295258, 0.036548759788274765, 0.0461474284529686, -0.019717078655958176, 0.08979116380214691, -0.04565099626779556, 0.20413710176944733, 0.04544687271118164, 0.029078975319862366, 0.000983554869890213, 0.5112385153770447, -0.24448390305042267, -0.06453905999660492, -0.3506300449371338, -0.14984454214572906, 0.05441678315401077, 0.3629710376262665, 0.010663431137800217, -0.3910157382488251, 0.319717675447464, -0.04728415235877037, 0.1141887754201889, 0.24019652605056763, -0.00012255208275746554, -0.002842336893081665, 0.4699680805206299, -0.21914023160934448, -0.2050269991159439, -0.2703421115875244, -0.21681585907936096, -0.2991623282432556, -0.04636288434267044, -0.21825683116912842, -0.06674200296401978, -0.06653846800327301, 0.07996359467506409, -0.05199211835861206, 0.3719485402107239, 0.29745855927467346, 0.125547394156456, 0.09593860059976578, -0.12729518115520477, 0.10268118977546692, 0.08485596626996994, 0.06652462482452393, -0.011884946376085281, -0.04185522720217705, -0.17200255393981934, -0.321993887424469, -0.12819787859916687, 0.5180577635765076, -0.19157950580120087, 0.12624672055244446, 0.18193864822387695, 0.0800839439034462, 0.2830647826194763, -0.11941677331924438, 0.06791115552186966, 0.3065376579761505, 0.4064788818359375, 0.10857941955327988, -0.2249048352241516, -0.5339201092720032, 0.007108952384442091, 0.05136070400476456, 0.45280009508132935, 0.34143397212028503, -0.197662353515625, 0.10054163634777069, -0.41586562991142273, -0.22632411122322083, -0.054598119109869, 0.007278576493263245, 0.2139008343219757, 0.20679163932800293, 0.03881848603487015, -0.010091401636600494, 0.1374928504228592, -0.037883661687374115, 0.22440919280052185, -0.09224051982164383, -0.14465267956256866, 0.08490939438343048, -0.21261771023273468, -0.15949317812919617, -0.12389127165079117, 0.03781522437930107, 0.1740897297859192, 0.153521329164505, 0.2372102290391922, -0.07823164016008377, 0.14463846385478973, 0.04192858189344406, 0.369392067193985, 0.17495565116405487, -0.001896597445011139, 0.42824146151542664, 0.14054054021835327, 0.033288851380348206, 0.11029240489006042, 0.003799423575401306, 0.2939819395542145, -0.15723007917404175, -0.0670633316040039, 0.045196548104286194, 0.20995980501174927, -0.543628990650177, -0.4165376126766205, 0.07986783981323242, -0.49174922704696655, -0.20235203206539154, -0.12295100837945938, 0.13204741477966309, -0.34991854429244995, 0.35234397649765015, -0.07884305715560913, 0.03780720755457878, -0.3732774257659912, -0.055995743721723557, -0.2014298439025879, 0.10900092124938965, -0.19364409148693085, 0.2363349348306656, 0.2714063227176666, -0.28212133049964905, 0.43705230951309204, -0.051476895809173584, -0.09336414933204651, -0.2561876177787781, 0.2646717131137848, -0.030780835077166557, -0.02329244464635849, 0.1904066503047943, -0.13137565553188324, 0.23993603885173798, 0.00858066976070404, -0.20084376633167267, -0.2873159646987915, 0.01637335494160652, -0.27687519788742065, -0.6388430595397949, 0.08123864233493805, 0.11715024709701538, -0.40461617708206177, -0.05964156240224838, 0.01749790459871292, 0.22032073140144348, 0.3705641031265259, -0.2604248821735382, -0.10591349750757217, -0.16875965893268585, -0.10133935511112213, -0.02985317073762417, -0.16532771289348602, 0.5042769312858582, 0.13272762298583984, -0.32241421937942505, 0.16070279479026794, -0.03138468414545059, 0.47419610619544983, 0.4351547956466675, -0.006501771509647369, 0.10734213143587112, -0.17257268726825714, 0.3553568720817566, 0.14280344545841217, -0.3949947953224182, -0.2550790309906006, 0.4000755250453949, -0.0010334476828575134, 0.2915744483470917, 0.18328091502189636, 0.13108772039413452, 0.09201548248529434, -0.0645950585603714, 0.05912280082702637, 0.472974956035614, 0.07429322600364685, 0.0992300882935524, -0.45281127095222473, -0.4169975221157074, 0.43306535482406616, 0.16881926357746124, 0.10977046191692352, 0.07702752947807312, -0.019928500056266785, 0.4880288541316986, 0.31506645679473877, -0.14438781142234802, 0.06338492035865784, 0.31531858444213867, -0.33134567737579346, 0.15865163505077362, 0.049391377717256546, -0.08946045488119125, -0.5167126059532166, 0.007718347012996674, -0.2730106711387634, 0.31955772638320923, 0.050448331981897354, -0.19811005890369415, -0.23508009314537048, -0.4627165198326111, -0.07891352474689484, -0.1806633174419403, 0.0007563978433609009, 0.23502175509929657, 0.409033864736557, 0.05455164238810539, -0.018292728811502457, 0.5882975459098816, -0.09648598730564117, 0.2448498159646988, -0.8845168948173523, 0.2083265334367752, -0.12776288390159607, -0.1689957082271576, 0.41228818893432617, 0.2858721911907196, 0.27037715911865234, -0.1416531503200531, -0.13194647431373596, 0.46327081322669983, -0.22580701112747192, 0.4788298010826111, -0.22793686389923096, -0.0022638291120529175, 0.27312564849853516, -0.2956171929836273, 0.021120639517903328, 0.2180158793926239, -0.046564821153879166, -0.055270787328481674, -0.06619317084550858, 0.3710591197013855, -0.3695761263370514, 0.13853028416633606, 0.017573252320289612, -0.04477160423994064, 0.10005359351634979, -0.2872810363769531, -0.14379769563674927, -0.4895384907722473, 0.5787147283554077, -0.1502460539340973, -0.2525634467601776, 0.021484743803739548, -0.1338326781988144, 0.016381271183490753, 0.3793991208076477, 0.06490480899810791, -0.09646495431661606, -0.03197631984949112, -0.01605333387851715, -0.20027032494544983, 0.05359116196632385, 0.24537457525730133, 0.5202927589416504, 0.136488676071167, 0.026374775916337967, 0.09233323484659195, -0.16415837407112122, -0.33067813515663147, 0.15244366228580475, -0.0643644779920578, 0.08235223591327667, 0.36921730637550354, 0.17514680325984955, -0.002143767662346363, -0.5159105062484741, 0.1042155772447586, 0.21197280287742615, 0.24466386437416077, -0.30409908294677734, -0.13484103977680206, -0.3648941218852997, -0.38661521673202515, -0.2667771875858307, -0.1904561072587967, -0.316261887550354, -0.3300526738166809, 0.20578353106975555, 0.2855917811393738, -0.11775757372379303, 0.15927450358867645, -0.09243506193161011, 0.0607672780752182, -0.32778340578079224, 0.009110844694077969, -0.10953319072723389, -0.2758680284023285, -0.20552194118499756, -0.05036255717277527, 0.10383964329957962, 0.3221498429775238, 0.18946270644664764, -0.356062650680542, -0.19915348291397095, -0.07048306614160538, -0.33795589208602905, 0.04746336489915848, -0.08756691217422485, -0.07165685296058655, 0.08050297200679779, -0.0683811753988266, 0.3928329348564148, -0.15853354334831238, 0.14054352045059204, -0.2609803080558777, -0.3093509078025818, -0.041610538959503174, 0.1713782101869583, 0.19560803472995758, -0.16764123737812042, -0.5419104695320129, -0.2437303215265274, -0.36349084973335266, -0.015157170593738556, 0.2141256183385849, 0.1851952075958252, 0.1718037724494934, -0.1324823498725891, -0.20916199684143066, -0.11133063584566116, 0.2509463429450989, -0.3925293982028961, -0.13909894227981567, 0.19587862491607666, -0.07682523876428604, -0.13032086193561554, -0.07141520082950592, 0.10173918306827545, -0.08580437302589417, 0.07061222195625305, -0.6237038969993591, -0.2201807200908661, -0.24198295176029205, -0.09389472007751465, -0.3028300702571869, -0.17801044881343842, 0.5007232427597046, -0.04541131854057312, -0.004132919013500214, 0.08095019310712814, -0.4596954882144928, 0.01794762909412384, 0.13317060470581055, 0.48731228709220886, -0.17907088994979858, 0.2381068766117096, -0.019556019455194473, 0.5533111691474915, 0.5295947194099426, 0.053176093846559525, -0.009989134967327118, 0.16884203255176544, 0.22626477479934692, 0.040080808103084564, -0.1329493373632431, 0.3330083191394806, -0.06891165673732758, 0.037263281643390656, 0.34564948081970215, 0.10086305439472198, -0.1492968648672104, -0.1453624665737152, -0.16622045636177063, -0.36227425932884216, -0.1801701933145523, 0.13737726211547852, -0.13920143246650696, 0.12042494118213654, 0.2959353029727936, 0.14354175329208374, -0.1756933033466339, -0.4716580808162689, -0.008708752691745758, 0.24278505146503448, 0.143654003739357, 0.19786418974399567, -0.4579145610332489, 0.17113882303237915, -0.5147838592529297, 0.2657478451728821, 0.15840087831020355, 0.26466140151023865, -0.13183258473873138, -0.14703072607517242, 0.09939620643854141, -0.17282649874687195, 0.39402198791503906, -0.404949426651001, -0.07724947482347488, -0.07967567443847656, -0.016715524718165398, -0.33635619282722473, -0.17841531336307526, -0.03545089811086655, 0.3083250820636749, 0.4751313626766205, 0.4908388555049896, -0.39141008257865906, -0.15796519815921783, 0.1868763118982315, -0.009669477120041847, -0.1367054283618927, -0.13703131675720215, -0.2299746721982956, -0.3890259265899658, -0.2763265073299408, -0.11392184346914291, -0.1300937682390213, 0.3114842176437378, 0.02537458948791027, 0.01682465150952339, 0.12226434051990509, 0.0988527312874794, 0.2510795593261719, 0.1897910237312317, 0.13215506076812744, 0.025374535471200943, 0.07778167724609375, 0.3831081986427307, 0.5103031396865845, 0.03691951185464859, 0.6296100616455078, -0.010878447443246841, -0.037916332483291626, -0.08799253404140472, -0.08143585920333862, 0.02740628272294998, 0.5505662560462952, 0.20019100606441498, 0.05380767583847046, 0.13748937845230103, 0.010431796312332153, -0.4142747223377228, -0.04083792492747307, 0.11529917269945145, -0.08835116773843765, -0.427870512008667, -0.2941102385520935, 0.3169788122177124, 0.23538543283939362, -0.1339426338672638, -0.13132132589817047, -0.1356361359357834, -0.27198246121406555, 0.24533598124980927, 0.055233847349882126, 1.041927695274353, 0.207284614443779, 0.04186401516199112, -0.055812858045101166, -0.1492011696100235, 0.9380415678024292, -0.30007869005203247, 0.4176374673843384, -0.351711243391037, -0.17800401151180267, -0.05750344693660736, -0.14746718108654022, 0.03603443503379822, 0.0908762663602829, -0.1540612280368805, 0.3749524652957916, 0.28883594274520874, 0.050439029932022095, -0.12973631918430328, 0.42349618673324585, -0.24959737062454224, -0.15231391787528992, -0.3030501902103424, 0.01705753430724144, -0.30228376388549805, 0.5400670766830444, -0.03814367577433586, -0.0711965411901474, -0.08948315680027008, -0.20423360168933868, -0.37206488847732544, 0.25258558988571167, 0.024172533303499222, 0.12936845421791077, 0.23180116713047028, -0.19619151949882507, -0.07764199376106262, 0.10560862720012665, 0.4005061388015747, -0.16487471759319305, -0.12909549474716187, -0.006210897117853165, -0.1460067629814148, -0.12728801369667053, 0.11892835050821304, 0.09156027436256409, 0.28809696435928345, -0.09807290136814117, -0.03008083999156952, -0.013803645968437195, -0.14492306113243103, -0.330866277217865, -0.24841278791427612, -0.05192255973815918, -0.16080883145332336, -0.2244488149881363, -0.014996785670518875, 0.1725156307220459, -0.028158828616142273, -0.1344454139471054, 0.020831778645515442, -0.04764583706855774, -0.1695941835641861, 0.11480950564146042, -0.09480074048042297, -0.3547867238521576, -0.1866084635257721, 0.46530085802078247, 0.41192328929901123, -0.13000470399856567, 0.4889170825481415, -0.10779237747192383, -0.04715222120285034, -0.16125038266181946, -0.18598607182502747, -0.2622641324996948, -0.42547574639320374, 0.20580711960792542, -0.2105516940355301, -0.17630591988563538, 0.2077464461326599, 0.26054129004478455, 0.23746593296527863, 0.31274643540382385, -0.013969458639621735, -0.4817531108856201, -0.03925495594739914, 0.13603372871875763, -0.011188969016075134, 0.09119077026844025, -0.024539876729249954, 0.3797798752784729, -0.24098224937915802, 0.318103164434433, -0.21519368886947632, 0.039288491010665894, -0.3406312167644501, 0.3029206693172455, 0.19331982731819153, -0.13923373818397522, 0.06780177354812622, 0.012921538203954697, -0.14105576276779175, 0.26413536071777344, -0.10220294445753098, -0.017113421112298965, -0.14058807492256165, 0.19693759083747864, 0.15148618817329407, -0.1118534505367279, -0.1557074338197708, 0.07982834428548813, 0.06451982259750366, 0.01601596549153328, 0.06329783797264099, 0.2837775647640228, 0.06752502173185349, 0.3050289452075958, -0.28709176182746887, -0.03953491151332855, 0.0005170963704586029, 0.19909065961837769, -0.24377396702766418, 0.22092701494693756, -0.039350274950265884, 0.21276786923408508, -0.16374057531356812, 0.05936986207962036, -0.09963305294513702, 0.0008988901972770691, 0.08639159053564072, 0.08966079354286194, 0.08942975103855133, -0.25115662813186646, 0.034797344356775284, 0.2602270543575287, 0.05379399657249451, 0.5116792917251587, -0.08315525203943253, -0.15430910885334015, 0.1248418316245079, 0.08711467683315277, -0.14101755619049072, -0.06204986944794655, 0.3102426528930664, -0.07737154513597488, -0.09637980163097382, 0.3810321092605591, 0.17304816842079163, -0.12971416115760803, -0.14074625074863434, 0.10686630755662918, 0.45320674777030945, -0.19162863492965698, 0.15098324418067932, 0.1268135905265808, -0.10557353496551514, -0.08674150705337524, 0.37435150146484375, -0.20107817649841309, 0.12214745581150055, 0.11047141253948212, 0.04505413770675659, 0.3720736503601074, 0.05472875386476517, 0.45312851667404175, -0.1651650071144104, -0.15896613895893097, -0.03961111977696419, 0.6585904955863953, 0.18797153234481812, 0.43030044436454773, 0.14218129217624664, 0.5052541494369507, -0.3357342779636383, -0.01521608792245388, -0.3209925889968872, 0.021360773593187332, -0.092747263610363, -0.16650784015655518, -0.30164259672164917, -0.11755897104740143, -0.10489378124475479, 0.15447767078876495, -0.15096940100193024, 0.055045656859874725, 0.4208003580570221, -0.06759683787822723, -0.29463064670562744, -0.5726760625839233, -0.003766532987356186, 0.05557895451784134, 0.27704954147338867, -0.2272971123456955, 0.2844380736351013, 0.0758509635925293, -0.20439988374710083, 0.319293349981308, 0.6775821447372437, 0.5208287835121155, 0.2321733683347702, -0.032789599150419235, -0.07751831412315369, 0.07033324241638184, 0.1269388347864151, 0.2107742726802826, 0.745038628578186, 0.29596900939941406, -0.31335198879241943, 0.2754243314266205, -0.038033515214920044, -0.11196877807378769, 0.3472534716129303, 0.025003623217344284, 0.07327782362699509, -0.03658036142587662, 0.38708585500717163, -0.1618291139602661, 0.08336413651704788, -0.06547261029481888, 0.09793051332235336, -0.3249247968196869, -0.2109721601009369, 0.16343875229358673, -0.17252646386623383, 0.30173978209495544, -0.08095744252204895, 0.0441569909453392, -0.2473849654197693, 0.5227413773536682, 0.5025137662887573, -0.3186624348163605, -0.15052732825279236, -0.026378393173217773, -0.526768684387207, 0.2962951958179474, -0.3916666507720947, -0.08135929703712463, -0.20274899899959564, 0.44821152091026306, -0.1969209611415863, 0.2801700532436371, -0.17557315528392792, 0.4149181544780731, -0.07420146465301514, -0.018937617540359497, -0.23684778809547424, 0.0765000432729721, -0.03134014457464218, -0.19340305030345917, 0.07157544791698456, -0.24616384506225586, 0.21737593412399292, 0.1577954888343811, -0.10438892245292664, 0.03160778433084488, 0.10152055323123932, -0.016133174300193787, 0.3900722563266754, -0.07530251145362854, 0.5313584804534912, 0.29228490591049194, -0.03974778950214386, -0.28395164012908936, -0.1452285498380661, -0.40259096026420593, -0.31096333265304565, 0.34221339225769043, 0.20279137790203094, 0.4863954484462738, -0.22398029267787933, 0.47227540612220764, -0.16622617840766907, 0.31102919578552246, 0.02751193940639496, -0.061902184039354324, -0.31948888301849365, -0.02634221501648426, -0.1740514636039734, 0.04495802894234657, 0.2053137868642807, 0.5501672625541687, 0.14749205112457275, 0.052830278873443604, -0.26446664333343506, -0.41413313150405884, 0.38899165391921997, -0.38547787070274353, 0.003921061754226685, 0.07362986356019974, 0.23610728979110718, 0.2745177447795868, -0.2053195834159851, -0.7636072635650635, 0.04344826191663742, 0.22158850729465485, -0.13191285729408264, -0.00814615748822689, 0.06734813749790192, -0.1885165274143219, -0.2726271450519562, -0.014860641211271286, -0.17716476321220398, 0.36956971883773804, -0.4083258807659149, 0.41735175251960754, -0.019796527922153473 ]
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)
[ -0.04786325991153717, 0.2781369984149933, 0.07765756547451019, 0.39152562618255615, -0.1035548746585846, 0.036403588950634, -0.08268314599990845, 0.24596023559570312, -0.08353319764137268, 0.04618596285581589, 0.08917642384767532, 0.17450417578220367, 0.04060813784599304, -0.15940970182418823, 0.01790878176689148, 0.05279986560344696, -0.06446699053049088, 0.1248314157128334, 0.10597938299179077, 0.042897261679172516, -0.010564025491476059, 0.39612406492233276, -0.19377824664115906, -0.040419064462184906, -0.30533477663993835, -0.23367762565612793, 0.029022345319390297, 0.23348161578178406, 0.04022801294922829, -0.4769929349422455, 0.24664773046970367, 0.05662955716252327, 0.10053175687789917, 0.16701549291610718, -0.00012176283780718222, -0.06326837837696075, 0.538947582244873, -0.19484302401542664, -0.25188663601875305, -0.029789388179779053, -0.383603036403656, -0.27975690364837646, 0.0038727768696844578, -0.28427672386169434, 0.01517852395772934, 0.023537248373031616, 0.17921487987041473, -0.1652388870716095, 0.22913643717765808, 0.30275896191596985, 0.12136118113994598, 0.10980746150016785, -0.18810929358005524, 0.16789482533931732, 0.17320461571216583, -0.027630262076854706, -0.06475889682769775, -0.15657681226730347, -0.1117100715637207, -0.28193017840385437, -0.24037232995033264, 0.4598936438560486, -0.23572421073913574, 0.1500595211982727, 0.26409223675727844, 0.08830877393484116, 0.32167696952819824, -0.08726488053798676, 0.013093648478388786, 0.39029115438461304, 0.5862076878547668, -0.04683353751897812, -0.28496623039245605, -0.4985653758049011, -0.12967389822006226, 0.028640015050768852, 0.5057663321495056, 0.27955061197280884, -0.15703776478767395, 0.05370159074664116, -0.46707478165626526, -0.17777815461158752, -0.05261141061782837, -0.0008108243346214294, 0.31687140464782715, 0.19807469844818115, 0.1256277710199356, -0.05163481831550598, 0.030755802989006042, -0.010176576673984528, 0.2684495449066162, -0.18169060349464417, -0.12499339878559113, 0.08544934540987015, -0.23216663300991058, -0.2051638960838318, -0.02672562748193741, 0.09116505086421967, 0.25675058364868164, 0.2174561768770218, 0.40052518248558044, 0.005886361002922058, 0.21397137641906738, 0.12421661615371704, 0.33585673570632935, 0.2084193080663681, -0.014871053397655487, 0.3932429850101471, 0.15307927131652832, -0.0024872946087270975, 0.13732904195785522, 0.05757558345794678, 0.3515794277191162, -0.20486615598201752, -0.011022193357348442, 0.20190489292144775, 0.15973474085330963, -0.4877755641937256, -0.3507683575153351, 0.100728340446949, -0.39977967739105225, -0.26050662994384766, -0.05867183580994606, 0.24248537421226501, -0.3720568120479584, 0.3226299285888672, -0.26906415820121765, -0.03352269530296326, -0.4499554932117462, -0.12790538370609283, -0.23330028355121613, 0.17981576919555664, -0.14685563743114471, 0.3382387161254883, 0.23913629353046417, -0.07944642007350922, 0.3943147361278534, -0.16760507225990295, -0.09943628311157227, -0.13192248344421387, 0.35655441880226135, -0.1095731258392334, 0.08933661878108978, 0.20857898890972137, -0.09280383586883545, 0.15524990856647491, -0.04847044125199318, -0.09950070828199387, -0.48516055941581726, 0.041018035262823105, -0.32748961448669434, -0.6717569828033447, 0.09786207228899002, 0.11474642902612686, -0.372028112411499, -0.0490010529756546, 0.1598801612854004, 0.3045908808708191, 0.5192116498947144, -0.3798099756240845, 0.0015788823366165161, -0.08771109580993652, -0.19829164445400238, -0.15863703191280365, -0.1705331802368164, 0.5231292247772217, 0.07261879742145538, -0.3068561553955078, 0.12426852434873581, 0.011717491783201694, 0.4826216399669647, 0.36976704001426697, -0.035665083676576614, 0.15336909890174866, -0.1641877144575119, 0.35587453842163086, 0.17119933664798737, -0.4163646101951599, -0.27680954337120056, 0.3803890645503998, 0.1063767820596695, 0.274901419878006, 0.24967363476753235, 0.12377060204744339, 0.24420641362667084, -0.19297713041305542, 0.016356170177459717, 0.42986148595809937, -0.06306517869234085, 0.09338276088237762, -0.5093783736228943, -0.41878780722618103, 0.4363306760787964, 0.11583659052848816, 0.0694446861743927, 0.036295562982559204, -0.0036746636033058167, 0.47427088022232056, 0.16346050798892975, -0.08560024201869965, 0.09851235151290894, 0.25446826219558716, -0.3477667570114136, 0.13968226313591003, 0.08394390344619751, -0.07341695576906204, -0.4933943748474121, -0.029929041862487793, -0.2795158624649048, 0.17815431952476501, 0.1481744945049286, -0.32019758224487305, -0.0982879176735878, -0.4774523079395294, -0.09374035894870758, -0.13652102649211884, 0.029480785131454468, 0.2034239023923874, 0.5127838253974915, 0.08227682113647461, 0.019414391368627548, 0.5455953478813171, -0.006968209519982338, 0.16135884821414948, -0.8287147283554077, 0.06388167291879654, -0.1368926465511322, -0.18384507298469543, 0.36261147260665894, 0.35950547456741333, 0.21638250350952148, -0.05094456672668457, -0.09464149922132492, 0.5340720415115356, -0.141417995095253, 0.4370478391647339, -0.22054386138916016, 0.15210270881652832, 0.24981924891471863, -0.2066890150308609, 0.1167721226811409, 0.21492500603199005, -0.13724833726882935, -0.0022523291409015656, -0.09489545971155167, 0.3331185579299927, -0.3746822774410248, 0.08620357513427734, 0.0674547478556633, -0.10661555081605911, 0.05747620761394501, -0.2557361125946045, -0.27272549271583557, -0.4599677324295044, 0.541471540927887, -0.05310304835438728, -0.23666363954544067, 0.013625881634652615, -0.11991279572248459, 0.10136975347995758, 0.4620301425457001, 0.10436475276947021, -0.015318512916564941, -0.06299625337123871, 0.09480446577072144, -0.2510460615158081, 0.20877355337142944, 0.21671460568904877, 0.4686732888221741, 0.15137289464473724, 0.07372000068426132, 0.019000276923179626, -0.18847380578517914, -0.33689552545547485, 0.10088314861059189, -0.01621163636445999, -0.024240173399448395, 0.27599823474884033, 0.011330806650221348, -0.001432274468243122, -0.3033420741558075, 0.0221082866191864, 0.19120846688747406, 0.1761496663093567, -0.33251288533210754, -0.061967164278030396, -0.3928716778755188, -0.3234999477863312, -0.24957427382469177, -0.1301068663597107, -0.20251011848449707, -0.3588606119155884, 0.3024477958679199, 0.16726481914520264, -0.25919559597969055, 0.1308690905570984, -0.25159937143325806, 0.18590371310710907, -0.2763071060180664, -0.10526414960622787, -0.07306021451950073, -0.27786558866500854, -0.2761032283306122, -0.06377295404672623, 0.11679285764694214, 0.3153984844684601, 0.3178683817386627, -0.4152183532714844, -0.21306505799293518, 0.04067284241318703, -0.3145899772644043, 0.06718403100967407, 0.05993776023387909, -0.10611526668071747, 0.10012204945087433, -0.21414674818515778, 0.3740346431732178, -0.22101722657680511, -0.055854637175798416, -0.22112038731575012, -0.25211021304130554, -0.06747736036777496, 0.20162172615528107, 0.24550239741802216, -0.19950510561466217, -0.5588045120239258, -0.32585474848747253, -0.3663262724876404, -0.057020142674446106, 0.10900698602199554, 0.29956427216529846, 0.0821918398141861, -0.21076981723308563, -0.21016119420528412, -0.18786519765853882, 0.31254082918167114, -0.37954747676849365, -0.16327399015426636, 0.35842156410217285, -0.1916780322790146, -0.07630971074104309, 0.04411069676280022, -0.042466700077056885, -0.07117582857608795, 0.12719348073005676, -0.5488820672035217, -0.1745838224887848, -0.1155957281589508, -0.04159938916563988, -0.21918216347694397, -0.2206537425518036, 0.6390057802200317, -0.044646356254816055, 0.04780697077512741, 0.0388738289475441, -0.36135271191596985, 0.07012835144996643, 0.03474840894341469, 0.5672551393508911, -0.041620321571826935, 0.23606443405151367, 0.06656758487224579, 0.5585212707519531, 0.4746897220611572, 0.1441454142332077, -0.024808205664157867, 0.30200421810150146, 0.1700400710105896, 0.08519993722438812, 0.09485109150409698, 0.3581472337245941, -0.07355281710624695, -0.20108471810817719, 0.3572823405265808, 0.14081478118896484, -0.14786866307258606, -0.08919966965913773, -0.04618319496512413, -0.2848127782344818, -0.26709040999412537, 0.17356158792972565, -0.32631292939186096, 0.20595723390579224, 0.27358636260032654, 0.18179109692573547, -0.16912174224853516, -0.4797634482383728, 0.051178354769945145, 0.3250443935394287, 0.2547907829284668, 0.2598891258239746, -0.4101262390613556, 0.18531832098960876, -0.548197865486145, 0.1673709750175476, 0.12330923229455948, 0.26960012316703796, -0.1897909939289093, -0.17715877294540405, 0.0442056879401207, -0.151191845536232, 0.29357317090034485, -0.5013008713722229, -0.010341182351112366, -0.11806616187095642, -0.018266094848513603, -0.19106534123420715, -0.175484761595726, 0.017090920358896255, 0.3075025975704193, 0.4353835880756378, 0.3108360767364502, -0.47793909907341003, -0.07290231436491013, 0.11803130060434341, -0.08828970789909363, -0.14250580966472626, -0.22670376300811768, -0.1783093810081482, -0.24675950407981873, -0.1737656444311142, -0.14634230732917786, -0.06844387203454971, 0.2047617882490158, -0.003669668920338154, 0.05763379856944084, 0.07807634025812149, 0.19451922178268433, 0.2609213590621948, 0.23615522682666779, 0.24638855457305908, -0.07365656644105911, 0.10551030933856964, 0.344176709651947, 0.3810717761516571, 0.0458010733127594, 0.4809633493423462, -0.04737747088074684, -0.03335289657115936, 0.03335140272974968, -0.15820325911045074, -0.055854521691799164, 0.46437501907348633, 0.1498102843761444, 0.05822744220495224, 0.046125754714012146, -0.11693403124809265, -0.37757667899131775, 0.0007064300589263439, -0.05507602542638779, -0.14003075659275055, -0.29990676045417786, -0.4009559452533722, 0.261601984500885, 0.34820178151130676, -0.14399400353431702, -0.1402910351753235, -0.17629612982273102, -0.3403513729572296, 0.2590388059616089, 0.23190838098526, 1.041952133178711, 0.2871282398700714, 0.07867705821990967, -0.14152073860168457, -0.19598761200904846, 0.8950437307357788, -0.4584696292877197, 0.3884287476539612, -0.3797582983970642, -0.16995228826999664, -0.049717869609594345, -0.13921773433685303, 0.06719174981117249, 0.07828167825937271, -0.34914612770080566, 0.44244837760925293, 0.23112909495830536, 0.06170216202735901, -0.14359596371650696, 0.3393726348876953, -0.24981385469436646, -0.18752291798591614, -0.19505421817302704, -0.005369901657104492, -0.23499362170696259, 0.3488823175430298, -0.009586093947291374, -0.0852256715297699, -0.05986133962869644, -0.18270446360111237, -0.3571862578392029, 0.11464259773492813, 0.09397121518850327, 0.012490633875131607, 0.23694093525409698, -0.20036742091178894, -0.11602862179279327, 0.09369586408138275, 0.3660138249397278, -0.1871987134218216, -0.1647212952375412, -0.05034508928656578, -0.13756588101387024, -0.0733359232544899, 0.11516939103603363, 0.06339465826749802, 0.3432280719280243, -0.0829625129699707, 0.06117624789476395, 0.1170477643609047, -0.19855327904224396, -0.44435131549835205, -0.21357160806655884, -0.13070234656333923, -0.14821791648864746, -0.11806496232748032, -0.023201357573270798, 0.08216432482004166, -0.01390138454735279, -0.05310804769396782, 0.04212629050016403, 0.003688342869281769, -0.15956947207450867, 0.3063167929649353, -0.2527936100959778, -0.23129945993423462, -0.16731902956962585, 0.3863880932331085, 0.3539524972438812, -0.02271563559770584, 0.38353440165519714, -0.1494131237268448, -0.16333270072937012, -0.1423272043466568, -0.22025412321090698, -0.24535731971263885, -0.3556593954563141, 0.11342094093561172, -0.32436907291412354, -0.23751303553581238, 0.09655298292636871, 0.19885854423046112, 0.17593811452388763, 0.398185670375824, -0.05935514718294144, -0.43293261528015137, -0.08512742072343826, 0.16189225018024445, -0.08562631905078888, 0.10745811462402344, -0.03459126502275467, 0.29530346393585205, -0.2341655194759369, 0.30818596482276917, -0.22415944933891296, -0.2144247442483902, -0.3940698206424713, 0.27270591259002686, 0.0848626047372818, -0.03858967125415802, 0.014369315467774868, 0.04511518031358719, -0.21619164943695068, 0.2673085033893585, -0.16531917452812195, 0.005659110844135284, -0.051704272627830505, 0.20624279975891113, 0.13152076303958893, -0.031795769929885864, -0.18519622087478638, 0.12904246151447296, 0.09241284430027008, 0.04547378793358803, 0.02897149696946144, 0.2819368839263916, 0.040343932807445526, 0.4037655293941498, -0.25698018074035645, 0.08450240641832352, -0.01948448270559311, 0.20341449975967407, -0.27178576588630676, 0.32474762201309204, -0.08645592629909515, 0.30650511384010315, -0.20353427529335022, 0.029180921614170074, -0.02076466754078865, 0.07677067816257477, 0.12723585963249207, 0.01133882999420166, 0.1532256305217743, -0.14263533055782318, 0.027951475232839584, 0.27671822905540466, -0.0762605294585228, 0.5579020977020264, -0.11840230226516724, -0.2716977596282959, 0.13532370328903198, 0.07329834997653961, -0.024312928318977356, -0.06569484621286392, 0.09087368845939636, 0.02758844569325447, -0.0758267492055893, 0.4246918857097626, 0.22875291109085083, -0.1360151767730713, -0.012512922286987305, 0.23883235454559326, 0.4896003305912018, -0.26902318000793457, 0.07844099402427673, 0.1427544802427292, -0.18673284351825714, 0.05787607282400131, 0.3953358829021454, -0.23366129398345947, 0.14236792922019958, 0.06332097202539444, 0.11303883790969849, 0.4625708758831024, 0.20257914066314697, 0.3750728666782379, -0.18873688578605652, -0.10802371799945831, -0.05696186050772667, 0.623226523399353, 0.27606868743896484, 0.3305191099643707, 0.19891810417175293, 0.3161473870277405, -0.14343422651290894, 0.010834918357431889, -0.3369036614894867, -0.18920423090457916, -0.0378134548664093, -0.14686629176139832, -0.37576279044151306, -0.09626250714063644, -0.08604558557271957, 0.08358070254325867, -0.13657602667808533, 0.10390137881040573, 0.5570069551467896, -0.17096583545207977, -0.1825268417596817, -0.5047012567520142, 0.22231590747833252, 0.003293905407190323, 0.20471422374248505, -0.22498595714569092, 0.30062878131866455, -0.04419136047363281, -0.20166704058647156, 0.28789833188056946, 0.5239021182060242, 0.42503392696380615, 0.2935965657234192, -0.040379539132118225, -0.17957088351249695, 0.07012112438678741, 0.17155423760414124, 0.19459757208824158, 0.7336778044700623, 0.22640828788280487, -0.36697906255722046, 0.2798475921154022, -0.07750452309846878, -0.04813341796398163, 0.47493553161621094, 0.010236386209726334, 0.03268764540553093, -0.02730133756995201, 0.42502889037132263, 0.010576346889138222, 0.11824994534254074, 0.11753836274147034, 0.14925113320350647, -0.22901606559753418, -0.2858244776725769, 0.1325470209121704, -0.17832809686660767, 0.28705668449401855, -0.1423882246017456, 0.0456337109208107, -0.20899464190006256, 0.609613835811615, 0.4575616121292114, -0.2356964647769928, -0.22168287634849548, 0.0019248947501182556, -0.5511803030967712, 0.3374861776828766, -0.3613298833370209, -0.20019188523292542, -0.17516092956066132, 0.5507522821426392, -0.34127503633499146, 0.32878273725509644, -0.30485665798187256, 0.3189481496810913, -0.0368739552795887, 0.030296996235847473, -0.27325257658958435, -0.09151332080364227, -0.08998356759548187, -0.14913113415241241, 0.08855604380369186, -0.2684873342514038, 0.251039057970047, 0.23709236085414886, -0.10101060569286346, -0.020666047930717468, -0.02704480290412903, 0.17944830656051636, 0.5358684062957764, -0.028986811637878418, 0.5525883436203003, 0.21866723895072937, -0.11766823381185532, -0.17617908120155334, -0.21538448333740234, -0.24386943876743317, -0.3098509609699249, 0.15333756804466248, 0.13062870502471924, 0.3093514144420624, -0.2756575047969818, 0.49836015701293945, -0.14632435142993927, 0.41652774810791016, -0.016830863431096077, 0.008334021084010601, -0.23042693734169006, -0.04797782748937607, -0.09354915469884872, 0.12309993803501129, 0.18964512646198273, 0.412678986787796, 0.17846408486366272, 0.14102651178836823, -0.19011104106903076, -0.33226096630096436, 0.3096321225166321, -0.3596898317337036, 0.08010396361351013, -0.03542779013514519, 0.2624869644641876, 0.27984440326690674, -0.24873369932174683, -0.731859564781189, -0.06884316354990005, 0.2349817305803299, -0.12555167078971863, 0.06184437498450279, 0.04573851078748703, -0.11327065527439117, -0.2084636092185974, -0.09276942908763885, -0.1714310348033905, 0.28795912861824036, -0.3046322166919708, 0.3468710482120514, -0.1572411060333252 ]
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 ```
[ 0.047822896391153336, -0.014993414282798767, 0.0999605804681778, 0.48276644945144653, 0.25262296199798584, 0.11827053874731064, 0.12425927817821503, 0.4629979133605957, 0.20459410548210144, -0.11430197954177856, 0.19549094140529633, 0.5628964900970459, -0.11228945106267929, -0.21451006829738617, -0.06854422390460968, -0.1865536868572235, 0.11651542037725449, 0.24228183925151825, -0.003951467573642731, -0.019452378153800964, -0.12144695222377777, 0.12436164915561676, -0.26443976163864136, 0.16095438599586487, -0.4509822726249695, -0.1343631148338318, 0.1006031185388565, 0.0705757588148117, -0.1665283888578415, -0.45291265845298767, 0.13311639428138733, -0.38631588220596313, -0.014494180679321289, 0.267690509557724, -0.00012025610340060666, -0.1023913249373436, 0.35268068313598633, 0.04843998700380325, -0.26101523637771606, -0.2479172796010971, -0.4819885492324829, 0.09865820407867432, 0.3085624873638153, -0.0510614737868309, 0.22519242763519287, 0.0243794247508049, 0.023879460990428925, -0.5974248647689819, 0.1822831928730011, 0.5890493988990784, 0.1811952143907547, 0.31175121665000916, -0.058616433292627335, 0.08240842819213867, 0.22701802849769592, 0.08643073588609695, -0.34749895334243774, 0.15860091149806976, -0.041865259408950806, -0.16374878585338593, -0.09725948423147202, 0.4248977303504944, -0.17378702759742737, 0.12876734137535095, 0.6042642593383789, 0.1013762578368187, 0.10312234610319138, -0.3009956479072571, 0.18946436047554016, 0.002262536436319351, 0.2738999128341675, -0.18591824173927307, -0.08067543059587479, 0.03091522306203842, -0.14413130283355713, -0.3694726824760437, 0.20961233973503113, 0.17653074860572815, -0.3138512670993805, 0.017662402242422104, 0.011315234936773777, 0.0840630978345871, -0.06323167681694031, -0.07367566227912903, -0.05972215160727501, 0.06742092221975327, 0.013082700781524181, 0.11426423490047455, -0.05431633070111275, -0.1860046535730362, -0.1897270679473877, -0.22003136575222015, -0.20488402247428894, 0.24149291217327118, -0.3378833830356598, 0.04840739071369171, -0.05792555958032608, -0.2089127004146576, -0.16488376259803772, -0.11893290281295776, 0.2784707546234131, -0.00025321170687675476, 0.14469991624355316, 0.12964114546775818, 0.3333240747451782, 0.2861374020576477, 0.32042112946510315, 0.24878515303134918, 0.0986189991235733, -0.3073054552078247, -0.1854362040758133, -0.013636887073516846, -0.011412401683628559, -0.009702622890472412, 0.5141108632087708, 0.016387635841965675, 0.3002511262893677, -0.3601059913635254, -0.3549935519695282, -0.011049035005271435, -0.735805332660675, 0.159963920712471, -0.16902358829975128, 0.10665593296289444, -0.008254090324044228, 0.4835464060306549, 0.24768327176570892, 0.2814023792743683, -0.3391999900341034, -0.19262823462486267, -0.11420821398496628, -0.004297148436307907, -0.29851973056793213, 0.015622427687048912, 0.305298775434494, -0.03636535629630089, 0.053923748433589935, 0.06570818275213242, -0.40770697593688965, -0.20890961587429047, 0.3739098012447357, -0.42265287041664124, 0.17886768281459808, -0.008707631379365921, -0.32873719930648804, 0.41525059938430786, 0.24119050800800323, -0.4089019298553467, -0.21291393041610718, 0.1715904176235199, -0.13049659132957458, 0.0368976853787899, -0.19280916452407837, 0.12478379905223846, 0.1592133790254593, -0.21862554550170898, -0.17026445269584656, -0.06532660126686096, 0.39254382252693176, -0.10832788050174713, 0.10371732711791992, -0.2929549813270569, -0.0403238981962204, -0.24107541143894196, -0.08629386126995087, 0.4788670539855957, -0.22314198315143585, 0.06661124527454376, -0.055577293038368225, 0.18355394899845123, 0.3641401529312134, 0.378109872341156, -0.020373918116092682, 0.1771153062582016, 0.1071867048740387, 0.28203481435775757, 0.4739597737789154, -0.30076542496681213, -0.10005518049001694, 0.09920661896467209, -0.028821874409914017, -0.007086681202054024, -0.041806064546108246, 0.12140318751335144, 0.07706896960735321, -0.1202663853764534, 0.1674463152885437, 0.19477063417434692, -0.06231947988271713, 0.23532545566558838, -0.4206896126270294, -0.10153254866600037, 0.5093355774879456, -0.17420120537281036, -0.0756608247756958, 0.026094786822795868, -0.07687561959028244, 0.05020181089639664, -0.0676264539361, -0.013059113174676895, 0.4281765818595886, 0.26007670164108276, -0.09044051170349121, -0.08258424699306488, 0.0832560732960701, -0.1882937252521515, -0.38731589913368225, -0.052016731351614, -0.10536511242389679, 0.13355886936187744, -0.17240871489048004, -0.040387071669101715, -0.3455455005168915, 0.07457055896520615, -0.02713197097182274, -0.19891293346881866, 0.1253451704978943, 0.10824747383594513, 0.15891964733600616, 0.05802837386727333, -0.22786827385425568, 0.1251097172498703, 0.00482417456805706, 0.12739752233028412, -0.3979262709617615, -0.14816661179065704, -0.08023757487535477, -0.3554534614086151, -0.09916677325963974, 0.17995890974998474, 0.2376292645931244, 0.16755416989326477, -0.26986411213874817, 0.1873168647289276, 0.018680404871702194, 0.2171434462070465, -0.2549818456172943, 0.02990114688873291, 0.01876700296998024, -0.4021495580673218, 0.2202015370130539, 0.32697343826293945, 0.36164918541908264, -0.2435903251171112, -0.3964940011501312, -0.02146250009536743, 0.01459936611354351, 0.58585524559021, -0.13255611062049866, 0.10853246599435806, -0.07677534222602844, -0.03623274341225624, 0.09410479664802551, -0.07976377010345459, 0.096011221408844, -0.11330640316009521, 0.19795340299606323, 0.29243698716163635, -0.30972325801849365, 0.27264833450317383, 0.42169538140296936, 0.3002501130104065, 0.13427668809890747, 0.15981075167655945, -0.33452171087265015, -0.0856642946600914, 0.2644166350364685, -0.012384925037622452, 0.46109122037887573, -0.0432867556810379, 0.21456070244312286, -0.08492761850357056, -0.15070253610610962, -0.14859648048877716, 0.35784003138542175, -0.09661878645420074, 0.01726612262427807, -0.009265869855880737, 0.4907113313674927, 0.010626139119267464, -0.17871445417404175, 0.20089170336723328, 0.005513790994882584, 0.2383062243461609, -0.3552829623222351, 0.3290385901927948, -0.3391128182411194, -0.16789516806602478, -0.38104143738746643, -0.2544400691986084, -0.018507950007915497, -0.502819299697876, -0.14053601026535034, 0.0792337954044342, 0.006112053990364075, 0.12038955837488174, -0.2541165053844452, -0.14185769855976105, 0.018085403367877007, -0.11071298271417618, -0.33473554253578186, -0.3540615737438202, -0.370501846075058, -0.06328783929347992, 0.46462759375572205, 0.32319629192352295, 0.1873820722103119, 0.053033873438835144, -0.0629505142569542, -0.06901051104068756, -0.3603874742984772, 0.03056313283741474, -0.19892199337482452, 0.05718579888343811, 0.12039291858673096, 0.07614035904407501, -0.13603101670742035, -0.12729725241661072, 0.42562082409858704, 0.0031829513609409332, -0.24060222506523132, 0.32919079065322876, -0.02077130228281021, 0.04902191460132599, -0.29279494285583496, -0.48070839047431946, 0.013537148013710976, -0.46828681230545044, 0.13702994585037231, -0.11211821436882019, 0.20932060480117798, 0.17763294279575348, 0.12377581000328064, 0.3106841444969177, -0.0557437539100647, 0.18512554466724396, 0.20702683925628662, 0.3653653860092163, 0.6162764430046082, -0.13313645124435425, -0.23794661462306976, 0.20846593379974365, -0.07696908712387085, 0.3138101398944855, 0.2074880450963974, -0.17882150411605835, 0.33277010917663574, -0.42187902331352234, 0.3207613527774811, -0.16330711543560028, -0.13050541281700134, 0.2217363715171814, 0.22102117538452148, -0.048052482306957245, -0.06043776869773865, -0.007517026737332344, 0.21831294894218445, -0.21776598691940308, 0.15179309248924255, 0.12625360488891602, 0.21420934796333313, -0.05345015227794647, 0.38171952962875366, 0.14399783313274384, -0.2284076064825058, 0.28303128480911255, 0.016808874905109406, -0.04759572818875313, -0.30222105979919434, -0.27842622995376587, -0.08813535422086716, -0.21157032251358032, 0.1277090311050415, -0.1773894876241684, 0.033944182097911835, -0.4855811297893524, -0.022265659645199776, 0.2889741063117981, -0.18293264508247375, -0.13431289792060852, 0.3022046387195587, -0.2138022780418396, 0.2901405990123749, 0.13461044430732727, 0.12581947445869446, -0.19792313873767853, -0.41243892908096313, 0.03293651342391968, -0.024656660854816437, 0.15686127543449402, 0.024755267426371574, -0.10193738341331482, -0.05289794132113457, -0.47397786378860474, 0.28840142488479614, 0.1751556694507599, 0.4326755702495575, -0.0764363631606102, -0.1432514786720276, 0.1917077898979187, -0.11352067440748215, 0.2986406683921814, -0.14612385630607605, -0.027595356106758118, 0.3758091628551483, 0.2723763883113861, -0.4777471125125885, 0.2693077623844147, -0.23111483454704285, 0.3493022322654724, 0.26166728138923645, 0.2834547162055969, -0.3896859288215637, -0.3131132423877716, 0.25790220499038696, 0.16102814674377441, 0.016652081161737442, 0.0899721160531044, -0.31698015332221985, -0.00351816788315773, -0.15553444623947144, -0.08673816174268723, 0.2757047116756439, 0.3106018602848053, 0.08262214064598083, 0.054902251809835434, -0.330187052488327, 0.17894992232322693, 0.18064190447330475, 0.026372727006673813, 0.428266704082489, -0.30367183685302734, 0.137691468000412, -0.1568317711353302, 0.20898042619228363, -0.08677200227975845, 0.3240392506122589, -0.06349006295204163, -0.19840139150619507, 0.26688045263290405, -0.05753713473677635, 0.36443638801574707, 0.3292663097381592, 0.03609280288219452, 0.3470538258552551, -0.12237447500228882, -0.09592166543006897, -0.2033008635044098, 0.029605166986584663, 0.16758917272090912, -0.05280951038002968, 0.04342714324593544, -0.2120436728000641, 0.4822666347026825, 0.2981495261192322, -0.033570606261491776, 0.6272773742675781, 0.1370314061641693, -0.2833763062953949, 0.23740094900131226, 0.1865319013595581, 1.1821577548980713, 0.1512257158756256, 0.1496681571006775, 0.13378293812274933, -0.4301357865333557, 0.33390843868255615, 0.06193799525499344, 0.32546019554138184, -0.4628365933895111, 0.3249562978744507, 0.024498580023646355, -0.1689324676990509, 0.06731470674276352, -0.03389614075422287, -0.3024047613143921, 0.0830451101064682, -0.19603002071380615, 0.3092464506626129, 0.11238547414541245, 0.058071210980415344, -0.3026752471923828, -0.31657859683036804, -0.37491852045059204, 0.08060061186552048, -0.25055015087127686, -0.0766252726316452, 0.11036089062690735, -0.21556557714939117, -0.49149301648139954, -0.04550528898835182, -0.06853950023651123, 0.1147971823811531, -0.1706717610359192, 0.046522919088602066, 0.23672963678836823, -0.15768678486347198, 0.1811174750328064, 0.0725666955113411, -0.10370728373527527, -0.06586925685405731, -0.1338338851928711, -0.08608773350715637, 0.19671598076820374, 0.10096611827611923, 0.12061899900436401, -0.07572063058614731, 0.011200949549674988, 0.07623331993818283, 0.02482161670923233, 0.2276177704334259, -0.3680090010166168, -0.01354152336716652, -0.06473123282194138, 0.16641730070114136, 0.11926662176847458, -0.09275055676698685, -0.08846518397331238, 0.09907546639442444, -0.06051851063966751, 0.0031132884323596954, 0.0932033360004425, 0.02566814050078392, 0.15167847275733948, 0.08057156950235367, -0.15691064298152924, -0.2600077986717224, 0.26901620626449585, 0.34133127331733704, 0.04775969684123993, 0.24060404300689697, 0.7169973850250244, 0.0595085434615612, -0.18554146587848663, -0.17794457077980042, 0.07761091738939285, -0.052395571023225784, -0.7310899496078491, 0.3293781280517578, -0.33141714334487915, -0.22107812762260437, 0.03403424471616745, 0.4544679522514343, -0.202938050031662, 0.08033965528011322, -0.1594124734401703, -0.2330784797668457, -0.21894444525241852, 0.17904852330684662, -0.030600447207689285, 0.03467164188623428, -0.45244812965393066, -0.057528574019670486, 0.09664352983236313, 0.1214955747127533, -0.22719335556030273, -0.010041768662631512, -0.17429344356060028, 0.3713476061820984, 0.09213243424892426, 0.03296457976102829, -0.050732433795928955, -0.18051032721996307, -0.011443930678069592, 0.515113115310669, -0.09715501219034195, -0.14438799023628235, -0.26212018728256226, 0.14715488255023956, 0.07113237679004669, -0.16618289053440094, 0.07055415213108063, 0.024171963334083557, 0.0784350335597992, -0.21286910772323608, -0.15107877552509308, -0.10288618505001068, -0.18629826605319977, 0.05603841319680214, -0.08770027756690979, 0.2029782235622406, -0.17871850728988647, 0.09925535321235657, -0.14713366329669952, 0.05682013928890228, -0.24345897138118744, 0.10793513804674149, 0.2107045203447342, 0.057123005390167236, -0.2081969976425171, 0.04877771437168121, -0.0946367084980011, -0.15886461734771729, 0.3047354519367218, -0.37726426124572754, -0.166188046336174, 0.056550346314907074, -0.1906721591949463, 0.41663122177124023, -0.0920555517077446, 0.0020520929247140884, -0.08149725943803787, 0.09775548428297043, 0.07033410668373108, 0.31662482023239136, -0.17869526147842407, -0.1631123125553131, -0.11346417665481567, 0.1911708414554596, -0.023984402418136597, -0.1549658477306366, -0.030290916562080383, 0.09489049762487411, 0.033047355711460114, -0.4421166777610779, -0.16407272219657898, 0.13605383038520813, -0.025477886199951172, -0.07160672545433044, 0.3569367825984955, -0.10453945398330688, 0.021537616848945618, 0.02030269056558609, 0.36647793650627136, 0.334046334028244, 0.43646442890167236, 0.20343390107154846, -0.06764330714941025, 0.11950598657131195, -0.02497245743870735, 0.38919922709465027, 0.24424739181995392, 0.15110091865062714, 0.17533531785011292, 0.5244446992874146, -0.298078715801239, -0.32812410593032837, 0.06573379039764404, 0.10832846164703369, -0.27272334694862366, -0.08761180937290192, -0.5703158378601074, -0.11522654443979263, -0.4419068396091461, -0.08809848129749298, -0.19053779542446136, 0.25448960065841675, 0.692865252494812, 0.23379063606262207, 0.016697607934474945, -0.5845615863800049, -0.12988480925559998, 0.13295458257198334, 0.39208438992500305, -0.2488599270582199, 0.40359482169151306, 0.10868524014949799, -0.0760030746459961, -0.13055269420146942, 0.49973490834236145, 0.4683450162410736, 0.5678133368492126, -0.2883794605731964, -0.06765200197696686, 0.05808182805776596, -0.18258091807365417, 0.23777779936790466, 0.5766581892967224, 0.33824294805526733, 0.14017975330352783, 0.03504382073879242, -0.03796571493148804, -0.1073760911822319, -0.09008006751537323, -0.029721643775701523, 0.277446985244751, -0.16376112401485443, 0.2949458956718445, -0.23841068148612976, -0.22263887524604797, -0.046252354979515076, -0.21092990040779114, -0.3650149405002594, -0.04621872305870056, 0.20551665127277374, 0.0495741181075573, 0.01566902920603752, -0.33418184518814087, 0.05742139369249344, 0.20212304592132568, 0.5171065926551819, 0.27114230394363403, 0.13861316442489624, -0.47538450360298157, -0.11695388704538345, -0.7315421104431152, 0.1763940453529358, -0.3149000406265259, -0.1567133218050003, -0.014518678188323975, 0.3632786273956299, 0.04523162543773651, 0.31469881534576416, -0.034921616315841675, 0.2289496809244156, 0.25130152702331543, 0.30536097288131714, -0.3416901230812073, -0.11979328095912933, 0.35304325819015503, -0.3037296235561371, 0.023086491972208023, -0.5888969302177429, 0.26860255002975464, 0.051241569221019745, 0.06511207669973373, -0.008692510426044464, -0.2844811677932739, 0.29853957891464233, 0.48924562335014343, 0.27550697326660156, 0.2637924551963806, 0.2856045365333557, 0.24758201837539673, -0.5557767748832703, -0.21760618686676025, -0.1734553575515747, -0.20772723853588104, -0.020305626094341278, 0.08986908197402954, 0.5805659294128418, -0.14044833183288574, 0.062437351793050766, -0.3035687804222107, 0.3277711570262909, -0.22651728987693787, 0.034377943724393845, -0.17557282745838165, 0.06700552254915237, -0.3490554690361023, 0.16302508115768433, 0.07218987494707108, 0.1204005628824234, 0.1978316456079483, 0.24254406988620758, -0.5347840189933777, -0.38742899894714355, 0.30438604950904846, -0.5194072723388672, -0.4466894865036011, 0.2931904196739197, 0.14298835396766663, 0.44685134291648865, -0.012415779754519463, -0.7071853280067444, -0.2688806653022766, 0.21321886777877808, -0.09643738716840744, 0.025330565869808197, 0.023005280643701553, 0.02659551426768303, 0.07402946054935455, -0.19160304963588715, 0.3619847893714905, 0.00830192118883133, 0.049281045794487, 0.04832902178168297, -0.22841106355190277 ]
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.
[ 0.1208232045173645, 0.08130066096782684, 0.08892324566841125, 0.4082986116409302, 0.2814409136772156, 0.05460158735513687, 0.11376447230577469, 0.33654892444610596, 0.18600942194461823, -0.18695004284381866, 0.24845752120018005, 0.5634832978248596, 0.0009862612932920456, -0.07831837981939316, -0.023558761924505234, -0.11811482906341553, 0.07306350767612457, 0.2508920431137085, 0.050629690289497375, -0.1244528666138649, -0.09246085584163666, 0.1616436094045639, -0.18193918466567993, 0.22253037989139557, -0.5130985379219055, -0.15872830152511597, 0.1441274732351303, 0.08026709407567978, -0.11963669210672379, -0.4990118741989136, 0.01551610417664051, -0.34734246134757996, 0.03081813082098961, 0.20836132764816284, -0.00011913953494513407, -0.16594409942626953, 0.3734931945800781, -0.052215129137039185, -0.1960316151380539, -0.21079610288143158, -0.5531389117240906, 0.16648231446743011, 0.3032819926738739, -0.06724217534065247, 0.23992854356765747, -0.10213456302881241, 0.0658058375120163, -0.5268751382827759, 0.09618109464645386, 0.49537739157676697, 0.19496870040893555, 0.15002773702144623, 0.11893051862716675, 0.08554054796695709, 0.26706463098526, 0.11917148530483246, -0.28745442628860474, 0.0432623028755188, -0.13040462136268616, -0.22419467568397522, -0.06938833743333817, 0.3475063741207123, -0.04919649660587311, -0.0045446306467056274, 0.5662893056869507, 0.044567935168743134, 0.19759419560432434, -0.18642769753932953, 0.25648245215415955, 0.058506954461336136, 0.3549572825431824, -0.061708323657512665, -0.1086406335234642, 0.09885508567094803, -0.12332411855459213, -0.3633821904659271, 0.28847557306289673, 0.22252202033996582, -0.2659066617488861, -0.012292016297578812, 0.11637657880783081, 0.06312420964241028, 0.009714841842651367, -0.12463657557964325, -0.15653665363788605, 0.1494547575712204, 0.018611852079629898, 0.11266791820526123, -0.05011356249451637, -0.1976815164089203, -0.1804715394973755, -0.07938490062952042, -0.30318981409072876, 0.2140204757452011, -0.28308483958244324, 0.011638686060905457, -0.05160389840602875, -0.16387298703193665, -0.20187893509864807, -0.08790966868400574, 0.41374561190605164, -0.0001445319503545761, -0.005352219566702843, 0.06754866987466812, 0.3144073188304901, 0.42886194586753845, 0.46972113847732544, 0.098272405564785, 0.13660791516304016, -0.30488911271095276, -0.18265269696712494, -0.03447582572698593, -0.025242093950510025, -0.009739932604134083, 0.43271714448928833, 0.024083491414785385, 0.07819658517837524, -0.3047500550746918, -0.25904086232185364, 0.026841796934604645, -0.7115640044212341, 0.2219676673412323, -0.21643343567848206, 0.13746324181556702, 0.06488527357578278, 0.4073716700077057, 0.29524290561676025, 0.2960963249206543, -0.2042863816022873, -0.20120768249034882, -0.11234623193740845, -0.09873443841934204, -0.3487413823604584, -0.1523420810699463, 0.19270963966846466, -0.0855012834072113, 0.08561816811561584, 0.02165371924638748, -0.5190795063972473, -0.22347429394721985, 0.29673436284065247, -0.2855612337589264, 0.2961820662021637, 0.06044945493340492, -0.27452200651168823, 0.45916247367858887, 0.21227796375751495, -0.5164345502853394, -0.09735013544559479, 0.0927971601486206, -0.204600989818573, 0.07067333161830902, -0.18194742500782013, 0.1397913545370102, 0.03730938956141472, -0.2744908034801483, -0.20205318927764893, -0.0635196715593338, 0.3884657919406891, -0.16859491169452667, 0.09392964094877243, -0.2046530842781067, -0.012688685208559036, -0.28410348296165466, -0.14891177415847778, 0.383408784866333, -0.18657824397087097, 0.10843682289123535, -0.0751800686120987, 0.1669815629720688, 0.3044077157974243, 0.34559154510498047, 0.0228829775005579, 0.19979768991470337, 0.10551220178604126, 0.3500438332557678, 0.39273685216903687, -0.2295798361301422, -0.009729590266942978, -0.0022537419572472572, -0.017602205276489258, -0.007658632472157478, -0.05828159302473068, -0.0028016925789415836, 0.042020607739686966, -0.11749880015850067, 0.08269516378641129, 0.12329777330160141, 0.01209910400211811, 0.259146511554718, -0.3936404287815094, -0.06219422072172165, 0.4800755977630615, -0.31273168325424194, -0.2060406655073166, -0.030394628643989563, 0.016499511897563934, 0.10142692923545837, -0.02642885595560074, -0.09374566376209259, 0.30804821848869324, 0.17479312419891357, -0.07066208124160767, -0.14800602197647095, 0.15718470513820648, -0.04129200056195259, -0.3996149003505707, -0.05961373448371887, -0.11000104248523712, 0.1341487169265747, -0.051153406500816345, 0.0029529668390750885, -0.2747977077960968, 0.05228060483932495, -0.007796440273523331, -0.25278154015541077, 0.14248062670230865, 0.06316795945167542, 0.16465048491954803, 0.12633134424686432, -0.21560579538345337, 0.06162596493959427, 0.021692467853426933, 0.12927879393100739, -0.36722826957702637, -0.2625303566455841, 0.010546255856752396, -0.3314328193664551, -0.19128312170505524, 0.23127031326293945, 0.2290375530719757, 0.13432130217552185, -0.2543701231479645, 0.19253799319267273, -0.009426010772585869, 0.3920418620109558, -0.21428406238555908, 0.07450925558805466, -0.01363501138985157, -0.5046822428703308, 0.19313804805278778, 0.4102119505405426, 0.2995215952396393, -0.2957395613193512, -0.4034363925457001, -0.053517699241638184, -0.03167383745312691, 0.5904877781867981, -0.09309862554073334, 0.25580811500549316, -0.02029421180486679, -0.06831356883049011, 0.07016144692897797, -0.12563331425189972, 0.19013722240924835, -0.16906914114952087, 0.17845749855041504, 0.2348817139863968, -0.23076646029949188, 0.31589844822883606, 0.38045749068260193, 0.32732710242271423, 0.21452650427818298, 0.03870785981416702, -0.3840383291244507, -0.060156673192977905, 0.27095603942871094, 0.06697294116020203, 0.3688619136810303, -0.059140823781490326, 0.22426293790340424, -0.13389982283115387, -0.15443645417690277, -0.17635606229305267, 0.321090430021286, -0.20176978409290314, 0.07100464403629303, 0.06617754697799683, 0.5125985145568848, 0.017989778891205788, -0.18942219018936157, 0.3008752167224884, -0.016682380810379982, 0.23172098398208618, -0.41760358214378357, 0.2389870434999466, -0.31163686513900757, -0.05878400430083275, -0.34087467193603516, -0.2117820829153061, -0.0825006514787674, -0.565778911113739, -0.030333183705806732, -0.033450912684202194, -0.07007120549678802, 0.03609158471226692, -0.11940206587314606, -0.14799292385578156, -0.10889899730682373, -0.07129647582769394, -0.3734090328216553, -0.4300118386745453, -0.25979503989219666, -0.04753976687788963, 0.3729802370071411, 0.23948872089385986, 0.10474350303411484, -0.012625178322196007, -0.07535828649997711, 0.00032676197588443756, -0.4714282155036926, -0.004124712198972702, -0.22297047078609467, 0.06794367730617523, 0.18282291293144226, 0.02107103168964386, -0.030382610857486725, -0.014437520876526833, 0.36036163568496704, -0.06673955172300339, -0.27643465995788574, 0.3501123785972595, 0.06635912507772446, 0.054407998919487, -0.2934790253639221, -0.31152215600013733, 0.08216521143913269, -0.466784805059433, 0.3006454408168793, -0.09647873044013977, 0.24163854122161865, 0.2114756554365158, -0.017977159470319748, 0.2288760095834732, -0.0653318464756012, 0.1617140769958496, 0.18838946521282196, 0.4239365756511688, 0.5407384634017944, -0.22937588393688202, -0.3174418807029724, 0.29581257700920105, -0.03764570504426956, 0.3881964087486267, 0.1468108892440796, -0.04843839257955551, 0.4057908058166504, -0.384377658367157, 0.131942018866539, -0.14741839468479156, -0.07091417163610458, 0.1809772104024887, 0.2352914959192276, -0.05996190011501312, -0.1455630660057068, 0.033156804740428925, 0.21858522295951843, -0.03075641393661499, 0.24850612878799438, 0.09110978245735168, 0.004534488543868065, -0.11354295909404755, 0.3820229172706604, 0.30407604575157166, -0.15708638727664948, 0.26299142837524414, 0.006089884787797928, -0.06353563070297241, -0.230089008808136, -0.27072858810424805, -0.1288834810256958, -0.17353969812393188, 0.15583141148090363, -0.15172062814235687, 0.033583350479602814, -0.38190969824790955, -0.031434834003448486, 0.33778005838394165, -0.18510079383850098, -0.1455889195203781, 0.2932654917240143, -0.2784809470176697, 0.23703891038894653, 0.08548419177532196, 0.19006088376045227, -0.12471485137939453, -0.40733543038368225, 0.12500068545341492, -0.0006934851408004761, 0.1178487092256546, 0.036328915506601334, -0.19866199791431427, -0.12969739735126495, -0.4102337658405304, 0.31972992420196533, 0.03993400186300278, 0.36018651723861694, -0.1612703651189804, -0.1563730239868164, 0.1840427666902542, -0.15219661593437195, 0.4350284934043884, -0.3664802610874176, -0.05906306952238083, 0.30513784289360046, 0.3421058654785156, -0.4739355742931366, 0.34191980957984924, -0.22176426649093628, 0.24682141840457916, 0.1647558957338333, 0.1495353877544403, -0.32348155975341797, -0.23489397764205933, 0.17474141716957092, 0.02016613818705082, 0.023735221475362778, 0.17043368518352509, -0.3350865840911865, 0.02666223794221878, -0.07697559148073196, -0.17390739917755127, 0.20773641765117645, 0.24202145636081696, 0.018690072000026703, 0.07589267194271088, -0.4179534316062927, 0.19375517964363098, 0.2174443006515503, 0.08719144761562347, 0.39977794885635376, -0.24007312953472137, 0.011527381837368011, -0.020229674875736237, 0.1994381695985794, -0.07045424729585648, 0.45299166440963745, -0.04165391996502876, -0.17850881814956665, 0.25579360127449036, -0.08369375020265579, 0.4723662734031677, 0.3834895193576813, 0.08568917214870453, 0.32462573051452637, -0.27232176065444946, -0.22830970585346222, -0.3209863603115082, 0.006949512753635645, 0.04117930680513382, 0.00267807487398386, 0.14303763210773468, -0.10382363200187683, 0.5234213471412659, 0.24864955246448517, -0.07634901255369186, 0.5133987665176392, 0.16703101992607117, -0.3357831835746765, 0.323159396648407, 0.2279377579689026, 1.1573790311813354, 0.058490168303251266, 0.10378345847129822, -0.058251798152923584, -0.46720439195632935, 0.44061124324798584, 0.06488735228776932, 0.2514752745628357, -0.47175145149230957, 0.35401394963264465, 0.004888011142611504, -0.2005733847618103, 0.04434853419661522, -0.051942937076091766, -0.30089566111564636, 0.11113596707582474, -0.21602872014045715, 0.24487365782260895, 0.09667222946882248, 0.09772033989429474, -0.3100961744785309, -0.3052619397640228, -0.2676105201244354, 0.09327099472284317, -0.1876256912946701, -0.007991142570972443, 0.11706545203924179, -0.2904031276702881, -0.5326499938964844, -0.05518368259072304, -0.08857408165931702, 0.19176533818244934, -0.1987987458705902, -0.07285838574171066, 0.25250235199928284, -0.2095436453819275, 0.23995491862297058, 0.23208938539028168, 0.08816669881343842, -0.06347925961017609, -0.17391392588615417, -0.04728449508547783, 0.29204028844833374, 0.15689215064048767, 0.12021410465240479, -0.033129528164863586, 0.014185480773448944, 0.10079529136419296, 0.04560539871454239, 0.25179731845855713, -0.4120064377784729, 0.04076303541660309, -0.08722396194934845, 0.20695391297340393, 0.09257963299751282, -0.05611642822623253, 0.058333199471235275, 0.056927867233753204, -0.05763889104127884, -0.024936577305197716, 0.10847321152687073, 0.013143669813871384, 0.16904377937316895, 0.03570611774921417, -0.11537586152553558, -0.2564198076725006, 0.31084921956062317, 0.31110742688179016, -0.011715319007635117, 0.26380234956741333, 0.7341855764389038, 0.027994778007268906, -0.15276385843753815, -0.19830745458602905, 0.02862986922264099, -0.02416124939918518, -0.7314627766609192, 0.3125096559524536, -0.3160664141178131, -0.36789238452911377, 0.020079785957932472, 0.5516512989997864, -0.23535113036632538, 0.1846444010734558, -0.185321643948555, -0.2600320875644684, -0.1918117254972458, 0.2942541241645813, -0.06804841756820679, 0.0068130940198898315, -0.4693588614463806, -0.10694636404514313, 0.029502471908926964, 0.1694573611021042, -0.2477700561285019, 0.035324156284332275, -0.11774788051843643, 0.36261871457099915, -0.046685829758644104, -0.005117494612932205, -0.004681721329689026, -0.25742170214653015, 0.010051561519503593, 0.5666002035140991, -0.12076975405216217, -0.1419093906879425, -0.28333717584609985, 0.14587776362895966, 0.08442181348800659, -0.1702071577310562, 0.16537953913211823, -0.017817551270127296, 0.06224697083234787, -0.2151457965373993, -0.07266784459352493, -0.2080886960029602, -0.18448589742183685, 0.07484998553991318, -0.021099571138620377, 0.24461886286735535, -0.2342158854007721, 0.08050147444009781, -0.07008995115756989, 0.06816443800926208, -0.3697085678577423, 0.007148392498493195, -0.014693065546452999, 0.0024236664175987244, -0.035316575318574905, 0.15229377150535583, -0.1864626258611679, -0.11169946938753128, 0.2823203206062317, -0.2953430414199829, -0.12911133468151093, -0.005005622282624245, -0.19424118101596832, 0.3379955291748047, -0.15246900916099548, 0.0480976402759552, -0.04230756685137749, 0.11493928730487823, 0.1366555243730545, 0.35715654492378235, -0.2620999217033386, 0.021108441054821014, -0.08318209648132324, 0.27575427293777466, 0.12450669705867767, -0.22823961079120636, -0.1093328595161438, 0.06607528775930405, -0.06436733156442642, -0.36997437477111816, -0.13122707605361938, 0.1212930977344513, -0.1415523886680603, -0.0027165487408638, 0.3157489597797394, -0.15956252813339233, -0.039288513362407684, -0.002072412520647049, 0.3984375298023224, 0.3103441596031189, 0.5081055760383606, 0.20689107477664948, 0.0034972578287124634, 0.09589074552059174, 0.008702009916305542, 0.24191135168075562, 0.34572988748550415, 0.36158615350723267, 0.2970229983329773, 0.6127428412437439, -0.3836427628993988, -0.2681463360786438, -0.026009079068899155, 0.10279293358325958, -0.2622458040714264, -0.1417003720998764, -0.5835599303245544, -0.12105023860931396, -0.44204020500183105, -0.08055077493190765, -0.2170771062374115, 0.19709840416908264, 0.5802704095840454, 0.21965691447257996, -0.04547487199306488, -0.6290245056152344, -0.20984312891960144, 0.058179207146167755, 0.5017368197441101, -0.2239837348461151, 0.45599672198295593, 0.016069352626800537, -0.05682431161403656, -0.018257025629281998, 0.5782515406608582, 0.43931031227111816, 0.5464798808097839, -0.3213457763195038, -0.013242482207715511, 0.05908244475722313, -0.17023105919361115, 0.28284281492233276, 0.6259995102882385, 0.37193670868873596, 0.07206597179174423, 0.049773093312978745, -0.028637081384658813, -0.09993836283683777, -0.26097196340560913, 0.021830879151821136, 0.4561338722705841, -0.009454039856791496, 0.21890728175640106, -0.29675382375717163, -0.13070404529571533, -0.07994214445352554, -0.24563893675804138, -0.30123287439346313, -0.07740287482738495, 0.13583429157733917, 0.01643802411854267, 0.020510993897914886, -0.34253883361816406, 0.07126212120056152, 0.16037490963935852, 0.5518673062324524, 0.2615603506565094, 0.12667232751846313, -0.4171103239059448, -0.11879702657461166, -0.5855780243873596, 0.2008875161409378, -0.23853275179862976, -0.1507560908794403, -0.09952849894762039, 0.3261135518550873, 0.10537063330411911, 0.24450406432151794, -0.027682185173034668, 0.20349770784378052, 0.33210206031799316, 0.3179226815700531, -0.3015897274017334, -0.06234607473015785, 0.3174629807472229, -0.3273674547672272, 0.00846301019191742, -0.5750225186347961, 0.2966243624687195, 0.06786385178565979, 0.06840582191944122, 0.1313176155090332, -0.301472008228302, 0.20186860859394073, 0.4438701570034027, 0.3714682459831238, 0.2160162627696991, 0.25399476289749146, 0.25643861293792725, -0.4523237645626068, -0.26916107535362244, -0.20937898755073547, -0.23863396048545837, 0.10153631865978241, 0.10521870851516724, 0.5344887971878052, -0.08086100220680237, 0.12656208872795105, -0.25590893626213074, 0.1533980369567871, -0.17736376821994781, 0.1092863604426384, -0.2637517750263214, 0.1466660499572754, -0.3174295723438263, 0.08098757266998291, 0.12022750079631805, 0.12272705882787704, 0.15131545066833496, 0.14416126906871796, -0.4850180745124817, -0.4537811875343323, 0.3317626416683197, -0.3799084722995758, -0.334623783826828, 0.3907817304134369, 0.21516934037208557, 0.43962419033050537, -0.1042729839682579, -0.7881215214729309, -0.35706236958503723, 0.1472138911485672, -0.13919532299041748, -0.050900354981422424, -0.11611431837081909, -0.05664430931210518, -0.005235496908426285, -0.21112194657325745, 0.3439016342163086, -0.008731506764888763, 0.0837300568819046, 0.09028655290603638, -0.16827616095542908 ]
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.
[ 0.09355707466602325, -0.05218075215816498, 0.09173013269901276, 0.49227476119995117, 0.2525184750556946, 0.1492106020450592, 0.10871350765228271, 0.4077000319957733, 0.1446879804134369, -0.0935792624950409, 0.3454059362411499, 0.5846510529518127, -0.12629714608192444, -0.08148074150085449, -0.03906689211726189, -0.18967239558696747, 0.15724416077136993, 0.2686881422996521, 0.08629145473241806, -0.06003156676888466, -0.14310383796691895, 0.15054133534431458, -0.27892497181892395, 0.1321714222431183, -0.4200155735015869, -0.10315186530351639, 0.07676132768392563, 0.0571163073182106, -0.08841769397258759, -0.45285388827323914, 0.07614743709564209, -0.3256979286670685, -0.00657503679394722, 0.2591463327407837, -0.00012003633310087025, -0.12074834853410721, 0.41301560401916504, -0.04033137857913971, -0.26557812094688416, -0.24955396354198456, -0.435833215713501, 0.03188975900411606, 0.33162352442741394, -0.04249082878232002, 0.18057310581207275, -0.06374594569206238, 0.02380255237221718, -0.5875162482261658, 0.20610764622688293, 0.5833624601364136, 0.17648330330848694, 0.2174154371023178, -0.06840868294239044, 0.130262553691864, 0.29421332478523254, 0.05315934866666794, -0.33830970525741577, -0.08270286023616791, 0.022632073611021042, -0.12178804725408554, -0.13316559791564941, 0.3214312493801117, -0.06175166368484497, 0.058570221066474915, 0.6247357726097107, 0.0932791531085968, 0.13128316402435303, -0.3116700053215027, 0.14749667048454285, 0.14453640580177307, 0.39628151059150696, -0.08866610378026962, -0.08990984410047531, -0.01776956021785736, -0.1631641685962677, -0.3528531789779663, 0.2785201370716095, 0.18064822256565094, -0.29843685030937195, -0.0033445656299591064, 0.07484915107488632, 0.14837554097175598, -0.08723288029432297, -0.06426304578781128, -0.11870856583118439, 0.15310712158679962, 0.039316605776548386, 0.06686507165431976, -0.10409830510616302, -0.20409660041332245, -0.10837486386299133, -0.23340731859207153, -0.2524468004703522, 0.3161037564277649, -0.4382798671722412, 0.037972576916217804, -0.0514424592256546, -0.13796302676200867, -0.14761598408222198, 0.06295932829380035, 0.38411492109298706, 0.09981708228588104, 0.027438730001449585, 0.0716606080532074, 0.3587127923965454, 0.3659059405326843, 0.31820422410964966, 0.13503791391849518, 0.06323648989200592, -0.22235745191574097, -0.08160152286291122, 0.017592262476682663, 0.01670028269290924, -0.03930993750691414, 0.40901291370391846, 0.009138813242316246, 0.14973634481430054, -0.33181336522102356, -0.2487674355506897, 0.058740902692079544, -0.7114651799201965, 0.2303270548582077, -0.19834312796592712, 0.09589575976133347, -0.024717159569263458, 0.5226476788520813, 0.2696346640586853, 0.3240772485733032, -0.2275407910346985, -0.10159376263618469, -0.14699961245059967, -0.13756303489208221, -0.24351471662521362, -0.04535190016031265, 0.21126894652843475, -0.07352477312088013, 0.16572391986846924, 0.024609260261058807, -0.4914286732673645, -0.20295216143131256, 0.26686957478523254, -0.3210102915763855, 0.17625238001346588, 0.0648171454668045, -0.32960668206214905, 0.45327848196029663, 0.1861165463924408, -0.39012306928634644, -0.18190482258796692, 0.10338769853115082, -0.19294854998588562, 0.09501944482326508, -0.15740565955638885, 0.12872956693172455, 0.024879056960344315, -0.14852550625801086, -0.30236488580703735, -0.032280486077070236, 0.3422371447086334, -0.11202162504196167, 0.15281787514686584, -0.24676944315433502, -0.10657308995723724, -0.2501501142978668, -0.06317590177059174, 0.40901368856430054, -0.19291436672210693, 0.08677572011947632, -0.018419388681650162, 0.14051750302314758, 0.2578383982181549, 0.2855258882045746, 0.016143830493092537, 0.12157110869884491, 0.06346698850393295, 0.39460426568984985, 0.43152397871017456, -0.27908146381378174, 0.00753962155431509, 0.13285386562347412, -0.03557674586772919, -0.01185564324259758, 0.0022213980555534363, 0.106452576816082, 0.033090755343437195, -0.2236127406358719, 0.10748087614774704, 0.19908706843852997, -0.06461169570684433, 0.2510938048362732, -0.4027029275894165, -0.06630875915288925, 0.49514514207839966, -0.20197103917598724, -0.18115761876106262, -0.05476035922765732, -0.05374499410390854, 0.1514836549758911, -0.014162302017211914, -0.028817620128393173, 0.38277533650398254, 0.3288469910621643, -0.14390335977077484, -0.16071803867816925, 0.1006358191370964, -0.10738269984722137, -0.46785426139831543, 0.011794470250606537, -0.18902337551116943, 0.2026108205318451, -0.08662518858909607, 0.028060778975486755, -0.3410986661911011, 0.12534892559051514, -0.10165326297283173, -0.2273106873035431, 0.14942359924316406, 0.054173775017261505, 0.16102643311023712, 0.14653602242469788, -0.12235642969608307, 0.11216219514608383, 0.013422539457678795, 0.08212637901306152, -0.46395158767700195, -0.08359160274267197, 0.0000034905970096588135, -0.2866653800010681, -0.08776189386844635, 0.2652673125267029, 0.27048325538635254, 0.11229990422725677, -0.2972472012042999, 0.21523022651672363, 0.11378934979438782, 0.27887803316116333, -0.40211978554725647, 0.07179568707942963, 0.06320392340421677, -0.42975902557373047, 0.1421480029821396, 0.4353228509426117, 0.36861947178840637, -0.34248340129852295, -0.38880297541618347, -0.04157602787017822, 0.022348709404468536, 0.680326521396637, -0.13089539110660553, 0.14410719275474548, -0.03299497812986374, -0.03734292834997177, 0.07075367867946625, -0.05491744726896286, 0.16543735563755035, 0.01798594556748867, 0.12332215905189514, 0.3098255395889282, -0.24867242574691772, 0.29118919372558594, 0.37837231159210205, 0.2824706435203552, 0.11098349094390869, 0.12216784060001373, -0.385292649269104, -0.21817061305046082, 0.31259840726852417, 0.025050241500139236, 0.375763863325119, -0.05269097909331322, 0.18641407787799835, -0.05416803061962128, -0.1827584058046341, -0.17576918005943298, 0.35379332304000854, -0.05842497944831848, 0.1938471794128418, 0.03641253709793091, 0.5011975765228271, -0.015893951058387756, -0.13138148188591003, 0.24980765581130981, -0.008867017924785614, 0.20772449672222137, -0.4109279215335846, 0.3374977111816406, -0.4077056050300598, -0.13143034279346466, -0.37055841088294983, -0.15277180075645447, -0.02439647540450096, -0.6165551543235779, -0.04489489644765854, -0.018378648906946182, -0.030627353116869926, 0.12242846935987473, -0.28436070680618286, -0.13166259229183197, 0.013293301686644554, -0.0776517242193222, -0.4024278521537781, -0.42999592423439026, -0.4094288945198059, -0.04795015603303909, 0.43011704087257385, 0.392078697681427, 0.10363443940877914, 0.03218571096658707, -0.10713545233011246, -0.03624807298183441, -0.412510484457016, 0.03437487781047821, -0.2577618956565857, 0.09031230211257935, 0.15104316174983978, 0.06090404838323593, 0.01968567818403244, -0.0868513435125351, 0.32464173436164856, 0.0006716586649417877, -0.3420940041542053, 0.39525100588798523, 0.053129956126213074, 0.14244522154331207, -0.3051873445510864, -0.4616950452327728, 0.039699941873550415, -0.4655473232269287, 0.1820293664932251, 0.006348218768835068, 0.18158909678459167, 0.23612701892852783, 0.05442696437239647, 0.28018122911453247, -0.04302285239100456, 0.1691674292087555, 0.19449341297149658, 0.47960275411605835, 0.5730381011962891, -0.14828313887119293, -0.36124667525291443, 0.2676405906677246, -0.13377967476844788, 0.2956109642982483, 0.17932796478271484, -0.15094295144081116, 0.3191587030887604, -0.3194129765033722, 0.15346959233283997, -0.13758273422718048, -0.12629760801792145, 0.2398141473531723, 0.1834496110677719, -0.04229190945625305, -0.10745668411254883, 0.03156432509422302, 0.25976672768592834, -0.18653300404548645, 0.1925482153892517, 0.1195826455950737, 0.20338138937950134, -0.0464816614985466, 0.3153981566429138, 0.1880693882703781, -0.2586812973022461, 0.33788976073265076, -0.03515106439590454, -0.008244846016168594, -0.28940531611442566, -0.10925959050655365, -0.08718876540660858, -0.14598098397254944, 0.15237297117710114, -0.08302415907382965, 0.07133284211158752, -0.4544132947921753, -0.02870737574994564, 0.2991611063480377, -0.1059684082865715, -0.14508497714996338, 0.250357985496521, -0.2715938091278076, 0.2326233685016632, 0.07304367423057556, 0.15341895818710327, -0.11898380517959595, -0.5184164047241211, 0.04418310150504112, -0.031994059681892395, 0.09283921867609024, -0.020921939983963966, -0.20190303027629852, 0.017810538411140442, -0.4196455180644989, 0.28542864322662354, 0.168317511677742, 0.4798279106616974, -0.04797328636050224, -0.19165828824043274, 0.1719038486480713, -0.15335027873516083, 0.3857506513595581, -0.2743549942970276, -0.01348680630326271, 0.36245664954185486, 0.36000773310661316, -0.48109689354896545, 0.26782646775245667, -0.2580377459526062, 0.26213786005973816, 0.17987318336963654, 0.19627155363559723, -0.41516295075416565, -0.30158761143684387, 0.14041119813919067, 0.19144894182682037, -0.021241571754217148, 0.16336633265018463, -0.2792511582374573, -0.023450009524822235, -0.06628596782684326, -0.17773979902267456, 0.2659930884838104, 0.3581661880016327, 0.03981295973062515, -0.0013968013226985931, -0.433009535074234, 0.11331077665090561, 0.1220482736825943, 0.020940713584423065, 0.4554500877857208, -0.29696622490882874, 0.12399284541606903, -0.20001757144927979, 0.16423341631889343, -0.06789224594831467, 0.34366923570632935, -0.06636166572570801, -0.0957203358411789, 0.30548208951950073, -0.12876884639263153, 0.3905342221260071, 0.3394402861595154, -0.02647443488240242, 0.3637796938419342, -0.17648416757583618, -0.1536535620689392, -0.2444365918636322, -0.025315474718809128, 0.14097058773040771, -0.0874008983373642, 0.10841048508882523, -0.19618093967437744, 0.5411815643310547, 0.3007131814956665, -0.03165445849299431, 0.4439757466316223, 0.18964070081710815, -0.32779401540756226, 0.3507964611053467, 0.21915343403816223, 1.112839698791504, 0.11890722066164017, 0.17537809908390045, 0.08425121009349823, -0.3473072648048401, 0.4513011574745178, 0.022426187992095947, 0.25419673323631287, -0.48580291867256165, 0.25739234685897827, -0.03554876148700714, -0.1875910460948944, 0.040305718779563904, 0.02434402145445347, -0.28639018535614014, 0.1932094246149063, -0.17377714812755585, 0.26287317276000977, 0.12903322279453278, 0.01991180330514908, -0.2860623300075531, -0.33544138073921204, -0.4839944839477539, 0.07514122128486633, -0.15866316854953766, -0.0643378347158432, 0.10971181094646454, -0.2614781856536865, -0.4933289885520935, -0.09190820157527924, -0.1271524727344513, 0.14361169934272766, -0.24136842787265778, -0.062289826571941376, 0.25829100608825684, -0.17976704239845276, 0.22817057371139526, 0.16256220638751984, -0.04886312782764435, -0.0574195496737957, -0.18875814974308014, -0.11977651715278625, 0.20588132739067078, 0.09073035418987274, 0.1580936759710312, -0.019870396703481674, 0.14279145002365112, 0.058030568063259125, 0.09111496061086655, 0.2642859220504761, -0.40775418281555176, -0.0016342736780643463, 0.06845372915267944, 0.16408002376556396, 0.03515498340129852, -0.021138448268175125, -0.12933418154716492, 0.10254503041505814, -0.01636495254933834, -0.03418709710240364, 0.0898296982049942, 0.04698125272989273, 0.08257467299699783, 0.05820866674184799, -0.10640527307987213, -0.22666838765144348, 0.2634595036506653, 0.3531939685344696, 0.05719932168722153, 0.14272743463516235, 0.7486613988876343, 0.1637575626373291, -0.1326281875371933, -0.19221460819244385, 0.08599641174077988, 0.018571345135569572, -0.7875833511352539, 0.28427761793136597, -0.27142956852912903, -0.3222104012966156, 0.03584709018468857, 0.6465661525726318, -0.23184804618358612, 0.08952032029628754, -0.16638681292533875, -0.2611517906188965, -0.17477571964263916, 0.17243973910808563, -0.03713047131896019, 0.061161190271377563, -0.509809136390686, -0.18457816541194916, 0.10252305865287781, 0.131935253739357, -0.21777275204658508, 0.0001780930906534195, -0.2947481870651245, 0.3535666763782501, -0.01486620306968689, 0.10168533027172089, 0.005318854004144669, -0.17168745398521423, -0.004655670374631882, 0.4534410834312439, -0.1637963354587555, -0.1482589691877365, -0.23846150934696198, 0.15354639291763306, 0.11453447490930557, -0.11689335107803345, 0.06979884952306747, -0.02155900187790394, 0.0305703766644001, -0.21219202876091003, -0.10736984014511108, -0.2103567123413086, -0.17519178986549377, -0.013519302941858768, 0.0763074979186058, 0.26232606172561646, -0.27176356315612793, 0.12272563576698303, -0.17236706614494324, 0.07838184386491776, -0.3097918927669525, 0.02072773315012455, 0.11779609322547913, 0.02143005281686783, -0.14642228186130524, 0.10597573220729828, -0.16352449357509613, -0.2105933129787445, 0.3242960274219513, -0.3493191599845886, -0.2074197381734848, 0.04359646514058113, -0.15914978086948395, 0.39598068594932556, -0.10549262166023254, 0.002634037286043167, -0.07950562983751297, 0.0834643617272377, 0.05427968502044678, 0.322162389755249, -0.19346070289611816, -0.11521845310926437, -0.04124796763062477, 0.2851087749004364, 0.1726256161928177, -0.21538770198822021, -0.026877500116825104, 0.11490350216627121, 0.02829228714108467, -0.41845881938934326, -0.08133640885353088, 0.18888001143932343, -0.048989567905664444, 0.01914885640144348, 0.26837441325187683, -0.19589760899543762, 0.02850089967250824, -0.07040534913539886, 0.43938636779785156, 0.4198613166809082, 0.4042418301105499, 0.16703370213508606, -0.024891190230846405, 0.08554995059967041, 0.054913394153118134, 0.3567952811717987, 0.34315094351768494, 0.21554216742515564, 0.25422748923301697, 0.5395474433898926, -0.17617838084697723, -0.26875847578048706, -0.05656963586807251, 0.12042126059532166, -0.22145402431488037, -0.13927096128463745, -0.4548976421356201, -0.15396666526794434, -0.48004013299942017, -0.2001102864742279, -0.16493968665599823, 0.13815337419509888, 0.568152666091919, 0.2202722281217575, 0.04264826327562332, -0.5250182747840881, -0.13786688446998596, 0.10644260793924332, 0.3856297731399536, -0.17447832226753235, 0.46047085523605347, 0.006540074944496155, -0.0706382691860199, -0.030537154525518417, 0.42286744713783264, 0.42817622423171997, 0.5327762961387634, -0.3272397220134735, -0.06885861605405807, 0.1260795146226883, -0.16419416666030884, 0.240712970495224, 0.5911158323287964, 0.2477376013994217, 0.06603623181581497, 0.06582185626029968, -0.05595836788415909, -0.08274547755718231, -0.10225661098957062, -0.022089514881372452, 0.34112438559532166, -0.12443605065345764, 0.421477735042572, -0.27773118019104004, -0.14602136611938477, -0.10044369101524353, -0.1861693561077118, -0.2864457666873932, 0.032042890787124634, 0.10367287695407867, 0.03099707141518593, 0.07790986448526382, -0.3275088369846344, 0.06784392893314362, 0.18376044929027557, 0.5826796889305115, 0.2241734117269516, 0.10328647494316101, -0.3924229145050049, -0.09169905632734299, -0.7618374824523926, 0.1753283143043518, -0.2416837513446808, -0.21087603271007538, 0.10276053845882416, 0.302072137594223, 0.08311302959918976, 0.30485838651657104, -0.04291003942489624, 0.27844712138175964, 0.252066433429718, 0.28670433163642883, -0.38265180587768555, -0.1970774233341217, 0.34452682733535767, -0.28109580278396606, -0.052883341908454895, -0.5603267550468445, 0.23074865341186523, 0.016348760575056076, 0.058767423033714294, 0.03245602548122406, -0.2147497981786728, 0.18569090962409973, 0.44472193717956543, 0.20759908854961395, 0.18357205390930176, 0.17585764825344086, 0.27285701036453247, -0.4320273995399475, -0.3502245247364044, -0.2179291844367981, -0.19331367313861847, 0.03420940786600113, 0.03560404106974602, 0.4834500849246979, -0.16845622658729553, 0.03575201705098152, -0.27571359276771545, 0.12094943970441818, -0.13554725050926208, 0.13369105756282806, -0.2686518430709839, 0.1832183301448822, -0.41397562623023987, 0.19967903196811676, 0.22928164899349213, 0.022406067699193954, 0.1284518539905548, 0.14924879372119904, -0.4297788143157959, -0.49896374344825745, 0.2569379210472107, -0.4704397916793823, -0.3264240026473999, 0.28185033798217773, 0.1456383913755417, 0.3511766791343689, -0.06056024134159088, -0.740278959274292, -0.2556817829608917, 0.2354605495929718, -0.12984564900398254, 0.017962586134672165, -0.04342929646372795, -0.0465824231505394, 0.026109915226697922, -0.23466283082962036, 0.23958729207515717, 0.0006647706031799316, 0.1402987837791443, 0.033131200820207596, -0.2577146291732788 ]
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.
[ -0.06821640580892563, 0.06093105673789978, 0.1053825318813324, 0.4737013578414917, 0.2520800828933716, 0.10438019782304764, 0.09826843440532684, 0.3812301754951477, 0.22328925132751465, -0.06250713765621185, 0.32705157995224, 0.4799831509590149, -0.08507631719112396, -0.022185668349266052, -0.04573952779173851, -0.1310695856809616, 0.16129447519779205, 0.32369378209114075, 0.06354093551635742, -0.12184248864650726, -0.1438349485397339, 0.08432061970233917, -0.23519453406333923, 0.07620376348495483, -0.5430987477302551, -0.16950677335262299, 0.05933573842048645, 0.10318537801504135, -0.15469522774219513, -0.4448223412036896, 0.014604160562157631, -0.2505454123020172, 0.13717079162597656, 0.31009209156036377, -0.00011774222366511822, -0.13116741180419922, 0.38236409425735474, -0.04525517672300339, -0.21709251403808594, -0.1553119570016861, -0.4626956284046173, 0.060809992253780365, 0.26330962777137756, -0.06844918429851532, 0.19009876251220703, -0.0022964784875512123, 0.06410683691501617, -0.4999206066131592, 0.10783122479915619, 0.5157179832458496, 0.2082083821296692, 0.2796901762485504, -0.01879967376589775, 0.03328923508524895, 0.32798734307289124, 0.021057460457086563, -0.36564770340919495, 0.0005636103451251984, -0.028315484523773193, -0.15549598634243011, -0.16190847754478455, 0.3987066149711609, -0.08816113322973251, 0.05234827101230621, 0.6084595322608948, 0.09695272892713547, 0.11834555864334106, -0.2912595272064209, 0.1959603726863861, 0.12539079785346985, 0.4451828598976135, -0.10498585551977158, -0.1137467697262764, -0.003936167806386948, -0.134548157453537, -0.39042460918426514, 0.23909559845924377, 0.19062307476997375, -0.25099968910217285, -0.0031804293394088745, 0.05959666147828102, 0.10651848465204239, -0.03581857308745384, -0.19298137724399567, -0.06624897569417953, 0.20389685034751892, 0.06159105896949768, 0.0426708459854126, -0.10467115044593811, -0.15812568366527557, -0.01255903672426939, -0.23138535022735596, -0.3124716281890869, 0.2501974403858185, -0.3363650441169739, 0.042031340301036835, 0.009231436997652054, -0.10786857455968857, -0.1433403491973877, -0.07465694099664688, 0.3727552592754364, 0.023687565699219704, 0.13817401230335236, 0.1114930510520935, 0.3602553904056549, 0.3803686499595642, 0.3734177052974701, 0.13893957436084747, 0.172773540019989, -0.25541582703590393, -0.20996034145355225, -0.018515706062316895, -0.025886449962854385, -0.026921186596155167, 0.4833763837814331, -0.018194954842329025, 0.1269228756427765, -0.34274405241012573, -0.2864153981208801, 0.037969790399074554, -0.6072550415992737, 0.1798172891139984, -0.216493159532547, 0.1809121072292328, 0.014957142062485218, 0.44045618176460266, 0.1411498636007309, 0.31975412368774414, -0.2866179943084717, -0.2129357010126114, -0.16906782984733582, -0.06719328463077545, -0.35299694538116455, -0.047406695783138275, 0.23163442313671112, 0.025345204398036003, 0.1709941029548645, 0.045751430094242096, -0.5268831253051758, -0.19393058121204376, 0.27525487542152405, -0.33649227023124695, 0.1010013073682785, 0.09800000488758087, -0.1817556768655777, 0.3801156282424927, 0.15239638090133667, -0.39502614736557007, -0.20507794618606567, 0.07577154040336609, -0.14228400588035583, -0.014657892286777496, -0.14263604581356049, 0.16258682310581207, 0.019487731158733368, -0.22843661904335022, -0.28548699617385864, -0.04417174682021141, 0.37683555483818054, -0.08247704803943634, 0.033103182911872864, -0.19425195455551147, -0.07683172076940536, -0.2546737790107727, -0.14971096813678741, 0.49143582582473755, -0.17050649225711823, 0.16371475160121918, 0.052067410200834274, 0.1588343381881714, 0.32009202241897583, 0.34647512435913086, -0.05128083378076553, 0.23606014251708984, 0.07445212453603745, 0.32859450578689575, 0.4783320128917694, -0.27670979499816895, 0.05458474159240723, 0.14386886358261108, -0.04974491149187088, -0.014119445346295834, 0.0557805560529232, 0.004746994003653526, 0.1131347045302391, -0.186081200838089, 0.11334642767906189, 0.17492221295833588, -0.10872144997119904, 0.2992888391017914, -0.44598719477653503, -0.049678459763526917, 0.3986054062843323, -0.2075345367193222, -0.08162309229373932, 0.059087470173835754, 0.010010667145252228, 0.055392369627952576, 0.017926130443811417, -0.08591845631599426, 0.3454481363296509, 0.2982359528541565, -0.12419264018535614, -0.10706953704357147, 0.08703393489122391, -0.0994408056139946, -0.4447139501571655, 0.03243211656808853, -0.15776506066322327, 0.12304919958114624, 0.05034473165869713, -0.09754765033721924, -0.294474333524704, 0.09615029394626617, -0.03822718933224678, -0.25051984190940857, 0.15453094244003296, 0.023196179419755936, 0.1731850802898407, 0.05925080552697182, -0.20210525393486023, 0.0661349892616272, -0.015744267031550407, 0.10741614550352097, -0.3512536287307739, -0.17887350916862488, -0.04425763338804245, -0.3509325087070465, -0.03988923504948616, 0.2398829162120819, 0.23701804876327515, 0.10021079331636429, -0.2585672438144684, 0.23891958594322205, -0.0016572964377701283, 0.3618732690811157, -0.24295124411582947, 0.08757425099611282, -0.04195791482925415, -0.26053377985954285, 0.1956302523612976, 0.2751908302307129, 0.35893163084983826, -0.2975814938545227, -0.4221460223197937, -0.006539434194564819, 0.03794514760375023, 0.5811728835105896, -0.044700462371110916, 0.1357862651348114, 0.03368104249238968, -0.07959751784801483, 0.13727305829524994, -0.10519817471504211, 0.17800775170326233, -0.07357253134250641, 0.06387844681739807, 0.3371363878250122, -0.2674139440059662, 0.31074604392051697, 0.42854443192481995, 0.2990175485610962, 0.11164859682321548, 0.057120949029922485, -0.29339706897735596, -0.13686709105968475, 0.35193371772766113, 0.13247989118099213, 0.4555588364601135, -0.015705600380897522, 0.20980580151081085, -0.07270664721727371, -0.16493405401706696, -0.20099247992038727, 0.31755495071411133, -0.14931783080101013, 0.19674572348594666, 0.06621268391609192, 0.45941370725631714, 0.02002319134771824, -0.21954345703125, 0.21167489886283875, 0.03651811182498932, 0.21878327429294586, -0.37831512093544006, 0.28381994366645813, -0.2859777808189392, -0.0772416889667511, -0.4313698709011078, -0.042443785816431046, -0.046468060463666916, -0.5650829076766968, -0.03836654871702194, 0.007924221456050873, -0.09303827583789825, 0.03654531389474869, -0.32838574051856995, -0.11925627291202545, -0.12170131504535675, -0.06536591798067093, -0.2936857044696808, -0.4041465222835541, -0.3502984941005707, -0.05848420411348343, 0.4213869571685791, 0.38563215732574463, 0.1997774988412857, 0.016428174450993538, -0.005452672950923443, -0.028151260688900948, -0.4237631857395172, 0.04701822251081467, -0.22188512980937958, 0.04329712688922882, 0.16059285402297974, -0.005856461822986603, 0.05179901421070099, -0.10299589484930038, 0.3697010576725006, -0.05113134905695915, -0.25189265608787537, 0.3521765470504761, 0.02231493405997753, -0.03763393312692642, -0.26293817162513733, -0.4058386981487274, 0.11667194962501526, -0.5667976140975952, 0.10078822076320648, -0.04693658649921417, 0.20174640417099, 0.171672061085701, 0.08207504451274872, 0.2255024015903473, -0.06525570154190063, 0.19515232741832733, 0.2803915739059448, 0.37839382886886597, 0.5403062105178833, -0.22598065435886383, -0.32307562232017517, 0.29508623480796814, -0.1108960509300232, 0.3608185052871704, 0.23071272671222687, -0.1520203948020935, 0.4052870571613312, -0.4103546440601349, 0.17004787921905518, -0.0837431252002716, -0.11723372340202332, 0.31236863136291504, 0.19045503437519073, -0.06837323307991028, -0.09682268649339676, 0.017994578927755356, 0.21361678838729858, -0.20148390531539917, 0.1650698333978653, 0.008490161970257759, 0.199805349111557, -0.10333198308944702, 0.3400725722312927, 0.12449967861175537, -0.1656520813703537, 0.3614374101161957, 0.01160019263625145, 0.012079384177923203, -0.25102946162223816, -0.14403466880321503, 0.019972141832113266, -0.21486935019493103, 0.06162186712026596, -0.04792577028274536, 0.056543897837400436, -0.4713309407234192, -0.034646663814783096, 0.23681128025054932, -0.14764034748077393, -0.16390953958034515, 0.322564959526062, -0.3387589156627655, 0.2803706228733063, 0.09475952386856079, 0.1365748643875122, -0.11946560442447662, -0.4064585566520691, 0.023601878434419632, -0.05851206183433533, 0.1087665855884552, -0.016414202749729156, -0.26465630531311035, 0.0026100880932062864, -0.534706175327301, 0.24957682192325592, 0.05217341333627701, 0.39608776569366455, -0.04335755109786987, -0.18558365106582642, 0.18215234577655792, -0.22928622364997864, 0.40755343437194824, -0.3126874268054962, 0.018774807453155518, 0.2859611511230469, 0.2986101806163788, -0.4982847571372986, 0.18487998843193054, -0.2240564525127411, 0.330209344625473, 0.22574888169765472, 0.18554359674453735, -0.37205302715301514, -0.32857340574264526, 0.22981742024421692, 0.04473920911550522, -0.0006509236991405487, 0.10367845743894577, -0.35246750712394714, -0.05329139903187752, -0.06974142789840698, -0.14215591549873352, 0.2665811777114868, 0.3564813733100891, -0.07092612981796265, 0.0079500712454319, -0.4361250102519989, 0.1583191156387329, 0.15794898569583893, 0.13093774020671844, 0.3880293667316437, -0.3398948907852173, 0.12485630810260773, -0.16583502292633057, 0.19165833294391632, -0.04550173133611679, 0.42883315682411194, -0.2038113921880722, -0.1900620460510254, 0.3096029460430145, -0.174551859498024, 0.48622891306877136, 0.37902361154556274, 0.0421503409743309, 0.31145578622817993, -0.2637413740158081, -0.17829649150371552, -0.2773716449737549, -0.043513596057891846, 0.13077571988105774, -0.08891058713197708, 0.09123104065656662, -0.23243579268455505, 0.5531400442123413, 0.30162313580513, -0.05554824322462082, 0.4027084708213806, 0.18881957232952118, -0.35648322105407715, 0.4658268094062805, 0.2532043159008026, 1.1186977624893188, 0.10575829446315765, 0.22237133979797363, 0.025404172018170357, -0.40893757343292236, 0.35305461287498474, -0.03306626155972481, 0.2725175619125366, -0.49555620551109314, 0.2272232323884964, 0.027789248153567314, -0.1229366809129715, 0.07964393496513367, 0.05283215641975403, -0.27910733222961426, 0.15774770081043243, -0.1973516345024109, 0.3361303210258484, 0.07588654011487961, 0.12650950253009796, -0.4026109278202057, -0.34524786472320557, -0.48598968982696533, 0.08268747478723526, -0.083915576338768, -0.08008445799350739, 0.026806198060512543, -0.2240903526544571, -0.40694770216941833, -0.13697338104248047, -0.13860897719860077, 0.13444149494171143, -0.15800276398658752, -0.11935241520404816, 0.2568897604942322, -0.154481440782547, 0.1984211951494217, 0.22811612486839294, 0.0449160672724247, -0.050611916929483414, -0.16760903596878052, -0.014997154474258423, 0.27013760805130005, 0.11929114162921906, 0.14295688271522522, -0.027162950485944748, 0.0714477002620697, 0.04230794310569763, 0.08588605374097824, 0.2309129536151886, -0.4100569188594818, 0.011083267629146576, -0.12113723158836365, 0.20614740252494812, -0.023422762751579285, -0.13388662040233612, -0.08782537281513214, 0.11561565101146698, -0.034164153039455414, -0.08314493298530579, 0.11682294309139252, 0.02967080846428871, 0.11227241158485413, 0.09994056075811386, -0.15056638419628143, -0.2468956857919693, 0.22061574459075928, 0.3188552260398865, 0.05649012699723244, 0.23681804537773132, 0.6745470762252808, 0.08035561442375183, -0.14073994755744934, -0.23257210850715637, 0.04807336628437042, 0.051203567534685135, -0.7669647932052612, 0.33888331055641174, -0.2679664194583893, -0.40249454975128174, 0.13678410649299622, 0.5917763113975525, -0.1855551302433014, 0.013741280883550644, -0.09446403384208679, -0.2665630280971527, -0.09995357692241669, 0.19291746616363525, 0.03185880184173584, 0.07356494665145874, -0.4767534136772156, -0.12634290754795074, 0.07701265066862106, 0.22448205947875977, -0.2680851221084595, -0.04891285300254822, -0.23274098336696625, 0.37962889671325684, 0.014470759779214859, 0.08027612417936325, -0.0298212468624115, -0.15083584189414978, -0.006759891286492348, 0.4313046336174011, -0.14375334978103638, -0.151163250207901, -0.27181369066238403, 0.1370278000831604, 0.07310191541910172, -0.02483992651104927, 0.1579185277223587, 0.029813691973686218, -0.01240551471710205, -0.2556418180465698, -0.0019757207483053207, -0.1860363781452179, -0.2229127734899521, -0.028440099209547043, 0.05852131545543671, 0.2825692296028137, -0.2086717188358307, 0.12727847695350647, -0.1877681314945221, 0.07987599074840546, -0.33941370248794556, 0.07165061682462692, 0.07958433032035828, 0.015731222927570343, 0.03195758908987045, 0.09080922603607178, -0.14252540469169617, -0.07515721768140793, 0.31408578157424927, -0.3454221487045288, -0.10453855246305466, 0.11357147991657257, -0.11710676550865173, 0.3987210690975189, -0.08429927378892899, 0.014218473806977272, -0.026213794946670532, 0.13289308547973633, 0.07714072614908218, 0.257689893245697, -0.2710583806037903, -0.12639226019382477, -0.05107753723859787, 0.30452218651771545, 0.17767737805843353, -0.21350863575935364, -0.04058536887168884, 0.12846454977989197, 0.05426420271396637, -0.3916771411895752, -0.14858072996139526, 0.08976948261260986, -0.09947685152292252, 0.007058609277009964, 0.33671271800994873, -0.11197775602340698, 0.021539010107517242, 0.08279454708099365, 0.49896591901779175, 0.4582202136516571, 0.39912155270576477, 0.16864843666553497, -0.11093109101057053, -0.001728832721710205, 0.021545136347413063, 0.31710419058799744, 0.31534647941589355, 0.3045441806316376, 0.15489012002944946, 0.5104095935821533, -0.3058182895183563, -0.26720643043518066, -0.09200514853000641, 0.06214548274874687, -0.2543644905090332, -0.14574915170669556, -0.5064579248428345, -0.14190487563610077, -0.42316168546676636, -0.10091476142406464, -0.1865212768316269, 0.2660888731479645, 0.5050715208053589, 0.17831434309482574, 0.0029863715171813965, -0.5828507542610168, -0.056132711470127106, 0.031109150499105453, 0.3965829312801361, -0.24684061110019684, 0.501538097858429, -0.03485250473022461, -0.09850385785102844, -0.09822677075862885, 0.5166679620742798, 0.46027395129203796, 0.6580979228019714, -0.27573996782302856, -0.07016037404537201, 0.07553207874298096, -0.1338510811328888, 0.15214858949184418, 0.5606761574745178, 0.3276126980781555, 0.15973082184791565, 0.11919431388378143, -0.03441876545548439, -0.09463390707969666, -0.11874830722808838, -0.03437354043126106, 0.2678053081035614, -0.09244757890701294, 0.35593241453170776, -0.15395544469356537, -0.2433713674545288, -0.08671636879444122, -0.15515628457069397, -0.28839296102523804, -0.05635576695203781, 0.13199466466903687, -0.0714786946773529, 0.05464477092027664, -0.4272519052028656, 0.07915197312831879, 0.20936818420886993, 0.6322185397148132, 0.2378581017255783, 0.16091462969779968, -0.4584116041660309, -0.07986677438020706, -0.7588373422622681, 0.1408190131187439, -0.39341655373573303, -0.1612657904624939, 0.026848401874303818, 0.35101792216300964, -0.014347352087497711, 0.2293814867734909, -0.05591113865375519, 0.32126232981681824, 0.17826001346111298, 0.2982586920261383, -0.3454146683216095, -0.17916637659072876, 0.2205488085746765, -0.34437063336372375, -0.049058496952056885, -0.6017604470252991, 0.32700303196907043, 0.050085149705410004, 0.08173493295907974, -0.0072828978300094604, -0.2222326546907425, 0.3362940847873688, 0.2882280945777893, 0.32577580213546753, 0.16702570021152496, 0.2884238064289093, 0.19261911511421204, -0.45870640873908997, -0.39807483553886414, -0.19980329275131226, -0.30666303634643555, -0.09984276443719864, 0.13190922141075134, 0.469083696603775, -0.17963804304599762, 0.11805972456932068, -0.2406577318906784, 0.25994157791137695, -0.10598036646842957, 0.1154140830039978, -0.15053386986255646, 0.13514043390750885, -0.3236255943775177, 0.11830688267946243, 0.2612653076648712, 0.16057133674621582, 0.1331453025341034, 0.1277531236410141, -0.47373223304748535, -0.4435843527317047, 0.28088921308517456, -0.3823678195476532, -0.38338327407836914, 0.25261789560317993, 0.22172556817531586, 0.3979001045227051, -0.09128394722938538, -0.7310698628425598, -0.3031177520751953, 0.13606199622154236, -0.08944415301084518, 0.0044166408479213715, 0.06323646008968353, 0.07703427970409393, -0.0773245319724083, -0.21661366522312164, 0.28396859765052795, -0.055230170488357544, 0.026483627036213875, 0.05415119603276253, -0.25348347425460815 ]
https://github.com/huggingface/datasets/issues/207
Remove test set from NLP viewer
Appears that [two thirds of those polled on Twitter](https://twitter.com/srush_nlp/status/1265734497632477185) are in favor of _some_ mechanism for averting eyeballs from the test data.
While the new [NLP viewer](https://huggingface.co/nlp/viewer/) is a great tool, I think it would be best to outright remove the option of looking at the test sets. At the very least, a warning should be displayed to users before showing the test set. Newcomers to the field might not be aware of best practices, and small things like this can help increase awareness.
22
Remove test set from NLP viewer While the new [NLP viewer](https://huggingface.co/nlp/viewer/) is a great tool, I think it would be best to outright remove the option of looking at the test sets. At the very least, a warning should be displayed to users before showing the test set. Newcomers to the field might not be aware of best practices, and small things like this can help increase awareness. Appears that [two thirds of those polled on Twitter](https://twitter.com/srush_nlp/status/1265734497632477185) are in favor of _some_ mechanism for averting eyeballs from the test data.
[ -0.13736289739608765, 0.17048227787017822, -0.1900717169046402, -0.2922053933143616, -0.17911425232887268, 0.044915564358234406, 0.2525577247142792, 0.5567147731781006, 0.055269014090299606, 0.05124489217996597, 0.014377285726368427, 0.17538653314113617, -0.3614470660686493, 0.08514592051506042, 0.07604803889989853, 0.1644962877035141, -0.30261021852493286, 0.5070188045501709, -0.1310981959104538, 0.1275256872177124, -0.22304917871952057, -0.20796175301074982, 0.0001344829797744751, 0.24395689368247986, 0.18622954189777374, -0.39478161931037903, 0.18797406554222107, -0.16565582156181335, -0.019231760874390602, -0.5875657796859741, -0.10718861222267151, 0.20037740468978882, -0.17094139754772186, -0.11505144089460373, -0.00009701006638351828, 0.0429435595870018, 0.16814804077148438, 0.12773200869560242, -0.289050430059433, 0.10696420818567276, -0.12031254917383194, -0.05883173644542694, 0.2786576449871063, -0.13759134709835052, -0.08296464383602142, -0.22412899136543274, 0.1421549767255783, -0.22430917620658875, 0.10305432975292206, 0.3524733781814575, 0.3387940526008606, 0.3524862229824066, -0.30852365493774414, 0.43407294154167175, 0.13730277121067047, -0.011704564094543457, -0.27527016401290894, 0.04518921300768852, 0.03301137313246727, 0.015701495110988617, -0.39855605363845825, 0.5405020117759705, -0.06409768760204315, -0.007953528314828873, -0.038286879658699036, -0.17734374105930328, 0.3933626711368561, -0.2694503664970398, -0.30618664622306824, 0.0869026929140091, 0.1984504759311676, -0.19477087259292603, -0.06235070526599884, -0.41920164227485657, 0.23830965161323547, 0.04672989994287491, 0.09231428802013397, 0.2671627700328827, -0.05867631360888481, 0.23186297714710236, -0.5527652502059937, -0.41144225001335144, -0.1109342873096466, 0.0838954895734787, 0.14198283851146698, -0.10454480350017548, 0.08363662660121918, -0.05657775700092316, 0.3543230891227722, -0.08692210912704468, 0.17274439334869385, -0.05581090599298477, -0.2580600678920746, 0.014445632696151733, -0.005055133253335953, -0.5017728805541992, 0.0999649241566658, -0.01905730366706848, -0.03359732776880264, 0.20497824251651764, 0.08029212802648544, 0.10392996668815613, -0.20463961362838745, 0.06354296207427979, 0.10912415385246277, 0.12340101599693298, 0.35100409388542175, 0.062386415898799896, 0.2597549855709076, 0.24740107357501984, 0.2432093620300293, 0.30436408519744873, 0.33762866258621216, -0.251872718334198, -0.2752167880535126, 0.12666171789169312, 0.354028582572937, -0.23466157913208008, -0.3460922837257385, 0.003445904701948166, -0.5296180248260498, 0.014234960079193115, 0.014480089768767357, 0.10928130894899368, -0.0006174850277602673, 0.12058684229850769, 0.03531207889318466, 0.03173815459012985, -0.3619564175605774, -0.35781922936439514, -0.0544237419962883, 0.10798418521881104, -0.09068240970373154, 0.11934177577495575, 0.4057035446166992, 0.04340828210115433, 0.16046184301376343, -0.0342019721865654, 0.09332332015037537, 0.04058919847011566, 0.3698306381702423, 0.024624507874250412, 0.346337229013443, -0.05756784602999687, -0.1173800528049469, -0.21008704602718353, -0.2158709615468979, -0.14201414585113525, -0.18773341178894043, 0.15288540720939636, 0.09354298561811447, -0.20848321914672852, 0.05266881734132767, 0.3505091369152069, -0.4395666718482971, -0.014217622578144073, 0.09178121387958527, 0.234715074300766, -0.3213294446468353, -0.04506167769432068, 0.14012736082077026, -0.17458771169185638, -0.13457608222961426, 0.07385968416929245, -0.17334511876106262, 0.29054293036460876, -0.09705110639333725, -0.08341526985168457, -0.03997647017240524, -0.09570494294166565, -0.207576185464859, -0.02417227253317833, -0.4004930555820465, 0.17513087391853333, -0.20396791398525238, -0.06718962639570236, 0.36196768283843994, -0.2131473869085312, -0.10236697643995285, 0.20034706592559814, -0.23160767555236816, 0.016960090026259422, -0.14347735047340393, 0.07153680175542831, -0.16221846640110016, -0.4502266049385071, -0.47924771904945374, 0.07163726538419724, 0.060160815715789795, 0.14337174594402313, -0.44935938715934753, -0.4256094694137573, 0.49176400899887085, 0.045784737914800644, -0.05306442454457283, -0.13597598671913147, -0.1581975519657135, 0.25878408551216125, 0.43748700618743896, 0.028907788917422295, -0.039323121309280396, -0.23834066092967987, 0.20831498503684998, -0.3265897035598755, -0.2815675735473633, 0.007924303412437439, -0.1326737105846405, -0.16728752851486206, -0.08347122371196747, 0.1869812160730362, 0.42072296142578125, -0.28292417526245117, -0.0900668129324913, -0.09678641706705093, -0.15172302722930908, -0.09776139259338379, 0.29000505805015564, -0.1413697898387909, -0.09831617772579193, 0.1374543309211731, 0.09653125703334808, -0.15791693329811096, -0.17216792702674866, 0.03815504163503647, 0.02072005905210972, -0.17654575407505035, -0.05305419862270355, 0.10111168771982193, 0.3118337094783783, 0.3026479482650757, 0.07121899724006653, 0.06072896346449852, 0.07649918645620346, 0.17240792512893677, 0.06634168326854706, 0.10490579158067703, 0.2324661910533905, 0.2782416343688965, 0.07447272539138794, -0.44166046380996704, -0.014351686462759972, 0.1355339139699936, -0.151292085647583, 0.23665392398834229, -0.31678566336631775, 0.2563723921775818, 0.16144399344921112, -0.07265856117010117, 0.0031079649925231934, -0.029503941535949707, 0.04893171787261963, -0.3398032486438751, -0.13817624747753143, -0.337873250246048, -0.026473138481378555, 0.2268938571214676, -0.18623097240924835, 0.23691917955875397, -0.33784234523773193, 0.17116916179656982, 0.09643452614545822, -0.015973366796970367, -0.0260864719748497, 0.10548362880945206, 0.20493507385253906, -0.35236141085624695, 0.23605813086032867, 0.07625585794448853, 0.10919326543807983, 0.3495016396045685, 0.22472359240055084, 0.17186972498893738, -0.21009546518325806, -0.27264276146888733, 0.3151495158672333, 0.3302733898162842, 0.2265135943889618, -0.024842336773872375, 0.24631303548812866, -0.166645348072052, -0.2424010932445526, 0.07095293700695038, -0.10728862881660461, -0.05060543864965439, -0.3737708628177643, 0.02483482100069523, -0.1493522822856903, -0.5540044903755188, -0.1692960262298584, 0.19525045156478882, -0.0923478901386261, -0.25069016218185425, 0.5380775332450867, -0.028283800929784775, -0.3674883246421814, 0.2202185094356537, -0.2902728021144867, 0.6584694981575012, -0.20973871648311615, 0.22307725250720978, -0.06582289189100266, -0.020816585049033165, -0.36914482712745667, 0.30891114473342896, 0.17983661592006683, 0.060834262520074844, 0.522704541683197, 0.10138292610645294, -0.17568466067314148, -0.1762026697397232, -0.27201029658317566, 0.08650858700275421, -0.03296918421983719, 0.3253641426563263, -0.0012542828917503357, -0.1657099574804306, -0.04760973900556564, -0.14837469160556793, -0.4186188578605652, -0.22404444217681885, -0.00045916810631752014, -0.15232473611831665, 0.14578016102313995, 0.38489559292793274, -0.24102428555488586, -0.46456465125083923, 0.0017172731459140778, -0.2199234515428543, 0.020160745829343796, -0.014971038326621056, 0.2074323296546936, 0.20418484508991241, -0.26563623547554016, 0.0050034308806061745, -0.14018674194812775, -0.05355343222618103, -0.11937104165554047, 0.04450136795639992, 0.01233486458659172, -0.27020204067230225, -0.28076624870300293, 0.019245881587266922, -0.2571013271808624, 0.1463606357574463, -0.1737016886472702, -0.6786444783210754, -0.32688552141189575, 0.18293766677379608, -0.09719103574752808, -0.045817092061042786, -0.5455358624458313, -0.11407436430454254, 0.17513547837734222, -0.2621878385543823, 0.1255512535572052, 0.08653382956981659, 0.25623390078544617, 0.00540316104888916, 0.14536041021347046, -0.10717704147100449, -0.10614816099405289, 0.267967164516449, 0.15631680190563202, 0.2747737765312195, 0.11752210557460785, -0.007231287658214569, 0.0892392173409462, 0.26060837507247925, -0.008756168186664581, -0.035113535821437836, 0.04626890644431114, 0.12895257771015167, 0.015699461102485657, 0.33214300870895386, -0.11922281235456467, -0.313361257314682, 0.05188624560832977, 0.08164504170417786, -0.2134508490562439, -0.4037645161151886, 0.14978067576885223, -0.40173640847206116, 0.13901287317276, 0.12182234227657318, 0.04311314970254898, 0.06124310567975044, -0.5748977065086365, 0.05878446623682976, -0.19712994992733002, 0.2268815040588379, -0.06648248434066772, -0.20710493624210358, -0.013696662150323391, -0.2750268280506134, -0.011230707168579102, 0.14623209834098816, 0.3845880925655365, 0.010643396526575089, -0.2336488664150238, 0.24429184198379517, 0.1683444231748581, 0.21813812851905823, 0.06173047423362732, 0.20189140737056732, 0.13025835156440735, -0.12362942099571228, -0.3014462888240814, 0.1879303902387619, -0.21928927302360535, -0.09623371809720993, 0.15762069821357727, -0.027110785245895386, 0.02971985936164856, -0.24706478416919708, 0.36989444494247437, 0.12154660373926163, -0.23160691559314728, 0.024665754288434982, -0.013284511864185333, 0.010629601776599884, -0.037181828171014786, 0.2463870644569397, -0.1701396256685257, -0.02812224254012108, -0.36340099573135376, -0.10158498585224152, -0.024530136957764626, -0.20282796025276184, 0.37643951177597046, 0.28827911615371704, -0.01591498777270317, 0.2726737856864929, 0.0973345935344696, 0.21220341324806213, 0.039156991988420486, -0.19112005829811096, 0.1898479163646698, 0.247857928276062, 0.3424505591392517, -0.04484070837497711, -0.20925547182559967, 0.34741029143333435, 0.41228631138801575, 0.13903403282165527, 0.0013980045914649963, -0.3875180780887604, 0.057237036526203156, -0.2808181643486023, 0.2900629937648773, 0.07467538118362427, -0.174647718667984, 0.053832121193408966, -0.12565749883651733, 0.24363866448402405, -0.1218833178281784, -0.3639516234397888, 0.000638376921415329, 0.11595022678375244, -0.18264447152614594, 0.20929919183254242, 0.4568058252334595, 0.7854485511779785, 0.04531314969062805, 0.19970180094242096, 0.47228047251701355, -0.15780213475227356, 0.43638378381729126, -0.15434899926185608, 0.2508730888366699, -0.13412988185882568, 0.09534639865159988, 0.05953202396631241, -0.11783868074417114, 0.11199340224266052, 0.20780609548091888, 0.002054903656244278, 0.0006368251051753759, 0.26226550340652466, -0.35835036635398865, -0.1193864494562149, 0.32153481245040894, 0.04634018987417221, -0.021235276013612747, -0.30379799008369446, 0.2726384997367859, -0.16115741431713104, -0.20504175126552582, 0.10787323862314224, 0.028086042031645775, 0.3095525801181793, 0.04540547728538513, 0.18522445857524872, 0.13581514358520508, 0.11219034343957901, 0.1330631673336029, -0.13726158440113068, -0.18840216100215912, 0.39658981561660767, 0.23549309372901917, -0.17092248797416687, 0.12436376512050629, -0.08929335325956345, -0.012547902762889862, -0.1636733114719391, 0.29650095105171204, 0.2138882428407669, 0.2833039164543152, 0.3150736391544342, -0.19931212067604065, -0.1702399104833603, 0.26491525769233704, 0.19701620936393738, 0.05972404032945633, -0.08262807130813599, -0.0058592259883880615, 0.2142532765865326, -0.47217684984207153, 0.03237347677350044, 0.039203714579343796, 0.11395150423049927, -0.1505127102136612, 0.28724610805511475, 0.04401278495788574, -0.03829532116651535, 0.043228570371866226, 0.3496764302253723, -0.23045490682125092, 0.10065092146396637, 0.05798114836215973, 0.09893631935119629, 0.10476025938987732, 0.43976637721061707, -0.16813921928405762, -0.24870668351650238, -0.4134778082370758, 0.25973135232925415, -0.12139610201120377, -0.2759031653404236, 0.1033058911561966, 0.12412773072719574, -0.05992032587528229, -0.036587972193956375, 0.2192913293838501, -0.10416138172149658, -0.006942488253116608, -0.09856215864419937, 0.004728846251964569, 0.002826118841767311, 0.16214457154273987, -0.13531595468521118, 0.2648948132991791, 0.23671147227287292, 0.1758655607700348, -0.07974191009998322, 0.16821028292179108, -0.45865800976753235, -0.2357047200202942, -0.3629606366157532, 0.16979244351387024, -0.33876797556877136, 0.16333013772964478, 0.025976408272981644, 0.0013712691143155098, 0.131124347448349, -0.2586054801940918, -0.23900258541107178, -0.3924419581890106, -0.11555568873882294, 0.0577705055475235, -0.014545109122991562, -0.1714857965707779, 0.020736847072839737, 0.011869490146636963, 0.10929164290428162, 0.05694784224033356, -0.11095632612705231, -0.36126941442489624, 0.18716436624526978, 0.654991865158081, -0.22639794647693634, -0.37684378027915955, 0.04805293679237366, 0.0021645687520503998, -0.3364020586013794, 0.1804552525281906, 0.19234485924243927, 0.2569877505302429, 0.07886850833892822, -0.07736150175333023, 0.09211748093366623, 0.14207780361175537, -0.0038993731141090393, 0.2430170774459839, -0.22079411149024963, 0.05613260716199875, 0.016851454973220825, -0.029933013021945953, -0.025893239304423332, -0.027973540127277374, -0.30157986283302307, -0.29457831382751465, 0.2904887795448303, 0.3381895124912262, -0.3103622794151306, 0.027729328721761703, 0.32519036531448364, 0.048447202891111374, 0.07958593964576721, 0.2135290503501892, 0.08040469884872437, 0.029684752225875854, 0.29514917731285095, 0.0309763140976429, -0.00004234188236296177, -0.3247302174568176, 0.08860582113265991, 0.16464705765247345, -0.41913896799087524, -0.2750462591648102, 0.12339773774147034, 0.12206247448921204, 0.17781221866607666, -0.01930566132068634, 0.19298282265663147, 0.6014686822891235, -0.04290518909692764, 0.3327934443950653, -0.029134601354599, -0.04987681284546852, -0.06779876351356506, 0.006361827254295349, -0.1729620099067688, 0.08005881309509277, 0.37211331725120544, 0.37689298391342163, 0.2283657193183899, -0.12036129087209702, -0.06467597186565399, 0.04856038838624954, 0.017171749845147133, 0.17235639691352844, -0.27359190583229065, 0.11902842670679092, -0.09038843214511871, 0.13026049733161926, 0.07079996168613434, -0.2096400409936905, 0.10960590839385986, 0.11861185729503632, -0.10826557874679565, -0.10142333060503006, 0.1704074740409851, -0.036531589925289154, 0.2522592842578888, -0.09298397600650787, -0.19044820964336395, -0.0768471509218216, 0.3444373309612274, 0.1669951230287552, 0.3699305057525635, -0.14906473457813263, -0.18440492451190948, -0.02608916535973549, -0.18871204555034637, -0.09965525567531586, -0.009150639176368713, -0.3066669702529907, 0.2665650248527527, 0.07575131207704544, 0.1416868269443512, 0.25459739565849304, 0.21736104786396027, -0.09045848995447159, 0.0599961057305336, 0.3468180298805237, -0.22026780247688293, -0.26668524742126465, 0.26403459906578064, 0.5363548994064331, -0.33223026990890503, -0.06163088232278824, -0.11595634371042252, 0.05984329432249069, 0.10731910169124603, 0.27525317668914795, -0.24124692380428314, 0.09731181710958481, -0.2812128961086273, 0.15977293252944946, 0.27768322825431824, 0.43047958612442017, 0.0057656243443489075, 0.1721988022327423, -0.09262100607156754, -0.07269566506147385, -0.5211904644966125, 0.18362881243228912, 0.22089821100234985, -0.14337538182735443, 0.259044885635376, -0.026720775291323662, -0.18254798650741577, 0.2621276378631592, -0.3430914878845215, 0.19633996486663818, -0.03144615888595581, -0.22313153743743896, -0.2514569163322449, 0.035640161484479904, -0.03186886012554169, 0.2519475817680359, -0.1788475662469864, -0.10400594770908356, 0.20293313264846802, -0.18828251957893372, 0.15878048539161682, 0.06955939531326294, 0.08495213091373444, 0.14256064593791962, 0.28576359152793884, 0.281242311000824, 0.1692962348461151, 0.12046939134597778, -0.03955983370542526, 0.09596652537584305, 0.000599101185798645, -0.018906811252236366, -0.18427985906600952, -0.2055191695690155, 0.09908117353916168, 0.19717489182949066, -0.0958147719502449, 0.27070245146751404, 0.12599588930606842, 0.1675502359867096, -0.02738126367330551, 0.22046245634555817, -0.11364427953958511, 0.1647748202085495, 0.032816462218761444, 0.2928231358528137, -0.1284547597169876, 0.22897553443908691, -0.09584259986877441, -0.14297911524772644, -0.07893221825361252, -0.19231939315795898, 0.25615233182907104, -0.2987458407878876, -0.04730776697397232, 0.028888415545225143, 0.372744083404541, 0.00969204306602478, -0.31185051798820496, -0.45354217290878296, -0.2849319577217102, 0.24580822885036469, -0.08283773809671402, 0.24263238906860352, 0.44049474596977234, -0.11160415410995483, -0.14189913868904114, -0.08976626396179199, 0.3100309371948242, 0.0736428052186966, 0.13642023503780365, 0.07103091478347778, -0.3468834161758423 ]
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.
[ -0.24451451003551483, 0.06389839947223663, 0.024110831320285797, 0.36282336711883545, 0.10962121933698654, 0.4387524425983429, 0.26998746395111084, 0.3603740930557251, 0.044979143887758255, -0.11644783616065979, -0.381481409072876, 0.12374326586723328, 0.003628678619861603, 0.2719852328300476, 0.27131691575050354, -0.5642147064208984, 0.05804482102394104, 0.21875345706939697, -0.5200099945068359, 0.18503700196743011, 0.007575150579214096, 0.005126262083649635, -0.16544681787490845, 0.012677192687988281, -0.20304375886917114, 0.2349441945552826, -0.3422628939151764, -0.22793927788734436, 0.04638442397117615, 0.15664580464363098, 0.3019716739654541, 0.16707965731620789, 0.21462075412273407, 0.31491559743881226, -0.00011992744111921638, -0.07949517667293549, -0.016488976776599884, -0.09052366763353348, 0.14189808070659637, -0.19872920215129852, 0.09682495146989822, -0.36194995045661926, -0.06335476040840149, -0.15658241510391235, 0.08227722346782684, -0.23953083157539368, -0.23318354785442352, 0.03752853721380234, 0.26808595657348633, -0.16602979600429535, 0.05411152169108391, -0.188320592045784, -0.251149445772171, 0.3409510850906372, -0.07127415388822556, -0.16056816279888153, 0.0614653155207634, -0.06943392753601074, 0.08189091831445694, -0.31951451301574707, 0.5779258608818054, 0.2133333683013916, -0.03511062264442444, -0.11987727880477905, 0.08233247697353363, 0.14248356223106384, 0.1433926224708557, -0.1095086932182312, 0.05533512309193611, 0.19000259041786194, 0.7741914987564087, -0.05964938551187515, 0.008376194164156914, -0.29153186082839966, 0.4038733243942261, -0.1315470039844513, -0.19321182370185852, 0.1430453211069107, 0.017043307423591614, 0.025498170405626297, -0.25481942296028137, -0.15632006525993347, -0.0164065882563591, 0.3023141622543335, -0.04316413402557373, 0.33158251643180847, 0.11024920642375946, 0.27707964181900024, -0.06372199952602386, -0.1514376997947693, 0.16603240370750427, -0.33642664551734924, 0.000991474837064743, 0.15475766360759735, -0.3073637783527374, -0.4046189486980438, -0.3330455422401428, -0.18422913551330566, 0.27262577414512634, -0.1534881740808487, -0.15438945591449738, -0.006971031427383423, -0.1090388372540474, 0.13872379064559937, 0.30268222093582153, 0.15827219188213348, 0.13861002027988434, 0.21253788471221924, -0.18847210705280304, -0.31628331542015076, -0.3603151738643646, 0.16161024570465088, 0.07011684775352478, -0.0728301927447319, -0.1233942061662674, -0.2811368703842163, -0.058452386409044266, -0.34144553542137146, 0.1455402374267578, -0.19353191554546356, -0.27553874254226685, -0.2310830056667328, -0.10256724059581757, -0.15160977840423584, -0.0181959867477417, 0.3452557921409607, -0.1489933580160141, 0.3091169595718384, -0.10607438534498215, -0.013937316834926605, -0.025030506774783134, 0.07381550967693329, -0.17078666388988495, 0.12827026844024658, 0.09456299245357513, 0.1703295111656189, -0.13089489936828613, 0.41861796379089355, -0.18310464918613434, 0.10155941545963287, 0.32089948654174805, -0.19957442581653595, -0.07107561081647873, -0.24224181473255157, 0.3129124045372009, 0.16381879150867462, 0.1630842238664627, 0.1007986068725586, -0.30066731572151184, 0.167317733168602, -0.173665851354599, -0.08247645199298859, 0.015731103718280792, 0.059499677270650864, -0.08394315838813782, -0.3012196719646454, -0.12381046265363693, 0.6368820667266846, 0.23763468861579895, -0.31690338253974915, -0.14158743619918823, 0.09555449336767197, -0.46418440341949463, -0.04789269343018532, 0.2085668444633484, 0.02511993795633316, -0.3501126766204834, -0.13585302233695984, 0.0010477718897163868, -0.006365634500980377, -0.05010829120874405, 0.28892162442207336, -0.2664487659931183, 0.20656630396842957, 0.09581834822893143, 0.0672091692686081, 0.6602531671524048, 0.017082706093788147, -0.21449172496795654, 0.22153334319591522, -0.15071627497673035, 0.13258711993694305, -0.10475029796361923, 0.11741265654563904, -0.37669920921325684, -0.047003574669361115, 0.18447914719581604, 0.8385096788406372, -0.2917744815349579, -0.15610283613204956, -0.06377512961626053, -0.3239149749279022, 0.3379635214805603, 0.15537415444850922, 0.10215353965759277, -0.2498815953731537, -0.14259573817253113, 0.16948066651821136, 0.16629913449287415, -0.356065958738327, 0.16821666061878204, 0.18711219727993011, -0.02912452444434166, 0.11733973026275635, -0.22763407230377197, -0.2857434153556824, -0.26597312092781067, -0.13448573648929596, 0.1124015748500824, 0.04761628434062004, 0.4460560083389282, -0.45091044902801514, 0.06634527444839478, -0.6603160500526428, -0.02074446901679039, 0.09107711911201477, -0.0005951300263404846, 0.3114056885242462, 0.10588116943836212, -0.2892327904701233, -0.175703763961792, 0.3350963890552521, -0.0520145557820797, -0.16662995517253876, -0.15280741453170776, 0.4420304596424103, 0.038530878722667694, 0.07375138998031616, 0.009351879358291626, 0.5092471837997437, 0.048488471657037735, 0.034645307809114456, 0.149419367313385, -0.12538093328475952, -0.5496699213981628, 0.1898004114627838, 0.183569997549057, 0.02857401967048645, -0.15538746118545532, -0.068069227039814, 0.3543192744255066, -0.3344201147556305, -0.0053182621486485004, -0.1602422446012497, -0.07414629310369492, 0.4110819697380066, -0.24898143112659454, 0.36992499232292175, -0.020390696823596954, -0.11749587208032608, 0.016925035044550896, -0.15341848134994507, 0.021901724860072136, -0.10193338245153427, 0.21405470371246338, -0.007762245833873749, 0.013551242649555206, 0.34956735372543335, -0.387805700302124, 0.2461588978767395, 0.28392383456230164, 0.10369329154491425, -0.046701379120349884, -0.014283496886491776, -0.14048606157302856, -0.047391608357429504, 0.03492948040366173, 0.17373257875442505, 0.5730593800544739, 0.19658580422401428, 0.201826810836792, 0.23324695229530334, -0.15427251160144806, 0.12579293549060822, 0.014814585447311401, 0.1872757375240326, 0.06608593463897705, 0.30791279673576355, 0.0905812680721283, 0.22131584584712982, 0.18971002101898193, 0.39874452352523804, 0.20331734418869019, -0.09153177589178085, -0.1926695704460144, -0.005747830495238304, -0.2850164473056793, -0.6398731470108032, -0.5628815293312073, -0.1282135248184204, -0.2211204171180725, -0.05306980013847351, -0.09376570582389832, -0.0052847787737846375, 0.00872943177819252, 0.1844758689403534, -0.11454105377197266, 0.42126092314720154, -0.47502192854881287, -0.2778404951095581, 0.2787609398365021, 0.041623201221227646, -0.18180063366889954, 0.0709051787853241, 0.4079386293888092, 0.21213699877262115, 0.2545481026172638, -0.057418711483478546, -0.23279772698879242, -0.06010568141937256, -0.5316919088363647, 0.05019490793347359, -0.36541086435317993, -0.14889386296272278, -0.2518877685070038, -0.10150045156478882, -0.21497701108455658, -0.3766782879829407, 0.051559530198574066, 0.3118175268173218, 0.05274157226085663, -0.1287008374929428, 0.05726409703493118, 0.2869410514831543, -0.10500974953174591, -0.29453229904174805, -0.42148053646087646, -0.05987602099776268, 0.042383067309856415, -0.16191205382347107, -0.020818598568439484, -0.36425983905792236, -0.05395589396357536, -0.036509353667497635, -0.2672067880630493, -0.011379342526197433, -0.2016235589981079, 0.0008076950907707214, 0.1787797063589096, -0.13474710285663605, -0.12834623456001282, -0.08461089432239532, -0.1897968202829361, 0.29974430799484253, -0.030816208571195602, -0.10249122232198715, -0.15170343220233917, -0.1399437040090561, -0.0023920759558677673, -0.14289137721061707, 0.3774227499961853, 0.17890051007270813, 0.017672274261713028, 0.14856798946857452, 0.07189612090587616, -0.2294684201478958, 0.021115601062774658, 0.09855327755212784, 0.45749837160110474, -0.1967516839504242, 0.25292760133743286, -0.14573286473751068, 0.6077174544334412, 0.5213463306427002, 0.16695120930671692, 0.06605766713619232, 0.25551122426986694, 0.14240209758281708, 0.0633198544383049, -0.20977532863616943, 0.2008449137210846, 0.19462987780570984, 0.05766982585191727, 0.2699241042137146, 0.10378659516572952, -0.053352147340774536, -0.3339765965938568, 0.09006898105144501, -0.3042641878128052, -0.11738692224025726, 0.37775230407714844, -0.272573858499527, 0.11911562830209732, 0.011026915162801743, -0.464156836271286, -0.3556422293186188, -0.13096469640731812, -0.21647487580776215, 0.11839114129543304, -0.2585562467575073, 0.13618622720241547, -0.5622233748435974, -0.10364512354135513, -0.140688955783844, 0.32386571168899536, 0.1514759510755539, 0.3083507716655731, 0.04973016306757927, 0.055047936737537384, -0.24609489738941193, -0.04382070153951645, 0.0648692399263382, -0.4924119710922241, -0.13473451137542725, 0.0649791806936264, 0.014051701873540878, -0.23107802867889404, -0.17589332163333893, 0.017375919967889786, -0.09139548987150192, -0.18299786746501923, 0.3364602327346802, -0.31847766041755676, -0.15623880922794342, 0.6944684982299805, 0.43872055411338806, -0.07578420639038086, -0.3850887417793274, 0.046434272080659866, -0.1232314258813858, -0.14339585602283478, -0.26000726222991943, -0.04751051962375641, 0.16320359706878662, -0.3689451813697815, 0.15044571459293365, 0.01641192100942135, 0.08038859069347382, 0.4504481554031372, 0.10332688689231873, 0.0721358209848404, 0.352663516998291, 0.3876456320285797, -0.24028700590133667, -0.0815337598323822, -0.2526187300682068, 0.4975552558898926, -0.07558071613311768, -0.3227156698703766, -0.1711563766002655, -0.14382445812225342, 0.7155787944793701, -0.06854936480522156, 0.15113170444965363, 0.231692373752594, -0.3995974361896515, -0.361508846282959, 0.001953156664967537, 0.20284631848335266, 0.2826133668422699, 0.32211485505104065, -0.3971283733844757, -0.4497554302215576, 0.6233009696006775, 0.4106392562389374, -0.020718932151794434, 0.2541210949420929, -0.25738000869750977, -0.17415764927864075, 0.09945577383041382, -0.10148287564516068, 0.8141263127326965, 0.021813631057739258, 0.3815930485725403, -0.00047905556857585907, -0.16811984777450562, 0.7002723217010498, -0.09663334488868713, 0.019921768456697464, 0.020172063261270523, -0.1437639445066452, -0.035385921597480774, -0.03009585104882717, -0.12921613454818726, 0.37022101879119873, -0.2077862173318863, 0.3171713054180145, -0.0004519149661064148, -0.00524304062128067, 0.0035987719893455505, 0.40919193625450134, 0.06669454276561737, -0.25877243280410767, -0.39313653111457825, -0.0303385891020298, -0.30272597074508667, -0.05387930944561958, -0.1444966197013855, 0.07147341966629028, -0.13137845695018768, 0.07047849893569946, -0.1482681930065155, 0.29996389150619507, 0.5066030025482178, 0.044584959745407104, -0.41170385479927063, -0.44156980514526367, 0.29387804865837097, 0.1413809210062027, 0.3556130528450012, 0.02908967062830925, 0.029953598976135254, 0.014914780855178833, -0.3565634787082672, -0.015754686668515205, -0.11483734846115112, -0.29389649629592896, 0.2594217360019684, 0.020187336951494217, -0.2533405125141144, 0.06901317089796066, 0.41276517510414124, -0.08236521482467651, -0.25440287590026855, -0.17855213582515717, 0.3314897418022156, -0.059461019933223724, -0.5873781442642212, 0.6571646928787231, 0.15120764076709747, -0.07791917026042938, 0.010054625570774078, -0.11371757090091705, -0.2892296314239502, 0.3845168650150299, -0.27283975481987, -0.25546637177467346, 0.060183294117450714, 0.1803913712501526, 0.3060007393360138, -0.36582598090171814, 0.4701862335205078, 0.1839190125465393, 0.2744100093841553, 0.02297355979681015, -0.06155161187052727, 0.26224833726882935, -0.07377277314662933, -0.3581317663192749, 0.10253649204969406, 0.20453248918056488, 0.11256218701601028, 0.1527307629585266, 0.1633155196905136, 0.12227499485015869, 0.20673184096813202, -0.06777942925691605, -0.3782421946525574, 0.2728555202484131, 0.200240820646286, 0.26178988814353943, 0.25701823830604553, 0.31280267238616943, 0.06446360796689987, 0.6031780242919922, -0.1696951985359192, -0.019097788259387016, 0.13622340559959412, 0.28524187207221985, 0.34651583433151245, 0.08450741320848465, -0.40706947445869446, 0.014368634670972824, -0.04422113299369812, 0.0018224380910396576, 0.11410652101039886, -0.14427253603935242, -0.07686243951320648, 0.17597177624702454, 0.34493160247802734, 0.001123265828937292, 0.06023344025015831, -0.2807413935661316, -0.011823095381259918, -0.022508585825562477, 0.06651268899440765, -0.09257775545120239, -0.00026310235261917114, 0.38064053654670715, -0.07365313172340393, 0.03403807058930397, -0.28564876317977905, 0.26426970958709717, -0.1323501467704773, -0.03567935526371002, 0.25900283455848694, 0.10173744708299637, 0.3916667401790619, -0.044483136385679245, -0.25726914405822754, -0.26786476373672485, 0.21714022755622864, 0.06505507230758667, -0.34551671147346497, -0.03496948629617691, 0.046090368181467056, -0.06114388257265091, 0.15212295949459076, 0.3250305950641632, -0.10721445083618164, -0.025068938732147217, 0.23458319902420044, 0.0007479563355445862, -0.1970008909702301, 0.07210490852594376, 0.23050764203071594, -0.06402743607759476, -0.14158520102500916, 0.21958449482917786, 0.09061235189437866, 0.3772614300251007, -0.3111916780471802, -0.012799031101167202, -0.007480906322598457, -0.03604806959629059, 0.3243543207645416, 0.5464572310447693, -0.032420575618743896, -0.10312484204769135, -0.012002123519778252, 0.08090406656265259, 0.058864183723926544, -0.05017808824777603, -0.14656272530555725, -0.0674607902765274, -0.0896884873509407, -0.07963159680366516, 0.23124241828918457, -0.08268290758132935, 0.06398656964302063, 0.29144909977912903, 0.22425898909568787, 0.13219547271728516, 0.059876590967178345, 0.27598100900650024, 0.005738468840718269, -0.4656766653060913, 0.027942942455410957, 0.41669270396232605, -0.07972194254398346, 0.0029158368706703186, -0.39661499857902527, -0.0038863271474838257, -0.1403137445449829, 0.20574165880680084, -0.12469911575317383, -0.33381330966949463, 0.5012160539627075, 0.09291692078113556, -0.14959858357906342, -0.22758518159389496, 0.1266917884349823, 0.5161339640617371, -0.03509516268968582, -0.048462413251399994, 0.0520235076546669, 0.16546931862831116, -0.1280071884393692, 0.037699319422245026, 0.25175678730010986, 0.11086656898260117, 0.029335781931877136, 0.0011414103209972382, 0.024302955716848373, 0.1404404491186142, 0.12212355434894562, 0.1034310832619667, 0.019177988171577454, 0.08414570987224579, -0.01620527356863022, 0.2771753966808319, 0.10079756379127502, -0.06125646084547043, 0.07313267886638641, -0.12367571890354156, 0.0440765842795372, -0.588858962059021, 0.26579609513282776, 0.31551527976989746, -0.023779835551977158, -0.05252765864133835, 0.17175768315792084, -0.32935717701911926, -0.18782618641853333, 0.5251398682594299, 0.2926930785179138, -0.15609651803970337, -0.15337051451206207, 0.03533812612295151, -0.02974071353673935, 0.31862959265708923, 0.3571052551269531, 0.08333572745323181, -0.42110368609428406, -0.1852589249610901, -0.7274479866027832, 0.11188363283872604, -0.1998959630727768, -0.11876031011343002, -0.26006925106048584, 0.0404849499464035, 0.26376107335090637, 0.2332354187965393, 0.06438417732715607, 0.43237608671188354, 0.14185306429862976, 0.1557856947183609, -0.43412360548973083, -0.15743917226791382, 0.08547855913639069, 0.18689313530921936, 0.05970897525548935, -0.11151805520057678, 0.024231597781181335, 0.35462501645088196, -0.1866980493068695, -0.06368330121040344, 0.01978754997253418, 0.13367462158203125, -0.04021917283535004, 0.2397705316543579, 0.03683649003505707, 0.6354694366455078, -0.16606959700584412, 0.09476259350776672, 0.1124473363161087, -0.050969451665878296, -0.5395124554634094, 0.35593634843826294, -0.15957559645175934, 0.019293449819087982, -0.6011696457862854, 0.6696318984031677, 0.12462136149406433, -0.28975358605384827, -0.08932241797447205, 0.12198374420404434, -0.053985126316547394, -0.13608352839946747, 0.2872995138168335, 0.0695641040802002, 0.2895844280719757, 0.6231246590614319, 0.044806335121393204, 0.08362887054681778, -0.18011710047721863, -0.09812998026609421, 0.2389068305492401, 0.10177017748355865, -0.4384743273258209, -0.09930215030908585, -0.1474403738975525, 0.0205276720225811, -0.08745379000902176, -0.5327368378639221, -0.03613046929240227, 0.3234035074710846, -0.17692339420318604, 0.3889912962913513, 0.5019958019256592, 0.01669117622077465, 0.0657181590795517, -0.2933851182460785, -0.06997093558311462, -0.07178293913602829, -0.07720504701137543, -0.1236257553100586, -0.2613339424133301 ]
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 ?
[ 0.34102776646614075, -0.050783343613147736, 0.04983740672469139, 0.20578475296497345, -0.012494008988142014, -0.09158316254615784, 0.13143377006053925, 0.24781447649002075, 0.07514068484306335, -0.21932382881641388, -0.1656389683485031, 0.10058028995990753, 0.06382811069488525, -0.29008498787879944, 0.24017472565174103, 0.0783972442150116, 0.0976623147726059, 0.11659547686576843, 0.14824892580509186, -0.1969805359840393, -0.11032861471176147, 0.46621453762054443, -0.11777040362358093, -0.12418785691261292, 0.02648148499429226, 0.05782896280288696, -0.25497737526893616, 0.22920289635658264, -0.17581428587436676, -0.2693144679069519, 0.22969691455364227, 0.46074485778808594, 0.01064758375287056, 0.24424946308135986, -0.00012602117203641683, -0.026637017726898193, 0.4085165858268738, 0.03749188035726547, -0.07534395158290863, -0.12848345935344696, 0.15254196524620056, -0.31279507279396057, 0.39784446358680725, -0.10654234141111374, 0.04228166490793228, -0.393091082572937, 0.17608490586280823, -0.046389978379011154, 0.17987260222434998, 0.0012874174863100052, 0.08718021214008331, 0.12423442304134369, 0.06380897760391235, 0.06727617979049683, 0.4933621883392334, 0.01011587679386139, 0.06612028926610947, 0.41989225149154663, 0.2697266936302185, -0.16339382529258728, 0.13654892146587372, 0.24459709227085114, -0.11495184898376465, 0.1563953310251236, 0.17969712615013123, 0.4194623827934265, 0.05964410677552223, -0.4671778976917267, -0.20219320058822632, 0.34553930163383484, 0.014854662120342255, 0.013018671423196793, -0.04609832167625427, -0.4677787721157074, -0.01220071129500866, 0.0002346821129322052, 0.20772784948349, 0.45637431740760803, -0.16081520915031433, -0.0648922324180603, 0.08118142932653427, -0.2500704526901245, -0.060462795197963715, 0.5617239475250244, 0.16432194411754608, 0.048341281712055206, -0.19733892381191254, 0.3637370467185974, 0.4003930985927582, 0.16857296228408813, -0.26249799132347107, -0.2875668406486511, 0.19754011929035187, 0.18750794231891632, -0.3029335141181946, 0.008081495761871338, 0.0892409011721611, -0.0668543353676796, 0.24016210436820984, 0.3428828716278076, 0.0869828388094902, -0.04315388202667236, 0.19389544427394867, 0.12176162004470825, 0.529692530632019, -0.06093313917517662, -0.020973533391952515, 0.1862221211194992, 0.3687583804130554, -0.016591491177678108, 0.0886765867471695, 0.09830331802368164, 0.09169559925794601, -0.025422867387533188, -0.23397573828697205, -0.598155677318573, 0.287977397441864, -0.2756761908531189, 0.24873152375221252, -0.023607876151800156, -0.011190991848707199, -0.05599594488739967, -0.2106602042913437, 0.004625678062438965, 0.018986862152814865, 0.2226305902004242, 0.11902976036071777, 0.28089165687561035, -0.20559236407279968, 0.20877200365066528, -0.0782691240310669, 0.005981238558888435, -0.3656753599643707, 0.03970438241958618, 0.27115416526794434, -0.27817502617836, 0.1672438383102417, 0.09166526049375534, -0.11322474479675293, -0.1840112805366516, 0.17918317019939423, -0.13153216242790222, 0.0667213574051857, 0.3698318600654602, 0.27781257033348083, 0.508072555065155, -0.0019681155681610107, -0.2545323967933655, -0.27548736333847046, -0.06114349514245987, -0.2797984778881073, -0.16831763088703156, -0.18680988252162933, 0.029858209192752838, -0.39023643732070923, 0.16661079227924347, 0.00011780112981796265, 0.28824135661125183, 0.1552387923002243, -0.17051096260547638, -0.11550761014223099, -0.004567872732877731, -0.08512463420629501, -0.37034308910369873, 0.12198788672685623, 0.34375470876693726, 0.19060109555721283, -0.1056913435459137, -0.12709030508995056, -0.1005403995513916, 0.08209993690252304, 0.311158686876297, -0.2761954665184021, 0.14208726584911346, -0.24165396392345428, 0.2310815006494522, 0.5904173254966736, -0.27101388573646545, -0.452669233083725, 0.31155821681022644, -0.024965107440948486, 0.07489564269781113, 0.15223035216331482, 0.2749258577823639, -0.1387363225221634, -0.02256181463599205, 0.16591569781303406, 0.3668280839920044, 0.09084153175354004, -0.1462748646736145, -0.30497613549232483, -0.194266214966774, 0.30231988430023193, -0.09074105322360992, 0.11595368385314941, 0.005938924849033356, -0.008974798023700714, 0.06970684230327606, 0.30660560727119446, 0.2294960767030716, 0.18607574701309204, 0.24630224704742432, -0.07493281364440918, -0.009011022746562958, 0.07086265832185745, -0.2000008523464203, -0.33421590924263, 0.05639561265707016, -0.3555225133895874, 0.2431671917438507, 0.26221126317977905, -0.042365819215774536, -0.23279789090156555, -0.4931893050670624, -0.1797831803560257, -0.26830679178237915, 0.004727199673652649, 0.21896950900554657, 0.12241218984127045, 0.16072684526443481, -0.26788291335105896, 0.17435473203659058, 0.009839585050940514, 0.08805970847606659, -0.8079344034194946, 0.3179464042186737, -0.017846236005425453, -0.07045114785432816, 0.0631343349814415, 0.1433001160621643, 0.001261923462152481, -0.16912347078323364, 0.32444727420806885, 0.43658122420310974, -0.13749289512634277, 0.22820736467838287, 0.22558555006980896, 0.05751761049032211, 0.2537793219089508, -0.11084961891174316, 0.15133501589298248, -0.1257915049791336, 0.11191115528345108, -0.23113122582435608, -0.3042531907558441, 0.11682837456464767, -0.0770876556634903, 0.37470802664756775, 0.14695371687412262, -0.011496390216052532, 0.010349219664931297, -0.23012566566467285, -0.47462472319602966, -0.21605578064918518, 0.20148727297782898, 0.22801299393177032, -0.1119728833436966, 0.14593082666397095, -0.2998264729976654, 0.07807661592960358, 0.07013916224241257, 0.017232604324817657, -0.020467184484004974, 0.17367807030677795, -0.27142301201820374, -0.2440958023071289, -0.18221423029899597, 0.516252338886261, 0.8718119859695435, 0.19785284996032715, 0.3229740858078003, 0.10880731791257858, -0.17633862793445587, -0.1969754993915558, 0.08697985857725143, 0.033732831478118896, -0.12805841863155365, 0.21404004096984863, 0.17132708430290222, -0.008642091415822506, -0.42317691445350647, 0.3805907964706421, 0.16046586632728577, 0.021289855241775513, -0.37637463212013245, -0.1903132200241089, -0.40123867988586426, -0.3012843132019043, -0.6075862050056458, -0.18438860774040222, -0.36245566606521606, -0.34663310647010803, -0.06475569307804108, 0.13995036482810974, -0.05048669874668121, 0.02463909424841404, -0.2065402865409851, 0.15557387471199036, -0.23960019648075104, 0.06841801851987839, 0.141157329082489, 0.1248280256986618, -0.3013819754123688, -0.080742746591568, 0.03651655092835426, 0.38875484466552734, 0.518622100353241, 0.054689645767211914, -0.2073940485715866, -0.08984847366809845, -0.4579673111438751, 0.25336790084838867, -0.2039085477590561, 0.4136425256729126, 0.2852550745010376, 0.0983385220170021, 0.36341434717178345, -0.08333992213010788, 0.2817317247390747, -0.14543606340885162, 0.030420979484915733, 0.008215336129069328, 0.08342495560646057, -0.020247697830200195, -0.2850768268108368, -0.5693515539169312, -0.3060920238494873, -0.2529955804347992, -0.11007168889045715, -0.05502310022711754, 0.12144863605499268, 0.24505096673965454, -0.0811234563589096, 0.12399712204933167, -0.4648562967777252, -0.045524708926677704, -0.36400437355041504, -0.08933573216199875, 0.24169158935546875, -0.22708259522914886, -0.4088027775287628, -0.04336240887641907, -0.02477332204580307, -0.09826163947582245, 0.15945127606391907, -0.38081425428390503, -0.41890549659729004, -0.21503809094429016, -0.5069783329963684, 0.17532096803188324, 0.41701313853263855, 0.30164095759391785, -0.3365061581134796, -0.10329846292734146, -0.08495710790157318, -0.1637304276227951, 0.18865779042243958, 0.2175230085849762, 0.3610684871673584, 0.20806372165679932, -0.09732384234666824, -0.015406223013997078, 0.5112146735191345, 0.45359480381011963, 0.1888437420129776, 0.2588818669319153, -0.030591147020459175, 0.6296035647392273, 0.08015412092208862, -0.352665513753891, 0.0025883882772177458, -0.023501619696617126, 0.2995643615722656, 0.3215644955635071, 0.2464163601398468, -0.17453664541244507, -0.09905517101287842, 0.1481444090604782, -0.35031506419181824, -0.3976816236972809, 0.22064201533794403, -0.11810313165187836, 0.03348477929830551, 0.21164000034332275, -0.3735966980457306, 0.05672074481844902, -0.05050106719136238, -0.3561001718044281, 0.39024847745895386, -0.13153076171875, 0.12844499945640564, -0.599489688873291, 0.11785094439983368, -0.43639665842056274, 0.13917550444602966, -0.12238942086696625, 0.16424347460269928, 0.11637763679027557, -0.06740288436412811, 0.3922004699707031, 0.028027335181832314, 0.5079417824745178, -0.1339224874973297, -0.21793539822101593, 0.1831878274679184, -0.09627951681613922, -0.8417407274246216, -0.16880621016025543, -0.07770348340272903, -0.11526864767074585, 0.2761756479740143, 0.30282431840896606, -0.5176626443862915, -0.25081342458724976, 0.2035353183746338, 0.1481359899044037, -0.17218632996082306, -0.2493516206741333, -0.263827383518219, -0.4485575556755066, -0.3102776110172272, -0.3914089500904083, -0.12274369597434998, -0.05234157666563988, -0.11759019643068314, -0.1287962943315506, 0.1574462354183197, -0.1702951192855835, 0.1424727588891983, -0.16090027987957, 0.2610478401184082, -0.020965980365872383, 0.03182746842503548, 0.465202271938324, 0.13367360830307007, -0.575064480304718, 0.5631838440895081, -0.18785931169986725, 0.2896803617477417, 0.1397072970867157, -0.17257237434387207, 0.3634226322174072, 0.53404700756073, -0.06888536363840103, 0.06709124147891998, 0.03953199088573456, -0.07136170566082001, -0.11491309851408005, 0.25564396381378174, 0.5111374258995056, -0.3491494655609131, -0.13880009949207306, -0.3376767933368683, 0.5596760511398315, 0.08166489005088806, -0.4084247350692749, 0.19925884902477264, 0.26665204763412476, -0.24415437877178192, 0.28216004371643066, -0.3538729250431061, 1.0832160711288452, 0.04487142711877823, 0.31874552369117737, 0.08139197528362274, -0.20796729624271393, 0.891269862651825, -0.019846737384796143, 0.025093605741858482, -0.2386833131313324, -0.04003006964921951, 0.016051342710852623, -0.18522723019123077, 0.05906523019075394, 0.0831565335392952, -0.026412058621644974, 0.29316970705986023, 0.22072936594486237, 0.09865710139274597, -0.10229625552892685, 0.5956615805625916, 0.04097191244363785, -0.08230295777320862, -0.3292580842971802, 0.04090885818004608, -0.013093538582324982, 0.31515443325042725, 0.06848505884408951, -0.14593453705310822, -0.22294293344020844, -0.4063076972961426, -0.26716148853302, 0.04904662072658539, 0.05897710099816322, 0.13230887055397034, 0.102867990732193, -0.11787711083889008, -0.29500049352645874, 0.02908182516694069, -0.04239260405302048, 0.019790973514318466, -0.25581473112106323, -0.0875733494758606, 0.003971949219703674, -0.3073880076408386, -0.13085803389549255, -0.22023822367191315, 0.2706191837787628, -0.1288914680480957, -0.18283438682556152, -0.23630553483963013, 0.04053318500518799, -0.15919379889965057, -0.46210959553718567, 0.31663966178894043, 0.2644232213497162, -0.0829138308763504, -0.29635560512542725, 0.3617565333843231, 0.11170418560504913, -0.0023642703890800476, 0.029360253363847733, 0.3525853753089905, 0.0783948078751564, 0.36199334263801575, 0.1353423148393631, -0.19166700541973114, -0.06823709607124329, 0.3341466188430786, 0.5289828777313232, -0.09188312292098999, 0.2554863691329956, -0.115418940782547, 0.051908038556575775, -0.05099864304065704, -0.16937878727912903, -0.08383148908615112, -0.3399846851825714, 0.12293001264333725, -0.01236810814589262, 0.06921026110649109, 0.1314322054386139, 0.043052248656749725, 0.09093345701694489, -0.21702653169631958, -0.10895167291164398, -0.45283451676368713, 0.06527463346719742, 0.29334086179733276, -0.1723695993423462, 0.40974053740501404, 0.0405631959438324, -0.12469850480556488, 0.25025713443756104, 0.31264162063598633, -0.16058337688446045, 0.2575591206550598, 0.0629950612783432, 0.24463985860347748, 0.6299228668212891, 0.01803286001086235, 0.08861178159713745, -0.11040416359901428, -0.08225207775831223, -0.057102978229522705, -0.17737193405628204, -0.09637583792209625, -0.2691378593444824, 0.22645141184329987, 0.14458200335502625, 0.05872973054647446, 0.0979403629899025, -0.23562291264533997, 0.0319424606859684, 0.008571362122893333, 0.20746645331382751, -0.06657180190086365, -0.039421163499355316, -0.025018073618412018, -0.16160482168197632, 0.03848223015666008, 0.11174441874027252, 0.34071171283721924, -0.36419492959976196, -0.04273565113544464, 0.3421165347099304, 0.008133165538311005, -0.21149592101573944, -0.09788179397583008, -0.28521832823753357, -0.030537590384483337, 0.1042071133852005, -0.04821782559156418, 0.11100776493549347, 0.006433602422475815, -0.02997584268450737, -0.1822148710489273, 0.3628309667110443, 0.5864512324333191, -0.26623332500457764, 0.08765033632516861, -0.05240938067436218, 0.02447781339287758, -0.1797003597021103, -0.038482461124658585, 0.22309213876724243, 0.2693849802017212, -0.08858191221952438, 0.268466055393219, 0.3248119652271271, 0.025378338992595673, -0.12275771796703339, -0.019717741757631302, 0.6393313407897949, -0.19335435330867767, 0.08591987192630768, -0.2401423454284668, -0.04187872260808945, -0.11093232035636902, -0.017787883058190346, -0.033447396010160446, 0.007564239203929901, -0.09867122769355774, 0.20377245545387268, 0.20150569081306458, -0.03543471544981003, 0.05255989357829094, 0.23083460330963135, -0.1845744103193283, 0.016510192304849625, 0.7022262811660767, 0.20640385150909424, 0.30672264099121094, 0.14233775436878204, 0.7357051968574524, -0.11739077419042587, -0.05003523454070091, -0.13992121815681458, 0.5564060211181641, 0.26854610443115234, -0.2376047670841217, 0.06799222528934479, -0.11369521915912628, -0.142375648021698, 0.048237983137369156, -0.04404212161898613, -0.328213632106781, 0.27146100997924805, -0.027660518884658813, -0.19461427628993988, -0.735490083694458, 0.16255098581314087, 0.07861082255840302, 0.1073671504855156, -0.15201492607593536, 0.18626035749912262, -0.07005301117897034, -0.050073251128196716, 0.2191651314496994, 0.5182143449783325, 0.2908780872821808, 0.3593866229057312, -0.0064812470227479935, -0.16204549372196198, 0.24868381023406982, 0.17708560824394226, -0.20950044691562653, 0.15459448099136353, 0.34004947543144226, 0.0780133605003357, 0.32010093331336975, 0.042092397809028625, -0.10348362475633621, -0.06722696125507355, 0.0774071142077446, 0.02938055619597435, -0.3207195997238159, 0.5537846684455872, -0.2837353050708771, 0.08010008186101913, -0.017824001610279083, 0.06254686415195465, -0.6169645190238953, -0.1906488686800003, 0.21398423612117767, 0.005209197290241718, 0.16936859488487244, -0.46065640449523926, 0.01407582312822342, 0.17706599831581116, 0.5300928354263306, 0.04728832468390465, 0.06634972244501114, -0.15457892417907715, -0.24412712454795837, -0.7898312211036682, 0.13781481981277466, -0.2290433943271637, 0.14415255188941956, -0.048181138932704926, 0.3934709429740906, 0.032874125987291336, 0.31124621629714966, -0.02222028374671936, 0.1306430697441101, -0.539971113204956, -0.13637644052505493, -0.17598344385623932, -0.04879867658019066, -0.21684114634990692, -0.058267854154109955, 0.005118973553180695, -0.11321879178285599, 0.10097970068454742, -0.2996004521846771, -0.07709407806396484, 0.09741338342428207, 0.0709075927734375, 0.0748334527015686, 0.05449429154396057, 0.40695086121559143, 0.23451271653175354, 0.41267135739326477, -0.000277101993560791, -0.2844558656215668, -0.5316896438598633, 0.04555456340312958, -0.416957288980484, 0.09936391562223434, 0.021548980847001076, 0.20746108889579773, -0.4415888488292694, 0.15417622029781342, -0.11783294379711151, -0.07033371180295944, 0.13117671012878418, 0.056523412466049194, 0.03648348152637482, -0.008407427929341793, 0.14452460408210754, 0.1382405012845993, 0.23347686231136322, 0.4183560013771057, -0.22531282901763916, 0.13117742538452148, -0.19558241963386536, -0.2739400565624237, 0.28979170322418213, -0.36235570907592773, -0.134887233376503, 0.12069552391767502, 0.24682091176509857, 0.07022978365421295, 0.06666940450668335, -0.6980173587799072, -0.2639472186565399, 0.3107894957065582, -0.04248134046792984, -0.1835707724094391, 0.2500549554824829, -0.13164961338043213, -0.4783773422241211, -0.0227043479681015, 0.044105131179094315, 0.15658116340637207, -0.3813154399394989, 0.011634185910224915, -0.23016799986362457 ]
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')` ```
[ 0.3337518572807312, 0.15182587504386902, 0.04038216546177864, 0.12209126353263855, -0.01158253476023674, 0.07017669826745987, 0.10876737534999847, 0.386903315782547, -0.11695444583892822, -0.22115789353847504, -0.12466732412576675, 0.3359684348106384, 0.04204697906970978, -0.3219139277935028, 0.3353402316570282, 0.059789542108774185, 0.17052797973155975, 0.11987831443548203, 0.24696116149425507, -0.3221699595451355, -0.15091925859451294, 0.36268121004104614, -0.20078310370445251, -0.07502949237823486, 0.17173171043395996, 0.17042206227779388, -0.13634681701660156, 0.2355842888355255, -0.28036168217658997, -0.2833782732486725, 0.3049052655696869, 0.41182857751846313, 0.12482453882694244, 0.3285561501979828, -0.0001224009320139885, -0.051074616611003876, 0.3996657729148865, -0.1145327091217041, -0.12407520413398743, -0.2439044862985611, 0.025619491934776306, -0.27727824449539185, 0.36111241579055786, -0.15762539207935333, 0.10323167592287064, -0.5018042922019958, 0.22668275237083435, -0.1019296795129776, 0.002705581486225128, 0.2615557014942169, 0.12145350873470306, -0.0823289230465889, 0.13635185360908508, 0.003647378645837307, 0.32640424370765686, 0.01863134279847145, 0.09062426537275314, 0.30791085958480835, 0.17882341146469116, 0.08364862948656082, 0.19623708724975586, 0.2178083211183548, -0.12764735519886017, 0.07411863654851913, 0.3337646722793579, 0.3773285150527954, 0.04918483644723892, -0.4533298909664154, -0.0953766331076622, 0.2512941360473633, 0.4695168435573578, -0.08259274810552597, 0.03704769164323807, -0.20686852931976318, -0.04010636359453201, -0.08870531618595123, 0.10810987651348114, 0.39001116156578064, -0.2233264446258545, -0.04825706407427788, 0.021009482443332672, -0.38639745116233826, -0.1963813453912735, 0.58647620677948, 0.16923373937606812, 0.15303432941436768, -0.08709702640771866, 0.3638680875301361, 0.21774187684059143, 0.13750764727592468, 0.033936306834220886, -0.32258543372154236, 0.25541025400161743, 0.2819863557815552, -0.2922053337097168, -0.007908500730991364, -0.035040438175201416, -0.2091566026210785, 0.18896672129631042, 0.326124370098114, 0.07502581179141998, -0.06704697757959366, 0.2005782276391983, 0.12096545845270157, 0.4643363952636719, -0.1520892232656479, 0.08678402006626129, 0.35399243235588074, 0.21561071276664734, -0.18660415709018707, 0.03401936963200569, 0.018244968727231026, -0.07721927762031555, 0.0717441588640213, -0.22967693209648132, -0.6142798066139221, 0.14092868566513062, -0.40558966994285583, 0.44782310724258423, -0.025238938629627228, 0.21384793519973755, -0.10629229247570038, -0.23445358872413635, 0.08172988146543503, 0.10517358034849167, 0.20921194553375244, 0.0853356346487999, 0.17645776271820068, -0.1176484003663063, 0.09517573565244675, -0.08684973418712616, 0.01650475151836872, -0.36104920506477356, -0.05815677344799042, 0.23637695610523224, -0.0132148377597332, 0.261093407869339, -0.09314660727977753, 0.07390698045492172, -0.090160071849823, 0.24405556917190552, -0.059681836515665054, 0.14490285515785217, 0.49642643332481384, 0.3115929663181305, 0.5195353627204895, -0.0001559816300868988, -0.19920705258846283, -0.27722200751304626, 0.1592060625553131, -0.3895842730998993, -0.12449295818805695, -0.11835961043834686, 0.07398621737957001, -0.3111487627029419, -0.06145257502794266, 0.10784704983234406, 0.2107418328523636, 0.18148663640022278, -0.35827332735061646, -0.10828778892755508, 0.02562979981303215, 0.04021844640374184, -0.3753201961517334, 0.17049986124038696, 0.1312364786863327, 0.06589071452617645, -0.038904186338186264, -0.285993754863739, -0.0642908364534378, 0.2266976237297058, 0.07826383411884308, -0.3078138828277588, 0.1558774709701538, -0.12100686132907867, 0.05395705997943878, 0.8645718097686768, -0.21998487412929535, -0.40715402364730835, 0.27943742275238037, 0.05443707853555679, -0.14355961978435516, 0.0917615070939064, 0.3109305799007416, -0.2846710681915283, 0.051498785614967346, 0.09130990505218506, 0.2885000705718994, -0.012590447440743446, -0.2461199164390564, -0.14480474591255188, -0.11279940605163574, 0.30863797664642334, -0.07593360543251038, 0.2943570017814636, 0.16808997094631195, -0.015602990984916687, 0.14979888498783112, 0.4100167155265808, 0.18755675852298737, 0.21378695964813232, 0.22659926116466522, -0.1147899329662323, -0.10925041139125824, 0.20860177278518677, -0.17600110173225403, -0.16421037912368774, 0.06642287224531174, -0.4373759329319, 0.2642475664615631, 0.3202393054962158, 0.0615609809756279, -0.29039326310157776, -0.520103931427002, -0.06402754783630371, -0.31061503291130066, -0.012993888929486275, 0.2909797132015228, -0.001455383375287056, 0.07865075767040253, -0.3556828796863556, 0.04744647443294525, -0.1270780712366104, 0.014778020791709423, -0.7680270075798035, 0.15683011710643768, -0.0025865230709314346, -0.014122537337243557, 0.08916129171848297, -0.030356526374816895, -0.04141630977392197, -0.0181819386780262, 0.2261602282524109, 0.41557687520980835, -0.03188009187579155, 0.3567318320274353, 0.3559616804122925, -0.10024566948413849, 0.11630096286535263, -0.2353438436985016, 0.15720008313655853, -0.10833920538425446, 0.2621355652809143, -0.2087302803993225, -0.34462541341781616, 0.10563212633132935, -0.15303215384483337, 0.3154360055923462, 0.15342338383197784, -0.002797182649374008, 0.06230337917804718, -0.12727221846580505, -0.40658944845199585, -0.14812877774238586, 0.17097873985767365, 0.27738890051841736, -0.3150275647640228, 0.0903300791978836, -0.26422256231307983, 0.03734511509537697, 0.12562444806098938, 0.006545018404722214, -0.17453709244728088, 0.1829851269721985, -0.3114658296108246, -0.218950092792511, -0.13686047494411469, 0.4669196307659149, 0.9032633900642395, 0.25674009323120117, 0.3886660039424896, 0.09151758253574371, -0.05875265598297119, -0.16813614964485168, 0.18907034397125244, 0.13167817890644073, 0.012559568509459496, 0.16735856235027313, -0.06914028525352478, -0.048744987696409225, -0.4017599821090698, 0.2793743312358856, 0.16855449974536896, 0.06223824620246887, -0.30279770493507385, -0.03771195560693741, -0.37835752964019775, -0.1974485218524933, -0.5116115212440491, -0.2571055591106415, -0.2551203668117523, -0.3293740749359131, -0.16789837181568146, 0.010062027722597122, 0.05495510622859001, 0.01971004530787468, -0.5319440364837646, 0.04521992802619934, -0.13294251263141632, -0.15076608955860138, 0.1229516863822937, -0.02294669859111309, -0.24956735968589783, 0.007366064935922623, -0.06586247682571411, 0.38983431458473206, 0.48786187171936035, -0.07025723159313202, -0.04292710870504379, 0.052866771817207336, -0.4029851257801056, 0.2302953153848648, -0.24395574629306793, 0.2364078313112259, 0.2415207326412201, 0.1290687620639801, 0.2934618592262268, -0.0029640784487128258, 0.15736204385757446, -0.18025970458984375, 0.024904688820242882, -0.04032796248793602, 0.10109657049179077, -0.10888008028268814, -0.27549323439598083, -0.8172608017921448, -0.3088199496269226, -0.18090634047985077, -0.23882095515727997, 0.18941515684127808, 0.1692734807729721, 0.3574632704257965, 0.020866934210062027, 0.1280636340379715, -0.32717815041542053, 0.057699061930179596, -0.36877599358558655, -0.1524684578180313, 0.4007762670516968, -0.27363458275794983, -0.37948542833328247, 0.10961058735847473, 0.14097359776496887, 0.038805801421403885, 0.14969386160373688, -0.27083054184913635, -0.05013028904795647, -0.007900236174464226, -0.5291979908943176, 0.1509532630443573, 0.33623063564300537, 0.2992240786552429, -0.3369194269180298, -0.11374468356370926, -0.13982738554477692, -0.06552587449550629, 0.11963070183992386, 0.24273642897605896, 0.5703086853027344, 0.13668487966060638, -0.06114605814218521, -0.18792985379695892, 0.47023582458496094, 0.3327806293964386, 0.2193659096956253, 0.13947364687919617, -0.08077389746904373, 0.5153934955596924, 0.07098846882581711, -0.2541893422603607, 0.17766226828098297, 0.03409871459007263, 0.11521978676319122, 0.4011799693107605, 0.14756056666374207, -0.31791743636131287, -0.2271197885274887, 0.029213115572929382, -0.3635048270225525, -0.29176682233810425, 0.22638431191444397, -0.16411001980304718, 0.1019243597984314, 0.26066869497299194, -0.41129791736602783, -0.10925649106502533, -0.021790388971567154, -0.42039117217063904, 0.4306926727294922, -0.15378358960151672, 0.1531316041946411, -0.6548850536346436, 0.16735491156578064, -0.4477037787437439, 0.11611578613519669, -0.2242058366537094, 0.17975777387619019, 0.013182636350393295, -0.10940734297037125, 0.20083558559417725, 0.07203806191682816, 0.3785322904586792, -0.09238990396261215, -0.27775925397872925, 0.16358672082424164, 0.11103224754333496, -0.6928816437721252, -0.05451362580060959, 0.03548400476574898, 0.06715419888496399, 0.07046359777450562, 0.15606406331062317, -0.32216528058052063, -0.2182283103466034, 0.2806127667427063, 0.12941719591617584, -0.2067832499742508, -0.23308983445167542, -0.3231203854084015, -0.5719004273414612, -0.23541997373104095, -0.4505121111869812, -0.14295253157615662, -0.04531502351164818, -0.13631458580493927, -0.09380374848842621, 0.21509875357151031, -0.1203148365020752, 0.25403082370758057, -0.06086587905883789, 0.26602500677108765, -0.037856221199035645, -0.03367999196052551, 0.4772062599658966, 0.09539496898651123, -0.27915599942207336, 0.4296419620513916, -0.14281487464904785, 0.2030206024646759, 0.13304805755615234, -0.372006893157959, 0.4784090220928192, 0.30285537242889404, -0.10998030006885529, -0.00590573251247406, 0.19600863754749298, -0.25283539295196533, -0.008956448175013065, 0.14065776765346527, 0.4020999073982239, -0.34685513377189636, -0.32030215859413147, -0.5828097462654114, 0.5492743253707886, 0.003125932067632675, -0.36882954835891724, 0.2551993131637573, -0.010728076100349426, -0.33157041668891907, 0.23260772228240967, -0.2238222360610962, 1.1526602506637573, 0.12742339074611664, 0.3265940845012665, -0.0009182412177324295, -0.10209798812866211, 0.82244473695755, -0.05445804446935654, 0.08067631721496582, -0.10956190526485443, -0.04315526783466339, -0.13305158913135529, -0.20595550537109375, 0.16717343032360077, 0.12997116148471832, -0.0915985256433487, 0.30789634585380554, 0.17555038630962372, 0.0021262802183628082, 0.07282951474189758, 0.47516173124313354, 0.1268334835767746, 0.004292736761271954, -0.35017240047454834, 0.06220252811908722, 0.031312499195337296, 0.15649764239788055, 0.04731954261660576, -0.23766162991523743, -0.48491808772087097, -0.41060131788253784, -0.09666375815868378, 0.09266923367977142, 0.3221284747123718, 0.07424497604370117, -0.07986213266849518, -0.1980777382850647, -0.2268858104944229, 0.20781366527080536, -0.008842572569847107, -0.05847974494099617, -0.33503925800323486, 0.17607887089252472, -0.14393207430839539, -0.27947548031806946, -0.20071381330490112, -0.1652783304452896, 0.26254016160964966, -0.25820496678352356, -0.2651193141937256, -0.010710440576076508, -0.07903056591749191, -0.2800616919994354, -0.3210082948207855, 0.29311150312423706, 0.42605656385421753, 0.010683368891477585, -0.2837388813495636, 0.40910011529922485, -0.024119526147842407, -0.004243135452270508, 0.056246668100357056, 0.40378648042678833, 0.024780165404081345, 0.47836121916770935, 0.1997634917497635, -0.30778151750564575, -0.11372819542884827, 0.30474233627319336, 0.7428837418556213, -0.07022078335285187, 0.3021067976951599, -0.09806780517101288, -0.011663965880870819, -0.10569634288549423, -0.10398510098457336, -0.2096780389547348, -0.1769353449344635, 0.03281928226351738, 0.031270582228899, 0.2512499690055847, 0.21842877566814423, 0.011708710342645645, -0.011573165655136108, -0.1802452802658081, -0.18839296698570251, -0.4614045023918152, -0.13243600726127625, 0.08870525658130646, -0.37249791622161865, 0.4005108177661896, 0.07215963304042816, -0.13302437961101532, 0.3243177533149719, 0.1654851734638214, -0.20654752850532532, 0.15494365990161896, -0.035978857427835464, 0.27045345306396484, 0.4610447287559509, 0.1112743392586708, 0.05570754408836365, -0.20335344970226288, -0.0635242611169815, -0.19280771911144257, -0.1160019040107727, -0.1025906354188919, -0.27869704365730286, 0.22386646270751953, 0.12650951743125916, 0.04898825287818909, 0.013911639340221882, -0.23256486654281616, -0.19723469018936157, 0.10693582147359848, 0.012789525091648102, -0.0910220593214035, 0.04635275900363922, 0.12139242887496948, -0.13883911073207855, 0.014663740992546082, 0.04517371952533722, 0.31722262501716614, -0.28538018465042114, 0.11752559244632721, 0.328976571559906, 0.11331665515899658, -0.2091713845729828, -0.13677528500556946, -0.2675720751285553, -0.014869920909404755, 0.12046246230602264, -0.17886057496070862, 0.309526264667511, 0.02365928888320923, 0.06650537252426147, -0.21033525466918945, 0.23484084010124207, 0.7377378344535828, -0.2688868045806885, 0.13990581035614014, -0.07228117436170578, 0.026591014117002487, -0.26007381081581116, -0.03565461188554764, 0.14888788759708405, 0.2221623808145523, -0.16084370017051697, 0.157526895403862, 0.1493343561887741, -0.07433685660362244, -0.1829710453748703, 0.010846085846424103, 0.7443419694900513, -0.16752219200134277, 0.10861487686634064, -0.13933515548706055, -0.006555989384651184, 0.04117734730243683, -0.015504902228713036, 0.17375347018241882, 0.021766699850559235, 0.10402995347976685, 0.09909937530755997, 0.29588207602500916, 0.046929530799388885, -0.08133494853973389, 0.156587153673172, -0.1968528926372528, -0.023894090205430984, 0.708240270614624, 0.10019351541996002, 0.3292528986930847, 0.21454080939292908, 0.5007230043411255, 0.04617156460881233, 0.08842470496892929, -0.15246064960956573, 0.45210152864456177, 0.30750763416290283, -0.12500478327274323, -0.1658221185207367, -0.2766610383987427, -0.06631577759981155, -0.10191549360752106, 0.004147782921791077, -0.2759506106376648, 0.16210857033729553, 0.013707369565963745, -0.20408134162425995, -0.619045615196228, 0.3473252058029175, 0.028992461040616035, -0.0881718322634697, -0.15126639604568481, 0.05593831092119217, -0.051576077938079834, -0.10623002052307129, 0.3673670291900635, 0.33350253105163574, 0.24257759749889374, 0.2939143478870392, 0.06989099830389023, -0.1643190234899521, 0.35103273391723633, 0.030839597806334496, -0.18330976366996765, 0.199112206697464, 0.3185024857521057, 0.09088282287120819, 0.3007443845272064, 0.10069775581359863, -0.036437030881643295, 0.03841108828783035, 0.27434471249580383, 0.0016558682546019554, -0.25471341609954834, 0.49166321754455566, -0.14734365046024323, 0.06510015577077866, -0.0114300437271595, 0.11638021469116211, -0.6328656673431396, -0.22419331967830658, 0.2930052578449249, 0.09583322703838348, 0.1150798574090004, -0.3418160080909729, 0.014299340546131134, 0.09833107888698578, 0.6334848403930664, -0.07817037403583527, 0.12740150094032288, -0.12272877246141434, -0.3408302068710327, -0.684195876121521, 0.22655139863491058, -0.11939598619937897, 0.13815149664878845, 0.0011202073656022549, 0.2716941237449646, 0.2850784659385681, 0.24463215470314026, 0.08168907463550568, 0.0556468740105629, -0.6260473728179932, 0.019000347703695297, -0.043697748333215714, -0.2273305505514145, -0.27663105726242065, 0.2152407020330429, -0.08741721510887146, -0.03948530554771423, 0.10128570348024368, -0.3512215316295624, -0.028956817463040352, -0.0786876305937767, -0.08384843170642853, 0.29225456714630127, 0.16823866963386536, 0.5404806137084961, 0.1115911453962326, 0.5250939130783081, -0.18571817874908447, -0.2845458388328552, -0.38820579648017883, 0.117737777531147, -0.37535855174064636, 0.10256737470626831, -0.0020772889256477356, 0.06304667145013809, -0.46579107642173767, 0.32165244221687317, -0.07286262512207031, 0.013991950079798698, -0.048585690557956696, -0.22922445833683014, -0.0022873161360621452, -0.051074616611003876, 0.07794699817895889, 0.1083177849650383, 0.18994644284248352, 0.2920578718185425, -0.20192834734916687, 0.07437711209058762, -0.18357303738594055, -0.13972805440425873, 0.38617220520973206, -0.20822550356388092, -0.194026380777359, 0.09874100238084793, 0.018720127642154694, -0.1719149947166443, 0.16188180446624756, -0.7090767621994019, -0.06251832842826843, 0.3703853189945221, -0.08114277571439743, -0.1896420121192932, 0.10166247189044952, -0.08538301289081573, -0.34090688824653625, -0.10702259838581085, 0.20711614191532135, 0.011787373572587967, -0.27496469020843506, -0.19342853128910065, -0.18986105918884277 ]
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.
[ -0.1073198914527893, -0.08736070990562439, 0.05800746753811836, 0.19675154983997345, 0.21142977476119995, 0.09753540903329849, 0.28022220730781555, 0.37640857696533203, -0.15019194781780243, 0.0037060528993606567, -0.06849081069231033, 0.5348143577575684, -0.11443067342042923, 0.0036037880927324295, 0.32549747824668884, -0.17846299707889557, -0.03447549790143967, 0.2562466263771057, -0.06642463803291321, -0.06997586786746979, -0.04355182498693466, -0.02456831745803356, -0.09649714827537537, -0.026774592697620392, 0.05681774765253067, 0.018075238913297653, 0.08552540838718414, -0.14649775624275208, 0.40811312198638916, 0.04641713947057724, 0.38986966013908386, 0.02334488183259964, -0.033621031790971756, 0.15604394674301147, -0.00011474287020973861, 0.40614014863967896, 0.035418152809143066, -0.19170302152633667, 0.07188375294208527, -0.11593537032604218, 0.16804373264312744, -0.16358737647533417, 0.1652325987815857, -0.2180953472852707, -0.05620696395635605, 0.02299460396170616, 0.21298514306545258, -0.06016510725021362, 0.2803554832935333, 0.13797186315059662, 0.09116256982088089, 0.5434460639953613, -0.17063158750534058, 0.26339152455329895, 0.14321106672286987, -0.15265360474586487, -0.23502619564533234, -0.25438031554222107, -0.19450947642326355, 0.020992005243897438, 0.11463582515716553, 0.08944624662399292, -0.23712721467018127, 0.1298714429140091, 0.46162647008895874, 0.16119210422039032, -0.06941086798906326, -0.36647167801856995, -0.2092972993850708, 0.4291154742240906, 0.26164674758911133, -0.06549926102161407, -0.17502015829086304, -0.446580171585083, -0.23368878662586212, -0.3110027611255646, -0.1049039289355278, 0.12665338814258575, -0.2806468904018402, 0.22548311948776245, -0.09822195768356323, -0.0873730480670929, -0.25664621591567993, 0.17651858925819397, 0.19647055864334106, -0.04481390118598938, 0.12743717432022095, 0.1107761561870575, 0.20774662494659424, 0.15451118350028992, 0.40833669900894165, 0.02135753631591797, 0.3032936751842499, 0.4022805988788605, -0.28941798210144043, 0.029115810990333557, 0.1779962182044983, -0.5676333904266357, -0.06889715045690536, -0.23024220764636993, 0.3916994631290436, 0.1674436330795288, -0.30582019686698914, 0.251703143119812, -0.01072559505701065, -0.09433722496032715, -0.4143868684768677, 0.1737426221370697, -0.1531386822462082, -0.5154545903205872, 0.05342916026711464, 0.044939037412405014, -0.254041850566864, -0.010109532624483109, 0.02233666554093361, -0.5388681292533875, -0.17005741596221924, 0.22495564818382263, -0.3130665421485901, -0.425286203622818, -0.25239044427871704, -0.049549102783203125, 0.0026699593290686607, 0.2106463611125946, -0.12489939481019974, 0.5658942461013794, 0.14857403934001923, -0.137321338057518, -0.3045380115509033, -0.0486360602080822, -0.19958896934986115, 0.08693213760852814, -0.06759446859359741, -0.08612370491027832, 0.1349666565656662, -0.175648033618927, 0.17783555388450623, -0.15483301877975464, -0.16657809913158417, 0.13682587444782257, 0.2998948097229004, -0.10216791927814484, 0.09434030205011368, 0.2443123757839203, 0.04629768803715706, -0.0560595765709877, 0.04743875190615654, 0.07257132232189178, -0.21072903275489807, 0.36221182346343994, -0.5235639810562134, -0.30007055401802063, -0.04146762937307358, 0.09681671112775803, 0.026772238314151764, -0.19203853607177734, -0.3231533169746399, 0.2882840931415558, 0.1580442488193512, -0.13978272676467896, -0.17260849475860596, -0.17757444083690643, -0.43399563431739807, -0.2991335690021515, 0.3465137779712677, 0.07501198351383209, -0.5027421116828918, -0.12278881669044495, -0.01291557028889656, -0.05792718753218651, 0.2622867822647095, 0.42779242992401123, -0.22205686569213867, 0.538048267364502, -0.08767727762460709, -0.027470942586660385, 0.801042914390564, -0.07477029412984848, -0.171042799949646, 0.24721184372901917, -0.11224529147148132, 0.16722844541072845, 0.02871318906545639, 0.3875944912433624, -0.0019190041348338127, -0.03081868588924408, 0.38124793767929077, 0.6379417181015015, -0.15168501436710358, 0.02548329345881939, -0.10614646971225739, -0.34956642985343933, 0.27074897289276123, 0.24196375906467438, 0.10449346899986267, -0.030045397579669952, -0.23651066422462463, 0.3986966013908386, 0.1312621533870697, -0.024340886622667313, -0.1928764283657074, 0.1461460292339325, -0.1458994448184967, 0.02518763765692711, -0.18390010297298431, 0.10199633985757828, -0.6190453767776489, 0.23352786898612976, 0.14448203146457672, 0.08388835191726685, 0.23681887984275818, -0.044368527829647064, 0.3555765450000763, 0.12261108309030533, 0.05143428593873978, 0.15910980105400085, -0.04666578024625778, 0.12642867863178253, -0.05838803946971893, -0.04362756758928299, -0.3507680296897888, 0.13079914450645447, -0.35051771998405457, -0.09408803284168243, -0.3933817148208618, 0.2961205840110779, 0.3782658278942108, -0.03691055625677109, -0.11612765491008759, 0.257525771856308, -0.33689430356025696, -0.10079985857009888, 0.05960436910390854, 0.2039336860179901, 0.008492151275277138, 0.3252086639404297, -0.7214286923408508, 0.4762921929359436, 0.057510197162628174, -0.017924971878528595, -0.08479515463113785, 0.13055159151554108, 0.009945487603545189, -0.2209624946117401, -0.018168173730373383, 0.2750522494316101, -0.23169951140880585, 0.1988259106874466, 0.28876668214797974, -0.28769582509994507, 0.14816169440746307, 0.0720500499010086, 0.1864931881427765, -0.1110144853591919, 0.47623977065086365, 0.17128829658031464, 0.17858904600143433, 0.3287806212902069, -0.5321705937385559, 0.20310641825199127, 0.6362370252609253, -0.01515454612672329, -0.1437099575996399, 0.17936590313911438, 0.018021419644355774, -0.2019813060760498, -0.03624867647886276, -0.07885129749774933, 0.2486165463924408, 0.10557541996240616, 0.04468761384487152, -0.018213583156466484, -0.09769757837057114, -0.28492289781570435, 0.21820439398288727, 0.10028055310249329, 0.4032105505466461, -0.09976091235876083, 0.026069751009345055, 0.23116254806518555, -0.17332789301872253, 0.08780363202095032, 0.249880850315094, 0.17551709711551666, -0.24388746917247772, 0.01786116696894169, -0.2885760962963104, -0.2792428433895111, -0.05579949915409088, 0.08746685087680817, 0.13963860273361206, -0.16931785643100739, -0.09134820103645325, 0.38729098439216614, -0.009227089583873749, -0.08408628404140472, -0.14456327259540558, 0.2974509596824646, 0.14736483991146088, -0.4759064018726349, -0.10680673271417618, -0.06507281959056854, -0.21300122141838074, 0.031523361802101135, 0.37497052550315857, 0.19188150763511658, 0.222330704331398, 0.0081336610019207, -0.13919644057750702, 0.18115459382534027, -0.08397331088781357, 0.2416813224554062, -0.031938329339027405, 0.13781845569610596, -0.046962663531303406, 0.11223617196083069, -0.0727233737707138, -0.26345980167388916, -0.061837464570999146, -0.429096519947052, 0.03967159241437912, 0.18683330714702606, 0.01748155802488327, 0.1602577567100525, -0.23621909320354462, -0.13863727450370789, -0.289046972990036, -0.39600279927253723, -0.01829865574836731, 0.18625009059906006, 0.32266682386398315, 0.4042637348175049, 0.003960270434617996, 0.2540595233440399, 0.3930632770061493, 0.06724469363689423, 0.08527881652116776, -0.15549039840698242, 0.49338218569755554, -0.27669817209243774, -0.266829252243042, -0.14344805479049683, -0.14479805529117584, 0.24469909071922302, 0.36844587326049805, -0.6950926184654236, 0.21136067807674408, -0.09673400223255157, 0.14602124691009521, -0.09841294586658478, -0.01586531288921833, 0.22807864844799042, -0.2083779126405716, 0.06874667853116989, -0.01793697476387024, 0.09904526174068451, 0.13033005595207214, 0.02588054910302162, -0.1436578333377838, 0.42202189564704895, 0.4722154438495636, 0.3167874217033386, 0.7827593088150024, -0.12301819026470184, -0.24631236493587494, 0.2838461399078369, -0.22802242636680603, 0.2599782943725586, -0.19186261296272278, -0.01625453680753708, 0.06628713756799698, 0.04867049306631088, -0.0657784715294838, 0.23008973896503448, -0.10858777165412903, -0.13981178402900696, -0.15558448433876038, 0.22440415620803833, 0.0955587700009346, -0.22741472721099854, 0.22539538145065308, -0.20803266763687134, 0.21618545055389404, 0.0942547470331192, 0.22961732745170593, -0.33768606185913086, -0.42883235216140747, -0.025798311457037926, -0.17720972001552582, 0.4478469789028168, -0.05506230518221855, -0.28194907307624817, 0.13209481537342072, -0.693756639957428, 0.09473717957735062, 0.04768528416752815, 0.3188590705394745, 0.0776636004447937, -0.24230387806892395, 0.004208177328109741, 0.0652078166604042, 0.6028183698654175, 0.04434838145971298, -0.08510814607143402, 0.21417254209518433, -0.19752860069274902, -0.5001963376998901, 0.02252393588423729, -0.03151191025972366, -0.08884158730506897, 0.2648521661758423, 0.12026447057723999, -0.2928028106689453, -0.5042575597763062, -0.07943674921989441, -0.00020402553491294384, -0.3595188856124878, -0.09357305616140366, -0.2238897979259491, -0.26167458295822144, -0.02819342352449894, -0.3826901614665985, 0.021888190880417824, 0.3010372817516327, -0.12246362119913101, 0.0201726071536541, -0.17321881651878357, 0.0016045160591602325, 0.23391839861869812, 0.011684682220220566, 0.14289602637290955, 0.4014410674571991, 0.040681470185518265, -0.061452604830265045, -0.0342998281121254, -0.09876605868339539, 0.2059071809053421, -0.09806300699710846, -0.16123175621032715, 0.03657034784555435, 0.15343116223812103, 0.3691772520542145, 0.03178824484348297, 0.023263178765773773, 0.12791937589645386, -0.14821909368038177, 0.03830118477344513, -0.14445967972278595, 0.37864598631858826, 0.09082312881946564, -0.34533339738845825, -0.07255448400974274, -0.5526954531669617, 0.5314421653747559, 0.2988010346889496, 0.15192611515522003, 0.48851919174194336, -0.1953694373369217, -0.35578614473342896, 0.4916248321533203, 0.3968748152256012, 0.5634973645210266, -0.37892574071884155, 0.488278329372406, 0.7028648853302002, 0.5218997597694397, 0.9219175577163696, -0.38402408361434937, 0.2837420105934143, -0.11130359023809433, -0.3852221369743347, -0.0714365616440773, -0.2753489315509796, -0.3265306055545807, 0.13125737011432648, 0.13309474289417267, 0.2455313801765442, 0.126722514629364, 0.2848149836063385, 0.24128463864326477, 0.35134434700012207, 0.3585633337497711, -0.4072931408882141, -0.21222932636737823, 0.013731896877288818, 0.051649563014507294, -0.14455965161323547, -0.11486605554819107, -0.20799045264720917, 0.21388474106788635, 0.05818121135234833, -0.014009371399879456, 0.15659651160240173, -0.2624652683734894, 0.2152727097272873, 0.13642607629299164, -0.13545162975788116, 0.06786555051803589, 0.2597355842590332, 0.1231052577495575, -0.18345579504966736, -0.07790296524763107, -0.07857638597488403, 0.061042480170726776, -0.011666816659271717, 0.14572271704673767, -0.04938455671072006, 0.6279585957527161, -0.05613863840699196, 0.1945367157459259, 0.16612496972084045, -0.08871294558048248, -0.5715486407279968, -0.1340985894203186, -0.14428310096263885, 0.07634425908327103, 0.18374980986118317, -0.5355894565582275, 0.2812327444553375, -0.2824206054210663, 0.20707973837852478, 0.05103122815489769, 0.13851380348205566, -0.2297205924987793, 0.039810460060834885, 0.01662156917154789, -0.19243469834327698, 0.20765329897403717, 0.03500457480549812, 0.09302018582820892, 0.031202398240566254, 0.3733731806278229, 0.22355763614177704, 0.020717144012451172, -0.017776809632778168, 0.0854412093758583, 0.2849075496196747, -0.07938229292631149, 0.08008500933647156, 0.026714550331234932, 0.2016976922750473, 0.07773049175739288, 0.3733219504356384, -0.18203888833522797, 0.06518024951219559, -0.29214492440223694, -0.16977250576019287, -0.2766077518463135, 0.339433491230011, -0.08160261809825897, 0.12775243818759918, -0.04945923388004303, 0.21621188521385193, -0.1064225435256958, 0.24101118743419647, -0.18489307165145874, -0.053981855511665344, 0.0022446569055318832, 0.04347361624240875, -0.14907661080360413, -0.017295964062213898, 0.04800468310713768, 0.11890891194343567, 0.027415594086050987, 0.12441618740558624, -0.32206061482429504, -0.14280161261558533, -0.02780255489051342, 0.20201469957828522, 0.051952119916677475, -0.25063204765319824, -0.023800689727067947, -0.37163621187210083, -0.33128997683525085, -0.20353341102600098, 0.11228162050247192, -0.11092804372310638, -0.2844444513320923, 0.022438805550336838, -0.0787317305803299, 0.36041101813316345, -0.25190919637680054, 0.06195436790585518, -0.08628301322460175, 0.34060928225517273, -0.09359625726938248, 0.25554755330085754, 0.15018978714942932, -0.022436991333961487, -0.37363651394844055, -0.12177456170320511, 0.17380215227603912, -0.15421348810195923, 0.12140513956546783, -0.328571081161499, -0.07203707098960876, -0.257598340511322, 0.02463843673467636, 0.31406277418136597, -0.19655224680900574, 0.12919332087039948, 0.22686833143234253, 0.0630623996257782, -0.10534346848726273, -0.1104499027132988, 0.21441768109798431, 0.026152312755584717, -0.03034992143511772, 0.22932729125022888, 0.07449840009212494, 0.15259431302547455, -0.10686443746089935, -0.030716732144355774, 0.04046764224767685, -0.5724344253540039, -0.18100467324256897, 0.4083300232887268, -0.07960902899503708, 0.06841670721769333, -0.17246660590171814, 0.10472419857978821, 0.1336464285850525, 0.13838425278663635, 0.3073665499687195, 0.17160212993621826, -0.10191697627305984, -0.08759215474128723, -0.04754035919904709, -0.3773829936981201, -0.4539465606212616, 0.48343297839164734, -0.041963785886764526, 0.571456253528595, -0.2652560770511627, 0.001651465892791748, 0.11766763031482697, -0.16160371899604797, -0.17609857022762299, 0.28459814190864563, -0.10150378197431564, -0.018113423138856888, -0.25577735900878906, 0.05828007310628891, -0.16195760667324066, 0.2735612988471985, -0.0926429033279419, -0.23047123849391937, -0.21142840385437012, 0.3268531560897827, 0.1237487941980362, 0.169804647564888, 0.08803112804889679, 0.08871430158615112, 0.20860150456428528, -0.14125806093215942, 0.24054531753063202, -0.28913414478302, -0.09723284840583801, -0.3500290811061859, 0.015348119661211967, 0.08870019018650055, 0.3145703375339508, -0.30286291241645813, 0.1288619339466095, 0.10812981426715851, -0.06667671352624893, -0.015008230693638325, 0.2600981891155243, 0.23029488325119019, 0.11595521867275238, 0.24547173082828522, -0.045211441814899445, -0.025874607264995575, 0.14597786962985992, 0.180082306265831, 0.16942837834358215, -0.5887806415557861, 0.6027829647064209, 0.18034836649894714, -0.0968208760023117, -0.008341532200574875, 0.18996262550354004, -0.05924639478325844, -0.03979812189936638, 0.489726185798645, -0.26842519640922546, 0.03991950303316116, 0.056890394538640976, 0.01327614113688469, -0.18620270490646362, 0.6324126124382019, 0.08844924718141556, 0.49170365929603577, -0.10871750861406326, 0.026488706469535828, -0.4697314500808716, 0.12125231325626373, -0.11352185904979706, 0.1386706531047821, 0.2599012553691864, 0.2758069932460785, 0.15131691098213196, 0.17945997416973114, -0.4360817074775696, -0.18453650176525116, 0.00546585489064455, 0.09761235117912292, -0.12733064591884613, -0.1538161337375641, -0.49346429109573364, 0.120275117456913, -0.15698720514774323, -0.49851176142692566, 0.054586682468652725, -0.4139026403427124, -0.10423582792282104, -0.09382889419794083, 0.253180593252182, 0.3283693194389343, 0.46398043632507324, 0.28149116039276123, -0.033510975539684296, -0.04955582320690155, 0.04295963793992996, -0.0872577354311943, -0.11083066463470459, -0.028888264670968056, -0.3553536832332611, -0.0496390163898468, -0.17677126824855804, 0.07457371056079865, -0.27969491481781006, -0.025637540966272354, -0.3028118908405304, -0.2440079003572464, 0.01799812912940979, -0.08510939031839371, -0.12512291967868805, 0.07936549186706543, -0.29182329773902893, 0.18006496131420135, 0.32479479908943176, 0.39626315236091614, -0.1290537714958191, -0.13008853793144226, -0.2926555275917053, 0.14886412024497986, 0.3681568503379822, -0.6422930955886841, -0.17142197489738464, 0.04930239915847778, -0.15187057852745056, 0.24562564492225647, 0.0641329288482666, -0.27930429577827454, -0.20046266913414001, 0.3068371117115021, -0.12597054243087769, -0.19656680524349213, 0.19951152801513672, 0.3359585702419281, -0.05168163776397705, -0.1160435825586319, 0.27684786915779114, -0.03130447492003441, -0.4617123305797577, -0.23161359131336212, -0.43983322381973267 ]
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 ?
[ -0.1686098277568817, -0.09398830682039261, 0.011844169348478317, 0.16997693479061127, 0.27237045764923096, 0.0823572501540184, 0.2256460040807724, 0.38973069190979004, -0.050376541912555695, 0.005149416625499725, -0.08119415491819382, 0.6080824732780457, -0.14292097091674805, -0.014412950724363327, 0.2508256435394287, -0.21652516722679138, -0.10149351507425308, 0.2596185803413391, -0.1496252566576004, -0.0935942679643631, -0.014412535354495049, -0.03481736034154892, -0.14135733246803284, -0.027524136006832123, -0.06012880802154541, -0.08392547070980072, 0.07913264632225037, -0.09566854685544968, 0.3711802363395691, 0.03904271125793457, 0.2916146218776703, 0.0025849156081676483, -0.02442087233066559, 0.16511137783527374, -0.00010320143337594345, 0.32157468795776367, 0.03437269479036331, -0.1477328985929489, 0.04883991926908493, -0.07913137972354889, 0.1653173267841339, -0.2739705741405487, 0.19196182489395142, -0.3028233051300049, -0.054379045963287354, 0.008168972097337246, 0.19384194910526276, -0.02881579101085663, 0.3141268789768219, 0.1588185429573059, 0.21697206795215607, 0.4294997453689575, -0.11771450936794281, 0.3017027974128723, 0.11722411215305328, -0.17708152532577515, -0.19004741311073303, -0.22906650602817535, -0.21755391359329224, -0.014429546892642975, 0.03290630131959915, 0.26630115509033203, -0.2041044682264328, 0.13051089644432068, 0.4554975926876068, 0.1430780291557312, -0.11107172816991806, -0.3703576326370239, -0.26494020223617554, 0.378965824842453, 0.22448182106018066, -0.08395039290189743, -0.1515456736087799, -0.4547223746776581, -0.17095226049423218, -0.30649858713150024, -0.15057682991027832, 0.11813399940729141, -0.2600310742855072, 0.12261305749416351, -0.10149882733821869, -0.0880129337310791, -0.23205605149269104, 0.1448359191417694, 0.16639269888401031, -0.042912617325782776, 0.08226311951875687, 0.10815662145614624, 0.15611283481121063, 0.04901379719376564, 0.24615411460399628, 0.05889111012220383, 0.35808175802230835, 0.36975133419036865, -0.34041744470596313, 0.05369757115840912, 0.17792092263698578, -0.5166316032409668, -0.0021290257573127747, -0.2505266070365906, 0.32315099239349365, 0.17214274406433105, -0.24499277770519257, 0.1835710108280182, -0.07256843149662018, 0.01231042668223381, -0.41781705617904663, 0.1140064001083374, -0.11787016689777374, -0.4833918809890747, -0.02392812818288803, 0.021701965481042862, -0.23941385746002197, -0.04856516048312187, 0.11936338245868683, -0.46755555272102356, -0.03389650210738182, 0.2352789044380188, -0.3758569359779358, -0.37095242738723755, -0.19662556052207947, 0.056530844420194626, 0.01388191431760788, 0.27162668108940125, -0.18506407737731934, 0.42958906292915344, 0.15240806341171265, -0.1067647784948349, -0.3398880064487457, 0.004123427905142307, -0.28825172781944275, 0.12291799485683441, -0.128438800573349, -0.11231770366430283, 0.14382165670394897, -0.07857148349285126, 0.202426016330719, -0.08083003759384155, -0.17189085483551025, 0.13403822481632233, 0.28055456280708313, -0.10108020901679993, 0.08383195102214813, 0.21504929661750793, 0.05136341229081154, -0.013148441910743713, 0.04464422166347504, 0.0967692881822586, -0.16510331630706787, 0.3000507652759552, -0.5144059062004089, -0.24342922866344452, -0.03408610820770264, 0.20138007402420044, 0.11621686816215515, -0.14932671189308167, -0.25425857305526733, 0.3058803975582123, 0.13064295053482056, -0.05816652625799179, -0.17962652444839478, -0.16915199160575867, -0.41828644275665283, -0.32790136337280273, 0.31852301955223083, 0.058431752026081085, -0.44932541251182556, -0.07593970000743866, -0.02289440482854843, -0.00007997453212738037, 0.2840339243412018, 0.41180548071861267, -0.2025938779115677, 0.4759204685688019, -0.06604333221912384, 0.06232486665248871, 0.72726970911026, -0.0797073096036911, -0.2046176642179489, 0.24727146327495575, -0.18754437565803528, 0.16651341319084167, -0.05649483948945999, 0.3881370425224304, 0.014578795060515404, 0.006611668039113283, 0.38259804248809814, 0.6268607378005981, -0.13027739524841309, 0.1262756735086441, -0.16897960007190704, -0.26857030391693115, 0.30446285009384155, 0.2909621000289917, 0.05505335330963135, -0.09941775351762772, -0.2690315246582031, 0.3530552089214325, 0.14091569185256958, -0.04823555424809456, -0.1705198884010315, 0.17430956661701202, -0.10140752792358398, 0.0338936410844326, -0.19647005200386047, 0.02353718876838684, -0.6025984883308411, 0.2567143738269806, 0.21885161101818085, 0.11899276822805405, 0.14820294082164764, -0.09703925251960754, 0.27089357376098633, 0.1100616380572319, 0.03228224813938141, 0.1563066691160202, 0.11244985461235046, 0.11767585575580597, -0.01731005497276783, -0.010448114946484566, -0.25076234340667725, 0.2139793187379837, -0.3762323558330536, -0.16274914145469666, -0.35741373896598816, 0.2085111141204834, 0.30177509784698486, -0.035390328615903854, 0.004790168255567551, 0.2599482834339142, -0.2557673454284668, -0.11474455147981644, 0.05725495144724846, 0.27530282735824585, -0.028167078271508217, 0.23908807337284088, -0.626108705997467, 0.3982264995574951, -0.01640932820737362, -0.06816980242729187, -0.05593923479318619, 0.14532607793807983, 0.04535271227359772, -0.17168757319450378, 0.016372691839933395, 0.22037409245967865, -0.1982620507478714, 0.0816420316696167, 0.21227948367595673, -0.25049859285354614, 0.1703198403120041, 0.02143106237053871, 0.17781007289886475, -0.09024903178215027, 0.3996802568435669, 0.05171949043869972, 0.12820933759212494, 0.2929065525531769, -0.5027151107788086, 0.24715286493301392, 0.6455687284469604, 0.0017064344137907028, -0.060609929263591766, 0.2191503643989563, -0.0046081021428108215, -0.18649226427078247, -0.11462296545505524, -0.0755285769701004, 0.22046300768852234, 0.23334906995296478, 0.004561223089694977, -0.04915616661310196, -0.17390264570713043, -0.23643594980239868, 0.22989504039287567, 0.10348096489906311, 0.3950205445289612, -0.09640935808420181, 0.04655270650982857, 0.15590442717075348, -0.21102258563041687, 0.05060579255223274, 0.22332820296287537, 0.2867368757724762, -0.20513232052326202, -0.03855033963918686, -0.2432168871164322, -0.2629225552082062, -0.07050161063671112, 0.01839357614517212, 0.10047151148319244, -0.17765039205551147, -0.06376825273036957, 0.32688218355178833, -0.005278110504150391, -0.0022929199039936066, -0.10702614486217499, 0.2660936713218689, 0.25394612550735474, -0.36759284138679504, -0.048126786947250366, -0.09654653817415237, -0.25975215435028076, 0.12326259911060333, 0.3457806706428528, 0.26507654786109924, 0.3060998022556305, 0.04240216314792633, -0.029359718784689903, 0.15714672207832336, -0.029308415949344635, 0.16931912302970886, -0.18968917429447174, 0.12898249924182892, -0.0720650851726532, 0.20253294706344604, -0.10920309275388718, -0.2199781984090805, -0.04013655334711075, -0.42298492789268494, 0.04551757872104645, 0.19250144064426422, 0.016064487397670746, 0.05312595143914223, -0.22644716501235962, -0.22007009387016296, -0.2942710816860199, -0.4235444664955139, -0.011317901313304901, 0.11504678428173065, 0.2867080271244049, 0.46687254309654236, -0.032427191734313965, 0.2744038701057434, 0.35955268144607544, 0.13477610051631927, 0.02100229822099209, -0.13184306025505066, 0.5142495632171631, -0.2914406657218933, -0.2923903167247772, -0.12593911588191986, -0.15778809785842896, 0.21387837827205658, 0.2645535469055176, -0.6998733878135681, 0.16331243515014648, -0.18016740679740906, 0.09063294529914856, -0.17192870378494263, -0.023034224286675453, 0.2796552777290344, -0.18283775448799133, -0.037739407271146774, -0.06421390175819397, 0.12822078168392181, 0.1593089997768402, -0.0026168711483478546, -0.17444071173667908, 0.2911268472671509, 0.43631649017333984, 0.2932071387767792, 0.6823476552963257, -0.049521900713443756, -0.24725846946239471, 0.24511629343032837, -0.20342181622982025, 0.17989686131477356, -0.15880486369132996, -0.12973527610301971, 0.06918278336524963, 0.04686060547828674, -0.019307183101773262, 0.2627427577972412, -0.07913021743297577, -0.19225265085697174, -0.08546385169029236, 0.19333986937999725, 0.02322906255722046, -0.2450512796640396, 0.2765367925167084, -0.19784586131572723, 0.1782020926475525, 0.10248073935508728, 0.17246127128601074, -0.38271018862724304, -0.43289726972579956, -0.02819056063890457, -0.2183409184217453, 0.36756956577301025, -0.05496903136372566, -0.3525266945362091, 0.12331695854663849, -0.6885601878166199, 0.07143362611532211, 0.09516589343547821, 0.2950817942619324, 0.044007617980241776, -0.26446810364723206, 0.02743583917617798, 0.06361382454633713, 0.5770883560180664, 0.08534595370292664, -0.0374402180314064, 0.19421914219856262, -0.06062142550945282, -0.5440307259559631, 0.027683336287736893, -0.083886057138443, -0.02656669169664383, 0.25266408920288086, 0.13720032572746277, -0.3163083791732788, -0.5172994136810303, -0.05490119010210037, 0.011884225532412529, -0.32387733459472656, -0.10303252935409546, -0.25590822100639343, -0.24362275004386902, -0.012446063570678234, -0.29287317395210266, 0.0692630335688591, 0.26373085379600525, -0.11182914674282074, -0.007646054029464722, -0.10520246624946594, -0.001755502074956894, 0.2264765202999115, 0.021024271845817566, 0.21798670291900635, 0.49741891026496887, 0.09195971488952637, -0.006784416735172272, 0.044760119169950485, -0.1746143400669098, 0.21001891791820526, -0.10321353375911713, -0.1212226003408432, -0.0029215235263109207, 0.11823958158493042, 0.3274335265159607, 0.04370652884244919, 0.04576556012034416, 0.08590245991945267, -0.2112955003976822, 0.11721108853816986, -0.2157902717590332, 0.4107756018638611, 0.16171136498451233, -0.28921425342559814, -0.10195816308259964, -0.45165905356407166, 0.4895883798599243, 0.22257594764232635, 0.10899336636066437, 0.4912770986557007, -0.17228461802005768, -0.36280930042266846, 0.4391489624977112, 0.33841994404792786, 0.5404884219169617, -0.3218809962272644, 0.4336530864238739, 0.7526194453239441, 0.48587948083877563, 0.8558774590492249, -0.3515866994857788, 0.28582507371902466, -0.15107737481594086, -0.3399323523044586, 0.00625995360314846, -0.22546721994876862, -0.40141764283180237, 0.06774863600730896, 0.068091481924057, 0.1856304109096527, 0.08648006618022919, 0.22489555180072784, 0.17575769126415253, 0.3943643271923065, 0.36009690165519714, -0.4310414791107178, -0.214525043964386, 0.14980143308639526, -0.03654587268829346, -0.15512952208518982, -0.05978808179497719, -0.20871591567993164, 0.18185418844223022, 0.06399624794721603, 0.0220143124461174, 0.15056127309799194, -0.18615037202835083, 0.2835206985473633, 0.05520600080490112, -0.06650624424219131, -0.013074557296931744, 0.2632400691509247, 0.12207894027233124, -0.17733296751976013, -0.0984831377863884, -0.057642001658678055, 0.02004125714302063, 0.0036127837374806404, 0.08650209754705429, -0.05334336310625076, 0.6400724053382874, -0.049178384244441986, 0.13697367906570435, 0.10900606215000153, -0.07215587794780731, -0.48971885442733765, -0.1959056705236435, -0.11551913619041443, 0.0858885869383812, 0.08319633454084396, -0.547953724861145, 0.23642683029174805, -0.20112018287181854, 0.18663056194782257, 0.1621534526348114, 0.12269800156354904, -0.19497346878051758, 0.09241079539060593, 0.09988820552825928, -0.21338355541229248, 0.2217307984828949, 0.03510220721364021, 0.10077407956123352, 0.0912485271692276, 0.3558182120323181, 0.2210889607667923, -0.025953620672225952, -0.07953093945980072, 0.08418352156877518, 0.3034347891807556, -0.06420763581991196, 0.035456523299217224, 0.04730710759758949, 0.288316011428833, 0.07017548382282257, 0.3478030860424042, -0.22790850698947906, 0.014587372541427612, -0.1992013156414032, -0.18210981786251068, -0.30683788657188416, 0.3347513675689697, -0.06750883162021637, 0.12765052914619446, -0.025996500626206398, 0.2155904620885849, -0.07443252205848694, 0.25938680768013, -0.3228645324707031, -0.04293408989906311, -0.022495359182357788, 0.0756106972694397, -0.10007309913635254, -0.12442082911729813, 0.09797748923301697, 0.09842254221439362, 0.14724940061569214, 0.1296602487564087, -0.358987033367157, -0.2738020122051239, -0.02206329256296158, 0.11938446760177612, 0.021814584732055664, -0.2305416762828827, -0.02456890419125557, -0.388022243976593, -0.3245272636413574, -0.17464721202850342, 0.10038813948631287, -0.14644457399845123, -0.29161930084228516, 0.05987213924527168, -0.11217352747917175, 0.27732694149017334, -0.09360376000404358, 0.00846327655017376, -0.17180174589157104, 0.28409406542778015, -0.06256428360939026, 0.1911998987197876, 0.13741609454154968, -0.07968723773956299, -0.3882512152194977, -0.13485881686210632, 0.1281261146068573, -0.10049576312303543, -0.0008420981466770172, -0.3645164370536804, -0.08045905828475952, -0.20343463122844696, 0.09064093977212906, 0.24004027247428894, -0.25075528025627136, 0.034920692443847656, 0.21327045559883118, 0.20128269493579865, -0.12772279977798462, -0.10396610200405121, 0.19983381032943726, 0.023196738213300705, 0.06183870509266853, 0.12107584625482559, -0.021087300032377243, 0.1645384281873703, -0.10702909529209137, 0.015114806592464447, 0.051866620779037476, -0.5421467423439026, -0.26469677686691284, 0.3222053050994873, -0.15680630505084991, -0.034058090299367905, -0.08664076775312424, 0.13565927743911743, 0.12472369521856308, 0.2581653892993927, 0.31806832551956177, 0.23370836675167084, -0.07991020381450653, -0.07542462646961212, -0.10472934693098068, -0.3642435371875763, -0.4094371497631073, 0.4181023836135864, -0.13467052578926086, 0.5021536350250244, -0.251481294631958, 0.0272148959338665, 0.1200004518032074, -0.2684727907180786, -0.1701836735010147, 0.30529215931892395, -0.21323728561401367, -0.01148919016122818, -0.24748308956623077, 0.0629420056939125, -0.1118248999118805, 0.22725602984428406, -0.06825117766857147, -0.19298408925533295, -0.18089815974235535, 0.29027456045150757, 0.0493139810860157, 0.19488759338855743, 0.07147763669490814, 0.18252313137054443, 0.23883718252182007, -0.12766583263874054, 0.2389977127313614, -0.2271493822336197, -0.09096384048461914, -0.40485456585884094, 0.02497946098446846, 0.10322326421737671, 0.33521679043769836, -0.3399101495742798, 0.12908819317817688, 0.0999295562505722, -0.10254407674074173, -0.06795084476470947, 0.2378438413143158, 0.2075454592704773, 0.1221877932548523, 0.29464390873908997, 0.07240201532840729, -0.12988510727882385, 0.14953750371932983, 0.19643737375736237, 0.17680440843105316, -0.46683070063591003, 0.49007347226142883, 0.19497531652450562, -0.020242858678102493, -0.10694965720176697, 0.1463768184185028, -0.1831081211566925, -0.089093416929245, 0.521966814994812, -0.24887987971305847, 0.04972730576992035, -0.088209368288517, 0.07930800318717957, -0.22433125972747803, 0.6939452290534973, 0.033030711114406586, 0.46245676279067993, -0.1568174660205841, 0.01007494330406189, -0.4859360158443451, 0.09705594927072525, -0.19249571859836578, 0.15192368626594543, 0.24726279079914093, 0.3328301012516022, 0.10673506557941437, 0.18085797131061554, -0.4369942545890808, -0.18489594757556915, -0.05393590033054352, 0.11275489628314972, -0.20269322395324707, -0.06505998969078064, -0.5096170902252197, 0.11405152827501297, -0.07577045261859894, -0.4709782004356384, 0.1125391274690628, -0.43344634771347046, 0.05631042271852493, -0.09103848785161972, 0.2783651649951935, 0.2807081639766693, 0.41367271542549133, 0.33369937539100647, -0.041489891707897186, -0.04722532629966736, -0.01979193463921547, -0.10569283366203308, -0.12665528059005737, 0.02345750294625759, -0.34460675716400146, -0.07571867853403091, -0.05020666494965553, 0.21799996495246887, -0.23613232374191284, -0.0374738909304142, -0.23950162529945374, -0.20596855878829956, 0.09328204393386841, -0.11421603709459305, -0.09886123985052109, 0.11413223296403885, -0.1716415286064148, 0.18321622908115387, 0.1843816041946411, 0.43261075019836426, -0.09274697303771973, -0.05656100809574127, -0.2720027565956116, -0.037574395537376404, 0.46004101634025574, -0.5970888733863831, -0.23472294211387634, 0.09746084362268448, -0.06007564067840576, 0.22572439908981323, 0.04289388656616211, -0.29877716302871704, -0.15917867422103882, 0.29100289940834045, -0.1312485933303833, -0.24028590321540833, 0.27734342217445374, 0.3229195475578308, -0.08262727409601212, -0.08083811402320862, 0.29049015045166016, -0.0606854110956192, -0.3934634029865265, -0.15429000556468964, -0.45488864183425903 ]
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']} ```
[ -0.12176664173603058, -0.02593163400888443, 0.047091420739889145, 0.1898202747106552, 0.19800370931625366, 0.15442749857902527, 0.24837557971477509, 0.45679691433906555, -0.037150345742702484, -0.05661487579345703, -0.14168617129325867, 0.6093761324882507, -0.1823762059211731, -0.03136494383215904, 0.24518440663814545, -0.16024306416511536, -0.05153343826532364, 0.2101798951625824, -0.08378109335899353, -0.05111920088529587, 0.04143429547548294, -0.10304632782936096, -0.11337640881538391, -0.08961835503578186, -0.08411601930856705, 0.037810683250427246, 0.02680412121117115, -0.15933364629745483, 0.4063459038734436, -0.015848200768232346, 0.34805411100387573, 0.0354810431599617, -0.061999354511499405, 0.10191073268651962, -0.00011394677858334035, 0.38300830125808716, 0.03163420408964157, -0.1697196662425995, 0.005115790292620659, -0.04868346452713013, 0.18722325563430786, -0.22094641625881195, 0.24761855602264404, -0.2797371447086334, 0.022960029542446136, -0.04330676421523094, 0.228247731924057, 0.02149239182472229, 0.274108350276947, 0.05634458735585213, 0.11358040571212769, 0.4625847637653351, -0.22616219520568848, 0.3665572702884674, 0.15347087383270264, -0.10670578479766846, -0.13823066651821136, -0.2763128876686096, -0.21094036102294922, 0.041313812136650085, 0.0462956503033638, 0.2148604542016983, -0.2967182695865631, 0.10010874271392822, 0.48837122321128845, 0.13461041450500488, -0.11587323993444443, -0.29013538360595703, -0.20416900515556335, 0.3537537455558777, 0.1901262253522873, -0.07160364091396332, -0.176860049366951, -0.5691706538200378, -0.2341982126235962, -0.21981756389141083, -0.214105486869812, 0.04874373972415924, -0.1875368356704712, 0.1386624127626419, -0.21671627461910248, -0.09149636328220367, -0.19212813675403595, 0.16961535811424255, 0.17300380766391754, 0.08847001194953918, 0.15939004719257355, 0.1527353972196579, 0.16368268430233002, 0.0402703583240509, 0.27683913707733154, -0.011180374771356583, 0.29415324330329895, 0.4132719337940216, -0.3397204875946045, -0.0049701593816280365, 0.14997586607933044, -0.5529506206512451, -0.002599624916911125, -0.21910403668880463, 0.31478384137153625, 0.21479743719100952, -0.25012966990470886, 0.17884549498558044, -0.006411362439393997, -0.06401263177394867, -0.396527498960495, 0.14390519261360168, -0.17764908075332642, -0.5946309566497803, 0.0709008276462555, 0.02544734627008438, -0.29087966680526733, 0.09226294606924057, 0.05414499714970589, -0.5405896306037903, -0.0903407409787178, 0.227428138256073, -0.3082317113876343, -0.45655399560928345, -0.24953967332839966, 0.03222159296274185, 0.02825886756181717, 0.26590344309806824, -0.17248530685901642, 0.4949045777320862, 0.0468011274933815, -0.12600718438625336, -0.2357965111732483, 0.07146383821964264, -0.20548024773597717, 0.08548416197299957, -0.05327289551496506, -0.0908421128988266, 0.18941828608512878, -0.0446491502225399, 0.23619991540908813, -0.18066555261611938, -0.15809719264507294, 0.1609501838684082, 0.2584719657897949, -0.09254209697246552, 0.07164009660482407, 0.2176823914051056, -0.014462517574429512, 0.06174687296152115, 0.047353651374578476, 0.11339722573757172, -0.23381265997886658, 0.35670915246009827, -0.5510233640670776, -0.22753360867500305, -0.004840852692723274, 0.08490916341543198, 0.10352516174316406, -0.23398065567016602, -0.3640980124473572, 0.35490846633911133, 0.10241933912038803, -0.12468791007995605, -0.14371708035469055, -0.14855515956878662, -0.4110416769981384, -0.26339447498321533, 0.2886543869972229, 0.11818336695432663, -0.37230175733566284, -0.1996258646249771, 0.013992603868246078, 0.03928866237401962, 0.2183590829372406, 0.44404229521751404, -0.22877204418182373, 0.4910328686237335, -0.06854419410228729, 0.17658208310604095, 0.7466118335723877, -0.1683417558670044, -0.21639244258403778, 0.2665832042694092, -0.10920847207307816, 0.20277346670627594, 0.007078424096107483, 0.4875006675720215, 0.05561169609427452, -0.028244778513908386, 0.327843576669693, 0.6070598363876343, -0.11004224419593811, 0.09091212600469589, -0.06263850629329681, -0.35401713848114014, 0.38336843252182007, 0.22360147535800934, 0.04943356290459633, -0.08908843249082565, -0.24513199925422668, 0.48383060097694397, 0.15792742371559143, 0.006073672324419022, -0.18220338225364685, 0.1302788257598877, -0.18597282469272614, 0.005775615572929382, -0.22100533545017242, 0.0439634770154953, -0.626213550567627, 0.2945079207420349, 0.2898067235946655, 0.019800759851932526, 0.2248508483171463, -0.1320510059595108, 0.3046323359012604, 0.08697177469730377, 0.05847340449690819, 0.15581576526165009, -0.03637566789984703, 0.15324527025222778, -0.008286310359835625, -0.05614299327135086, -0.3398273289203644, 0.2135774791240692, -0.35561296343803406, -0.08592253923416138, -0.2873382270336151, 0.26861265301704407, 0.4462813138961792, 0.006337590515613556, -0.13960552215576172, 0.2558356821537018, -0.23948130011558533, -0.09090784192085266, 0.09431416541337967, 0.22145769000053406, -0.03742823749780655, 0.3102300763130188, -0.6426635384559631, 0.5175512433052063, -0.00729936920106411, -0.05198519304394722, -0.03803402930498123, 0.15125687420368195, 0.04083907604217529, -0.22987133264541626, -0.04504137486219406, 0.20381388068199158, -0.18118439614772797, 0.252112478017807, 0.22757010161876678, -0.29739847779273987, 0.08644019067287445, 0.09497058391571045, 0.14749155938625336, -0.10604110360145569, 0.38184425234794617, 0.0820869654417038, 0.2589591443538666, 0.29437315464019775, -0.4201216995716095, 0.20113766193389893, 0.5873521566390991, -0.03383904695510864, -0.08823416382074356, 0.2429153025150299, -0.08980477601289749, -0.21951758861541748, -0.04968205839395523, -0.1579289585351944, 0.23874813318252563, 0.15569506585597992, 0.04756339266896248, -0.01836027018725872, -0.08513352274894714, -0.189578115940094, 0.23301160335540771, 0.13817881047725677, 0.34429094195365906, -0.10489457100629807, 0.07945506274700165, 0.24394959211349487, -0.20069316029548645, 0.12063638865947723, 0.20413264632225037, 0.16148771345615387, -0.18216632306575775, 0.020874956622719765, -0.2740585505962372, -0.3233441710472107, -0.06881661713123322, -0.01563415303826332, 0.11983832716941833, -0.1766922026872635, -0.08395637571811676, 0.35642576217651367, 0.03898804262280464, -0.026908492669463158, -0.12038229405879974, 0.3014894723892212, 0.1829606145620346, -0.4852120280265808, -0.1409328579902649, -0.0722895935177803, -0.1876290738582611, 0.012351103127002716, 0.341431200504303, 0.24763786792755127, 0.21397319436073303, 0.1156187579035759, -0.10640104860067368, 0.09337572008371353, -0.13714498281478882, 0.22248098254203796, -0.14118948578834534, 0.07441708445549011, -0.09138825535774231, 0.0858772024512291, -0.13170884549617767, -0.23741337656974792, -0.13842526078224182, -0.4001227915287018, 0.015284059569239616, 0.178102508187294, 0.04398305341601372, 0.14078150689601898, -0.17305465042591095, -0.054111726582050323, -0.31138139963150024, -0.368055135011673, 0.025136403739452362, 0.164383202791214, 0.2630552053451538, 0.42665642499923706, -0.09771574288606644, 0.2875244617462158, 0.2841937243938446, 0.05660056322813034, 0.0843755453824997, -0.059109944850206375, 0.5510886907577515, -0.21195615828037262, -0.20756618678569794, -0.1972452700138092, -0.1682773232460022, 0.12386772036552429, 0.2879939675331116, -0.7350272536277771, 0.25820624828338623, -0.14762330055236816, 0.05228989198803902, -0.136073037981987, -0.04323390871286392, 0.3010389506816864, -0.24622800946235657, 0.05798213928937912, -0.01921362616121769, 0.042981427162885666, 0.07409325987100601, 0.015041571110486984, -0.21246114373207092, 0.2835451662540436, 0.42533111572265625, 0.35637781023979187, 0.8198503255844116, -0.08152447640895844, -0.19571217894554138, 0.21599642932415009, -0.2431856095790863, 0.2042168378829956, -0.2124851942062378, -0.07977622747421265, 0.073217011988163, 0.014921128749847412, -0.019206777215003967, 0.30555546283721924, -0.1208544597029686, -0.2351139485836029, -0.06570623815059662, 0.16245457530021667, 0.07112404704093933, -0.18955563008785248, 0.2537774443626404, -0.23828265070915222, 0.12286156415939331, 0.14369387924671173, 0.2271645963191986, -0.39819368720054626, -0.5145755410194397, -0.052926596254110336, -0.19155599176883698, 0.34394755959510803, -0.038793012499809265, -0.40476492047309875, 0.18230606615543365, -0.7144542932510376, 0.13730914890766144, 0.05638710781931877, 0.3346385359764099, 0.10766465961933136, -0.2996092736721039, 0.02329104393720627, 0.07969065755605698, 0.6119204759597778, 0.16484849154949188, -0.07098306715488434, 0.17361631989479065, -0.15104737877845764, -0.5380707383155823, 0.012870658189058304, -0.05156503990292549, 0.03713851794600487, 0.2749985158443451, 0.1597067415714264, -0.30154356360435486, -0.5077016949653625, -0.04973512887954712, 0.006742892786860466, -0.35283520817756653, -0.06370021402835846, -0.22928370535373688, -0.15957221388816833, -0.036567918956279755, -0.31081509590148926, 0.07019171863794327, 0.2330520749092102, -0.05486302077770233, 0.035344574600458145, -0.23602719604969025, -0.07313291728496552, 0.2959987223148346, -0.007531940471380949, 0.17190706729888916, 0.555413544178009, 0.0686982125043869, -0.05782347917556763, 0.028072133660316467, -0.2228824645280838, 0.2243756800889969, -0.046551648527383804, -0.17790314555168152, 0.020541710779070854, 0.13925306499004364, 0.4212951064109802, 0.0915336161851883, 0.04267314821481705, 0.1779651641845703, -0.18196235597133636, 0.02648220583796501, -0.23252315819263458, 0.45168939232826233, 0.10564132034778595, -0.2878836691379547, -0.11276016384363174, -0.41713830828666687, 0.5051949620246887, 0.26663684844970703, 0.09778295457363129, 0.4054182767868042, -0.1845548301935196, -0.37959763407707214, 0.474712610244751, 0.3681877851486206, 0.5930814146995544, -0.3568763732910156, 0.522830069065094, 0.6306702494621277, 0.513515293598175, 0.9229066967964172, -0.36990708112716675, 0.25785937905311584, -0.10761429369449615, -0.2891255021095276, -0.03414952754974365, -0.270114004611969, -0.36774057149887085, 0.09762480854988098, 0.14890345931053162, 0.2554055154323578, 0.05960749089717865, 0.3281601369380951, 0.22872522473335266, 0.4213240146636963, 0.400400310754776, -0.4190787374973297, -0.1393458992242813, 0.04883725941181183, -0.0004682987928390503, -0.1998964101076126, -0.08649679273366928, -0.13067443668842316, 0.22034922242164612, -0.03784758970141411, -0.06312507390975952, 0.09271541237831116, -0.19131918251514435, 0.27331072092056274, 0.08860552310943604, -0.022878140211105347, -0.030652809888124466, 0.2359609305858612, 0.05160657316446304, -0.16254596412181854, -0.07398046553134918, -0.07821480929851532, 0.020682163536548615, 0.012831945903599262, 0.13242071866989136, -0.0732160210609436, 0.5777001976966858, -0.0037469007074832916, 0.23270893096923828, 0.20195773243904114, -0.07018744200468063, -0.6475852727890015, -0.2420337200164795, -0.0774628072977066, 0.16460154950618744, 0.12083669006824493, -0.5740079283714294, 0.27751532196998596, -0.216440349817276, 0.20163749158382416, 0.04024412855505943, 0.05791104584932327, -0.18905197083950043, 0.057202037423849106, 0.08815030008554459, -0.22422657907009125, 0.23804382979869843, 0.07987484335899353, 0.080214723944664, 0.00434209406375885, 0.4003141224384308, 0.09414893388748169, 0.026444852352142334, 0.0004956014454364777, 0.140666663646698, 0.318990558385849, -0.053173262625932693, -0.035519614815711975, -0.03209904208779335, 0.22842811048030853, -0.0022147579584270716, 0.31702908873558044, -0.2210991531610489, 0.04519498348236084, -0.3271239399909973, -0.22691549360752106, -0.27978166937828064, 0.40973973274230957, -0.03227889910340309, 0.17035385966300964, -0.029661305248737335, 0.23653820157051086, -0.21207377314567566, 0.25990185141563416, -0.19452978670597076, -0.00035261688753962517, 0.055743128061294556, 0.06459634006023407, -0.20630134642124176, -0.037182509899139404, 0.08591098338365555, 0.1474553346633911, 0.029786797240376472, 0.14584200084209442, -0.28631365299224854, -0.16016757488250732, -0.05519580841064453, 0.21042361855506897, 0.06889985501766205, -0.18370765447616577, -0.07681185752153397, -0.35104912519454956, -0.30541983246803284, -0.21098463237285614, 0.07133463025093079, -0.11096087098121643, -0.2526506781578064, -0.01770791783928871, -0.016698386520147324, 0.3033115565776825, -0.18731701374053955, 0.05895691365003586, -0.03881070017814636, 0.3205208480358124, -0.06789536029100418, 0.2631838917732239, 0.17212143540382385, -0.016811057925224304, -0.3609761893749237, -0.11104582995176315, 0.2617635130882263, -0.12928399443626404, -0.019321009516716003, -0.354259192943573, -0.028004365041851997, -0.24651046097278595, 0.019432909786701202, 0.38171759247779846, -0.2598596215248108, 0.011901510879397392, 0.25529196858406067, 0.0760611817240715, -0.069309301674366, -0.13709065318107605, 0.289845734834671, 0.0156412310898304, -0.01702931523323059, 0.21574857831001282, 0.01905597373843193, 0.23018860816955566, -0.07639765739440918, -0.059450481086969376, 0.05092654377222061, -0.45813047885894775, -0.24831125140190125, 0.3442261219024658, -0.14797531068325043, -0.04453134164214134, -0.156599760055542, 0.1595495194196701, 0.1588200032711029, 0.20096920430660248, 0.268885999917984, 0.27205371856689453, -0.13097810745239258, -0.08041979372501373, -0.10907384008169174, -0.29884976148605347, -0.41599294543266296, 0.4332541525363922, -0.030950263142585754, 0.5543476343154907, -0.3014596402645111, -0.04429031163454056, 0.21874772012233734, -0.2889840006828308, -0.14263536036014557, 0.28795987367630005, -0.1215730682015419, -0.09826506674289703, -0.18458379805088043, 0.0626409575343132, -0.2162010222673416, 0.2093060165643692, -0.12552742660045624, -0.24214600026607513, -0.18847361207008362, 0.28286221623420715, 0.12130404263734818, 0.14187608659267426, 0.03304450586438179, 0.13925668597221375, 0.26242363452911377, -0.1684403270483017, 0.28532490134239197, -0.3038650155067444, -0.023454107344150543, -0.30802375078201294, 0.05243360996246338, 0.030013471841812134, 0.32330068945884705, -0.34771162271499634, 0.13549581170082092, 0.07323811948299408, -0.05230707302689552, -0.021849362179636955, 0.25487399101257324, 0.21410419046878815, 0.11187854409217834, 0.24433214962482452, -0.04741958528757095, -0.038062963634729385, 0.1554596871137619, 0.13566061854362488, 0.10434865951538086, -0.6335431337356567, 0.5223554968833923, 0.2589547038078308, -0.03485376015305519, 0.04822508245706558, 0.2361012101173401, -0.1764226257801056, -0.11756715178489685, 0.49291664361953735, -0.26624596118927, 0.07922807335853577, 0.08962001651525497, 0.02324390783905983, -0.23603755235671997, 0.6891621947288513, 0.015659432858228683, 0.5115365982055664, -0.12947995960712433, 0.08085672557353973, -0.4833618402481079, 0.04604011029005051, -0.13137918710708618, 0.13447974622249603, 0.24390091001987457, 0.28714609146118164, 0.154861181974411, 0.21748216450214386, -0.37715381383895874, -0.2930711507797241, -0.07903548330068588, 0.005899231880903244, -0.12049275636672974, -0.09162646532058716, -0.4695655107498169, 0.20530220866203308, -0.06458187103271484, -0.4875369668006897, 0.05354753136634827, -0.31423690915107727, -0.10486914962530136, -0.06824911385774612, 0.20545005798339844, 0.27082759141921997, 0.3716528117656708, 0.3245227336883545, -0.05926666408777237, 0.002298116683959961, 0.06226669251918793, -0.10009273886680603, -0.1329045593738556, -0.028408346697688103, -0.3497496247291565, -0.08344780653715134, -0.16258271038532257, 0.16817989945411682, -0.33418509364128113, 0.013359403237700462, -0.21613335609436035, -0.3234879970550537, 0.05041768401861191, -0.2010858803987503, -0.14042305946350098, 0.09030617028474808, -0.30964651703834534, 0.31978514790534973, 0.13747623562812805, 0.47105610370635986, -0.1315186321735382, -0.038961559534072876, -0.19754278659820557, 0.1254427134990692, 0.47626370191574097, -0.696999192237854, -0.18452942371368408, 0.08059538900852203, -0.13366712629795074, 0.28647270798683167, 0.08193138241767883, -0.34681081771850586, -0.21591895818710327, 0.36738553643226624, -0.0630364716053009, -0.2189217507839203, 0.23914088308811188, 0.3046948313713074, -0.10874490439891815, -0.11808691918849945, 0.30738404393196106, -0.08267609775066376, -0.3657676577568054, -0.19985421001911163, -0.5045340657234192 ]
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
[ -0.1305495947599411, -0.03085968643426895, 0.03630451485514641, 0.12274200469255447, 0.22812069952487946, 0.09177563339471817, 0.27305594086647034, 0.4065890908241272, -0.11111760139465332, 0.0027905330061912537, -0.08880870044231415, 0.5513769388198853, -0.15211911499500275, 0.04721108078956604, 0.3196074664592743, -0.26276540756225586, -0.06650402396917343, 0.24885836243629456, -0.1298072785139084, -0.014312822371721268, -0.04055114462971687, -0.053230561316013336, -0.09556209295988083, -0.016439586877822876, 0.015059834346175194, -0.03353513032197952, 0.051149748265743256, -0.1520296335220337, 0.4169294536113739, 0.021076910197734833, 0.35533690452575684, 0.0021883435547351837, -0.03409393131732941, 0.11227062344551086, -0.00010922730143647641, 0.38452446460723877, 0.03214535117149353, -0.14960576593875885, 0.06209937110543251, -0.07406571507453918, 0.20812636613845825, -0.22736459970474243, 0.1988503485918045, -0.2661628723144531, -0.04680870845913887, -0.005252291448414326, 0.23685553669929504, -0.03428581357002258, 0.257424920797348, 0.11936882883310318, 0.15609541535377502, 0.5252828598022461, -0.173379048705101, 0.2786279022693634, 0.1325308084487915, -0.15955960750579834, -0.21689723432064056, -0.27933022379875183, -0.13236470520496368, -0.010954213328659534, 0.06574587523937225, 0.14796696603298187, -0.22829397022724152, 0.09629996120929718, 0.4325977861881256, 0.15464548766613007, -0.11124914139509201, -0.3790856897830963, -0.2610156834125519, 0.4085841774940491, 0.2015363872051239, -0.07298093289136887, -0.17706947028636932, -0.4851957857608795, -0.17577533423900604, -0.2934437394142151, -0.17836517095565796, 0.08686387538909912, -0.2625954747200012, 0.18190957605838776, -0.10118096321821213, -0.12041530758142471, -0.21891944110393524, 0.12010437995195389, 0.21201223134994507, -0.03557460010051727, 0.16041889786720276, 0.11686260998249054, 0.22441290318965912, 0.1082771047949791, 0.349454402923584, 0.08066010475158691, 0.354869544506073, 0.36756548285484314, -0.2840503752231598, -0.011491380631923676, 0.16927161812782288, -0.5806731581687927, -0.044434115290641785, -0.25259649753570557, 0.35710659623146057, 0.18533852696418762, -0.29192599654197693, 0.21650972962379456, -0.013307638466358185, -0.07356643676757812, -0.4224946200847626, 0.1319459229707718, -0.13990892469882965, -0.5050941705703735, 0.04702812060713768, 0.04630619287490845, -0.22082199156284332, -0.01992803066968918, 0.060599204152822495, -0.5112510323524475, -0.14770323038101196, 0.2690739929676056, -0.2849743962287903, -0.4418402910232544, -0.2301689088344574, 0.028392238542437553, -0.01433276291936636, 0.1737464815378189, -0.12791681289672852, 0.5364341735839844, 0.13731825351715088, -0.12483684718608856, -0.2813413143157959, -0.02957770600914955, -0.23230744898319244, 0.1130899041891098, -0.06146424263715744, -0.09602802246809006, 0.15568780899047852, -0.16067297756671906, 0.17979377508163452, -0.12268876284360886, -0.12777183949947357, 0.181620255112648, 0.2595548629760742, -0.08195087313652039, 0.1001741960644722, 0.257914662361145, 0.014886215329170227, -0.06456946581602097, 0.0382283553481102, 0.07491455972194672, -0.21613088250160217, 0.3539184033870697, -0.5225857496261597, -0.28902292251586914, -0.048797063529491425, 0.13496698439121246, 0.08778078109025955, -0.2136901319026947, -0.3331679701805115, 0.3261956572532654, 0.12096337974071503, -0.1192808598279953, -0.18521875143051147, -0.20750415325164795, -0.42386937141418457, -0.3011671304702759, 0.30410706996917725, 0.04064252972602844, -0.46133387088775635, -0.11523310840129852, 0.006508614402264357, -0.01659354940056801, 0.2493128478527069, 0.44560661911964417, -0.23294807970523834, 0.4801928400993347, -0.05343689024448395, 0.0012748390436172485, 0.7598667144775391, -0.08542155474424362, -0.16154108941555023, 0.22563154995441437, -0.12333734333515167, 0.20281225442886353, -0.004924893379211426, 0.40537595748901367, 0.00034402497112751007, -0.029118189588189125, 0.3868638575077057, 0.6318569183349609, -0.15841248631477356, 0.07676413655281067, -0.10929078608751297, -0.3459326922893524, 0.30011415481567383, 0.279721200466156, 0.09528972208499908, -0.05658244341611862, -0.26506099104881287, 0.3954322636127472, 0.13062487542629242, -0.0343218557536602, -0.18207566440105438, 0.0670047253370285, -0.1286899298429489, 0.00043360888957977295, -0.17078568041324615, 0.06576729565858841, -0.6022964119911194, 0.2533167898654938, 0.1912776380777359, 0.09608078002929688, 0.1893984079360962, -0.11190374940633774, 0.36263155937194824, 0.10908060520887375, 0.11826247721910477, 0.15904927253723145, 0.019094761461019516, 0.0754808783531189, -0.028345119208097458, -0.011725282296538353, -0.34432217478752136, 0.1589709371328354, -0.37544140219688416, -0.12695880234241486, -0.34896767139434814, 0.26449695229530334, 0.37324219942092896, 0.01796679198741913, -0.08456045389175415, 0.30133238434791565, -0.3664577305316925, -0.12099925428628922, 0.11764286458492279, 0.21522921323776245, -0.030569130554795265, 0.29321372509002686, -0.6672239303588867, 0.4424225687980652, 0.05159900337457657, -0.022755734622478485, -0.10921411961317062, 0.12416113913059235, 0.0025713806971907616, -0.19744905829429626, 0.010840348899364471, 0.28054291009902954, -0.22586210072040558, 0.1471133828163147, 0.2610127925872803, -0.2442212700843811, 0.15271605551242828, 0.0638124942779541, 0.1342633068561554, -0.11270657181739807, 0.3706355690956116, 0.09925232827663422, 0.13777045905590057, 0.2951819598674774, -0.5292934775352478, 0.23122292757034302, 0.6706404685974121, -0.022198140621185303, -0.11366260796785355, 0.21687638759613037, 0.03733411058783531, -0.21183159947395325, -0.08162235468626022, -0.08116958290338516, 0.20193099975585938, 0.1457822173833847, 0.041635867208242416, 0.0004894677549600601, -0.1641395390033722, -0.25839439034461975, 0.24762287735939026, 0.13116759061813354, 0.38085076212882996, -0.1395995318889618, 0.030044760555028915, 0.20352737605571747, -0.19155406951904297, 0.08323949575424194, 0.2106477916240692, 0.17547309398651123, -0.2056153416633606, 0.03021705336868763, -0.27456989884376526, -0.30668652057647705, -0.08983340859413147, 0.035019807517528534, 0.14405672252178192, -0.18368183076381683, -0.049924641847610474, 0.3294704854488373, -0.04942644387483597, -0.05361642688512802, -0.11888295412063599, 0.3149208426475525, 0.16494840383529663, -0.4449191689491272, -0.06582370400428772, -0.105502188205719, -0.18507158756256104, 0.09022971987724304, 0.36076802015304565, 0.17744196951389313, 0.2429792881011963, 0.026393840089440346, -0.08753322809934616, 0.168130025267601, -0.07798132300376892, 0.2226693034172058, -0.11164140701293945, 0.11009330302476883, -0.04960907995700836, 0.07181691378355026, -0.12877157330513, -0.2588932514190674, -0.11553194373846054, -0.4312821328639984, 0.053853221237659454, 0.1550912708044052, -0.0024249479174613953, 0.14857839047908783, -0.2426670342683792, -0.17587366700172424, -0.30553802847862244, -0.39437296986579895, 0.018097572028636932, 0.16430489718914032, 0.3020918369293213, 0.42609962821006775, -0.029929064214229584, 0.28222182393074036, 0.38556820154190063, 0.05240762233734131, 0.08026604354381561, -0.13831505179405212, 0.47845780849456787, -0.27619481086730957, -0.25517192482948303, -0.15608848631381989, -0.17423512041568756, 0.1930471509695053, 0.3186531066894531, -0.7124727964401245, 0.14556282758712769, -0.08327458798885345, 0.08658705651760101, -0.08990909159183502, -0.04109260439872742, 0.2317248433828354, -0.2033177763223648, 0.020511122420430183, 0.004608724266290665, 0.1650746613740921, 0.14517617225646973, 0.024806426838040352, -0.20415616035461426, 0.370391309261322, 0.423714816570282, 0.33127135038375854, 0.7575738430023193, -0.08429160714149475, -0.21429266035556793, 0.23427140712738037, -0.1822497695684433, 0.1970280408859253, -0.23281565308570862, -0.06427554786205292, 0.042938943952322006, -0.013692915439605713, -0.005353173241019249, 0.22368408739566803, -0.10000864416360855, -0.1953769326210022, -0.11733990907669067, 0.22071102261543274, 0.11836301535367966, -0.23836761713027954, 0.25822973251342773, -0.21444028615951538, 0.1831321120262146, 0.1111203283071518, 0.21261000633239746, -0.3299848735332489, -0.39972004294395447, -0.019310519099235535, -0.23272426426410675, 0.4178357422351837, -0.04035956785082817, -0.3286675810813904, 0.10396482050418854, -0.7131766676902771, 0.06532415747642517, 0.11960602551698685, 0.3085826337337494, 0.08994682133197784, -0.29901477694511414, -0.0069963932037353516, 0.0677560344338417, 0.54835045337677, 0.0983370915055275, -0.08745630085468292, 0.2097495198249817, -0.17732767760753632, -0.4892943799495697, 0.034078385680913925, -0.05583801865577698, -0.0789330005645752, 0.24943751096725464, 0.14017444849014282, -0.2619260549545288, -0.49131250381469727, -0.024556774646043777, -0.01641920953989029, -0.33952704071998596, -0.048820044845342636, -0.1868024468421936, -0.2434476912021637, 0.0010784752666950226, -0.3156525492668152, 0.05761398375034332, 0.2791344225406647, -0.1710854470729828, 0.020696353167295456, -0.15080365538597107, -0.0014762766659259796, 0.21223333477973938, 0.0035507548600435257, 0.16318711638450623, 0.46328669786453247, 0.0473945327103138, -0.04891367256641388, -0.008087970316410065, -0.13080638647079468, 0.22405441105365753, -0.09222035109996796, -0.15226924419403076, 0.01875913515686989, 0.15853534638881683, 0.3719961941242218, 0.04845401644706726, 0.036095090210437775, 0.12852442264556885, -0.17415043711662292, 0.0331273153424263, -0.19866780936717987, 0.36705324053764343, 0.08676327019929886, -0.34173792600631714, -0.08246486634016037, -0.5109409093856812, 0.5615847706794739, 0.29596754908561707, 0.12883765995502472, 0.4830287992954254, -0.19704554975032806, -0.36169514060020447, 0.5212909579277039, 0.38263067603111267, 0.5099002718925476, -0.3703688383102417, 0.5261975526809692, 0.7229405045509338, 0.4783591032028198, 0.9239943027496338, -0.3590673804283142, 0.271091103553772, -0.12388913333415985, -0.35866448283195496, -0.028797602280974388, -0.2556568682193756, -0.37506231665611267, 0.07888748496770859, 0.10311451554298401, 0.21673324704170227, 0.13108590245246887, 0.27752795815467834, 0.2179107964038849, 0.3121587336063385, 0.3466190695762634, -0.4222288429737091, -0.1941184550523758, 0.06759946048259735, 0.029670272022485733, -0.23409509658813477, -0.08185210824012756, -0.1964038461446762, 0.23074984550476074, 0.06485637277364731, 0.02035634219646454, 0.12005943804979324, -0.18482458591461182, 0.28192728757858276, 0.13958941400051117, -0.09903085976839066, 0.0404784195125103, 0.2417283058166504, 0.12007851898670197, -0.20490717887878418, -0.10917830467224121, -0.08963549137115479, 0.06531937420368195, 0.010170526802539825, 0.14681871235370636, -0.044765181839466095, 0.6814284920692444, -0.021972253918647766, 0.19641372561454773, 0.16399000585079193, -0.07861166447401047, -0.5933405160903931, -0.18238696455955505, -0.11446219682693481, 0.08533680438995361, 0.16856367886066437, -0.5447580814361572, 0.2875629663467407, -0.19816642999649048, 0.18102926015853882, 0.0979793518781662, 0.11959955096244812, -0.20461025834083557, 0.054966479539871216, 0.05523791164159775, -0.18702730536460876, 0.22112686932086945, 0.02921537682414055, 0.12724709510803223, 0.03463389724493027, 0.3529255986213684, 0.22735239565372467, 0.004736330360174179, -0.025911346077919006, 0.10443621128797531, 0.2849634885787964, -0.04618552699685097, 0.06443589180707932, 0.03664505109190941, 0.22891287505626678, 0.06418915092945099, 0.3598421514034271, -0.22694502770900726, 0.031078461557626724, -0.2919079661369324, -0.17317941784858704, -0.3018862009048462, 0.36624377965927124, -0.05979655310511589, 0.1284087598323822, -0.0037668123841285706, 0.21896661818027496, -0.12371939420700073, 0.22889870405197144, -0.24168671667575836, -0.02274310775101185, 0.005892595276236534, 0.06760427355766296, -0.12601542472839355, -0.0586392804980278, 0.06964524835348129, 0.13090012967586517, 0.06438510864973068, 0.1422179788351059, -0.3747166693210602, -0.20188695192337036, -0.04005315527319908, 0.16559568047523499, 0.032665714621543884, -0.23869407176971436, -0.004422037396579981, -0.34728121757507324, -0.33256155252456665, -0.16482393443584442, 0.1165752112865448, -0.12507416307926178, -0.2715752124786377, 0.060311850160360336, -0.059461016207933426, 0.2978954017162323, -0.13933756947517395, 0.03453737124800682, -0.1070035994052887, 0.313271164894104, -0.0834675282239914, 0.24981437623500824, 0.09919039905071259, -0.05009351670742035, -0.3136667013168335, -0.11679358035326004, 0.13764718174934387, -0.11599984019994736, 0.06822923570871353, -0.31335949897766113, -0.07454542815685272, -0.2488783597946167, 0.042872995138168335, 0.2715889811515808, -0.23715418577194214, 0.09041175246238708, 0.2457917034626007, 0.11498203873634338, -0.09760560840368271, -0.1507219523191452, 0.20687197148799896, 0.08234623819589615, -0.02431715466082096, 0.20025314390659332, 0.08809411525726318, 0.1603432446718216, -0.06396119296550751, -0.040288254618644714, 0.010186681523919106, -0.5695468187332153, -0.1938687264919281, 0.4081520438194275, -0.13632269203662872, -0.01425013318657875, -0.1612708568572998, 0.09546278417110443, 0.1648012101650238, 0.21101638674736023, 0.2794186770915985, 0.22885382175445557, -0.071799635887146, -0.09621991217136383, -0.06139986962080002, -0.35142725706100464, -0.42889073491096497, 0.4405210316181183, -0.061253905296325684, 0.5847461819648743, -0.2274622917175293, 0.03551597520709038, 0.1482982188463211, -0.21972991526126862, -0.16932755708694458, 0.2957035303115845, -0.14875149726867676, -0.06481284648180008, -0.23912541568279266, 0.06481736153364182, -0.15066874027252197, 0.2991454601287842, -0.10230936110019684, -0.21432216465473175, -0.16765889525413513, 0.2694036364555359, 0.09658157080411911, 0.22632597386837006, 0.11800218373537064, 0.07471579313278198, 0.24854061007499695, -0.11725829541683197, 0.19789016246795654, -0.2634506821632385, -0.10788576304912567, -0.3696136176586151, 0.011613085865974426, 0.0912146270275116, 0.2826961874961853, -0.333754301071167, 0.12661029398441315, 0.07012587785720825, -0.045848943293094635, -0.02080637589097023, 0.2440701723098755, 0.2368536740541458, 0.09861889481544495, 0.30468040704727173, -0.0035823434591293335, -0.05681198090314865, 0.1515500694513321, 0.15705440938472748, 0.1753876507282257, -0.5788911581039429, 0.5717279314994812, 0.2261790782213211, -0.055577121675014496, -0.005985397845506668, 0.20326685905456543, -0.08927394449710846, -0.12076389789581299, 0.5214148759841919, -0.2620495855808258, 0.0367177277803421, -0.00250809732824564, 0.046714432537555695, -0.19307464361190796, 0.6902192234992981, 0.05742118880152702, 0.4553859531879425, -0.08694545924663544, 0.04564303159713745, -0.44194960594177246, 0.09158547222614288, -0.10719521343708038, 0.211973637342453, 0.2606280446052551, 0.2983134686946869, 0.1464911252260208, 0.20516307651996613, -0.45577120780944824, -0.2137388437986374, -0.016207600012421608, 0.04839964210987091, -0.1488637924194336, -0.10379485785961151, -0.5044936537742615, 0.12454164028167725, -0.13039083778858185, -0.442399799823761, 0.08776954561471939, -0.42137792706489563, -0.034125976264476776, -0.08691303431987762, 0.2229066789150238, 0.32336729764938354, 0.4466139078140259, 0.3336485028266907, -0.039631858468055725, -0.10491986572742462, 0.049597591161727905, -0.0795430913567543, -0.1129736453294754, -0.027450663968920708, -0.31767064332962036, -0.07864468544721603, -0.1492992490530014, 0.1286468505859375, -0.2904026210308075, -0.047470126301050186, -0.22104136645793915, -0.26274269819259644, 0.025212854146957397, -0.10299608111381531, -0.13584566116333008, 0.09924516826868057, -0.26050934195518494, 0.17538069188594818, 0.2303081601858139, 0.4477683901786804, -0.14069747924804688, -0.1299131214618683, -0.21534103155136108, 0.08180087059736252, 0.38260042667388916, -0.6043994426727295, -0.20278362929821014, 0.06831440329551697, -0.16580957174301147, 0.25549808144569397, 0.06379912793636322, -0.27773869037628174, -0.21997889876365662, 0.27987343072891235, -0.12841495871543884, -0.21454517543315887, 0.25627386569976807, 0.3233982026576996, -0.10076102614402771, -0.12213346362113953, 0.2780953645706177, -0.061141014099121094, -0.43306151032447815, -0.21358869969844818, -0.4865185022354126 ]
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)
[ -0.1738838404417038, -0.06410497426986694, -0.045787423849105835, 0.06958923488855362, -0.3965015411376953, 0.13637560606002808, 0.06761038303375244, 0.40038391947746277, -0.068243607878685, -0.3222419023513794, 0.13134296238422394, 0.18338912725448608, -0.08250486105680466, 0.07896033674478531, 0.39143094420433044, -0.04079137742519379, -0.0588856115937233, 0.2828744053840637, -0.03286244720220566, -0.03655622899532318, 0.16092781722545624, -0.0212873462587595, -0.10444429516792297, 0.007256589829921722, -0.19171856343746185, -0.34966567158699036, -0.16615350544452667, 0.15690958499908447, -0.1471918672323227, -0.4131222069263458, 0.11186324059963226, 0.0859130322933197, 0.5545092821121216, 0.1522887796163559, -0.00012087303912267089, -0.46484100818634033, 0.1461741030216217, -0.07211586833000183, -0.15252892673015594, -0.20889602601528168, -0.232600599527359, -0.09182374179363251, 0.17215833067893982, -0.3390498459339142, 0.0773116946220398, -0.4477842450141907, 0.17121951282024384, -0.2712026834487915, 0.6657806634902954, 0.3493711054325104, 0.07335950434207916, -0.1018950417637825, 0.18934036791324615, 0.2285752296447754, 0.47995370626449585, 0.542945146560669, -0.3052915334701538, 0.1529545933008194, 0.26077890396118164, -0.20762118697166443, -0.28251877427101135, 0.15879952907562256, -0.09765521436929703, -0.23909716308116913, 0.33057811856269836, 0.1180301383137703, -0.2703995108604431, -0.33307674527168274, 0.048938315361738205, 0.39643773436546326, 0.36962535977363586, -0.4938056766986847, 0.048539161682128906, -0.05016287416219711, 0.18940657377243042, -0.48908114433288574, -0.14875277876853943, 0.6180800795555115, -0.5020691752433777, 0.3597387969493866, 0.3344428539276123, -0.38164591789245605, -0.43398380279541016, 0.1284179389476776, 0.45758846402168274, -0.07156996428966522, -0.04713990166783333, -0.07582772523164749, -0.050290171056985855, 0.17449459433555603, 0.21123643219470978, -0.24556133151054382, -0.14875848591327667, 0.3470054566860199, -0.1437053680419922, -0.16275882720947266, -0.4877951145172119, -0.20781898498535156, 0.2758224308490753, -0.029754452407360077, 0.3700706660747528, -0.1301804780960083, -0.03909512981772423, -0.09167703241109848, 0.039463020861148834, -0.09501941502094269, 0.4510572552680969, 0.20620964467525482, -0.009002260863780975, -0.26278287172317505, 0.16223491728305817, -0.0809464082121849, -0.20558717846870422, 0.0966666042804718, -0.1564088761806488, 0.03321831673383713, 0.04962415248155594, -0.1927904188632965, 0.011112412437796593, 0.006019183434545994, -0.4366477131843567, -0.024570120498538017, -0.2640628516674042, 0.23352792859077454, 0.18255864083766937, -0.09135620296001434, 0.22798101603984833, 0.3980581760406494, 0.002583298832178116, -0.1910053789615631, 0.12297531217336655, 0.07351367920637131, -0.2823683023452759, 0.28089261054992676, 0.24723418056964874, 0.04876583814620972, -0.002006618771702051, -0.09525186568498611, 0.04778182506561279, -0.3019627630710602, 0.19040703773498535, -0.011526014655828476, 0.16694147884845734, -0.1192820817232132, 0.2679125964641571, 0.04077550023794174, -0.14994093775749207, -0.12755176424980164, -0.10135650634765625, 0.3692103326320648, -0.3687538504600525, -0.1211438775062561, -0.13666413724422455, -0.0014926977455615997, 0.055035948753356934, -0.14150623977184296, -0.2777971029281616, 0.25129154324531555, 0.13343475759029388, 0.19284114241600037, 0.05564342439174652, 0.2801027297973633, -0.013974577188491821, -0.25124961137771606, 0.034251868724823, 0.47603195905685425, -0.23271846771240234, 0.1828201711177826, -0.1303548663854599, 0.4220266342163086, 0.04189300164580345, 0.19998136162757874, -0.25090548396110535, 0.3373119533061981, 0.028532909229397774, 0.37845587730407715, 0.7308472394943237, -0.07829384505748749, 0.22644926607608795, 0.40336307883262634, 0.18155469000339508, -0.3522214889526367, 0.14907340705394745, -0.043225377798080444, 0.10294309258460999, -0.09913067519664764, -0.30004069209098816, 0.49250930547714233, 0.11881767213344574, 0.03809336572885513, -0.07898508757352829, -0.2024594098329544, -0.18061646819114685, 0.013829093426465988, -0.14248503744602203, -0.1521226167678833, 0.09050919860601425, -0.45889031887054443, 0.4787382483482361, -0.29847607016563416, 0.4021228551864624, 0.2903975248336792, 0.041525423526763916, -0.014199167490005493, 0.24639397859573364, -0.11176923662424088, -0.18240374326705933, -0.31211769580841064, -0.40653055906295776, 0.29062819480895996, -0.2542751431465149, -0.2743143141269684, -0.25395360589027405, -0.11944717168807983, 0.07688561081886292, -0.030594658106565475, -0.03145664557814598, -0.24361345171928406, -0.21933585405349731, 0.017764132469892502, -0.20738831162452698, 0.2594844698905945, -0.2797715365886688, 0.13498452305793762, -0.23029448091983795, 0.21568074822425842, 0.08199449628591537, -0.19584107398986816, 0.24660933017730713, 0.23807460069656372, -0.2799947261810303, 0.14895713329315186, 0.0656597912311554, 0.1418319046497345, -0.05355462804436684, 0.20256268978118896, 0.4277292490005493, -0.28081637620925903, 0.08477296680212021, -0.6659767031669617, 0.30370938777923584, -0.04134868085384369, 0.22573737800121307, -0.11210650950670242, -0.3082040846347809, 0.44492775201797485, -0.06242334470152855, 0.19242992997169495, -0.16103288531303406, 0.018283164128661156, -0.07411972433328629, -0.030854687094688416, -0.08382117003202438, 0.17987780272960663, 0.055171363055706024, 0.43464672565460205, 0.156639963388443, -0.15134316682815552, 0.027362637221813202, 0.08514806628227234, -0.12751756608486176, -0.17111240327358246, 0.2346557080745697, 0.23757702112197876, -0.3029535114765167, -0.10359880328178406, -0.1672840416431427, 0.04839325323700905, 0.09259171783924103, 0.18483607470989227, -0.15063875913619995, -0.04035823792219162, 0.08050712198019028, -0.09687428176403046, 0.3708855211734772, -0.030180826783180237, 0.2931443750858307, -0.16108748316764832, -0.0907881110906601, -0.02335536479949951, -0.17951402068138123, -0.28069770336151123, -0.04755205661058426, 0.1384667009115219, -0.2195429801940918, 0.028408849611878395, -0.06600518524646759, -0.37388554215431213, -0.3311775028705597, 0.12961038947105408, -0.04142492637038231, 0.25273483991622925, 0.012834236025810242, -0.24280181527137756, 0.16066354513168335, 0.0032888036221265793, -0.015388000756502151, 0.5403857231140137, 0.09072727710008621, 0.07187342643737793, -0.18865998089313507, -0.5223940014839172, -0.18206366896629333, 0.05640018358826637, 0.24623338878154755, 0.42041847109794617, 0.3654952347278595, -0.18795721232891083, 0.12019790709018707, -0.037212222814559937, -0.015586551278829575, 0.03923356533050537, -0.17580151557922363, -0.27563244104385376, -0.18567413091659546, 0.12907779216766357, -0.5696601867675781, -0.10519050806760788, -0.2161608785390854, 0.04164506494998932, -0.17576496303081512, -0.21181301772594452, -0.07265765964984894, -0.040174324065446854, 0.19620949029922485, -0.5246601104736328, -0.15463665127754211, -0.1268382966518402, 0.35743266344070435, 0.30721616744995117, 0.2926774024963379, -0.024484626948833466, 0.23975211381912231, 0.11661984771490097, -0.3075913190841675, -0.05353686213493347, 0.19105185568332672, -0.06540259718894958, 0.37870287895202637, -0.3172715902328491, -0.5370230078697205, 0.011981863528490067, -0.05331190302968025, -0.07275433838367462, 0.31094807386398315, -0.3548154830932617, 0.4196316599845886, 0.08230727165937424, 0.02953271195292473, -0.1032710075378418, -0.012858429923653603, 0.24516265094280243, -0.00042297039180994034, 0.06135670840740204, -0.05839308723807335, 0.09048902988433838, 0.356862872838974, -0.1284237802028656, 0.3276352286338806, 0.34489214420318604, 0.1456562876701355, -0.046614307910203934, 0.5362757444381714, 0.25174787640571594, 0.06070580333471298, 0.17636626958847046, -0.11871372908353806, 0.07755326479673386, 0.10546062886714935, 0.03186950832605362, -0.07428266108036041, 0.12698684632778168, 0.037481941282749176, 0.18022382259368896, 0.31824931502342224, -0.3227258324623108, -0.20063084363937378, -0.24340973794460297, -0.37224990129470825, -0.19315633177757263, 0.39509591460227966, -0.2958453595638275, 0.2528686821460724, -0.11756265163421631, -0.21010136604309082, -0.17794978618621826, -0.47201061248779297, -0.10274523496627808, 0.23868027329444885, 0.2641415596008301, 0.17290504276752472, -0.04200541228055954, 0.36136147379875183, -0.6206266283988953, 0.2994306683540344, 0.18296527862548828, -0.11638809740543365, 0.1148841381072998, 0.053787678480148315, 0.21332398056983948, 0.08678186684846878, -0.116925448179245, 0.1683700680732727, -0.05905257165431976, -0.014313681051135063, 0.33650946617126465, -0.3689159154891968, 0.3253720700740814, -0.28265267610549927, 0.037240538746118546, 0.22013762593269348, -0.11274115741252899, -0.4396894872188568, -0.22016751766204834, 0.17403529584407806, 0.14907994866371155, -0.11088638007640839, -0.4020352065563202, 0.12999650835990906, 0.005802426487207413, -0.3960942029953003, -0.14598968625068665, 0.39685720205307007, -0.016033083200454712, -0.12497647851705551, -0.15539239346981049, -0.0196640957146883, 0.10079912096261978, 0.22278183698654175, -0.23059993982315063, 0.2011769711971283, -0.2508758008480072, 0.02177169919013977, -0.08185725659132004, 0.3278181254863739, 0.06736361235380173, -0.001511555165052414, 0.18697330355644226, -0.22211626172065735, -0.0643140897154808, -0.3994571268558502, 0.5800670385360718, 0.1728685200214386, 0.11208881437778473, 0.15619589388370514, 0.1852518916130066, -0.3044200539588928, -0.11992905288934708, -0.011086886748671532, 0.13509565591812134, -0.11089462041854858, -0.4100247919559479, -0.3892832100391388, 0.5749382376670837, 0.04820920154452324, -0.03414585441350937, 0.2953094244003296, 0.05651567503809929, -0.29029685258865356, 0.6706217527389526, 0.20020261406898499, 1.225579023361206, -0.1759890615940094, 0.2871074974536896, 0.05487573891878128, -0.10766854882240295, 0.5923519134521484, -0.42020469903945923, -0.21309912204742432, -0.5059723258018494, 0.30423209071159363, -0.11025388538837433, -0.1454433649778366, 0.15465322136878967, 0.33180296421051025, -0.06618227809667587, 0.1046360656619072, 0.2753223776817322, -0.045742105692625046, 0.1352178156375885, 0.32489001750946045, -0.21979480981826782, -0.4144164025783539, -0.15021663904190063, -0.09407122433185577, -0.12854498624801636, 0.04692883789539337, 0.11305602639913559, -0.4174148440361023, -0.32015812397003174, 0.07482137531042099, -0.42436736822128296, 0.10035192221403122, -0.00813352968543768, -0.22221902012825012, -0.26612773537635803, -0.6077929139137268, 0.3085108697414398, 0.3429756462574005, 0.09421686083078384, -0.0077797845005989075, -0.285155713558197, 0.09172770380973816, -0.2546539306640625, -0.3908626437187195, 0.27973127365112305, -0.12988488376140594, 0.07501401752233505, -0.15044385194778442, -0.20562174916267395, 0.15109014511108398, 0.05339084565639496, -0.2871481776237488, -0.006581559777259827, -0.2250330001115799, 0.4990307092666626, -0.12148292362689972, -0.06034606322646141, 0.32706499099731445, -0.41001591086387634, 0.005483098328113556, 0.07348687946796417, 0.3840419054031372, 0.10894711315631866, 0.05277860164642334, 0.26296380162239075, -0.2531402111053467, -0.04520687088370323, 0.17500489950180054, 0.3665909469127655, -0.22578546404838562, 0.500103235244751, 0.41960784792900085, -0.15374994277954102, -0.2953294515609741, 0.276152640581131, 0.3888964056968689, -0.8429332971572876, -0.1055259183049202, -0.008339785039424896, 0.09867812693119049, 0.016229508444666862, 0.10067287087440491, -0.07750094681978226, -0.04655391722917557, 0.010311707854270935, -0.5420810580253601, -0.26967689394950867, 0.14991000294685364, -0.0966624766588211, 0.07544510066509247, 0.3981233239173889, 0.09652493894100189, 0.04153711348772049, 0.13556744158267975, -0.20735105872154236, 0.08010713011026382, 0.5916277766227722, 0.17521585524082184, 0.10622365027666092, 0.09708917140960693, 0.019881684333086014, 0.02555643394589424, -0.024785522371530533, 0.1171044334769249, 0.11115194857120514, -0.13142049312591553, -0.015419403091073036, 0.20284497737884521, -0.06086132675409317, -0.2965807318687439, -0.30982348322868347, -0.45137453079223633, -0.2555306553840637, -0.19112567603588104, -0.14729651808738708, -0.15576708316802979, -0.13286037743091583, 0.13090789318084717, 0.07797189801931381, 0.014390937983989716, -0.4827423691749573, 0.22568193078041077, -0.21510633826255798, 0.25157713890075684, -0.09725858271121979, 0.1495199054479599, -0.0002763529773801565, -0.05023433268070221, 0.04456475004553795, 0.12908262014389038, 0.24697478115558624, 0.18958786129951477, 0.5016099214553833, -0.11936318874359131, -0.04702064394950867, -0.3728875517845154, 0.20446962118148804, 0.2573812007904053, 0.09038460999727249, 0.06380896270275116, 0.2950482666492462, 0.09182357043027878, -0.2226516455411911, 0.2888721823692322, -0.23050370812416077, -0.20139402151107788, -0.18630316853523254, 0.12676364183425903, -0.021069806069135666, -0.19755569100379944, -0.31091368198394775, -0.16034024953842163, 0.18026570975780487, -0.015163816511631012, -0.3553696870803833, 0.19000335037708282, 0.22559238970279694, -0.04182204231619835, -0.1311609297990799, 0.22774817049503326, 0.2554951012134552, -0.06322768330574036, -0.05395831912755966, 0.1503247320652008, 0.7298564314842224, 0.2667449414730072, 0.1422620713710785, -0.16784663498401642, 0.041503921151161194, 0.4384757876396179, 0.2494744211435318, 0.014322001487016678, 0.23090054094791412, 0.5041629076004028, 0.6058549880981445, 0.10812121629714966, -0.04871394485235214, 0.29230350255966187, 0.12374924123287201, 0.03654119372367859, -0.4599583148956299, -0.06597040593624115, -0.04848200082778931, -0.2736518681049347, -0.19247521460056305, -0.0760037899017334, -0.14734040200710297, 0.37071287631988525, 0.15805155038833618, -0.05577407032251358, 0.16497400403022766, 0.07142594456672668, 0.324858158826828, -0.2492869645357132, 0.08210805058479309, -0.12004059553146362, 0.01536385715007782, 0.03747386485338211, 0.43583157658576965, 0.3315373957157135, 0.17218945920467377, -0.13928698003292084, 0.14918887615203857, -0.0002274494618177414, -0.10717981308698654, 0.1763480007648468, 0.19603392481803894, 0.4006015956401825, 0.03843868523836136, 0.07309971749782562, 0.045242540538311005, 0.02798810601234436, -0.12604179978370667, 0.31556347012519836, -0.21342405676841736, 0.09988749027252197, -0.15060198307037354, 0.07479558885097504, -0.3740050792694092, -0.16975471377372742, 0.09984663873910904, -0.16131079196929932, 0.3633503019809723, 0.6061455011367798, -0.23281389474868774, -0.04642936587333679, -0.11821354180574417, 0.011206597089767456, -0.10335047543048859, 0.41821733117103577, 0.3434796929359436, 0.12700296938419342, -0.29066649079322815, -0.2813267707824707, -0.047038279473781586, 0.10857333987951279, -0.03955468162894249, -0.23724059760570526, -0.05329872667789459, 0.03458254039287567, 0.2613576650619507, 0.41253042221069336, 0.04884767532348633, 0.13425299525260925, -0.11785807460546494, 0.47180449962615967, -0.28801384568214417, -0.04197234287858009, 0.27318787574768066, 0.5164781808853149, -0.1628139615058899, -0.10243487358093262, 0.15229006111621857, 0.3586495816707611, -0.06448974460363388, -0.23673920333385468, -0.08922378718852997, 0.32758402824401855, -0.17799578607082367, 0.22377021610736847, -0.09077423065900803, 0.15293294191360474, -0.05167493224143982, -0.12333330512046814, 0.004621267318725586, -0.08744807541370392, 0.10571654886007309, 0.4439932107925415, 0.3223770558834076, -0.1910894215106964, -0.26522544026374817, -0.0015395544469356537, 0.04711391031742096, 0.27495089173316956, -0.28760039806365967, -0.050848063081502914, -0.27267876267433167, -0.023377373814582825, -0.26177310943603516, 0.14743778109550476, -0.03487565740942955, 0.28765204548835754, -0.21054619550704956, 0.11583209037780762, -0.12690198421478271, -0.13028763234615326, 0.09160180389881134, -0.25090840458869934, 0.03996610268950462, 0.26852038502693176, 0.21695390343666077, 0.5681498050689697, 0.3674054443836212, -0.5660695433616638, 0.21065393090248108, 0.23423747718334198, 0.07025489211082458, 0.1036049872636795, 0.06293847411870956, 0.00258641317486763, 0.10403327643871307, -0.016969434916973114, 0.3965064287185669, -0.05848296731710434, -0.0382046215236187, -0.46145692467689514, -0.49836641550064087 ]
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 :)
[ -0.10410341620445251, -0.0871485024690628, -0.09965090453624725, 0.15440313518047333, -0.3756243586540222, 0.15951833128929138, 0.11732514202594757, 0.42783257365226746, -0.043595489114522934, -0.29375743865966797, 0.07305074483156204, 0.11510443687438965, -0.08337914943695068, 0.1475035399198532, 0.25153082609176636, -0.15385308861732483, -0.0880170688033104, 0.21654415130615234, -0.2933834493160248, 0.015711992979049683, 0.27406173944473267, -0.024749254807829857, -0.08778530359268188, 0.012240864336490631, -0.17144320905208588, -0.23625531792640686, -0.12816700339317322, -0.0033138543367385864, -0.14507731795310974, -0.3376830518245697, -0.02538456954061985, -0.02935335412621498, 0.5116754770278931, 0.20511534810066223, -0.00012170407717349008, -0.34656012058258057, 0.12741808593273163, -0.004115752875804901, -0.2411179542541504, -0.20824997127056122, -0.151987686753273, -0.009040888398885727, 0.19447427988052368, -0.3356022834777832, -0.06002156808972359, -0.3797878324985504, 0.030287371948361397, -0.46885737776756287, 0.8392431735992432, 0.3928324580192566, 0.07577163726091385, -0.16022609174251556, 0.1887146681547165, 0.2179344743490219, 0.34135088324546814, 0.5511065721511841, -0.34023723006248474, -0.07742129266262054, 0.14995825290679932, 0.03337541222572327, -0.16508617997169495, 0.19850076735019684, 0.04398871213197708, -0.153246209025383, 0.3540644347667694, -0.03162287548184395, -0.2806325852870941, -0.22104308009147644, 0.12315778434276581, 0.32636189460754395, 0.49540793895721436, -0.47507354617118835, -0.045036979019641876, -0.011971943080425262, 0.2050464153289795, -0.42492926120758057, -0.22572112083435059, 0.5759365558624268, -0.4589052200317383, 0.4854774475097656, 0.1496037244796753, -0.40089502930641174, -0.4561852514743805, -0.07455518841743469, 0.3925391137599945, -0.11834357678890228, -0.10440413653850555, -0.06381229311227798, -0.14212647080421448, 0.16296061873435974, 0.22700129449367523, -0.33754250407218933, -0.12779399752616882, 0.2690401077270508, -0.04277494177222252, -0.21517270803451538, -0.5973444581031799, -0.19733577966690063, 0.18537640571594238, 0.07732850313186646, 0.3857082426548004, -0.112907774746418, -0.04760902002453804, -0.2783753573894501, 0.07719891518354416, -0.07277937233448029, 0.43398335576057434, 0.13293291628360748, -0.007070891559123993, -0.2632562220096588, 0.17075678706169128, -0.11871539801359177, -0.09468083083629608, 0.017398886382579803, -0.1367938220500946, 0.20157131552696228, -0.010499807074666023, -0.2780986428260803, -0.03335587680339813, -0.019908487796783447, -0.3864952623844147, -0.07700014114379883, -0.2138030230998993, 0.026128772646188736, 0.20136471092700958, -0.14675667881965637, 0.22471027076244354, 0.37763166427612305, 0.05205434560775757, -0.28021350502967834, 0.17620937526226044, -0.01802181825041771, -0.23208202421665192, 0.2816554605960846, 0.08700843155384064, 0.1653803437948227, -0.13530714809894562, -0.06850451976060867, -0.058296531438827515, -0.14523959159851074, 0.25249144434928894, -0.036170680075883865, 0.21758876740932465, -0.06669114530086517, 0.20696479082107544, 0.016298608854413033, -0.12016589939594269, -0.007247496396303177, -0.07581132650375366, 0.35496294498443604, -0.18190181255340576, 0.035340290516614914, -0.04450523108243942, 0.015734387561678886, 0.07062423974275589, -0.08456777036190033, -0.2357397973537445, 0.16044221818447113, -0.0032508820295333862, 0.33366259932518005, 0.07029986381530762, 0.28684478998184204, 0.06052299961447716, -0.14232565462589264, -0.06134990602731705, 0.44516944885253906, -0.33855628967285156, 0.14817200601100922, -0.09775698930025101, 0.2030450850725174, -0.11506327241659164, 0.3096967041492462, -0.2214798778295517, 0.28552699089050293, -0.021732572466135025, 0.2480594962835312, 0.7609075903892517, -0.0031955912709236145, 0.1953616440296173, 0.49882906675338745, 0.032613471150398254, -0.3793385624885559, 0.09234792739152908, -0.08416327834129333, -0.08584770560264587, -0.029711073264479637, -0.2868659794330597, 0.5281493067741394, 0.11185752600431442, -0.0169500932097435, -0.11940550804138184, -0.17205464839935303, -0.11659656465053558, -0.053938183933496475, -0.23231525719165802, -0.05483003705739975, 0.09218277782201767, -0.4823828935623169, 0.48952716588974, -0.335669606924057, 0.45325058698654175, 0.23785249888896942, 0.025121541693806648, -0.03932981193065643, 0.17496195435523987, -0.03196074441075325, -0.0986606702208519, -0.3560510277748108, -0.3879697024822235, 0.2141292244195938, -0.32696717977523804, -0.23821283876895905, -0.2696836590766907, -0.12529027462005615, -0.06862235069274902, 0.08229755610227585, -0.01979624107480049, -0.23528045415878296, -0.15579433739185333, 0.014363888651132584, -0.21273356676101685, 0.3180287480354309, -0.11018781363964081, 0.14913520216941833, -0.22364474833011627, 0.11636769771575928, 0.1057070791721344, -0.004393856041133404, 0.2375989556312561, 0.13671763241291046, -0.23089119791984558, 0.052421558648347855, -0.054583676159381866, 0.0032957755029201508, -0.07111269235610962, 0.17900681495666504, 0.5197829604148865, -0.2920801639556885, 0.07352743297815323, -0.6826451420783997, 0.3104526400566101, 0.01753075420856476, 0.2184138298034668, -0.10102784633636475, -0.4654025137424469, 0.4807800054550171, -0.08094540238380432, 0.19033676385879517, -0.1441078782081604, 0.043296992778778076, -0.13720957934856415, -0.11262868344783783, -0.16778352856636047, 0.14391864836215973, -0.06783577799797058, 0.3569745719432831, 0.2148139476776123, -0.09206975251436234, 0.09972019493579865, 0.03115183860063553, -0.1916438192129135, -0.1972682625055313, 0.2329721748828888, 0.21440386772155762, -0.2332383096218109, -0.07656899094581604, -0.061181455850601196, 0.040917832404375076, 0.11759556829929352, 0.14910176396369934, 0.021691661328077316, 0.07660005986690521, 0.22029925882816315, -0.1371527761220932, 0.33411848545074463, -0.1321944296360016, 0.3828599154949188, -0.15953022241592407, -0.16526348888874054, 0.021199064329266548, -0.09979048371315002, -0.193245530128479, -0.1010817438364029, 0.14014826714992523, -0.08165815472602844, -0.038220860064029694, -0.13772624731063843, -0.40400800108909607, -0.3725435733795166, 0.26179951429367065, -0.03137803450226784, 0.2539513111114502, -0.0742642879486084, -0.424617737531662, 0.14733347296714783, -0.1559280902147293, -0.07536538690328598, 0.5126533508300781, 0.07902292907238007, 0.06046071648597717, -0.28361693024635315, -0.3982830345630646, -0.16304075717926025, 0.07741984724998474, 0.30698493123054504, 0.40575915575027466, 0.34915289282798767, -0.08209364116191864, 0.19486449658870697, -0.1915404498577118, 0.09424611181020737, 0.027147049084305763, -0.03498043864965439, -0.30891913175582886, -0.15635044872760773, 0.16644491255283356, -0.5549853444099426, 0.11955296993255615, -0.18406939506530762, 0.1168874204158783, -0.23893514275550842, -0.2792016863822937, -0.07527559995651245, -0.021391257643699646, 0.22103792428970337, -0.4112202525138855, -0.13717827200889587, -0.08414331823587418, 0.42231398820877075, 0.23396296799182892, 0.2500331997871399, -0.04165847972035408, 0.1320330649614334, 0.1708654910326004, -0.23383069038391113, -0.0723307654261589, 0.19056569039821625, -0.15384632349014282, 0.2510853707790375, -0.36824554204940796, -0.5960063338279724, -0.0636395588517189, -0.015842750668525696, 0.021004999056458473, 0.3126481771469116, -0.4215537905693054, 0.5755350589752197, -0.035364747047424316, 0.053930897265672684, -0.14537562429904938, -0.005995715036988258, 0.20333749055862427, -0.033618539571762085, 0.044098325073719025, -0.030090264976024628, 0.08547812700271606, 0.3154035210609436, -0.16932356357574463, 0.3524860143661499, 0.41920360922813416, 0.13491728901863098, 0.08188965916633606, 0.6222638487815857, 0.24559468030929565, 0.10145352780818939, 0.23656949400901794, -0.0248213279992342, 0.006416585296392441, 0.026079073548316956, 0.051901571452617645, -0.14044982194900513, 0.19511981308460236, 0.1238490492105484, 0.07361648231744766, 0.2893804609775543, -0.0958741158246994, -0.29619133472442627, -0.26517927646636963, -0.3516789674758911, -0.2011638581752777, 0.29699963331222534, -0.3599572479724884, 0.2795429825782776, -0.15597039461135864, -0.21718445420265198, -0.06307139992713928, -0.31133943796157837, 0.05996665731072426, 0.12868838012218475, 0.3190666437149048, 0.14906051754951477, -0.2706942558288574, 0.34821566939353943, -0.6118109822273254, 0.34974855184555054, 0.051624152809381485, -0.10029660165309906, 0.05672121420502663, 0.030564218759536743, 0.1468581110239029, 0.08642438799142838, -0.17882390320301056, 0.11098543554544449, -0.02531621605157852, 0.009351830929517746, 0.2899273633956909, -0.4226803779602051, 0.3903345763683319, -0.228807270526886, 0.11045856028795242, 0.012456780299544334, 0.1247149258852005, -0.3796062767505646, -0.2781963348388672, 0.13440150022506714, 0.24216262996196747, -0.14169588685035706, -0.36481356620788574, 0.2568521499633789, 0.008763179183006287, -0.18793970346450806, -0.11055953055620193, 0.37304532527923584, 0.020873036235570908, -0.07553976774215698, -0.19689372181892395, -0.05982205271720886, 0.15742221474647522, 0.12727102637290955, -0.3288094103336334, 0.10629283636808395, -0.27296242117881775, 0.026045743376016617, 0.05754533410072327, 0.38840243220329285, 0.008848275989294052, 0.08767350763082504, 0.17699673771858215, -0.21610409021377563, -0.06899258494377136, -0.30543607473373413, 0.6498403549194336, 0.22233113646507263, 0.12618371844291687, 0.1377880573272705, 0.21072715520858765, -0.2504120469093323, -0.07103036344051361, -0.1461544930934906, 0.04115205258131027, -0.06889695674180984, -0.4261358380317688, -0.34913817048072815, 0.5717188119888306, 0.053953658789396286, 0.0027203261852264404, 0.19814103841781616, 0.08373086154460907, -0.2717004716396332, 0.5873885154724121, 0.27925077080726624, 1.185522437095642, -0.27428150177001953, 0.27304890751838684, -0.1355077624320984, -0.316017746925354, 0.4253915250301361, -0.27867889404296875, -0.2547183632850647, -0.5003959536552429, 0.44879671931266785, -0.07695700973272324, -0.15314775705337524, 0.429432213306427, 0.41808271408081055, -0.04471581056714058, 0.023450668901205063, 0.21344710886478424, -0.19627436995506287, 0.11166626214981079, 0.44475460052490234, -0.3399270176887512, -0.30095478892326355, -0.09718866646289825, -0.05439300835132599, -0.12668243050575256, 0.1494974046945572, 0.12932434678077698, -0.4405778646469116, -0.37254294753074646, 0.08562347292900085, -0.2790433466434479, 0.06759664416313171, -0.16045573353767395, -0.2300846427679062, -0.30945369601249695, -0.6526288986206055, 0.38459211587905884, 0.30310294032096863, 0.16113793849945068, 0.0180328581482172, -0.1456814706325531, 0.06527063995599747, -0.3059377670288086, -0.21463371813297272, 0.14000774919986725, -0.09702987968921661, 0.014985751360654831, -0.17341473698616028, -0.14731064438819885, 0.1327158510684967, 0.07229085266590118, -0.22374700009822845, 0.036948010325431824, -0.2066386193037033, 0.5542380809783936, -0.22024859488010406, -0.057372454553842545, 0.20829910039901733, -0.3871699273586273, 0.0955580472946167, 0.07478738576173782, 0.34375059604644775, 0.010053078643977642, -0.019843056797981262, 0.2923174500465393, -0.19841399788856506, 0.06222645193338394, 0.19615569710731506, 0.20381462574005127, -0.19679836928844452, 0.43999484181404114, 0.38451290130615234, -0.09716885536909103, -0.32262328267097473, 0.5589426755905151, 0.48946595191955566, -0.7103240489959717, -0.05827130004763603, -0.08733179420232773, 0.17481128871440887, 0.0075123547576367855, 0.07378527522087097, -0.04773649573326111, -0.024275466799736023, -0.041590385138988495, -0.45077306032180786, -0.3029094338417053, 0.02130497619509697, -0.0532931424677372, 0.0931830033659935, 0.31392353773117065, 0.018152650445699692, 0.0753161758184433, 0.04191654548048973, -0.20519816875457764, 0.15909433364868164, 0.6357861757278442, 0.17778651416301727, 0.027765974402427673, 0.04229686036705971, -0.056716982275247574, -0.0001512644812464714, -0.004828028380870819, 0.13855107128620148, 0.0704178512096405, -0.174381822347641, -0.03296966105699539, 0.2171880006790161, -0.09015798568725586, -0.3297184407711029, -0.33929261565208435, -0.44494307041168213, -0.15494011342525482, -0.03558608517050743, -0.13536983728408813, -0.08628912270069122, -0.06770709156990051, 0.16228163242340088, 0.17669163644313812, -0.09132010489702225, -0.5623161196708679, 0.1294589340686798, -0.10235713422298431, 0.21732443571090698, -0.13209490478038788, 0.16996757686138153, 0.09857195615768433, -0.05798983946442604, 0.04792109131813049, 0.009129434823989868, 0.3277367353439331, 0.16313984990119934, 0.35374221205711365, -0.17176230251789093, 0.04795057699084282, -0.2062690705060959, 0.3029593527317047, 0.14168061316013336, -0.008553537540137768, 0.07447546720504761, 0.2950400114059448, 0.07731816917657852, -0.20862732827663422, 0.2393229603767395, -0.21360789239406586, -0.22185061872005463, -0.2649485170841217, 0.059707723557949066, 0.004555646330118179, 0.026559151709079742, -0.4137255549430847, -0.2439277172088623, 0.04004896059632301, 0.009618261829018593, -0.2730497121810913, 0.3426245152950287, 0.1418381780385971, -0.0367220900952816, -0.012312350794672966, 0.22880738973617554, 0.16126631200313568, -0.057614587247371674, -0.10032588988542557, 0.1408086121082306, 0.7164052724838257, 0.25883302092552185, 0.22971674799919128, -0.14957675337791443, 0.10609418153762817, 0.3407496213912964, 0.17358361184597015, 0.06045181676745415, 0.30843037366867065, 0.501695990562439, 0.49053341150283813, -0.06533706933259964, 0.10396670550107956, 0.2332955002784729, 0.02620002068579197, 0.09245093166828156, -0.2975195050239563, -0.09762099385261536, -0.07974141836166382, -0.2705274224281311, -0.26703470945358276, -0.0659562349319458, -0.12805064022541046, 0.37830907106399536, 0.22421573102474213, 0.006519787944853306, 0.09486521780490875, 0.0762881189584732, 0.30902552604675293, -0.19678980112075806, -0.08400505036115646, -0.047924816608428955, 0.06425436586141586, 0.014819752424955368, 0.44661566615104675, 0.14851871132850647, 0.1166367381811142, -0.18219128251075745, 0.21957474946975708, -0.032384149730205536, -0.13474971055984497, 0.25331494212150574, 0.24931135773658752, 0.43743470311164856, 0.024659302085638046, 0.16236242651939392, 0.0377783477306366, 0.06002858281135559, 0.010893892496824265, 0.19675016403198242, -0.1606857180595398, -0.10829699039459229, -0.048386819660663605, -0.07437437772750854, -0.45578721165657043, -0.21377207338809967, 0.07236243784427643, -0.1384688913822174, 0.20172850787639618, 0.6342796087265015, -0.3023315668106079, -0.0542244054377079, -0.2243732064962387, 0.012459781020879745, -0.11517803370952606, 0.42381569743156433, 0.3466522991657257, 0.143425852060318, -0.2957482635974884, -0.2863423228263855, -0.06824139505624771, 0.11662796139717102, -0.023076195269823074, -0.04253184050321579, -0.09624232351779938, -0.031166289001703262, 0.2795169949531555, 0.5127824544906616, 0.13443022966384888, 0.12040817737579346, -0.17573872208595276, 0.48935070633888245, -0.23794102668762207, -0.019993234425783157, 0.4501163959503174, 0.5878046154975891, -0.15019407868385315, -0.1882663369178772, 0.3224354684352875, 0.615074634552002, -0.06042851135134697, -0.1997968554496765, 0.00783056765794754, 0.33421847224235535, -0.17057214677333832, 0.20960336923599243, -0.01346089132130146, 0.09708046913146973, 0.02499491721391678, -0.09996960312128067, 0.14133583009243011, 0.032871756702661514, 0.12586301565170288, 0.5581203699111938, 0.25094497203826904, -0.17178665101528168, -0.20902101695537567, -0.06832177937030792, 0.015021784231066704, 0.2349604070186615, -0.33540862798690796, -0.0859619528055191, -0.31260308623313904, -0.09726447612047195, -0.2712610065937042, 0.04189324751496315, 0.01688387617468834, 0.3037467896938324, -0.17290370166301727, 0.012211451306939125, -0.20690622925758362, 0.023852596059441566, 0.11828340590000153, -0.2757697105407715, 0.030768264085054398, 0.364524245262146, 0.10293464362621307, 0.6225420236587524, 0.15047995746135712, -0.4848259389400482, 0.14814814925193787, 0.3166601061820984, -0.022683419287204742, 0.26371610164642334, 0.09053975343704224, 0.05713647976517677, 0.13081464171409607, 0.010113336145877838, 0.5547727942466736, -0.056815773248672485, 0.00499417819082737, -0.5158873200416565, -0.4087069034576416 ]
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)
[ -0.11623531579971313, -0.18411743640899658, -0.07161258906126022, 0.17512014508247375, -0.3280264139175415, 0.13923606276512146, 0.11795933544635773, 0.38174307346343994, 0.0022522900253534317, -0.27272242307662964, 0.056914299726486206, 0.09839931130409241, -0.07723778486251831, 0.2192150205373764, 0.3109802007675171, -0.05802804231643677, -0.10293128341436386, 0.18729205429553986, -0.2899806499481201, 0.03419706970453262, 0.24861109256744385, 0.02355029247701168, -0.06273826211690903, 0.014808647334575653, -0.18604975938796997, -0.21685835719108582, -0.12376022338867188, 0.10292959213256836, -0.11955898255109787, -0.40562549233436584, 0.04538264870643616, -0.02501906082034111, 0.5199143886566162, 0.17655107378959656, -0.00012255784531589597, -0.3012934923171997, 0.14918234944343567, -0.008922338485717773, -0.2559577226638794, -0.3521430492401123, -0.09115904569625854, -0.05141822248697281, 0.3022270202636719, -0.27631571888923645, -0.13665299117565155, -0.36661404371261597, 0.044753964990377426, -0.4632765054702759, 0.894417405128479, 0.38236698508262634, 0.06688161194324493, -0.06973031163215637, 0.12017402052879333, 0.17469841241836548, 0.29358938336372375, 0.6498956084251404, -0.29851973056793213, 0.01733425073325634, 0.22698600590229034, -0.0276035163551569, -0.18706870079040527, 0.21038703620433807, 0.03350234031677246, -0.16015806794166565, 0.41826844215393066, 0.03294292464852333, -0.3197293281555176, -0.26673150062561035, 0.10459907352924347, 0.29186373949050903, 0.44357216358184814, -0.5081267356872559, -0.06863570958375931, -0.1588287353515625, 0.19593681395053864, -0.40478408336639404, -0.20237398147583008, 0.5912532806396484, -0.45607396960258484, 0.4850004315376282, 0.11115680634975433, -0.3301267623901367, -0.4546760320663452, -0.023367010056972504, 0.41302016377449036, -0.11923055350780487, -0.1424250602722168, -0.03422095626592636, -0.08781935274600983, 0.11395935714244843, 0.14466159045696259, -0.34028249979019165, -0.08196960389614105, 0.2702951431274414, -0.030417662113904953, -0.14818450808525085, -0.5688720941543579, -0.11655572801828384, 0.2962394058704376, 0.19481061398983002, 0.3581240773200989, -0.14922824501991272, -0.08302667737007141, -0.24126338958740234, 0.11707843840122223, -0.035984721034765244, 0.4621579349040985, 0.12294962257146835, -0.022878237068653107, -0.10338762402534485, 0.21610522270202637, -0.11804764717817307, -0.07146736234426498, -0.007160442881286144, -0.19580762088298798, 0.1252121478319168, 0.06821299344301224, -0.35838767886161804, -0.06369633227586746, -0.06358198821544647, -0.3160603940486908, -0.06485778838396072, -0.16414423286914825, 0.12130085378885269, 0.15002983808517456, -0.1485426127910614, 0.20260143280029297, 0.4209895133972168, 0.002455327659845352, -0.23466061055660248, 0.1313377469778061, -0.01091739907860756, -0.2892898917198181, 0.3412545621395111, 0.15145495533943176, 0.04102557897567749, -0.033676959574222565, -0.06513373553752899, 0.041875869035720825, -0.18664006888866425, 0.2114461213350296, 0.0035024136304855347, 0.25545647740364075, -0.027918193489313126, 0.22552108764648438, 0.05125468969345093, -0.08152379095554352, -0.015555758029222488, -0.1415528506040573, 0.3194226622581482, -0.19716235995292664, -0.0013497546315193176, -0.01955215446650982, -0.012566233985126019, -0.07126863300800323, -0.1170920580625534, -0.29664430022239685, 0.16671687364578247, -0.05532965809106827, 0.3410831391811371, 0.05154557526111603, 0.30411356687545776, -0.004431478679180145, -0.15689964592456818, 0.06064585596323013, 0.5751582980155945, -0.27983880043029785, 0.0774809867143631, -0.06367764621973038, 0.1921238899230957, -0.13203993439674377, 0.32104748487472534, -0.2403281182050705, 0.20451784133911133, -0.10772127658128738, 0.21508730947971344, 0.6738172769546509, -0.111568883061409, 0.16387313604354858, 0.5188359022140503, 0.056450217962265015, -0.2370753288269043, 0.11921419203281403, -0.11641399562358856, -0.07434406876564026, -0.0028064223006367683, -0.3140353560447693, 0.5571510791778564, 0.08260932564735413, 0.01825597696006298, -0.12995746731758118, -0.21523694694042206, -0.09700989723205566, 0.00468393974006176, -0.23405607044696808, -0.014258526265621185, 0.060363639146089554, -0.5500681400299072, 0.4852308928966522, -0.32607460021972656, 0.400387167930603, 0.23203349113464355, 0.0025297626852989197, 0.02298756316304207, 0.1399737000465393, -0.031166348606348038, -0.26830238103866577, -0.25660666823387146, -0.31450095772743225, 0.2239762395620346, -0.45106619596481323, -0.24881407618522644, -0.3250754177570343, -0.13213957846164703, -0.10724616050720215, -0.021121937781572342, -0.06251736730337143, -0.2079959213733673, -0.10984879732131958, 0.053021226078271866, -0.21073785424232483, 0.36422213912010193, -0.05453042685985565, 0.24731077253818512, -0.369183212518692, 0.1250729113817215, 0.131422221660614, -0.010291158221662045, 0.23380392789840698, 0.17634448409080505, -0.1896401047706604, 0.03401071950793266, 0.007023070938885212, 0.08414427936077118, 0.010062401182949543, 0.2041340172290802, 0.4599419832229614, -0.2706068158149719, 0.12420658022165298, -0.7130149602890015, 0.2794916331768036, 0.007425929419696331, 0.1846461296081543, -0.09204725921154022, -0.4825463593006134, 0.5747973918914795, -0.14493034780025482, 0.23111018538475037, -0.1459645926952362, 0.027441812679171562, -0.13340207934379578, -0.07772533595561981, -0.22561107575893402, 0.06281210482120514, -0.032138075679540634, 0.32594117522239685, 0.3570249378681183, -0.12372724711894989, 0.017253518104553223, 0.03389055281877518, -0.22877836227416992, -0.2201501429080963, 0.15034902095794678, 0.26336804032325745, -0.24480053782463074, -0.07236368954181671, -0.03271276503801346, -0.008668195456266403, 0.114156574010849, 0.1704612672328949, -0.025074291974306107, 0.12046820670366287, 0.2024989277124405, -0.14061719179153442, 0.3045223653316498, -0.10037079453468323, 0.4107745289802551, -0.17753532528877258, -0.13777637481689453, -0.01287070196121931, -0.12090450525283813, -0.23075637221336365, -0.16143406927585602, 0.10699710994958878, -0.17705467343330383, -0.008562078699469566, -0.16584168374538422, -0.49550169706344604, -0.41691046953201294, 0.23199589550495148, -0.06793010234832764, 0.2515646815299988, -0.16291965544223785, -0.33356720209121704, 0.12842419743537903, -0.1341254860162735, -0.10172967612743378, 0.473461389541626, -0.0070494599640369415, 0.01629512384533882, -0.2771504819393158, -0.38154298067092896, -0.16488194465637207, 0.06597517430782318, 0.3686506152153015, 0.41833606362342834, 0.3976905047893524, -0.12005145847797394, 0.13771796226501465, -0.19119052588939667, 0.014920510351657867, 0.048584356904029846, -0.0908251702785492, -0.21377716958522797, -0.1240314245223999, 0.13249337673187256, -0.5243587493896484, 0.0738309845328331, -0.11730518937110901, 0.08368150889873505, -0.23695914447307587, -0.287192702293396, -0.09819870442152023, 0.013112182728946209, 0.1493162214756012, -0.3908492922782898, -0.1651652455329895, -0.10185348987579346, 0.5026170015335083, 0.25109153985977173, 0.19040639698505402, 0.04375794902443886, 0.10605809837579727, 0.1319570392370224, -0.2570900022983551, -0.10649888217449188, 0.18408675491809845, -0.20155015587806702, 0.2977464199066162, -0.39212673902511597, -0.5845625400543213, -0.11088474094867706, -0.006884768605232239, 0.03418830409646034, 0.28568094968795776, -0.4618057608604431, 0.41268226504325867, -0.07110361754894257, 0.07117271423339844, -0.11788609623908997, -0.01039588637650013, 0.20418310165405273, -0.025251850485801697, 0.0849703848361969, -0.020421545952558517, -0.017280973494052887, 0.298691987991333, -0.20029914379119873, 0.2878692150115967, 0.4254693388938904, 0.1723223626613617, 0.058060433715581894, 0.5829074382781982, 0.3013733923435211, 0.1089276522397995, 0.2840297222137451, -0.08005814254283905, 0.05360933393239975, -0.02631373703479767, 0.013504425063729286, -0.2094152867794037, 0.12492375075817108, 0.14099834859371185, 0.10728881508111954, 0.30725088715553284, -0.03692931681871414, -0.3321259319782257, -0.28665441274642944, -0.41476643085479736, -0.21308201551437378, 0.30389174818992615, -0.38708120584487915, 0.23486050963401794, -0.18260973691940308, -0.18359503149986267, -0.07116365432739258, -0.31488892436027527, 0.05090334266424179, 0.22340431809425354, 0.3517563045024872, 0.2050117701292038, -0.2665773034095764, 0.42818528413772583, -0.6509031653404236, 0.36899155378341675, 0.05844457447528839, -0.1497977375984192, 0.050097446888685226, -0.004401609301567078, 0.17925038933753967, 0.07588078081607819, -0.10500721633434296, 0.169464573264122, -0.018786177039146423, -0.022987715899944305, 0.1463407278060913, -0.5188398957252502, 0.39623939990997314, -0.1713441163301468, 0.02779264748096466, 0.012690221890807152, 0.19453568756580353, -0.43167951703071594, -0.3492641746997833, 0.11258301883935928, 0.2232145220041275, -0.12154300510883331, -0.3311595916748047, 0.21770144999027252, -0.06027301028370857, -0.24204589426517487, -0.0690440908074379, 0.3726750910282135, 0.03390030562877655, -0.047613754868507385, -0.13076594471931458, -0.04701988399028778, 0.12143319100141525, 0.11338669061660767, -0.26113712787628174, 0.13783738017082214, -0.2840893566608429, 0.09319926798343658, 0.14680469036102295, 0.42589765787124634, 0.07082612812519073, 0.21173246204853058, 0.13413259387016296, -0.340861976146698, -0.12266609817743301, -0.22335781157016754, 0.6217302083969116, 0.2724975347518921, 0.15294653177261353, 0.09791561961174011, 0.2817009389400482, -0.2173740416765213, -0.17630131542682648, -0.09220097959041595, 0.029144084081053734, -0.06997953355312347, -0.421385794878006, -0.37580588459968567, 0.5728219747543335, 0.03647741302847862, -0.021538957953453064, 0.15765061974525452, 0.14217323064804077, -0.2886929512023926, 0.5798212289810181, 0.24263042211532593, 1.1859325170516968, -0.2814680337905884, 0.31370726227760315, -0.03596274554729462, -0.360617458820343, 0.5188565254211426, -0.201727956533432, -0.27785593271255493, -0.4951661229133606, 0.3837963938713074, -0.04490648955106735, -0.18553081154823303, 0.3748287558555603, 0.3764258027076721, -0.030155284330248833, 0.1241893395781517, 0.16666489839553833, -0.12160615622997284, 0.05953522026538849, 0.47545504570007324, -0.312310129404068, -0.34697553515434265, -0.15084673464298248, -0.07249189913272858, -0.1424747258424759, 0.2328319400548935, 0.12784799933433533, -0.45531749725341797, -0.36892133951187134, 0.06390170753002167, -0.25991761684417725, 0.1060086116194725, -0.21665330231189728, -0.24640588462352753, -0.2637867033481598, -0.6389373540878296, 0.41278719902038574, 0.36399346590042114, 0.220606729388237, -0.04784080758690834, -0.2109031081199646, 0.08878776431083679, -0.3631051182746887, -0.18496039509773254, 0.12916597723960876, -0.14812685549259186, 0.03022698499262333, -0.1457628458738327, -0.11702010035514832, 0.0354306623339653, 0.11575955152511597, -0.17766138911247253, -0.019819527864456177, -0.1916959434747696, 0.45271193981170654, -0.26642051339149475, -0.10804013907909393, 0.19887083768844604, -0.2783833146095276, 0.08723224699497223, 0.03405511751770973, 0.31136322021484375, 0.049848657101392746, -0.0752512738108635, 0.2784467041492462, -0.22621800005435944, 0.027817463502287865, 0.228594571352005, 0.16525056958198547, -0.264872282743454, 0.4430118799209595, 0.425874799489975, -0.09706966578960419, -0.254608690738678, 0.4642239212989807, 0.581470251083374, -0.8045464754104614, -0.02388336881995201, -0.05021890625357628, 0.1676592379808426, 0.02494833618402481, 0.11201559752225876, -0.013125497847795486, -0.09906893223524094, 0.03518950566649437, -0.4654947817325592, -0.3229876458644867, 0.0847470685839653, -0.05763581767678261, 0.10016071796417236, 0.33518677949905396, 0.10849267244338989, 0.05705278366804123, 0.0697697326540947, -0.19205892086029053, 0.1754726767539978, 0.6514856219291687, 0.16727936267852783, 0.12899954617023468, 0.044357120990753174, 0.05866865813732147, -0.010351757518947124, -0.029424188658595085, 0.15425318479537964, 0.033484265208244324, -0.14044460654258728, -0.05531327426433563, 0.22434090077877045, -0.11151164025068283, -0.3583466410636902, -0.30706509947776794, -0.3529408574104309, -0.0504787378013134, -0.14498578011989594, -0.0487508662045002, -0.06242793798446655, -0.048514626920223236, 0.11967781186103821, 0.09116675704717636, -0.06606367975473404, -0.5188183188438416, 0.16587643325328827, -0.12332658469676971, 0.21302790939807892, -0.027788789942860603, 0.16786661744117737, -0.014582027681171894, -0.0924047976732254, 0.06336070597171783, 0.019591279327869415, 0.347075879573822, 0.2004651129245758, 0.3311263918876648, -0.11110356450080872, -0.019029829651117325, -0.20461252331733704, 0.34920892119407654, 0.1420440375804901, 0.04396386072039604, 0.04862738400697708, 0.3234454393386841, 0.06121771037578583, -0.1930483877658844, 0.2455327957868576, -0.1727200746536255, -0.18861542642116547, -0.2519422173500061, 0.0988527461886406, 0.03157603368163109, 0.10778237879276276, -0.4496346116065979, -0.23619261384010315, 0.17203710973262787, 0.06814950704574585, -0.258975625038147, 0.2703651189804077, 0.14384567737579346, -0.03995649889111519, -0.007922982797026634, 0.27942895889282227, 0.22925084829330444, -0.08278056979179382, -0.1410936713218689, 0.1737515926361084, 0.6221064329147339, 0.35927021503448486, 0.29032212495803833, -0.23674072325229645, 0.06427210569381714, 0.33668479323387146, 0.18424676358699799, 0.03932691365480423, 0.2048325389623642, 0.6484082937240601, 0.5068117380142212, -0.07449229806661606, 0.11960945278406143, 0.29356828331947327, 0.04900941997766495, 0.03089677169919014, -0.2590872049331665, -0.08743606507778168, -0.12445281445980072, -0.19977828860282898, -0.26839759945869446, -0.1559889018535614, -0.07701228559017181, 0.32390719652175903, 0.19428834319114685, -0.028334075585007668, 0.028100455179810524, 0.09170132130384445, 0.30897948145866394, -0.1727255880832672, 0.022006860002875328, 0.0015087723731994629, 0.01093374565243721, 0.040897201746702194, 0.43185803294181824, 0.2473566234111786, 0.09488613903522491, -0.12778662145137787, 0.2288220226764679, 0.026229051873087883, -0.09294761717319489, 0.2152305394411087, 0.19966132938861847, 0.3968440890312195, 0.03586225211620331, 0.18723459541797638, 0.03754754364490509, 0.06753221154212952, -0.055728521198034286, 0.18198063969612122, -0.15538348257541656, -0.07866621017456055, -0.150333434343338, -0.051529377698898315, -0.3914935290813446, -0.19860784709453583, 0.06892629712820053, -0.17805668711662292, 0.2038511484861374, 0.5880892276763916, -0.2470196932554245, -0.04993647709488869, -0.16663672029972076, 0.00043440423905849457, -0.03191450983285904, 0.45689618587493896, 0.3662309944629669, 0.14163066446781158, -0.30244672298431396, -0.32491812109947205, -0.13995331525802612, 0.09046223759651184, -0.014410216361284256, -0.09643401205539703, -0.05399131029844284, -0.05557709187269211, 0.22304940223693848, 0.45404696464538574, 0.02330099046230316, 0.1714804321527481, -0.1257140189409256, 0.4537195563316345, -0.22176793217658997, -0.010113108903169632, 0.37255823612213135, 0.5218807458877563, -0.10128220170736313, -0.13284358382225037, 0.25423502922058105, 0.5032428503036499, -0.09317536652088165, -0.1573285311460495, 0.05690664052963257, 0.23135694861412048, -0.25062325596809387, 0.20016826689243317, 0.035372328013181686, 0.13309989869594574, -0.007061762735247612, -0.15031678974628448, 0.09966996312141418, -0.02799461968243122, 0.05913616344332695, 0.6108744144439697, 0.24947255849838257, -0.10200604051351547, -0.21705719828605652, -0.11341233551502228, -0.05578731745481491, 0.2859748899936676, -0.2583238482475281, -0.12872783839702606, -0.3091915249824524, -0.062091320753097534, -0.24192306399345398, 0.025370411574840546, 0.0015793079510331154, 0.40263399481773376, -0.155255526304245, 0.06434876471757889, -0.16876152157783508, -0.04738273471593857, 0.2202804833650589, -0.31012842059135437, 0.08213500678539276, 0.32826653122901917, 0.12444241344928741, 0.6768010258674622, 0.10875971615314484, -0.5213270783424377, 0.1987709403038025, 0.2741400897502899, 0.004109051078557968, 0.219305157661438, 0.09368613362312317, 0.053228069096803665, 0.07330302894115448, -0.03502937778830528, 0.4586232900619507, -0.029533907771110535, -0.029490791261196136, -0.42908352613449097, -0.4309493899345398 ]
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"
[ 0.189832866191864, -0.4081777036190033, -0.02667148783802986, -0.07893197238445282, 0.17554613947868347, -0.11440110206604004, 0.2933264672756195, 0.23464734852313995, -0.11313799023628235, -0.006701231002807617, -0.07507719844579697, 0.5407452583312988, -0.25638607144355774, 0.21988382935523987, 0.0012711333110928535, -0.44989344477653503, 0.16272661089897156, 0.13243573904037476, -0.12136062979698181, -0.11142566800117493, -0.2248261719942093, -0.18424540758132935, -0.2530255913734436, -0.1031172126531601, -0.15764784812927246, 0.06863262504339218, -0.3666492998600006, -0.21603578329086304, -0.37540403008461, -0.333639919757843, -0.1936650276184082, 0.2823522686958313, -0.21831154823303223, -0.03667692095041275, -0.00012514082482084632, -0.4015752077102661, -0.1773592233657837, -0.17478589713573456, -0.11023940145969391, -0.11791358888149261, -0.8842170238494873, -0.007036093156784773, 0.38440045714378357, 0.018586695194244385, 0.007791262120008469, 0.026044782251119614, 0.026937825605273247, -0.3077804744243622, 0.4810143709182739, -0.008772121742367744, 0.07200632989406586, -0.037221767008304596, 0.03685818985104561, 0.0641988143324852, 0.1308983862400055, 0.5086036324501038, 0.01867283694446087, -0.04318332299590111, 0.6300451755523682, 0.003555065020918846, -0.20570045709609985, 0.3613415062427521, 0.09690244495868683, 0.10612984001636505, 0.34404945373535156, -0.02620140090584755, -0.05275914818048477, -0.08939734101295471, -0.2308887541294098, 0.3512212038040161, 0.7839898467063904, -0.0999956950545311, -0.26931700110435486, -0.037071987986564636, 0.02138654515147209, 0.06871125102043152, -0.1864115297794342, 0.01460917480289936, -0.1401163786649704, -0.06115397438406944, -0.48590776324272156, -0.3253767192363739, -0.3561829626560211, 0.06938893347978592, -0.2764469087123871, 0.4265614151954651, 0.008110159076750278, -0.035178884863853455, -0.05221099033951759, -0.2725643515586853, -0.021641220897436142, -0.4392715394496918, 0.39041703939437866, 0.1420876681804657, 0.1063588410615921, -0.37915337085723877, -0.1874583661556244, 0.07733557373285294, 0.37103471159935, -0.2946627140045166, -0.0931793674826622, -0.0026956871151924133, -0.5204649567604065, -0.0124722421169281, 0.5711627006530762, 0.0022489195689558983, -0.04572527855634689, 0.07216678559780121, -0.3457534909248352, -0.29139602184295654, -0.2567962408065796, 0.04164757579565048, 0.12521547079086304, 0.2284185141324997, 0.07164223492145538, 0.1218288317322731, 0.05681435763835907, -0.321562260389328, 0.06382282078266144, 0.07790405303239822, -0.3623915910720825, -0.20421423017978668, 0.03367452695965767, 0.154516339302063, -0.08497203886508942, 0.5170227289199829, -0.1964961290359497, 0.20776639878749847, 0.08788135647773743, -0.35689577460289, 0.1389586180448532, -0.10198041796684265, -0.5272906422615051, -0.2540569603443146, 0.19491790235042572, 0.2871021330356598, -0.010821458883583546, 0.36924538016319275, 0.37839585542678833, -0.2300274521112442, 0.07069475948810577, 0.0249626524746418, 0.5642063617706299, -0.18388137221336365, -0.35279953479766846, 0.3267550468444824, 0.29061374068260193, 0.3297920227050781, -0.08868296444416046, 0.5011613368988037, -0.21087926626205444, -0.014086801558732986, -0.153348907828331, 0.017584221437573433, 0.21666352450847626, -0.21159717440605164, 0.04111845791339874, 0.3512011468410492, 0.045034557580947876, -0.3355676531791687, 0.13014280796051025, -0.5346509218215942, 0.04989193007349968, -0.2956683039665222, 0.002575365826487541, 0.4049500823020935, -0.46270552277565, 0.09012217819690704, -0.19625224173069, 0.23597614467144012, 0.2237938642501831, 0.26253271102905273, -0.07536926865577698, -0.059609390795230865, -0.011268031783401966, 0.356981098651886, 0.31552308797836304, -0.16380952298641205, -0.31837132573127747, 0.22978602349758148, 0.16081301867961884, 0.11740314215421677, -0.036376602947711945, 0.23446999490261078, 0.1547362357378006, 0.2690247595310211, 0.05115576088428497, -0.1387476921081543, -0.30224430561065674, 0.36625227332115173, -0.018001213669776917, -0.04290368780493736, 0.12097165733575821, -0.04484010487794876, -0.07994233071804047, -0.14226484298706055, -0.08018431812524796, 0.11475404351949692, 0.21867242455482483, -0.20548701286315918, -0.05747019499540329, -0.42732083797454834, 0.15273521840572357, 0.029008686542510986, -0.12003201246261597, -0.17264282703399658, -0.6358430981636047, -0.07989707589149475, -0.028847303241491318, -0.1870037466287613, 0.22696946561336517, -0.2735062837600708, -0.2580797076225281, 0.058287858963012695, -0.33636143803596497, 0.10470384359359741, -0.013620542362332344, 0.020014546811580658, -0.2457858920097351, -0.29985085129737854, -0.27718693017959595, -0.06258738040924072, 0.5232810378074646, 0.018669795244932175, 0.05305280163884163, 0.11633943021297455, 0.09669698029756546, 0.0029886425472795963, 0.13223706185817719, 0.3557763993740082, 0.19228729605674744, 0.2523917853832245, -0.03136170655488968, 0.2370242476463318, 0.45958828926086426, -0.048325315117836, -0.03865766525268555, 0.11578787118196487, 0.4056984782218933, -0.23331999778747559, 0.32311785221099854, 0.17757375538349152, 0.09546326100826263, -0.013860311359167099, -0.8099335432052612, 0.189357727766037, -0.07852838933467865, 0.11697982251644135, -0.14940418303012848, 0.2087828814983368, -0.1929720640182495, 0.2378268837928772, -0.06035556644201279, -0.4787614345550537, -0.12163159251213074, 0.26565828919410706, 0.1503821611404419, 0.07501419633626938, -0.4771554172039032, 0.2781411111354828, 0.37942156195640564, -0.017873769626021385, 0.2281641960144043, -0.06022486835718155, -0.01726694405078888, -0.09874344617128372, 0.41970616579055786, -0.21944649517536163, 0.23195873200893402, 0.1562119424343109, -0.06431920826435089, -0.14587201178073883, 0.008576968684792519, -0.12822508811950684, -0.002887815237045288, -0.03820955008268356, -0.10356047749519348, 0.10288208723068237, 0.1906561404466629, -0.10545209050178528, 0.0022766776382923126, 0.07329021394252777, -0.05676286667585373, -0.09499042481184006, -0.12194760143756866, 0.005622925236821175, -0.5228880047798157, -0.01635923981666565, -0.15502753853797913, -0.5842819213867188, -0.19491051137447357, -0.18189498782157898, 0.20081683993339539, -0.1969064623117447, -0.08773282170295715, 0.006360802333801985, 0.0872836709022522, 0.3509705066680908, 0.13916301727294922, 0.0905078649520874, 0.07922898232936859, -0.0174418892711401, 0.06981420516967773, 0.04823223873972893, 0.004030964337289333, 0.03277851268649101, 0.19473370909690857, 0.10559012740850449, 0.011595755815505981, -0.35372409224510193, 0.015375129878520966, 0.19428357481956482, -0.04946019500494003, 0.010338237509131432, 0.33580487966537476, 0.08736909180879593, -0.495445191860199, 0.13991816341876984, 0.14058928191661835, 0.3530597984790802, -0.05232655256986618, 0.030834835022687912, 0.3473846912384033, 0.017780210822820663, 0.09395411610603333, -0.3488008379936218, -0.1854475885629654, 0.04607253149151802, 0.09167945384979248, -0.024083953350782394, 0.10844962298870087, -0.23966284096240997, 0.17827773094177246, -0.1272117793560028, -0.34878531098365784, -0.247926265001297, -0.20571362972259521, 0.11752523481845856, 0.048901788890361786, -0.14117513597011566, -0.01738835498690605, -0.2172677367925644, 0.14096347987651825, -0.13892053067684174, 0.18003641068935394, 0.27049028873443604, -0.27786785364151, -0.04547499865293503, 0.12244537472724915, 0.0982365608215332, 0.4216066300868988, 0.3028852343559265, 0.3918447494506836, 0.08133647590875626, -0.07759658992290497, 0.10318781435489655, 0.11008711159229279, 0.3211497962474823, 0.009788133203983307, 0.14529214799404144, 0.04953548312187195, -0.1308119148015976, 0.7850809693336487, 0.47630518674850464, 0.14479035139083862, 0.1125241294503212, 0.002753874287009239, -0.16256743669509888, -0.13791394233703613, -0.10453397035598755, 0.2698311507701874, -0.07678432762622833, 0.19053786993026733, 0.45777904987335205, 0.08809159696102142, -0.06342166662216187, 0.03595363348722458, 0.2923191487789154, -0.23634710907936096, -0.2863630950450897, 0.42527344822883606, -0.17180877923965454, 0.25057095289230347, 0.014177825301885605, -0.11715613305568695, -0.2502304017543793, -0.12013731896877289, -0.06141114979982376, -0.11164245009422302, 0.34356555342674255, -0.39867067337036133, -0.09759130328893661, -0.256198525428772, -0.15704861283302307, 0.29891878366470337, 0.47733354568481445, 0.10383498668670654, -0.01350841298699379, -0.07692837715148926, 0.16502931714057922, 0.3424927592277527, 0.517250120639801, -0.03934948891401291, -0.11622586846351624, -0.09208141267299652, 0.05629953742027283, 0.17325051128864288, -0.03681115061044693, -0.6540710926055908, 0.32207465171813965, -0.2926149070262909, 0.1334902048110962, -0.22548875212669373, -0.35192811489105225, 0.17181181907653809, 0.14538055658340454, -0.10473287105560303, -0.22611120343208313, 0.13815724849700928, 0.5807726979255676, 0.0008972045034170151, 0.2521328330039978, 0.2287890464067459, -0.15927039086818695, 0.19195635616779327, -0.09325660765171051, 0.18506526947021484, -0.006664203479886055, 0.3489795923233032, 0.21433041989803314, 0.06443063169717789, 0.16485711932182312, -0.3918599486351013, 0.2294236570596695, 0.1500619798898697, 0.19573739171028137, 0.20301252603530884, -0.6772517561912537, -0.060861051082611084, 0.08496478945016861, -0.11696672439575195, 0.6384943127632141, 0.3110847771167755, 0.2727363705635071, -0.08074901252985, 0.09046341478824615, -0.014859922230243683, -0.18671002984046936, -0.07817530632019043, 0.3036341369152069, 0.08140749484300613, -0.2078104019165039, -0.416168749332428, 0.38397493958473206, -0.11780276894569397, -0.26804718375205994, -0.13686548173427582, 0.5329189300537109, 0.006374480202794075, 0.03252534568309784, 0.41089504957199097, 0.8881409764289856, -0.18266119062900543, 0.3090890645980835, 0.13335850834846497, 0.07515868544578552, 0.3711303472518921, -0.13379468023777008, 0.22359661757946014, -0.22098954021930695, 0.105465367436409, -0.18438202142715454, -0.19609878957271576, -0.000308893620967865, 0.5590388774871826, -0.693390965461731, 0.2738444209098816, -0.154904305934906, 0.15904928743839264, 0.14156168699264526, -0.0622471459209919, 0.3114883601665497, -0.1365651935338974, -0.1344507783651352, -0.012796031311154366, -0.2086646407842636, -0.13177862763404846, -0.15400856733322144, -0.04845869541168213, 0.03196627274155617, 0.005609109997749329, -0.3148159086704254, -0.06546637415885925, -0.5026320815086365, -0.22931347787380219, 0.0704682394862175, -0.13271108269691467, 0.49281880259513855, 0.4730241298675537, 0.11013013124465942, 0.12637798488140106, -0.08461261540651321, -0.18574948608875275, 0.1338845044374466, 0.28642597794532776, -0.23144900798797607, 0.10738244652748108, 0.5054354071617126, -0.028761200606822968, -0.30972820520401, 0.23615384101867676, -0.12689198553562164, -0.08921197056770325, -0.48502466082572937, 0.11372119188308716, 0.7107093930244446, -0.5381376147270203, -0.4216567277908325, -0.03158755600452423, 0.1221124529838562, 0.02823345735669136, -0.08895732462406158, 0.2638345956802368, 0.11011446267366409, 0.5779937505722046, 0.23609013855457306, 0.1575019657611847, 0.24454277753829956, -0.13318315148353577, 0.14268717169761658, -0.1016162782907486, 0.31638839840888977, 0.20700936019420624, -0.1430155485868454, -0.13768139481544495, 0.08183915168046951, 0.22919079661369324, -0.10338245332241058, 0.1376062035560608, -0.25277161598205566, -0.08193708956241608, -0.03888702765107155, 0.3396036922931671, 0.3437861204147339, 0.30378568172454834, -0.009304389357566833, 0.04974660277366638, -0.11958939582109451, 0.2682490944862366, -0.22673340141773224, 0.1715279519557953, 0.19719451665878296, -0.1525161862373352, 0.06678889691829681, 0.260731965303421, -0.17686200141906738, -0.09415048360824585, 0.11468591541051865, 0.3973570764064789, -0.0025256630033254623, -0.2013346254825592, -0.2250729501247406, -0.05165468156337738, -0.04072563350200653, -0.22952820360660553, -0.22013938426971436, -0.12144707143306732, -0.1096401959657669, 0.13896621763706207, 0.09552144259214401, -0.07645708322525024, -0.12439693510532379, -0.12791964411735535, -0.008751053363084793, -0.12493596971035004, -0.002041764557361603, -0.06993503123521805, -0.16730575263500214, 0.3306974470615387, 0.1385970115661621, -0.3417626917362213, -0.19411854445934296, 0.12415020167827606, 0.32734930515289307, -0.046312421560287476, 0.1761833131313324, 0.1268651783466339, 0.06882397085428238, -0.21992753446102142, -0.07351990044116974, 0.43982794880867004, 0.2794691324234009, 0.0344989113509655, 0.028155870735645294, 0.09866347163915634, 0.16357484459877014, 0.0910554826259613, 0.29285088181495667, 0.05899415537714958, 0.04810664430260658, 0.09166944772005081, 0.04486217349767685, 0.08739277720451355, -0.10868410021066666, -0.21807761490345, 0.5362933874130249, 0.10885204374790192, 0.16727371513843536, 0.35398173332214355, 0.058678146451711655, 0.017523393034934998, 0.37596338987350464, -0.028894171118736267, -0.35815781354904175, 0.0637386292219162, 0.01486562192440033, 0.6164768934249878, -0.09675905853509903, 0.06004120409488678, 0.1579797863960266, 0.21127848327159882, 0.06314335763454437, -0.00310448557138443, 0.07378453016281128, 0.21520675718784332, 0.119562529027462, -0.13529027998447418, 0.49863874912261963, -0.016309211030602455, 0.21602973341941833, 0.22185903787612915, 0.1069485992193222, 0.44822293519973755, 0.35138222575187683, 0.1383226215839386, -0.26116451621055603, -0.5070874691009521, -0.005821077153086662, 0.08295352756977081, -0.14927971363067627, -0.18763838708400726, -0.023021284490823746, -0.2203642725944519, -0.34096357226371765, -0.03082037903368473, -0.6939731240272522, -0.13618916273117065, 0.4685724377632141, 0.323300838470459, 0.503394603729248, -0.12030547857284546, -0.049626708030700684, 0.10586360841989517, 0.2920260429382324, -0.12352420389652252, 0.18553775548934937, 0.03374519944190979, 0.02798856794834137, 0.10734626650810242, 0.3419782519340515, -0.0009043291211128235, 0.21667702496051788, -0.18185119330883026, -0.0319611132144928, -0.3361881673336029, -0.08698029071092606, 0.2088313102722168, -0.23594453930854797, 0.22233769297599792, -0.0785495787858963, 0.1523190289735794, 0.011357923969626427, 0.051598161458969116, 0.15603037178516388, 0.06995577365159988, 0.006510185077786446, -0.2356921285390854, -0.5836465358734131, 0.2164982706308365, -0.25015977025032043, 0.08592202514410019, -0.031399063766002655, -0.6951836943626404, -0.0269295834004879, 0.056830644607543945, -0.1017041802406311, -0.08000939339399338, 0.14819777011871338, -0.012619556859135628, 0.26396527886390686, -0.02980627492070198, 0.36169663071632385, 0.014449577778577805, -0.3209570646286011, -0.17841556668281555, -0.2791803777217865, -0.11175758391618729, 0.2492118626832962, -0.12218450754880905, 0.07240916043519974, -0.02137771062552929, 0.05432358384132385, 0.3683212995529175, 0.05567166954278946, -0.5797141790390015, -0.25978291034698486, 0.473615437746048, -0.1560894399881363, 0.1644119769334793, 0.28888195753097534, 0.2590021789073944, -0.028212402015924454, -0.06761834025382996, 0.27827516198158264, 0.310769647359848, -0.151582270860672, -0.23049874603748322, 0.010738961398601532, -0.24945521354675293, -0.05991800129413605, -0.227616548538208, 0.36596599221229553, 0.2763817608356476, 0.10707373172044754, -0.20322701334953308, 0.28647246956825256, 0.40573084354400635, 0.11927688121795654, 0.09432676434516907, 0.11116792261600494, 0.6024980545043945, -0.1494073122739792, 0.2883363664150238, -0.05454804003238678, 0.023511825129389763, 0.09809628129005432, -0.1638512909412384, -0.6133045554161072, -0.02244396135210991, -0.023944184184074402, -0.3201178014278412, 0.138483926653862, 0.41740283370018005, -0.1201280951499939, 0.1861407607793808, -0.24419832229614258, 0.2617798149585724, 0.13630354404449463, -0.22383365035057068, -0.34966403245925903, -0.30819401144981384, -0.04592210054397583, -0.2481653094291687, -0.014777405187487602, -0.7189196348190308, -0.241082102060318, 0.2345632165670395, -0.11553273350000381, -0.14388951659202576, 0.04211653769016266, -0.039306651800870895, -0.16625675559043884, -0.3629511594772339, 0.6250200271606445, 0.036941591650247574, 0.10133044421672821, -0.3119325041770935, -0.45195573568344116 ]
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) ```
[ 0.1992310732603073, -0.575800895690918, 0.010819882154464722, -0.0699361115694046, 0.2859188914299011, -0.15366774797439575, 0.24460144340991974, 0.24167530238628387, -0.009871283546090126, 0.018018953502178192, -0.15490840375423431, 0.455800861120224, -0.2713589370250702, 0.0753546729683876, 0.12558721005916595, -0.37995249032974243, 0.15262764692306519, 0.06899324059486389, -0.20162361860275269, -0.06752023845911026, -0.11300843954086304, 0.02811495028436184, -0.16656044125556946, -0.02064725011587143, -0.3513280153274536, 0.04645802080631256, -0.36015385389328003, -0.0656294897198677, -0.30250850319862366, -0.3981378674507141, -0.1327248066663742, 0.19807399809360504, -0.25676780939102173, 0.11358477175235748, -0.00013058973127044737, -0.31119462847709656, -0.21208496391773224, -0.21448419988155365, -0.09534168243408203, -0.031543880701065063, -0.7208173274993896, -0.0117108765989542, 0.39389553666114807, 0.10779359191656113, -0.025061124935746193, -0.024407606571912766, 0.0035536643117666245, -0.1321898251771927, 0.6255216598510742, -0.02035348489880562, 0.009229175746440887, 0.04004726558923721, 0.08175075054168701, 0.09900170564651489, 0.1613849401473999, 0.5654871463775635, -0.07321750372648239, 0.08077193796634674, 0.569547712802887, -0.0895318016409874, -0.2318439483642578, 0.5064967274665833, 0.03164707124233246, 0.07505442947149277, 0.2941820025444031, 0.11394406110048294, -0.015605650842189789, -0.08051556348800659, -0.22278180718421936, 0.41991183161735535, 0.5713568329811096, -0.30700618028640747, -0.44495052099227905, -0.2811705470085144, 0.07755261659622192, -0.05330924317240715, -0.1852957308292389, 0.044255949556827545, -0.2463555932044983, 0.0007777474820613861, -0.3870481550693512, -0.14347663521766663, -0.24800752103328705, 0.03583591803908348, -0.2692349851131439, 0.511880099773407, -0.05074486881494522, 0.021876029670238495, -0.006453309208154678, -0.23134665191173553, -0.15191029012203217, -0.3797209858894348, 0.30629488825798035, 0.20800743997097015, -0.04424024000763893, -0.39680713415145874, -0.13806238770484924, -0.07668542116880417, 0.4508575201034546, -0.24334202706813812, -0.00857329647988081, 0.023109758272767067, -0.4934563934803009, -0.022741883993148804, 0.648693859577179, 0.04183975234627724, -0.011994697153568268, -0.0025547584518790245, -0.23035399615764618, -0.08692910522222519, -0.10430460423231125, 0.06293801963329315, 0.14950506389141083, 0.1489628553390503, 0.12328298389911652, 0.10730724781751633, 0.024612929672002792, -0.2623036205768585, -0.03937054052948952, 0.012350460514426231, -0.32854601740837097, -0.13844557106494904, 0.13970710337162018, 0.24745827913284302, -0.127646803855896, 0.503680944442749, -0.1274145245552063, 0.3086261749267578, -0.009251781739294529, -0.09967132657766342, 0.09143596142530441, -0.1582871973514557, -0.4034566283226013, -0.10639635473489761, 0.21177919209003448, 0.33019858598709106, 0.09043743461370468, 0.4027234613895416, 0.5290735960006714, -0.2610943913459778, 0.02161410264670849, 0.01453854888677597, 0.4841368496417999, 0.02228257805109024, -0.33286750316619873, 0.40923282504081726, 0.35748252272605896, 0.270408034324646, -0.1167823001742363, 0.37184426188468933, -0.17706051468849182, -0.03748030960559845, -0.016465483233332634, -0.05915633961558342, 0.1264021247625351, -0.16889415681362152, -0.15660089254379272, 0.36347857117652893, 0.12413916736841202, -0.1976221203804016, 0.08903414756059647, -0.5058838129043579, -0.01267833262681961, -0.18312636017799377, 0.15960529446601868, 0.5324588418006897, -0.2871987521648407, -0.14317461848258972, -0.023123497143387794, 0.14544807374477386, 0.0770607516169548, 0.36676448583602905, -0.08329935371875763, -0.09566760808229446, -0.028542842715978622, 0.4020654559135437, 0.20190644264221191, -0.2790754437446594, -0.2886229157447815, 0.18329249322414398, 0.32033413648605347, 0.30419206619262695, -0.12321639806032181, 0.13023139536380768, 0.020967885851860046, 0.22887706756591797, 0.10841445624828339, -0.16214489936828613, -0.17627912759780884, 0.4218747615814209, -0.05920543894171715, -0.13189488649368286, 0.07925218343734741, -0.036505427211523056, -0.10871562361717224, -0.2211264967918396, -0.21198472380638123, 0.1571851372718811, 0.3730960488319397, -0.13576793670654297, -0.15732991695404053, -0.19157469272613525, 0.14541824162006378, -0.08411453664302826, -0.03216903656721115, -0.18415400385856628, -0.6730773448944092, 0.03187338262796402, -0.058820322155952454, -0.013976752758026123, -0.023560382425785065, -0.23712123930454254, -0.3311190903186798, 0.15139460563659668, -0.35942086577415466, -0.06123196706175804, -0.06678400933742523, -0.011039458215236664, -0.09729538857936859, -0.21554264426231384, -0.23209570348262787, 0.05262405425310135, 0.48802900314331055, 0.11189083009958267, -0.011933045461773872, 0.19364546239376068, 0.19373174011707306, -0.12392780184745789, 0.06306765973567963, 0.3638549745082855, 0.2385554313659668, 0.18986479938030243, 0.015624390915036201, 0.21068793535232544, 0.533371090888977, -0.07160782814025879, -0.12715297937393188, 0.23659689724445343, 0.4539567232131958, -0.19625253975391388, 0.1511325240135193, 0.20681197941303253, 0.19266006350517273, -0.04783997684717178, -0.7790422439575195, 0.2532617449760437, 0.013713613152503967, 0.25994306802749634, -0.09325248003005981, 0.192877396941185, -0.07523898035287857, 0.3315355181694031, -0.059525735676288605, -0.39366692304611206, -0.20147694647312164, 0.12461215257644653, 0.14805004000663757, -0.029560385271906853, -0.4988958537578583, 0.3210471272468567, 0.37384963035583496, -0.052154675126075745, 0.25738486647605896, -0.057632751762866974, -0.10829807072877884, -0.15855783224105835, 0.3733244240283966, -0.35355350375175476, 0.28380057215690613, 0.06916701793670654, -0.08368012309074402, -0.12158340960741043, -0.04242310672998428, -0.03475860506296158, 0.027157500386238098, 0.06743291020393372, -0.07191870361566544, 0.1330578774213791, 0.1848296970129013, -0.22043871879577637, -0.08031675219535828, -0.005193222314119339, -0.08777441084384918, -0.0895499438047409, -0.20640704035758972, 0.09672456979751587, -0.5124544501304626, 0.1038118526339531, -0.2801769971847534, -0.554701030254364, -0.36573559045791626, -0.2290543168783188, 0.26497071981430054, -0.08991865813732147, -0.16554860770702362, 0.05860508233308792, 0.11672475934028625, 0.09931112080812454, 0.19559450447559357, -0.0060081202536821365, -0.08945012092590332, 0.04601384326815605, 0.10184025764465332, -0.05130700021982193, -0.04768429696559906, -0.04785206541419029, 0.037647977471351624, 0.12882810831069946, -0.11320019513368607, -0.4533537030220032, -0.008424080908298492, 0.3111928105354309, -0.11082722246646881, 0.1549980491399765, 0.3773556053638458, 0.1029159426689148, -0.48735517263412476, 0.048792775720357895, 0.212270125746727, 0.2985989451408386, -0.0036084335297346115, 0.11281442642211914, 0.30032792687416077, -0.022947629913687706, -0.012343760579824448, -0.19157731533050537, -0.13206130266189575, -0.003897925838828087, 0.10018597543239594, 0.026955600827932358, 0.15248650312423706, -0.16705459356307983, 0.26794713735580444, -0.03599873557686806, -0.33557772636413574, -0.2415345013141632, -0.14004871249198914, 0.08695998787879944, 0.14672958850860596, -0.08437187224626541, -0.05225079879164696, -0.2838633954524994, 0.04827546328306198, -0.018866952508687973, 0.11450206488370895, 0.18315814435482025, -0.35383525490760803, -0.0870286300778389, -0.06288972496986389, 0.1746329367160797, 0.48489096760749817, 0.28579121828079224, 0.2647237479686737, 0.11941230297088623, -0.056173358112573624, -0.00612759031355381, 0.1639111042022705, 0.34851157665252686, 0.04665456712245941, 0.004913097247481346, -0.09494320303201675, 0.0010067038238048553, 1.0838128328323364, 0.48438388109207153, 0.24819226562976837, -0.003678448498249054, -0.11817046254873276, -0.1263032853603363, -0.15671616792678833, -0.09003377705812454, 0.11810848861932755, -0.07367570698261261, 0.30270129442214966, 0.46514344215393066, 0.2442467212677002, 0.01206075306981802, 0.053465522825717926, 0.38583505153656006, -0.16218233108520508, -0.4523179829120636, 0.3965875506401062, -0.0745377242565155, 0.14584703743457794, 0.01808437705039978, 0.06413006037473679, -0.18320032954216003, -0.1223573312163353, -0.0762113481760025, 0.04760737344622612, 0.31014183163642883, -0.19234776496887207, -0.04028281196951866, -0.2107447385787964, -0.25974369049072266, 0.2814566195011139, 0.522388219833374, 0.1269291490316391, -0.0026252418756484985, -0.1497012972831726, 0.2083989381790161, 0.23716433346271515, 0.635626494884491, 0.178530752658844, -0.20981650054454803, -0.00905932392925024, -0.12181360274553299, -0.010507297702133656, -0.12758637964725494, -0.6342929005622864, 0.18389475345611572, -0.2725788950920105, 0.22634418308734894, -0.39750564098358154, -0.46626895666122437, -0.0781315416097641, 0.11758805066347122, -0.1353408247232437, -0.10770022124052048, -0.05650631710886955, 0.4158494174480438, -0.15804964303970337, 0.32521000504493713, 0.3609256148338318, -0.14264832437038422, 0.17127522826194763, -0.06450578570365906, 0.2282078117132187, -0.05428936332464218, 0.34357163310050964, 0.33272436261177063, 0.13303759694099426, 0.09885770082473755, -0.36606472730636597, 0.22141101956367493, 0.08414459228515625, -0.009226266294717789, 0.3217577636241913, -0.5721666812896729, -0.29042747616767883, 0.14899960160255432, -0.04958136007189751, 0.4967501759529114, 0.44231799244880676, 0.21683450043201447, -0.05097954720258713, 0.2642033100128174, 0.07970675826072693, -0.13998578488826752, 0.1115492433309555, 0.2770272195339203, 0.08728804439306259, -0.1706182211637497, -0.3983020782470703, 0.4744041860103607, -0.06491909176111221, -0.28020185232162476, 0.013689447194337845, 0.5781557559967041, -0.08784236013889313, -0.0790267363190651, 0.2935151755809784, 0.9808769822120667, -0.23133574426174164, 0.39982396364212036, 0.40193745493888855, 0.026256516575813293, 0.31943967938423157, -0.282145231962204, 0.2668992877006531, -0.22960402071475983, 0.11477533727884293, -0.12683071196079254, -0.18965332210063934, 0.09107086062431335, 0.48549652099609375, -0.6954213380813599, 0.40398919582366943, -0.06320933252573013, 0.07957585901021957, 0.06719355285167694, -0.02080794982612133, 0.16373121738433838, -0.2564390003681183, -0.26492947340011597, -0.06672534346580505, -0.2586745023727417, 0.00040293484926223755, -0.10235649347305298, -0.06116139143705368, -0.00667085126042366, -0.12094180285930634, -0.3932931423187256, -0.014796941541135311, -0.45530563592910767, -0.17547589540481567, 0.29111674427986145, -0.15258720517158508, 0.3953568637371063, 0.43452224135398865, 0.0012733116745948792, 0.14556869864463806, -0.12557163834571838, -0.028710704296827316, -0.08400887995958328, 0.4058331251144409, -0.26161664724349976, 0.09170582890510559, 0.5665050745010376, -0.006685782223939896, -0.3421810269355774, 0.15743711590766907, -0.08092371374368668, -0.10101303458213806, -0.43462657928466797, 0.18062655627727509, 0.6312641501426697, -0.5566819310188293, -0.27094703912734985, 0.1057053655385971, 0.11234088242053986, -0.04159419611096382, -0.1443416178226471, 0.3904123902320862, -0.03776446357369423, 0.5050055384635925, 0.15269379317760468, 0.07537214457988739, 0.15192118287086487, -0.024606160819530487, 0.04468595236539841, -0.2201678305864334, 0.35747721791267395, 0.2970626950263977, -0.1157662644982338, -0.10506823658943176, -0.057065773755311966, 0.20738916099071503, -0.19489699602127075, 0.3421352207660675, -0.21411162614822388, -0.22223147749900818, -0.09522446990013123, 0.30014461278915405, 0.18653343617916107, 0.1520581692457199, 0.04173438623547554, -0.03233838826417923, -0.11665630340576172, 0.328619122505188, -0.14338849484920502, 0.2544211447238922, 0.19086512923240662, -0.03195108100771904, 0.028394566848874092, 0.1957789808511734, -0.1289578676223755, -0.08749967813491821, 0.008275063708424568, 0.34924253821372986, -0.11322985589504242, -0.26551559567451477, -0.029820891097187996, 0.003573790192604065, -0.03897588327527046, -0.09551601111888885, -0.2530175447463989, -0.06898951530456543, -0.2502748966217041, 0.16109591722488403, 0.04794320464134216, -0.12595948576927185, -0.02653845027089119, 0.03126091882586479, -0.023183010518550873, -0.23883140087127686, 0.19180499017238617, -0.08689837157726288, -0.14084836840629578, 0.3501245677471161, 0.007580507546663284, -0.2158450484275818, -0.13044050335884094, 0.02710792049765587, 0.18230727314949036, -0.03597831726074219, 0.33490630984306335, 0.15295855700969696, -0.03547601401805878, -0.23685064911842346, 0.0391065776348114, 0.4817042350769043, 0.15812793374061584, -0.013921603560447693, 0.07354781031608582, 0.08150287717580795, 0.07890522480010986, 0.017839156091213226, 0.2860039174556732, 0.00811094418168068, 0.07826099544763565, 0.0466495081782341, 0.14516785740852356, -0.006222911179065704, -0.10724980384111404, -0.29134562611579895, 0.4584799110889435, 0.10486221313476562, 0.2002466320991516, 0.4067821204662323, 0.017292674630880356, -0.003218837082386017, 0.3959064483642578, -0.0531291626393795, -0.2769777476787567, 0.06952840089797974, 0.011238678358495235, 0.6116731762886047, -0.16785180568695068, 0.12471422553062439, 0.07526944577693939, 0.21174637973308563, 0.1798262596130371, 0.03575913980603218, -0.18006432056427002, 0.20812660455703735, -0.13592606782913208, -0.15321236848831177, 0.541154146194458, -0.1362389177083969, 0.014910653233528137, -0.006787106394767761, 0.021940521895885468, 0.4590047299861908, 0.2312006950378418, 0.25780022144317627, -0.3163483440876007, -0.3504883646965027, 0.0007417444139719009, 0.10561448335647583, -0.11325456947088242, -0.19692258536815643, 0.20105041563510895, -0.11095333099365234, -0.30959609150886536, -0.06364308297634125, -0.5783081650733948, -0.3299182057380676, 0.48669734597206116, 0.22485026717185974, 0.5121278762817383, -0.1676066368818283, -0.11800263077020645, 0.14495772123336792, 0.36797523498535156, -0.15258798003196716, 0.15408723056316376, 0.30220744013786316, -0.0390835702419281, 0.018621642142534256, 0.20260384678840637, 0.0895346850156784, 0.13583555817604065, -0.20486100018024445, -0.01434855256229639, -0.3012128472328186, -0.027429647743701935, 0.18805748224258423, -0.1548427939414978, 0.18413934111595154, 0.026713328436017036, 0.22319234907627106, -0.06094777584075928, 0.06802726536989212, -0.11772581934928894, 0.031707536429166794, 0.1584734469652176, -0.18284845352172852, -0.3853932023048401, -0.013146482408046722, -0.17111173272132874, 0.04769604653120041, -0.000627957284450531, -0.7032938599586487, -0.06327411532402039, 0.07264988124370575, -0.06937970966100693, -0.03406525403261185, 0.09923778474330902, -0.043120697140693665, 0.20425191521644592, 0.15444724261760712, 0.24497471749782562, 0.1417865753173828, -0.2718694806098938, 0.12096333503723145, -0.2757836580276489, -0.18676051497459412, 0.28453966975212097, -0.23186273872852325, 0.03978751599788666, 0.020781049504876137, 0.10187821835279465, 0.34108859300613403, 0.06822305917739868, -0.5816965103149414, -0.113857202231884, 0.4657232463359833, -0.20126885175704956, 0.1362157016992569, 0.18447740375995636, 0.26688218116760254, 0.10629057884216309, -0.12843114137649536, 0.212474524974823, 0.35382887721061707, -0.21196269989013672, -0.170796200633049, -0.07130181044340134, -0.33269795775413513, -0.05884702503681183, -0.2403455674648285, 0.5293843150138855, 0.20186175405979156, 0.1224137619137764, -0.2680337727069855, 0.1406514048576355, 0.2860543131828308, 0.05670357123017311, -0.028978899121284485, 0.027318062260746956, 0.7622185945510864, -0.09328220784664154, 0.10506893694400787, -0.11119595170021057, 0.215980663895607, 0.12799906730651855, -0.3299756944179535, -0.49485650658607483, 0.08940304815769196, -0.02528863400220871, -0.2356468141078949, 0.10793600976467133, 0.4834584593772888, -0.12256108224391937, 0.13673877716064453, -0.26708289980888367, 0.1285669058561325, 0.31999218463897705, -0.2631489634513855, -0.3871843218803406, -0.39997532963752747, -0.07116074860095978, -0.03558336943387985, -0.03292090445756912, -0.8910266160964966, -0.17213231325149536, 0.11249079555273056, -0.03815922886133194, -0.2536053955554962, 0.08522751927375793, -0.022356484085321426, -0.2716106176376343, -0.37771695852279663, 0.6619982719421387, 0.05689779669046402, 0.03396216034889221, -0.22671574354171753, -0.5398955345153809 ]
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.
[ -0.2745407223701477, -0.07471400499343872, -0.2828441560268402, 0.08163279294967651, 0.10131534188985825, 0.08511572331190109, 0.16226986050605774, 0.18157170712947845, 0.15496572852134705, 0.028999298810958862, 0.0906955748796463, 0.2542612552642822, -0.24301375448703766, 0.09133551269769669, 0.2556442618370056, -0.34201547503471375, 0.2013435810804367, -0.07195033133029938, -0.24810531735420227, -0.35922640562057495, -0.10869914293289185, 0.0939246341586113, -0.26665621995925903, -0.21100129187107086, -0.13362134993076324, -0.060056932270526886, -0.2486087530851364, -0.04428119212388992, -0.42535167932510376, -0.23080095648765564, 0.06616612523794174, 0.2980559170246124, 0.23175828158855438, 0.3571215569972992, -0.00010253637447021902, -0.1791471242904663, 0.1952541172504425, 0.06595970690250397, -0.09675733745098114, -0.3587573766708374, -0.36296626925468445, -0.2668955326080322, 0.17846550047397614, 0.10846651345491409, 0.09927225857973099, 0.0903831422328949, -0.03913361579179764, -0.40964069962501526, 0.027926310896873474, 0.35471290349960327, 0.23870010673999786, 0.356764554977417, 0.4147942364215851, -0.1960522085428238, 0.23967823386192322, 0.0669640302658081, -0.18212391436100006, 0.08796869218349457, 0.4211682081222534, 0.1152283251285553, 0.30466559529304504, 0.14688749611377716, 0.1961059272289276, 0.10997584462165833, 0.3800206184387207, 0.01832268200814724, 0.013787861913442612, -0.5149791240692139, -0.326249897480011, 0.28526046872138977, 0.7866032719612122, -0.23449641466140747, -0.23747442662715912, -0.1045805960893631, -0.03371289744973183, -0.28094711899757385, 0.03654807060956955, 0.03145764395594597, 0.21467262506484985, -0.16369502246379852, 0.06738597899675369, -0.4039888083934784, -0.33536893129348755, 0.4025501608848572, 0.1397748440504074, 0.48646020889282227, 0.123641736805439, 0.12610046565532684, 0.09881323575973511, -0.14917506277561188, 0.395193874835968, -0.0036636944860219955, 0.21630744636058807, 0.26635605096817017, -0.21953915059566498, -0.21650680899620056, -0.09907501190900803, -0.08511698991060257, 0.01853322610259056, -0.1848432719707489, -0.14839716255664825, 0.35971978306770325, -0.30942848324775696, 0.25821059942245483, 0.21343915164470673, -0.15015840530395508, -0.1043485626578331, -0.0047554695047438145, 0.4263310432434082, -0.033694855868816376, 0.0713232159614563, 0.19717857241630554, -0.1961682140827179, 0.0076111555099487305, -0.27579036355018616, 0.027275169268250465, 0.02952425181865692, -0.012821812182664871, 0.12252480536699295, -0.3832065761089325, 0.09562390297651291, -0.36199119687080383, -0.04778095707297325, -0.08453317731618881, -0.3333871066570282, 0.1714133322238922, -0.009538624435663223, 0.027227267622947693, -0.14987407624721527, -0.023156672716140747, -0.0725000873208046, 0.15157699584960938, -0.2152206152677536, 0.01989271305501461, 0.23342444002628326, -0.10189911723136902, 0.0881504938006401, 0.012576811946928501, 0.20164771378040314, 0.15531834959983826, 0.18760861456394196, 0.1086168885231018, -0.1332702934741974, 0.22911305725574493, -0.16917988657951355, 0.03806190937757492, -0.26937001943588257, 0.18624171614646912, -0.0961536169052124, -0.050221510231494904, -0.14967170357704163, -0.21687638759613037, -0.16860432922840118, 0.2626044452190399, -0.14511266350746155, -0.3540809154510498, -0.22397267818450928, 0.37762829661369324, -0.28075385093688965, 0.2039029896259308, -0.1772482991218567, -0.12753510475158691, -0.09540987014770508, -0.33919641375541687, 0.005437646061182022, 0.0747956857085228, -0.4191731810569763, 0.15557458996772766, -0.13233238458633423, -0.058336492627859116, -0.23669740557670593, -0.05418844893574715, -0.19494976103305817, 0.015208423137664795, -0.09529176354408264, -0.12077134847640991, 0.45636481046676636, -0.552367091178894, -0.06465820223093033, -0.16999679803848267, 0.22772087156772614, -0.2298593372106552, -0.16411623358726501, 0.19302158057689667, 0.039700623601675034, -0.1482827514410019, -0.2039141058921814, 0.32110273838043213, -0.14256274700164795, -0.2917487621307373, -0.20709387958049774, -0.11056298762559891, 0.05193526670336723, 0.11169503629207611, 0.3645804524421692, 0.01750713586807251, 0.16154718399047852, 0.2132958471775055, 0.014443308115005493, 0.10343044996261597, 0.03156667202711105, 0.20741131901741028, 0.414486825466156, 0.21465058624744415, 0.46539101004600525, -0.4261690080165863, -0.43927866220474243, 0.1380772590637207, 0.19536946713924408, 0.18815842270851135, 0.10498815029859543, 0.08624672889709473, -0.004376244731247425, 0.10294608771800995, -0.12055496871471405, 0.3178800642490387, 0.14472611248493195, -0.04055717587471008, 0.21057137846946716, -0.14622390270233154, -0.5968180298805237, 0.09261275082826614, -0.10412328690290451, 0.41401198506355286, -0.4701024889945984, 0.30474185943603516, 0.015341789461672306, -0.15350066125392914, 0.16933949291706085, 0.2441728115081787, 0.005732301622629166, -0.4595033824443817, 0.0803297832608223, 0.296469122171402, 0.1613711416721344, 0.2357262521982193, -0.03380635380744934, 0.44494009017944336, 0.4259870946407318, -0.06091940402984619, 0.3119790256023407, -0.31688180565834045, -0.026507500559091568, -0.056937284767627716, 0.010205831378698349, 0.2147311419248581, -0.0372154675424099, -0.09473753720521927, 0.268436074256897, 0.016563652083277702, 0.16712196171283722, 0.26641350984573364, -0.04216377064585686, -0.21579426527023315, 0.030043838545680046, -0.008427707478404045, -0.021350311115384102, -0.13367782533168793, -0.2969609498977661, 0.22897610068321228, 0.6034261584281921, -0.12870319187641144, 0.03560430929064751, 0.06194142997264862, -0.24993255734443665, -0.008935527876019478, 0.09163682162761688, 0.27020007371902466, 0.19163954257965088, 0.14972513914108276, 0.2187141478061676, -0.018969638273119926, -0.29128900170326233, -0.12581434845924377, 0.03649461269378662, 0.19965317845344543, 0.14305704832077026, -0.06021957844495773, -0.013700097799301147, 0.005899482406675816, -0.09948026388883591, -0.019224274903535843, -0.04581395536661148, 0.4715065062046051, 0.07942873239517212, -0.18269290030002594, -0.4099806249141693, -0.18739193677902222, -0.07155609130859375, -0.17729948461055756, -0.15517424046993256, -0.30651596188545227, 0.03282032534480095, -0.01460038498044014, 0.19896554946899414, 0.016423966735601425, 0.19826912879943848, 0.09923149645328522, -0.2671225666999817, 0.36756622791290283, -0.27087801694869995, 0.002239215886220336, -0.21750178933143616, 0.19029970467090607, 0.2565094530582428, -0.08611162006855011, 0.41823428869247437, -0.03400958701968193, 0.23329398036003113, -0.13853555917739868, -0.64036625623703, 0.17638804018497467, -0.1495903730392456, -0.02689194120466709, -0.02024998515844345, -0.08343067765235901, 0.1799279749393463, 0.11929908394813538, 0.07279106974601746, -0.2918451130390167, -0.05149553716182709, -0.044087156653404236, -0.1970873326063156, -0.11120231449604034, -0.13188080489635468, -0.5581414103507996, -0.45936694741249084, -0.2616689205169678, 0.1764208823442459, 0.3262961208820343, 0.08081508427858353, 0.1624584197998047, 0.08413614332675934, 0.1643996685743332, -0.06314384937286377, 0.03947480395436287, 0.11278688907623291, -0.25843706727027893, 0.40675827860832214, -0.30711066722869873, -0.3843969404697418, 0.4961083233356476, -0.10421452671289444, 0.3492160439491272, 0.2127215713262558, -0.3054114878177643, -0.19868513941764832, -0.02530968002974987, -0.01599104329943657, 0.16639338433742523, -0.019606439396739006, 0.23422180116176605, -0.009280534461140633, 0.05265606939792633, -0.25241121649742126, -0.1440179944038391, -0.08615788072347641, 0.3769686222076416, 0.029445581138134003, 0.14342643320560455, 0.334758460521698, -0.03322427347302437, 0.430533230304718, 0.03644873574376106, -0.2452239841222763, 0.050405580550432205, 0.16883547604084015, 0.10867685079574585, 0.011381909251213074, -0.050217658281326294, 0.07712594419717789, 0.03132672607898712, 0.29908913373947144, 0.04891547933220863, 0.12088093906641006, -0.10746306926012039, -0.31403422355651855, 0.3371352255344391, -0.17844313383102417, -0.0018074177205562592, -0.0018808655440807343, -0.25169336795806885, 0.3295738101005554, 0.08878345787525177, -0.19609886407852173, -0.28800487518310547, -0.14876821637153625, 0.1574668139219284, 0.3545747995376587, 0.24198195338249207, -0.12902605533599854, -0.5378569960594177, -0.1852819323539734, -0.1731891632080078, 0.41747599840164185, -0.10197825729846954, 0.17927797138690948, -0.19229258596897125, 0.11066804081201553, 0.11774341017007828, 0.0859801396727562, 0.18783017992973328, -0.4766133427619934, -0.527123212814331, 0.279164582490921, -0.13627249002456665, -0.23418812453746796, -0.024256978183984756, -0.3106381893157959, 0.11367247253656387, -0.02908310294151306, 0.44889017939567566, 0.015480495989322662, -0.3676788806915283, 0.24308724701404572, 0.1308567374944687, 0.08591009676456451, -0.1405642330646515, 0.26299381256103516, -0.010136343538761139, -0.11732552200555801, -0.2620445787906647, -0.10521437972784042, -0.11233888566493988, -0.11672763526439667, -0.5168623328208923, 0.07123604416847229, -0.33272644877433777, 0.1847611665725708, -0.047048330307006836, -0.05658993870019913, 0.27492934465408325, 0.03683147206902504, 0.3168187141418457, -0.14537106454372406, 0.46666744351387024, -0.011513415724039078, -0.38040339946746826, -0.16965311765670776, 0.17406600713729858, -0.09162214398384094, 0.19564741849899292, 0.15047915279865265, -0.05298173800110817, 0.18283626437187195, -0.06081804633140564, -0.031034313142299652, -0.1255171000957489, 0.16413477063179016, 0.41934022307395935, 0.2279086709022522, -0.28524813055992126, -0.607187807559967, 0.1147230863571167, 0.016152273863554, -0.0485958456993103, 0.4207155108451843, -0.13392092287540436, -0.17450818419456482, 0.11098363250494003, 0.04105674847960472, 0.5863262414932251, -0.2631884217262268, -0.29278111457824707, 0.034749798476696014, -0.0863000750541687, 0.4732598662376404, -0.6111388802528381, 0.1742394119501114, -0.03426749259233475, 0.10870567709207535, -0.08511227369308472, -0.2548709809780121, 0.2109452486038208, 0.0722760558128357, -0.20136064291000366, 0.09954236447811127, 0.17709039151668549, 0.13807952404022217, -0.1942109316587448, 0.659774661064148, -0.06312782317399979, -0.14037330448627472, -0.09899745136499405, 0.2516936659812927, -0.028308488428592682, 0.052181921899318695, -0.060260724276304245, -0.19137771427631378, -0.03187745064496994, 0.18113330006599426, -0.28495216369628906, -0.16876009106636047, -0.22131681442260742, 0.008319931104779243, 0.09131792932748795, -0.04329487308859825, 0.3468938171863556, 0.09623097628355026, 0.5915396809577942, -0.03202419355511665, -0.2376018762588501, 0.27973926067352295, -0.2682269215583801, -0.11788816750049591, -0.1228887066245079, 0.017207536846399307, 0.37216290831565857, -0.29674217104911804, -0.012082837522029877, -0.07735525816679001, -0.0002788277342915535, -0.24259760975837708, -0.36513471603393555, -0.07142254710197449, 0.3298625946044922, 0.09708575904369354, 0.19230224192142487, 0.30334579944610596, -0.09278278797864914, 0.00755995512008667, 0.24786651134490967, 0.45117902755737305, 0.014724215492606163, 0.4076642096042633, 0.04798474907875061, 0.000529654324054718, -0.1513432413339615, -0.15054932236671448, 0.27086058259010315, -0.2877451181411743, 0.07257498800754547, 0.43703028559684753, -0.19491265714168549, -0.2097458839416504, -0.19884449243545532, 0.1838003695011139, 0.05712958425283432, 0.07818350195884705, 0.1559590846300125, 0.09627260267734528, 0.19285620748996735, 0.09991782903671265, -0.023509129881858826, -0.4111431837081909, -0.4108959436416626, -0.29322004318237305, -0.22436915338039398, 0.2566511332988739, 0.04946180060505867, 0.27945002913475037, -0.1963750422000885, -0.07223781943321228, 0.17411194741725922, 0.14937160909175873, -0.37918296456336975, 0.0570145845413208, 0.15133535861968994, 0.03618868067860603, 0.22612495720386505, -0.0725005641579628, -0.03612145036458969, -0.07173995673656464, 0.09386391192674637, -0.027947835624217987, -0.46794483065605164, -0.18472224473953247, 0.04202614352107048, 0.1002144068479538, -0.042637549340724945, -0.2287178486585617, 0.1266716569662094, -0.31415069103240967, -0.28279295563697815, -0.024908896535634995, -0.0397677943110466, -0.3050770163536072, -0.1421203762292862, 0.3311747610569, 0.18199796974658966, 0.1702476441860199, 0.5582084059715271, 0.09513989090919495, 0.21245083212852478, 0.20485800504684448, 0.402187317609787, 0.18952004611492157, 0.2601071000099182, -0.08211477845907211, -0.14819790422916412, 0.07816367596387863, -0.13820554316043854, -0.17945772409439087, 0.3982885479927063, -0.16925735771656036, 0.15359097719192505, 0.03235531225800514, 0.10143978893756866, 0.2642422914505005, 0.021073685958981514, 0.03696414828300476, 0.14142945408821106, 0.21311019361019135, -0.13443012535572052, -0.055899783968925476, 0.40270641446113586, 0.49797600507736206, 0.10187976062297821, 0.076864093542099, 0.5620503425598145, 0.1583326905965805, 0.28498029708862305, 0.17854461073875427, 0.23345711827278137, -0.14217440783977509, 0.40867558121681213, -0.1030537486076355, -0.053733132779598236, 0.07344022393226624, 0.08991994708776474, 0.38149261474609375, 0.09467067569494247, 0.3289194703102112, 0.18405809998512268, 0.4397299289703369, 0.14604389667510986, -0.08007971942424774, 0.5982280969619751, -0.3191629648208618, 0.06604903936386108, 0.16271120309829712, 0.06078612431883812, 0.09383036196231842, 0.07741852104663849, 0.4690662920475006, -0.009389805607497692, -0.13898442685604095, -0.05403551459312439, -0.1166994571685791, -0.10878206044435501, -0.005931571591645479, -0.04140736907720566, -0.17655441164970398, -0.20635242760181427, -0.09554574638605118, 0.042455997318029404, -0.24463890492916107, 0.10841155052185059, -0.0067840442061424255, 0.02507495880126953, -0.1830291897058487, 0.3499510884284973, -0.0880805179476738, -0.06958433240652084, 0.20820173621177673, 0.4416111409664154, -0.2254699021577835, -0.20558935403823853, -0.09230738878250122, -0.07757948338985443, -0.06999090313911438, 0.02039320021867752, -0.3597542941570282, 0.3094954788684845, 0.051701620221138, -0.19295810163021088, 0.20771804451942444, 0.40192317962646484, 0.07114528864622116, 0.36640802025794983, -0.1109786331653595, 0.15951362252235413, -0.05355575680732727, -0.022820819169282913, -0.2833409607410431, -0.021147258579730988, -0.41630253195762634, -0.08433752506971359, 0.11024495214223862, -0.15187138319015503, -0.045349545776844025, -0.04589655250310898, -0.47221639752388, -0.5669896006584167, 0.5651452541351318, -0.18957000970840454, -0.15351352095603943, -0.25356125831604004, 0.10814587771892548, -0.09533166885375977, 0.2511746883392334, 0.12147380411624908, 0.45501282811164856, 0.05822417885065079, -0.37744593620300293, -0.2873685956001282, 0.0056760311126708984, 0.149158775806427, 0.2815351188182831, -0.11302030831575394, 0.19985252618789673, 0.2401162087917328, -0.03342721611261368, 0.4303882122039795, -0.2973402440547943, 0.09230461716651917, -0.004097919911146164, -0.32758745551109314, 0.00985543429851532, -0.20289473235607147, 0.04155024141073227, 0.1104595810174942, 0.25907841324806213, 0.07082263380289078, -0.2688300609588623, -0.025742657482624054, -0.41214805841445923, -0.02335938811302185, 0.33951470255851746, 0.0814245343208313, 0.4212687015533447, 0.14328189194202423, 0.29541245102882385, 0.18000543117523193, -0.045552950352430344, -0.41668328642845154, -0.1639951467514038, -0.01975807175040245, 0.23975995182991028, -0.12995515763759613, 0.0806007981300354, -0.19504308700561523, 0.01444610022008419, -0.16572193801403046, 0.14214080572128296, -0.16518990695476532, -0.35498902201652527, 0.03663385659456253, 0.06656566262245178, -0.08965340256690979, 0.03277633339166641, 0.4078938066959381, 0.13496807217597961, 0.233576238155365, -0.24015268683433533, -0.21379780769348145, 0.1258505880832672, 0.05555248260498047, -0.07777585834264755, -0.24822504818439484, -0.1938065141439438, -0.18507090210914612, 0.26290974020957947, 0.1135978028178215, -0.875744104385376, -0.21726350486278534, 0.2978818118572235, -0.09050773829221725, 0.06773225963115692, 0.26117268204689026, 0.4608120918273926, -0.01169823482632637, -0.1992034763097763, 0.08760400116443634, 0.004510505124926567, -0.05850180238485336, -0.1302710324525833, -0.15369963645935059 ]
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 ?
[ -0.15169493854045868, 0.00311848521232605, -0.01721992902457714, 0.3954119086265564, -0.06618136167526245, 0.05573359876871109, 0.04334799572825432, 0.33450645208358765, 0.34885430335998535, 0.1168280839920044, 0.319930762052536, 0.17322245240211487, -0.1478523463010788, -0.08189839869737625, -0.01621365174651146, -0.45966222882270813, -0.051987938582897186, 0.24634328484535217, 0.036583516746759415, -0.055987875908613205, -0.2965039908885956, 0.26790350675582886, -0.29956990480422974, 0.1040332168340683, -0.06252715736627579, -0.14766909182071686, 0.150971457362175, 0.02395021542906761, -0.2917691171169281, -0.37569132447242737, 0.04113136976957321, -0.22728359699249268, 0.27645933628082275, 0.149385467171669, -0.00011269281822023913, -0.03307892754673958, 0.666763961315155, 0.013357171788811684, -0.44841063022613525, -0.39789509773254395, -0.1365269124507904, -0.40732404589653015, 0.45148611068725586, -0.2783578336238861, 0.010811049491167068, -0.1832435578107834, 0.3193396329879761, 0.17982672154903412, 0.41272199153900146, 0.29741448163986206, 0.2151002287864685, 0.013293188065290451, 0.31618157029151917, 0.02897235006093979, 0.43313151597976685, 0.023455746471881866, 0.037681788206100464, -0.08287952840328217, -0.13199631869792938, 0.034157395362854004, -0.12720948457717896, 0.15120500326156616, -0.2554621398448944, 0.08563673496246338, 0.34953808784484863, -0.19634762406349182, -0.0711800679564476, -0.5439115762710571, 0.13230383396148682, 0.14068442583084106, 0.6836259961128235, -0.046239934861660004, 0.16607248783111572, 0.09736485034227371, 0.11649180203676224, 0.05789187550544739, 0.23870918154716492, 0.3242081105709076, -0.4263734519481659, -0.30791041254997253, -0.21799717843532562, -0.24967512488365173, -0.2927876114845276, 0.46477794647216797, 0.04294971376657486, 0.5524114370346069, -0.06267865002155304, 0.22764942049980164, -0.04183034226298332, -0.023902785032987595, -0.06886428594589233, -0.1292952299118042, 0.028790326789021492, 0.30135416984558105, 0.060495082288980484, 0.09357406198978424, 0.021500572562217712, 0.22951245307922363, 0.19686871767044067, -0.043417222797870636, -0.17343655228614807, -0.1454145312309265, 0.14635173976421356, 0.14654338359832764, 0.03337777405977249, 0.25672271847724915, 0.1367921382188797, -0.16623704135417938, 0.1898491382598877, 0.21545089781284332, -0.06594294309616089, 0.08795030415058136, -0.03179530054330826, -0.17879846692085266, -0.46619170904159546, 0.02353104203939438, 0.2609488368034363, -0.17632582783699036, -0.036259714514017105, 0.04170362651348114, -0.42591845989227295, -0.24169892072677612, -0.02936335653066635, 0.32280653715133667, -0.21127139031887054, 0.3524998724460602, 0.374053955078125, 0.07673802971839905, -0.3170037269592285, -0.3271934688091278, 0.020153403282165527, 0.2933025062084198, -0.2867259681224823, 0.08495920896530151, 0.23058660328388214, 0.30473241209983826, 0.35114482045173645, -0.017778804525732994, -0.07731509208679199, 0.008208584040403366, 0.16384051740169525, -0.08172617852687836, -0.21706679463386536, 0.2386733442544937, 0.31429746747016907, 0.27012401819229126, 0.058486949652433395, -0.33255624771118164, -0.07982198894023895, 0.25485438108444214, -0.3396589457988739, -0.07505985349416733, -0.03394258767366409, 0.16418899595737457, -0.10578513890504837, 0.013759659603238106, -0.34825336933135986, 0.24756541848182678, 0.03083135187625885, -0.2846430242061615, -0.17346858978271484, -0.1604103446006775, -0.08583105355501175, -0.35577037930488586, 0.2626807987689972, 0.33504220843315125, 0.13158604502677917, -0.10666563361883163, -0.08848176151514053, 0.1678844690322876, 0.22503505647182465, 0.056384846568107605, -0.12499351054430008, 0.4726501703262329, -0.04442140460014343, -0.07877199351787567, 0.7422001957893372, -0.13415753841400146, -0.17248471081256866, 0.23400962352752686, 0.13185544312000275, 0.019067810848355293, 0.0614241324365139, -0.06255611777305603, 0.12045732885599136, 0.08603453636169434, -0.051530465483665466, 0.566815197467804, -0.011338306590914726, 0.0343194380402565, -0.40983110666275024, -0.2514777183532715, 0.17665739357471466, 0.06843884289264679, 0.3860551118850708, 0.15072938799858093, 0.010939329862594604, 0.5279035568237305, 0.08195846527814865, -0.03969378396868706, 0.026423433795571327, 0.3199971318244934, -0.297061026096344, -0.09037400782108307, -0.021785132586956024, 0.06996960937976837, -0.1751367151737213, -0.0254768468439579, -0.3948405385017395, 0.15085041522979736, 0.15849989652633667, 0.10617592930793762, -0.3540581166744232, -0.1429709494113922, -0.052457261830568314, -0.1548895239830017, 0.1619727909564972, 0.1855083703994751, 0.09486676752567291, 0.15069285035133362, 0.054137613624334335, 0.06682412326335907, -0.1275254189968109, -0.04091872274875641, -0.6120949387550354, 0.1855083703994751, -0.27361801266670227, -0.07680457085371017, 0.11377494037151337, 0.2102164924144745, 0.24915528297424316, 0.011264992877840996, -0.1234433725476265, 0.006876267492771149, -0.20115044713020325, -0.06263680756092072, -0.027232900261878967, -0.07050861418247223, 0.20225463807582855, -0.3002311587333679, 0.31451135873794556, 0.2933093011379242, 0.1526009440422058, -0.1323699653148651, 0.08773380517959595, -0.07689173519611359, 0.11841444671154022, 0.3404713273048401, 0.07403960824012756, 0.10327670723199844, 0.004244735464453697, -0.06270985305309296, -0.0017370712012052536, 0.2277480959892273, 0.3068396747112274, 0.42919909954071045, 0.0985184907913208, -0.10778805613517761, 0.0011934638023376465, -0.20937365293502808, 0.36993667483329773, 0.04473312571644783, 0.04405396431684494, 0.26717716455459595, -0.4356001019477844, -0.015432308427989483, 0.1121867448091507, -0.16287536919116974, 0.16692376136779785, 0.2249116748571396, -0.1555088311433792, -0.03987376391887665, -0.0813923254609108, -0.12739378213882446, 0.20037701725959778, 0.1784820258617401, 0.36961862444877625, 0.11934078484773636, 0.08972068876028061, 0.07794459909200668, -0.07833760231733322, -0.1367141604423523, -0.14669492840766907, 0.317047655582428, -0.3040972650051117, 0.07823999971151352, -0.023409854620695114, -0.47671717405319214, -0.24302992224693298, 0.04412386193871498, -0.45233333110809326, -0.3893639147281647, -0.04385986179113388, 0.26631495356559753, 0.10461795330047607, 0.09521710127592087, 0.34473323822021484, 0.14382530748844147, -0.01657220348715782, -0.09474824368953705, -0.16188697516918182, -0.2388000786304474, -0.46269989013671875, 0.06032678112387657, 0.3643651306629181, 0.2716529369354248, 0.26252681016921997, -0.03581193834543228, 0.0011127106845378876, -0.016719801351428032, -0.26169857382774353, 0.035663191229104996, -0.08692272007465363, 0.19745075702667236, -0.1058526337146759, 0.4281997084617615, -0.06742101907730103, -0.19357021152973175, 0.26940399408340454, -0.03542928025126457, -0.09465430676937103, 0.0787811279296875, -0.13136334717273712, 0.01989472657442093, -0.018719865009188652, -0.3938398063182831, -0.19605626165866852, -0.5722573399543762, 0.03628219664096832, 0.5017025470733643, 0.15746326744556427, 0.1577184647321701, 0.22311870753765106, 0.07921328395605087, 0.37221410870552063, 0.3003673851490021, -0.07484946399927139, 0.08280737698078156, 0.3569962978363037, -0.24258480966091156, -0.45372629165649414, 0.006873439997434616, -0.25352829694747925, 0.20978479087352753, 0.1102360263466835, -0.5407999753952026, -0.06307719647884369, -0.03617759048938751, -0.09409025311470032, -0.07280644029378891, 0.09220913052558899, 0.4026174247264862, -0.02519320696592331, 0.0642644464969635, 0.03355094790458679, -0.004459993913769722, -0.15063700079917908, -0.2979053556919098, 0.539756178855896, 0.3464975655078888, 0.3857057988643646, -0.12387114763259888, 0.9763141870498657, 0.11937181651592255, -0.010839805006980896, 0.21248789131641388, 0.1578991711139679, 0.04472491145133972, -0.20025968551635742, -0.21275107562541962, 0.1421237289905548, -0.015238747000694275, -0.051308535039424896, 0.37091290950775146, -0.09053631126880646, -0.25643083453178406, -0.28977763652801514, -0.05816591531038284, -0.2621939778327942, -0.12282593548297882, 0.2391219139099121, 0.04297163337469101, 0.1968320608139038, -0.11754602193832397, 0.04727274924516678, -0.060972217470407486, -0.3704538345336914, 0.08413924276828766, 0.2635555863380432, -0.059675805270671844, 0.08931107074022293, 0.2353031486272812, -0.1747625172138214, -0.500149667263031, 0.40322381258010864, 0.02955075167119503, 0.09460225701332092, -0.2325175553560257, -0.02591796964406967, 0.172696053981781, -0.04392104223370552, 0.22944006323814392, -0.17277568578720093, -0.18007849156856537, 0.3131050765514374, 0.27893176674842834, -0.4931185841560364, -0.0071612149477005005, -0.22605517506599426, 0.21649278700351715, 0.31887030601501465, -0.06501565873622894, -0.4468342065811157, 0.011840201914310455, 0.35860732197761536, 0.2625586986541748, -0.16119277477264404, -0.06786547601222992, -0.4241856336593628, -0.10542401671409607, -0.35802534222602844, -0.16013994812965393, -0.1432648003101349, 0.3781591057777405, 0.014349144883453846, 0.08232946693897247, -0.051163285970687866, -0.07533728331327438, 0.03009631484746933, 0.056553084403276443, 0.1316145658493042, 0.15762071311473846, -0.3129262924194336, -0.17733995616436005, 0.2570785880088806, -0.3667410910129547, 0.207405686378479, 0.33988094329833984, 0.022439733147621155, 0.009034428745508194, 0.1605154275894165, 0.10108733922243118, 0.1211695596575737, 0.038661349564790726, -0.09372284263372421, -0.09350200742483139, -0.02033752202987671, -0.10446882992982864, 0.04888502135872841, 0.20298326015472412, -0.1898670792579651, -0.19062067568302155, -0.1861119568347931, 0.9091108441352844, 0.11542610824108124, -0.08426406234502792, 0.39913371205329895, -0.10438928008079529, -0.30044353008270264, 0.4643959403038025, 0.2923504114151001, 1.033750057220459, -0.05213094502687454, 0.07410755753517151, 0.35504287481307983, 0.14539143443107605, 0.5975180864334106, -0.5634698271751404, 0.24681450426578522, -0.46771374344825745, 0.19664128124713898, -0.042075254023075104, -0.0888727456331253, 0.18185505270957947, -0.025462642312049866, -0.42296040058135986, 0.30799660086631775, 0.21473200619220734, 0.05316548049449921, -0.004552369937300682, 0.4972430467605591, -0.007547469809651375, -0.2034151405096054, -0.2026687115430832, 0.11879855394363403, -0.45273542404174805, 0.4196888506412506, -0.31372305750846863, -0.07739031314849854, -0.03726926073431969, -0.26645174622535706, -0.32122600078582764, 0.23819094896316528, -0.28426429629325867, 0.2605060935020447, -0.0788804292678833, -0.4040435552597046, 0.13645246624946594, 0.20011821389198303, -0.07382701337337494, 0.22187985479831696, -0.36752042174339294, 0.30768585205078125, -0.3412056267261505, -0.3970913589000702, -0.16703787446022034, 0.30388855934143066, 0.31267955899238586, -0.20412343740463257, -0.09111820161342621, 0.26975297927856445, -0.1574624627828598, -0.19343270361423492, -0.11627297103404999, -0.04585834592580795, 0.4036433696746826, -0.08228058367967606, -0.28408175706863403, 0.08862954378128052, -0.11740091443061829, 0.10527443885803223, 0.12881626188755035, -0.15789106488227844, -0.023842837661504745, -0.1117863580584526, 0.4168771505355835, -0.12245699763298035, 0.15307927131652832, 0.21601274609565735, 0.351275771856308, -0.0435095876455307, 0.6242598295211792, 0.2489510029554367, -0.12671306729316711, -0.19333383440971375, -0.08216258138418198, -0.36340129375457764, -0.22032541036605835, -0.13269227743148804, 0.06104392185807228, 0.026035770773887634, -0.16175705194473267, 0.31193411350250244, 0.045710865408182144, 0.14814773201942444, 0.08542022109031677, -0.47049370408058167, -0.08573375642299652, 0.3661576509475708, -0.21597419679164886, -0.02671365812420845, 0.1270245611667633, -0.11447963118553162, 0.46333280205726624, 0.13523633778095245, -0.2516554594039917, -0.19684122502803802, -0.2700388729572296, 0.20194363594055176, 0.33353686332702637, -0.0315210297703743, -0.09638059884309769, 0.18977464735507965, 0.038723282516002655, -0.18776559829711914, -0.1967306137084961, -0.19243746995925903, 0.07814830541610718, 0.16214875876903534, 0.20964843034744263, -0.0580848827958107, -0.10040753334760666, -0.2537480890750885, 0.040457434952259064, -0.17626824975013733, -0.17049220204353333, -0.2846740782260895, -0.16244207322597504, 0.07903767377138138, -0.0626453310251236, 0.28314512968063354, -0.39182209968566895, 0.21595892310142517, -0.15372951328754425, 0.45376354455947876, 0.1176493689417839, 0.04202556237578392, 0.5127455592155457, -0.055491093546152115, -0.742988646030426, 0.03110482543706894, 0.013880766928195953, 0.11196262389421463, 0.25134673714637756, -0.10180933773517609, 0.07591325044631958, -0.050089895725250244, -0.18661344051361084, 0.017425913363695145, -0.12645600736141205, 0.07031650841236115, 0.11335647851228714, 0.15643487870693207, -0.16165699064731598, 0.36316293478012085, 0.39460474252700806, 0.1727810800075531, -0.17091768980026245, 0.0786660835146904, 0.03425268456339836, -0.0917048379778862, -0.0037516653537750244, 0.2260856032371521, -0.005977984517812729, -0.055716536939144135, 0.023208700120449066, -0.12723307311534882, -0.02531924471259117, -0.013456884771585464, 0.011188994161784649, -0.020251654088497162, 0.12814384698867798, 0.11533625423908234, -0.1465819776058197, 0.184061661362648, 0.054378442466259, 0.006913401186466217, 0.01574249565601349, -0.18640601634979248, 0.23530103266239166, 0.15203259885311127, 0.37113508582115173, 0.04706276208162308, 0.0498722605407238, 0.04481049254536629, 0.11762683093547821, -0.14441266655921936, -0.4065524935722351, 0.1806311309337616, -0.12935304641723633, 0.12624359130859375, -0.4547829329967499, -0.07528595626354218, -0.28914397954940796, 0.14150762557983398, 0.107052743434906, -0.2766270637512207, 0.057172488421201706, 0.28942739963531494, -0.34862011671066284, -0.408365935087204, 0.43975356221199036, 0.14451634883880615, 0.02715018391609192, -0.15341520309448242, 0.36439910531044006, -0.1512814313173294, 0.0181087926030159, -0.08264751732349396, 0.383638471364975, 0.3436254858970642, 0.5497421622276306, -0.47595930099487305, -0.05491851642727852, 0.05561330169439316, -0.15051956474781036, -0.07915543764829636, 0.11907316744327545, 0.4115956425666809, 0.23406927287578583, 0.18227368593215942, 0.13737617433071136, -0.23148688673973083, -0.03167343512177467, 0.18226854503154755, -0.10571850836277008, -0.08261997997760773, 0.1635473072528839, 0.002595210447907448, -0.1768183410167694, -0.24999567866325378, -0.02985338866710663, -0.678904116153717, 0.08470211923122406, -0.017861846834421158, 0.19374920427799225, -0.09923888742923737, -0.31724512577056885, 0.028229273855686188, -0.19291368126869202, 0.453159362077713, 0.33714932203292847, 0.3244171142578125, -0.3690621852874756, -0.31667032837867737, -0.3890501856803894, 0.1598982959985733, -0.3162230849266052, -0.3201473355293274, -0.09974776208400726, 0.23386363685131073, -0.15514898300170898, 0.3409181237220764, -0.23970501124858856, -0.0007637313101440668, 0.07648145407438278, 0.029246600344777107, -0.2659042179584503, -0.16083340346813202, 0.0497715026140213, -0.11939367651939392, -0.3061857521533966, -0.12973952293395996, -0.004993867129087448, -0.2757503390312195, 0.06248512119054794, -0.026224423199892044, -0.25990456342697144, 0.2788187861442566, 0.053317029029130936, 0.2739656865596771, 0.07670781761407852, 0.626677393913269, -0.10283482819795609, -0.06840270757675171, -0.0048550330102443695, -0.06971421837806702, -0.06355053931474686, 0.3697381615638733, 0.03681907430291176, 0.5085330009460449, -0.08119191974401474, 0.17832893133163452, -0.3596469461917877, 0.49303311109542847, -0.26348695158958435, 0.14966356754302979, -0.1138262078166008, -0.1301417201757431, -0.2357620894908905, 0.315569132566452, -0.03986366465687752, 0.08727344125509262, -0.11603307723999023, 0.07417353987693787, -0.4722554087638855, -0.21190263330936432, 0.31642910838127136, -0.7915248274803162, -0.3263425827026367, 0.11869138479232788, 0.09361734986305237, -0.08433306962251663, 0.023439951241016388, -0.5125057101249695, -0.06366849690675735, 0.2648855447769165, -0.07161727547645569, -0.10451541841030121, 0.03642389550805092, -0.11945199966430664, -0.058210816234350204, -0.11786192655563354, -0.22919020056724548, -0.12916311621665955, -0.22626130282878876, -0.18604321777820587, -0.30095940828323364 ]
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
[ 0.3151267468929291, -0.1353006213903427, -0.005951644852757454, 0.20902259647846222, 0.09835072606801987, -0.18703681230545044, 0.2608935534954071, 0.3269939124584198, -0.1120341569185257, -0.1279546171426773, -0.10739805549383163, 0.4277702271938324, 0.081903837621212, -0.20389296114444733, 0.09390927851200104, 0.06840936839580536, 0.16564157605171204, 0.11475609987974167, -0.11452038586139679, -0.04316331818699837, -0.17623743414878845, 0.10245218127965927, -0.08972574025392532, -0.11245466768741608, -0.4616798460483551, 0.14604823291301727, -0.2929147779941559, -0.15815600752830505, 0.21726326644420624, -0.39209818840026855, 0.3604980707168579, 0.07840589433908463, -0.25562795996665955, 0.2644164264202118, -0.00011643479956546798, -0.03943786397576332, 0.1898857206106186, -0.06046994775533676, -0.09947286546230316, 0.1508258432149887, 0.02018778771162033, -0.14972227811813354, 0.16033540666103363, -0.3118852376937866, -0.01663912460207939, 0.11477603018283844, 0.05911436304450035, 0.13279888033866882, 0.5603553056716919, 0.0672883465886116, 0.19152891635894775, 0.2768821120262146, -0.3045634329319, 0.1994514763355255, 0.15829500555992126, -0.0923035740852356, -0.19662734866142273, -0.07651291787624359, 0.018527425825595856, -0.2760593891143799, -0.03462202101945877, 0.3285541534423828, -0.0990368127822876, 0.09711770713329315, -0.01888156682252884, 0.09337882697582245, 0.16540852189064026, -0.19152851402759552, 0.08571898192167282, 0.06397505104541779, -0.12460622936487198, -0.3299974203109741, -0.2282962203025818, -0.36594727635383606, -0.22737908363342285, -0.3725855350494385, 0.13380682468414307, 0.1214892566204071, -0.151002898812294, -0.11508496105670929, -0.28873348236083984, 0.2583758533000946, 0.18000319600105286, 0.06546595692634583, 0.14850790798664093, 0.34907928109169006, 0.17052243649959564, 0.003995656967163086, 0.2054346650838852, 0.1646483838558197, -0.022904343903064728, -0.167671337723732, 0.039331141859292984, 0.38251471519470215, -0.32864588499069214, -0.1350691169500351, 0.25804200768470764, -0.037615031003952026, -0.02088424563407898, -0.035171620547771454, 0.2935587465763092, -0.05806347727775574, 0.030032485723495483, 0.03780946135520935, 0.14594821631908417, 0.44085782766342163, -0.05845873802900314, 0.26226210594177246, 0.03631453216075897, -0.1119014173746109, -0.47820383310317993, 0.07153686881065369, 0.3043478727340698, 0.08766844123601913, 0.30457645654678345, 0.055590637028217316, -0.07608941942453384, -0.09867473691701889, 0.04273633658885956, 0.09191568940877914, -0.3604728877544403, -0.16655229032039642, 0.04151648283004761, 0.12388285994529724, 0.010513169690966606, 0.18547239899635315, -0.2612515091896057, 0.041237086057662964, -0.5030217170715332, 0.16908606886863708, -0.31585007905960083, 0.011568843387067318, -0.47468942403793335, 0.30212485790252686, 0.2482239156961441, -0.20883436501026154, 0.27461838722229004, 0.3141807019710541, -0.2436191588640213, -0.2717688977718353, 0.29848527908325195, -0.20965208113193512, 0.355764776468277, 0.09198476374149323, -0.11236394941806793, 0.2093500941991806, -0.05222238227725029, 0.1795106679201126, -0.20357465744018555, 0.10834099352359772, -0.10235651582479477, -0.1805284470319748, 0.34438660740852356, 0.16676528751850128, -0.2625376582145691, 0.15600115060806274, 0.325206995010376, 0.21365125477313995, 0.7869170308113098, -0.38525667786598206, 0.1696178913116455, -0.07720290124416351, -0.2634733319282532, -0.3530149459838867, -0.03340950608253479, 0.2723201513290405, 0.12401340901851654, -0.3135712146759033, 0.16955462098121643, 0.23233549296855927, 0.0021243467926979065, 0.5360867381095886, -0.16725273430347443, 0.15477803349494934, -0.32080078125, 0.03022192418575287, 0.4241330027580261, -0.29382696747779846, -0.46321871876716614, 0.1235976591706276, -0.010485541075468063, 0.175144225358963, 0.11865317821502686, 0.19418875873088837, 0.11643929779529572, -0.28288182616233826, 0.20412296056747437, 0.04053041338920593, -0.10490642488002777, 0.09167332202196121, -0.37813282012939453, 0.01463225670158863, 0.3601006269454956, -0.24554426968097687, 0.000285957008600235, 0.06188816577196121, -0.17130637168884277, -0.0000895848497748375, 0.2635815441608429, 0.0021530501544475555, 0.13179726898670197, 0.021445900201797485, -0.0550762377679348, -0.1804734319448471, 0.07052115350961685, -0.18271583318710327, 0.010329119861125946, 0.2068817913532257, -0.29167336225509644, -0.03090876340866089, 0.24033492803573608, -0.11940084397792816, 0.10904709249734879, -0.27650582790374756, -0.26970374584198, -0.49192121624946594, 0.17864221334457397, 0.05129951611161232, 0.48496031761169434, -0.008755091577768326, -0.08701664954423904, 0.2997877597808838, 0.39124879240989685, -0.11880902945995331, -0.3205811679363251, 0.11166104674339294, -0.0021976064890623093, -0.16492031514644623, -0.35923901200294495, 0.5168806910514832, 0.21017169952392578, -0.008701598271727562, -0.1274029016494751, -0.04566941410303116, 0.005072888918220997, 0.08217810839414597, -0.25943881273269653, 0.08567062765359879, 0.10755650699138641, 0.14894163608551025, 0.032679490745067596, 0.1141880452632904, -0.03314945101737976, -0.2737235128879547, 0.1267448216676712, 0.48298293352127075, 0.3349759876728058, 0.10381120443344116, -0.0001738220453262329, -0.08665530383586884, -0.06366563588380814, -0.15970763564109802, -0.22713595628738403, -0.34744158387184143, 0.25259459018707275, -0.03174920007586479, 0.4256989657878876, 0.2751820981502533, -0.04971015825867653, 0.2771766781806946, 0.5075843930244446, 0.18606296181678772, -0.022791534662246704, -0.09178218990564346, -0.01519213616847992, -0.26299023628234863, -0.09401455521583557, 0.21855495870113373, 0.5466564893722534, 0.0993313267827034, 0.14615941047668457, 0.18917642533779144, -0.3211837410926819, -0.20500105619430542, -0.029941104352474213, -0.12286625802516937, 0.06329655647277832, -0.042650334537029266, 0.37207135558128357, -0.08435273915529251, -0.2066480815410614, 0.34665051102638245, 0.1877124160528183, -0.1285448521375656, 0.037768322974443436, 0.2113196849822998, -0.23826144635677338, -0.4322883188724518, -0.3985510468482971, -0.20139983296394348, -0.12059199810028076, -0.30527982115745544, 0.1275995969772339, 0.30709323287010193, 0.09445148706436157, 0.28941023349761963, -0.20783571898937225, 0.08653531968593597, -0.17033296823501587, -0.15077237784862518, 0.0021730661392211914, -0.34527069330215454, -0.31340864300727844, 0.0322800874710083, -0.016492675989866257, -0.2743602395057678, 0.2854227125644684, 0.018409589305520058, -0.1734153777360916, -0.1414489448070526, -0.39166513085365295, 0.20791564881801605, -0.008682787418365479, -0.20053015649318695, -0.19862213730812073, -0.36847949028015137, 0.09001211822032928, 0.012220765464007854, 0.3374744951725006, -0.3655262291431427, -0.24376414716243744, 0.055022671818733215, -0.08614634722471237, 0.10925304889678955, -0.27967268228530884, -0.10740382224321365, -0.1448773294687271, -0.04877806454896927, 0.20083214342594147, -0.28256672620773315, 0.25015515089035034, 0.3470872640609741, -0.2549815773963928, -0.09772158414125443, -0.1596534103155136, -0.16263175010681152, -0.4625169634819031, -0.036857955157756805, 0.21641843020915985, -0.31619617342948914, -0.189056396484375, -0.38392314314842224, -0.3472077250480652, 0.26691100001335144, 0.17019525170326233, -0.43564140796661377, -0.433349609375, -0.061189547181129456, 0.073504239320755, 0.17087775468826294, 0.04932292923331261, 0.38772496581077576, -0.17816810309886932, -0.01823333650827408, -0.1533551812171936, -0.3007078170776367, 0.20788156986236572, 0.39638322591781616, 0.35678884387016296, -0.17569974064826965, 0.22636640071868896, 0.13107015192508698, 0.5557451844215393, 0.631463348865509, 0.06918303668498993, 0.3732171654701233, -0.013673929497599602, 0.21526825428009033, -0.26185548305511475, -0.1999867707490921, 0.1839580535888672, -0.1526450514793396, -0.2572273910045624, 0.1832875907421112, 0.17619872093200684, -0.5797118544578552, -0.06663401424884796, 0.16522251069545746, -0.27863410115242004, -0.3541930317878723, 0.09116494655609131, -0.3137050271034241, 0.22862017154693604, 0.31083017587661743, 0.23395311832427979, -0.41848793625831604, -0.11604499816894531, -0.1248803436756134, -0.3066101670265198, 0.14552704989910126, 0.09425295889377594, -0.09625327587127686, -0.24743661284446716, -0.4561161994934082, 0.20617571473121643, 0.24228781461715698, 0.1278308480978012, 0.02266521379351616, -0.23083919286727905, 0.0699959322810173, 0.01940467767417431, 0.4736870527267456, -0.03395918756723404, -0.003952924162149429, 0.1012185662984848, -0.0711037889122963, -0.40342977643013, -0.2792932689189911, 0.23190394043922424, 0.4614938795566559, 0.4238910377025604, 0.2808866500854492, -0.21179400384426117, -0.10709019005298615, -0.0721648633480072, 0.13284051418304443, -0.23876851797103882, -0.39052367210388184, -0.23795627057552338, -0.06291764229536057, 0.06605026870965958, 0.07129471004009247, -0.12635430693626404, -0.18954648077487946, -0.19810256361961365, -0.0718088448047638, 0.19534069299697876, -0.03914891555905342, 0.03154459595680237, 0.24644117057323456, 0.2672863006591797, -0.1262178122997284, 0.17920629680156708, 0.17825369536876678, -0.10764861106872559, -0.4886571168899536, 0.4111676514148712, -0.18163657188415527, 0.3603590130805969, 0.18541860580444336, -0.05674787610769272, -0.020053081214427948, 0.4066050350666046, 0.19946922361850739, 0.14544135332107544, -0.07756063342094421, -0.07732506096363068, -0.026072043925523758, -0.22639605402946472, 0.32063502073287964, 0.0029674163088202477, -0.2382083386182785, -0.455353319644928, 0.23632293939590454, 0.15792149305343628, -0.14968989789485931, 0.228380486369133, -0.6113734245300293, -0.11635462939739227, 0.3909173011779785, -0.19715052843093872, 0.7841208577156067, 0.02420204132795334, 0.47903695702552795, 0.22325578331947327, -0.05244646593928337, 0.38019251823425293, -0.046632565557956696, 0.36362510919570923, -0.3955558240413666, 0.18653632700443268, 0.12171317636966705, -0.17433662712574005, -0.33306577801704407, 0.02130279317498207, 0.06636716425418854, 0.4315521717071533, 0.2073076218366623, 0.48412686586380005, 0.020280152559280396, -0.21137654781341553, 0.20552682876586914, 0.020140070468187332, -0.32142817974090576, 0.17260445654392242, 0.051495250314474106, 0.24724841117858887, -0.06316383928060532, -0.22587215900421143, -0.07192140817642212, -0.1654232293367386, -0.031881194561719894, 0.2956010401248932, -0.13948577642440796, 0.4080139994621277, 0.3562556803226471, -0.2011677324771881, -0.6638270616531372, 0.09609924256801605, 0.13754159212112427, -0.39595749974250793, -0.01546795479953289, -0.11183474957942963, 0.47419649362564087, 0.4154973030090332, -0.05152217671275139, -0.2530066967010498, 0.36641138792037964, -0.07084521651268005, -0.04830094426870346, -0.0550028532743454, -0.00859941728413105, -0.4087691903114319, -0.2352360486984253, 0.31752216815948486, -0.3334673047065735, 0.05075538158416748, 0.13472510874271393, 0.183037668466568, -0.07473671436309814, -0.11483460664749146, 0.11003774404525757, 0.07397682964801788, -0.12934610247612, 0.35186007618904114, -0.3545050323009491, -0.10786082595586777, 0.03488881140947342, 0.2857210636138916, 0.28677165508270264, 0.08036516606807709, 0.36261308193206787, -0.2709963619709015, 0.06172808259725571, -0.03672278672456741, -0.3326627016067505, -0.31491518020629883, -0.3201562762260437, 0.10790367424488068, -0.2535044252872467, -0.18748363852500916, 0.1025049015879631, 0.0071214064955711365, 0.12069103866815567, 0.37156054377555847, -0.08053518831729889, -0.1893053650856018, -0.5517704486846924, -0.053989216685295105, -0.10335403680801392, 0.1609768122434616, -0.017603058367967606, 0.6225500106811523, 0.07026728987693787, 0.393542617559433, -0.2894585132598877, -0.020572105422616005, -0.015657247975468636, 0.22764621675014496, -0.1329174041748047, -0.40480247139930725, -0.054158762097358704, 0.14537955820560455, 0.04365754872560501, 0.33607935905456543, -0.48687177896499634, -0.2018183469772339, -0.16914990544319153, 0.1019672229886055, 0.1467081606388092, 0.1432410031557083, -0.25145378708839417, 0.1305127888917923, 0.3261987268924713, -0.34623968601226807, 0.1161704808473587, 0.16987422108650208, 0.022485069930553436, 0.19154703617095947, 0.16412422060966492, 0.06718472391366959, 0.14418566226959229, 0.11707310378551483, -0.14140509068965912, -0.05994516611099243, 0.16280727088451385, 0.18861621618270874, -0.13985148072242737, 0.15012964606285095, -0.1367889642715454, -0.16258308291435242, -0.11337688565254211, -0.07932806015014648, -0.100755475461483, -0.2714381814002991, -0.12925609946250916, 0.07303940504789352, 0.2117285579442978, 0.22056759893894196, -0.08022291213274002, -0.1748630702495575, -0.10348450392484665, 0.16781757771968842, 0.06923595070838928, 0.10698181390762329, 0.3573373556137085, 0.20533929765224457, 0.07064777612686157, 0.3598752021789551, 0.41936129331588745, -0.00026047229766845703, 0.4682632088661194, 0.034650932997465134, -0.08014567941427231, -0.4715932011604309, 0.26750680804252625, 0.010570377111434937, 0.19505776464939117, 0.08407611399888992, 0.34856468439102173, -0.2840741276741028, -0.07184205204248428, 0.06369417160749435, 0.2873607277870178, 0.16066107153892517, -0.029453180730342865, -0.07150281965732574, 0.08202341943979263, 0.019564740359783173, -0.2752048671245575, 0.02920675277709961, -0.09377898275852203, 0.2505137324333191, 0.04363768547773361, 0.3715869188308716, -0.15463951230049133, -0.18234413862228394, -0.42075103521347046, 0.31654927134513855, -0.2892088294029236, -0.28602924942970276, 0.2699078917503357, -0.3006247580051422, 0.2749161124229431, 0.14489153027534485, -0.04462624713778496, 0.1926157921552658, 0.7117822170257568, 0.060089126229286194, -0.027329452335834503, -0.15319854021072388, -0.31560954451560974, 0.2999636232852936, 0.3962077498435974, -0.10624181479215622, 0.34747663140296936, 0.06683951616287231, -0.34605440497398376, -0.15965640544891357, 0.1851087510585785, 0.3952050507068634, 0.2746729254722595, -0.28492045402526855, 0.16531166434288025, 0.31191620230674744, 0.0437198132276535, -0.14279255270957947, 0.28440606594085693, 0.46444469690322876, -0.275371253490448, 0.10013259947299957, 0.07736312597990036, -0.16198547184467316, 0.33745381236076355, 0.09224112331867218, 0.262637734413147, -0.2047155797481537, 0.5761634707450867, -0.055060602724552155, -0.07937566936016083, 0.008969021961092949, 0.20302975177764893, -0.5066409111022949, -0.1221209466457367, 0.06195889785885811, 0.057623136788606644, 0.1631636917591095, -0.22196754813194275, 0.09016378223896027, 0.10871443152427673, 0.3901248276233673, 0.1252451092004776, 0.13043203949928284, -0.2856214940547943, -0.016744092106819153, -0.5698694586753845, 0.33293038606643677, -0.38900381326675415, 0.24580830335617065, -0.18367385864257812, 0.27481552958488464, -0.26683565974235535, 0.0394250825047493, 0.14697884023189545, -0.2107890099287033, 0.18485155701637268, 0.10188579559326172, -0.45321252942085266, 0.025243131443858147, 0.07228796184062958, -0.19550500810146332, 0.2549091577529907, -0.23168110847473145, 0.021740082651376724, -0.019082754850387573, 0.13105076551437378, 0.003731008619070053, -0.003355499356985092, 0.2672176957130432, 0.5061380863189697, 0.2889290750026703, 0.4251261055469513, 0.20019441843032837, 0.040501877665519714, -0.18851475417613983, -0.3905116319656372, 0.03260090947151184, -0.03158155456185341, -0.24561229348182678, -0.06935802102088928, 0.3341864347457886, -0.31259843707084656, -0.1130521297454834, -0.28429922461509705, 0.22645722329616547, 0.08093833923339844, -0.39334407448768616, 0.1780988872051239, -0.03644590079784393, -0.12942680716514587, 0.06052616238594055, 0.030793573707342148, 0.7254053354263306, 0.016145382076501846, 0.4419616758823395, -0.18536314368247986, -0.2571273148059845, 0.381931334733963, -0.6724093556404114, -0.2614430785179138, 0.003796197474002838, 0.3030821681022644, 0.09342258423566818, -0.17738240957260132, -0.530910849571228, -0.12437977641820908, 0.17773990333080292, -0.16756942868232727, -0.12121240794658661, 0.18059678375720978, -0.19370520114898682, -0.035032834857702255, -0.010602939873933792, 0.16422662138938904, 0.0944303572177887, -0.26849400997161865, 0.27279290556907654, -0.2511070966720581 ]
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, -0.3640769124031067, -0.061697736382484436, 0.13931091129779816, -0.06844886392354965, -0.07843451201915741, 0.34460073709487915, 0.2630481719970703, 0.37012946605682373, 0.19107209146022797, -0.27151840925216675, 0.684173583984375, -0.06559266149997711, 0.11889613419771194, 0.1820763349533081, -0.112140953540802, -0.10166793316602707, 0.3471294045448303, -0.15045125782489777, -0.12176007032394409, -0.26120519638061523, 0.37858524918556213, -0.25765156745910645, 0.14147379994392395, -0.179623082280159, -0.10173977166414261, -0.15014256536960602, 0.07673320919275284, -0.0330740362405777, -0.40146899223327637, 0.10599252581596375, 0.14746442437171936, -0.10761848837137222, -0.12033400684595108, -0.0001057189583661966, -0.28695642948150635, 0.2894555628299713, 0.0716744065284729, 0.0324568934738636, 0.12928733229637146, -0.14674878120422363, -0.29723060131073, -0.03970751166343689, -0.38646867871284485, -0.05714061111211777, -0.0009069684892892838, 0.045928142964839935, -0.1283685564994812, 0.16652294993400574, 0.16479021310806274, 0.24084722995758057, 0.1582946479320526, 0.08807916939258575, 0.13617266714572906, 0.4565761387348175, -0.47681695222854614, 0.04431554675102234, 0.2460300773382187, 0.09880662709474564, -0.34286826848983765, -0.016030214726924896, 0.45544931292533875, 0.12964686751365662, 0.12535879015922546, 0.03968103229999542, 0.14423763751983643, 0.03318917006254196, -0.35992300510406494, -0.08387000858783722, 0.42646485567092896, 0.1315077543258667, -0.35245177149772644, -0.31000691652297974, -0.20165613293647766, 0.21391212940216064, -0.31251105666160583, -0.06473945826292038, 0.14203594624996185, -0.016297195106744766, 0.009707011282444, 0.03766291216015816, 0.10280214250087738, -0.12196803838014603, 0.13918480277061462, 0.1389048546552658, 0.58426833152771, -0.020311597734689713, 0.06995426118373871, 0.369263619184494, -0.029156841337680817, -0.1850247085094452, 0.09218508005142212, -0.020080607384443283, 0.17599333822727203, -0.28078657388687134, -0.1972288191318512, 0.10954360663890839, 0.10077453404664993, 0.021074973046779633, 0.04459695890545845, 0.2250237762928009, 0.03807013854384422, 0.20733889937400818, 0.164203479886055, 0.0804491639137268, 0.28569498658180237, 0.3848165273666382, 0.2837710678577423, 0.028996579349040985, 0.06244755908846855, -0.12825216352939606, 0.208902508020401, -0.014154279604554176, -0.10459042340517044, 0.2733452618122101, -0.09539482742547989, -0.07485391199588776, -0.12386596202850342, -0.31944164633750916, -0.025793492794036865, -0.2851245403289795, 0.002068644855171442, -0.179657980799675, 0.24445262551307678, 0.02786293253302574, 0.015347061678767204, -0.034648630768060684, 0.09597478061914444, -0.15768174827098846, -0.1615324318408966, -0.29697129130363464, 0.37302109599113464, -0.2329164296388626, -0.25091123580932617, 0.14201562106609344, 0.2079273760318756, 0.28334131836891174, 0.02848684787750244, -0.13069741427898407, 0.03727928549051285, 0.14972732961177826, -0.020442385226488113, 0.2775792181491852, 0.1166960820555687, -0.06863987445831299, 0.10494784265756607, -0.010465957224369049, -0.3798036575317383, -0.27660971879959106, -0.0801539346575737, 0.09214653819799423, 0.07429817318916321, 0.023173747584223747, 0.2386457622051239, -0.17580591142177582, 0.027032241225242615, 0.18054483830928802, 0.09859307110309601, -0.09635759145021439, 0.2063305526971817, 0.14499212801456451, -0.14029499888420105, -0.11530124396085739, -0.189780130982399, 0.29160141944885254, 0.20183679461479187, -0.30105796456336975, -0.11395061016082764, 0.08180488646030426, -0.035246532410383224, -0.12665051221847534, 0.05246239900588989, 0.22159940004348755, 0.005408667027950287, -0.26821568608283997, 0.22214002907276154, 0.2920093238353729, -0.5699788331985474, -0.4601467549800873, 0.026117650792002678, -0.09821595996618271, -0.11541444063186646, 0.08403828740119934, 0.07993172854185104, -0.02733754739165306, 0.20734506845474243, 0.3057432174682617, 0.2759208679199219, 0.1464020311832428, -0.01405496895313263, -0.33543702960014343, 0.14553837478160858, 0.3818058371543884, 0.14466899633407593, -0.172483429312706, -0.1629219949245453, -0.20003029704093933, -0.16491568088531494, 0.26382991671562195, 0.11475236713886261, -0.08811499178409576, 0.26813000440597534, 0.012011706829071045, -0.35355842113494873, 0.12525106966495514, -0.14601200819015503, -0.23448625206947327, 0.24857288599014282, -0.23344555497169495, 0.39037951827049255, 0.2793833911418915, -0.13543885946273804, -0.39104709029197693, -0.009539967402815819, -0.021591123193502426, -0.22261112928390503, 0.27640199661254883, 0.022156011313199997, 0.3353922665119171, 0.06124608591198921, -0.09688166528940201, 0.3609965443611145, 0.09802190959453583, 0.04160770773887634, -0.36970967054367065, 0.09687409549951553, -0.059088364243507385, -0.0400339812040329, 0.2472638189792633, 0.5104691386222839, 0.2441360056400299, -0.053404368460178375, 0.13028821349143982, 0.3582194745540619, -0.023660508915781975, 0.012895934283733368, 0.08358973264694214, -0.017382413148880005, -0.042964912950992584, -0.29577168822288513, 0.09626911580562592, 0.167236328125, 0.08046390861272812, -0.04020093381404877, -0.15581396222114563, 0.3673761188983917, 0.09383880347013474, 0.051863525062799454, 0.11298194527626038, 0.09805752336978912, -0.09670708328485489, -0.29650333523750305, -0.3145450949668884, -0.13945330679416656, 0.33484065532684326, -0.22544245421886444, 0.35534876585006714, 0.0405212827026844, -0.2502377927303314, 0.20897047221660614, 0.4467064142227173, 0.02467864751815796, 0.20379221439361572, -0.11887644976377487, -0.28534436225891113, -0.10063870996236801, 0.05481969565153122, 0.521816074848175, 0.16740722954273224, 0.4208611845970154, 0.017650090157985687, -0.059704579412937164, -0.39394035935401917, -0.073459692299366, 0.1677398681640625, -0.08496889472007751, 0.07969314604997635, -0.08924082666635513, 0.20133481919765472, -0.14826007187366486, -0.18963688611984253, 0.055628422647714615, -0.24803262948989868, 0.23431752622127533, -0.262809693813324, -0.1954481601715088, -0.08484265208244324, -0.49717023968696594, -0.2795579731464386, -0.2954830527305603, -0.14680330455303192, -0.5272935032844543, 0.1358085572719574, -0.2413761019706726, -0.020685074850916862, 0.2252945899963379, 0.05436110123991966, 0.16111983358860016, -0.1830400824546814, 0.07843844592571259, -0.20593924820423126, -0.13555596768856049, -0.21072593331336975, 0.1599218249320984, -0.2740514576435089, 0.5203291773796082, 0.3989133834838867, -0.09670913964509964, 0.07538843154907227, 0.319004088640213, -0.21436789631843567, -0.0005631670355796814, -0.1737833470106125, 0.24654264748096466, 0.41369199752807617, 0.001874573528766632, -0.1277114450931549, -0.012992790900170803, 0.24294744431972504, -0.4388088583946228, -0.005683007650077343, -0.06721718609333038, 0.20725557208061218, -0.03914029896259308, -0.14822210371494293, -0.4048267602920532, -0.3648366630077362, -0.22121040523052216, 0.23954297602176666, 0.2358979433774948, 0.2596205472946167, 0.11519569158554077, -0.33117005228996277, 0.32700514793395996, -0.16804878413677216, -0.03360198438167572, -0.15168629586696625, -0.05473499745130539, 0.24902060627937317, -0.10521747171878815, -0.29829567670822144, -0.24963794648647308, -0.1970977634191513, 0.22305834293365479, -0.29266560077667236, -0.21160583198070526, -0.17701636254787445, -0.07415030896663666, -0.140194371342659, 0.027845121920108795, 0.22281573712825775, 0.11768673360347748, 0.032634586095809937, -0.15178067982196808, -0.14525485038757324, -0.24856722354888916, 0.33657920360565186, 0.05205577611923218, -0.14087173342704773, 0.03853772208094597, 0.035232655704021454, 0.0140007846057415, 0.17996056377887726, 0.11784464120864868, -0.12312307953834534, 0.19788864254951477, -0.27885881066322327, 0.2952500879764557, 0.23055246472358704, -0.2540300190448761, 0.010372409597039223, 0.31672489643096924, 0.2669713497161865, 0.13920292258262634, 0.11716970056295395, -0.32901597023010254, -0.0998944491147995, 0.11832112073898315, -0.2628355324268341, -0.05908047407865524, 0.03934847563505173, -0.1033320426940918, -0.21532243490219116, 0.020679865032434464, -0.03607349470257759, -0.10807126760482788, -0.1725042164325714, -0.042756009846925735, -0.39986807107925415, -0.08388235419988632, -0.0383346863090992, -0.6121565103530884, -0.12144096195697784, -0.519475519657135, 0.15431669354438782, 0.04839468374848366, 0.47348707914352417, -0.010718386620283127, -0.09485802054405212, -0.0313771516084671, 0.030421389266848564, 0.5715874433517456, -0.41469699144363403, 0.2131909281015396, 0.4603511393070221, 0.18213927745819092, -0.6949966549873352, 0.0056409090757369995, -0.5489997267723083, -0.1169113740324974, 0.628441572189331, 0.26226311922073364, -0.2489963173866272, -0.15028288960456848, 0.27377182245254517, 0.01633814349770546, -0.09192781150341034, -0.1665167510509491, -0.29358014464378357, -0.2230232059955597, -0.03679610416293144, 0.14788085222244263, -0.009105484001338482, 0.2636672258377075, 0.1972573697566986, -0.07814471423625946, 0.05830570310354233, -0.2125874161720276, 0.14469707012176514, 0.22399543225765228, 0.29998311400413513, 0.0681437999010086, -0.03851299360394478, 0.11713027954101562, 0.21222658455371857, -0.6786831021308899, 0.12109183520078659, -0.520142138004303, 0.3546111583709717, -0.21757519245147705, -0.2540557086467743, 0.46517059206962585, 0.16192726790905, 0.04661724343895912, -0.1614825427532196, -0.5788977146148682, 0.21016110479831696, -0.04784177243709564, 0.020870104432106018, 0.20848537981510162, 0.16397525370121002, -0.23401446640491486, -0.07566914707422256, 0.5497063994407654, 0.06847737729549408, -0.22469213604927063, 0.33281785249710083, 0.19439521431922913, -0.3568422496318817, 0.41191208362579346, -0.08889146894216537, 0.9092602133750916, -0.0902237594127655, -0.007592473179101944, 0.25458449125289917, -0.36367470026016235, 0.6806232929229736, 0.14083079993724823, 0.12344540655612946, -0.5268474817276001, -0.3948061466217041, 0.06874201446771622, -0.08320694416761398, -0.1278975009918213, 0.27551499009132385, -0.22262030839920044, 0.41558215022087097, 0.07301349937915802, -0.19334717094898224, 0.022332414984703064, 0.3862771689891815, 0.0031906133517622948, -0.10484971106052399, -0.34154435992240906, 0.15480966866016388, -0.03297057002782822, 0.1674894392490387, -0.061462417244911194, -0.019678330048918724, -0.1323699802160263, -0.21145224571228027, -0.0483417809009552, 0.19770094752311707, -0.2182677984237671, -0.16570232808589935, 0.3732075095176697, -0.030512623488903046, -0.4524395167827606, 0.13093191385269165, 0.18288502097129822, 0.12453879415988922, 0.006406888365745544, 0.05262470245361328, 0.1765984296798706, 0.0018503903411328793, 0.5183125734329224, 0.2572227120399475, 0.29605740308761597, -0.10177106410264969, -0.057917073369026184, -0.10081515461206436, 0.08597951382398605, -0.1053326278924942, -0.06399112939834595, 0.058669161051511765, -0.05259709805250168, -0.14442995190620422, -0.34001415967941284, 0.06040734052658081, -0.013500288128852844, -0.1976650208234787, 0.19822053611278534, 0.09399652481079102, -0.41108354926109314, 0.27004361152648926, 0.09358696639537811, -0.3252710700035095, -0.03686972334980965, 0.3700680732727051, -0.0598650723695755, 0.4367310106754303, 0.17269816994667053, 0.02936570718884468, -0.016895389184355736, -0.28269121050834656, -0.06539274007081985, 0.3214573860168457, -0.22965207695960999, 0.0666491910815239, -0.12342483550310135, -0.10889684408903122, -0.02172115258872509, 0.0352230966091156, 0.19642804563045502, 0.17936421930789948, 0.09442651271820068, -0.46983906626701355, -0.12708443403244019, 0.16168776154518127, 0.06652915477752686, 0.3171401023864746, 0.0377875417470932, -0.09414584934711456, 0.2179001122713089, 0.35001426935195923, -0.4358421862125397, -0.02367055043578148, -0.20669178664684296, -0.07812841981649399, 0.4145662784576416, -0.023621082305908203, 0.07423576712608337, 0.05426991358399391, 0.31571969389915466, 0.07997889816761017, -0.09340900182723999, -0.3097279965877533, -0.18688155710697174, 0.01974952220916748, -0.07820963859558105, -0.011730499565601349, 0.02576456591486931, -0.12001354992389679, -0.07672470808029175, -0.06595675647258759, -0.09149377048015594, -0.1465875208377838, 0.007814858108758926, 0.44732120633125305, 0.00041973963379859924, 0.09699159860610962, 0.06402929127216339, 0.11668185889720917, -0.02109866589307785, -0.2281688004732132, 0.03219006955623627, 0.1621513068675995, -0.051699306815862656, -0.09866438060998917, -0.15055562555789948, -0.04716537147760391, -0.3882819414138794, -0.06876557320356369, 0.006776086986064911, -0.1605016142129898, -0.09582380950450897, 0.07739389687776566, 0.6340235471725464, 0.010784368962049484, -0.29655274748802185, -0.19216448068618774, 0.02798072062432766, 0.3456474244594574, -0.3949342370033264, -0.12690258026123047, 0.45424744486808777, 0.24674122035503387, 0.26460975408554077, 0.3529655337333679, 0.1391114592552185, -0.08405223488807678, -0.1915549784898758, 0.004000389948487282, 0.29133230447769165, -0.4078513979911804, -0.16344201564788818, 0.17371073365211487, -0.14265118539333344, -0.16981899738311768, 0.32067832350730896, 0.057864993810653687, 0.16803519427776337, 0.0838930606842041, 0.3906581699848175, 0.5761868953704834, 0.1416967213153839, 0.03042634017765522, -0.057727061212062836, 0.11515700817108154, -0.07071809470653534, 0.5745166540145874, -0.027649804949760437, 0.2495676577091217, 0.15678000450134277, 0.5897642374038696, -0.2613559365272522, -0.21765758097171783, -0.04092089459300041, 0.21259534358978271, -0.03424641489982605, -0.21949529647827148, 0.13715526461601257, -0.07145743072032928, -0.3883407711982727, -0.004371333867311478, 0.014580883085727692, -0.23136065900325775, 0.3803726136684418, -0.04845938831567764, 0.12761947512626648, -0.2602132260799408, -0.11909803003072739, 0.024867886677384377, 0.4030858874320984, -0.39799317717552185, 0.06432206183671951, -0.25062912702560425, 0.09554294496774673, -0.017598174512386322, 0.3547402024269104, 0.1716810166835785, 0.01775120571255684, -0.251244455575943, -0.08426274359226227, -0.1949606090784073, -0.11999738961458206, -0.20278146862983704, 0.22799235582351685, 0.24091899394989014, 0.018384048715233803, 0.4225250482559204, 0.1816043108701706, -0.22556337714195251, 0.04913727194070816, 0.16723473370075226, 0.09844847023487091, -0.007039850112050772, 0.4931473731994629, -0.10184995830059052, -0.2557195723056793, -0.2666807770729065, -0.051310501992702484, -0.40269023180007935, -0.06665518134832382, -0.11750084161758423, -0.03085355833172798, 0.1739559769630432, -0.3835698366165161, 0.14161598682403564, -0.04973557963967323, 0.44530877470970154, 0.012515820562839508, 0.2989167273044586, -0.2112409472465515, -0.21040134131908417, -0.6262401342391968, -0.02299164980649948, -0.09374164044857025, -0.5266145467758179, 0.21520569920539856, 0.029598353430628777, -0.06426141411066055, 0.020876767113804817, 0.042776159942150116, 0.5819962620735168, -0.1631898134946823, -0.10652980208396912, -0.42306703329086304, 0.11542114615440369, -0.18298102915287018, -0.05020345002412796, 0.14122669398784637, -0.29087498784065247, 0.011113103479146957, -0.23848843574523926, 0.19092661142349243, 0.07287663221359253, 0.025542572140693665, 0.06769707798957825, 0.27018457651138306, 0.1387310028076172, 0.04367171972990036, 0.26167649030685425, -0.14459040760993958, 0.08573254197835922, -0.1028868705034256, -0.016846153885126114, -0.2982347309589386, 0.03944222629070282, 0.10759301483631134, 0.33385246992111206, 0.04120609164237976, 0.021992461755871773, -0.072879359126091, 0.09334103018045425, 0.1682973951101303, 0.2568713128566742, 0.03382617235183716, 0.19437940418720245, 0.14239341020584106, -0.041015639901161194, 0.09539620578289032, 0.3407485783100128, 0.00537433847784996, 0.055003829300403595, -0.012975223362445831, -0.3489196300506592, 0.4604702591896057, -0.32872405648231506, -0.1328895092010498, 0.0940818265080452, 0.3477376103401184, 0.26250338554382324, -0.24081814289093018, -0.5395858287811279, -0.1369069516658783, 0.3368121087551117, 0.042628079652786255, -0.29551711678504944, 0.3126370310783386, -0.04802190884947777, 0.12107129395008087, 0.054204311221838, -0.015983998775482178, 0.2987123727798462, -0.2579358220100403, -0.10434575378894806, -0.25383806228637695 ]
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.
[ 0.13189557194709778, -0.12678326666355133, 0.05668975040316582, -0.02853645384311676, 0.11699807643890381, -0.15414652228355408, 0.39699190855026245, -0.062105778604745865, -0.2148279994726181, 0.2518129348754883, -0.08670718222856522, 0.1588992327451706, -0.17179957032203674, 0.2942127287387848, 0.3824805021286011, 0.1044841781258583, 0.09072622656822205, 0.31321364641189575, 0.039623599499464035, -0.1279827356338501, -0.2627106010913849, 0.05864565074443817, -0.0162038616836071, -0.07316631078720093, -0.0752854198217392, -0.186767116189003, -0.13502033054828644, 0.2678755223751068, -0.22584228217601776, -0.3100969195365906, 0.4256783425807953, 0.20961831510066986, 0.34658485651016235, 0.5284520983695984, -0.00012068286014255136, -0.06405109167098999, 0.17346450686454773, -0.24845093488693237, -0.5683867335319519, -0.3589860796928406, 0.08358649164438248, -0.13413137197494507, 0.08103309571743011, -0.286552757024765, 0.22721758484840393, 0.05291748046875, 0.050873033702373505, 0.13236527144908905, 0.5366085171699524, 0.40950727462768555, 0.12447778880596161, 0.06690997630357742, -0.21695269644260406, -0.10151143372058868, 0.008197188377380371, 0.16872510313987732, -0.03125625476241112, 0.29346412420272827, 0.31920769810676575, -0.4081132709980011, 0.29175901412963867, -0.24268227815628052, -0.03904249891638756, -0.007286045700311661, 0.5370670557022095, 0.05455385893583298, -0.05379808694124222, -0.4257262647151947, 0.004693884402513504, 0.15714409947395325, 0.28184232115745544, 0.03364906460046768, -0.32898184657096863, -0.1129494309425354, 0.11158498376607895, -0.5599852204322815, 0.20915251970291138, 0.30534711480140686, -0.425655335187912, -0.03852710500359535, -0.2067476511001587, 0.33666813373565674, -0.3919665515422821, 0.47593575716018677, 0.11568908393383026, 0.3350135385990143, -0.037230268120765686, 0.1301805078983307, 0.2861829102039337, -0.09340396523475647, -0.009625946171581745, 0.07325896620750427, 0.025191180408000946, 0.29447171092033386, -0.1664421558380127, -0.18143440783023834, -0.030567705631256104, -0.2799895703792572, -0.11502810567617416, 0.09093773365020752, 0.4687071442604065, -0.18926206231117249, -0.21843667328357697, 0.1630101203918457, 0.11493547260761261, 0.3294888436794281, -0.02343813329935074, 0.05227089673280716, 0.10152753442525864, -0.0250001922249794, 0.31006181240081787, -0.08968452364206314, -0.2538239657878876, -0.10262657701969147, -0.1475948989391327, -0.023878667503595352, 0.27968502044677734, 0.12876714766025543, 0.012593811377882957, -0.1362917274236679, 0.055091097950935364, 0.06330811977386475, -0.16161300241947174, 0.3230098783969879, -0.08835256844758987, -0.2684031128883362, 0.0020211413502693176, 0.27830955386161804, 0.16045379638671875, -0.29687362909317017, 0.1016654521226883, 0.33483922481536865, -0.10520203411579132, -0.15593469142913818, 0.38596123456954956, 0.026123058050870895, 0.18523725867271423, -0.05975525453686714, 0.22655539214611053, -0.14647191762924194, 0.022025955840945244, -0.03487211465835571, -0.16940507292747498, 0.01491536945104599, 0.23729723691940308, 0.3928617835044861, -0.03909881040453911, -0.5896609425544739, -0.11377360671758652, 0.09599334746599197, -0.4078333377838135, -0.3410404324531555, -0.45826658606529236, 0.05670671910047531, 0.047931917011737823, -0.18805880844593048, -0.35080015659332275, -0.13403929769992828, 0.10075756162405014, -0.2789129614830017, -0.09075621515512466, 0.01032448373734951, -0.08747869729995728, -0.1320457011461258, 0.06136983633041382, 0.04111156612634659, -0.5356754660606384, 0.26288190484046936, -0.08810169249773026, 0.5520693063735962, 0.3589295446872711, 0.29697832465171814, -0.31327173113822937, 0.3023737967014313, -0.043867308646440506, 0.17262260615825653, 0.5356279015541077, -0.4726383090019226, -0.22461479902267456, 0.007903169840574265, -0.036810435354709625, -0.1761104315519333, 0.11352550983428955, 0.19710120558738708, 0.12219541519880295, -0.06313978135585785, 0.09179738909006119, 0.38595670461654663, 0.050373680889606476, 0.018724409863352776, -0.1974925547838211, -0.16739946603775024, 0.2501932382583618, -0.026913408190011978, -0.168757826089859, 0.11825580894947052, 0.26363605260849, 0.028966180980205536, 0.030211206525564194, -0.3133379817008972, 0.1195906326174736, 0.21816334128379822, 0.336930513381958, -0.18201304972171783, -0.1540909856557846, 0.2823061943054199, -0.40281227231025696, 0.16599223017692566, 0.1726563572883606, 0.34748575091362, -0.055268168449401855, -0.04414612054824829, -0.4338451027870178, -0.11113899946212769, -0.020370300859212875, -0.02535172551870346, 0.045775607228279114, 0.1155954971909523, 0.02293524518609047, -0.23246537148952484, -0.2718803584575653, 0.10688915103673935, -0.26044517755508423, 0.04938431829214096, -0.21144266426563263, -0.039563290774822235, -0.3254668414592743, -0.0634855404496193, 0.25599977374076843, 0.11972060054540634, 0.20380091667175293, -0.012565569020807743, -0.1673872470855713, 0.10784434527158737, -0.42461252212524414, 0.2669956088066101, -0.030855342745780945, 0.11985543370246887, 0.007746268063783646, -0.006525103002786636, -0.25970226526260376, -0.07866546511650085, 0.22222746908664703, -0.20514318346977234, -0.38404104113578796, 0.16649064421653748, -0.00962015613913536, 0.0653088241815567, 0.12813536822795868, 0.00844607688486576, 0.2979525625705719, -0.18777604401111603, 0.06347161531448364, 0.1433505415916443, 0.3599367141723633, 0.11930444836616516, 0.380616158246994, -0.02155371569097042, -0.12040473520755768, -0.15498527884483337, 0.16857118904590607, 0.010845199227333069, 0.13222718238830566, 0.058391839265823364, 0.06193959712982178, -0.07862596213817596, 0.32469508051872253, 0.5969869494438171, 0.384693443775177, 0.035543277859687805, -0.06499974429607391, -0.008059078827500343, -0.10091213881969452, -0.23450496792793274, -0.025495924055576324, -0.04307097941637039, 0.042327675968408585, 0.16191382706165314, 0.3972325325012207, 0.22959718108177185, -0.09001210331916809, 0.027445081621408463, 0.13210558891296387, 0.42635661363601685, -0.2129417210817337, 0.2630130648612976, -0.10497573018074036, -0.41205835342407227, 0.28124183416366577, -0.24620558321475983, -0.06726276874542236, -0.15471282601356506, -0.1818927675485611, 0.13285918533802032, 0.10234750807285309, -0.14528951048851013, 0.15610329806804657, 0.3768150210380554, -0.11184241622686386, -0.25870564579963684, -0.23002561926841736, -0.09443660825490952, -0.20480453968048096, 0.022891487926244736, 0.28228190541267395, -0.07962040603160858, 0.3960428535938263, 0.29778948426246643, 0.05912502110004425, 0.0036392370238900185, -0.30800023674964905, -0.018483616411685944, -0.044878169894218445, 0.10120940953493118, 0.09834799915552139, 0.3818156123161316, 0.0118325874209404, -0.528057873249054, 0.14642761647701263, 0.07743369042873383, -0.27657240629196167, 0.24847862124443054, -0.021304728463292122, -0.35785606503486633, -0.48331716656684875, -0.025970757007598877, -0.23736806213855743, -0.3723588287830353, 0.23394285142421722, 0.34234634041786194, 0.20104345679283142, 0.6263754963874817, 0.033080752938985825, 0.10552358627319336, 0.36732491850852966, 0.15832485258579254, 0.0377814918756485, -0.14254775643348694, 0.48901185393333435, -0.21970714628696442, -0.41547125577926636, 0.14479434490203857, -0.15794141590595245, 0.5447531342506409, -0.07129579782485962, -0.35194599628448486, -0.1823221743106842, 0.08118394762277603, 0.14644116163253784, -0.03778723627328873, -0.03920012712478638, 0.3170751631259918, -0.2450891137123108, 0.06349536031484604, -0.07407765090465546, -0.3456256091594696, 0.11119922995567322, 0.6317628026008606, 0.26698553562164307, 0.001901073381304741, 0.46685704588890076, 0.0014615021646022797, 0.7395951151847839, 0.07472380995750427, -0.2926449179649353, 0.4109627306461334, -0.1004679724574089, 0.25575876235961914, -0.17619848251342773, -0.3271777331829071, 0.02762286178767681, 0.18094253540039062, -0.12938284873962402, 0.1223585233092308, -0.056772276759147644, -0.07398992776870728, -0.4756854772567749, 0.30633044242858887, -0.2941302955150604, 0.049686674028635025, 0.010089896619319916, 0.10700774192810059, -0.009691186249256134, -0.0785011351108551, -0.2286146879196167, -0.33661964535713196, -0.48897460103034973, -0.18894588947296143, 0.2726052403450012, -0.21989968419075012, 0.17207551002502441, -0.6565858721733093, -0.13269878923892975, -0.6450006365776062, 0.30821236968040466, -0.1679040789604187, 0.3023477792739868, -0.18335309624671936, 0.18022412061691284, 0.17021030187606812, -0.06942994892597198, 0.14790436625480652, -0.3889729082584381, -0.20179714262485504, -0.1040051281452179, 0.3181891143321991, -0.4269338548183441, -0.06536213308572769, -0.07279998809099197, 0.0820639505982399, 0.48182615637779236, 0.20355457067489624, -0.014485858380794525, 0.0970020741224289, -0.19211557507514954, -0.1825750172138214, -0.20688699185848236, -0.21783868968486786, -0.08103419840335846, -0.6945075988769531, -0.36786046624183655, -0.501621425151825, 0.12273717671632767, 0.1390097588300705, 0.14689327776432037, -0.03024246171116829, 0.2109568566083908, 0.08789141476154327, 0.4853430390357971, 0.004931243136525154, -0.0599362850189209, 0.6008954644203186, -0.1651570200920105, -0.4717911183834076, 0.49630558490753174, -0.11771067976951599, 0.3781425654888153, 0.09147711098194122, -0.15693682432174683, -0.13820087909698486, -0.32531216740608215, 0.4744024872779846, 0.23392300307750702, 0.1031668409705162, 0.28175362944602966, -0.0628662034869194, 0.02600932866334915, -0.15068195760250092, 0.4732237756252289, 0.34279608726501465, -0.1364433914422989, -0.3455304205417633, -0.14745070040225983, 0.48506808280944824, -0.2086663395166397, -0.06152798607945442, 0.22016724944114685, -0.20480971038341522, -0.2317996472120285, 0.21763405203819275, 0.1648198515176773, 1.2438884973526, -0.05959595739841461, 0.4256584346294403, 0.07456724345684052, -0.03276018798351288, 0.7743026614189148, -0.1663348376750946, 0.06815848499536514, -0.23189301788806915, 0.1349938064813614, -0.14140063524246216, -0.06125308573246002, -0.11390981823205948, 0.1102670356631279, -0.22976121306419373, 0.1084144115447998, -0.11354091763496399, -0.04289843514561653, 0.0006191423162817955, 0.4512530565261841, 0.10172717273235321, -0.34147435426712036, -0.17380934953689575, 0.061344318091869354, -0.2538866400718689, 0.31335511803627014, -0.17983421683311462, -0.21803924441337585, -0.3014184534549713, -0.23261982202529907, -0.1817784309387207, 0.24452084302902222, -0.26989036798477173, 0.07795178890228271, 0.0035698264837265015, -0.32298940420150757, 0.17310534417629242, 0.4630565643310547, 0.5105307102203369, -0.06357783079147339, -0.06871569901704788, -0.009666115045547485, -0.0855763778090477, -0.042368341237306595, -0.09295277297496796, 0.06018524616956711, 0.13630156219005585, -0.29558858275413513, 0.07367678731679916, 0.3211551606655121, -0.08515811711549759, -0.37327325344085693, -0.21020714938640594, 0.19486314058303833, 0.05941276624798775, -0.1340414434671402, -0.3536511957645416, 0.06596338003873825, -0.21248839795589447, -0.01574072428047657, 0.004534289240837097, 0.033151160925626755, 0.08300954848527908, -0.09015639871358871, 0.08961185812950134, -0.19795191287994385, 0.09447424113750458, 0.06312532722949982, -0.04164844751358032, -0.04899197071790695, 0.700760006904602, -0.12110793590545654, 0.08024638146162033, -0.04312053322792053, -0.20529380440711975, -0.3344261050224304, -0.14639444649219513, -0.07787778973579407, 0.31912145018577576, -0.030446283519268036, 0.03915878385305405, 0.6615256667137146, 0.39813387393951416, 0.18721811473369598, 0.06805628538131714, -0.1111069917678833, -0.32315707206726074, -0.3133423924446106, -0.03441209718585014, 0.20288443565368652, -0.025104040279984474, 0.2131599634885788, 0.20983272790908813, 0.29661574959754944, -0.21475595235824585, -0.1815822720527649, -0.2030603140592575, 0.17764772474765778, 0.7618895769119263, -0.013123616576194763, -0.11110197752714157, -0.01039477065205574, -0.033159658312797546, 0.009593192487955093, -0.14806410670280457, -0.10335216671228409, -0.3793177306652069, 0.18479835987091064, 0.5306533575057983, -0.1196054145693779, -0.014834612607955933, -0.23938070237636566, -0.11199536919593811, -0.12136735022068024, -0.32977530360221863, 0.04165046662092209, -0.014645189046859741, 0.20553617179393768, -0.3102738857269287, 0.3925665318965912, -0.1536860167980194, 0.1730085164308548, 0.020328618586063385, 0.2511637806892395, 0.06056476756930351, -0.052163735032081604, 0.15323297679424286, 0.0744905024766922, -0.39298906922340393, 0.006554029881954193, -0.17752008140087128, -0.09909840673208237, 0.025729719549417496, -0.4181019067764282, 0.15793122351169586, -0.22159242630004883, 0.022271227091550827, 0.04608551412820816, -0.5020123720169067, 0.2189575880765915, 0.4416903257369995, 0.08857264369726181, 0.04070451855659485, 0.16635048389434814, 0.07060461491346359, 0.21389895677566528, 0.024117764085531235, 0.2750561535358429, -0.0007002539932727814, -0.3031476140022278, -0.04899744689464569, -0.22703415155410767, -0.11426685750484467, -0.2721330523490906, 0.043390631675720215, -0.060880016535520554, -0.4242880344390869, 0.004609048366546631, 0.17042632400989532, 0.1590813249349594, -0.10351177304983139, 0.15583407878875732, -0.19883909821510315, 0.09525345265865326, 0.05639883130788803, -0.11588490009307861, 0.18386025726795197, -0.23433218896389008, 0.12833602726459503, -0.04518181085586548, -0.007981568574905396, 0.33807313442230225, 0.16224393248558044, 0.6163569688796997, -0.34986600279808044, -0.011332513764500618, -0.26698699593544006, 0.0580330416560173, 0.023866988718509674, 0.17970961332321167, -0.66041499376297, -0.005375497043132782, 0.0684242695569992, -0.08650712668895721, 0.03860250860452652, -0.3267977833747864, -0.05037975311279297, 0.2629709243774414, -0.1634490042924881, 0.10786832123994827, -0.1304248869419098, 0.19259408116340637, 0.13164880871772766, -0.11263614892959595, 0.3684447705745697, 0.003190934658050537, -0.03597176820039749, -0.06830348074436188, 0.05866541713476181, 0.5559294819831848, 0.6175879836082458, 0.16811425983905792, 0.27055463194847107, 0.23017127811908722, -0.0069065578281879425, 0.0774819478392601, 0.16673247516155243, 0.4616715610027313, 0.30198201537132263, 0.12015050649642944, 0.030812913551926613, -0.12133515626192093, 0.14843037724494934, 0.053318850696086884, 0.09564624726772308, -0.3158126771450043, 0.22804582118988037, -0.029106397181749344, -0.39975622296333313, -0.15640880167484283, 0.40171706676483154, -0.45790690183639526, -0.014089287258684635, 0.1900971531867981, 0.1823270320892334, 0.3674246668815613, -0.23387150466442108, 0.022305496037006378, -0.1849544197320938, 0.3975655138492584, 0.4656144678592682, -0.036127567291259766, -0.14645622670650482, -0.21958984434604645, -0.48715439438819885, 0.17760160565376282, -0.22294235229492188, -0.2007375806570053, -0.37588945031166077, 0.17404340207576752, 0.06222350895404816, -0.01866282895207405, 0.07831208407878876, 0.08619526028633118, -0.2741774618625641, 0.023239172995090485, -0.3622928857803345, -0.02059972658753395, -0.12250666320323944, -0.002512466162443161, -0.07682991772890091, -0.16656631231307983, 0.039569176733493805, 0.1100979670882225, 0.01287674531340599, 0.024147765710949898, 0.24240949749946594, 0.1370498687028885, 0.3673099875450134, 0.4786677956581116, -0.07187215238809586, 0.7685264348983765, 0.06704555451869965, 0.02057444490492344, 0.04930085316300392, -0.04225849360227585, -0.1771257072687149, -0.0018190070986747742, 0.17840667068958282, 0.37813428044319153, -0.058838121592998505, 0.34986403584480286, -0.19776242971420288, 0.06250464171171188, -0.022588256746530533, -0.5348160862922668, 0.05529928207397461, -0.006339428015053272, -0.08719682693481445, 0.23023784160614014, 0.42383113503456116, 0.08845043182373047, 0.07060220837593079, -0.1437549591064453, -0.43358665704727173, -0.2854061424732208, 0.4140363931655884, -0.5086387991905212, -0.30031147599220276, 0.2236095815896988, 0.23953387141227722, -0.02934328466653824, 0.05272877961397171, -0.22304341197013855, -0.0850505456328392, 0.135616272687912, -0.2508601248264313, -0.05188954994082451, 0.32419660687446594, -0.5186461210250854, 0.01627274975180626, -0.055287428200244904, 0.3049897849559784, 0.20241886377334595, -0.37965095043182373, -0.07080521434545517, -0.21507976949214935 ]
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' ```
[ 0.05260493978857994, -0.02810990810394287, 0.05533044412732124, -0.0020758435130119324, 0.14557740092277527, -0.18475821614265442, 0.33769673109054565, -0.04772265627980232, -0.19266782701015472, 0.1797221302986145, -0.1303306519985199, 0.24369043111801147, -0.21156124770641327, 0.2791500985622406, 0.3693567216396332, 0.06986403465270996, 0.013265587389469147, 0.28722038865089417, 0.0126219242811203, -0.1083281934261322, -0.25431978702545166, 0.09193598479032516, -0.01930972933769226, -0.08095788955688477, -0.12342613190412521, -0.19380062818527222, -0.08119546622037888, 0.2602364122867584, -0.22382165491580963, -0.36161258816719055, 0.491118848323822, 0.234791561961174, 0.36471027135849, 0.5793417096138, -0.00011945605365326628, -0.048161353915929794, 0.16207453608512878, -0.22125189006328583, -0.6242919564247131, -0.407953679561615, 0.020531892776489258, -0.151094451546669, 0.07041076570749283, -0.25698909163475037, 0.18926122784614563, 0.023185856640338898, 0.02170673757791519, 0.047051042318344116, 0.48964202404022217, 0.4066980481147766, 0.13775835931301117, 0.12786923348903656, -0.22636933624744415, -0.1397721767425537, 0.007812626659870148, 0.17320725321769714, -0.058755140751600266, 0.2595108449459076, 0.3348256051540375, -0.40046218037605286, 0.2734430432319641, -0.19609782099723816, 0.012751845642924309, -0.03790637105703354, 0.4887402653694153, -0.021754182875156403, -0.06072574108839035, -0.3897158205509186, -0.011177632957696915, 0.1542045772075653, 0.3593454658985138, -0.017493337392807007, -0.390593022108078, -0.12496654689311981, 0.13513121008872986, -0.4883081614971161, 0.24852344393730164, 0.2740962505340576, -0.40215224027633667, -0.027209419757127762, -0.2619933784008026, 0.3479141592979431, -0.4030788242816925, 0.44457709789276123, 0.10771401226520538, 0.37084731459617615, -0.03705957159399986, 0.09139961749315262, 0.2884838581085205, -0.15400105714797974, 0.022261077538132668, 0.038142286241054535, 0.028013093397021294, 0.2681345045566559, -0.21565796434879303, -0.1543159782886505, 0.011639691889286041, -0.23738548159599304, -0.08751183748245239, 0.11485089361667633, 0.42785075306892395, -0.1259317696094513, -0.24697113037109375, 0.14052718877792358, 0.10184241831302643, 0.3375077545642853, 0.023061472922563553, 0.1245047077536583, 0.13154982030391693, 0.0036603191401809454, 0.24482887983322144, -0.07746654003858566, -0.29436543583869934, -0.0886225551366806, -0.11200898885726929, -0.05211998522281647, 0.26885515451431274, 0.07469941675662994, -0.00757841020822525, -0.10847916454076767, 0.07862361520528793, 0.04231101647019386, -0.08323121815919876, 0.3526007831096649, -0.0594031997025013, -0.236591637134552, -0.020166635513305664, 0.2576301097869873, 0.1414344310760498, -0.2609562575817108, 0.06589063256978989, 0.29550161957740784, -0.12431095540523529, -0.14895284175872803, 0.4141158163547516, 0.023555435240268707, 0.19150908291339874, -0.06857933849096298, 0.21296444535255432, -0.11405246704816818, 0.05686989426612854, 0.029537595808506012, -0.164823979139328, 0.017245955765247345, 0.22356146574020386, 0.3928877115249634, -0.05233893170952797, -0.5404902100563049, -0.1403445154428482, 0.02345065027475357, -0.34771981835365295, -0.35778993368148804, -0.42381608486175537, 0.07509461790323257, 0.020578566938638687, -0.24025288224220276, -0.3375001549720764, -0.09723281860351562, 0.0914815217256546, -0.28204506635665894, -0.09507935494184494, -0.045188285410404205, -0.15627175569534302, -0.11636000871658325, 0.11207837611436844, 0.06296399980783463, -0.49924236536026, 0.21195179224014282, -0.07937294244766235, 0.5196447372436523, 0.27601712942123413, 0.3038299083709717, -0.3427072763442993, 0.23301392793655396, -0.06637683510780334, 0.18478022515773773, 0.5762104392051697, -0.49689874053001404, -0.31175607442855835, 0.046711888164281845, -0.03056374564766884, -0.11799081414937973, 0.07224185764789581, 0.17157939076423645, 0.11363708227872849, -0.08424391597509384, 0.10218079388141632, 0.43852120637893677, 0.06725264340639114, 0.04018282890319824, -0.22083240747451782, -0.15429767966270447, 0.2543807625770569, 0.03698397055268288, -0.15067903697490692, 0.08312258124351501, 0.26047420501708984, 0.06287041306495667, 0.01045648381114006, -0.32887956500053406, 0.14159715175628662, 0.21016967296600342, 0.35805806517601013, -0.1686781942844391, -0.12786760926246643, 0.21730640530586243, -0.4310542643070221, 0.21843907237052917, 0.1891905814409256, 0.34858307242393494, -0.13365566730499268, -0.03646417707204819, -0.43643516302108765, -0.10224456340074539, -0.1047709584236145, -0.04856086149811745, 0.04172849655151367, 0.12958484888076782, 0.07387520372867584, -0.18000899255275726, -0.22281411290168762, 0.15982894599437714, -0.2588525712490082, 0.0683702826499939, -0.21001917123794556, 0.03586208075284958, -0.2738802134990692, -0.09568630158901215, 0.2633073329925537, 0.167036771774292, 0.21565210819244385, -0.015994008630514145, -0.2017110288143158, 0.09329313039779663, -0.4534428119659424, 0.285800039768219, 0.017558999359607697, 0.10015590488910675, 0.036241769790649414, -0.055206049233675, -0.21381981670856476, -0.1098799854516983, 0.17156237363815308, -0.15249350666999817, -0.3677257001399994, 0.17456935346126556, -0.0044889748096466064, 0.03608394041657448, 0.1666366308927536, 0.010147880762815475, 0.29598456621170044, -0.1779492199420929, 0.09211891889572144, 0.0983789712190628, 0.4174831211566925, 0.09317228198051453, 0.40335720777511597, -0.011018628254532814, -0.1426611691713333, -0.08572553843259811, 0.16825538873672485, -0.0006488393992185593, 0.1505850851535797, 0.07491092383861542, 0.024227645248174667, -0.0512586385011673, 0.30666765570640564, 0.6624006628990173, 0.3770543932914734, 0.08588339388370514, -0.026365477591753006, -0.05196020007133484, -0.12508489191532135, -0.2106570303440094, -0.04003196209669113, -0.07790277898311615, 0.05889182165265083, 0.1829155683517456, 0.44176435470581055, 0.23638848960399628, -0.13617059588432312, -0.048412010073661804, 0.15327373147010803, 0.4534132778644562, -0.19562536478042603, 0.2583773136138916, -0.16094234585762024, -0.42994067072868347, 0.2424887865781784, -0.25551632046699524, -0.06268537044525146, -0.18169525265693665, -0.21708352863788605, 0.14613737165927887, 0.10717156529426575, -0.08416202664375305, 0.10070210695266724, 0.33156874775886536, -0.10592744499444962, -0.28663909435272217, -0.25440698862075806, -0.11374913901090622, -0.19180315732955933, 0.027764219790697098, 0.32500773668289185, -0.08840072154998779, 0.41247642040252686, 0.3248803913593292, 0.08864571154117584, -0.04613993316888809, -0.32683849334716797, -0.004596970975399017, -0.03740953654050827, 0.12089180946350098, 0.12138736248016357, 0.35637176036834717, 0.00866468995809555, -0.5445621609687805, 0.1948816180229187, 0.07846416532993317, -0.24362555146217346, 0.25166013836860657, -0.032538991421461105, -0.3485890030860901, -0.4676506221294403, -0.020827285945415497, -0.2737122178077698, -0.3687703311443329, 0.23967094719409943, 0.29409345984458923, 0.20991727709770203, 0.5754039287567139, 0.01771211251616478, 0.12528333067893982, 0.3206275403499603, 0.12391521036624908, 0.06834226846694946, -0.18604694306850433, 0.5255687236785889, -0.2455926537513733, -0.42524829506874084, 0.10256612300872803, -0.16349107027053833, 0.5601885318756104, -0.009718913584947586, -0.3501392602920532, -0.17843422293663025, 0.059479184448719025, 0.21830520033836365, -0.023407310247421265, -0.06744977831840515, 0.2789643406867981, -0.20497199892997742, 0.06126336753368378, -0.0819353312253952, -0.33696842193603516, 0.1344735324382782, 0.6249410510063171, 0.2558962404727936, -0.04228174686431885, 0.44199538230895996, 0.03808283433318138, 0.7028172612190247, 0.10719972848892212, -0.33854764699935913, 0.4494321942329407, -0.03851468861103058, 0.18772760033607483, -0.13236775994300842, -0.369392991065979, 0.06506086885929108, 0.22544598579406738, -0.06130904331803322, 0.13972710072994232, -0.02224118635058403, -0.10212033987045288, -0.4349319040775299, 0.23016834259033203, -0.25660964846611023, 0.04861299321055412, -0.005603514611721039, 0.10512038320302963, 0.05551033467054367, -0.06783266365528107, -0.21281301975250244, -0.3788440525531769, -0.48297104239463806, -0.16287319362163544, 0.2749961018562317, -0.23916330933570862, 0.1747974306344986, -0.6903297305107117, -0.09328945726156235, -0.6602327823638916, 0.3001747131347656, -0.1387222707271576, 0.3314149081707001, -0.20235322415828705, 0.20676812529563904, 0.11336015164852142, -0.0766509547829628, 0.1280318647623062, -0.46501556038856506, -0.16797992587089539, -0.07704339921474457, 0.27476781606674194, -0.472359299659729, -0.04191281273961067, -0.027604926377534866, 0.07491299510002136, 0.43653425574302673, 0.23509682714939117, 0.018999427556991577, 0.07955583930015564, -0.1766325831413269, -0.20306070148944855, -0.15370070934295654, -0.22335879504680634, -0.03689347952604294, -0.6711584329605103, -0.35238784551620483, -0.4475812315940857, 0.06721355766057968, 0.1456536054611206, 0.14058202505111694, -0.037643399089574814, 0.1860169619321823, 0.11092222481966019, 0.5076406598091125, 0.042842406779527664, -0.04625977948307991, 0.629524290561676, -0.09305115044116974, -0.45713672041893005, 0.472504198551178, -0.009942961856722832, 0.38947993516921997, 0.03559476137161255, -0.19427186250686646, -0.1696309745311737, -0.2850334048271179, 0.45783284306526184, 0.1723031848669052, 0.08548004180192947, 0.21727722883224487, -0.10695026814937592, 0.028125490993261337, -0.1549404263496399, 0.46995577216148376, 0.33896535634994507, -0.0914730653166771, -0.4181351363658905, -0.18395131826400757, 0.44957301020622253, -0.17907273769378662, -0.03373924270272255, 0.24281132221221924, -0.19249631464481354, -0.2763460576534271, 0.2006179541349411, 0.11939811706542969, 1.091148018836975, -0.07678566873073578, 0.37884485721588135, 0.12213318794965744, -0.0487479567527771, 0.7742531299591064, -0.14554890990257263, 0.045702870935201645, -0.23358836770057678, 0.1004330962896347, -0.11330718547105789, -0.05880850553512573, -0.06760469824075699, 0.1411709487438202, -0.23739971220493317, 0.08419743925333023, -0.04000561311841011, -0.11120772361755371, 0.0072576808743178844, 0.5523278117179871, 0.07271003723144531, -0.39419955015182495, -0.22025981545448303, 0.0838499367237091, -0.31129831075668335, 0.30938011407852173, -0.16995719075202942, -0.2196555733680725, -0.28573909401893616, -0.18248267471790314, -0.20725451409816742, 0.2760235071182251, -0.25142884254455566, 0.09939274936914444, 0.04887313395738602, -0.35540109872817993, 0.3192008137702942, 0.4633541405200958, 0.524256706237793, -0.07650457322597504, -0.09716552495956421, 0.04989174008369446, -0.1302095651626587, -0.014176782220602036, -0.03877822682261467, 0.030591942369937897, 0.19029146432876587, -0.26959821581840515, 0.04275324195623398, 0.25302842259407043, -0.02574026957154274, -0.3122658133506775, -0.17742237448692322, 0.14603468775749207, 0.046734996140003204, -0.16329124569892883, -0.3530453145503998, 0.0866529569029808, -0.17310136556625366, -0.04445986822247505, 0.027808688580989838, 0.028734590858221054, 0.07372549176216125, -0.06803662329912186, 0.09252879023551941, -0.20078012347221375, 0.0808694064617157, 0.10644624382257462, -0.11565140634775162, -0.05883951112627983, 0.7411611080169678, -0.06006147339940071, 0.07929187268018723, -0.085643470287323, -0.19852131605148315, -0.3181910216808319, -0.14075693488121033, -0.08977296203374863, 0.2795100510120392, -0.030942685902118683, 0.08318276703357697, 0.6403139233589172, 0.3384789824485779, 0.10922135412693024, 0.16181769967079163, -0.14827968180179596, -0.3228500783443451, -0.3060607314109802, -0.04380908980965614, 0.1931377351284027, -0.10730212926864624, 0.19159775972366333, 0.20194736123085022, 0.29810240864753723, -0.22825750708580017, -0.1637888103723526, -0.20056785643100739, 0.1670415699481964, 0.754091739654541, -0.011614255607128143, -0.06655439734458923, -0.036773066967725754, -0.01890009641647339, 0.0023950282484292984, -0.16107885539531708, -0.1169770359992981, -0.3814639151096344, 0.17557679116725922, 0.4809395670890808, -0.18953633308410645, -0.04248196631669998, -0.24042126536369324, -0.0995074212551117, -0.1417813003063202, -0.2967749238014221, 0.0675632581114769, -0.03371056914329529, 0.24088989198207855, -0.26430997252464294, 0.33600613474845886, -0.09664463251829147, 0.21474194526672363, 0.006567113101482391, 0.1791645586490631, 0.0874752402305603, -0.022731199860572815, 0.19590207934379578, 0.077484130859375, -0.37948134541511536, -0.038136936724185944, -0.20470307767391205, -0.06822095066308975, 0.028972072526812553, -0.4525882601737976, 0.13075196743011475, -0.11226868629455566, 0.04397562891244888, 0.035784728825092316, -0.4939000904560089, 0.16325348615646362, 0.3917882442474365, 0.10325777530670166, 0.014925528317689896, 0.11850061267614365, 0.16980460286140442, 0.20384256541728973, 0.020077425986528397, 0.244547039270401, 0.03487822785973549, -0.29139673709869385, -0.05173569917678833, -0.20435792207717896, -0.05848463997244835, -0.26753944158554077, 0.121254563331604, -0.09284739196300507, -0.3782729208469391, -0.043263692408800125, 0.1533951312303543, 0.18866701424121857, -0.1023828312754631, 0.19348876178264618, -0.20576730370521545, 0.1222844272851944, 0.03275245428085327, -0.07289662957191467, 0.2181718647480011, -0.2738104462623596, 0.11131155490875244, -0.06723922491073608, -0.02377428114414215, 0.3217807710170746, 0.1499323844909668, 0.6510177254676819, -0.35149484872817993, -0.005461741704493761, -0.26742199063301086, 0.11025910079479218, 0.009168919175863266, 0.1439594179391861, -0.6295010447502136, -0.025503255426883698, 0.054359883069992065, -0.08338461071252823, 0.0133439302444458, -0.3383152484893799, -0.03636631369590759, 0.2841479182243347, -0.14115643501281738, 0.05001092702150345, -0.10629687458276749, 0.2101648449897766, 0.12682895362377167, -0.15633127093315125, 0.41823020577430725, 0.017136946320533752, -0.1046929806470871, -0.04810387268662453, 0.07154323160648346, 0.5066620111465454, 0.5816254019737244, 0.13121725618839264, 0.28301841020584106, 0.18137682974338531, 0.03374314308166504, 0.06061127781867981, 0.19215212762355804, 0.4136762022972107, 0.2930328845977783, 0.14628487825393677, 0.02923169545829296, -0.11405077576637268, 0.07913732528686523, 0.04990002140402794, 0.0570765882730484, -0.3188937306404114, 0.2812897562980652, -0.031187504529953003, -0.336191862821579, -0.16785858571529388, 0.3929540812969208, -0.4424988627433777, 0.011144447140395641, 0.2605806887149811, 0.22482039034366608, 0.3537737727165222, -0.2567600905895233, 0.024390485137701035, -0.16960233449935913, 0.4160141050815582, 0.48843076825141907, -0.07816924899816513, -0.14322052896022797, -0.2934461832046509, -0.5263962745666504, 0.16106991469860077, -0.2162843942642212, -0.23151743412017822, -0.3568779230117798, 0.18111085891723633, 0.05725971236824989, -0.09626960754394531, 0.0581553652882576, 0.14493922889232635, -0.23265406489372253, -0.004405803978443146, -0.44343671202659607, 0.014719972386956215, -0.153797909617424, -0.014265863224864006, -0.038493234664201736, -0.16828292608261108, 0.029524514451622963, 0.0920230969786644, 0.018767893314361572, -0.0066742803901433945, 0.22007080912590027, 0.1332624852657318, 0.3443257510662079, 0.4730617105960846, -0.09442443400621414, 0.7783323526382446, 0.09284857660531998, 0.004618408158421516, -0.006963711231946945, -0.06584223359823227, -0.16452109813690186, 0.02516850084066391, 0.13609915971755981, 0.4669879674911499, -0.08531725406646729, 0.37317198514938354, -0.2298850268125534, 0.03581182658672333, -0.03140724450349808, -0.5221918821334839, 0.07673229277133942, 0.009968927130103111, -0.06821577996015549, 0.24942481517791748, 0.4241013824939728, 0.13799181580543518, 0.0984186977148056, -0.10654041916131973, -0.42790061235427856, -0.3111042082309723, 0.41429752111434937, -0.5499926209449768, -0.33335116505622864, 0.19432814419269562, 0.22382161021232605, -0.05248164385557175, 0.022693302482366562, -0.28314080834388733, -0.07024005055427551, 0.15611465275287628, -0.23596206307411194, -0.019298363476991653, 0.3056490123271942, -0.431890070438385, 0.07046492397785187, -0.052911415696144104, 0.31289142370224, 0.19471442699432373, -0.3801869750022888, -0.04024699330329895, -0.16324961185455322 ]
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
[ 0.010165128856897354, -0.014697805047035217, 0.06399784982204437, -0.04725117236375809, 0.13680770993232727, -0.1440621018409729, 0.3703484535217285, -0.0352640338242054, -0.17155125737190247, 0.1538785994052887, -0.10310790687799454, 0.25897887349128723, -0.23088781535625458, 0.31687358021736145, 0.37554576992988586, 0.09489145874977112, 0.019372418522834778, 0.27177363634109497, -0.020116537809371948, -0.09850377589464188, -0.22678780555725098, 0.07013359665870667, -0.002222903072834015, -0.09574879705905914, -0.14261746406555176, -0.21162621676921844, -0.090327687561512, 0.2581224739551544, -0.23455409705638885, -0.34122946858406067, 0.48685991764068604, 0.23599064350128174, 0.34359246492385864, 0.5804957151412964, -0.00011795770842581987, -0.051996178925037384, 0.17183172702789307, -0.21570520102977753, -0.6393759250640869, -0.3554881811141968, -0.006130743771791458, -0.13421714305877686, 0.060677528381347656, -0.25174057483673096, 0.19287467002868652, 0.027462057769298553, 0.03491725027561188, 0.03088005632162094, 0.4822887182235718, 0.4076220393180847, 0.14891520142555237, 0.145722433924675, -0.17812110483646393, -0.15002450346946716, -0.0028320401906967163, 0.18243561685085297, -0.05897195637226105, 0.25395676493644714, 0.35869622230529785, -0.4081469774246216, 0.2644426226615906, -0.19958709180355072, 0.03678107261657715, -0.04424070194363594, 0.48748522996902466, -0.009249689988791943, -0.07491926103830338, -0.4059666693210602, -0.004839859902858734, 0.17713764309883118, 0.3769998848438263, -0.06002657115459442, -0.3945048451423645, -0.14569073915481567, 0.15505413711071014, -0.4790176451206207, 0.25372353196144104, 0.27886900305747986, -0.42470577359199524, -0.022530261427164078, -0.27115684747695923, 0.3137443959712982, -0.409356027841568, 0.4221656918525696, 0.10564929246902466, 0.35269659757614136, -0.03370986506342888, 0.0666821226477623, 0.24374662339687347, -0.1536940336227417, 0.011768709868192673, 0.06547368317842484, 0.021466810256242752, 0.2782633602619171, -0.2166644185781479, -0.1950293630361557, 0.002793125808238983, -0.24058520793914795, -0.05818124860525131, 0.12915544211864471, 0.4047309458255768, -0.12968289852142334, -0.24420079588890076, 0.15063321590423584, 0.13368673622608185, 0.32412484288215637, 0.0014830529689788818, 0.12236370891332626, 0.12233429402112961, -0.005845783744007349, 0.2769152522087097, -0.051742419600486755, -0.2742708921432495, -0.050802651792764664, -0.11301049590110779, -0.06884722411632538, 0.2920278310775757, 0.06772629916667938, -0.00554061122238636, -0.12934675812721252, 0.09735853224992752, 0.016586599871516228, -0.10256070643663406, 0.3439236879348755, -0.08194633573293686, -0.261303186416626, -0.05380438268184662, 0.2326335608959198, 0.13708706200122833, -0.290010005235672, 0.05705765262246132, 0.27582848072052, -0.11549268662929535, -0.1544322967529297, 0.43567249178886414, 0.012411205098032951, 0.1956508308649063, -0.054452311247587204, 0.2021941840648651, -0.06608017534017563, 0.07224203646183014, 0.011986557394266129, -0.16288024187088013, 0.008681721985340118, 0.21959316730499268, 0.3847878873348236, -0.06849811971187592, -0.5333330631256104, -0.14358171820640564, 0.02616080269217491, -0.3532375395298004, -0.3473283052444458, -0.45941486954689026, 0.08847416192293167, 0.0037650354206562042, -0.2817482054233551, -0.32444483041763306, -0.07150520384311676, 0.08965375274419785, -0.25171083211898804, -0.09521683305501938, -0.04630006104707718, -0.14219067990779877, -0.12248457223176956, 0.12956251204013824, 0.08803869783878326, -0.5273423790931702, 0.18487200140953064, -0.08601373434066772, 0.489278644323349, 0.27127566933631897, 0.2785509526729584, -0.3086808919906616, 0.24388495087623596, -0.0801139548420906, 0.18975184857845306, 0.5997768044471741, -0.48072585463523865, -0.3346844017505646, 0.04278600960969925, -0.03731606900691986, -0.09002774953842163, 0.03727872669696808, 0.1934349685907364, 0.10801922529935837, -0.09636010229587555, 0.09529425948858261, 0.4479227066040039, 0.04567858204245567, 0.016692688688635826, -0.216155543923378, -0.16694766283035278, 0.2624109983444214, 0.040792372077703476, -0.1555020660161972, 0.07853058725595474, 0.2734294533729553, 0.08999688178300858, 0.024154886603355408, -0.3324548900127411, 0.1537162959575653, 0.1996031403541565, 0.3573833107948303, -0.17576824128627777, -0.1370927095413208, 0.20602956414222717, -0.469605028629303, 0.217020183801651, 0.18055641651153564, 0.34908393025398254, -0.12102635204792023, -0.01830347627401352, -0.4437151849269867, -0.10057184100151062, -0.12403099238872528, -0.058668527752161026, 0.049087829887866974, 0.11639866232872009, 0.05498087778687477, -0.17006069421768188, -0.2535091042518616, 0.18731676042079926, -0.2769075036048889, 0.061252862215042114, -0.2197422981262207, 0.04505927860736847, -0.26968783140182495, -0.12417362630367279, 0.2587276101112366, 0.18245284259319305, 0.16576704382896423, 0.006094574462622404, -0.18364930152893066, 0.10668765753507614, -0.47848546504974365, 0.2623986601829529, 0.033690545707941055, 0.1265527456998825, 0.0412624217569828, -0.06761729717254639, -0.18328449130058289, -0.14366233348846436, 0.1708056628704071, -0.14655131101608276, -0.4013840854167938, 0.14321789145469666, -0.027938369661569595, 0.004520796239376068, 0.16433124244213104, 0.00042167119681835175, 0.2795955538749695, -0.1793309599161148, 0.11594351381063461, 0.10681833326816559, 0.3852042555809021, 0.06651086360216141, 0.4020828604698181, -0.04134408384561539, -0.12613633275032043, -0.03238102048635483, 0.2109304815530777, 0.019254609942436218, 0.1892593950033188, 0.07365372776985168, 0.013792533427476883, -0.08014346659183502, 0.2915307581424713, 0.6511791348457336, 0.3804545998573303, 0.11107791215181351, -0.03202156350016594, -0.05627492070198059, -0.10829666256904602, -0.18812237679958344, -0.06301780790090561, -0.11955749988555908, 0.07156787067651749, 0.1802181601524353, 0.42369550466537476, 0.25452080368995667, -0.13882160186767578, -0.03602667152881622, 0.1298334151506424, 0.4580765962600708, -0.19086547195911407, 0.23381786048412323, -0.13966137170791626, -0.43804100155830383, 0.24693119525909424, -0.22983141243457794, -0.09044496715068817, -0.16917309165000916, -0.20334893465042114, 0.13439375162124634, 0.14506080746650696, -0.0874411016702652, 0.11892323195934296, 0.337837815284729, -0.08473843336105347, -0.28900524973869324, -0.26066452264785767, -0.09300477802753448, -0.2001829743385315, 0.053276125341653824, 0.3222638964653015, -0.10707593709230423, 0.4432598948478699, 0.3364589214324951, 0.08638136088848114, -0.05491837114095688, -0.3232802450656891, 0.00857200101017952, -0.06074152886867523, 0.12722769379615784, 0.10323647409677505, 0.34343206882476807, -0.0030989758670330048, -0.5138470530509949, 0.1799706071615219, 0.06420611590147018, -0.2702476382255554, 0.22155006229877472, -0.02242940478026867, -0.31593844294548035, -0.44382137060165405, -0.05534747987985611, -0.2909521460533142, -0.33917292952537537, 0.2574867904186249, 0.32961487770080566, 0.19983582198619843, 0.583056628704071, 0.027132075279951096, 0.13877174258232117, 0.321448415517807, 0.10150288045406342, 0.04909181594848633, -0.2018526792526245, 0.5142779350280762, -0.2508528530597687, -0.46681079268455505, 0.12727035582065582, -0.1354219615459442, 0.5388175845146179, -0.026185207068920135, -0.3549956679344177, -0.20387734472751617, 0.06519965827465057, 0.2458014190196991, -0.036645978689193726, -0.05192917585372925, 0.2837883234024048, -0.19145289063453674, 0.04438886046409607, -0.07229115813970566, -0.33189117908477783, 0.1452784538269043, 0.6041277647018433, 0.2712108790874481, -0.058542825281620026, 0.43397361040115356, 0.043424587696790695, 0.7070649862289429, 0.132707417011261, -0.3311067223548889, 0.44850122928619385, -0.06909222900867462, 0.20655706524848938, -0.14012011885643005, -0.37472620606422424, 0.13324420154094696, 0.23546567559242249, -0.033650998026132584, 0.11279840022325516, -0.03097517415881157, -0.08227454870939255, -0.41568055748939514, 0.22088682651519775, -0.23390769958496094, 0.04839261993765831, -0.016792142763733864, 0.11899576336145401, 0.04606027528643608, -0.08111147582530975, -0.22336795926094055, -0.3637562096118927, -0.5096311569213867, -0.12451796233654022, 0.24588915705680847, -0.24647220969200134, 0.21042367815971375, -0.6700917482376099, -0.05532325804233551, -0.6859089732170105, 0.2873028516769409, -0.1499042809009552, 0.3086010217666626, -0.21619319915771484, 0.19936713576316833, 0.094112828373909, -0.08231562376022339, 0.150537371635437, -0.47544893622398376, -0.15888743102550507, -0.08539849519729614, 0.2848472595214844, -0.47482746839523315, -0.026424815878272057, -0.05132071301341057, 0.07265022397041321, 0.4124598503112793, 0.19209975004196167, 0.03950861841440201, 0.07596605271100998, -0.14139607548713684, -0.21616972982883453, -0.17694474756717682, -0.22804751992225647, -0.04577102139592171, -0.6439384818077087, -0.33588743209838867, -0.46366727352142334, 0.056370772421360016, 0.1376708745956421, 0.16398774087429047, -0.06866595149040222, 0.15550364553928375, 0.09471671283245087, 0.5056183338165283, 0.047923874109983444, -0.066242516040802, 0.638486385345459, -0.09328266978263855, -0.42609381675720215, 0.485016405582428, 0.0034874696284532547, 0.3892187774181366, 0.028055738657712936, -0.14984765648841858, -0.15592554211616516, -0.2998329997062683, 0.47131356596946716, 0.13734591007232666, 0.08705980330705643, 0.18480390310287476, -0.09927674382925034, 0.02812296152114868, -0.19187159836292267, 0.5012048482894897, 0.3349218964576721, -0.08053115010261536, -0.4222892224788666, -0.18815913796424866, 0.4366787374019623, -0.16746795177459717, -0.002293229103088379, 0.24263334274291992, -0.18027828633785248, -0.26839780807495117, 0.1748557984828949, 0.14205573499202728, 1.0450079441070557, -0.09819291532039642, 0.3577185273170471, 0.11459758877754211, -0.030823538079857826, 0.7806406617164612, -0.1800108700990677, 0.04540562629699707, -0.2553688883781433, 0.09904008358716965, -0.11995568126440048, -0.04338899254798889, -0.04568786919116974, 0.13680106401443481, -0.2528732120990753, 0.08574341982603073, -0.029485352337360382, -0.14704395830631256, 0.005725708790123463, 0.5664041638374329, 0.06178762763738632, -0.3942602276802063, -0.20463372766971588, 0.10850328207015991, -0.31704938411712646, 0.2979927957057953, -0.1628500521183014, -0.20761960744857788, -0.28265658020973206, -0.15511317551136017, -0.22094181180000305, 0.30353090167045593, -0.23986969888210297, 0.07949524372816086, 0.05891779810190201, -0.319892942905426, 0.3326181173324585, 0.46785664558410645, 0.5250080823898315, -0.08236932754516602, -0.10083150863647461, 0.037575770169496536, -0.167093425989151, -0.027053849771618843, 0.009089290164411068, 0.05029284209012985, 0.22911930084228516, -0.290698379278183, 0.03021179884672165, 0.2814527750015259, -0.0009340867400169373, -0.30204153060913086, -0.17566703259944916, 0.13425405323505402, 0.07122749090194702, -0.16748592257499695, -0.3050804138183594, 0.10473191738128662, -0.13321731984615326, -0.039233628660440445, 0.0401461198925972, 0.04761671647429466, 0.06888840347528458, -0.028613664209842682, 0.07775548100471497, -0.18398180603981018, 0.057182833552360535, 0.11787647008895874, -0.11957670003175735, -0.07269173860549927, 0.7520657777786255, -0.053994070738554, 0.09325073659420013, -0.10994230210781097, -0.18213312327861786, -0.273678183555603, -0.13815313577651978, -0.09949136525392532, 0.2768901586532593, -0.020947933197021484, 0.09042224287986755, 0.6383234858512878, 0.3213026523590088, 0.09763477742671967, 0.13834625482559204, -0.1498948335647583, -0.3523445427417755, -0.33584898710250854, -0.020538460463285446, 0.1858692467212677, -0.12017836421728134, 0.14587904512882233, 0.2181803286075592, 0.31043580174446106, -0.25146135687828064, -0.1950041800737381, -0.19985800981521606, 0.15835464000701904, 0.7369635105133057, -0.01766325533390045, -0.012900841422379017, -0.06114561855792999, 0.006987877190113068, -0.03010920062661171, -0.17383694648742676, -0.1364712119102478, -0.36932259798049927, 0.16456398367881775, 0.46165916323661804, -0.19141323864459991, -0.056230101734399796, -0.21856941282749176, -0.12847858667373657, -0.13379023969173431, -0.2693403959274292, 0.06040240824222565, -0.057562947273254395, 0.2415003627538681, -0.22885139286518097, 0.32027676701545715, -0.08825058490037918, 0.17375975847244263, 0.03331265226006508, 0.16573955118656158, 0.07580462098121643, -0.019646011292934418, 0.19066877663135529, 0.07480838894844055, -0.3352307677268982, -0.02620074898004532, -0.1774321049451828, -0.0726977214217186, 0.017780333757400513, -0.47883278131484985, 0.10806423425674438, -0.13837334513664246, 0.05096103996038437, 0.022260237485170364, -0.4982840418815613, 0.13783924281597137, 0.38783401250839233, 0.11959435045719147, -0.0020742658525705338, 0.11066970229148865, 0.20888091623783112, 0.2025485336780548, 0.013452615588903427, 0.2371889054775238, 0.0218781940639019, -0.2497704029083252, -0.0744701474905014, -0.19819730520248413, -0.07921246439218521, -0.28159695863723755, 0.09921444952487946, -0.11246882379055023, -0.3867097496986389, -0.0492035411298275, 0.1662590652704239, 0.19723787903785706, -0.09329438954591751, 0.21389196813106537, -0.20517244935035706, 0.11387556046247482, 0.032077886164188385, -0.06195525452494621, 0.2281331717967987, -0.2797221541404724, 0.1353696882724762, -0.09851956367492676, 0.005143940448760986, 0.3340870440006256, 0.14589402079582214, 0.6321766972541809, -0.31623148918151855, -0.01611216738820076, -0.24867364764213562, 0.07460862398147583, 0.01759660243988037, 0.16787521541118622, -0.6061632037162781, -0.02913285791873932, 0.05296070873737335, -0.08525784313678741, 0.01077720895409584, -0.36782461404800415, -0.04822459816932678, 0.2759411931037903, -0.15788498520851135, 0.05883438140153885, -0.09755999594926834, 0.19536921381950378, 0.13777554035186768, -0.1273939162492752, 0.40262725949287415, -0.008652724325656891, -0.09197324514389038, -0.05486853048205376, 0.11255726963281631, 0.46995019912719727, 0.579823911190033, 0.14910875260829926, 0.3191978931427002, 0.20295307040214539, 0.03397442400455475, 0.06528585404157639, 0.183114156126976, 0.3685245215892792, 0.2624918520450592, 0.13088898360729218, 0.040757112205028534, -0.10797838866710663, 0.059256549924612045, 0.027628939598798752, 0.07058069109916687, -0.308167964220047, 0.2848705053329468, -0.01995892822742462, -0.3238849937915802, -0.17790156602859497, 0.397659033536911, -0.42802703380584717, 0.008006664924323559, 0.3059368133544922, 0.2120257019996643, 0.3255811631679535, -0.27194514870643616, 0.033614348620176315, -0.18920110166072845, 0.45011478662490845, 0.4820158779621124, -0.09710150212049484, -0.10160064697265625, -0.25178587436676025, -0.5138525366783142, 0.1598954200744629, -0.19470663368701935, -0.26225218176841736, -0.3417937159538269, 0.1898447722196579, 0.036116696894168854, -0.08899536728858948, 0.10701452940702438, 0.13900624215602875, -0.21020136773586273, -0.02003651112318039, -0.46240222454071045, 0.013604216277599335, -0.15904930233955383, 0.009875490330159664, -0.02020547166466713, -0.17562425136566162, 0.043981000781059265, 0.06408044695854187, 0.01580042392015457, -0.0002460945397615433, 0.18995270133018494, 0.11839324235916138, 0.32662031054496765, 0.4968593418598175, -0.10089100152254105, 0.7589023113250732, 0.10915186256170273, -0.008116360753774643, -0.012253377586603165, -0.06070781126618385, -0.2116783708333969, 0.07008427381515503, 0.12792526185512543, 0.4384544789791107, -0.0824115201830864, 0.3603610694408417, -0.22094301879405975, 0.07509887963533401, -0.033126525580883026, -0.5146056413650513, 0.08451104164123535, 0.02284649945795536, -0.04442755877971649, 0.2764641344547272, 0.41493144631385803, 0.09248562157154083, 0.09413083642721176, -0.09509802609682083, -0.4262844920158386, -0.2746139168739319, 0.39774632453918457, -0.5510282516479492, -0.29202577471733093, 0.1532038152217865, 0.21421507000923157, -0.032668761909008026, -0.009405512362718582, -0.3166782259941101, -0.06710265576839447, 0.17874471843242645, -0.22816726565361023, -0.011713536456227303, 0.2962084412574768, -0.4524925947189331, 0.09072346985340118, -0.046880509704351425, 0.3302346169948578, 0.17732536792755127, -0.3714454174041748, -0.016200726851820946, -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>`
[ 0.07664860785007477, -0.05189128220081329, 0.05664337798953056, -0.013380460441112518, 0.1621718555688858, -0.1869213581085205, 0.3156285285949707, -0.02782139927148819, -0.19460216164588928, 0.19591796398162842, -0.1271069049835205, 0.23585718870162964, -0.22072944045066833, 0.292635053396225, 0.3533962070941925, 0.07266484946012497, 0.005575597286224365, 0.2951040267944336, 0.009004920721054077, -0.09385786950588226, -0.21986123919487, 0.09480921179056168, 0.019349902868270874, -0.07987074553966522, -0.13481725752353668, -0.1881403625011444, -0.07919929921627045, 0.27564123272895813, -0.19723261892795563, -0.33476564288139343, 0.4486151337623596, 0.2227426916360855, 0.3325243592262268, 0.6002951264381409, -0.00011811569856945425, -0.05863648280501366, 0.16570237278938293, -0.22807450592517853, -0.6185938715934753, -0.37377268075942993, 0.07301405817270279, -0.14274385571479797, 0.0919017344713211, -0.2383846640586853, 0.18933576345443726, 0.03741716966032982, 0.016469644382596016, 0.06538467854261398, 0.52815842628479, 0.37782976031303406, 0.14735276997089386, 0.1031377837061882, -0.17057590186595917, -0.09499382227659225, 0.027344077825546265, 0.18931761384010315, -0.05604257434606552, 0.29447856545448303, 0.33864349126815796, -0.3992345631122589, 0.2927800118923187, -0.19877883791923523, -0.00343981571495533, -0.06038891151547432, 0.48242419958114624, -0.027946360409259796, 0.0078292116522789, -0.4104381203651428, -0.029902271926403046, 0.17739379405975342, 0.30182331800460815, -0.030178628861904144, -0.3591754734516144, -0.11387863755226135, 0.16878533363342285, -0.4841402471065521, 0.25605514645576477, 0.27259162068367004, -0.41021379828453064, -0.03852648660540581, -0.25709426403045654, 0.32797667384147644, -0.4180815815925598, 0.44269436597824097, 0.1196066290140152, 0.3393747806549072, -0.022680915892124176, 0.07667822390794754, 0.2550285756587982, -0.1277889609336853, 0.007880826480686665, 0.09094315767288208, 0.021747460588812828, 0.2719120383262634, -0.19632475078105927, -0.18465973436832428, -0.02068992704153061, -0.2897568941116333, -0.10948570817708969, 0.09983215481042862, 0.4507439434528351, -0.1309795081615448, -0.2473733127117157, 0.14947140216827393, 0.09313434362411499, 0.32917824387550354, -0.007071547210216522, 0.09553524106740952, 0.12827558815479279, 0.021282576024532318, 0.27535539865493774, -0.066107377409935, -0.2871696352958679, -0.07212265580892563, -0.12615810334682465, -0.04732624441385269, 0.27471923828125, 0.08483269810676575, 0.011768626049160957, -0.09488480538129807, 0.07253137230873108, 0.060445889830589294, -0.0801716074347496, 0.3419950604438782, -0.11174273490905762, -0.24170127511024475, 0.005972001701593399, 0.23593328893184662, 0.13921090960502625, -0.26889768242836, 0.07524190843105316, 0.30232560634613037, -0.11196798086166382, -0.14454683661460876, 0.4047555923461914, 0.008399376645684242, 0.1989605575799942, -0.046806368976831436, 0.18748703598976135, -0.13644537329673767, 0.06247740238904953, 0.006645042449235916, -0.1763249635696411, 0.03343529999256134, 0.22955000400543213, 0.38577964901924133, -0.06581887602806091, -0.5420820116996765, -0.13114003837108612, 0.03096667304635048, -0.37687376141548157, -0.3573206067085266, -0.44525375962257385, 0.08834989368915558, 0.04142218828201294, -0.2367771863937378, -0.307137131690979, -0.1186676174402237, 0.09791091084480286, -0.2934206426143646, -0.09514971822500229, -0.04661904275417328, -0.1209682747721672, -0.13289496302604675, 0.11520536988973618, 0.04702848941087723, -0.47550541162490845, 0.21962383389472961, -0.08374617993831635, 0.512432336807251, 0.32623007893562317, 0.29876676201820374, -0.3122524321079254, 0.271157830953598, -0.06317339092493057, 0.17326487600803375, 0.5793401002883911, -0.47124627232551575, -0.2747647166252136, 0.027949489653110504, -0.03406146541237831, -0.10284287482500076, 0.05465691164135933, 0.17109650373458862, 0.08083709329366684, -0.09934062510728836, 0.07489281892776489, 0.44040703773498535, 0.07226992398500443, 0.04624674469232559, -0.23941278457641602, -0.16745544970035553, 0.258391797542572, 0.03397069498896599, -0.16789847612380981, 0.07635146379470825, 0.24125459790229797, 0.08969418704509735, 0.0238095261156559, -0.33165550231933594, 0.1337258666753769, 0.20273829996585846, 0.3554898500442505, -0.1569676250219345, -0.14370864629745483, 0.2355039119720459, -0.40783828496932983, 0.18856623768806458, 0.12364427745342255, 0.3522772490978241, -0.10161960124969482, -0.010192330926656723, -0.41855716705322266, -0.0969458520412445, -0.09851209819316864, -0.023751594126224518, 0.06040218472480774, 0.10851233452558517, 0.05050450190901756, -0.18064525723457336, -0.22650733590126038, 0.14043653011322021, -0.30637800693511963, 0.05729084834456444, -0.2142498642206192, 0.0430721640586853, -0.28264400362968445, -0.11205196380615234, 0.253186970949173, 0.1754760891199112, 0.1947956085205078, -0.015051346272230148, -0.19718411564826965, 0.0930892676115036, -0.4591653347015381, 0.3227218985557556, -0.002345513552427292, 0.13259637355804443, 0.03050420992076397, -0.022738121449947357, -0.23679900169372559, -0.09179329872131348, 0.1783350557088852, -0.16936469078063965, -0.33602118492126465, 0.13825173676013947, -0.029993291944265366, 0.016736023128032684, 0.15873317420482635, 0.0029885340481996536, 0.28925013542175293, -0.18946008384227753, 0.11126898974180222, 0.13147734105587006, 0.4015832841396332, 0.09714187681674957, 0.3815452754497528, -0.04512950778007507, -0.13328233361244202, -0.08486033231019974, 0.18308381736278534, 0.023189224302768707, 0.16448454558849335, 0.07439347356557846, 0.020963281393051147, -0.0921829491853714, 0.30198606848716736, 0.5998583436012268, 0.3605665862560272, 0.07620026916265488, -0.05645529553294182, -0.06078985333442688, -0.1149052232503891, -0.22583210468292236, -0.03621252626180649, -0.08589228987693787, 0.06461302936077118, 0.2199166715145111, 0.43957406282424927, 0.26594144105911255, -0.13194306194782257, -0.011397767812013626, 0.10586436837911606, 0.4606073498725891, -0.19288797676563263, 0.24376700818538666, -0.1272561252117157, -0.4152642488479614, 0.24683064222335815, -0.2743428349494934, -0.08276297152042389, -0.17546343803405762, -0.199441596865654, 0.15085017681121826, 0.09969285130500793, -0.11422117799520493, 0.17351354658603668, 0.32977747917175293, -0.10300754010677338, -0.2672959268093109, -0.28693437576293945, -0.08144953846931458, -0.20950135588645935, 0.043896693736314774, 0.33913376927375793, -0.10610567033290863, 0.42552557587623596, 0.33081158995628357, 0.05794121325016022, -0.031525470316410065, -0.3174516558647156, -0.012977927923202515, -0.038392044603824615, 0.09033845365047455, 0.08890723437070847, 0.3747074007987976, 0.01147523894906044, -0.5478421449661255, 0.19541172683238983, 0.05433341860771179, -0.25022077560424805, 0.2478959858417511, -0.031557921320199966, -0.31541746854782104, -0.47585931420326233, 0.002425380051136017, -0.26761865615844727, -0.3631879389286041, 0.24924412369728088, 0.3301575183868408, 0.2313889116048813, 0.5911978483200073, 0.019704103469848633, 0.13168871402740479, 0.3197172284126282, 0.1506917029619217, 0.045514725148677826, -0.18698321282863617, 0.49666300415992737, -0.2360728681087494, -0.42654773592948914, 0.10766628384590149, -0.14385582506656647, 0.5702795386314392, -0.04440854489803314, -0.3675757646560669, -0.18725848197937012, 0.08487427979707718, 0.17660599946975708, -0.07252096384763718, -0.07167656719684601, 0.29713717103004456, -0.19216713309288025, 0.0524093434214592, -0.05853378027677536, -0.29756587743759155, 0.1311010718345642, 0.6311604380607605, 0.2633405029773712, -0.05882040411233902, 0.4399247169494629, 0.01797991245985031, 0.6985477805137634, 0.10503318905830383, -0.3126603066921234, 0.44941115379333496, -0.038286007940769196, 0.18113131821155548, -0.1402672827243805, -0.3474927544593811, 0.08872711658477783, 0.23850512504577637, -0.042701371014118195, 0.1445247232913971, -0.0436849370598793, -0.0786023736000061, -0.44182536005973816, 0.24700278043746948, -0.2635873556137085, 0.05894371494650841, 0.009580896236002445, 0.10640180110931396, 0.037593960762023926, -0.10671721398830414, -0.24084511399269104, -0.34414172172546387, -0.5165199041366577, -0.1217706948518753, 0.26317548751831055, -0.24034827947616577, 0.18618836998939514, -0.6647051572799683, -0.12223074585199356, -0.6593774557113647, 0.3021116554737091, -0.1535392850637436, 0.2957099378108978, -0.20639866590499878, 0.18639734387397766, 0.1276567280292511, -0.0775057002902031, 0.1092493087053299, -0.4198964238166809, -0.1877129077911377, -0.09577645361423492, 0.2693746089935303, -0.4779631197452545, -0.05513397231698036, -0.0379420630633831, 0.08147379010915756, 0.4557524621486664, 0.2085588425397873, 0.012649916112422943, 0.09506552666425705, -0.16743123531341553, -0.22884663939476013, -0.17120373249053955, -0.24247993528842926, -0.047817058861255646, -0.638192355632782, -0.34167706966400146, -0.48311924934387207, 0.04853607714176178, 0.16847625374794006, 0.15164008736610413, -0.0745968297123909, 0.1877347081899643, 0.09917943924665451, 0.4985141158103943, 0.021204393357038498, -0.057446740567684174, 0.6246477365493774, -0.11789751052856445, -0.44503387808799744, 0.4890882968902588, -0.035323888063430786, 0.35839998722076416, 0.06729531288146973, -0.14457662403583527, -0.1746768057346344, -0.2812870144844055, 0.44760602712631226, 0.1807999610900879, 0.10797969251871109, 0.21889814734458923, -0.11514140665531158, 0.04811500757932663, -0.1877969205379486, 0.4967009127140045, 0.3341120183467865, -0.09505445510149002, -0.3954765796661377, -0.15638874471187592, 0.4564093351364136, -0.19788669049739838, -0.0036152824759483337, 0.2535567581653595, -0.22835589945316315, -0.281291127204895, 0.20379826426506042, 0.1643548309803009, 1.138471245765686, -0.09754566848278046, 0.3732185661792755, 0.12780284881591797, -0.0374407134950161, 0.7614163756370544, -0.18077364563941956, 0.05255858227610588, -0.256600022315979, 0.14564953744411469, -0.10637672990560532, -0.05241217091679573, -0.08326361328363419, 0.08725977689027786, -0.23801003396511078, 0.07538589090108871, -0.06170376390218735, -0.13530021905899048, -0.005106111988425255, 0.5711607336997986, 0.10040029883384705, -0.3970246911048889, -0.21205256879329681, 0.09203217178583145, -0.3342668414115906, 0.3082551956176758, -0.16845157742500305, -0.23394064605236053, -0.3076724410057068, -0.14399370551109314, -0.18095362186431885, 0.28737279772758484, -0.2535722553730011, 0.07502125203609467, 0.06265754997730255, -0.33089619874954224, 0.2630811631679535, 0.4393528699874878, 0.5037201642990112, -0.07716363668441772, -0.09232001006603241, 0.02560356631875038, -0.13528934121131897, -0.04990985617041588, -0.026900609955191612, 0.07493158429861069, 0.191163569688797, -0.31154924631118774, 0.07954157143831253, 0.2915780544281006, -0.02502903901040554, -0.3022928833961487, -0.1862049400806427, 0.15034270286560059, 0.0951574295759201, -0.14537625014781952, -0.33030492067337036, 0.09154897928237915, -0.18308109045028687, -0.03186061233282089, 0.040296826511621475, 0.05976008251309395, 0.04360241815447807, -0.06491034477949142, 0.09557205438613892, -0.18126699328422546, 0.07074012607336044, 0.09366314113140106, -0.09329038113355637, -0.05471434444189072, 0.755597710609436, -0.07153280079364777, 0.09503118693828583, -0.08538438379764557, -0.21003076434135437, -0.3262542486190796, -0.15901440382003784, -0.0998723953962326, 0.2788127064704895, -0.051781632006168365, 0.08019936084747314, 0.6546950936317444, 0.3361327648162842, 0.12761613726615906, 0.1409851461648941, -0.1391994059085846, -0.3442883789539337, -0.333631694316864, -0.02599933370947838, 0.17864908277988434, -0.10025950521230698, 0.20802272856235504, 0.19337491691112518, 0.3369990289211273, -0.2439340502023697, -0.18857669830322266, -0.19681774079799652, 0.1518172323703766, 0.7084043025970459, -0.03436862677335739, -0.06733915954828262, -0.02548118680715561, -0.0026783663779497147, -0.02516811341047287, -0.17077355086803436, -0.12747356295585632, -0.37970876693725586, 0.16888132691383362, 0.48321598768234253, -0.1887608915567398, -0.06686544418334961, -0.21625450253486633, -0.10766865313053131, -0.1371471881866455, -0.3264840841293335, 0.03570879250764847, -0.04568508267402649, 0.20570193231105804, -0.2262984663248062, 0.3478628396987915, -0.09009246528148651, 0.2073937952518463, 0.017075777053833008, 0.18031837046146393, 0.05808906629681587, -0.04084564745426178, 0.17120876908302307, 0.06229694187641144, -0.3547435402870178, -0.03789883852005005, -0.18474037945270538, -0.10040662437677383, 0.025786247104406357, -0.43826913833618164, 0.1099516749382019, -0.15907278656959534, 0.043411001563072205, 0.03248404338955879, -0.5039010643959045, 0.15125025808811188, 0.39393386244773865, 0.10969085991382599, 0.04253868758678436, 0.12714213132858276, 0.1619914472103119, 0.20843978226184845, -0.017391681671142578, 0.22560808062553406, 0.039039161056280136, -0.2658630311489105, -0.052080437541007996, -0.16882598400115967, -0.12271874397993088, -0.2723003029823303, 0.07296204566955566, -0.14059127867221832, -0.4185332953929901, -0.059193868190050125, 0.1482294648885727, 0.16802509129047394, -0.10553344339132309, 0.15312913060188293, -0.20561042428016663, 0.08979486674070358, 0.018970036879181862, -0.08495743572711945, 0.23121878504753113, -0.2662840187549591, 0.07315874844789505, -0.08859777450561523, -0.032751381397247314, 0.33546701073646545, 0.14241033792495728, 0.6379615664482117, -0.32824769616127014, -0.02502352185547352, -0.24423660337924957, 0.1070641428232193, 0.005122607573866844, 0.1822264939546585, -0.6148508191108704, -0.019356362521648407, 0.04265167564153671, -0.10466176271438599, 0.030356280505657196, -0.3352682590484619, -0.06210431456565857, 0.3056540787220001, -0.17846320569515228, 0.09636440128087997, -0.14240166544914246, 0.2184900939464569, 0.16331420838832855, -0.16216038167476654, 0.40879136323928833, 0.03424454480409622, -0.10550744831562042, -0.08757491409778595, 0.08006279915571213, 0.49889031052589417, 0.5946881175041199, 0.11233998835086823, 0.30674824118614197, 0.21389813721179962, 0.03758193552494049, 0.0789858028292656, 0.20481666922569275, 0.43595924973487854, 0.28090909123420715, 0.114513099193573, 0.04079467058181763, -0.1263175755739212, 0.07522377371788025, 0.031144484877586365, 0.1062246561050415, -0.292772501707077, 0.3019169569015503, -0.010781370103359222, -0.3658282160758972, -0.17470884323120117, 0.38556450605392456, -0.4677879810333252, 0.023127175867557526, 0.24933668971061707, 0.21936671435832977, 0.3385199308395386, -0.2580972909927368, 0.03067830577492714, -0.17712639272212982, 0.4321080446243286, 0.48943403363227844, -0.08823614567518234, -0.12464950233697891, -0.28623396158218384, -0.47932541370391846, 0.18860946595668793, -0.1928073912858963, -0.21018359065055847, -0.3527606129646301, 0.18883012235164642, 0.04558246582746506, -0.10253976285457611, 0.07314872741699219, 0.0991477370262146, -0.22263728082180023, -0.02926824986934662, -0.45445024967193604, 0.049344249069690704, -0.17575670778751373, -0.03394394367933273, -0.028623457998037338, -0.1621444970369339, 0.07061208039522171, 0.059837374836206436, 0.03442052751779556, 0.020242618396878242, 0.22016996145248413, 0.11978629231452942, 0.3783562481403351, 0.47854748368263245, -0.09864504635334015, 0.7240331172943115, 0.0831502377986908, -0.009448787197470665, 0.026825491338968277, -0.05517902225255966, -0.17598721385002136, 0.03951149433851242, 0.1294780969619751, 0.4713501036167145, -0.07734829932451248, 0.35965675115585327, -0.24028164148330688, 0.07416467368602753, -0.05003424361348152, -0.537612795829773, 0.08918238431215286, -0.004219084046781063, -0.044237785041332245, 0.2662244439125061, 0.4257829487323761, 0.11622225493192673, 0.11474072933197021, -0.12512779235839844, -0.46675992012023926, -0.28307726979255676, 0.41881105303764343, -0.5340771675109863, -0.3015989363193512, 0.19961611926555634, 0.2269032597541809, -0.06853408366441727, 0.050434861332178116, -0.30186235904693604, -0.06729238480329514, 0.163137286901474, -0.22798311710357666, -0.01558799110352993, 0.31345662474632263, -0.47143709659576416, 0.0794144868850708, -0.03801819682121277, 0.279806911945343, 0.20968550443649292, -0.3522428274154663, -0.04154529049992561, -0.15494854748249054 ]
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)
[ 0.09444819390773773, -0.03021540492773056, 0.04769307002425194, -0.02157311886548996, 0.15865458548069, -0.19878631830215454, 0.3093152642250061, -0.028614450246095657, -0.2004597783088684, 0.17955026030540466, -0.14158895611763, 0.26486989855766296, -0.21903859078884125, 0.2876114845275879, 0.3653639853000641, 0.08437439799308777, -0.012491606175899506, 0.2811937928199768, -0.014133051037788391, -0.09843245148658752, -0.21194221079349518, 0.10448246449232101, 0.03202156722545624, -0.04935638606548309, -0.11954204738140106, -0.17534485459327698, -0.09639161080121994, 0.2744736969470978, -0.20591849088668823, -0.32282862067222595, 0.44809550046920776, 0.22257763147354126, 0.30274730920791626, 0.5770042538642883, -0.0001172610791400075, -0.04289437085390091, 0.14301031827926636, -0.22344990074634552, -0.6293326020240784, -0.382637083530426, 0.07990779727697372, -0.13676342368125916, 0.07053136080503464, -0.25299200415611267, 0.19080787897109985, 0.04170547053217888, 0.0007065478712320328, 0.0639595165848732, 0.5112634301185608, 0.3838019371032715, 0.1552751660346985, 0.08335160464048386, -0.19444243609905243, -0.0838209018111229, 0.03142506256699562, 0.18943168222904205, -0.04444887861609459, 0.2973892390727997, 0.36469703912734985, -0.3999795615673065, 0.28799811005592346, -0.19582900404930115, -0.004944635555148125, -0.09528373181819916, 0.4774853587150574, -0.02443009614944458, 0.002210259437561035, -0.4087410867214203, -0.004733026027679443, 0.14790241420269012, 0.2761026620864868, -0.044267818331718445, -0.34744542837142944, -0.10563983023166656, 0.1462830901145935, -0.48935994505882263, 0.25155404210090637, 0.3014514744281769, -0.42473962903022766, -0.029061753302812576, -0.2817878723144531, 0.32458654046058655, -0.4137301445007324, 0.43699783086776733, 0.11532680690288544, 0.354594886302948, -0.021374192088842392, 0.07461950927972794, 0.23075909912586212, -0.13691918551921844, -0.022248752415180206, 0.07378741353750229, 0.012209251523017883, 0.2807027995586395, -0.18891213834285736, -0.18761013448238373, -0.018634609878063202, -0.2808479964733124, -0.09787112474441528, 0.11185141652822495, 0.4561227858066559, -0.13783910870552063, -0.2397993803024292, 0.15594863891601562, 0.09088344871997833, 0.3234420120716095, -0.006549075245857239, 0.09779709577560425, 0.12052936106920242, 0.00033426645677536726, 0.28369003534317017, -0.0605228915810585, -0.2795400023460388, -0.07793854922056198, -0.14036278426647186, -0.03552112728357315, 0.2655668258666992, 0.08274981379508972, -0.0020336154848337173, -0.0889303907752037, 0.06567001342773438, 0.06152743846178055, -0.08327458053827286, 0.3429432809352875, -0.08083135634660721, -0.25687673687934875, -0.004063107073307037, 0.25254517793655396, 0.14727067947387695, -0.26811474561691284, 0.0705912634730339, 0.322195827960968, -0.12113502621650696, -0.14960487186908722, 0.40277257561683655, 0.016711529344320297, 0.20685160160064697, -0.05475427210330963, 0.16550320386886597, -0.14268727600574493, 0.06256703287363052, 0.01568053849041462, -0.1528286337852478, 0.03313666954636574, 0.2373732328414917, 0.3591506779193878, -0.07919710874557495, -0.5343109965324402, -0.12540274858474731, 0.04936794191598892, -0.38119059801101685, -0.3526667356491089, -0.4539191722869873, 0.09591706842184067, 0.04416888952255249, -0.22488164901733398, -0.32030734419822693, -0.10148513317108154, 0.09989187121391296, -0.2611241340637207, -0.07177598029375076, -0.026748213917016983, -0.1189921498298645, -0.12630946934223175, 0.12081434577703476, 0.03390306979417801, -0.4636046886444092, 0.20927631855010986, -0.04615643620491028, 0.5230317115783691, 0.2966090440750122, 0.28856390714645386, -0.3173558712005615, 0.26712968945503235, -0.06915789097547531, 0.18908892571926117, 0.5893789529800415, -0.4831956624984741, -0.25311359763145447, 0.0503125824034214, -0.037071049213409424, -0.1145578920841217, 0.02929721772670746, 0.1607622355222702, 0.08552613109350204, -0.0990605279803276, 0.07839181274175644, 0.42903363704681396, 0.07837413996458054, 0.04569486528635025, -0.24622958898544312, -0.1808585375547409, 0.22457720339298248, 0.03091520071029663, -0.1713448166847229, 0.07902853190898895, 0.2619319558143616, 0.05616098642349243, 0.02804422751069069, -0.3400058150291443, 0.14545096457004547, 0.19515591859817505, 0.33123913407325745, -0.16132642328739166, -0.1434578150510788, 0.24710294604301453, -0.40864309668540955, 0.1843581199645996, 0.11572976410388947, 0.361627459526062, -0.14487531781196594, 0.0029637105762958527, -0.4288193881511688, -0.09242089837789536, -0.08533722162246704, -0.03767777606844902, 0.06950782984495163, 0.09063811600208282, 0.06854411959648132, -0.16730563342571259, -0.22225067019462585, 0.11700507998466492, -0.30204761028289795, 0.05609985440969467, -0.18788982927799225, 0.06197626143693924, -0.28403085470199585, -0.11223077774047852, 0.22809478640556335, 0.18706172704696655, 0.20362785458564758, -0.01500997319817543, -0.1789759397506714, 0.0905495136976242, -0.4720057249069214, 0.3271068334579468, 0.045911043882369995, 0.16948187351226807, 0.023193463683128357, -0.026097919791936874, -0.21184656023979187, -0.11862611770629883, 0.18914373219013214, -0.15958642959594727, -0.32982558012008667, 0.15665610134601593, -0.022238533943891525, 0.008161164820194244, 0.16950006783008575, 0.012336524203419685, 0.28302568197250366, -0.1977064460515976, 0.11460226029157639, 0.13609112799167633, 0.4101678729057312, 0.09705832600593567, 0.37629246711730957, -0.03750362992286682, -0.139716237783432, -0.07505517452955246, 0.15518182516098022, 0.0004627462476491928, 0.1710679829120636, 0.07220088690519333, 0.025278333574533463, -0.09470263868570328, 0.2884228229522705, 0.5951279997825623, 0.35346925258636475, 0.07945821434259415, -0.05248225852847099, -0.04475101828575134, -0.1115134209394455, -0.22755970060825348, -0.016284100711345673, -0.07748311758041382, 0.04854606091976166, 0.2117173969745636, 0.43869125843048096, 0.2605527639389038, -0.13261163234710693, -0.01688234694302082, 0.11479970067739487, 0.4540281295776367, -0.19068242609500885, 0.266332745552063, -0.09921854734420776, -0.4326513111591339, 0.25927120447158813, -0.2881087064743042, -0.08621688187122345, -0.18278421461582184, -0.18338346481323242, 0.12537230551242828, 0.0859253853559494, -0.11140725016593933, 0.1523824781179428, 0.35105255246162415, -0.09527852386236191, -0.27169257402420044, -0.2880839705467224, -0.10307908803224564, -0.20516344904899597, 0.04893950745463371, 0.3304883539676666, -0.11064367741346359, 0.4104601740837097, 0.3433223366737366, 0.06626521050930023, -0.04624151438474655, -0.32678037881851196, -0.0032144710421562195, -0.04425811022520065, 0.09379170089960098, 0.0913999155163765, 0.3395487666130066, 0.004957783967256546, -0.5441511869430542, 0.18702395260334015, 0.058574043214321136, -0.24736620485782623, 0.25977322459220886, -0.03784012421965599, -0.31450504064559937, -0.4980628192424774, 0.010082259774208069, -0.2758822739124298, -0.3460386395454407, 0.2394309937953949, 0.33021262288093567, 0.23633964359760284, 0.5939531922340393, 0.018188927322626114, 0.13191567361354828, 0.3163865804672241, 0.1342409998178482, 0.03354417160153389, -0.20578573644161224, 0.4849308729171753, -0.2360694408416748, -0.4347623586654663, 0.10699346661567688, -0.13748374581336975, 0.5780161023139954, -0.04567144066095352, -0.3929758071899414, -0.19886109232902527, 0.09783370047807693, 0.20639614760875702, -0.06614221632480621, -0.08298256993293762, 0.2729760408401489, -0.19738008081912994, 0.039316438138484955, -0.06291241943836212, -0.2984917163848877, 0.15413407981395721, 0.6249303221702576, 0.2360731065273285, -0.0714835599064827, 0.4156576693058014, 0.050727155059576035, 0.6994937658309937, 0.0851803570985794, -0.3150876462459564, 0.452667772769928, -0.031725890934467316, 0.177195742726326, -0.13752251863479614, -0.3402392268180847, 0.0954924076795578, 0.237472265958786, -0.04053609445691109, 0.15654617547988892, -0.028311729431152344, -0.08319411426782608, -0.4500623941421509, 0.2251792550086975, -0.25933727622032166, 0.04975133761763573, 0.014600490219891071, 0.08145266771316528, 0.0388733446598053, -0.09079229831695557, -0.2421228587627411, -0.3565264940261841, -0.5098557472229004, -0.1295813024044037, 0.2505680322647095, -0.2341906726360321, 0.18313723802566528, -0.652407705783844, -0.10370753705501556, -0.6693298816680908, 0.29620012640953064, -0.14688852429389954, 0.275445818901062, -0.21646232903003693, 0.18662288784980774, 0.13347949087619781, -0.08640041947364807, 0.08926220238208771, -0.4160813093185425, -0.17348675429821014, -0.09920832514762878, 0.2953800559043884, -0.4609887897968292, -0.04646923020482063, -0.03333190828561783, 0.0832534208893776, 0.48458757996559143, 0.21635566651821136, 0.020407967269420624, 0.09795549511909485, -0.17281484603881836, -0.2555910646915436, -0.16451434791088104, -0.22752930223941803, -0.03326965123414993, -0.6557900309562683, -0.34370648860931396, -0.4793667197227478, 0.0711589902639389, 0.1532273292541504, 0.15192006528377533, -0.07051819562911987, 0.22515353560447693, 0.11375264078378677, 0.49375730752944946, 0.017786361277103424, -0.08321520686149597, 0.6168342232704163, -0.12393531203269958, -0.46655580401420593, 0.4901573956012726, -0.06203443557024002, 0.38743916153907776, 0.07472115755081177, -0.13771258294582367, -0.15648922324180603, -0.30095186829566956, 0.4412569999694824, 0.1783326119184494, 0.10966116189956665, 0.21357375383377075, -0.13211598992347717, 0.052149541676044464, -0.19166411459445953, 0.47335904836654663, 0.33841753005981445, -0.09716370701789856, -0.3771507740020752, -0.13972985744476318, 0.44671115279197693, -0.2157982885837555, -0.00792153924703598, 0.23950056731700897, -0.21200771629810333, -0.28229597210884094, 0.21089628338813782, 0.17036598920822144, 1.136423945426941, -0.07740247249603271, 0.37175992131233215, 0.13204941153526306, -0.054556190967559814, 0.756166398525238, -0.1896362453699112, 0.02476353570818901, -0.24850235879421234, 0.1745956540107727, -0.1003190279006958, -0.04481516778469086, -0.08898066729307175, 0.11526520550251007, -0.22819115221500397, 0.05740367993712425, -0.04674191027879715, -0.16363725066184998, 0.007388378493487835, 0.541745126247406, 0.09941393882036209, -0.3780415654182434, -0.20321010053157806, 0.10226583480834961, -0.33484190702438354, 0.3006332814693451, -0.1681475043296814, -0.23469935357570648, -0.32431986927986145, -0.16216182708740234, -0.1864946484565735, 0.30506080389022827, -0.23749369382858276, 0.08102436363697052, 0.05311499908566475, -0.3184688687324524, 0.2553342878818512, 0.4358457326889038, 0.5033403635025024, -0.10633139312267303, -0.07909460365772247, 0.030860640108585358, -0.13777166604995728, -0.02995627373456955, 0.00021415879018604755, 0.07229562103748322, 0.17591863870620728, -0.3124390244483948, 0.065276600420475, 0.29665830731391907, -0.014302889816462994, -0.3096221089363098, -0.17401351034641266, 0.1471676081418991, 0.07625976204872131, -0.14346034824848175, -0.31247609853744507, 0.0992087796330452, -0.17771460115909576, -0.02474302425980568, 0.04121081903576851, 0.06617677956819534, 0.03593987599015236, -0.0725613459944725, 0.0739264264702797, -0.1873590499162674, 0.06512042880058289, 0.07385537773370743, -0.1152709499001503, -0.06169385835528374, 0.7477415800094604, -0.06718678772449493, 0.09660480916500092, -0.09921880066394806, -0.23556292057037354, -0.3396826982498169, -0.18870624899864197, -0.10450927913188934, 0.2684950828552246, -0.045154228806495667, 0.0695589929819107, 0.6650158166885376, 0.3328905701637268, 0.1416175812482834, 0.16352134943008423, -0.15161074697971344, -0.33256766200065613, -0.340668261051178, -0.033449482172727585, 0.16924771666526794, -0.08269958198070526, 0.19785895943641663, 0.20175224542617798, 0.3437700569629669, -0.2551347017288208, -0.17512626945972443, -0.21311801671981812, 0.1651635468006134, 0.7187833786010742, -0.05416584759950638, -0.061099231243133545, -0.019157908856868744, 0.003239557147026062, -0.04416728764772415, -0.16698211431503296, -0.13342903554439545, -0.37222737073898315, 0.16226866841316223, 0.4754083454608917, -0.20718491077423096, -0.07411116361618042, -0.2015528529882431, -0.10357825458049774, -0.1250372678041458, -0.31520020961761475, 0.040920644998550415, -0.05833368003368378, 0.21087133884429932, -0.24682410061359406, 0.3526257276535034, -0.08808554708957672, 0.1950051486492157, 0.019292578101158142, 0.18320316076278687, 0.07253194600343704, -0.024480484426021576, 0.16411766409873962, 0.05056805908679962, -0.3545888662338257, -0.031485170125961304, -0.20351509749889374, -0.09130924195051193, 0.047117121517658234, -0.4444262981414795, 0.11381736397743225, -0.16836635768413544, 0.04783853143453598, 0.02164560556411743, -0.5189085602760315, 0.14512787759304047, 0.38480237126350403, 0.12011395394802094, 0.04832223057746887, 0.12300294637680054, 0.16844356060028076, 0.21478132903575897, -0.010158903896808624, 0.23202203214168549, 0.062191080302000046, -0.27599048614501953, -0.05973254144191742, -0.17764469981193542, -0.10463622957468033, -0.2630261182785034, 0.0702865868806839, -0.1269129514694214, -0.40820831060409546, -0.051093656569719315, 0.13153283298015594, 0.17131702601909637, -0.11073245853185654, 0.1462876796722412, -0.21450954675674438, 0.09462955594062805, 0.02808244712650776, -0.08880074322223663, 0.2282724678516388, -0.2685781419277191, 0.0781366154551506, -0.10730597376823425, -0.04195278882980347, 0.3354443311691284, 0.14839155972003937, 0.6253676414489746, -0.30057933926582336, -0.02767845056951046, -0.24202121794223785, 0.14442509412765503, -0.0007230639457702637, 0.176571786403656, -0.6177634000778198, -0.02887548878788948, 0.05070360004901886, -0.10464958846569061, 0.028326144441962242, -0.34961193799972534, -0.06082187592983246, 0.30982959270477295, -0.1774180382490158, 0.11428431421518326, -0.15285083651542664, 0.21429786086082458, 0.14840368926525116, -0.15133216977119446, 0.42992523312568665, 0.05962705612182617, -0.10391737520694733, -0.08217445015907288, 0.07708925753831863, 0.514744222164154, 0.582910418510437, 0.11593963205814362, 0.31674543023109436, 0.21503444015979767, 0.028208373114466667, 0.07165069878101349, 0.21679043769836426, 0.43436098098754883, 0.2830864191055298, 0.12656116485595703, 0.047754548490047455, -0.1217484325170517, 0.07318644225597382, 0.05191686376929283, 0.12302322685718536, -0.27411791682243347, 0.3007913827896118, 0.001267947256565094, -0.37161657214164734, -0.192124605178833, 0.3825390338897705, -0.46314090490341187, 0.03657793626189232, 0.21874840557575226, 0.2108299285173416, 0.3433954417705536, -0.27728167176246643, 0.03888614475727081, -0.16345824301242828, 0.436238557100296, 0.4967517554759979, -0.0677364245057106, -0.12502843141555786, -0.28918662667274475, -0.4506284296512604, 0.18131105601787567, -0.19864442944526672, -0.22995272278785706, -0.35685494542121887, 0.19610658288002014, 0.04848911613225937, -0.10085813701152802, 0.053834907710552216, 0.09990924596786499, -0.21706204116344452, -0.016972936689853668, -0.4612075686454773, 0.06106337159872055, -0.15878146886825562, -0.01481538824737072, -0.028728902339935303, -0.14931774139404297, 0.0639244019985199, 0.07336713373661041, 0.04292559623718262, 0.02961556613445282, 0.21092569828033447, 0.10708600282669067, 0.38921472430229187, 0.47182729840278625, -0.10158546268939972, 0.7334182262420654, 0.08890098333358765, -0.0027491040527820587, 0.012416567653417587, -0.05101092532277107, -0.17123188078403473, 0.029724419116973877, 0.14156678318977356, 0.4371802806854248, -0.0835019201040268, 0.3541299104690552, -0.21917669475078583, 0.05322762951254845, -0.0519561730325222, -0.5288751125335693, 0.10355675965547562, -0.00581617746502161, -0.03416029363870621, 0.2721140384674072, 0.40909668803215027, 0.11103549599647522, 0.11265715211629868, -0.10881724953651428, -0.4588083028793335, -0.28008833527565, 0.39896130561828613, -0.5402222275733948, -0.30159831047058105, 0.19581717252731323, 0.23403993248939514, -0.06005135923624039, 0.04783665016293526, -0.3058381676673889, -0.059672996401786804, 0.16777807474136353, -0.21494102478027344, -0.00905650109052658, 0.31976640224456787, -0.47192636132240295, 0.07732565701007843, -0.032319843769073486, 0.273898184299469, 0.20475369691848755, -0.3582354485988617, -0.03778167441487312, -0.1672145426273346 ]
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.
[ 0.1084665060043335, -0.044079914689064026, 0.0504009984433651, -0.02760537713766098, 0.138132244348526, -0.18117070198059082, 0.32669156789779663, -0.009185023605823517, -0.19502247869968414, 0.1800151765346527, -0.12995530664920807, 0.23764531314373016, -0.23067624866962433, 0.291965514421463, 0.3679998517036438, 0.07485648989677429, 0.006218485534191132, 0.2890029549598694, 0.01184871792793274, -0.10124002397060394, -0.21110135316848755, 0.10817155241966248, 0.03599921613931656, -0.0665363222360611, -0.12294409424066544, -0.17405103147029877, -0.10020700842142105, 0.2709645926952362, -0.2011130154132843, -0.3351973295211792, 0.4402619004249573, 0.2242947369813919, 0.3215375244617462, 0.5771903991699219, -0.00011768119293265045, -0.05328431352972984, 0.13914427161216736, -0.22553709149360657, -0.6261972188949585, -0.3673126697540283, 0.08496671169996262, -0.1735265552997589, 0.08267300575971603, -0.2572464644908905, 0.18322902917861938, 0.025091547518968582, 0.008598476648330688, 0.07702229917049408, 0.5234442353248596, 0.36876606941223145, 0.154951274394989, 0.0958213210105896, -0.18301622569561005, -0.09871860593557358, 0.029122356325387955, 0.18727988004684448, -0.04247978329658508, 0.2965947985649109, 0.36872488260269165, -0.39618468284606934, 0.30293774604797363, -0.1861633062362671, -0.008991852402687073, -0.0887327790260315, 0.48444414138793945, -0.022030655294656754, 0.00685080885887146, -0.40278056263923645, -0.008083105087280273, 0.1580434888601303, 0.2721641957759857, -0.03827735036611557, -0.34664639830589294, -0.11687016487121582, 0.1563587188720703, -0.4871675968170166, 0.24212101101875305, 0.283277690410614, -0.4218556582927704, -0.04651624336838722, -0.2582417130470276, 0.3114124834537506, -0.4107930660247803, 0.4234354496002197, 0.12531885504722595, 0.3503040373325348, -0.026141155511140823, 0.0610152892768383, 0.25930115580558777, -0.1453033983707428, -0.021243542432785034, 0.0925709456205368, 0.017637968063354492, 0.28195831179618835, -0.19193212687969208, -0.1905704289674759, -0.021574191749095917, -0.2899375557899475, -0.10397440195083618, 0.1091003343462944, 0.4545277953147888, -0.1308465600013733, -0.22314785420894623, 0.15165236592292786, 0.08718313276767731, 0.3312406539916992, 0.008624903857707977, 0.07960406690835953, 0.11092212796211243, 0.01273290440440178, 0.2880062758922577, -0.0693008080124855, -0.27576762437820435, -0.07849974185228348, -0.13718947768211365, -0.048890382051467896, 0.2654901146888733, 0.0865570604801178, 0.0027185995131731033, -0.08611766248941422, 0.07163922488689423, 0.03991120681166649, -0.07255170494318008, 0.3346664309501648, -0.09973806887865067, -0.2767598330974579, 0.008246410638093948, 0.24410714209079742, 0.12901820242404938, -0.2690316140651703, 0.07090892642736435, 0.3014501631259918, -0.12323862314224243, -0.14110568165779114, 0.41318556666374207, 0.0026035141199827194, 0.19098681211471558, -0.0497686043381691, 0.17416974902153015, -0.11572152376174927, 0.07259533554315567, 0.026877503842115402, -0.1545821726322174, 0.026117458939552307, 0.23558557033538818, 0.3750022351741791, -0.07226434350013733, -0.5426658987998962, -0.1255454421043396, 0.04687352478504181, -0.37620729207992554, -0.3481234908103943, -0.44942307472229004, 0.09366775304079056, 0.03533300384879112, -0.2343597114086151, -0.2903655171394348, -0.0844537615776062, 0.09920261800289154, -0.2737710475921631, -0.08804372698068619, -0.01781938597559929, -0.10808943212032318, -0.14024388790130615, 0.11493652313947678, 0.05398008972406387, -0.48611342906951904, 0.20890986919403076, -0.06266934424638748, 0.5137913823127747, 0.3013725280761719, 0.2810112237930298, -0.3043970465660095, 0.25999101996421814, -0.06236653029918671, 0.1773788183927536, 0.5754057168960571, -0.4743399918079376, -0.2507370710372925, 0.03427904471755028, -0.0365610271692276, -0.10301307588815689, 0.05468665063381195, 0.16820453107357025, 0.08836285769939423, -0.10266026854515076, 0.07413911074399948, 0.40953683853149414, 0.07595818489789963, 0.030012978240847588, -0.22451412677764893, -0.17478139698505402, 0.2542687654495239, 0.012387309223413467, -0.1504034847021103, 0.07483947277069092, 0.2523055374622345, 0.08474501222372055, 0.03077039122581482, -0.33520933985710144, 0.1500527262687683, 0.20149922370910645, 0.33596470952033997, -0.13525721430778503, -0.13722960650920868, 0.2380940318107605, -0.4048444926738739, 0.19656231999397278, 0.12716268002986908, 0.35962817072868347, -0.11083972454071045, 0.006236918270587921, -0.42009249329566956, -0.11388057470321655, -0.09573781490325928, -0.031772542744874954, 0.06666681170463562, 0.0963793694972992, 0.0702921450138092, -0.1855904757976532, -0.23020364344120026, 0.12882034480571747, -0.3153885006904602, 0.0610581673681736, -0.1971268504858017, 0.05917049199342728, -0.2863607108592987, -0.11160078644752502, 0.2181827425956726, 0.17532090842723846, 0.2073209285736084, -0.01801915466785431, -0.17255359888076782, 0.08949821442365646, -0.46536678075790405, 0.3087373375892639, 0.012427110224962234, 0.1415526568889618, 0.03356895595788956, -0.021583683788776398, -0.1996210217475891, -0.11788231134414673, 0.1772724688053131, -0.1705971658229828, -0.3431077003479004, 0.14997297525405884, -0.03314923122525215, 0.019479043781757355, 0.166823610663414, -0.002482302486896515, 0.27056747674942017, -0.1924484372138977, 0.1081031784415245, 0.11901304125785828, 0.3931277096271515, 0.1005677580833435, 0.37359166145324707, -0.06235010176897049, -0.12617242336273193, -0.08031881600618362, 0.17376305162906647, 0.014593619853258133, 0.15850797295570374, 0.0684889629483223, 0.033259306102991104, -0.07810258865356445, 0.286007821559906, 0.6209808588027954, 0.3515610992908478, 0.09832146763801575, -0.04840387776494026, -0.050611890852451324, -0.11802095919847488, -0.22260472178459167, -0.03233370929956436, -0.08653303980827332, 0.0450492687523365, 0.2136896252632141, 0.429272323846817, 0.2628838121891022, -0.13860444724559784, 0.002830352634191513, 0.11219204217195511, 0.44106993079185486, -0.18067596852779388, 0.24720104038715363, -0.11076143383979797, -0.4309421181678772, 0.23817774653434753, -0.28240883350372314, -0.08930563926696777, -0.16441914439201355, -0.19531381130218506, 0.11842580139636993, 0.1135321706533432, -0.12695875763893127, 0.16629987955093384, 0.3391779661178589, -0.11040570586919785, -0.2544940710067749, -0.27940356731414795, -0.0882849171757698, -0.20852869749069214, 0.052959781140089035, 0.3282940983772278, -0.11787300556898117, 0.4269642233848572, 0.3558061420917511, 0.06698235869407654, -0.0482543408870697, -0.34660592675209045, -0.012719795107841492, -0.06370146572589874, 0.07861366868019104, 0.08492837101221085, 0.34389036893844604, -0.0008335113525390625, -0.5517429709434509, 0.1905890703201294, 0.06303217262029648, -0.2525031268596649, 0.2460184097290039, -0.02194218523800373, -0.3063017725944519, -0.4941287636756897, -0.007458776235580444, -0.2783518433570862, -0.3438790738582611, 0.2504080533981323, 0.35123640298843384, 0.2351839542388916, 0.6135790348052979, 0.02082943171262741, 0.13948358595371246, 0.3170113265514374, 0.15296348929405212, 0.03316979855298996, -0.1874237060546875, 0.4657979905605316, -0.253059983253479, -0.43299925327301025, 0.11046834290027618, -0.13290530443191528, 0.5748026371002197, -0.04160178080201149, -0.3659118413925171, -0.19975851476192474, 0.09144090861082077, 0.18816472589969635, -0.07725890725851059, -0.06613492965698242, 0.285658597946167, -0.18216541409492493, 0.042108096182346344, -0.05911857634782791, -0.31170493364334106, 0.15016894042491913, 0.6340498924255371, 0.2549777328968048, -0.07658519595861435, 0.42161300778388977, 0.04242275282740593, 0.7029532790184021, 0.08945327997207642, -0.3097898066043854, 0.4549664556980133, -0.02821935899555683, 0.19479677081108093, -0.13296857476234436, -0.35792699456214905, 0.09596803784370422, 0.23466458916664124, -0.030786124989390373, 0.13775259256362915, -0.027219802141189575, -0.06776422262191772, -0.4466431438922882, 0.24750655889511108, -0.25212863087654114, 0.06079145893454552, -0.001671820878982544, 0.10914751887321472, 0.04245893284678459, -0.10231339931488037, -0.23655301332473755, -0.36565423011779785, -0.5041987895965576, -0.1292748749256134, 0.24986596405506134, -0.2540324926376343, 0.196914941072464, -0.6588920950889587, -0.12374027073383331, -0.6575385332107544, 0.2880603075027466, -0.14811068773269653, 0.259455144405365, -0.21375088393688202, 0.17713740468025208, 0.13104259967803955, -0.08934599161148071, 0.11717312783002853, -0.41821685433387756, -0.1867833435535431, -0.09565195441246033, 0.2866274416446686, -0.47739729285240173, -0.053471412509679794, -0.043761562556028366, 0.08510807156562805, 0.46540355682373047, 0.22015412151813507, 0.03126198798418045, 0.09527502954006195, -0.1582922786474228, -0.24406841397285461, -0.1730845719575882, -0.23456303775310516, -0.04078027978539467, -0.6392326354980469, -0.337543785572052, -0.48867690563201904, 0.04947779327630997, 0.15704035758972168, 0.16081547737121582, -0.07531023025512695, 0.2124437689781189, 0.11879842728376389, 0.4977753758430481, 0.022433698177337646, -0.06972028315067291, 0.6355297565460205, -0.12254385650157928, -0.44986554980278015, 0.48968544602394104, -0.05355869233608246, 0.3719123303890228, 0.050603754818439484, -0.12606185674667358, -0.18033042550086975, -0.2989444136619568, 0.46419769525527954, 0.18745748698711395, 0.1153121218085289, 0.2137036919593811, -0.11608603596687317, 0.04945141077041626, -0.18874646723270416, 0.47350189089775085, 0.3472863435745239, -0.09278992563486099, -0.37483134865760803, -0.16618876159191132, 0.44636306166648865, -0.18877369165420532, -0.0016305968165397644, 0.2298571616411209, -0.20053787529468536, -0.28849026560783386, 0.19627487659454346, 0.17903289198875427, 1.135075330734253, -0.09616318345069885, 0.3786777853965759, 0.13602018356323242, -0.05242539197206497, 0.7858077883720398, -0.18570801615715027, 0.03413783758878708, -0.24208010733127594, 0.16550885140895844, -0.102665975689888, -0.03763774782419205, -0.08714558929204941, 0.10502208769321442, -0.239122673869133, 0.0672181248664856, -0.057275936007499695, -0.1655508577823639, -0.006034396588802338, 0.5472075939178467, 0.09571055322885513, -0.3948436379432678, -0.19912922382354736, 0.10654103755950928, -0.33136850595474243, 0.29732832312583923, -0.1884973645210266, -0.2347911149263382, -0.2989426553249359, -0.15512239933013916, -0.1712501049041748, 0.3010401129722595, -0.23888637125492096, 0.08907851576805115, 0.05478129908442497, -0.32584255933761597, 0.2452925741672516, 0.42029932141304016, 0.5057651400566101, -0.09477053582668304, -0.07163000851869583, 0.015904512256383896, -0.12948498129844666, -0.058059338480234146, -0.0169940534979105, 0.07856112718582153, 0.20451369881629944, -0.3050251603126526, 0.06459299474954605, 0.2832258343696594, -0.004202867858111858, -0.2958987057209015, -0.1927383840084076, 0.17007604241371155, 0.1059855967760086, -0.1450621485710144, -0.32288140058517456, 0.10752259194850922, -0.1985766589641571, -0.023928284645080566, 0.043140217661857605, 0.052143506705760956, 0.02309335023164749, -0.06285258382558823, 0.07580910623073578, -0.18461138010025024, 0.06752069294452667, 0.09131070226430893, -0.08641165494918823, -0.08536329865455627, 0.7531559467315674, -0.06545592844486237, 0.10210259258747101, -0.09430181980133057, -0.23572054505348206, -0.34444108605384827, -0.16373026371002197, -0.10431723296642303, 0.2887050211429596, -0.04541238397359848, 0.10916142910718918, 0.6310181021690369, 0.3479662537574768, 0.11863772571086884, 0.14224323630332947, -0.13004641234874725, -0.33897319436073303, -0.3504675626754761, -0.02279360219836235, 0.17645163834095, -0.09358379244804382, 0.19203655421733856, 0.20667418837547302, 0.33863088488578796, -0.25417274236679077, -0.18684646487236023, -0.18249282240867615, 0.1629965603351593, 0.7249685525894165, -0.04661892354488373, -0.06739471852779388, -0.02037668228149414, 0.001918857917189598, -0.04075629264116287, -0.16854192316532135, -0.13428905606269836, -0.3807846009731293, 0.16128723323345184, 0.479666143655777, -0.20250077545642853, -0.07540307939052582, -0.20506326854228973, -0.09480388462543488, -0.11984419822692871, -0.30565738677978516, 0.04392385482788086, -0.04808367043733597, 0.22156155109405518, -0.23215968906879425, 0.3371523320674896, -0.0559341162443161, 0.18995903432369232, 0.030670665204524994, 0.17264893651008606, 0.07241877168416977, -0.034020863473415375, 0.16806046664714813, 0.05748865008354187, -0.35415589809417725, -0.032752133905887604, -0.1912262886762619, -0.10384615510702133, 0.029014982283115387, -0.4462992548942566, 0.10618840157985687, -0.17603501677513123, 0.0497995987534523, 0.022434327751398087, -0.4888646602630615, 0.13809168338775635, 0.38951027393341064, 0.11818037182092667, 0.03648409992456436, 0.11728770285844803, 0.1646391749382019, 0.21876661479473114, -0.018171388655900955, 0.22868025302886963, 0.04256518557667732, -0.26980364322662354, -0.0645267516374588, -0.1723543405532837, -0.12308385223150253, -0.2726885676383972, 0.0867752730846405, -0.11697190999984741, -0.3999551832675934, -0.06926880776882172, 0.15473796427249908, 0.1715613603591919, -0.09740515798330307, 0.1438145637512207, -0.2051553726196289, 0.10718590021133423, 0.03388480469584465, -0.08608280122280121, 0.2209549844264984, -0.27044281363487244, 0.08013659715652466, -0.08750225603580475, -0.02710743248462677, 0.3488439619541168, 0.14098060131072998, 0.6430832147598267, -0.3007907271385193, -0.03844211995601654, -0.23728394508361816, 0.12414363026618958, -0.0008916668593883514, 0.19300581514835358, -0.6118700504302979, -0.016115702688694, 0.040175892412662506, -0.10860534012317657, 0.03986580669879913, -0.354142963886261, -0.06759509444236755, 0.30947357416152954, -0.17172743380069733, 0.113868348300457, -0.16982093453407288, 0.20617780089378357, 0.14696474373340607, -0.1484517753124237, 0.39706525206565857, 0.05424065515398979, -0.103056401014328, -0.07449620962142944, 0.08739249408245087, 0.5002301931381226, 0.6005567312240601, 0.11638952791690826, 0.32311075925827026, 0.201610267162323, 0.038133375346660614, 0.07256466895341873, 0.20181415975093842, 0.4283694922924042, 0.2745617926120758, 0.10998043417930603, 0.04795938730239868, -0.12619179487228394, 0.06361710280179977, 0.04795245826244354, 0.09902055561542511, -0.29639700055122375, 0.29422900080680847, 0.0014723017811775208, -0.3734475374221802, -0.19348110258579254, 0.38225844502449036, -0.4880560636520386, 0.0280253104865551, 0.2486959993839264, 0.21553656458854675, 0.33565667271614075, -0.2708441913127899, 0.03663018345832825, -0.16707256436347961, 0.4374946355819702, 0.482622891664505, -0.09399565309286118, -0.12573982775211334, -0.2834654152393341, -0.4656837582588196, 0.17053553462028503, -0.1989501565694809, -0.22324541211128235, -0.3570995330810547, 0.20041115581989288, 0.04286929592490196, -0.1253056675195694, 0.07186217606067657, 0.0930996760725975, -0.22871442139148712, -0.058963850140571594, -0.4724920392036438, 0.05648234859108925, -0.13548052310943604, -0.021281642839312553, -0.02528080716729164, -0.137778639793396, 0.050760988146066666, 0.05550025776028633, 0.034974582493305206, 0.02594171278178692, 0.21244016289710999, 0.10449182987213135, 0.3972402811050415, 0.4888581931591034, -0.1169951856136322, 0.7418338060379028, 0.09344252198934555, -0.010376179590821266, 0.01193920150399208, -0.037465810775756836, -0.1898796558380127, 0.05390319973230362, 0.12369830906391144, 0.45055848360061646, -0.09263259917497635, 0.34791991114616394, -0.2200065404176712, 0.07006774097681046, -0.04054585099220276, -0.5381631851196289, 0.11118822544813156, -0.003120427019894123, -0.031219709664583206, 0.26711276173591614, 0.4321092367172241, 0.12411181628704071, 0.11082868278026581, -0.10501448065042496, -0.4677984118461609, -0.2764584422111511, 0.40621381998062134, -0.5333334803581238, -0.29643017053604126, 0.18773044645786285, 0.23693424463272095, -0.08291321247816086, 0.03613553196191788, -0.2994180917739868, -0.07017689943313599, 0.16325369477272034, -0.22941243648529053, -0.012820586562156677, 0.32174599170684814, -0.4839877784252167, 0.06313890218734741, -0.036177389323711395, 0.24127407371997833, 0.22410348057746887, -0.35660433769226074, -0.03303779661655426, -0.15160374343395233 ]
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.
[ 0.043388914316892624, -0.198110893368721, 0.030185792595148087, 0.21069124341011047, 0.1068333089351654, 0.08471900969743729, 0.549275815486908, 0.19509577751159668, -0.014409476891160011, 0.15259361267089844, 0.0023246072232723236, 0.45175233483314514, -0.27085140347480774, 0.3255721926689148, 0.019077958539128304, -0.4061383306980133, -0.06452716141939163, 0.06130611523985863, 0.23302070796489716, 0.10228583216667175, 0.09767566621303558, 0.22732265293598175, -0.03829368203878403, 0.2814406752586365, -0.23800937831401825, -0.2281857430934906, -0.20587792992591858, -0.16539832949638367, 0.018803222104907036, -0.4043419063091278, 0.1754712015390396, 0.1728925257921219, -0.2417057603597641, 0.03427813947200775, -0.00011900685785803944, -0.1262153685092926, 0.21351443231105804, -0.009097915142774582, -0.1567249447107315, -0.17348860204219818, -0.5583314895629883, -0.4940270185470581, 0.4417521059513092, -0.5981582403182983, -0.004088187590241432, -0.06665784120559692, -0.009362993761897087, 0.2533454895019531, 0.5818644165992737, -0.16109618544578552, 0.09534082561731339, -0.3728165924549103, -0.0064572542905807495, 0.12188389152288437, 0.30312296748161316, 0.1706283688545227, -0.32061243057250977, -0.33287787437438965, -0.19216209650039673, 0.01641998067498207, -0.014343135058879852, 0.2186739593744278, 0.0672658309340477, -0.14753015339374542, 0.12505751848220825, 0.20889079570770264, -0.18132513761520386, -0.16720527410507202, -0.06251548230648041, 0.3708769381046295, -0.165843665599823, -0.1862097680568695, -0.21817892789840698, -0.7446471452713013, 0.192808598279953, -0.3991025388240814, 0.009476587176322937, 0.2820800840854645, -0.02325771376490593, 0.056905366480350494, 0.2280941605567932, -0.3026104271411896, -0.17334744334220886, 0.07937277853488922, 0.15809093415737152, 0.614372968673706, 0.08293669670820236, 0.20643454790115356, 0.19808414578437805, 0.0196453295648098, 0.19974549114704132, 0.13025815784931183, -0.04197729006409645, 0.18368537724018097, -0.2608940303325653, -0.3455762267112732, -0.39284220337867737, -0.08230053633451462, -0.21530193090438843, 0.1477324217557907, 0.2048051804304123, -0.037261031568050385, 0.08624116331338882, 0.1348695158958435, 0.2460688203573227, 0.07950559258460999, 0.430012971162796, 0.6358774900436401, 0.11269094794988632, -0.01923733949661255, -0.360803484916687, -0.02425123006105423, -0.06769773364067078, 0.061923712491989136, -0.17635348439216614, 0.10095915198326111, -0.09962873160839081, 0.23635709285736084, -0.0581112802028656, -0.2936745285987854, -0.11544933915138245, -0.29221248626708984, 0.27776002883911133, -0.007407099008560181, 0.13234083354473114, 0.0555955171585083, -0.3250805139541626, 0.13766491413116455, -0.19032758474349976, -0.4007329046726227, -0.10605251789093018, 0.12324117124080658, -0.3039294183254242, 0.2852862775325775, -0.025047605857253075, 0.08957436680793762, 0.139749675989151, -0.0239226296544075, -0.17118701338768005, -0.26183968782424927, 0.19998784363269806, -0.030773375183343887, 0.2510841488838196, 0.0799311101436615, 0.016892150044441223, 0.05753538757562637, -0.09818610548973083, -0.11058638989925385, -0.3132472336292267, -0.004342339932918549, -0.15221214294433594, -0.3459358811378479, 0.14371636509895325, 0.05959426239132881, 0.14884132146835327, 0.06756255030632019, 0.28562501072883606, 0.6215516924858093, 0.347048819065094, -0.11526219546794891, 0.13643139600753784, 0.05871979892253876, -0.22138524055480957, -0.2414395958185196, -0.02082010544836521, -0.055000826716423035, 0.04045463725924492, -0.4513779878616333, -0.2863801419734955, 0.10820047557353973, -0.10519058257341385, 0.09117934107780457, -0.22937315702438354, 0.11082299053668976, -0.026686040684580803, 0.600189208984375, 0.7154178023338318, -0.2713155448436737, 0.07817188650369644, 0.33046507835388184, 0.07260444760322571, -0.34571951627731323, 0.4756770133972168, 0.0852038636803627, 0.1321296989917755, -0.05762256309390068, -0.3968377113342285, 0.18206700682640076, -0.38011738657951355, -0.1711622029542923, 0.20397420227527618, -0.5321476459503174, 0.16839471459388733, -0.09095190465450287, -0.05552608147263527, -0.3709447383880615, 0.02002197504043579, 0.2963106632232666, 0.46555712819099426, 0.03446836769580841, 0.23717832565307617, -0.17417502403259277, -0.07773013412952423, 0.28939032554626465, -0.17316144704818726, -0.36931225657463074, -0.27011463046073914, -0.010613486170768738, -0.1732831746339798, 0.05678073689341545, 0.013177722692489624, -0.40739625692367554, 0.043092742562294006, -0.27191102504730225, -0.16464731097221375, -0.28500720858573914, 0.046643905341625214, -0.2125971019268036, -0.05866728723049164, -0.34646645188331604, -0.33430343866348267, 0.17123740911483765, -0.0637914389371872, 0.28379327058792114, -0.31552478671073914, -0.1585671454668045, 0.1928495317697525, 0.008334740065038204, -0.23998507857322693, 0.5741865038871765, 0.2506142258644104, -0.025338763371109962, 0.0648970976471901, 0.35172754526138306, 0.38948309421539307, -0.07242327928543091, 0.023375311866402626, -0.12618188560009003, 0.25394028425216675, -0.06732706725597382, -0.08393135666847229, -0.10601688921451569, -0.2638683617115021, -0.14377805590629578, -0.10177119821310043, 0.5019598007202148, 0.12187337875366211, 0.5909390449523926, 0.2901533842086792, 0.05820636451244354, -0.06445089727640152, -0.25854024291038513, -0.1919446587562561, -0.4637323319911957, -0.2635042667388916, -0.37583598494529724, 0.08257988095283508, -0.043832436203956604, -0.2890101671218872, 0.24587520956993103, 0.4475729167461395, 0.14510788023471832, 0.01608213409781456, -0.23104313015937805, 0.11973906308412552, 0.17330697178840637, -0.2040422409772873, 0.7559658885002136, 0.1979500949382782, 0.16951493918895721, -0.09351575374603271, 0.05726432055234909, -0.04233295097947121, -0.18957491219043732, 0.27606847882270813, 0.2820304036140442, 0.07955451309680939, -0.027139633893966675, -0.16981947422027588, -0.14319168031215668, 0.10303781181573868, -0.03333811089396477, 0.22930604219436646, -0.32859787344932556, -0.08614987134933472, -0.07723227888345718, -0.12501002848148346, -0.33749672770500183, -0.6012640595436096, -0.031781863421201706, -0.23478984832763672, -0.24172693490982056, 0.15794692933559418, -0.08859357237815857, -0.21088071167469025, 0.14174984395503998, -0.13971684873104095, 0.13323985040187836, -0.37389951944351196, 0.039972033351659775, 0.14731816947460175, -0.38086020946502686, -0.06988393515348434, 0.02591230720281601, 0.05599711835384369, 0.034937791526317596, 0.26271501183509827, 0.3256538510322571, -0.280883252620697, 0.1503276228904724, -0.335971474647522, 0.17685173451900482, -0.3270226716995239, -0.28942108154296875, 0.22199629247188568, -0.3101726770401001, -0.004219852387905121, -0.46895065903663635, 0.3010503053665161, -0.1061420813202858, -0.10014080256223679, -0.3176913857460022, 0.32188063859939575, 0.21165882050991058, -0.12713532149791718, -0.31906259059906006, -0.4289039671421051, -0.14181466400623322, 0.34559130668640137, -0.07541993260383606, 0.29651618003845215, 0.09804403781890869, -0.42088618874549866, -0.029327325522899628, -0.09281198680400848, -0.11690005660057068, 0.04345462843775749, 0.1445825845003128, -0.08969876170158386, -0.41021329164505005, 0.15662699937820435, 0.05233420804142952, -0.1368255913257599, 0.33396780490875244, -0.17611588537693024, -0.011461332440376282, 0.18883337080478668, -0.013530848547816277, -0.034934185445308685, 0.035726144909858704, 0.11508945375680923, 0.15593788027763367, 0.2736527621746063, 0.06651613116264343, -0.07385837286710739, -0.23240259289741516, 0.4348661005496979, 0.15245786309242249, 0.15153172612190247, -0.0900646522641182, 0.024157926440238953, -0.07993441075086594, 0.6646888256072998, 0.2682938277721405, 0.19962088763713837, -0.1777927577495575, -0.0047391075640916824, -0.34178397059440613, 0.0782979354262352, -0.21145600080490112, 0.1659894585609436, -0.003022134304046631, -0.2993083596229553, 0.3410925269126892, 0.030007313936948776, -0.061492741107940674, -0.25075435638427734, -0.06884309649467468, -0.03625013306736946, -0.5644195675849915, 0.0686144083738327, -0.18404866755008698, 0.21568161249160767, -0.19991736114025116, 0.05245441198348999, 0.003842836245894432, 0.06245703995227814, -0.14635202288627625, -0.15383009612560272, -0.1487346589565277, -0.07147318869829178, -0.11421895027160645, -0.6203206181526184, -0.053815729916095734, -0.042243584990501404, 0.2887144386768341, -0.1428101807832718, -0.021646961569786072, -0.22285938262939453, 0.2955179214477539, 0.30176863074302673, 0.2860284447669983, -0.09012575447559357, -0.16662849485874176, 0.09333692491054535, 0.04214923828840256, 0.19225487112998962, -0.19830764830112457, 0.09100930392742157, 0.029229935258626938, 0.07483071833848953, 0.0393587127327919, -0.572121798992157, -0.29482290148735046, 0.5887439250946045, -0.24235469102859497, -0.3340775668621063, 0.03839493915438652, 0.03843754157423973, 0.13238531351089478, 0.07342898845672607, -0.040468014776706696, 0.10366358608007431, 0.12703490257263184, 0.13721337914466858, 0.08338874578475952, -0.032854288816452026, -0.23436832427978516, 0.08977630734443665, 0.13432177901268005, 0.021047186106443405, -0.05372077226638794, 0.08779174089431763, 0.29639896750450134, 0.265729695558548, -0.3124951422214508, 0.4403763711452484, -0.4000658392906189, -0.09892058372497559, 0.009537859819829464, -0.21707414090633392, 0.538526713848114, 0.3315562605857849, 0.1370118260383606, 0.44719982147216797, 0.005818815901875496, 0.08080242574214935, -0.20520927011966705, -0.06846999377012253, 0.2784636616706848, -0.1718561351299286, -0.4090304970741272, -0.7786413431167603, 0.39444875717163086, 0.09553095698356628, -0.1922818124294281, 0.2213871330022812, 0.1355135291814804, -0.40848296880722046, 0.575567901134491, 0.01621134579181671, 0.884846031665802, -0.07890202105045319, 0.21045660972595215, -0.05240240693092346, -0.2898176312446594, 0.6745566725730896, -0.09591862559318542, 0.13303452730178833, -0.3664886951446533, 0.2945716679096222, 0.057935670018196106, 0.10396677255630493, 0.23142299056053162, 0.4899129271507263, 0.0067724622786045074, 0.23215807974338531, 0.08443000912666321, 0.2629137337207794, -0.3335978090763092, 0.33951884508132935, 0.14325380325317383, -0.11598354578018188, -0.08670628815889359, -0.04819542542099953, 0.001482393592596054, -0.24703145027160645, 0.20888113975524902, -0.19910165667533875, 0.016625676304101944, 0.17032156884670258, -0.25539788603782654, -0.006950175389647484, -0.12027204781770706, 0.2119198739528656, -0.12011604011058807, -0.415666401386261, 0.39140909910202026, 0.17767199873924255, 0.23396167159080505, -0.16502225399017334, 0.019226256757974625, 0.4578539729118347, -0.042343251407146454, 0.09356638044118881, -0.1717168092727661, -0.01251368410885334, 0.3011681139469147, 0.053446538746356964, 0.027400396764278412, 0.031027600169181824, 0.018673676997423172, -0.08838097751140594, -0.22427859902381897, 0.23940733075141907, 0.44946640729904175, -0.515265166759491, 0.20778274536132812, 0.031566496938467026, 0.001956336200237274, -0.17893292009830475, 0.02122216485440731, -0.21939076483249664, -0.24939391016960144, 0.04010827839374542, -0.17997564375400543, -0.06843467801809311, 0.004961658269166946, -0.319285124540329, 0.1781570017337799, 0.007905263453722, 0.38825592398643494, -0.18148966133594513, 0.1604386866092682, -0.23996278643608093, 0.14632421731948853, 0.15137812495231628, -0.2061508595943451, 0.2905280292034149, 0.1530693769454956, -0.26518458127975464, 0.13528470695018768, -0.019382257014513016, 0.0580952987074852, 0.6687648892402649, -0.1800166666507721, 0.03185361623764038, -0.2208247184753418, -0.04019639641046524, -0.10357333719730377, 0.09732288867235184, 0.0068293847143650055, 0.4960913062095642, -0.12097837775945663, 0.40659427642822266, -0.1664101928472519, -0.09653301537036896, 0.20132718980312347, 0.25393733382225037, -0.10842792689800262, 0.017032146453857422, 0.15272291004657745, 0.1292618364095688, 0.020380837842822075, -0.05497113615274429, 0.1774236112833023, -0.08827248215675354, -0.3262885808944702, 0.11676814407110214, 0.09450903534889221, 0.20274858176708221, -0.36953088641166687, -0.14116649329662323, 0.10343480110168457, -0.15560398995876312, 0.20738093554973602, 0.06280753016471863, 0.26048657298088074, 0.5012200474739075, 0.27072247862815857, 0.010650262236595154, -0.37880629301071167, 0.21698929369449615, -0.12480489909648895, 0.3370317816734314, 0.25255095958709717, 0.3109065294265747, -0.2106352150440216, 0.07289381325244904, 0.13318601250648499, -0.05577515810728073, 0.1861095130443573, -0.27954769134521484, -0.07307413965463638, 0.036065809428691864, -0.09714667499065399, -0.14056943356990814, 0.21551014482975006, 0.4737994372844696, 0.13721752166748047, -0.03496857360005379, 0.22364720702171326, 0.11047765612602234, -0.07229302823543549, 0.21312935650348663, 0.2800595760345459, -0.17977799475193024, 0.09948089718818665, 0.5531473755836487, 0.11321595311164856, 0.2622583508491516, -0.4264291524887085, -0.06307681649923325, -0.10696753114461899, -0.030634194612503052, 0.5049614906311035, 0.8284043073654175, 0.21459437906742096, -0.06918095052242279, 0.4637245833873749, -0.18284429609775543, 0.26808011531829834, 0.45255982875823975, -0.18141058087348938, 0.3613757789134979, 0.3433874845504761, -0.17254184186458588, 0.5454283952713013, 0.10762101411819458, -0.3695584535598755, 0.04664392024278641, -0.058181554079055786, -0.24876540899276733, -0.12223467230796814, 0.7737444639205933, 0.04387817531824112, -0.3321119248867035, 0.17119184136390686, 0.3155762553215027, 0.1680946946144104, -0.25017011165618896, -0.1361531913280487, 0.11348176747560501, -0.3541952073574066, 0.0006027892231941223, -0.054300371557474136, -0.22777579724788666, 0.35432401299476624, 0.08226707577705383, 0.30091047286987305, -0.09902681410312653, -0.06769798696041107, 0.21079182624816895, 0.4752942621707916, -0.3517262637615204, -0.13512469828128815, -0.01743301749229431, -0.29152241349220276, 0.1551801860332489, -0.07194473594427109, 0.0081624835729599, -0.13468003273010254, -0.25497835874557495, -0.2885667085647583, 0.2460688203573227, -0.10548322647809982, 0.023284662514925003, -0.06977830082178116, 0.12184923142194748, 0.29119667410850525, 0.21185970306396484, -0.02308325469493866, -0.03540932014584541, -0.20721466839313507, 0.2606799006462097, -0.018992993980646133, -0.05991321802139282, 0.5498867034912109, 0.2398156374692917, -0.187866672873497, -0.025995541363954544, 0.22880128026008606, -0.29212087392807007, -0.01780359074473381, 0.35891595482826233, -0.06609861552715302, 0.08894039690494537, -0.16790629923343658, -0.022721940651535988, -0.021172359585762024, 0.4855484962463379, -0.22243981063365936, -0.20367199182510376, -0.27540645003318787, -0.3702969551086426, -0.022414423525333405, -0.14031383395195007, -0.12521381676197052, -0.2741914391517639, -0.13138236105442047, 0.16895127296447754, -0.15706154704093933, -0.16794607043266296, 0.16192831099033356, 0.04043976590037346, 0.4329414367675781, 0.33852577209472656, -0.23166562616825104, 0.0057669468224048615, -0.02158721536397934, 0.14577631652355194, 0.1738491654396057, 0.13576942682266235, -0.07442143559455872, 0.3656339645385742, -0.2971503734588623, 0.15661509335041046, -0.022885922342538834, 0.5849363803863525, 0.13453072309494019, 0.0016774162650108337, -0.02073596976697445, 0.2275300770998001, 0.029862627387046814, 0.03711678087711334, 0.040696509182453156, 0.012698505073785782, -0.5554834008216858, 0.4114270806312561, 0.06802884489297867, 0.1626555323600769, -0.27482008934020996, 0.41206029057502747, -0.14148388803005219, 0.0037246302235871553, -0.23200519382953644, 0.10678331553936005, -0.01392391137778759, 0.19690781831741333, -0.02301153540611267, 0.29645177721977234, 0.37596166133880615, 0.2861540913581848, -0.1174602061510086, -0.06586804240942001, -0.13472887873649597, -0.3783159852027893, 0.16204290091991425, -0.036627303808927536, -0.09991857409477234, 0.08973997831344604, 0.33894962072372437, 0.19934052228927612, 0.11615044623613358, -0.7652568221092224, -0.2747171223163605, 0.18879801034927368, -0.24032485485076904, -0.024294890463352203, 0.14281335473060608, 0.2829599678516388, -0.16623690724372864, 0.01470622792840004, -0.13825127482414246, 0.27970483899116516, -0.06312927603721619, 0.030536312609910965, -0.14354068040847778 ]
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?
[ 0.10458482801914215, -0.09824596345424652, 0.027599211782217026, 0.22247891128063202, 0.11074329912662506, 0.1121133491396904, 0.5969955921173096, 0.23253045976161957, -0.035822078585624695, 0.11198772490024567, 0.01879112422466278, 0.4395623803138733, -0.21577893197536469, 0.25080645084381104, -0.04807367920875549, -0.3708149194717407, -0.031473927199840546, 0.099583201110363, 0.3009316027164459, 0.110613614320755, 0.07535482197999954, 0.24139271676540375, -0.04599466919898987, 0.2722318470478058, -0.21119655668735504, -0.1831187903881073, -0.23876826465129852, -0.15462404489517212, 0.02841603010892868, -0.4011644721031189, 0.19092701375484467, 0.17102372646331787, -0.2977710962295532, 0.11503446847200394, -0.00012200132186990231, -0.1493762731552124, 0.2697799801826477, -0.008835399523377419, -0.17719951272010803, -0.23157073557376862, -0.5108581781387329, -0.47503483295440674, 0.4412977993488312, -0.6071307063102722, 0.006469763815402985, -0.16628794372081757, 0.009862657636404037, 0.23457130789756775, 0.5989477634429932, -0.14206349849700928, 0.07940220832824707, -0.39332544803619385, 0.004251584410667419, 0.05900344252586365, 0.2985043227672577, 0.25531595945358276, -0.35816827416419983, -0.3896653950214386, -0.13251174986362457, 0.019385769963264465, 0.013584814965724945, 0.19017522037029266, 0.04804027080535889, -0.15089289844036102, 0.05642222985625267, 0.18820162117481232, -0.24059054255485535, -0.10517488420009613, -0.031474657356739044, 0.3205353021621704, -0.14700540900230408, -0.1331748366355896, -0.22489915788173676, -0.7789202332496643, 0.12296392768621445, -0.44449716806411743, 0.023178912699222565, 0.3083992302417755, -0.08555446565151215, 0.0007131583988666534, 0.15985266864299774, -0.2577974796295166, -0.15868929028511047, 0.07997286319732666, 0.1872967928647995, 0.5755347013473511, 0.042686834931373596, 0.21303340792655945, 0.1306547373533249, 0.05146577209234238, 0.24592836201190948, 0.09748344123363495, 0.03707391396164894, 0.14819671213626862, -0.2402217537164688, -0.34626346826553345, -0.36292824149131775, -0.06122704595327377, -0.24780452251434326, 0.07276201248168945, 0.1947951763868332, -0.09304316341876984, 0.03792031481862068, 0.10942426323890686, 0.33250829577445984, 0.0307522751390934, 0.36637812852859497, 0.5928639769554138, 0.09704606235027313, 0.0739007443189621, -0.33025893568992615, -0.018299344927072525, -0.055823132395744324, 0.10930235683917999, -0.2353014200925827, 0.08426624536514282, -0.10003908723592758, 0.16355499625205994, -0.045883048325777054, -0.25977039337158203, -0.09913961589336395, -0.24285975098609924, 0.22444310784339905, 0.03780568391084671, 0.11807219684123993, 0.03906617313623428, -0.3129241466522217, 0.20334045588970184, -0.12280918657779694, -0.4388540983200073, -0.08893860876560211, 0.09969562292098999, -0.3002474009990692, 0.2296498417854309, 0.10692672431468964, 0.1424671858549118, 0.18240663409233093, 0.008428710512816906, -0.14765872061252594, -0.24537387490272522, 0.15634943544864655, 0.006643891334533691, 0.23175549507141113, 0.04239439591765404, 0.03246494010090828, 0.05623100697994232, -0.03774191066622734, -0.07236993312835693, -0.30212584137916565, 0.017728086560964584, -0.07856012135744095, -0.3099417984485626, 0.18636199831962585, 0.040432706475257874, 0.1440340131521225, 0.0619366429746151, 0.36202529072761536, 0.6019584536552429, 0.4055771231651306, -0.16478456556797028, 0.10734278708696365, 0.03777235746383667, -0.2596316337585449, -0.19345679879188538, -0.04369520768523216, -0.013956189155578613, -0.010729439556598663, -0.4486927092075348, -0.38156411051750183, 0.16156989336013794, -0.1280699074268341, 0.07277113199234009, -0.14765974879264832, 0.05844838172197342, -0.0028754714876413345, 0.6331416368484497, 0.5671981573104858, -0.2285178154706955, 0.08923420310020447, 0.33646729588508606, 0.11590470373630524, -0.3352690041065216, 0.48383769392967224, 0.05015658214688301, 0.1074584349989891, -0.10666864365339279, -0.42390814423561096, 0.18046195805072784, -0.5068798065185547, -0.22542394697666168, 0.1737489402294159, -0.5565367937088013, 0.1767486333847046, -0.07062986493110657, -0.03899857774376869, -0.33048513531684875, -0.002800464630126953, 0.2975156307220459, 0.484963595867157, 0.034014418721199036, 0.2780068516731262, -0.193300262093544, -0.06715768575668335, 0.26429852843284607, -0.1216094046831131, -0.3572371006011963, -0.31178978085517883, -0.04289236664772034, -0.12129057198762894, 0.03108002245426178, -0.037856221199035645, -0.38954925537109375, -0.012586912140250206, -0.27908098697662354, -0.16043080389499664, -0.27749621868133545, 0.031025007367134094, -0.2263089418411255, -0.06190350279211998, -0.3348512053489685, -0.3435286283493042, 0.1144687682390213, -0.08758200705051422, 0.2657090425491333, -0.32785436511039734, -0.16811028122901917, 0.2228049337863922, -0.025595609098672867, -0.18210676312446594, 0.6047660708427429, 0.28620004653930664, -0.008092899806797504, 0.062073979526758194, 0.3274427056312561, 0.39119067788124084, -0.030648797750473022, 0.005734089761972427, -0.12919311225414276, 0.18323741853237152, -0.1230173110961914, -0.052457764744758606, -0.043942611664533615, -0.2799319326877594, -0.20842567086219788, -0.20216104388237, 0.5459465384483337, 0.02758134715259075, 0.6292759776115417, 0.29110682010650635, 0.04996940493583679, -0.012831151485443115, -0.33546561002731323, -0.1722605675458908, -0.5019679069519043, -0.2116854041814804, -0.41578930616378784, 0.11526750028133392, -0.13372960686683655, -0.2776995599269867, 0.23551073670387268, 0.43055224418640137, 0.17799270153045654, 0.03364701569080353, -0.23415637016296387, 0.13413673639297485, 0.1435289978981018, -0.19840002059936523, 0.6893718838691711, 0.1891099214553833, 0.15224871039390564, -0.04077437147498131, 0.08033354580402374, 0.033018745481967926, -0.18132691085338593, 0.2855430543422699, 0.2572939991950989, 0.08319474756717682, -0.00174027681350708, -0.1671253740787506, -0.13735799491405487, 0.053016968071460724, 0.060680363327264786, 0.22533902525901794, -0.33681875467300415, -0.05381298437714577, -0.07482229918241501, -0.12956325709819794, -0.31246545910835266, -0.554850161075592, -0.03342394158244133, -0.18639399111270905, -0.2262297421693802, 0.20903968811035156, 0.014331009238958359, -0.21730944514274597, 0.15765073895454407, -0.1592097282409668, 0.16644924879074097, -0.4105944037437439, 0.08151502907276154, 0.05014517903327942, -0.41217392683029175, -0.03657117486000061, 0.022978369146585464, 0.06295384466648102, 0.013407085090875626, 0.2887689173221588, 0.36025840044021606, -0.2948746085166931, 0.18425501883029938, -0.37269872426986694, 0.18451961874961853, -0.39978164434432983, -0.2675513029098511, 0.22958457469940186, -0.3035169839859009, 0.014626704156398773, -0.4285724461078644, 0.3069131076335907, -0.16745927929878235, -0.13079622387886047, -0.33435046672821045, 0.32829394936561584, 0.19106030464172363, -0.17773805558681488, -0.31340306997299194, -0.4024410545825958, -0.14577072858810425, 0.39054131507873535, -0.06979222595691681, 0.27505964040756226, 0.07240894436836243, -0.47120171785354614, -0.028921205550432205, -0.0028675044886767864, -0.12118388712406158, 0.06887870281934738, 0.2207588255405426, -0.07444769144058228, -0.4078602194786072, 0.12987999618053436, 0.08459217846393585, -0.12469129264354706, 0.3680254817008972, -0.24733681976795197, -0.06267524510622025, 0.2727823257446289, 0.0512867346405983, -0.0616641566157341, 0.16137829422950745, 0.04686148464679718, 0.16062408685684204, 0.2921682298183441, 0.06726719439029694, -0.1720879077911377, -0.24137789011001587, 0.3609776198863983, 0.12154173851013184, 0.1340135782957077, -0.07184866070747375, 0.09102082252502441, -0.0506044365465641, 0.6642796397209167, 0.2965461015701294, 0.2599329948425293, -0.15238459408283234, 0.0683135911822319, -0.3416401743888855, 0.08138628304004669, -0.2282859981060028, 0.14147549867630005, -0.05142064392566681, -0.3744940161705017, 0.3664928078651428, 0.03204856067895889, -0.13655821979045868, -0.26294875144958496, -0.012271646410226822, -0.08253288269042969, -0.5328514575958252, 0.008546222932636738, -0.18302999436855316, 0.26944786310195923, -0.13070543110370636, 0.06471652537584305, 0.032201122492551804, -0.03419853374361992, -0.14848749339580536, -0.19269607961177826, -0.09611973911523819, -0.14421384036540985, -0.1351083666086197, -0.5458652973175049, -0.05232853442430496, -0.01371203362941742, 0.24540723860263824, -0.13100899755954742, -0.04341259226202965, -0.19963476061820984, 0.2979271411895752, 0.27644407749176025, 0.2861585021018982, -0.07402962446212769, -0.1080961674451828, 0.14678698778152466, 0.04686491936445236, 0.19412098824977875, -0.1502501219511032, 0.07547254860401154, 0.025733720511198044, 0.040008898824453354, 0.02105630189180374, -0.5988472700119019, -0.2771078646183014, 0.5790029764175415, -0.2179686576128006, -0.3166865408420563, 0.02838323265314102, 0.054655272513628006, 0.1528545618057251, 0.08562319725751877, -0.028498567640781403, 0.12638422846794128, 0.102929025888443, 0.0625976175069809, 0.13151976466178894, -0.03761439770460129, -0.18430200219154358, 0.12430372834205627, 0.12849843502044678, 0.00018762890249490738, -0.08899036794900894, 0.15551355481147766, 0.23965558409690857, 0.2841665744781494, -0.3276495933532715, 0.49771326780319214, -0.3636684715747833, -0.10032550990581512, 0.02168770134449005, -0.21436487138271332, 0.5283656120300293, 0.41589781641960144, 0.14213867485523224, 0.3851136267185211, 0.04732169210910797, 0.04736083745956421, -0.19367410242557526, -0.11665048450231552, 0.3290862441062927, -0.2601209878921509, -0.46221598982810974, -0.7009153366088867, 0.4129931628704071, 0.07184499502182007, -0.22511830925941467, 0.2465706467628479, 0.1570991575717926, -0.3763379156589508, 0.571111261844635, -0.0197211354970932, 0.8757331967353821, 0.0014640577137470245, 0.1962801069021225, 0.01991446688771248, -0.28456056118011475, 0.6315013766288757, -0.015095246955752373, 0.10312175750732422, -0.3777364492416382, 0.30199331045150757, 0.0610177144408226, 0.08488969504833221, 0.22855684161186218, 0.5355200171470642, 0.013650249689817429, 0.18368718028068542, 0.0637408047914505, 0.27864792943000793, -0.2928551137447357, 0.3380335867404938, 0.20744100213050842, -0.08355380594730377, -0.14133700728416443, -0.04221634939312935, 0.026887496933341026, -0.2585872709751129, 0.1981571614742279, -0.18899011611938477, -0.02284853532910347, 0.1410934180021286, -0.276981920003891, 0.06596143543720245, -0.004031449090689421, 0.11964206397533417, -0.1257592886686325, -0.4483075737953186, 0.4320882558822632, 0.16505305469036102, 0.3249998092651367, -0.18011312186717987, -0.038074590265750885, 0.47148019075393677, -0.05818856507539749, 0.09149451553821564, -0.16750508546829224, 0.02509332448244095, 0.22142207622528076, 0.02951090782880783, 0.10628586262464523, 0.026310499757528305, -0.02527446672320366, -0.06378035992383957, -0.2644971013069153, 0.21506088972091675, 0.4529537558555603, -0.5559493899345398, 0.17467321455478668, -0.012984134256839752, 0.052493173629045486, -0.23507465422153473, 0.01745961606502533, -0.23933292925357819, -0.21969947218894958, 0.016380183398723602, -0.18049491941928864, -0.04825238510966301, -0.02668430283665657, -0.2955889403820038, 0.23804041743278503, -0.015308177098631859, 0.42816048860549927, -0.11671416461467743, 0.1651087999343872, -0.2321612536907196, 0.07606162875890732, 0.17264424264431, -0.26263248920440674, 0.28448647260665894, 0.11356765031814575, -0.27455276250839233, 0.15061388909816742, -0.015604149550199509, 0.052263043820858, 0.6398313641548157, -0.15392321348190308, 0.015465021133422852, -0.20868000388145447, 0.004173249006271362, -0.06663264334201813, 0.06332933902740479, -0.04518792778253555, 0.46698451042175293, -0.1915590614080429, 0.4138619601726532, -0.13755114376544952, -0.01859290339052677, 0.2807840406894684, 0.2623925507068634, -0.15228036046028137, 0.009632933884859085, 0.19519390165805817, 0.07528987526893616, 0.0053584277629852295, -0.006513435393571854, 0.2042776346206665, -0.060487858951091766, -0.31543147563934326, 0.12557697296142578, 0.1368718445301056, 0.19400639832019806, -0.3965693712234497, -0.16864928603172302, 0.07181931287050247, -0.15368957817554474, 0.1787271946668625, 0.0668719932436943, 0.28282541036605835, 0.4819812774658203, 0.305769681930542, 0.11160007119178772, -0.36445021629333496, 0.184103861451149, -0.1227378249168396, 0.3808993101119995, 0.2149304449558258, 0.2762942910194397, -0.16513989865779877, 0.09467662870883942, 0.1048569604754448, -0.04907936602830887, 0.1903131604194641, -0.3088618218898773, 0.013765502721071243, 0.032846011221408844, -0.030747124925255775, -0.14969176054000854, 0.2588711380958557, 0.5672738552093506, 0.12643210589885712, -0.002587677910923958, 0.22976338863372803, 0.08614946156740189, -0.09101273119449615, 0.25636425614356995, 0.28159111738204956, -0.15940628945827484, 0.08424220979213715, 0.554146409034729, 0.12478646636009216, 0.265510231256485, -0.4000856280326843, -0.03358306735754013, -0.05661609396338463, -0.04436315596103668, 0.5013856887817383, 0.8824547529220581, 0.20604358613491058, -0.01995519921183586, 0.4297121465206146, -0.19077296555042267, 0.24227681756019592, 0.49159663915634155, -0.17309698462486267, 0.33046823740005493, 0.31470343470573425, -0.1226273626089096, 0.5907777547836304, 0.19488196074962616, -0.4043387472629547, -0.0030668526887893677, -0.01677120476961136, -0.2341189682483673, -0.1521470695734024, 0.7617432475090027, -0.011581952683627605, -0.3254048824310303, 0.09411849081516266, 0.26180654764175415, 0.1801457405090332, -0.2026377022266388, -0.16221767663955688, 0.055055610835552216, -0.35550713539123535, 0.00861787423491478, -0.08892384171485901, -0.20625172555446625, 0.35118499398231506, 0.03732084482908249, 0.26485681533813477, -0.15261128544807434, -0.03247227519750595, 0.17793399095535278, 0.5029484033584595, -0.3182084858417511, -0.10205728560686111, -0.0024594739079475403, -0.2936454713344574, 0.1971895545721054, -0.0927339568734169, -0.0056500546634197235, -0.1405806541442871, -0.3290480673313141, -0.3568286597728729, 0.2028418779373169, -0.14652951061725616, -0.00009953230619430542, 0.036620721220970154, 0.12525323033332825, 0.30373483896255493, 0.18481414020061493, -0.04838898032903671, -0.026214726269245148, -0.15177975594997406, 0.28189215064048767, -0.07489265501499176, -0.04711500555276871, 0.6482118368148804, 0.13582047820091248, -0.18381652235984802, 0.02307552471756935, 0.17055025696754456, -0.23163998126983643, -0.0665179118514061, 0.3410789966583252, -0.0482519268989563, 0.11722389608621597, -0.11295083910226822, -0.03360505402088165, -0.03483925387263298, 0.40514323115348816, -0.20681490004062653, -0.28671562671661377, -0.26146480441093445, -0.42668598890304565, -0.08076158910989761, -0.07397984713315964, -0.14937064051628113, -0.24868905544281006, -0.09933724254369736, 0.1381153166294098, -0.08727002888917923, -0.16104337573051453, 0.11898640543222427, -0.07627593725919724, 0.3919476866722107, 0.362347275018692, -0.22487092018127441, 0.006812773644924164, -0.03206918388605118, 0.1607152670621872, 0.12720413506031036, 0.11123877018690109, -0.1088411808013916, 0.34477901458740234, -0.3298475742340088, 0.17180845141410828, -0.07115071266889572, 0.5625976920127869, 0.20206698775291443, -0.059335678815841675, 0.01238972321152687, 0.23775723576545715, 0.02985571324825287, -0.06352116912603378, 0.035342663526535034, -0.002857648301869631, -0.539322018623352, 0.5002475380897522, 0.017259713262319565, 0.17635081708431244, -0.23865637183189392, 0.44471848011016846, -0.14618802070617676, -0.04464228078722954, -0.25948840379714966, 0.165245920419693, -0.06921359896659851, 0.2242836058139801, -0.06041460484266281, 0.3944738805294037, 0.4205290973186493, 0.2640945017337799, -0.12344498932361603, -0.02524401806294918, -0.2117692530155182, -0.35545220971107483, 0.2351643592119217, -0.0745387151837349, -0.13716071844100952, 0.10333753377199173, 0.3187505602836609, 0.171633780002594, 0.01623907871544361, -0.7228697538375854, -0.26910266280174255, 0.1650715470314026, -0.2762211263179779, 0.020327262580394745, 0.038657110184431076, 0.23195235431194305, -0.23970797657966614, 0.013042990118265152, -0.045807257294654846, 0.2986750900745392, -0.07945685088634491, 0.06468236446380615, -0.11545078456401825 ]
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
[ -0.11933845281600952, 0.08862784504890442, -0.00897780992090702, 0.21019737422466278, 0.3692678213119507, 0.026059836149215698, 0.4487815499305725, 0.4131021499633789, 0.4330812692642212, 0.012566991150379181, 0.12384359538555145, 0.395429790019989, -0.15054914355278015, 0.021962707862257957, 0.12090519070625305, -0.42898982763290405, -0.03373972326517105, 0.3636728823184967, -0.041466742753982544, 0.06297364085912704, -0.1876835823059082, 0.10851715505123138, -0.2651914358139038, 0.25500577688217163, -0.30572283267974854, -0.05702577531337738, 0.2159208357334137, 0.08206366002559662, -0.07891867309808731, -0.5590065121650696, 0.2770831286907196, -0.2170492708683014, 0.06445172429084778, 0.11474639922380447, -0.00010995571938110515, 0.20530645549297333, 0.33269721269607544, 0.030212881043553352, -0.5799458026885986, -0.18957795202732086, -0.4773316979408264, -0.11893881857395172, 0.29743173718452454, -0.2497076690196991, -0.028513483703136444, 0.05389726161956787, 0.035910993814468384, -0.07441209256649017, 0.29674679040908813, 0.531506359577179, 0.237888902425766, 0.5236923694610596, 0.15454915165901184, 0.03449363261461258, -0.07856784015893936, -0.17057108879089355, -0.08211500197649002, 0.22396822273731232, 0.07387731224298477, -0.39154377579689026, -0.013373926281929016, 0.1269245445728302, 0.15117238461971283, 0.26370200514793396, 0.5083709359169006, 0.014393925666809082, 0.2405235767364502, -0.15062691271305084, 0.014946820214390755, 0.07640057057142258, 0.3353826403617859, -0.1649332046508789, -0.2715998888015747, -0.2243589460849762, 0.2476661205291748, -0.33475619554519653, 0.2365303635597229, -0.03667646646499634, -0.16220436990261078, -0.008197136223316193, -0.26922789216041565, -0.028147468343377113, -0.24182075262069702, 0.3526870608329773, -0.058214087039232254, 0.49964913725852966, 0.043583355844020844, 0.1427474468946457, -0.05806552246212959, -0.07545772939920425, 0.14037704467773438, -0.08709109574556351, -0.1293005496263504, 0.25781309604644775, -0.38315531611442566, -0.036467600613832474, -0.034946344792842865, -0.11697594076395035, -0.031624749302864075, 0.26352018117904663, 0.21998973190784454, -0.22487351298332214, 0.3238390386104584, 0.3154153525829315, 0.22298845648765564, 0.42046889662742615, 0.07276557385921478, 0.06359237432479858, -0.08280497789382935, 0.1360260546207428, -0.23372812569141388, -0.15548521280288696, -0.1555362045764923, -0.1765558421611786, 0.09289798140525818, -0.15460222959518433, 0.22602073848247528, -0.07111581414937973, -0.4984738230705261, 0.11472072452306747, -0.14340484142303467, 0.046290118247270584, 0.20511485636234283, 0.17948494851589203, -0.12763413786888123, 0.23223412036895752, 0.14412455260753632, 0.2825669050216675, -0.25197696685791016, -0.04313553497195244, -0.16628961265087128, 0.249213308095932, -0.23438286781311035, -0.18224890530109406, 0.24275390803813934, 0.193400040268898, 0.3994784653186798, -0.10116767883300781, 0.11126405000686646, -0.2017553448677063, 0.2273758500814438, -0.04551952704787254, -0.1288793981075287, 0.25665315985679626, -0.21606102585792542, 0.1548980474472046, 0.20159120857715607, -0.2824234366416931, -0.0848238617181778, -0.10174446552991867, -0.27263572812080383, -0.30809223651885986, -0.1262514442205429, 0.28141558170318604, 0.147381991147995, -0.24296480417251587, -0.15226313471794128, -0.06118778511881828, 0.14749255776405334, -0.36872372031211853, -0.11810629814863205, -0.4612586498260498, -0.22063903510570526, -0.023005738854408264, 0.3976435363292694, 0.20490935444831848, 0.13138674199581146, -0.1089726984500885, -0.11542053520679474, 0.07785429060459137, -0.031797103583812714, 0.1227136105298996, -0.43835723400115967, 0.5421629548072815, -0.06105594336986542, 0.118863046169281, 0.64134681224823, -0.3427872657775879, -0.4578544795513153, 0.12356892228126526, -0.014279965311288834, 0.20932787656784058, -0.22536897659301758, -0.01672065444290638, -0.013226330280303955, -0.1845511645078659, 0.21231377124786377, 0.6192131042480469, -0.015042958781123161, 0.08247934281826019, -0.12689706683158875, -0.01554920244961977, 0.5002195239067078, 0.10985688865184784, 0.09252531826496124, -0.007928885519504547, -0.04170481115579605, 0.49664852023124695, -0.00131293386220932, -0.12597057223320007, -0.014860082417726517, -0.013630155473947525, -0.06468622386455536, -0.10460402071475983, -0.2178265005350113, -0.07388680428266525, -0.2776644825935364, 0.30430400371551514, -0.11754798144102097, -0.05853193253278732, -0.04715031385421753, 0.10618656873703003, -0.34911325573921204, 0.16848869621753693, -0.42326927185058594, -0.3773188889026642, 0.17209337651729584, 0.15513363480567932, -0.01082286611199379, 0.23844115436077118, -0.20474110543727875, 0.21943674981594086, -0.2740277051925659, 0.055500876158475876, -0.21892870962619781, 0.16128991544246674, -0.05588398873806, -0.04242202267050743, -0.009479247033596039, 0.30879634618759155, 0.03833160549402237, -0.017996493726968765, -0.3062889277935028, 0.28603029251098633, -0.12548455595970154, 0.08035629242658615, -0.03226495534181595, -0.14337028563022614, 0.06880630552768707, -0.07154297828674316, -0.17518453299999237, 0.20819191634655, 0.33195099234580994, -0.18012642860412598, -0.04314299672842026, 0.11287609487771988, 0.1272123008966446, 0.0814221054315567, 0.09886433184146881, 0.16397228837013245, 0.14399199187755585, -0.06582621484994888, 0.1012795940041542, 0.07743129134178162, 0.23370946943759918, -0.09498715400695801, 0.20471079647541046, -0.15645107626914978, -0.1671449989080429, -0.11880645900964737, 0.3131202161312103, 0.17005868256092072, 0.16741719841957092, 0.08148638904094696, -0.2752525210380554, -0.09487432986497879, 0.07855785638093948, 0.14604909718036652, 0.40096497535705566, -0.0518328957259655, -0.10834059119224548, -0.05257079750299454, -0.4165968596935272, 0.07112551480531693, 0.17070795595645905, 0.03662925213575363, 0.20912905037403107, 0.30185645818710327, 0.04131554812192917, 0.0873207226395607, -0.2952670454978943, -0.35701698064804077, 0.2824336886405945, 0.4405299723148346, -0.22848191857337952, 0.20788007974624634, -0.03053159825503826, -0.0025112517178058624, -0.06784830242395401, -0.33363378047943115, -0.18862853944301605, -0.342993825674057, -0.12133617699146271, 0.433775395154953, 0.024119451642036438, 0.181004598736763, -0.3977614641189575, -0.022051095962524414, 0.19475087523460388, -0.11478548496961594, -0.03137951344251633, -0.3157978653907776, -0.2693289816379547, 0.04594217613339424, 0.3818606734275818, 0.09107664972543716, 0.3510980010032654, -0.06677059084177017, -0.0925356075167656, -0.007035041227936745, -0.05841507762670517, -0.07556726038455963, -0.18372085690498352, 0.2440476268529892, 0.009521827101707458, 0.2855835258960724, -0.17702968418598175, -0.5220333337783813, 0.5017467737197876, -0.25570181012153625, -0.057785287499427795, 0.31510666012763977, 0.27111297845840454, -0.24919047951698303, -0.07947683334350586, -0.37468862533569336, -0.12645035982131958, -0.28060582280158997, -0.05946876108646393, 0.07482925057411194, 0.062020909041166306, 0.39746737480163574, 0.07085651159286499, 0.21601088345050812, -0.0025735688395798206, 0.1134379655122757, -0.04826278239488602, -0.09006398171186447, 0.1859847605228424, -0.31003400683403015, -0.31731829047203064, -0.12134282290935516, 0.00993594154715538, 0.3347616195678711, 0.11073551326990128, -0.3489593267440796, -0.21413181722164154, -0.10495822131633759, 0.08531276881694794, -0.2213253676891327, -0.04006670415401459, 0.18286007642745972, 0.07230357825756073, -0.06572244316339493, -0.08664388954639435, 0.03598948195576668, 0.07124010473489761, 0.15234598517417908, 0.13496194779872894, -0.006304748356342316, 0.08922688663005829, -0.08488400280475616, 0.6008991599082947, 0.21869370341300964, -0.21437688171863556, 0.30028587579727173, -0.10327673703432083, 0.005144510418176651, -0.208683580160141, -0.4398045539855957, 0.26416414976119995, -0.11523276567459106, 0.18292897939682007, 0.17231473326683044, -0.07949357479810715, -0.2247839719057083, -0.1925591230392456, 0.06885170936584473, -0.17844745516777039, -0.27277958393096924, 0.14616836607456207, -0.21739450097084045, 0.2787109315395355, -0.06886422634124756, 0.22253704071044922, -0.20522335171699524, -0.16798070073127747, 0.06985078752040863, 0.16693411767482758, -0.09054567664861679, 0.07967513799667358, -0.24400167167186737, -0.27610862255096436, -0.4799582362174988, -0.02783966064453125, 0.24103179574012756, 0.643750011920929, -0.08638451993465424, 0.03718879818916321, 0.08072253316640854, 0.08174140751361847, 0.13737376034259796, -0.2301362007856369, 0.1822635680437088, 0.2891308069229126, 0.03182467073202133, -0.31974661350250244, -0.06433149427175522, -0.14011231064796448, 0.35801658034324646, 0.0766519159078598, -0.054431721568107605, -0.03723123297095299, -0.1676540970802307, 0.1398320496082306, 0.2599392831325531, -0.1872951090335846, -0.044113077223300934, -0.09173431247472763, -0.08981046825647354, -0.30776458978652954, -0.006154969334602356, -0.009947758167982101, 0.33566054701805115, 0.12296926230192184, 0.26572471857070923, -0.5462992191314697, 0.11665705591440201, 0.09016524255275726, 0.17458684742450714, 0.09595315158367157, 0.057944219559431076, 0.18724678456783295, -0.16663400828838348, 0.18321000039577484, -0.14590442180633545, 0.5633803606033325, 0.14737597107887268, -0.3307987451553345, -0.31952953338623047, 0.02166324108839035, 0.16743338108062744, 0.18398034572601318, -0.01858552172780037, -0.14112097024917603, -0.2769453525543213, -0.06146351993083954, -0.008708748035132885, 0.3045704662799835, 0.5610411167144775, 0.0026276633143424988, -0.3271808922290802, -0.2585407495498657, 0.4828394949436188, -0.2850031852722168, 0.11732904613018036, 0.3215572237968445, 0.05337715521454811, -0.3131922781467438, 0.08413887023925781, 0.07783256471157074, 0.6767752766609192, 0.16953882575035095, 0.3111467659473419, 0.3826707899570465, 0.18392826616764069, 0.4142706096172333, -0.043089497834444046, 0.2847142815589905, -0.3927357494831085, -0.13342586159706116, -0.08068586140871048, -0.06546270102262497, 0.10708172619342804, -0.008873388171195984, -0.02543865703046322, 0.150621235370636, 0.15688925981521606, -0.0041420795023441315, 0.05714789777994156, 0.4492526650428772, -0.03878391534090042, -0.1968192607164383, -0.3470253050327301, 0.15601636469364166, -0.21799974143505096, 0.2975102663040161, -0.09073382616043091, -0.06648421287536621, 0.05520467087626457, -0.06735203415155411, -0.25555843114852905, 0.2574385404586792, -0.29947492480278015, 0.4594579339027405, 0.05782182514667511, -0.4756813049316406, 0.1984151303768158, 0.20709790289402008, 0.1756638139486313, 0.0056989602744579315, -0.29959404468536377, 0.23877887427806854, -0.2900899052619934, -0.22186626493930817, -0.06728766113519669, 0.11194741725921631, 0.4200734496116638, -0.17255154252052307, -0.14637798070907593, 0.14794065058231354, -0.12065417319536209, -0.24150244891643524, 0.1039588451385498, 0.036866605281829834, 0.3524894714355469, -0.3925756514072418, -0.5152195692062378, -0.0024207234382629395, -0.18207629024982452, -0.17491021752357483, 0.17450003325939178, 0.02345065027475357, -0.18903149664402008, 0.15189561247825623, 0.14982710778713226, -0.19898000359535217, 0.11640655994415283, 0.24804237484931946, -0.2495403289794922, 0.14628303050994873, 0.7637573480606079, 0.03394555673003197, 0.01280517503619194, -0.3138304352760315, -0.05428845435380936, 0.12508459389209747, -0.48199987411499023, -0.1171272024512291, 0.12550412118434906, 0.16912586987018585, -0.1442466676235199, 0.49836599826812744, 0.06302111595869064, -0.01884474605321884, 0.24718548357486725, -0.5424140095710754, -0.10621928423643112, -0.043021634221076965, 0.017753619700670242, 0.23777522146701813, -0.0026239529252052307, 0.11488902568817139, 0.13550958037376404, 0.1033887192606926, -0.2726374566555023, 0.014254427514970303, -0.15762582421302795, 0.1379965841770172, 0.21783430874347687, -0.1565312147140503, 0.22328168153762817, -0.07694721221923828, 0.15638616681098938, -0.09584659337997437, -0.12303443253040314, -0.24829328060150146, -0.25483784079551697, 0.10011143982410431, 0.24556711316108704, -0.1645212471485138, -0.030679188668727875, -0.10233347862958908, 0.02367764711380005, -0.23110505938529968, -0.00846264511346817, -0.11772798001766205, -0.2933427095413208, 0.38943231105804443, 0.04389800876379013, 0.27836859226226807, -0.11169030517339706, 0.12635567784309387, -0.30075740814208984, 0.14912036061286926, -0.01849590800702572, -0.07522959262132645, 0.27581319212913513, 0.09365326166152954, -0.4658869206905365, 0.0054205358028411865, -0.40260517597198486, 0.03714786097407341, 0.3698212504386902, -0.4551103711128235, 0.15995843708515167, -0.1316281408071518, 0.09546546638011932, 0.4656829833984375, -0.27140846848487854, -0.0036669299006462097, 0.3496653139591217, 0.23784254491329193, -0.29741764068603516, -0.1892606019973755, 0.2399907410144806, 0.061345674097537994, 0.03245936706662178, 0.02723889797925949, 0.04666122421622276, -0.3037486672401428, -0.32123446464538574, 0.08927963674068451, -0.029340408742427826, -0.2656012773513794, -0.01140776090323925, 0.33932653069496155, -0.002364441752433777, -0.05951501801609993, -0.0500493124127388, 0.21102115511894226, 0.03539281338453293, 0.38269317150115967, -0.29260092973709106, 0.2069980353116989, 0.05926353484392166, 0.04391039162874222, -0.11114301532506943, -0.21921727061271667, 0.05030050128698349, -0.15784099698066711, 0.03383420407772064, 0.2952827513217926, 0.09981479495763779, 0.1530807912349701, -0.2855033874511719, -0.12453707307577133, -0.20802274346351624, 0.22792446613311768, -0.17904222011566162, -0.04890655726194382, -0.24299554526805878, 0.02101702243089676, -0.02587638422846794, -0.09469917416572571, -0.14167022705078125, -0.0381786972284317, 0.3095560371875763, 0.2532220482826233, -0.1722705215215683, -0.45553502440452576, 0.10826653242111206, 0.3432941436767578, 0.18004727363586426, -0.24588416516780853, 0.3368532359600067, 0.19381549954414368, -0.06877827644348145, -0.01412438228726387, 0.3971101641654968, 0.43576839566230774, 0.3383936583995819, -0.364220529794693, 0.08116793632507324, -0.20338064432144165, -0.05682189017534256, -0.023260291665792465, 0.2695123851299286, 0.10607316344976425, 0.11737622320652008, 0.46242257952690125, 0.12148632109165192, -0.1247272789478302, 0.04363914206624031, -0.08805669844150543, -0.2605326473712921, -0.26294705271720886, 0.5191487073898315, 0.1857946217060089, -0.1143883466720581, -0.32065102458000183, 0.1899377703666687, -0.27498704195022583, -0.08091447502374649, 0.06285165250301361, 0.24058954417705536, 0.11315394192934036, -0.21382759511470795, 0.07050890475511551, 0.00031514279544353485, 0.7061734795570374, 0.48772192001342773, -0.017719194293022156, -0.3658052682876587, -0.11652375012636185, -0.5989872217178345, 0.21783632040023804, -0.3639681041240692, -0.26901817321777344, 0.2833365499973297, -0.07233157753944397, -0.288082480430603, 0.24852731823921204, 0.1276196390390396, 0.15819591283798218, -0.05727645754814148, 0.09834779053926468, -0.4907650053501129, -0.14860455691814423, -0.030650503933429718, -0.007675016298890114, -0.08318105340003967, -0.25705307722091675, 0.14532306790351868, -0.4349946677684784, 0.11610303074121475, 0.01198504213243723, 0.06934809684753418, -0.00023945607244968414, 0.09664029628038406, 0.35442453622817993, 0.1653239130973816, 0.5012482404708862, -0.03556128963828087, 0.0036574434489011765, -0.1911543905735016, -0.1566217541694641, -0.2289923131465912, 0.13182267546653748, 0.08762918412685394, 0.5133118629455566, -0.02610740251839161, -0.06854520738124847, -0.26402729749679565, 0.40316787362098694, -0.08100667595863342, 0.08831582963466644, -0.1350807249546051, 0.0773145779967308, -0.0703495442867279, 0.2275840938091278, 0.2643648684024811, 0.42454031109809875, -0.1783217340707779, 0.2022092193365097, -0.2722919285297394, -0.46024608612060547, 0.32893532514572144, -0.661480724811554, -0.517790675163269, 0.0751224160194397, 0.1161930188536644, -0.1423342525959015, 0.05819430202245712, -0.6006339192390442, 0.07056698948144913, 0.24261271953582764, -0.1114056184887886, -0.317119836807251, 0.21794511377811432, 0.016542082652449608, 0.08000141382217407, 0.04357365146279335, 0.4483563303947449, -0.14847293496131897, -0.15976761281490326, -0.0049346014857292175, -0.163252055644989 ]
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).
[ -0.11933845281600952, 0.08862784504890442, -0.00897780992090702, 0.21019737422466278, 0.3692678213119507, 0.026059836149215698, 0.4487815499305725, 0.4131021499633789, 0.4330812692642212, 0.012566991150379181, 0.12384359538555145, 0.395429790019989, -0.15054914355278015, 0.021962707862257957, 0.12090519070625305, -0.42898982763290405, -0.03373972326517105, 0.3636728823184967, -0.041466742753982544, 0.06297364085912704, -0.1876835823059082, 0.10851715505123138, -0.2651914358139038, 0.25500577688217163, -0.30572283267974854, -0.05702577531337738, 0.2159208357334137, 0.08206366002559662, -0.07891867309808731, -0.5590065121650696, 0.2770831286907196, -0.2170492708683014, 0.06445172429084778, 0.11474639922380447, -0.00010995571938110515, 0.20530645549297333, 0.33269721269607544, 0.030212881043553352, -0.5799458026885986, -0.18957795202732086, -0.4773316979408264, -0.11893881857395172, 0.29743173718452454, -0.2497076690196991, -0.028513483703136444, 0.05389726161956787, 0.035910993814468384, -0.07441209256649017, 0.29674679040908813, 0.531506359577179, 0.237888902425766, 0.5236923694610596, 0.15454915165901184, 0.03449363261461258, -0.07856784015893936, -0.17057108879089355, -0.08211500197649002, 0.22396822273731232, 0.07387731224298477, -0.39154377579689026, -0.013373926281929016, 0.1269245445728302, 0.15117238461971283, 0.26370200514793396, 0.5083709359169006, 0.014393925666809082, 0.2405235767364502, -0.15062691271305084, 0.014946820214390755, 0.07640057057142258, 0.3353826403617859, -0.1649332046508789, -0.2715998888015747, -0.2243589460849762, 0.2476661205291748, -0.33475619554519653, 0.2365303635597229, -0.03667646646499634, -0.16220436990261078, -0.008197136223316193, -0.26922789216041565, -0.028147468343377113, -0.24182075262069702, 0.3526870608329773, -0.058214087039232254, 0.49964913725852966, 0.043583355844020844, 0.1427474468946457, -0.05806552246212959, -0.07545772939920425, 0.14037704467773438, -0.08709109574556351, -0.1293005496263504, 0.25781309604644775, -0.38315531611442566, -0.036467600613832474, -0.034946344792842865, -0.11697594076395035, -0.031624749302864075, 0.26352018117904663, 0.21998973190784454, -0.22487351298332214, 0.3238390386104584, 0.3154153525829315, 0.22298845648765564, 0.42046889662742615, 0.07276557385921478, 0.06359237432479858, -0.08280497789382935, 0.1360260546207428, -0.23372812569141388, -0.15548521280288696, -0.1555362045764923, -0.1765558421611786, 0.09289798140525818, -0.15460222959518433, 0.22602073848247528, -0.07111581414937973, -0.4984738230705261, 0.11472072452306747, -0.14340484142303467, 0.046290118247270584, 0.20511485636234283, 0.17948494851589203, -0.12763413786888123, 0.23223412036895752, 0.14412455260753632, 0.2825669050216675, -0.25197696685791016, -0.04313553497195244, -0.16628961265087128, 0.249213308095932, -0.23438286781311035, -0.18224890530109406, 0.24275390803813934, 0.193400040268898, 0.3994784653186798, -0.10116767883300781, 0.11126405000686646, -0.2017553448677063, 0.2273758500814438, -0.04551952704787254, -0.1288793981075287, 0.25665315985679626, -0.21606102585792542, 0.1548980474472046, 0.20159120857715607, -0.2824234366416931, -0.0848238617181778, -0.10174446552991867, -0.27263572812080383, -0.30809223651885986, -0.1262514442205429, 0.28141558170318604, 0.147381991147995, -0.24296480417251587, -0.15226313471794128, -0.06118778511881828, 0.14749255776405334, -0.36872372031211853, -0.11810629814863205, -0.4612586498260498, -0.22063903510570526, -0.023005738854408264, 0.3976435363292694, 0.20490935444831848, 0.13138674199581146, -0.1089726984500885, -0.11542053520679474, 0.07785429060459137, -0.031797103583812714, 0.1227136105298996, -0.43835723400115967, 0.5421629548072815, -0.06105594336986542, 0.118863046169281, 0.64134681224823, -0.3427872657775879, -0.4578544795513153, 0.12356892228126526, -0.014279965311288834, 0.20932787656784058, -0.22536897659301758, -0.01672065444290638, -0.013226330280303955, -0.1845511645078659, 0.21231377124786377, 0.6192131042480469, -0.015042958781123161, 0.08247934281826019, -0.12689706683158875, -0.01554920244961977, 0.5002195239067078, 0.10985688865184784, 0.09252531826496124, -0.007928885519504547, -0.04170481115579605, 0.49664852023124695, -0.00131293386220932, -0.12597057223320007, -0.014860082417726517, -0.013630155473947525, -0.06468622386455536, -0.10460402071475983, -0.2178265005350113, -0.07388680428266525, -0.2776644825935364, 0.30430400371551514, -0.11754798144102097, -0.05853193253278732, -0.04715031385421753, 0.10618656873703003, -0.34911325573921204, 0.16848869621753693, -0.42326927185058594, -0.3773188889026642, 0.17209337651729584, 0.15513363480567932, -0.01082286611199379, 0.23844115436077118, -0.20474110543727875, 0.21943674981594086, -0.2740277051925659, 0.055500876158475876, -0.21892870962619781, 0.16128991544246674, -0.05588398873806, -0.04242202267050743, -0.009479247033596039, 0.30879634618759155, 0.03833160549402237, -0.017996493726968765, -0.3062889277935028, 0.28603029251098633, -0.12548455595970154, 0.08035629242658615, -0.03226495534181595, -0.14337028563022614, 0.06880630552768707, -0.07154297828674316, -0.17518453299999237, 0.20819191634655, 0.33195099234580994, -0.18012642860412598, -0.04314299672842026, 0.11287609487771988, 0.1272123008966446, 0.0814221054315567, 0.09886433184146881, 0.16397228837013245, 0.14399199187755585, -0.06582621484994888, 0.1012795940041542, 0.07743129134178162, 0.23370946943759918, -0.09498715400695801, 0.20471079647541046, -0.15645107626914978, -0.1671449989080429, -0.11880645900964737, 0.3131202161312103, 0.17005868256092072, 0.16741719841957092, 0.08148638904094696, -0.2752525210380554, -0.09487432986497879, 0.07855785638093948, 0.14604909718036652, 0.40096497535705566, -0.0518328957259655, -0.10834059119224548, -0.05257079750299454, -0.4165968596935272, 0.07112551480531693, 0.17070795595645905, 0.03662925213575363, 0.20912905037403107, 0.30185645818710327, 0.04131554812192917, 0.0873207226395607, -0.2952670454978943, -0.35701698064804077, 0.2824336886405945, 0.4405299723148346, -0.22848191857337952, 0.20788007974624634, -0.03053159825503826, -0.0025112517178058624, -0.06784830242395401, -0.33363378047943115, -0.18862853944301605, -0.342993825674057, -0.12133617699146271, 0.433775395154953, 0.024119451642036438, 0.181004598736763, -0.3977614641189575, -0.022051095962524414, 0.19475087523460388, -0.11478548496961594, -0.03137951344251633, -0.3157978653907776, -0.2693289816379547, 0.04594217613339424, 0.3818606734275818, 0.09107664972543716, 0.3510980010032654, -0.06677059084177017, -0.0925356075167656, -0.007035041227936745, -0.05841507762670517, -0.07556726038455963, -0.18372085690498352, 0.2440476268529892, 0.009521827101707458, 0.2855835258960724, -0.17702968418598175, -0.5220333337783813, 0.5017467737197876, -0.25570181012153625, -0.057785287499427795, 0.31510666012763977, 0.27111297845840454, -0.24919047951698303, -0.07947683334350586, -0.37468862533569336, -0.12645035982131958, -0.28060582280158997, -0.05946876108646393, 0.07482925057411194, 0.062020909041166306, 0.39746737480163574, 0.07085651159286499, 0.21601088345050812, -0.0025735688395798206, 0.1134379655122757, -0.04826278239488602, -0.09006398171186447, 0.1859847605228424, -0.31003400683403015, -0.31731829047203064, -0.12134282290935516, 0.00993594154715538, 0.3347616195678711, 0.11073551326990128, -0.3489593267440796, -0.21413181722164154, -0.10495822131633759, 0.08531276881694794, -0.2213253676891327, -0.04006670415401459, 0.18286007642745972, 0.07230357825756073, -0.06572244316339493, -0.08664388954639435, 0.03598948195576668, 0.07124010473489761, 0.15234598517417908, 0.13496194779872894, -0.006304748356342316, 0.08922688663005829, -0.08488400280475616, 0.6008991599082947, 0.21869370341300964, -0.21437688171863556, 0.30028587579727173, -0.10327673703432083, 0.005144510418176651, -0.208683580160141, -0.4398045539855957, 0.26416414976119995, -0.11523276567459106, 0.18292897939682007, 0.17231473326683044, -0.07949357479810715, -0.2247839719057083, -0.1925591230392456, 0.06885170936584473, -0.17844745516777039, -0.27277958393096924, 0.14616836607456207, -0.21739450097084045, 0.2787109315395355, -0.06886422634124756, 0.22253704071044922, -0.20522335171699524, -0.16798070073127747, 0.06985078752040863, 0.16693411767482758, -0.09054567664861679, 0.07967513799667358, -0.24400167167186737, -0.27610862255096436, -0.4799582362174988, -0.02783966064453125, 0.24103179574012756, 0.643750011920929, -0.08638451993465424, 0.03718879818916321, 0.08072253316640854, 0.08174140751361847, 0.13737376034259796, -0.2301362007856369, 0.1822635680437088, 0.2891308069229126, 0.03182467073202133, -0.31974661350250244, -0.06433149427175522, -0.14011231064796448, 0.35801658034324646, 0.0766519159078598, -0.054431721568107605, -0.03723123297095299, -0.1676540970802307, 0.1398320496082306, 0.2599392831325531, -0.1872951090335846, -0.044113077223300934, -0.09173431247472763, -0.08981046825647354, -0.30776458978652954, -0.006154969334602356, -0.009947758167982101, 0.33566054701805115, 0.12296926230192184, 0.26572471857070923, -0.5462992191314697, 0.11665705591440201, 0.09016524255275726, 0.17458684742450714, 0.09595315158367157, 0.057944219559431076, 0.18724678456783295, -0.16663400828838348, 0.18321000039577484, -0.14590442180633545, 0.5633803606033325, 0.14737597107887268, -0.3307987451553345, -0.31952953338623047, 0.02166324108839035, 0.16743338108062744, 0.18398034572601318, -0.01858552172780037, -0.14112097024917603, -0.2769453525543213, -0.06146351993083954, -0.008708748035132885, 0.3045704662799835, 0.5610411167144775, 0.0026276633143424988, -0.3271808922290802, -0.2585407495498657, 0.4828394949436188, -0.2850031852722168, 0.11732904613018036, 0.3215572237968445, 0.05337715521454811, -0.3131922781467438, 0.08413887023925781, 0.07783256471157074, 0.6767752766609192, 0.16953882575035095, 0.3111467659473419, 0.3826707899570465, 0.18392826616764069, 0.4142706096172333, -0.043089497834444046, 0.2847142815589905, -0.3927357494831085, -0.13342586159706116, -0.08068586140871048, -0.06546270102262497, 0.10708172619342804, -0.008873388171195984, -0.02543865703046322, 0.150621235370636, 0.15688925981521606, -0.0041420795023441315, 0.05714789777994156, 0.4492526650428772, -0.03878391534090042, -0.1968192607164383, -0.3470253050327301, 0.15601636469364166, -0.21799974143505096, 0.2975102663040161, -0.09073382616043091, -0.06648421287536621, 0.05520467087626457, -0.06735203415155411, -0.25555843114852905, 0.2574385404586792, -0.29947492480278015, 0.4594579339027405, 0.05782182514667511, -0.4756813049316406, 0.1984151303768158, 0.20709790289402008, 0.1756638139486313, 0.0056989602744579315, -0.29959404468536377, 0.23877887427806854, -0.2900899052619934, -0.22186626493930817, -0.06728766113519669, 0.11194741725921631, 0.4200734496116638, -0.17255154252052307, -0.14637798070907593, 0.14794065058231354, -0.12065417319536209, -0.24150244891643524, 0.1039588451385498, 0.036866605281829834, 0.3524894714355469, -0.3925756514072418, -0.5152195692062378, -0.0024207234382629395, -0.18207629024982452, -0.17491021752357483, 0.17450003325939178, 0.02345065027475357, -0.18903149664402008, 0.15189561247825623, 0.14982710778713226, -0.19898000359535217, 0.11640655994415283, 0.24804237484931946, -0.2495403289794922, 0.14628303050994873, 0.7637573480606079, 0.03394555673003197, 0.01280517503619194, -0.3138304352760315, -0.05428845435380936, 0.12508459389209747, -0.48199987411499023, -0.1171272024512291, 0.12550412118434906, 0.16912586987018585, -0.1442466676235199, 0.49836599826812744, 0.06302111595869064, -0.01884474605321884, 0.24718548357486725, -0.5424140095710754, -0.10621928423643112, -0.043021634221076965, 0.017753619700670242, 0.23777522146701813, -0.0026239529252052307, 0.11488902568817139, 0.13550958037376404, 0.1033887192606926, -0.2726374566555023, 0.014254427514970303, -0.15762582421302795, 0.1379965841770172, 0.21783430874347687, -0.1565312147140503, 0.22328168153762817, -0.07694721221923828, 0.15638616681098938, -0.09584659337997437, -0.12303443253040314, -0.24829328060150146, -0.25483784079551697, 0.10011143982410431, 0.24556711316108704, -0.1645212471485138, -0.030679188668727875, -0.10233347862958908, 0.02367764711380005, -0.23110505938529968, -0.00846264511346817, -0.11772798001766205, -0.2933427095413208, 0.38943231105804443, 0.04389800876379013, 0.27836859226226807, -0.11169030517339706, 0.12635567784309387, -0.30075740814208984, 0.14912036061286926, -0.01849590800702572, -0.07522959262132645, 0.27581319212913513, 0.09365326166152954, -0.4658869206905365, 0.0054205358028411865, -0.40260517597198486, 0.03714786097407341, 0.3698212504386902, -0.4551103711128235, 0.15995843708515167, -0.1316281408071518, 0.09546546638011932, 0.4656829833984375, -0.27140846848487854, -0.0036669299006462097, 0.3496653139591217, 0.23784254491329193, -0.29741764068603516, -0.1892606019973755, 0.2399907410144806, 0.061345674097537994, 0.03245936706662178, 0.02723889797925949, 0.04666122421622276, -0.3037486672401428, -0.32123446464538574, 0.08927963674068451, -0.029340408742427826, -0.2656012773513794, -0.01140776090323925, 0.33932653069496155, -0.002364441752433777, -0.05951501801609993, -0.0500493124127388, 0.21102115511894226, 0.03539281338453293, 0.38269317150115967, -0.29260092973709106, 0.2069980353116989, 0.05926353484392166, 0.04391039162874222, -0.11114301532506943, -0.21921727061271667, 0.05030050128698349, -0.15784099698066711, 0.03383420407772064, 0.2952827513217926, 0.09981479495763779, 0.1530807912349701, -0.2855033874511719, -0.12453707307577133, -0.20802274346351624, 0.22792446613311768, -0.17904222011566162, -0.04890655726194382, -0.24299554526805878, 0.02101702243089676, -0.02587638422846794, -0.09469917416572571, -0.14167022705078125, -0.0381786972284317, 0.3095560371875763, 0.2532220482826233, -0.1722705215215683, -0.45553502440452576, 0.10826653242111206, 0.3432941436767578, 0.18004727363586426, -0.24588416516780853, 0.3368532359600067, 0.19381549954414368, -0.06877827644348145, -0.01412438228726387, 0.3971101641654968, 0.43576839566230774, 0.3383936583995819, -0.364220529794693, 0.08116793632507324, -0.20338064432144165, -0.05682189017534256, -0.023260291665792465, 0.2695123851299286, 0.10607316344976425, 0.11737622320652008, 0.46242257952690125, 0.12148632109165192, -0.1247272789478302, 0.04363914206624031, -0.08805669844150543, -0.2605326473712921, -0.26294705271720886, 0.5191487073898315, 0.1857946217060089, -0.1143883466720581, -0.32065102458000183, 0.1899377703666687, -0.27498704195022583, -0.08091447502374649, 0.06285165250301361, 0.24058954417705536, 0.11315394192934036, -0.21382759511470795, 0.07050890475511551, 0.00031514279544353485, 0.7061734795570374, 0.48772192001342773, -0.017719194293022156, -0.3658052682876587, -0.11652375012636185, -0.5989872217178345, 0.21783632040023804, -0.3639681041240692, -0.26901817321777344, 0.2833365499973297, -0.07233157753944397, -0.288082480430603, 0.24852731823921204, 0.1276196390390396, 0.15819591283798218, -0.05727645754814148, 0.09834779053926468, -0.4907650053501129, -0.14860455691814423, -0.030650503933429718, -0.007675016298890114, -0.08318105340003967, -0.25705307722091675, 0.14532306790351868, -0.4349946677684784, 0.11610303074121475, 0.01198504213243723, 0.06934809684753418, -0.00023945607244968414, 0.09664029628038406, 0.35442453622817993, 0.1653239130973816, 0.5012482404708862, -0.03556128963828087, 0.0036574434489011765, -0.1911543905735016, -0.1566217541694641, -0.2289923131465912, 0.13182267546653748, 0.08762918412685394, 0.5133118629455566, -0.02610740251839161, -0.06854520738124847, -0.26402729749679565, 0.40316787362098694, -0.08100667595863342, 0.08831582963466644, -0.1350807249546051, 0.0773145779967308, -0.0703495442867279, 0.2275840938091278, 0.2643648684024811, 0.42454031109809875, -0.1783217340707779, 0.2022092193365097, -0.2722919285297394, -0.46024608612060547, 0.32893532514572144, -0.661480724811554, -0.517790675163269, 0.0751224160194397, 0.1161930188536644, -0.1423342525959015, 0.05819430202245712, -0.6006339192390442, 0.07056698948144913, 0.24261271953582764, -0.1114056184887886, -0.317119836807251, 0.21794511377811432, 0.016542082652449608, 0.08000141382217407, 0.04357365146279335, 0.4483563303947449, -0.14847293496131897, -0.15976761281490326, -0.0049346014857292175, -0.163252055644989 ]
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.
[ -0.11933845281600952, 0.08862784504890442, -0.00897780992090702, 0.21019737422466278, 0.3692678213119507, 0.026059836149215698, 0.4487815499305725, 0.4131021499633789, 0.4330812692642212, 0.012566991150379181, 0.12384359538555145, 0.395429790019989, -0.15054914355278015, 0.021962707862257957, 0.12090519070625305, -0.42898982763290405, -0.03373972326517105, 0.3636728823184967, -0.041466742753982544, 0.06297364085912704, -0.1876835823059082, 0.10851715505123138, -0.2651914358139038, 0.25500577688217163, -0.30572283267974854, -0.05702577531337738, 0.2159208357334137, 0.08206366002559662, -0.07891867309808731, -0.5590065121650696, 0.2770831286907196, -0.2170492708683014, 0.06445172429084778, 0.11474639922380447, -0.00010995571938110515, 0.20530645549297333, 0.33269721269607544, 0.030212881043553352, -0.5799458026885986, -0.18957795202732086, -0.4773316979408264, -0.11893881857395172, 0.29743173718452454, -0.2497076690196991, -0.028513483703136444, 0.05389726161956787, 0.035910993814468384, -0.07441209256649017, 0.29674679040908813, 0.531506359577179, 0.237888902425766, 0.5236923694610596, 0.15454915165901184, 0.03449363261461258, -0.07856784015893936, -0.17057108879089355, -0.08211500197649002, 0.22396822273731232, 0.07387731224298477, -0.39154377579689026, -0.013373926281929016, 0.1269245445728302, 0.15117238461971283, 0.26370200514793396, 0.5083709359169006, 0.014393925666809082, 0.2405235767364502, -0.15062691271305084, 0.014946820214390755, 0.07640057057142258, 0.3353826403617859, -0.1649332046508789, -0.2715998888015747, -0.2243589460849762, 0.2476661205291748, -0.33475619554519653, 0.2365303635597229, -0.03667646646499634, -0.16220436990261078, -0.008197136223316193, -0.26922789216041565, -0.028147468343377113, -0.24182075262069702, 0.3526870608329773, -0.058214087039232254, 0.49964913725852966, 0.043583355844020844, 0.1427474468946457, -0.05806552246212959, -0.07545772939920425, 0.14037704467773438, -0.08709109574556351, -0.1293005496263504, 0.25781309604644775, -0.38315531611442566, -0.036467600613832474, -0.034946344792842865, -0.11697594076395035, -0.031624749302864075, 0.26352018117904663, 0.21998973190784454, -0.22487351298332214, 0.3238390386104584, 0.3154153525829315, 0.22298845648765564, 0.42046889662742615, 0.07276557385921478, 0.06359237432479858, -0.08280497789382935, 0.1360260546207428, -0.23372812569141388, -0.15548521280288696, -0.1555362045764923, -0.1765558421611786, 0.09289798140525818, -0.15460222959518433, 0.22602073848247528, -0.07111581414937973, -0.4984738230705261, 0.11472072452306747, -0.14340484142303467, 0.046290118247270584, 0.20511485636234283, 0.17948494851589203, -0.12763413786888123, 0.23223412036895752, 0.14412455260753632, 0.2825669050216675, -0.25197696685791016, -0.04313553497195244, -0.16628961265087128, 0.249213308095932, -0.23438286781311035, -0.18224890530109406, 0.24275390803813934, 0.193400040268898, 0.3994784653186798, -0.10116767883300781, 0.11126405000686646, -0.2017553448677063, 0.2273758500814438, -0.04551952704787254, -0.1288793981075287, 0.25665315985679626, -0.21606102585792542, 0.1548980474472046, 0.20159120857715607, -0.2824234366416931, -0.0848238617181778, -0.10174446552991867, -0.27263572812080383, -0.30809223651885986, -0.1262514442205429, 0.28141558170318604, 0.147381991147995, -0.24296480417251587, -0.15226313471794128, -0.06118778511881828, 0.14749255776405334, -0.36872372031211853, -0.11810629814863205, -0.4612586498260498, -0.22063903510570526, -0.023005738854408264, 0.3976435363292694, 0.20490935444831848, 0.13138674199581146, -0.1089726984500885, -0.11542053520679474, 0.07785429060459137, -0.031797103583812714, 0.1227136105298996, -0.43835723400115967, 0.5421629548072815, -0.06105594336986542, 0.118863046169281, 0.64134681224823, -0.3427872657775879, -0.4578544795513153, 0.12356892228126526, -0.014279965311288834, 0.20932787656784058, -0.22536897659301758, -0.01672065444290638, -0.013226330280303955, -0.1845511645078659, 0.21231377124786377, 0.6192131042480469, -0.015042958781123161, 0.08247934281826019, -0.12689706683158875, -0.01554920244961977, 0.5002195239067078, 0.10985688865184784, 0.09252531826496124, -0.007928885519504547, -0.04170481115579605, 0.49664852023124695, -0.00131293386220932, -0.12597057223320007, -0.014860082417726517, -0.013630155473947525, -0.06468622386455536, -0.10460402071475983, -0.2178265005350113, -0.07388680428266525, -0.2776644825935364, 0.30430400371551514, -0.11754798144102097, -0.05853193253278732, -0.04715031385421753, 0.10618656873703003, -0.34911325573921204, 0.16848869621753693, -0.42326927185058594, -0.3773188889026642, 0.17209337651729584, 0.15513363480567932, -0.01082286611199379, 0.23844115436077118, -0.20474110543727875, 0.21943674981594086, -0.2740277051925659, 0.055500876158475876, -0.21892870962619781, 0.16128991544246674, -0.05588398873806, -0.04242202267050743, -0.009479247033596039, 0.30879634618759155, 0.03833160549402237, -0.017996493726968765, -0.3062889277935028, 0.28603029251098633, -0.12548455595970154, 0.08035629242658615, -0.03226495534181595, -0.14337028563022614, 0.06880630552768707, -0.07154297828674316, -0.17518453299999237, 0.20819191634655, 0.33195099234580994, -0.18012642860412598, -0.04314299672842026, 0.11287609487771988, 0.1272123008966446, 0.0814221054315567, 0.09886433184146881, 0.16397228837013245, 0.14399199187755585, -0.06582621484994888, 0.1012795940041542, 0.07743129134178162, 0.23370946943759918, -0.09498715400695801, 0.20471079647541046, -0.15645107626914978, -0.1671449989080429, -0.11880645900964737, 0.3131202161312103, 0.17005868256092072, 0.16741719841957092, 0.08148638904094696, -0.2752525210380554, -0.09487432986497879, 0.07855785638093948, 0.14604909718036652, 0.40096497535705566, -0.0518328957259655, -0.10834059119224548, -0.05257079750299454, -0.4165968596935272, 0.07112551480531693, 0.17070795595645905, 0.03662925213575363, 0.20912905037403107, 0.30185645818710327, 0.04131554812192917, 0.0873207226395607, -0.2952670454978943, -0.35701698064804077, 0.2824336886405945, 0.4405299723148346, -0.22848191857337952, 0.20788007974624634, -0.03053159825503826, -0.0025112517178058624, -0.06784830242395401, -0.33363378047943115, -0.18862853944301605, -0.342993825674057, -0.12133617699146271, 0.433775395154953, 0.024119451642036438, 0.181004598736763, -0.3977614641189575, -0.022051095962524414, 0.19475087523460388, -0.11478548496961594, -0.03137951344251633, -0.3157978653907776, -0.2693289816379547, 0.04594217613339424, 0.3818606734275818, 0.09107664972543716, 0.3510980010032654, -0.06677059084177017, -0.0925356075167656, -0.007035041227936745, -0.05841507762670517, -0.07556726038455963, -0.18372085690498352, 0.2440476268529892, 0.009521827101707458, 0.2855835258960724, -0.17702968418598175, -0.5220333337783813, 0.5017467737197876, -0.25570181012153625, -0.057785287499427795, 0.31510666012763977, 0.27111297845840454, -0.24919047951698303, -0.07947683334350586, -0.37468862533569336, -0.12645035982131958, -0.28060582280158997, -0.05946876108646393, 0.07482925057411194, 0.062020909041166306, 0.39746737480163574, 0.07085651159286499, 0.21601088345050812, -0.0025735688395798206, 0.1134379655122757, -0.04826278239488602, -0.09006398171186447, 0.1859847605228424, -0.31003400683403015, -0.31731829047203064, -0.12134282290935516, 0.00993594154715538, 0.3347616195678711, 0.11073551326990128, -0.3489593267440796, -0.21413181722164154, -0.10495822131633759, 0.08531276881694794, -0.2213253676891327, -0.04006670415401459, 0.18286007642745972, 0.07230357825756073, -0.06572244316339493, -0.08664388954639435, 0.03598948195576668, 0.07124010473489761, 0.15234598517417908, 0.13496194779872894, -0.006304748356342316, 0.08922688663005829, -0.08488400280475616, 0.6008991599082947, 0.21869370341300964, -0.21437688171863556, 0.30028587579727173, -0.10327673703432083, 0.005144510418176651, -0.208683580160141, -0.4398045539855957, 0.26416414976119995, -0.11523276567459106, 0.18292897939682007, 0.17231473326683044, -0.07949357479810715, -0.2247839719057083, -0.1925591230392456, 0.06885170936584473, -0.17844745516777039, -0.27277958393096924, 0.14616836607456207, -0.21739450097084045, 0.2787109315395355, -0.06886422634124756, 0.22253704071044922, -0.20522335171699524, -0.16798070073127747, 0.06985078752040863, 0.16693411767482758, -0.09054567664861679, 0.07967513799667358, -0.24400167167186737, -0.27610862255096436, -0.4799582362174988, -0.02783966064453125, 0.24103179574012756, 0.643750011920929, -0.08638451993465424, 0.03718879818916321, 0.08072253316640854, 0.08174140751361847, 0.13737376034259796, -0.2301362007856369, 0.1822635680437088, 0.2891308069229126, 0.03182467073202133, -0.31974661350250244, -0.06433149427175522, -0.14011231064796448, 0.35801658034324646, 0.0766519159078598, -0.054431721568107605, -0.03723123297095299, -0.1676540970802307, 0.1398320496082306, 0.2599392831325531, -0.1872951090335846, -0.044113077223300934, -0.09173431247472763, -0.08981046825647354, -0.30776458978652954, -0.006154969334602356, -0.009947758167982101, 0.33566054701805115, 0.12296926230192184, 0.26572471857070923, -0.5462992191314697, 0.11665705591440201, 0.09016524255275726, 0.17458684742450714, 0.09595315158367157, 0.057944219559431076, 0.18724678456783295, -0.16663400828838348, 0.18321000039577484, -0.14590442180633545, 0.5633803606033325, 0.14737597107887268, -0.3307987451553345, -0.31952953338623047, 0.02166324108839035, 0.16743338108062744, 0.18398034572601318, -0.01858552172780037, -0.14112097024917603, -0.2769453525543213, -0.06146351993083954, -0.008708748035132885, 0.3045704662799835, 0.5610411167144775, 0.0026276633143424988, -0.3271808922290802, -0.2585407495498657, 0.4828394949436188, -0.2850031852722168, 0.11732904613018036, 0.3215572237968445, 0.05337715521454811, -0.3131922781467438, 0.08413887023925781, 0.07783256471157074, 0.6767752766609192, 0.16953882575035095, 0.3111467659473419, 0.3826707899570465, 0.18392826616764069, 0.4142706096172333, -0.043089497834444046, 0.2847142815589905, -0.3927357494831085, -0.13342586159706116, -0.08068586140871048, -0.06546270102262497, 0.10708172619342804, -0.008873388171195984, -0.02543865703046322, 0.150621235370636, 0.15688925981521606, -0.0041420795023441315, 0.05714789777994156, 0.4492526650428772, -0.03878391534090042, -0.1968192607164383, -0.3470253050327301, 0.15601636469364166, -0.21799974143505096, 0.2975102663040161, -0.09073382616043091, -0.06648421287536621, 0.05520467087626457, -0.06735203415155411, -0.25555843114852905, 0.2574385404586792, -0.29947492480278015, 0.4594579339027405, 0.05782182514667511, -0.4756813049316406, 0.1984151303768158, 0.20709790289402008, 0.1756638139486313, 0.0056989602744579315, -0.29959404468536377, 0.23877887427806854, -0.2900899052619934, -0.22186626493930817, -0.06728766113519669, 0.11194741725921631, 0.4200734496116638, -0.17255154252052307, -0.14637798070907593, 0.14794065058231354, -0.12065417319536209, -0.24150244891643524, 0.1039588451385498, 0.036866605281829834, 0.3524894714355469, -0.3925756514072418, -0.5152195692062378, -0.0024207234382629395, -0.18207629024982452, -0.17491021752357483, 0.17450003325939178, 0.02345065027475357, -0.18903149664402008, 0.15189561247825623, 0.14982710778713226, -0.19898000359535217, 0.11640655994415283, 0.24804237484931946, -0.2495403289794922, 0.14628303050994873, 0.7637573480606079, 0.03394555673003197, 0.01280517503619194, -0.3138304352760315, -0.05428845435380936, 0.12508459389209747, -0.48199987411499023, -0.1171272024512291, 0.12550412118434906, 0.16912586987018585, -0.1442466676235199, 0.49836599826812744, 0.06302111595869064, -0.01884474605321884, 0.24718548357486725, -0.5424140095710754, -0.10621928423643112, -0.043021634221076965, 0.017753619700670242, 0.23777522146701813, -0.0026239529252052307, 0.11488902568817139, 0.13550958037376404, 0.1033887192606926, -0.2726374566555023, 0.014254427514970303, -0.15762582421302795, 0.1379965841770172, 0.21783430874347687, -0.1565312147140503, 0.22328168153762817, -0.07694721221923828, 0.15638616681098938, -0.09584659337997437, -0.12303443253040314, -0.24829328060150146, -0.25483784079551697, 0.10011143982410431, 0.24556711316108704, -0.1645212471485138, -0.030679188668727875, -0.10233347862958908, 0.02367764711380005, -0.23110505938529968, -0.00846264511346817, -0.11772798001766205, -0.2933427095413208, 0.38943231105804443, 0.04389800876379013, 0.27836859226226807, -0.11169030517339706, 0.12635567784309387, -0.30075740814208984, 0.14912036061286926, -0.01849590800702572, -0.07522959262132645, 0.27581319212913513, 0.09365326166152954, -0.4658869206905365, 0.0054205358028411865, -0.40260517597198486, 0.03714786097407341, 0.3698212504386902, -0.4551103711128235, 0.15995843708515167, -0.1316281408071518, 0.09546546638011932, 0.4656829833984375, -0.27140846848487854, -0.0036669299006462097, 0.3496653139591217, 0.23784254491329193, -0.29741764068603516, -0.1892606019973755, 0.2399907410144806, 0.061345674097537994, 0.03245936706662178, 0.02723889797925949, 0.04666122421622276, -0.3037486672401428, -0.32123446464538574, 0.08927963674068451, -0.029340408742427826, -0.2656012773513794, -0.01140776090323925, 0.33932653069496155, -0.002364441752433777, -0.05951501801609993, -0.0500493124127388, 0.21102115511894226, 0.03539281338453293, 0.38269317150115967, -0.29260092973709106, 0.2069980353116989, 0.05926353484392166, 0.04391039162874222, -0.11114301532506943, -0.21921727061271667, 0.05030050128698349, -0.15784099698066711, 0.03383420407772064, 0.2952827513217926, 0.09981479495763779, 0.1530807912349701, -0.2855033874511719, -0.12453707307577133, -0.20802274346351624, 0.22792446613311768, -0.17904222011566162, -0.04890655726194382, -0.24299554526805878, 0.02101702243089676, -0.02587638422846794, -0.09469917416572571, -0.14167022705078125, -0.0381786972284317, 0.3095560371875763, 0.2532220482826233, -0.1722705215215683, -0.45553502440452576, 0.10826653242111206, 0.3432941436767578, 0.18004727363586426, -0.24588416516780853, 0.3368532359600067, 0.19381549954414368, -0.06877827644348145, -0.01412438228726387, 0.3971101641654968, 0.43576839566230774, 0.3383936583995819, -0.364220529794693, 0.08116793632507324, -0.20338064432144165, -0.05682189017534256, -0.023260291665792465, 0.2695123851299286, 0.10607316344976425, 0.11737622320652008, 0.46242257952690125, 0.12148632109165192, -0.1247272789478302, 0.04363914206624031, -0.08805669844150543, -0.2605326473712921, -0.26294705271720886, 0.5191487073898315, 0.1857946217060089, -0.1143883466720581, -0.32065102458000183, 0.1899377703666687, -0.27498704195022583, -0.08091447502374649, 0.06285165250301361, 0.24058954417705536, 0.11315394192934036, -0.21382759511470795, 0.07050890475511551, 0.00031514279544353485, 0.7061734795570374, 0.48772192001342773, -0.017719194293022156, -0.3658052682876587, -0.11652375012636185, -0.5989872217178345, 0.21783632040023804, -0.3639681041240692, -0.26901817321777344, 0.2833365499973297, -0.07233157753944397, -0.288082480430603, 0.24852731823921204, 0.1276196390390396, 0.15819591283798218, -0.05727645754814148, 0.09834779053926468, -0.4907650053501129, -0.14860455691814423, -0.030650503933429718, -0.007675016298890114, -0.08318105340003967, -0.25705307722091675, 0.14532306790351868, -0.4349946677684784, 0.11610303074121475, 0.01198504213243723, 0.06934809684753418, -0.00023945607244968414, 0.09664029628038406, 0.35442453622817993, 0.1653239130973816, 0.5012482404708862, -0.03556128963828087, 0.0036574434489011765, -0.1911543905735016, -0.1566217541694641, -0.2289923131465912, 0.13182267546653748, 0.08762918412685394, 0.5133118629455566, -0.02610740251839161, -0.06854520738124847, -0.26402729749679565, 0.40316787362098694, -0.08100667595863342, 0.08831582963466644, -0.1350807249546051, 0.0773145779967308, -0.0703495442867279, 0.2275840938091278, 0.2643648684024811, 0.42454031109809875, -0.1783217340707779, 0.2022092193365097, -0.2722919285297394, -0.46024608612060547, 0.32893532514572144, -0.661480724811554, -0.517790675163269, 0.0751224160194397, 0.1161930188536644, -0.1423342525959015, 0.05819430202245712, -0.6006339192390442, 0.07056698948144913, 0.24261271953582764, -0.1114056184887886, -0.317119836807251, 0.21794511377811432, 0.016542082652449608, 0.08000141382217407, 0.04357365146279335, 0.4483563303947449, -0.14847293496131897, -0.15976761281490326, -0.0049346014857292175, -0.163252055644989 ]
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)
[ 0.15341989696025848, -0.037310101091861725, -0.06269726157188416, -0.12907017767429352, 0.17716169357299805, 0.08376287668943405, 0.2535592317581177, 0.2304861843585968, -0.21282333135604858, 0.33027225732803345, -0.07075794786214828, 0.7361391186714172, -0.22251440584659576, -0.2018119841814041, 0.023780349642038345, -0.08247566223144531, 0.2442183494567871, -0.21178573369979858, -0.1650036722421646, 0.027072928845882416, -0.11413007974624634, -0.38973113894462585, 0.023244716227054596, -0.14219342172145844, -0.19263318181037903, -0.0498332604765892, -0.13309884071350098, 0.04740720987319946, -0.24616168439388275, -0.25444748997688293, -0.46736636757850647, 0.7014355659484863, -0.15274126827716827, 0.449937105178833, -0.00011275450378889218, -0.2929297387599945, 0.3552502989768982, 0.032352328300476074, -0.3119528293609619, 0.07475161552429199, -0.1722264140844345, -0.07203903049230576, 0.06607111543416977, -0.20177561044692993, 0.21404674649238586, 0.05777452886104584, 0.056991904973983765, -0.1298425793647766, 0.16466911137104034, -0.0894339457154274, 0.17472481727600098, -0.09979449957609177, -0.30139559507369995, -0.01625042036175728, 0.5358918905258179, 0.5136167407035828, -0.15510791540145874, 0.11450314521789551, 0.5382506251335144, 0.3613971173763275, -0.1880669891834259, 0.156181201338768, -0.09367482364177704, -0.07671932876110077, 0.48861464858055115, -0.29398173093795776, -0.5896617770195007, 0.04415927454829216, 0.12085763365030289, 0.1790134310722351, 0.6000992059707642, -0.20126229524612427, -0.41691845655441284, -0.24042877554893494, 0.26787981390953064, -0.13337336480617523, 0.05386372655630112, -0.19137075543403625, -0.012277732603251934, -0.1094486266374588, -0.13056577742099762, -0.15746831893920898, -0.005668371915817261, -0.05863003432750702, 0.4977310597896576, 0.529700517654419, 0.12314771115779877, -0.019478455185890198, 0.1149400919675827, 0.004155231639742851, 0.5591570734977722, -0.10236585140228271, 0.014495473355054855, 0.2615671157836914, -0.11023664474487305, -0.10021746903657913, -0.007673762738704681, 0.2563236355781555, 0.2494794726371765, 0.11320117115974426, 0.3054129481315613, 0.2701375186443329, -0.06867581605911255, 0.07461673766374588, -0.06943447887897491, -0.07609111070632935, 0.0326872393488884, -0.04483646899461746, 0.2888600826263428, -0.3153234124183655, 0.179962158203125, -0.11023575812578201, -0.15132713317871094, -0.05018090084195137, 0.03848663344979286, 0.12834730744361877, -0.3014702796936035, -0.10199998319149017, 0.19806785881519318, -0.5291101932525635, 0.06318304687738419, -0.4272432327270508, 0.07826334983110428, 0.11914832144975662, 0.29036223888397217, 0.11539510637521744, -0.15124547481536865, 0.049449652433395386, 0.29246804118156433, -0.1264243870973587, -0.06676331162452698, 0.0325259193778038, -0.474370539188385, 0.2319803237915039, 0.1612490564584732, -0.2171570211648941, 0.10984321683645248, 0.08951105922460556, 0.15006762742996216, 0.22949722409248352, 0.16365483403205872, 0.2171189785003662, 0.32886338233947754, -0.1775379329919815, -0.44875964522361755, -0.06935002654790878, -0.09622713923454285, 0.29492223262786865, -0.2965107858181, 0.2620725631713867, -0.27368491888046265, -0.2621244490146637, -0.06397712230682373, 0.12658850848674774, -0.025156602263450623, -0.10126557946205139, -0.2362121343612671, 0.20045337080955505, 0.11881725490093231, -0.3033382296562195, 0.3808515965938568, 0.1356823444366455, 0.18717367947101593, -0.10355515033006668, -0.030979741364717484, 0.259865403175354, -0.09520679712295532, -0.020418502390384674, -0.3389224112033844, -0.10294423997402191, 0.18109747767448425, 0.11230626702308655, -0.16153833270072937, 0.023971259593963623, -0.12279897183179855, -0.1310102790594101, 0.48198848962783813, 0.21702727675437927, -0.17160694301128387, -0.03480614349246025, -0.15349331498146057, -0.16826380789279938, 0.5057481527328491, 0.32181769609451294, -0.07960885018110275, -0.2573795020580292, 0.28341472148895264, -0.09259987622499466, -0.4131712019443512, 0.037119559943675995, -0.23320911824703217, -0.13467749953269958, 0.38062286376953125, 0.2380649596452713, -0.0918971449136734, 0.001127682626247406, 0.0451047383248806, -0.4092564880847931, 0.18580007553100586, -0.1324031800031662, -0.12164705246686935, -0.252774715423584, -0.029326658695936203, 0.3391715884208679, -0.08386562019586563, -0.09441967308521271, -0.03489300608634949, -0.15264181792736053, 0.2866232097148895, -0.10843174159526825, 0.028408020734786987, -0.29122602939605713, 0.038068488240242004, -0.3927571177482605, -0.004562634974718094, -0.12355299293994904, 0.03388376533985138, 0.24871128797531128, -0.24041326344013214, -0.23405371606349945, -0.2861441969871521, -0.19298109412193298, -0.1690485179424286, -0.1284978985786438, -0.22818197309970856, -0.05662815272808075, -0.04576103389263153, -0.046052467077970505, -0.27629995346069336, 0.20862233638763428, -0.030198633670806885, -0.08184047043323517, -0.17129185795783997, 0.26135411858558655, 0.08897054195404053, -0.13293933868408203, 0.43658536672592163, 0.345750629901886, -0.0626278668642044, 0.09248360991477966, 0.1802336722612381, -0.056374941021203995, -0.01670355722308159, -0.09874457120895386, -0.2994474768638611, 0.2603796124458313, -0.2665768563747406, 0.09091302007436752, -0.2621859610080719, -0.318847119808197, -0.047044798731803894, 0.37200817465782166, -0.26494839787483215, 0.1697956919670105, 0.1168682873249054, 0.2092730849981308, 0.35136839747428894, 0.16266946494579315, -0.3234311044216156, -0.139518141746521, 0.26496991515159607, -0.159983828663826, -0.02203412353992462, 0.34199294447898865, 0.20328062772750854, -0.24940231442451477, 0.14528779685497284, 0.06851378083229065, 0.37311041355133057, 0.3408048152923584, 0.012890361249446869, 0.07192341238260269, 0.279762327671051, -0.1501164585351944, -0.04826968163251877, 0.16651786863803864, -0.28262168169021606, -0.01117776334285736, 0.17486003041267395, 0.12027426064014435, -0.1523294895887375, -0.38304707407951355, 0.07604926824569702, 0.032207027077674866, -0.03617866337299347, -0.07769874483346939, -0.018525294959545135, -0.07378381490707397, -0.2008935809135437, -0.1971311718225479, -0.3390846252441406, -0.2624276578426361, 0.26744744181632996, 0.05849086120724678, -0.29521071910858154, 0.38806024193763733, 0.18326818943023682, 0.5487815737724304, -0.2152482271194458, -0.10795969516038895, -0.35612156987190247, -0.4411761164665222, 0.12635695934295654, 0.040955934673547745, 0.22586002945899963, -0.1359720677137375, 0.7025216817855835, -0.051004648208618164, -0.07998791337013245, -0.382279634475708, -0.4698299467563629, -0.08981539309024811, -0.0074404142796993256, -0.25710999965667725, 0.382809042930603, -0.32705962657928467, 0.08849550783634186, -0.47281864285469055, -0.06896976381540298, -0.2552247941493988, -0.010260610841214657, 0.007352788001298904, -0.010249239392578602, 0.009276734665036201, -0.18420101702213287, -0.29757723212242126, -0.13124610483646393, -0.40497976541519165, 0.074064239859581, -0.6318063735961914, -0.0500297024846077, 0.012037180364131927, -0.28683409094810486, -0.1324150413274765, -0.22056880593299866, 0.16137036681175232, -0.13669763505458832, -0.7596741318702698, 0.19291149079799652, -0.20223425328731537, 0.057574350386857986, -0.1422053724527359, -0.06971137225627899, 0.028892938047647476, 0.6446990370750427, 0.2754821181297302, -0.16692981123924255, 0.14853470027446747, 0.20165473222732544, -0.0016144923865795135, 0.15614524483680725, 0.3470441699028015, 0.28477585315704346, -0.12292183190584183, 0.06234151870012283, -0.23541176319122314, -0.11293720453977585, -0.1072697564959526, 0.31545379757881165, 0.10719215124845505, 0.3211008310317993, 0.01664702221751213, 0.6867965459823608, -0.6602864861488342, -0.09058082103729248, 0.01934916153550148, 0.20021802186965942, 0.13724634051322937, -0.06796775013208389, -0.09096982330083847, 0.5424612760543823, 0.24561837315559387, -0.3071576952934265, 0.19216035306453705, -0.04839305579662323, -0.16012327373027802, 0.037762805819511414, 0.1311073750257492, -0.14749892055988312, -0.14253079891204834, 0.34279873967170715, -0.32730862498283386, 0.08625137805938721, -0.04006420075893402, 0.17369800806045532, -0.29519936442375183, 0.004260499030351639, 0.14684420824050903, 0.20264247059822083, 0.32880860567092896, -0.27043113112449646, -0.43609973788261414, -0.021180186420679092, -0.12659403681755066, 0.29588377475738525, 0.2474624216556549, 0.12883858382701874, -0.013604741543531418, -0.18180373311042786, 0.122365303337574, 0.2576231360435486, 0.4825328588485718, -0.24698953330516815, -0.28336113691329956, -0.005569253116846085, -0.016896996647119522, -0.11249476671218872, 0.015213456004858017, -0.075557641685009, 0.14184683561325073, 0.07289402186870575, 0.09383468329906464, -0.22726130485534668, -0.17683012783527374, 0.11363396048545837, 0.19612252712249756, 0.16109314560890198, -0.33927249908447266, -0.08296407759189606, -0.16535870730876923, -0.15990211069583893, 0.19259142875671387, -0.023795748129487038, -0.30083730816841125, -0.055608510971069336, 0.05883060023188591, -0.03348209708929062, 0.3097732365131378, 0.026684239506721497, 0.12922780215740204, 0.03208727389574051, 0.1218257024884224, 0.04497941955924034, 0.01223859004676342, 0.37701907753944397, 0.42766568064689636, -0.20563459396362305, -0.48697629570961, -0.170944482088089, 0.01451437920331955, -0.5441601276397705, 0.24406981468200684, 0.23828548192977905, 0.2514321208000183, 0.29675230383872986, -0.14040754735469818, 0.1760656088590622, -0.218506321310997, 0.07353851944208145, 0.05243784189224243, 0.007404102012515068, -0.5293942093849182, -0.9117103815078735, 0.5416220426559448, 0.37718644738197327, -0.32241392135620117, 0.5554148554801941, 0.0005213767290115356, -0.2684897184371948, 0.4150931239128113, 0.2951989471912384, 1.2055474519729614, -0.08146865665912628, 0.18745803833007812, -0.0292167030274868, -0.3767351508140564, 0.27660760283470154, -0.18090739846229553, 0.14939947426319122, -0.1293390691280365, -0.39364707469940186, 0.05398126691579819, -0.08897186815738678, 0.07474078983068466, 0.3279804587364197, -0.12489880621433258, 0.34531208872795105, -0.3932633399963379, 0.17307855188846588, 0.1445833146572113, 0.12142482399940491, -0.24394631385803223, -0.29141977429389954, -0.08227938413619995, 0.07282465696334839, -0.09159129858016968, -0.21026615798473358, -0.033486489206552505, -0.005069621838629246, -0.06700876355171204, 0.14088626205921173, -0.14437377452850342, -0.3057858645915985, 0.15224632620811462, -0.0800418108701706, -0.2814944088459015, -0.17594867944717407, 0.4433322846889496, 0.10858981311321259, 0.2535843253135681, 0.07985029369592667, 0.16678963601589203, 0.2920089066028595, 0.08756481856107712, 0.40555620193481445, -0.0955696776509285, -0.08928123861551285, 0.10746953636407852, 0.08212121576070786, -0.048722848296165466, 0.030048537999391556, 0.11088299751281738, -0.42860132455825806, -0.5298833847045898, -0.08485546708106995, 0.5617538690567017, -0.1269344538450241, 0.09261015057563782, 0.0187520831823349, -0.10598739981651306, -0.01749984733760357, 0.13704358041286469, 0.278785765171051, -0.2586563229560852, 0.36239713430404663, -0.21821726858615875, 0.019990622997283936, -0.1200646460056305, 0.10490235686302185, -0.2883315980434418, -0.12774547934532166, 0.21792252361774445, -0.2194971889257431, -0.19734559953212738, -0.16577619314193726, 0.11748946458101273, 0.233148455619812, -0.004312954843044281, 0.0334511324763298, -0.13740471005439758, -0.27819085121154785, -0.1548794060945511, -0.15670913457870483, 0.21951740980148315, 0.3490118682384491, -0.03535391390323639, 0.11037425696849823, -0.4601903259754181, 0.07673440128564835, 0.1705334335565567, 0.3398687541484833, -0.16756445169448853, -0.04664931073784828, -0.11511196941137314, -0.019233787432312965, -0.3286210894584656, -0.1494593471288681, 0.2203085720539093, 0.27189552783966064, 0.009397382847964764, 0.22101452946662903, -0.2014082372188568, -0.24054953455924988, 0.06552565842866898, 0.06110464408993721, -0.018643217161297798, -0.14807040989398956, -0.13654419779777527, 0.11171083897352219, -0.07000520825386047, 0.2294294387102127, 0.20375722646713257, -0.0818171575665474, -0.002820374444127083, -0.25039294362068176, 0.02648245543241501, 0.29474928975105286, 0.005087465047836304, 0.4968271255493164, 0.6184066534042358, 0.0818525031208992, -0.21739962697029114, 0.20662236213684082, -0.025502480566501617, 0.031439997255802155, 0.12063337862491608, 0.14772596955299377, -0.026562681421637535, -0.020382672548294067, 0.11907152831554413, 0.3216317892074585, 0.18479996919631958, 0.06783601641654968, -0.09052736312150955, 0.2885022461414337, 0.193085715174675, 0.45367830991744995, 0.09952227026224136, 0.4589996635913849, 0.0012196345487609506, -0.35758090019226074, 0.43149542808532715, 0.2065579891204834, -0.019028961658477783, -0.07362893968820572, -0.2535402774810791, -0.005758973769843578, 0.0805494636297226, -0.005397506058216095, -0.22965003550052643, 0.015039943158626556, -0.3449506163597107, -0.07759958505630493, 0.14512397348880768, -0.478800892829895, 0.3961937129497528, 0.3324783146381378, -0.12170538306236267, -0.02795560285449028, 0.3708794414997101, 0.009444952011108398, -0.26959097385406494, 0.13245993852615356, 0.4628448486328125, 0.9553411602973938, 0.3267765939235687, 0.16063372790813446, 0.13742747902870178, 0.14922748506069183, 0.16059593856334686, 0.6927061676979065, 0.341178834438324, 0.0890379548072815, 0.3850887715816498, 0.22731852531433105, 0.14771747589111328, -0.3402741849422455, 0.2771570682525635, 0.07764579355716705, 0.25607356429100037, 0.030500028282403946, -0.06635604798793793, -0.19421157240867615, -0.2519179582595825, -0.08347742259502411, 0.012616932392120361, 0.051535092294216156, 0.5799620151519775, 0.2349970042705536, 0.25914323329925537, -0.3528333902359009, 0.126642644405365, 0.2718268632888794, 0.4411761164665222, -0.15010902285575867, -0.40551549196243286, 0.14202538132667542, -0.17151226103305817, 0.745175838470459, -0.01977100782096386, -0.001275382936000824, 0.2458317130804062, 0.029737230390310287, 0.1183306872844696, -0.3746950328350067, -0.03710135072469711, -0.01562034897506237, 0.12326828390359879, 0.14358177781105042, -0.17993007600307465, -0.012090180069208145, 0.17657999694347382, -0.14344370365142822, 0.07518965005874634, -0.07518407702445984, -0.04627443477511406, -0.1835743933916092, 0.344208300113678, 0.49961671233177185, -0.30887603759765625, 0.1062355786561966, -0.08491865545511246, -0.1956767439842224, -0.15276408195495605, 0.04055929183959961, 0.16102313995361328, -0.20532777905464172, 0.09134699404239655, 0.05906745046377182, 0.2930762767791748, 0.28379178047180176, 0.2766534090042114, -0.007268339395523071, -0.14981020987033844, -0.39398545026779175, -0.3894713819026947, -0.1301598846912384, 0.4760110080242157, 0.0036115478724241257, -0.03547752648591995, -0.1831997185945511, 0.03684106469154358, -0.06707943230867386, 0.14567506313323975, -0.1224084123969078, 0.05333206057548523, 0.19794237613677979, -0.357058048248291, -0.1212211549282074, -0.13139519095420837, -0.022572332993149757, 0.1836550533771515, -0.061004143208265305, -0.30679693818092346, 0.41008299589157104, 0.04413165897130966, -0.3097118139266968, 0.1448230892419815, 0.06988412141799927, 0.11397361755371094, 0.21174103021621704, -0.019559109583497047, 0.388078898191452, -0.0024604089558124542, -0.1821300983428955, 0.026386026293039322, -0.14211679995059967, -0.5724015831947327, 0.12119083851575851, 0.007544107735157013, -0.011270652525126934, -0.03836609795689583, -0.20877541601657867, -0.44213399291038513, 0.548570454120636, -0.35532158613204956, 0.08550754934549332, -0.7277730703353882, 0.15805087983608246, 0.24726364016532898, 0.08188671618700027, -0.04858000949025154, 0.11439962685108185, 0.05100638046860695, 0.27076050639152527, -0.22954663634300232, 0.19431617856025696, 0.5219244956970215, 0.16950620710849762, -0.24273917078971863, -0.32140734791755676, 0.21311473846435547, -0.08961277455091476, 0.029533321037888527, -0.31957271695137024, -0.08377373218536377, 0.23225055634975433, 0.003435719758272171, 0.07984821498394012, 0.38395240902900696, 0.1284504383802414, -0.0466567799448967, -0.22277195751667023, -0.001449093222618103, -0.1201058104634285, -0.030343756079673767, -0.11485505104064941, -0.2884330153465271 ]
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.
[ 0.07046830654144287, 0.016060546040534973, -0.081076979637146, -0.13309326767921448, 0.18675069510936737, -0.018765434622764587, 0.3612537086009979, 0.2494378238916397, -0.1316787302494049, 0.39812397956848145, -0.11517098546028137, 0.745942234992981, -0.32974594831466675, -0.21496982872486115, -0.04452400282025337, -0.07981567084789276, 0.2653793692588806, -0.1031481921672821, -0.2052428424358368, 0.0562237948179245, -0.12470757961273193, -0.34814876317977905, 0.02367430180311203, -0.14052945375442505, -0.08968804776668549, 0.01218579150736332, -0.11303740739822388, 0.0714677944779396, -0.2733202874660492, -0.218613401055336, -0.4450688660144806, 0.7157349586486816, -0.1049821525812149, 0.566881000995636, -0.00010911502613453194, -0.25887179374694824, 0.341471791267395, 0.032926589250564575, -0.35604017972946167, 0.1413135975599289, -0.13315832614898682, -0.07831971347332001, 0.038680098950862885, -0.272943913936615, 0.22207000851631165, -0.03439842164516449, 0.11589889973402023, -0.11278656870126724, 0.11099021136760712, -0.1473732888698578, 0.1819334626197815, -0.09936526417732239, -0.3846816420555115, 0.002091312548145652, 0.5785199999809265, 0.46103423833847046, -0.2091200351715088, 0.18245258927345276, 0.5745741724967957, 0.3853928744792938, -0.33382806181907654, 0.1915493905544281, -0.13094337284564972, -0.1019473671913147, 0.5779687762260437, -0.30015552043914795, -0.539605975151062, 0.06476444005966187, -0.03939133137464523, 0.20368918776512146, 0.6476193070411682, -0.19023606181144714, -0.4480341076850891, -0.1510586142539978, 0.18167032301425934, -0.04967717081308365, 0.03078971803188324, -0.24211889505386353, -0.01957586407661438, -0.08532571792602539, -0.03737848252058029, -0.11297060549259186, -0.002562493085861206, -0.13636167347431183, 0.4217982590198517, 0.5597809553146362, 0.15498246252536774, -0.0286143496632576, 0.23609453439712524, -0.022240005433559418, 0.5372042655944824, -0.06967476010322571, 0.07333254814147949, 0.23729828000068665, -0.21732141077518463, -0.18081368505954742, -0.015533536672592163, 0.211710125207901, 0.18797332048416138, 0.10574356466531754, 0.30266353487968445, 0.2721402049064636, 0.013140109367668629, 0.1095646321773529, -0.11448945105075836, -0.03855254873633385, 0.033464789390563965, -0.10943488031625748, 0.37629491090774536, -0.2285958230495453, 0.15266044437885284, -0.08162569999694824, -0.26237422227859497, -0.01783922128379345, 0.10215289890766144, 0.14140647649765015, -0.413308322429657, 0.056872230023145676, 0.17915107309818268, -0.491369366645813, 0.15620630979537964, -0.42619937658309937, 0.10317113250494003, 0.18962308764457703, 0.18701566755771637, 0.13203954696655273, -0.15503767132759094, -0.0127902552485466, 0.21155117452144623, -0.18474049866199493, -0.10223613679409027, 0.04918678104877472, -0.507145345211029, 0.2194851040840149, 0.10138165950775146, -0.2031901627779007, 0.11597613245248795, 0.03873898833990097, 0.13651734590530396, 0.3439450263977051, 0.1533035933971405, 0.20069317519664764, 0.19247902929782867, -0.2556219696998596, -0.4391716718673706, -0.12956403195858002, -0.14438550174236298, 0.3152363896369934, -0.2803958058357239, 0.24230843782424927, -0.2930678427219391, -0.23662276566028595, -0.11380833387374878, 0.17173609137535095, -0.001863684505224228, -0.06199924647808075, -0.2293267548084259, 0.2463952898979187, 0.07400448620319366, -0.30563846230506897, 0.4155254364013672, 0.07851982861757278, 0.17123425006866455, -0.12333601713180542, -0.07254890352487564, 0.2618194818496704, -0.11126986145973206, -0.018436145037412643, -0.23025205731391907, -0.16228759288787842, 0.2056494504213333, 0.11422605812549591, -0.19200213253498077, 0.00789506733417511, -0.1419527530670166, -0.14744582772254944, 0.4907086193561554, 0.18126294016838074, -0.15881508588790894, -0.012354780919849873, -0.11443603038787842, -0.10257536917924881, 0.6056382656097412, 0.3467119634151459, 0.05649430304765701, -0.33416399359703064, 0.2879277765750885, -0.06120084226131439, -0.3102132976055145, 0.09815352410078049, -0.22475816309452057, -0.11841296404600143, 0.38146328926086426, 0.26114925742149353, -0.12281814217567444, -0.009176865220069885, 0.0056611523032188416, -0.3373616337776184, 0.19471287727355957, -0.17169393599033356, -0.07312556356191635, -0.23634667694568634, -0.08199551701545715, 0.30896878242492676, -0.015054728835821152, -0.06498883664608002, -0.06311679631471634, -0.1157575398683548, 0.29175838828086853, 0.0176498144865036, 0.18580931425094604, -0.281282901763916, 0.029021091759204865, -0.2989603579044342, -0.01543935015797615, -0.17469657957553864, 0.061628445982933044, 0.26166319847106934, -0.22212836146354675, -0.2475365847349167, -0.2646145820617676, -0.1795038878917694, -0.27028512954711914, -0.16231746971607208, -0.25778427720069885, -0.10035745054483414, -0.04210995137691498, 0.027301959693431854, -0.2138468325138092, 0.2204703390598297, -0.13025018572807312, -0.07318171113729477, -0.15486472845077515, 0.24804913997650146, 0.11062881350517273, 0.0012296810746192932, 0.30390554666519165, 0.3766377568244934, -0.023014439269900322, 0.07872435450553894, 0.30762720108032227, -0.08393964171409607, 0.012866853736341, -0.11360392719507217, -0.27618739008903503, 0.2542322874069214, -0.23281978070735931, 0.08788233995437622, -0.2713022232055664, -0.3496970534324646, -0.008261360228061676, 0.35968276858329773, -0.33686205744743347, 0.23720219731330872, 0.12871089577674866, 0.27883729338645935, 0.4074001610279083, 0.2508475184440613, -0.3290272057056427, -0.13281577825546265, 0.3740355968475342, -0.12329480051994324, -0.07980761677026749, 0.3697061836719513, 0.12849664688110352, -0.2949621081352234, 0.11040480434894562, 0.08891589939594269, 0.3699168860912323, 0.34708428382873535, 0.07254628837108612, -0.033813491463661194, 0.331180214881897, -0.13317978382110596, -0.010054424405097961, 0.09691943973302841, -0.13696503639221191, -0.02087973803281784, 0.10166246443986893, 0.058118030428886414, -0.1998406946659088, -0.5052339434623718, 0.033564887940883636, 0.09178837388753891, 0.03976184502243996, -0.07727980613708496, 0.05568966269493103, -0.05265963077545166, -0.14146248996257782, -0.17583584785461426, -0.2801251709461212, -0.27149176597595215, 0.2531289756298065, 0.10748499631881714, -0.3382929563522339, 0.33368679881095886, 0.18314342200756073, 0.5316723585128784, -0.15840722620487213, -0.05841858685016632, -0.37355005741119385, -0.447194367647171, 0.15386386215686798, 0.07673262059688568, 0.2285132259130478, -0.1827123463153839, 0.6215428113937378, -0.035527683794498444, -0.06387980282306671, -0.3513753414154053, -0.3830469250679016, -0.06246766448020935, -0.004240013659000397, -0.2293926179409027, 0.3913438320159912, -0.22182045876979828, 0.11684173345565796, -0.4907088279724121, -0.005510867573320866, -0.3185173571109772, -0.10767782479524612, 0.027148324996232986, 0.025267956778407097, 0.013321095146238804, -0.16682873666286469, -0.24212759733200073, -0.17180229723453522, -0.4843711853027344, 0.03587126359343529, -0.5416005253791809, -0.002558000385761261, 0.01305378321558237, -0.2118820697069168, -0.10356827825307846, -0.11099404841661453, 0.15742120146751404, -0.15449339151382446, -0.816232442855835, 0.22808286547660828, -0.18251317739486694, 0.07005289196968079, -0.22892799973487854, -0.11542637646198273, 0.08359460532665253, 0.6077487468719482, 0.22179462015628815, -0.06396794319152832, 0.13452968001365662, 0.13789765536785126, 0.03936556726694107, 0.16842582821846008, 0.26220953464508057, 0.30263376235961914, -0.1816999614238739, 0.0207793191075325, -0.2253594547510147, -0.08990485221147537, -0.12235292047262192, 0.22594287991523743, 0.17040753364562988, 0.36077162623405457, -0.015084872022271156, 0.7392673492431641, -0.6775937080383301, -0.053755078464746475, -0.02820456027984619, 0.16907565295696259, 0.13554586470127106, -0.03121870756149292, -0.14977528154850006, 0.559766411781311, 0.21935686469078064, -0.3056308627128601, 0.23087020218372345, -0.087122343480587, -0.24886958301067352, -0.011342715471982956, 0.06998847424983978, -0.041594069451093674, -0.057806506752967834, 0.3630118668079376, -0.30870136618614197, 0.01324857771396637, -0.06105046346783638, 0.12251811474561691, -0.32691770792007446, 0.009557018056511879, 0.15242917835712433, 0.22981412708759308, 0.265505313873291, -0.24354392290115356, -0.4098946452140808, 0.044229451566934586, -0.2146960198879242, 0.16246865689754486, 0.2266128957271576, 0.0795103907585144, 0.09746839106082916, -0.2776513695716858, 0.06662573665380478, 0.2089672088623047, 0.5011021494865417, -0.12202250212430954, -0.2938186526298523, -0.01813543401658535, -0.03504668548703194, -0.08999653160572052, -0.0249855387955904, -0.11665748059749603, 0.1255786269903183, 0.02066122367978096, 0.09768398106098175, -0.2659021317958832, -0.16656813025474548, 0.07446404546499252, 0.13256365060806274, 0.07663963735103607, -0.335916668176651, -0.1441165655851364, -0.0007815398275852203, -0.1548815667629242, 0.17788606882095337, -0.09627924859523773, -0.28561320900917053, -0.06143783777952194, -0.021943816915154457, 0.036408379673957825, 0.29397955536842346, 0.07353497296571732, 0.2143486738204956, 0.057643432170152664, 0.07084514200687408, 0.04822836443781853, -0.06505827605724335, 0.2817290723323822, 0.43656712770462036, -0.1428242325782776, -0.46511268615722656, -0.10632561147212982, 0.13579952716827393, -0.48085305094718933, 0.24971437454223633, 0.11656621098518372, 0.21496297419071198, 0.19017577171325684, -0.14731937646865845, 0.21787649393081665, -0.2683399021625519, -0.01057024858891964, 0.061114661395549774, 0.028212152421474457, -0.4887728691101074, -0.9323238730430603, 0.5213586091995239, 0.36476394534111023, -0.258247971534729, 0.574323296546936, 0.004262492060661316, -0.2041110247373581, 0.41235750913619995, 0.3018031418323517, 1.1519230604171753, -0.09614545106887817, 0.218508780002594, 0.10118705779314041, -0.21098646521568298, 0.29550737142562866, -0.2298135906457901, 0.15573906898498535, -0.15916036069393158, -0.3745524287223816, 0.06611093133687973, -0.13276295363903046, 0.04989742860198021, 0.2634168267250061, -0.25604724884033203, 0.25369787216186523, -0.3084162473678589, 0.23445823788642883, 0.020436149090528488, 0.19144318997859955, -0.28746283054351807, -0.3286157548427582, -0.05082404613494873, 0.0998365730047226, -0.05817987769842148, -0.30008700489997864, -0.047375887632369995, -0.05468501150608063, -0.09370657801628113, 0.2028755247592926, -0.10154703259468079, -0.2596939504146576, 0.1305818408727646, -0.05969154089689255, -0.2657773494720459, -0.09929939359426498, 0.36674004793167114, 0.03436394780874252, 0.1254948526620865, -0.014877259731292725, 0.15132880210876465, 0.24826131761074066, 0.03901910036802292, 0.46660685539245605, -0.045037683099508286, -0.13180314004421234, 0.24407753348350525, 0.06336642801761627, 0.005415201187133789, 0.04518643021583557, 0.11685143411159515, -0.48749643564224243, -0.42851921916007996, -0.04317504167556763, 0.5997399091720581, -0.15401963889598846, 0.11646611988544464, 0.04915709048509598, -0.19502615928649902, -0.048161327838897705, 0.1584094613790512, 0.2610589563846588, -0.23258623480796814, 0.2867871820926666, -0.1204795241355896, 0.04268600046634674, -0.1251039355993271, 0.08967898786067963, -0.1612808108329773, -0.11361165344715118, 0.19274240732192993, -0.23611801862716675, -0.280558705329895, -0.19184023141860962, 0.15163633227348328, 0.21312810480594635, -0.02919405698776245, 0.1103457510471344, -0.23565185070037842, -0.3742779493331909, -0.07946967333555222, -0.19264337420463562, 0.23270660638809204, 0.3861812353134155, -0.07074115425348282, 0.12319789826869965, -0.4934324026107788, 0.08805854618549347, 0.09681490063667297, 0.3548377752304077, -0.24070832133293152, -0.09278306365013123, -0.10960809886455536, -0.01845471002161503, -0.3788088262081146, -0.2516290843486786, 0.15548188984394073, 0.1993105262517929, 0.00818982720375061, 0.11616284400224686, -0.22874048352241516, -0.10670122504234314, 0.09685687720775604, 0.08086872100830078, -0.07961455732584, -0.17176371812820435, -0.17849420011043549, 0.09644432365894318, -0.09953518956899643, 0.21651527285575867, 0.21649432182312012, -0.06108008697628975, -0.0931975394487381, -0.24933061003684998, 0.051228735595941544, 0.2615835964679718, -0.059606075286865234, 0.4868055284023285, 0.6710013747215271, 0.0972437933087349, -0.20602212846279144, 0.16008427739143372, 0.07366210967302322, 0.05982646346092224, 0.08756186813116074, 0.12772925198078156, -0.036417748779058456, -0.09296658635139465, 0.10473387688398361, 0.4397200345993042, 0.15501439571380615, 0.13724616169929504, -0.002952631562948227, 0.28087615966796875, 0.19764989614486694, 0.3586438000202179, 0.18620380759239197, 0.43419361114501953, 0.041121114045381546, -0.39689844846725464, 0.412551611661911, 0.23424574732780457, -0.0356530100107193, -0.13068056106567383, -0.2073366641998291, -0.11624658852815628, 0.13849030435085297, 0.010381326079368591, -0.14836148917675018, 0.006765358150005341, -0.21708612143993378, -0.13100191950798035, 0.09854527562856674, -0.4322226047515869, 0.36990851163864136, 0.24644212424755096, -0.1015394926071167, -0.07392777502536774, 0.442527174949646, 0.01766747608780861, -0.18382644653320312, 0.16107571125030518, 0.3885459899902344, 0.8879134654998779, 0.3873404860496521, 0.14933861792087555, 0.09435451030731201, 0.17374691367149353, 0.14424824714660645, 0.7236194610595703, 0.23297666013240814, 0.11240538954734802, 0.3068799376487732, 0.2533043622970581, 0.10200060158967972, -0.2223387360572815, 0.27811363339424133, 0.14876356720924377, 0.13228145241737366, 0.011391542851924896, -0.03571192920207977, -0.22587159276008606, -0.3367294669151306, -0.09032933413982391, -0.013227701187133789, 0.0849902480840683, 0.5246771574020386, 0.19994014501571655, 0.29892781376838684, -0.2885637581348419, 0.14000903069972992, 0.20761984586715698, 0.4345974624156952, -0.21771685779094696, -0.42913076281547546, 0.1927124559879303, -0.11657305061817169, 0.6738446950912476, 0.05358833074569702, -0.03115168586373329, 0.2896183431148529, 0.054473698139190674, 0.07489553838968277, -0.3837040662765503, -0.03711394593119621, -0.06795172393321991, 0.10829363018274307, 0.10009673237800598, -0.23522530496120453, 0.04172064736485481, 0.17708905041217804, -0.199933260679245, 0.016714664176106453, 0.03517799824476242, -0.14711062610149384, -0.1254277229309082, 0.23322010040283203, 0.5211758613586426, -0.23741364479064941, 0.10508299618959427, -0.02069588005542755, -0.31418731808662415, -0.1590208113193512, -0.029609132558107376, 0.07011150568723679, -0.1647290587425232, 0.047691188752651215, 0.08573614805936813, 0.24409614503383636, 0.3486331105232239, 0.2794223725795746, 0.009575147181749344, -0.1195920929312706, -0.41816824674606323, -0.40806230902671814, -0.13012906908988953, 0.44738849997520447, 0.02127697505056858, -0.02329091727733612, -0.036524027585983276, 0.03883352875709534, -0.03901775926351547, 0.16553083062171936, -0.13707776367664337, 0.01628265157341957, 0.11755913496017456, -0.36923402547836304, -0.10993663966655731, -0.09451670944690704, -0.05248396843671799, 0.18723328411579132, -0.12909914553165436, -0.2749139070510864, 0.31997203826904297, 0.1062961295247078, -0.24907808005809784, 0.0901089459657669, 0.15677806735038757, 0.1838327795267105, 0.2313612699508667, -0.10574929416179657, 0.22225704789161682, -0.013972029089927673, -0.18763671815395355, 0.06642089039087296, -0.08607544004917145, -0.48912352323532104, 0.009830731898546219, 0.09424670040607452, 0.0846363976597786, -0.02032383903861046, -0.22575189173221588, -0.45795518159866333, 0.5146231651306152, -0.30500298738479614, 0.08379412442445755, -0.7182446718215942, 0.21887990832328796, 0.23858317732810974, 0.051791317760944366, -0.15571333467960358, 0.12745359539985657, 0.10186199843883514, 0.2214876413345337, -0.30647704005241394, 0.16515307128429413, 0.4485177993774414, 0.10767001658678055, -0.23511822521686554, -0.28615713119506836, 0.1609083116054535, -0.043487586081027985, 0.17592792212963104, -0.3066343665122986, 0.00241696834564209, 0.21606038510799408, 0.011468365788459778, 0.10717856884002686, 0.41457149386405945, 0.05682273954153061, 0.005426015704870224, -0.20953309535980225, 0.02015519142150879, -0.14458653330802917, -0.0305701345205307, -0.11404657363891602, -0.28600752353668213 ]
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
[ 0.11682143807411194, 0.047994956374168396, -0.008072571828961372, -0.2174365073442459, 0.12888166308403015, 0.028689570724964142, 0.1363588273525238, 0.360643208026886, -0.2647383213043213, 0.2979813814163208, 0.011635839007794857, 0.851158618927002, -0.35360410809516907, -0.2320978194475174, 0.20139333605766296, -0.07050041109323502, 0.15841257572174072, -0.10356594622135162, -0.24149152636528015, 0.0523638054728508, -0.13913841545581818, -0.38281068205833435, -0.07822690159082413, -0.2000361829996109, -0.13133971393108368, -0.09400457888841629, -0.08385643362998962, 0.13230089843273163, -0.18401826918125153, -0.29359671473503113, -0.4079645276069641, 0.6355051398277283, -0.10218510776758194, 0.14192354679107666, -0.00010813646076712757, -0.27224457263946533, 0.3262631893157959, 0.13107790052890778, -0.43213951587677, 0.044686466455459595, -0.08185350894927979, -0.0996280163526535, 0.00863670278340578, -0.15263867378234863, 0.06552892178297043, 0.13389600813388824, 0.16543443500995636, -0.1464117467403412, 0.24235886335372925, 0.050310708582401276, 0.20914514362812042, -0.03290870040655136, -0.19009999930858612, 0.15847425162792206, 0.5000157952308655, 0.40686893463134766, -0.10122678428888321, 0.028628256171941757, 0.43952375650405884, 0.22933261096477509, -0.2657989263534546, 0.2744673192501068, -0.06643892079591751, -0.18149206042289734, 0.3078671991825104, -0.28791874647140503, -0.568967878818512, 0.07428516447544098, 0.0025422051548957825, 0.17878586053848267, 0.5037114024162292, -0.28533515334129333, -0.3903697729110718, -0.35262978076934814, 0.3302518129348755, -0.18850016593933105, 0.01999017596244812, -0.09641744941473007, -0.05681595578789711, -0.06574320793151855, -0.24524720013141632, -0.15989506244659424, 0.03680569678544998, -0.021753549575805664, 0.5161433219909668, 0.6055986881256104, 0.20056715607643127, -0.003385476768016815, 0.24499811232089996, 0.08339643478393555, 0.5854985117912292, -0.009341668337583542, 0.03303344547748566, 0.2996443808078766, -0.14720936119556427, -0.20837321877479553, 0.055003777146339417, 0.29483646154403687, 0.2364351451396942, 0.14461249113082886, 0.32707205414772034, 0.2851001024246216, -0.04130205139517784, 0.03735366836190224, 0.005551239475607872, 0.03812050819396973, -0.05310115963220596, 0.05495560169219971, 0.2810693383216858, -0.3015080690383911, 0.23124808073043823, 0.003970738500356674, -0.04903753101825714, -0.08059713989496231, -0.010975509881973267, 0.13244429230690002, -0.19454528391361237, -0.06849871575832367, 0.03613021969795227, -0.5815836787223816, -0.04963257908821106, -0.32450762391090393, 0.14889642596244812, 0.10779420286417007, 0.27703583240509033, 0.1407214254140854, -0.10792308300733566, -0.048173874616622925, 0.18350951373577118, -0.23938734829425812, -0.10505374521017075, 0.025838693603873253, -0.4939265847206116, 0.17078769207000732, 0.17374013364315033, -0.07522200047969818, 0.1390148103237152, 0.1129436269402504, 0.06702779978513718, 0.23749428987503052, 0.07884752750396729, 0.2658466696739197, 0.4057009220123291, -0.20365069806575775, -0.4380703270435333, -0.13343045115470886, -0.10840372741222382, 0.2805593013763428, -0.3744527995586395, 0.2528225779533386, -0.2730662226676941, -0.2303438037633896, -0.0922333374619484, 0.1565236747264862, -0.09696942567825317, -0.2698223888874054, -0.17591553926467896, 0.3530561923980713, 0.06264600157737732, -0.4234463572502136, 0.35462215542793274, 0.07788464426994324, 0.062279801815748215, -0.0040478818118572235, 0.013162774965167046, 0.317663311958313, 0.04695447161793709, 0.020499326288700104, -0.24038124084472656, 0.0018358547240495682, 0.23789846897125244, 0.2391386330127716, -0.19989801943302155, -0.02130229026079178, -0.16164132952690125, -0.061269547790288925, 0.5229777097702026, 0.1896456927061081, -0.20807000994682312, 0.013185974210500717, -0.20950418710708618, -0.09727853536605835, 0.40916791558265686, 0.3883783519268036, -0.12241677939891815, -0.33623144030570984, 0.20757310092449188, 0.11739514768123627, -0.41046637296676636, 0.1484217792749405, -0.3570765256881714, -0.07999233901500702, 0.4687560200691223, 0.31442922353744507, -0.24346338212490082, 0.000865645706653595, 0.07365517318248749, -0.1594797521829605, 0.2940753102302551, -0.1337454915046692, -0.1370370090007782, -0.23566274344921112, -0.06926561892032623, 0.128549724817276, -0.17178975045681, -0.1168624758720398, -0.10402866452932358, -0.23739498853683472, 0.22362768650054932, -0.05178038030862808, 0.11644144356250763, -0.2859424948692322, 0.08745760470628738, -0.3385698199272156, 0.07401012629270554, -0.16828955709934235, 0.09142883121967316, 0.11526472866535187, -0.10391446948051453, -0.16962818801403046, -0.22857993841171265, -0.0651368796825409, -0.14694315195083618, -0.1881207376718521, -0.3230237066745758, 0.029987024143338203, -0.11075789481401443, -0.05123117193579674, -0.07592548429965973, 0.46679234504699707, -0.08376158028841019, -0.062399785965681076, -0.0960649847984314, 0.29552963376045227, -0.00012784404680132866, -0.2677119970321655, 0.37398555874824524, 0.2960720658302307, 0.033665966242551804, 0.035146210342645645, 0.09655119478702545, 0.033621616661548615, -0.05559798330068588, -0.002862803637981415, -0.3500950336456299, 0.31631022691726685, -0.1683957725763321, 0.02897990494966507, -0.18219545483589172, -0.25976240634918213, 0.06748025119304657, 0.2224581241607666, -0.3342282772064209, 0.1662607342004776, 0.21991662681102753, 0.21182863414287567, 0.2813102900981903, 0.22115714848041534, -0.475361168384552, 0.067124143242836, 0.49271923303604126, -0.05404485762119293, 0.018465764820575714, 0.34422391653060913, 0.25495633482933044, -0.332937628030777, 0.12259738892316818, 0.12657177448272705, 0.2766236960887909, 0.4469537138938904, 0.05266153812408447, 0.14309260249137878, 0.09512905031442642, -0.1651533991098404, -0.044339440762996674, -0.06922054290771484, -0.17121653258800507, -0.02813558280467987, 0.2366146445274353, 0.14595496654510498, -0.28768274188041687, -0.32734712958335876, -0.04693536460399628, 0.18919777870178223, -0.1102178692817688, -0.05883181840181351, -0.13308194279670715, -0.30620077252388, -0.3187927007675171, -0.09166985750198364, -0.3395278751850128, -0.3069959878921509, 0.3917925953865051, 0.010386575013399124, -0.32073935866355896, 0.40423503518104553, 0.14643113315105438, 0.5348381996154785, -0.27478501200675964, -0.011294098570942879, -0.2897605895996094, -0.43033847212791443, -0.08097264170646667, 0.07382319867610931, 0.28863680362701416, -0.047781944274902344, 0.712497353553772, 0.07893392443656921, -0.13275235891342163, -0.30444157123565674, -0.5318078994750977, -0.1408616453409195, 0.03139061853289604, -0.2137415111064911, 0.3421265482902527, -0.39242321252822876, 0.09010970592498779, -0.38826945424079895, -0.18767426908016205, -0.34289199113845825, -0.08024118840694427, 0.12296096980571747, 0.0008933634962886572, 0.02239943854510784, -0.29992955923080444, -0.34487882256507874, -0.07238560169935226, -0.4497673213481903, 0.07709676027297974, -0.6393797993659973, 0.07205920666456223, 0.05661386623978615, -0.2364700734615326, -0.16641592979431152, -0.37279024720191956, 0.007765317335724831, -0.12797074019908905, -0.5969664454460144, 0.16613708436489105, -0.11037240922451019, -0.061282236129045486, -0.2556629180908203, -0.05603986978530884, 0.01381276547908783, 0.6228564977645874, 0.19484542310237885, -0.13971500098705292, 0.10687891393899918, 0.07648119330406189, -0.08116407692432404, 0.15028256177902222, 0.43438106775283813, 0.16205082833766937, -0.12345730513334274, 0.11876928061246872, -0.18264272809028625, 0.042115334421396255, -0.015398278832435608, 0.32869142293930054, 0.034825649112463, 0.2597433030605316, 0.13255898654460907, 0.5310877561569214, -0.5287259817123413, -0.08033180236816406, 0.05685708671808243, 0.17602059245109558, 0.25830376148223877, -0.09530036896467209, 0.0022557098418474197, 0.5387944579124451, 0.29065459966659546, -0.253852516412735, 0.25488215684890747, -0.0269765667617321, -0.30920788645744324, 0.12093102931976318, 0.11350800096988678, -0.0975208729505539, -0.1581999510526657, 0.4676395356655121, -0.33594322204589844, 0.02855132520198822, 0.05402672663331032, 0.12251193076372147, -0.2675395905971527, -0.1677858531475067, 0.007246201857924461, 0.04410355165600777, 0.3468672037124634, -0.13314172625541687, -0.4249151349067688, -0.045601293444633484, -0.29521673917770386, 0.19519026577472687, 0.26279452443122864, 0.27447035908699036, -0.015354800969362259, -0.17939487099647522, 0.1379479169845581, 0.1270141303539276, 0.3991087079048157, -0.1852073073387146, -0.1568528115749359, -0.08336517214775085, 0.1259494572877884, -0.1888723522424698, 0.053663674741983414, -0.11359268426895142, 0.10668960958719254, 0.16241836547851562, 0.05929304659366608, -0.26366710662841797, -0.19537679851055145, 0.13751189410686493, 0.08628317713737488, 0.1633119285106659, -0.3234401047229767, -0.07841929793357849, -0.19659286737442017, -0.1358698308467865, 0.2577889859676361, -0.006182191893458366, -0.2898067533969879, -0.03264281153678894, 0.08358713984489441, -0.05086030066013336, 0.22913530468940735, -0.008727006614208221, 0.17341579496860504, 0.05967771261930466, 0.21080222725868225, -0.02444550395011902, -0.008577316999435425, 0.3802028000354767, 0.201516255736351, -0.17109259963035583, -0.35532546043395996, 0.0014370828866958618, -0.080220527946949, -0.3615971505641937, 0.2507231831550598, 0.21086831390857697, 0.21007803082466125, 0.2614561915397644, -0.14829213917255402, 0.15075713396072388, -0.2337452471256256, -0.030188381671905518, -0.06791020929813385, 0.007533682510256767, -0.5339679718017578, -0.8507195711135864, 0.5450387597084045, 0.3243734836578369, -0.3054797649383545, 0.4437330663204193, -0.024501033127307892, -0.35627874732017517, 0.3782590627670288, 0.32997459173202515, 1.0034143924713135, -0.07321757078170776, 0.2827591300010681, 0.16532638669013977, -0.3735780119895935, 0.42987287044525146, -0.18536461889743805, 0.2949046492576599, -0.19988366961479187, -0.26987776160240173, 0.10410730540752411, -0.0036480799317359924, -0.028021305799484253, 0.28348779678344727, -0.23576804995536804, 0.31240272521972656, -0.16876855492591858, 0.0294245183467865, 0.12953603267669678, 0.2141418755054474, -0.26988521218299866, -0.22724027931690216, -0.24185234308242798, 0.13649655878543854, -0.22734926640987396, -0.21245168149471283, 0.02145717665553093, -0.15386484563350677, 0.007626291364431381, 0.06421861052513123, -0.1773071587085724, -0.11582428216934204, 0.11999904364347458, -0.13377799093723297, -0.19474442303180695, -0.06841358542442322, 0.5051286816596985, 0.015424598008394241, 0.2500437796115875, 0.07930870354175568, 0.14135032892227173, 0.23059959709644318, 0.07139067351818085, 0.34186220169067383, -0.04602603241801262, -0.03625039383769035, 0.1730341613292694, 0.02574169635772705, 0.0010225772857666016, -0.059546031057834625, 0.13033847510814667, -0.6264157891273499, -0.47416695952415466, -0.008948206901550293, 0.320462703704834, -0.025614703074097633, 0.1331462562084198, 0.23690786957740784, -0.07707677036523819, -0.08411955088376999, 0.15529413521289825, 0.23977577686309814, -0.2477659285068512, 0.3586961030960083, -0.11311745643615723, -0.03689306974411011, -0.07945579290390015, 0.15876634418964386, -0.23023149371147156, 0.015225932002067566, 0.3005438446998596, -0.23293356597423553, -0.13038402795791626, -0.23513087630271912, -0.019167520105838776, 0.23810626566410065, -0.06477047502994537, 0.1519562005996704, -0.13655030727386475, -0.22427085041999817, -0.06621584296226501, -0.13610884547233582, 0.05216206982731819, 0.2807183265686035, -0.05464257299900055, 0.013736039400100708, -0.39536499977111816, 0.020800083875656128, 0.09897324442863464, 0.3021586835384369, -0.1149682030081749, 0.14674429595470428, -0.15087030827999115, 0.0054872483015060425, -0.37829846143722534, -0.24021054804325104, 0.11466757208108902, 0.319722443819046, 0.07316653430461884, 0.15754017233848572, -0.2105623334646225, -0.21844366192817688, 0.06116180866956711, 0.043341051787137985, -0.049846943467855453, -0.2375105619430542, -0.13929516077041626, 0.06823790818452835, -0.003146030008792877, 0.2037644386291504, 0.15675580501556396, -0.12760591506958008, -0.03967002406716347, -0.20938272774219513, -0.11454802751541138, 0.17652040719985962, -0.01926197111606598, 0.5665624737739563, 0.5050373673439026, 0.10235260426998138, 0.009776581078767776, 0.18582399189472198, -0.1405273675918579, -0.11232700943946838, 0.01957671530544758, 0.2767694592475891, -0.013695345260202885, 0.051071107387542725, 0.024323076009750366, 0.2786233723163605, 0.14185400307178497, 0.08936851471662521, -0.08146516233682632, 0.19317325949668884, 0.10954712331295013, 0.46652019023895264, 0.061829812824726105, 0.3582407236099243, 0.08235479146242142, -0.4855670928955078, 0.43374964594841003, 0.2181694358587265, -0.16130027174949646, -0.002209446160122752, -0.17344042658805847, 0.07349109649658203, 0.16542167961597443, -0.003153584897518158, -0.10503259301185608, -0.06873240321874619, -0.28761184215545654, -0.07359515130519867, 0.14936880767345428, -0.5486273765563965, 0.28238940238952637, 0.2594435214996338, -0.12507213652133942, -0.1608252227306366, 0.21707992255687714, -0.09827996790409088, -0.1600293517112732, 0.16166973114013672, 0.36011219024658203, 1.0041594505310059, 0.3350103497505188, 0.21191102266311646, 0.024365626275539398, 0.14911051094532013, 0.09338098764419556, 0.643322229385376, 0.3787447214126587, 0.1495162546634674, 0.429431289434433, 0.23128503561019897, 0.18977764248847961, -0.30213499069213867, 0.260072261095047, 0.1661073863506317, 0.165843665599823, 0.06580869853496552, -0.1154681071639061, -0.1193484514951706, -0.26598140597343445, -0.007739661261439323, -0.053818006068468094, 0.10528263449668884, 0.530920684337616, 0.18618106842041016, 0.16036461293697357, -0.41128939390182495, 0.20345550775527954, 0.2584473192691803, 0.4700574576854706, -0.18684901297092438, -0.24186114966869354, 0.025070182979106903, -0.21610897779464722, 0.7542920112609863, 0.04195919260382652, 0.02450689859688282, 0.28092706203460693, 0.04929617419838905, 0.03278752788901329, -0.4148055911064148, 0.009417077526450157, -0.15027835965156555, 0.17937560379505157, 0.2596067786216736, -0.12145720422267914, 0.021811317652463913, 0.13871459662914276, -0.11151222139596939, 0.20139601826667786, -0.04670361056923866, -0.07099401950836182, -0.08586189150810242, 0.3180975019931793, 0.5659241676330566, -0.2598892152309418, 0.04079608991742134, 0.029231250286102295, -0.13603153824806213, -0.06514067947864532, 0.00668475404381752, 0.14587558805942535, -0.17006808519363403, -0.056863535195589066, 0.09269674867391586, 0.35246074199676514, 0.4031998813152313, 0.2684458792209625, -0.10742684453725815, -0.11384580284357071, -0.36881470680236816, -0.5009536743164062, -0.1890978217124939, 0.3571246266365051, -0.051589444279670715, -0.028194501996040344, -0.022273287177085876, -0.01885136216878891, -0.018899455666542053, -0.04477250576019287, 0.035463664680719376, 0.013845345936715603, 0.07077226042747498, -0.3659396767616272, -0.013891540467739105, -0.20218102633953094, -0.07325318455696106, 0.14310477674007416, -0.1260942816734314, -0.2787930965423584, 0.314702570438385, 0.04463823139667511, -0.22550278902053833, 0.25094741582870483, -0.08069898188114166, 0.14602085947990417, 0.03393852710723877, 0.13528819382190704, 0.18925467133522034, 0.049447715282440186, -0.11370118707418442, -0.15378084778785706, -0.10955610871315002, -0.6963850855827332, 0.00513111799955368, 0.05858930945396423, 0.05486719310283661, -0.06912816315889359, -0.1053767055273056, -0.41169583797454834, 0.5068252682685852, -0.26187863945961, 0.0811077430844307, -0.5822639465332031, 0.17405733466148376, 0.19211417436599731, 0.1904579997062683, -0.14750538766384125, 0.21940958499908447, 0.09773136675357819, 0.18745581805706024, -0.2305743247270584, 0.10078959912061691, 0.46335476636886597, 0.0216558575630188, -0.1791427880525589, -0.21649959683418274, 0.2422758936882019, 0.01254146546125412, -0.14014916121959686, -0.36441338062286377, -0.1711253821849823, 0.17598265409469604, -0.04166674613952637, 0.2560969293117523, 0.46521297097206116, 0.04604590684175491, -0.04631691426038742, -0.18348293006420135, 0.0035772770643234253, -0.06137688457965851, -0.045355893671512604, 0.005076345056295395, -0.4476947784423828 ]
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
[ -0.2522469758987427, -0.2662777900695801, -0.12816189229488373, -0.382615327835083, -0.41568905115127563, -0.10087786614894867, 0.019268110394477844, 0.2192726880311966, -0.08714987337589264, 0.38095980882644653, 0.057025887072086334, 0.376335471868515, -0.01554610300809145, 0.38528168201446533, -0.07647743076086044, 0.14052680134773254, 0.08472877740859985, 0.36588799953460693, -0.08661046624183655, -0.1383289098739624, -0.20194710791110992, 0.11041399836540222, -0.06942584365606308, -0.0544472336769104, 0.14053839445114136, 0.17084470391273499, -0.1972423642873764, -0.07244310528039932, -0.2242700457572937, -0.2077137976884842, 0.35505497455596924, 0.4555104672908783, 0.22513926029205322, 0.06983283162117004, -0.00010584223491605371, -0.2381841242313385, 0.40260040760040283, -0.07296501845121384, -0.33391594886779785, -0.23844747245311737, -0.41790148615837097, -0.37712976336479187, 0.06569432467222214, -0.0959780290722847, 0.06359253823757172, 0.09290831536054611, 0.1650215983390808, -0.1966177076101303, 0.0641363337635994, 0.011999520473182201, 0.3026267886161804, 0.17412491142749786, -0.0875844731926918, -0.34514904022216797, 0.33494824171066284, 0.44910669326782227, -0.35373738408088684, -0.17945295572280884, 0.3167716860771179, 0.013063503429293633, -0.0673423781991005, 0.41108134388923645, 0.22772830724716187, -0.1817956268787384, -0.02522088587284088, -0.06428276002407074, 0.18126511573791504, 0.007419601082801819, 0.14173555374145508, 0.1168770045042038, 0.5061972737312317, 0.12980133295059204, -0.4155387878417969, 0.28998398780822754, 0.0042793480679392815, -0.4935304820537567, 0.2665930390357971, -0.10623874515295029, -0.029614266008138657, 0.06423866003751755, 0.18722155690193176, -0.33883601427078247, -0.49577194452285767, 0.20651434361934662, -0.07795708626508713, 0.09057144075632095, -0.15790341794490814, -0.15367048978805542, 0.12449894845485687, 0.10767865180969238, -0.043482277542352676, 0.041367921978235245, -0.1763661652803421, 0.204386368393898, -0.24313990771770477, -0.259062796831131, 0.2833954691886902, 0.0010785162448883057, 0.4588758945465088, 0.08281131088733673, 0.15400458872318268, 0.04369566962122917, -0.07180526852607727, 0.10415858030319214, 0.15376776456832886, 0.20157016813755035, 0.46074309945106506, -0.059857212007045746, 0.3992583155632019, 0.2871793508529663, 0.4768282175064087, 0.042006347328424454, 0.15518613159656525, 0.03422975912690163, -0.06027243658900261, -0.03193368390202522, 0.07255344837903976, -0.581581711769104, -0.0960402861237526, 0.050024598836898804, -0.01638912409543991, 0.01660105027258396, -0.03814193233847618, 0.09351342171430588, 0.2691534161567688, 0.22098153829574585, 0.2916286587715149, -0.06439065933227539, -0.035680435597896576, -0.4872015416622162, -0.2953397035598755, -0.07494331896305084, -0.2445649951696396, 0.2993691563606262, 0.32775935530662537, -0.42228981852531433, 0.13820597529411316, -0.31309938430786133, 0.2833760380744934, 0.04933442175388336, -0.18007932603359222, 0.33908727765083313, -0.22517728805541992, 0.12475208938121796, 0.057131920009851456, 0.08531032502651215, -0.27858254313468933, 0.5125010013580322, -0.1971522718667984, 0.1246589794754982, -0.33828338980674744, -0.4485531151294708, -0.386896550655365, 0.2870852053165436, -0.07405190914869308, -0.15551364421844482, -0.15182101726531982, 0.3666447699069977, -0.23981431126594543, -0.16440415382385254, -0.08050500601530075, 0.3010808229446411, -0.27182331681251526, -0.12088780105113983, 0.027140622958540916, 0.14325252175331116, -0.12814584374427795, -0.04009613022208214, -0.5082143545150757, -0.0936475545167923, -0.14256691932678223, -0.029935766011476517, -0.12085288017988205, -0.27354881167411804, -0.1699310541152954, -0.17351999878883362, 0.39941591024398804, -0.3332187235355377, -0.14639170467853546, 0.23286615312099457, -0.02869914099574089, -0.24974213540554047, 0.14834274351596832, -0.15650637447834015, 0.046719059348106384, -0.2703949213027954, -0.22819778323173523, 0.04900597780942917, -0.14814750850200653, -0.10404182225465775, -0.14730152487754822, -0.44127345085144043, -0.18274010717868805, 0.26036566495895386, 0.24740354716777802, 0.033463962376117706, 0.22792422771453857, -0.39961373805999756, 0.09001308679580688, -0.16994045674800873, 0.17561841011047363, -0.019764039665460587, 0.7601090669631958, 0.032792869955301285, 0.1584606170654297, 0.06392212212085724, -0.39667025208473206, 0.07201751321554184, 0.09055445343255997, 0.1692185401916504, 0.22833292186260223, -0.37668633460998535, -0.03511722758412361, -0.049209270626306534, 0.0924886167049408, -0.0949603021144867, 0.20418097078800201, 0.12047629058361053, -0.13210266828536987, -0.015266526490449905, -0.028222447261214256, 0.25926682353019714, -0.24022437632083893, -0.15056917071342468, -0.07579462230205536, 0.16699109971523285, -0.18648108839988708, 0.039145804941654205, -0.1825120449066162, 0.12863759696483612, -0.0777057483792305, -0.06598882377147675, -0.12712767720222473, 0.39034274220466614, -0.09755895286798477, 0.23098550736904144, 0.055191829800605774, 0.32341456413269043, 0.14523333311080933, -0.17544777691364288, 0.07596822828054428, -0.1623838245868683, -0.10134578496217728, 0.18791499733924866, -0.22592774033546448, 0.38410699367523193, 0.06922245770692825, -0.15123489499092102, 0.009792573750019073, -0.00877450779080391, 0.06245529651641846, -0.09672942012548447, -0.3642008900642395, -0.11718857288360596, 0.18756024539470673, 0.11140085756778717, -0.44378751516342163, -0.2857009470462799, -0.30886125564575195, -0.08802192658185959, 0.032737839967012405, -0.06471402943134308, 0.11028396338224411, 0.20297113060951233, -0.0012311190366744995, -0.23741555213928223, 0.3081510365009308, -0.10598641633987427, 0.27346473932266235, 0.3854374587535858, -0.1669514775276184, -0.021352270618081093, -0.20481492578983307, 0.11400623619556427, 0.1383962631225586, 0.28159111738204956, 0.04978880286216736, 0.10224251449108124, 0.3348848521709442, 0.09427326172590256, -0.35023343563079834, -0.30894413590431213, -0.04541849344968796, 0.05769772082567215, -0.15502676367759705, -0.19572332501411438, -0.033697210252285004, 0.07117685675621033, 0.0014656614512205124, 0.14784641563892365, 0.16372083127498627, -0.1914488673210144, 0.40738844871520996, 0.2692486047744751, -0.22280721366405487, 0.3243142068386078, 0.03676152601838112, 0.8047561049461365, -0.12738491594791412, 0.18538598716259003, -0.24488967657089233, 0.05357445776462555, -0.3132985532283783, 0.25463640689849854, -0.22676576673984528, -0.22785374522209167, 0.46085360646247864, -0.10080918669700623, 0.24583815038204193, -0.1181473508477211, -0.7101554870605469, 0.09934358298778534, -0.10333864390850067, -0.09342781454324722, 0.030385147780179977, -0.2146865874528885, -0.07387551665306091, 0.019448693841695786, 0.05901806429028511, -0.4485960006713867, -0.48253872990608215, -0.10999599099159241, -0.2480621039867401, 0.13051103055477142, -0.3649929463863373, -0.48683181405067444, 0.14573653042316437, -0.26064589619636536, -0.20767129957675934, -0.05296815186738968, 0.1615414321422577, 0.26981139183044434, 0.12356339395046234, -0.0581197552382946, 0.06886543333530426, 0.05131299048662186, -0.38603103160858154, 0.04582185670733452, 0.34243717789649963, -0.3188911974430084, -0.41584312915802, 0.1217738538980484, 0.011761464178562164, 0.15939754247665405, 0.17577895522117615, -0.3779391646385193, -0.39495590329170227, 0.27527758479118347, 0.12953297793865204, -0.1465403437614441, 0.06901110708713531, 0.5344862341880798, 0.06983810663223267, -0.21253933012485504, -0.07770728319883347, -0.4181554615497589, 0.18073105812072754, 0.4679791033267975, 0.1432115137577057, -0.294879287481308, 0.2655492424964905, 0.19312970340251923, 0.3202052414417267, 0.1647084504365921, -0.07552215456962585, 0.12141621112823486, 0.13372129201889038, 0.21813711524009705, -0.20591655373573303, 0.010713692754507065, 0.016583044081926346, 0.052660830318927765, 0.17323461174964905, 0.5349007844924927, 0.19639050960540771, -0.407214492559433, -0.372956246137619, -0.22823132574558258, 0.09779674559831619, 0.0007368624210357666, -0.1938171535730362, 0.1649022400379181, 0.44048231840133667, -0.06908129155635834, -0.15247789025306702, -0.3290403187274933, -0.10789782553911209, 0.09796702861785889, 0.34180572628974915, 0.15771353244781494, -0.08698377013206482, 0.09859806299209595, 0.17732566595077515, -0.2839728891849518, 0.3561050295829773, 0.14722788333892822, -0.0044860271736979485, 0.005454372614622116, -0.10187941044569016, 0.15186414122581482, 0.01983906701207161, 0.18591216206550598, -0.21365287899971008, -0.16913338005542755, -0.14002712070941925, -0.0866285115480423, 0.08970752358436584, -0.061822038143873215, -0.11171072721481323, -0.280530720949173, -0.09050488471984863, 0.05194654315710068, 0.027261562645435333, -0.14532747864723206, 0.6831696033477783, -0.17081420123577118, 0.010883715003728867, -0.2852770984172821, 0.05051552131772041, -0.2335626184940338, -0.0009023305028676987, -0.017939038574695587, -0.1901804506778717, 0.0005650762468576431, -0.10173087567090988, -0.15057452023029327, 0.2393195480108261, 0.013858072459697723, -0.006709679961204529, 0.14193856716156006, -0.14012393355369568, 0.011414280161261559, 0.11479316651821136, 0.08129552006721497, 0.04784281179308891, 0.3729478418827057, 0.12629103660583496, 0.22759857773780823, -0.26654544472694397, 0.1760542392730713, -0.19377730786800385, 0.06537304818630219, 0.2491941899061203, 0.030140243470668793, -0.06442669779062271, -0.2971714735031128, -0.22939170897006989, -0.123458631336689, -0.20089203119277954, 0.26194342970848083, 0.04154439643025398, -0.372658371925354, 0.0283846165984869, 0.3245108425617218, 0.02175328880548477, -0.2729002833366394, 0.2598053216934204, 0.21533317863941193, -0.24798910319805145, 0.3315582871437073, 0.0006595999002456665, 0.7347702383995056, 0.24774335324764252, -0.06344842910766602, 0.23774367570877075, -0.0478474497795105, 0.49948808550834656, -0.1533694863319397, 0.01370753813534975, 0.07889296114444733, -0.13275238871574402, -0.09385041147470474, 0.29843810200691223, 0.3770507276058197, -0.09206093847751617, -0.30580034852027893, 0.17827288806438446, 0.36761474609375, 0.24887485802173615, 0.08150078356266022, 0.18558502197265625, 0.3359355628490448, -0.3099146783351898, 0.1831713169813156, 0.188380628824234, 0.007580742239952087, 0.1698165386915207, 0.000819103792309761, -0.036344677209854126, 0.014058288186788559, -0.32612311840057373, -0.22565679252147675, -0.14075130224227905, 0.3442858159542084, -0.00929962657392025, -0.035762470215559006, -0.07222989201545715, 0.4353812336921692, -0.37264421582221985, 0.20798258483409882, 0.19100767374038696, -0.2896561026573181, 0.12577463686466217, 0.16576635837554932, 0.053724415600299835, -0.06959030777215958, -0.09448534995317459, 0.07702066004276276, -0.16798225045204163, -0.09140501171350479, 0.18939551711082458, 0.2116372287273407, -0.15778538584709167, -0.08024898171424866, -0.1400250792503357, -0.032127514481544495, 0.048933904618024826, 0.3648069202899933, 0.5734384059906006, 0.19827184081077576, -0.28032034635543823, 0.24688145518302917, 0.2291853427886963, -0.11168717592954636, -0.07789973169565201, 0.15205621719360352, -0.10007680952548981, -0.20936422049999237, 0.41510429978370667, 0.4110593795776367, -0.40592584013938904, -0.04722730815410614, -0.08289749920368195, -0.08956558257341385, -0.27430734038352966, -0.04644796997308731, 0.12019743025302887, -0.5605071783065796, -0.19973480701446533, 0.06941761821508408, -0.20039404928684235, 0.18112991750240326, 0.7872427701950073, 0.1409822404384613, 0.17524513602256775, -0.05200928449630737, -0.018956638872623444, -0.18221917748451233, 0.06492462754249573, -0.4313209652900696, 0.302185595035553, -0.026819203048944473, -0.1548498421907425, -0.11379803717136383, 0.0398104190826416, -0.4568730294704437, -0.19220370054244995, 0.2623075544834137, -0.13239438831806183, 0.37219303846359253, 0.31053996086120605, 0.26454198360443115, 0.0035354867577552795, 0.2119165062904358, -0.1414116472005844, -0.3338809013366699, -0.20281726121902466, 0.06465190649032593, 0.09212342649698257, 0.06677166372537613, 0.1733732521533966, -0.038244180381298065, -0.1149311438202858, -0.2642989158630371, 0.032552070915699005, 0.26225921511650085, -0.029773134738206863, -0.15817293524742126, 0.18510571122169495, -0.24483980238437653, -0.07198566943407059, 0.15404579043388367, 0.3823603689670563, 0.18938875198364258, 0.08267536759376526, -0.06526093930006027, 0.42594367265701294, -0.007294267416000366, -0.11266477406024933, 0.046603307127952576, 0.3781132400035858, -0.27740901708602905, 0.3002438545227051, 0.018674682825803757, -0.02458496391773224, -0.08572185039520264, -0.03785388171672821, 0.13398222625255585, -0.4113802909851074, -0.07079374045133591, 0.028906991705298424, 0.03366852179169655, 0.3719574809074402, -0.2801135182380676, 0.052501458674669266, -0.05817509815096855, -0.055172398686409, 0.21768569946289062, 0.19678078591823578, 0.34751248359680176, 0.015417419373989105, 0.7643715143203735, 0.032585494220256805, 0.01245035044848919, -0.053611353039741516, 0.4525500535964966, 0.1095074936747551, 0.0451970212161541, 0.09280000627040863, 0.08463910967111588, 0.06753060221672058, 0.1761351376771927, -0.06873731315135956, 0.11972801387310028, -0.08243213593959808, 0.4220132529735565, 0.5082911849021912, 0.009178526699543, -0.3382551372051239, 0.07828802615404129, 0.28291958570480347, 0.08967144042253494, 0.03318367153406143, 0.3009582757949829, -0.16590338945388794, 0.011286738328635693, 0.07916280627250671, -0.7445873618125916, -0.1521623581647873, 0.020007815212011337, -0.1553732007741928, -0.15849386155605316, -0.13477876782417297, -0.40616241097450256, -0.06609366834163666, 0.16439232230186462, 0.00754677876830101, 0.15479014813899994, -0.1995062679052353, 0.015690237283706665, -0.10653702914714813, -0.00798344612121582, 0.11110593378543854, 0.2240457534790039, 0.026626918464899063, 0.36857226490974426, -0.4948239326477051, -0.27720123529434204, 0.3008655607700348, 0.32901614904403687, -0.016461435705423355, -0.07080584019422531, 0.38324445486068726, 0.12766481935977936, 0.08333250880241394, -0.10145267099142075, 0.04092054069042206, 0.08495280891656876, 0.09764710813760757, 0.1466628611087799, 0.209085613489151, 0.20400598645210266, -0.13698942959308624, 0.24193733930587769, -0.07487781345844269, -0.13943180441856384, -0.28207170963287354, 0.11750312149524689, 0.1690332144498825, 0.09013193100690842, 0.030282072722911835, 0.16643111407756805, 0.041157566010951996, -0.37090572714805603, -0.0729445219039917, -0.05760568007826805, -0.12956909835338593, -0.16902801394462585, 0.13554269075393677, -0.1693432331085205, -0.04853004589676857, 0.22995777428150177, 0.007201503962278366, -0.013600717298686504, -0.40249717235565186, 0.053143538534641266, 0.3580959439277649, 0.1434372365474701, -0.07688236981630325, 0.06701886653900146, 0.2265019416809082, 0.2597028613090515, -0.1298811286687851, 0.1573919802904129, -0.2579093873500824, 0.36999547481536865, 0.09451119601726532, -0.3097192943096161, 0.16183717548847198, 0.45948630571365356, -0.02836923860013485, -0.1124441921710968, -0.1568191647529602, -0.10699570178985596, 0.04783381521701813, 0.0735948458313942, -0.24181495606899261, 0.07018890976905823, 0.13543979823589325, -0.001816488802433014, 0.5354022979736328, 0.02570386789739132, 0.17436665296554565, -0.2132638692855835, -0.04778806120157242, -0.44986969232559204, -0.20575019717216492, -0.08338751643896103, 0.028016671538352966, 0.08583544939756393, 0.08767642825841904, -0.004705161787569523, 0.05368794873356819, -0.03178022801876068, 0.27985110878944397, -0.014120722189545631, 0.0002252983395010233, -0.2494332194328308, 0.4703889787197113, -0.37131360173225403, -0.27084335684776306, -0.049185097217559814, -0.15531057119369507, 0.11308008432388306, -0.1719784140586853, -0.10465776920318604, -0.176669180393219, 0.2870733141899109, -0.07113653421401978, 0.05178259685635567, -0.18545502424240112, 0.3238450884819031, 0.3122841715812683, -0.00006301887333393097, -0.33270472288131714, 0.06873547285795212, 0.1956883668899536, 0.179451584815979, -0.13641557097434998, 0.26792481541633606, -0.07479532063007355, 0.023500200361013412, -0.18617749214172363, -0.01118343323469162, -0.19011551141738892, -0.06747450679540634, 0.03285990282893181, -0.3748263120651245 ]
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
[ -0.02323286421597004, -0.186763197183609, -0.17846335470676422, -0.3313103914260864, -0.5435771942138672, 0.06571724265813828, -0.17831747233867645, 0.12594062089920044, -0.3440611958503723, 0.46897250413894653, 0.23541782796382904, 0.1240529790520668, 0.07490070164203644, 0.4212980270385742, 0.01564176194369793, 0.08941159397363663, 0.11703037470579147, 0.40649154782295227, 0.1564115434885025, -0.07623115926980972, -0.046965986490249634, 0.3286643326282501, -0.26885986328125, -0.012009188532829285, 0.16894929111003876, 0.11247776448726654, -0.3255740702152252, -0.2680037021636963, -0.14709439873695374, -0.16988055408000946, 0.43014833331108093, 0.2873970866203308, 0.35420387983322144, -0.008296296000480652, -0.00011804488894995302, -0.24944938719272614, 0.26780998706817627, -0.0638856366276741, -0.22480155527591705, -0.2884771227836609, -0.3987266421318054, -0.3756754398345947, -0.1216316670179367, -0.01598048210144043, 0.13137945532798767, 0.05806165933609009, 0.2008039802312851, 0.015675485134124756, -0.12592141330242157, 0.10718555003404617, 0.1894945651292801, 0.1807601898908615, 0.14113670587539673, -0.3795875906944275, 0.1705751121044159, 0.3191806375980377, -0.2734740078449249, -0.22853103280067444, 0.26993754506111145, 0.025856537744402885, 0.3077666461467743, 0.4816244840621948, 0.28743380308151245, -0.3409464359283447, -0.1720893383026123, -0.013270949013531208, 0.116734579205513, -0.1250803917646408, 0.27835139632225037, -0.04526615887880325, 0.5306007862091064, 0.17983481287956238, -0.38571879267692566, 0.41310515999794006, 0.04533282667398453, -0.4731230139732361, 0.2673749029636383, -0.013816200196743011, 0.06820408999919891, 0.11895711719989777, 0.23744802176952362, -0.3053065538406372, -0.43842706084251404, 0.16605892777442932, -0.16326305270195007, 0.005223505198955536, -0.11229757964611053, -0.18681848049163818, 0.06563088297843933, 0.04510565102100372, -0.048754796385765076, 0.18406324088573456, -0.25097814202308655, 0.260393887758255, -0.22966523468494415, -0.3830537497997284, 0.28179800510406494, 0.021808292716741562, 0.32971617579460144, -0.03462202101945877, 0.15862037241458893, -0.21944329142570496, -0.24751384556293488, 0.0007928647100925446, 0.12062594294548035, 0.32286378741264343, 0.35899585485458374, 0.052947867661714554, 0.14591898024082184, 0.2353777140378952, 0.44060006737709045, -0.04040244221687317, 0.0712786614894867, 0.012920827604830265, -0.16717565059661865, 0.056895043700933456, 0.06021244078874588, -0.6049146056175232, -0.19651488959789276, -0.004578499123454094, -0.23373374342918396, 0.06281173974275589, -0.14380230009555817, 0.03192576393485069, 0.41804495453834534, 0.21806281805038452, 0.2200229912996292, -0.19355091452598572, -0.009806827642023563, -0.6159113645553589, -0.2765553593635559, -0.014677830040454865, -0.06726313382387161, 0.3063664436340332, 0.2101641148328781, -0.3618153929710388, 0.2205507904291153, -0.3328656852245331, 0.22424012422561646, 0.055308446288108826, -0.15858860313892365, 0.4364677369594574, -0.22660881280899048, 0.18580719828605652, -0.02298659458756447, 0.060771942138671875, -0.28177493810653687, 0.4902442693710327, -0.21706965565681458, 0.18422415852546692, -0.38244685530662537, -0.4308209717273712, -0.5014219284057617, 0.14038515090942383, 0.05071000009775162, -0.19810694456100464, 0.036736004054546356, 0.24654097855091095, -0.2765113115310669, -0.1712322235107422, 0.02272365242242813, 0.2964550852775574, -0.17472504079341888, -0.13263116776943207, -0.00553295761346817, 0.17904698848724365, -0.2631228268146515, -0.062196988612413406, -0.39297613501548767, 0.07705947011709213, -0.11440915614366531, 0.05869714170694351, -0.06112319231033325, -0.3164398968219757, -0.14322927594184875, -0.25975000858306885, 0.3386319577693939, -0.6128677725791931, -0.5238038897514343, 0.14918261766433716, -0.031318195164203644, -0.2901611328125, 0.2189878523349762, 0.005281124729663134, 0.00720395939424634, -0.16219674050807953, -0.20951777696609497, -0.0362691693007946, -0.2873406410217285, -0.33313676714897156, 0.021973613649606705, -0.4311455488204956, -0.07969623804092407, 0.10742993652820587, 0.09247314929962158, 0.09353873133659363, 0.23239237070083618, -0.6215121150016785, 0.1444459855556488, -0.01956053450703621, 0.010354232043027878, 0.06936415284872055, 0.6898937821388245, 0.07786283642053604, 0.2611396312713623, 0.05161260813474655, -0.3183964788913727, 0.056200072169303894, -0.21524792909622192, 0.06954513490200043, 0.26794856786727905, -0.37729018926620483, -0.11547413468360901, 0.0034696999937295914, 0.1692008376121521, -0.28243619203567505, 0.0478886142373085, 0.04885255545377731, 0.10802213847637177, 0.0012774355709552765, 0.013500649482011795, 0.33671051263809204, -0.15173646807670593, -0.17996656894683838, -0.10303992033004761, 0.17719830572605133, -0.1950802206993103, 0.0085162203758955, 0.016285471618175507, 0.22205933928489685, 0.10666133463382721, 0.019013242796063423, -0.08943860977888107, 0.1999126672744751, -0.16442206501960754, 0.25171518325805664, 0.25995245575904846, 0.4301052987575531, 0.22773393988609314, -0.22602127492427826, 0.0658654049038887, -0.09690937399864197, -0.19786733388900757, 0.35074323415756226, -0.010787330567836761, 0.2520740032196045, 0.14984852075576782, -0.07521001249551773, -0.007431790232658386, 0.20284217596054077, 0.03159359097480774, 0.041656870394945145, -0.20685186982154846, -0.11898766458034515, 0.2147410809993744, 0.10000599920749664, -0.34825098514556885, -0.3811192810535431, -0.20631006360054016, 0.16081619262695312, 0.13865166902542114, -0.11609688401222229, -0.000791594386100769, 0.11394792050123215, -0.020683571696281433, -0.0809202790260315, 0.34602686762809753, -0.10363257676362991, 0.0956856980919838, 0.2869535982608795, -0.24085912108421326, -0.1860807090997696, -0.009621341712772846, 0.16023385524749756, 0.15206804871559143, 0.27990031242370605, -0.24120919406414032, -0.17410489916801453, 0.37920939922332764, 0.17880699038505554, -0.12696203589439392, -0.23240919411182404, -0.014871317893266678, 0.05603492259979248, 0.028574872761964798, -0.09625128656625748, -0.13883979618549347, 0.005258549004793167, -0.06992299109697342, -0.05941121652722359, 0.18317663669586182, -0.19776906073093414, 0.2772122323513031, 0.1787654161453247, 0.07255779206752777, 0.2747689485549927, -0.0002000369131565094, 0.7763862013816833, -0.22260740399360657, 0.3033178150653839, -0.07584991306066513, 0.13949936628341675, -0.4241236746311188, 0.13569271564483643, -0.2917243540287018, -0.17309853434562683, 0.41069963574409485, -0.09878095984458923, 0.3334307074546814, -0.08852609992027283, -0.8045454621315002, -0.0008278489112854004, -0.10774485766887665, -0.10973872244358063, 0.11922863125801086, -0.31333667039871216, -0.09177760779857635, 0.024535104632377625, -0.02578258328139782, -0.2734338939189911, -0.48534756898880005, -0.16465319693088531, -0.39834660291671753, 0.08785322308540344, -0.17175862193107605, -0.38223913311958313, -0.04387309402227402, -0.07672463357448578, -0.12428030371665955, -0.12437795102596283, 0.2652050256729126, 0.1253896802663803, -0.00854033138602972, -0.07314734160900116, 0.17700539529323578, -0.04666546732187271, -0.1972220242023468, 0.20906862616539001, 0.33739566802978516, -0.3322140574455261, -0.3296048939228058, 0.3415127098560333, -0.02170230634510517, 0.2813384532928467, 0.01147801335901022, -0.4279921054840088, -0.15206898748874664, 0.06688883900642395, 0.13095764815807343, 0.051632437855005264, -0.03655090928077698, 0.7915472984313965, -0.11600484699010849, -0.07611781358718872, -0.02755453810095787, -0.5734846591949463, 0.10225138068199158, 0.5704912543296814, 0.24227428436279297, -0.11164764314889908, 0.317426860332489, 0.15854166448116302, 0.5127480030059814, 0.35929909348487854, 0.010425470769405365, 0.02273433655500412, 0.1787559986114502, 0.041327670216560364, -0.12431816011667252, -0.2332146018743515, 0.030882054939866066, 0.08355062454938889, 0.0810873880982399, 0.3579851984977722, 0.17127439379692078, -0.2575511932373047, -0.2342398315668106, -0.43134695291519165, -0.048631127923727036, 0.1353089064359665, -0.34676092863082886, 0.0899704247713089, 0.666692852973938, -0.057271551340818405, -0.1794433891773224, -0.4841732084751129, -0.05204256996512413, 0.19989363849163055, 0.44825083017349243, 0.31226590275764465, -0.002331764902919531, 0.159017875790596, -0.06324783712625504, -0.2990369498729706, 0.3822367787361145, 0.09644601494073868, 0.0762844979763031, -0.02362120896577835, 0.0570530891418457, 0.1606551706790924, 0.0039883749559521675, 0.207462877035141, -0.11883360892534256, -0.20663686096668243, -0.08955511450767517, -0.04198877513408661, 0.32143867015838623, -0.03235849738121033, 0.03982267901301384, -0.2484743446111679, -0.10593543946743011, -0.18759527802467346, 0.1950107216835022, -0.03170004114508629, 0.661746621131897, -0.27993783354759216, -0.1786489635705948, -0.11406160891056061, -0.05655476450920105, -0.07455011457204819, 0.04568089544773102, -0.05885583907365799, -0.261006236076355, -0.01762521266937256, -0.03928036987781525, -0.17231272161006927, 0.21923214197158813, -0.20522719621658325, 0.00713571161031723, 0.048595529049634933, -0.06891921907663345, -0.01868925616145134, 0.08036014437675476, 0.10861791670322418, -0.028513826429843903, 0.23418131470680237, 0.003001254051923752, 0.39899730682373047, -0.31781503558158875, 0.17326241731643677, -0.04418456554412842, 0.22356343269348145, 0.15306678414344788, -0.031643930822610855, 0.0737861692905426, -0.17802104353904724, -0.3110802173614502, -0.2239454686641693, -0.4165325164794922, 0.3540852665901184, 0.09462454915046692, -0.2084658443927765, 0.08699817955493927, 0.2448115050792694, 0.09113173186779022, -0.2714416980743408, 0.33740535378456116, 0.39424216747283936, -0.013790847733616829, 0.17862308025360107, -0.040981538593769073, 0.7828719615936279, 0.3878049850463867, -0.11633850634098053, 0.08909569680690765, -0.1653100848197937, 0.30338600277900696, 0.027321740984916687, 0.10152038931846619, 0.1656833291053772, -0.1250874102115631, -0.13380074501037598, 0.1904686540365219, 0.3938691318035126, -0.19231148064136505, -0.2761598825454712, 0.3090684413909912, 0.2963230013847351, 0.37913772463798523, 0.18517258763313293, 0.04660651832818985, 0.308780699968338, -0.012174827978014946, 0.22883759438991547, 0.0856974646449089, 0.1700183004140854, 0.44463080167770386, -0.015017283149063587, -0.1496933251619339, -0.07364707440137863, -0.24480487406253815, -0.31907719373703003, -0.24239102005958557, 0.10532356798648834, -0.12534353137016296, 0.06371273100376129, -0.13501372933387756, 0.47955790162086487, -0.18675996363162994, 0.4332156777381897, 0.019766826182603836, -0.3183548152446747, 0.11428603529930115, 0.2496323585510254, -0.24592366814613342, 0.05072131007909775, -0.03206104040145874, -0.2735680043697357, -0.19744345545768738, -0.12815316021442413, 0.4675574004650116, 0.26430127024650574, -0.18208615481853485, -0.29033780097961426, -0.17927472293376923, -0.009771496057510376, 0.16735342144966125, 0.3781090974807739, 0.4931943714618683, 0.21740108728408813, -0.23251913487911224, 0.13885162770748138, 0.14542095363140106, -0.12462020665407181, -0.33353573083877563, -0.10845135152339935, 0.12606149911880493, -0.15928465127944946, 0.20721027255058289, 0.42981356382369995, -0.3224424123764038, 0.07744616270065308, -0.12384441494941711, 0.035477541387081146, -0.13220752775669098, -0.2717389464378357, 0.030616700649261475, -0.4341999888420105, -0.1563023030757904, 0.03174485266208649, -0.17103588581085205, 0.19185855984687805, 0.9666599631309509, 0.061495013535022736, 0.20077848434448242, -0.18781906366348267, -0.015118543058633804, -0.2095966637134552, 0.1748201549053192, -0.5250986814498901, 0.21809221804141998, -0.17597243189811707, -0.07979463040828705, -0.16315673291683197, 0.26689305901527405, -0.2849743366241455, -0.228902205824852, 0.17995263636112213, -0.15777570009231567, 0.33274638652801514, 0.28982308506965637, 0.17223277688026428, 0.11206085979938507, 0.043336525559425354, -0.13135674595832825, -0.1408834010362625, -0.047134608030319214, 0.10009497404098511, 0.12241396307945251, 0.08467115461826324, 0.33936193585395813, -0.037495218217372894, -0.1875881552696228, -0.19110731780529022, 0.19971676170825958, 0.24985697865486145, -0.09309854358434677, 0.0867888331413269, 0.24760381877422333, -0.2960225045681, -0.1283235251903534, -0.08523381501436234, 0.5458859205245972, 0.24163523316383362, 0.13734978437423706, 0.15754488110542297, 0.37560027837753296, -0.09079910069704056, -0.03890293836593628, -0.07435857504606247, 0.4465688467025757, -0.3893502950668335, 0.2567908465862274, 0.03685351461172104, 0.09243886917829514, -0.13186967372894287, 0.038229480385780334, -0.26250752806663513, -0.5318465828895569, -0.012972288765013218, 0.19005686044692993, -0.057240135967731476, 0.20679312944412231, -0.36436301469802856, 0.022032814100384712, -0.2610281705856323, -0.09930095076560974, -0.0296506155282259, 0.21123558282852173, 0.37755751609802246, 0.0707448273897171, 0.8590601682662964, 0.1181165799498558, -0.11029167473316193, -0.07216499000787735, 0.4866432845592499, 0.22771112620830536, 0.3900809586048126, 0.13891121745109558, 0.09025733917951584, 0.0017621535807847977, 0.10719466954469681, -0.16587089002132416, -0.07671058923006058, -0.13972161710262299, 0.4415704011917114, 0.45498254895210266, 0.3275451958179474, -0.20021550357341766, -0.20803408324718475, 0.24367785453796387, 0.18482714891433716, 0.014333982020616531, 0.23309193551540375, -0.08115050941705704, 0.1278199702501297, 0.094309963285923, -0.7621111273765564, -0.09891968965530396, 0.07419556379318237, -0.16120028495788574, -0.1580151915550232, -0.12280513346195221, -0.26581406593322754, -0.11310794949531555, 0.36796367168426514, 0.0979597419500351, 0.22673699259757996, -0.27596282958984375, -0.14950010180473328, -0.05910411477088928, -0.20407924056053162, 0.2876177430152893, 0.2594887912273407, 0.06134823337197304, 0.48437702655792236, -0.564453661441803, -0.3007173240184784, 0.4297741949558258, 0.21083779633045197, 0.01913793385028839, -0.2560524642467499, 0.2916184365749359, 0.055539846420288086, -0.02176825702190399, -0.26670563220977783, 0.23909780383110046, 0.18271811306476593, -0.14026705920696259, 0.11083655059337616, -0.014656071551144123, 0.062215305864810944, 0.04342533275485039, -0.05880115553736687, 0.012008767575025558, -0.08062687516212463, -0.4757099151611328, 0.13423489034175873, -0.20503276586532593, 0.16521883010864258, -0.12193162739276886, 0.44938090443611145, 0.08370696753263474, -0.38023364543914795, -0.2866361141204834, -0.044458650052547455, -0.2227209210395813, 0.014352894388139248, 0.055022262036800385, -0.4033523499965668, -0.16033096611499786, 0.3848738968372345, 0.19997671246528625, -0.12101326137781143, -0.6013121604919434, 0.06163422018289566, 0.25735101103782654, 0.1335514634847641, -0.13102898001670837, 0.11507102102041245, 0.2602618336677551, 0.14775578677654266, -0.2571094036102295, 0.2359616607427597, -0.014066595584154129, 0.2910211682319641, 0.09884433448314667, -0.16290105879306793, 0.08296118676662445, 0.3361518383026123, 0.003927494399249554, -0.24153657257556915, -0.013702761381864548, -0.12546321749687195, 0.2838474214076996, -0.12509283423423767, -0.148213192820549, -0.029125213623046875, 0.0956340879201889, 0.12166111171245575, 0.46083852648735046, 0.0902075469493866, 0.021488837897777557, -0.06572086364030838, -0.14753443002700806, -0.40121519565582275, -0.060062553733587265, 0.03288842365145683, -0.1326179802417755, 0.014206446707248688, 0.05187322199344635, -0.135909304022789, 0.3132942318916321, -0.009249239228665829, 0.4369445741176605, 0.1265651136636734, -0.024377649649977684, -0.14826227724552155, 0.4578133225440979, -0.46626439690589905, -0.16879814863204956, 0.2816956639289856, -0.1104644164443016, 0.0842432975769043, -0.11323767900466919, -0.10509461909532547, -0.0178554505109787, 0.17952176928520203, -0.15866148471832275, 0.2223753184080124, -0.20011001825332642, 0.40533655881881714, 0.2907789945602417, -0.031212368980050087, -0.3527666926383972, -0.07439643144607544, 0.18313199281692505, 0.27748650312423706, -0.017893413081765175, 0.38302019238471985, 0.1775818169116974, 0.05988417938351631, -0.12405893206596375, 0.2663041651248932, -0.18348199129104614, -0.0043844785541296005, 0.04803791269659996, -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.
[ -0.24925146996974945, -0.10364571958780289, -0.13912147283554077, -0.42108476161956787, -0.5307990312576294, -0.0687481164932251, -0.07875014841556549, 0.07906541228294373, -0.21976329386234283, 0.4390559196472168, 0.1513829380273819, 0.29369914531707764, 0.06534174829721451, 0.43753260374069214, 0.006778089329600334, 0.2519569396972656, 0.15442757308483124, 0.3981415927410126, -0.035902880132198334, -0.08111092448234558, -0.098997563123703, 0.30109986662864685, -0.22698141634464264, -0.05949544906616211, 0.06923174858093262, -0.020368127152323723, -0.27286192774772644, -0.1656876504421234, -0.15396252274513245, -0.15395793318748474, 0.41139307618141174, 0.3626805543899536, 0.20779532194137573, 0.004311591386795044, -0.0001105137198464945, -0.21097853779792786, 0.28249117732048035, -0.1277121603488922, -0.23003864288330078, -0.12380580604076385, -0.33321240544319153, -0.32515519857406616, 0.008640028536319733, 0.0028272345662117004, 0.0867995023727417, 0.1362747699022293, 0.12899629771709442, -0.10482463240623474, 0.0015904083847999573, 0.17070774734020233, 0.26446282863616943, 0.2665337324142456, -0.0055653005838394165, -0.39218854904174805, 0.16724231839179993, 0.27682894468307495, -0.3164585828781128, -0.14915382862091064, 0.3068573474884033, -0.017131198197603226, 0.08296903967857361, 0.5250522494316101, 0.26207688450813293, -0.17759650945663452, -0.1293550431728363, -0.053835153579711914, 0.12355461716651917, -0.11475835740566254, 0.32602939009666443, 0.05549313873052597, 0.5153299570083618, 0.14862272143363953, -0.4204435646533966, 0.3833313286304474, -0.0019985903054475784, -0.389568567276001, 0.257610023021698, -0.05857372283935547, 0.006152037531137466, 0.05968878045678139, 0.14559219777584076, -0.15127874910831451, -0.3991320729255676, 0.1424819976091385, -0.1709086298942566, 0.031247396022081375, -0.13222801685333252, -0.21965506672859192, 0.08026755601167679, 0.09614323079586029, -0.10330730676651001, 0.20984381437301636, -0.2582564651966095, 0.18841113150119781, -0.2771468758583069, -0.28245383501052856, 0.3214116394519806, 0.0659700259566307, 0.4131602942943573, -0.06235169619321823, 0.151973158121109, -0.0995105728507042, -0.11116892844438553, 0.14640206098556519, 0.05179961770772934, 0.25040289759635925, 0.23542797565460205, -0.053371794521808624, 0.32314854860305786, 0.2254091054201126, 0.48004674911499023, -0.03999726474285126, 0.16129684448242188, 0.0416819304227829, -0.21857786178588867, 0.012427378445863724, -0.026502275839447975, -0.6751197576522827, -0.2794654965400696, 0.002901901490986347, -0.07580031454563141, 0.05903176963329315, -0.07884785532951355, 0.14038969576358795, 0.38729918003082275, 0.1108703687787056, 0.2383928745985031, -0.03799448907375336, -0.06861913204193115, -0.5039106607437134, -0.3114447295665741, 0.049900367856025696, -0.08856596797704697, 0.2114315629005432, 0.29866084456443787, -0.2297256886959076, 0.3060239255428314, -0.3308388292789459, 0.3673429489135742, 0.11446791887283325, -0.20475706458091736, 0.3666544258594513, -0.19268497824668884, 0.2855261564254761, 0.0603473000228405, 0.042868368327617645, -0.22038297355175018, 0.3571574091911316, -0.22282853722572327, 0.09572034329175949, -0.2677157521247864, -0.42350250482559204, -0.47084885835647583, 0.19991570711135864, 0.05588305741548538, -0.22937512397766113, 0.049219515174627304, 0.32721760869026184, -0.20584681630134583, -0.17716988921165466, -0.02569682151079178, 0.33732426166534424, -0.10757425427436829, -0.12294257432222366, 0.11131118983030319, 0.18788516521453857, -0.2905878722667694, -0.009798012673854828, -0.3686021864414215, 0.044304005801677704, -0.04414796084165573, 0.04356708377599716, -0.10054714977741241, -0.22311890125274658, -0.11662528663873672, -0.23496821522712708, 0.3060913681983948, -0.3721950054168701, -0.41131892800331116, 0.13090969622135162, 0.058612048625946045, -0.34151044487953186, 0.128153458237648, -0.153347447514534, 0.019392352551221848, -0.19477085769176483, -0.17797908186912537, 0.1028158888220787, -0.18041346967220306, -0.16533173620700836, -0.023646719753742218, -0.4142703711986542, -0.12499180436134338, 0.2005959451198578, 0.12091828882694244, 0.06461642682552338, 0.27017322182655334, -0.6116054058074951, 0.05576079338788986, -0.1391737461090088, 0.11108908802270889, 0.08375389873981476, 0.7464858889579773, -0.04615722596645355, 0.1895795315504074, 0.14584100246429443, -0.3134401738643646, 0.02355070412158966, -0.26641911268234253, 0.14026635885238647, 0.11664465069770813, -0.34082651138305664, -0.0927940085530281, 0.002391614019870758, 0.06440754979848862, -0.33980435132980347, 0.13861142098903656, 0.09465945512056351, 0.0508364774286747, 0.09796687960624695, 0.1258644461631775, 0.3094979524612427, -0.19311818480491638, -0.22526568174362183, 0.001833360642194748, 0.1426396667957306, -0.2699939012527466, -0.10479746758937836, 0.14433252811431885, 0.2866573929786682, 0.025029432028532028, -0.04019121825695038, -0.1668752133846283, 0.26575028896331787, -0.16327416896820068, 0.27370142936706543, 0.09590202569961548, 0.38307997584342957, 0.14369438588619232, -0.14276359975337982, 0.07344484329223633, -0.1170787364244461, -0.1200573518872261, 0.3190228044986725, -0.11595211178064346, 0.34866467118263245, 0.08622618764638901, -0.08114559203386307, -0.035293981432914734, 0.18887203931808472, 0.18004760146141052, -0.03710181266069412, -0.21818643808364868, -0.10062845051288605, 0.19473564624786377, 0.09217852354049683, -0.3050864040851593, -0.3667620122432709, -0.24421106278896332, 0.05284672975540161, 0.18153175711631775, -0.008825916796922684, 0.09464175999164581, 0.16038444638252258, -0.038517698645591736, -0.16388607025146484, 0.4497125446796417, -0.09222950041294098, 0.10848797857761383, 0.3049633204936981, -0.20059725642204285, -0.15271013975143433, -0.14874960482120514, 0.047722503542900085, 0.12991148233413696, 0.2641792893409729, -0.0851147249341011, -0.04902441054582596, 0.33934900164604187, 0.14000369608402252, -0.3533046245574951, -0.22798293828964233, -0.05839638411998749, 0.02828158251941204, 0.0008941106498241425, -0.0952368900179863, -0.13379213213920593, 0.0065200552344322205, 0.030234232544898987, 0.0011708508245646954, 0.13668008148670197, -0.23664966225624084, 0.36741703748703003, 0.1476510912179947, -0.11774073541164398, 0.3822777271270752, -0.007208691909909248, 0.7500996589660645, -0.11174765974283218, 0.3082488477230072, -0.18575958907604218, 0.12192968279123306, -0.4235031306743622, 0.2297055423259735, -0.2892768681049347, -0.1285896599292755, 0.4955853521823883, -0.09794572740793228, 0.47120773792266846, -0.10985422879457474, -0.7647414207458496, 0.06731861084699631, -0.12290874123573303, -0.015936559066176414, 0.014705538749694824, -0.2569577693939209, -0.1591472029685974, -0.01935558393597603, -0.04918188229203224, -0.33155807852745056, -0.43553027510643005, -0.05172509700059891, -0.48017874360084534, -0.005570529028773308, -0.3519701659679413, -0.4595109522342682, -0.021146250888705254, -0.21627861261367798, -0.1986929327249527, -0.21795067191123962, 0.17308765649795532, 0.25137925148010254, 0.025441572070121765, 0.05033515766263008, -0.015435701236128807, 0.007242961786687374, -0.2543773651123047, 0.1297796219587326, 0.49869659543037415, -0.3865443766117096, -0.5030509829521179, 0.29115578532218933, -0.04333081841468811, 0.3502500355243683, -0.006292922422289848, -0.3598858118057251, -0.2641659080982208, 0.000894526019692421, 0.14793285727500916, -0.009467877447605133, 0.03826868161559105, 0.7621444463729858, -0.11581352353096008, -0.19106866419315338, -0.11626658588647842, -0.5392231941223145, 0.19554352760314941, 0.4982726275920868, 0.24280890822410583, -0.27890506386756897, 0.3810690641403198, 0.20991508662700653, 0.31938087940216064, 0.351168692111969, -0.056519005447626114, 0.12163647264242172, 0.16339358687400818, 0.14771616458892822, -0.19633936882019043, -0.016693828627467155, 0.14289188385009766, 0.07544223964214325, 0.14860877394676208, 0.5334334373474121, 0.12728244066238403, -0.3797418475151062, -0.25314009189605713, -0.27874472737312317, -0.05657918006181717, 0.052457865327596664, -0.2920558452606201, 0.015738388523459435, 0.5925155282020569, -0.02757169120013714, -0.16437718272209167, -0.29470545053482056, -0.024257853627204895, 0.18267162144184113, 0.4032709002494812, 0.22562989592552185, -0.011568616144359112, 0.08430549502372742, -0.07599983364343643, -0.36049914360046387, 0.38195037841796875, 0.05675133317708969, 0.10147261619567871, -0.12819619476795197, 0.061539679765701294, 0.1697923243045807, 0.010395808145403862, 0.23315387964248657, -0.19572284817695618, -0.270768940448761, -0.09753899276256561, -0.03323427587747574, 0.20889689028263092, -0.07565906643867493, -0.060405682772397995, -0.17605644464492798, -0.05845494195818901, -0.17866334319114685, 0.0345807746052742, -0.09470606595277786, 0.5656789541244507, -0.28112345933914185, -0.04575753211975098, -0.0969703420996666, -0.017560925334692, -0.2522439956665039, -0.0532970055937767, 0.018999725580215454, -0.1461751013994217, -0.029696844518184662, -0.06324397027492523, -0.2030780166387558, 0.3560207188129425, -0.13447648286819458, 0.005193449556827545, 0.20223788917064667, 0.013026426546275616, 0.053487882018089294, 0.06713122129440308, 0.03821226954460144, 0.03194700554013252, 0.20771418511867523, 0.02288118191063404, 0.3638218939304352, -0.14250923693180084, 0.26192307472229004, -0.2598850131034851, 0.25790271162986755, 0.17956404387950897, -0.053154557943344116, 0.04724732041358948, -0.2090412825345993, -0.2319692224264145, -0.11197467148303986, -0.2654051184654236, 0.28373512625694275, 0.03603268787264824, -0.14099259674549103, 0.12018948793411255, 0.24428707361221313, 0.026496976613998413, -0.2426859587430954, 0.23383943736553192, 0.3414636254310608, -0.10660884529352188, 0.3187498450279236, 0.014410741627216339, 0.7144322991371155, 0.2847661077976227, -0.1513303518295288, 0.1777021288871765, -0.09784714877605438, 0.2846790850162506, -0.08520574867725372, 0.03084205836057663, 0.14362411201000214, -0.029414750635623932, -0.10803273320198059, 0.22408488392829895, 0.47330978512763977, -0.1807129979133606, -0.4074785113334656, 0.23934654891490936, 0.37372463941574097, 0.25265413522720337, 0.17917318642139435, 0.02582484483718872, 0.2792445719242096, -0.1941184550523758, 0.22286474704742432, 0.16126205027103424, 0.11865907907485962, 0.24784404039382935, -0.007930041290819645, -0.12176413834095001, -0.10707306861877441, -0.33475154638290405, -0.2222701907157898, -0.15550276637077332, 0.17588695883750916, -0.1371014267206192, 0.12958724796772003, -0.13968020677566528, 0.5002411603927612, -0.21630284190177917, 0.3512284755706787, 0.07694332301616669, -0.393901526927948, 0.19849856197834015, 0.12374664098024368, -0.09371845424175262, -0.04847056418657303, 0.011539828963577747, -0.09307029098272324, -0.31500595808029175, -0.040459439158439636, 0.27459079027175903, 0.13504843413829803, -0.2294904589653015, -0.150488018989563, -0.15534141659736633, -0.05174735188484192, 0.19557903707027435, 0.3932056725025177, 0.4967649281024933, 0.20558060705661774, -0.2776392698287964, 0.21186195313930511, 0.19321778416633606, -0.09961581975221634, -0.09029289335012436, 0.001679297536611557, 0.052381500601768494, -0.209777370095253, 0.20240315794944763, 0.3447622060775757, -0.2052677422761917, 0.11227018386125565, -0.06963338702917099, -0.0756426453590393, -0.23330840468406677, -0.3203873336315155, -0.013944464735686779, -0.5064802765846252, -0.21499544382095337, 0.04626106098294258, -0.2646346688270569, 0.16258925199508667, 0.9665590524673462, 0.01933588832616806, 0.08566275238990784, -0.15413010120391846, -0.12778069078922272, -0.27205634117126465, 0.09145689755678177, -0.6348620653152466, 0.3025992214679718, -0.09700276702642441, -0.010685129091143608, -0.05887637287378311, 0.2275681495666504, -0.40056437253952026, -0.23105879127979279, 0.08580842614173889, -0.07548955082893372, 0.17271468043327332, 0.18505537509918213, 0.3002164363861084, 0.20005735754966736, 0.10802093148231506, -0.03642883151769638, -0.2978534996509552, -0.14092481136322021, 0.06706100702285767, 0.09022772312164307, 0.08517469465732574, 0.265106201171875, 0.015573612414300442, -0.09612858295440674, -0.3636767566204071, 0.14022549986839294, 0.31589752435684204, -0.25541505217552185, -0.08340774476528168, 0.21389128267765045, -0.2956829071044922, -0.15318617224693298, 0.0740920901298523, 0.268174946308136, 0.20325908064842224, 0.00018320977687835693, 0.022653240710496902, 0.40176713466644287, -0.07402580231428146, -0.12795908749103546, 0.06694352626800537, 0.39901527762413025, -0.3243759274482727, 0.23266419768333435, 0.05787014216184616, 0.021776720881462097, -0.07467720657587051, -0.027419010177254677, -0.11537861078977585, -0.5985841155052185, -0.05257910490036011, 0.08984836935997009, 0.027268800884485245, 0.2741132080554962, -0.4083036780357361, 0.10977049171924591, -0.23368915915489197, -0.06540708988904953, 0.17333264648914337, 0.1268070489168167, 0.538725733757019, -0.04353984817862511, 0.9180060625076294, 0.10789883881807327, -0.07989365607500076, -0.07315390557050705, 0.35224682092666626, 0.09844856709241867, 0.19314135611057281, 0.2026144564151764, 0.1697167009115219, 0.017392292618751526, 0.08312458544969559, -0.06407191604375839, 0.010316044092178345, -0.09518469870090485, 0.3211269974708557, 0.4013398587703705, 0.15263217687606812, -0.31373360753059387, 0.03715774789452553, 0.16140078008174896, 0.1964500993490219, 0.016423720866441727, 0.21400709450244904, -0.09558103233575821, 0.12567318975925446, 0.11404748260974884, -0.8520107865333557, -0.14799919724464417, -0.0002811327576637268, -0.15192794799804688, -0.15474072098731995, -0.12153647094964981, -0.26129478216171265, -0.10409194231033325, 0.2915908098220825, 0.1082492545247078, 0.29016727209091187, -0.29568856954574585, -0.08855624496936798, -0.07152862101793289, -0.08061780780553818, 0.14213575422763824, 0.14287446439266205, 0.08947965502738953, 0.5019301772117615, -0.4905027747154236, -0.25601014494895935, 0.30503740906715393, 0.3041679263114929, 0.16102735698223114, -0.20795701444149017, 0.2815849184989929, 0.18219362199306488, 0.019643504172563553, -0.26642152667045593, 0.16778211295604706, 0.26305466890335083, -0.0011916086077690125, 0.12151727080345154, 0.1015295535326004, 0.11109834909439087, -0.06895503401756287, 0.01900523342192173, -0.11042600870132446, -0.004637842997908592, -0.2969089448451996, 0.14319083094596863, -0.07266214489936829, 0.12516340613365173, -0.08303329348564148, 0.24224305152893066, 0.0606016144156456, -0.3382854759693146, -0.2355436086654663, -0.14563843607902527, -0.2185368537902832, -0.14484553039073944, 0.11242605745792389, -0.2741743326187134, 0.02028036117553711, 0.24042868614196777, 0.2540437579154968, -0.02682773396372795, -0.5908101201057434, 0.002972930669784546, 0.27418720722198486, 0.13339383900165558, 0.03358721733093262, 0.08341849595308304, 0.25750401616096497, 0.208232581615448, -0.2613700032234192, 0.29738879203796387, -0.1402307003736496, 0.3001266121864319, -0.008812136948108673, -0.35600730776786804, 0.06686928868293762, 0.26842188835144043, -0.008375920355319977, -0.18063223361968994, -0.04371894896030426, -0.11608526110649109, 0.2443636953830719, 0.0062119849026203156, -0.2181481271982193, 0.014315061271190643, 0.08671729266643524, 0.04670063406229019, 0.47451052069664, 0.2091076374053955, 0.1296912431716919, -0.1403009593486786, -0.13859014213085175, -0.41770774126052856, -0.09693159908056259, 0.04019985720515251, -0.1644831895828247, 0.07427644729614258, 0.22511287033557892, -0.0425955131649971, 0.13968177139759064, -0.0036886464804410934, 0.4698243737220764, 0.07916218787431717, -0.0848124772310257, -0.12159688770771027, 0.5496423840522766, -0.4163467586040497, -0.17564962804317474, 0.12344086170196533, -0.134454607963562, 0.16413360834121704, -0.1847333312034607, -0.1444331854581833, -0.10510910302400589, 0.27850863337516785, -0.22749774158000946, 0.09104298055171967, -0.0843522921204567, 0.31331199407577515, 0.11957132071256638, -0.010894568637013435, -0.33837202191352844, 0.04173070937395096, 0.11309703439474106, 0.23142507672309875, -0.08588773012161255, 0.44417470693588257, 0.10862861573696136, 0.011421862989664078, -0.14661401510238647, 0.3638548254966736, -0.15389755368232727, -0.040989309549331665, 0.04374947026371956, -0.3198181390762329 ]
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 !
[ -0.25247395038604736, -0.2647853493690491, -0.1142226904630661, -0.4864540696144104, -0.37947577238082886, -0.20802780985832214, 0.08028712868690491, 0.1892700344324112, 0.11779683828353882, 0.3005794882774353, -0.155775249004364, 0.5226978063583374, -0.1476927548646927, 0.4125770926475525, 0.1462078094482422, 0.10172688215970993, 0.03274606913328171, 0.36265653371810913, -0.14609387516975403, -0.20175470411777496, -0.24998591840267181, 0.22434023022651672, -0.17047974467277527, -0.0546533465385437, 0.2192804217338562, 0.07695111632347107, -0.29594311118125916, 0.0793561264872551, -0.06169799715280533, -0.19533966481685638, 0.32915693521499634, 0.329232782125473, 0.18306970596313477, 0.06467734277248383, -0.0001042149233398959, -0.23515938222408295, 0.257485032081604, -0.12841549515724182, -0.28690019249916077, -0.18150989711284637, -0.1894627809524536, -0.4737764894962311, 0.2212621569633484, -0.21766060590744019, 0.10919704288244247, 0.12459304183721542, 0.20315438508987427, -0.1243731826543808, 0.1270555704832077, 0.2281511127948761, 0.3232874870300293, 0.4082866907119751, -0.03426554799079895, -0.18497490882873535, 0.20779341459274292, 0.2706398069858551, -0.20863227546215057, -0.012949122115969658, 0.15916483104228973, -0.275456041097641, -0.1968260407447815, 0.40579548478126526, 0.14011336863040924, -0.09584838151931763, -0.1321592926979065, -0.06573273986577988, 0.16082273423671722, 0.021586338058114052, 0.010288204997777939, 0.03543519228696823, 0.23504433035850525, -0.049714528024196625, -0.3959871828556061, 0.14017510414123535, 0.028912823647260666, -0.5668572783470154, 0.09331750869750977, -0.08714375644922256, -0.12451334297657013, -0.14305618405342102, 0.207477867603302, -0.162346750497818, -0.4726060926914215, 0.2572448253631592, 0.00482989102602005, 0.29475849866867065, -0.06364817917346954, -0.2017003893852234, 0.22664369642734528, 0.04250162094831467, -0.07998193055391312, 0.15150271356105804, -0.032829832285642624, 0.3442155718803406, -0.532577395439148, -0.2802596092224121, 0.34057557582855225, 0.09036564081907272, 0.2678717374801636, 0.042602624744176865, 0.2928159832954407, 0.14166155457496643, -0.003866479266434908, 0.19317743182182312, 0.030217591673135757, 0.24012567102909088, 0.292665034532547, -0.04129859805107117, 0.42817866802215576, 0.2653530538082123, 0.3137470483779907, 0.05636422336101532, 0.0704406276345253, -0.10898222774267197, -0.14499956369400024, -0.06849263608455658, 0.008154410868883133, -0.4685887098312378, -0.2638154923915863, -0.05357527732849121, -0.055846989154815674, 0.14551283419132233, -0.013918506912887096, 0.2732974886894226, 0.1701270341873169, 0.24480241537094116, 0.43004345893859863, 0.009215237572789192, -0.22485925257205963, -0.3951990306377411, -0.29168108105659485, 0.09738127887248993, -0.3190457224845886, 0.08400389552116394, 0.373189777135849, -0.20060120522975922, 0.27423256635665894, -0.43340110778808594, 0.25433433055877686, 0.071140818297863, -0.1991911232471466, 0.23665927350521088, -0.11447153240442276, 0.06143748760223389, 0.2251134216785431, -0.04297841712832451, -0.23835982382297516, 0.20623402297496796, -0.20838817954063416, -0.08082080632448196, -0.3151925504207611, -0.382951557636261, -0.40688368678092957, 0.27562108635902405, -0.055168330669403076, -0.314896821975708, -0.13434967398643494, 0.3234364986419678, -0.17393061518669128, -0.1865476667881012, -0.03619306534528732, 0.13921815156936646, -0.25417569279670715, -0.12859782576560974, 0.09105675667524338, 0.11252589523792267, 0.054606158286333084, -0.1368643343448639, -0.5214046835899353, 0.0013879556208848953, -0.010001927614212036, 0.049277059733867645, 0.02189880609512329, -0.12459664791822433, -0.1309000849723816, -0.1309100091457367, 0.5791587233543396, -0.36789417266845703, -0.06391140818595886, 0.1252528727054596, 0.05375072360038757, -0.28811267018318176, 0.09406547248363495, -0.17142541706562042, -0.14280207455158234, -0.10669790208339691, -0.1012190580368042, 0.3418888449668884, -0.07049553096294403, 0.09957312792539597, -0.11652086675167084, -0.30355843901634216, -0.012261595577001572, 0.3013630211353302, 0.29123976826667786, -0.08109045773744583, 0.23300936818122864, -0.15536600351333618, 0.20717063546180725, -0.2782922685146332, 0.034141480922698975, 0.0545126348733902, 0.5859828591346741, 0.04629027843475342, 0.09907401353120804, -0.05174317955970764, -0.3171609342098236, 0.01667911931872368, -0.11427298933267593, 0.29519736766815186, 0.11283303797245026, -0.29259783029556274, -0.03223046660423279, 0.05407194048166275, 0.11803597211837769, -0.11149375140666962, 0.2614137828350067, 0.11125195026397705, 0.023204389959573746, 0.07920382916927338, 0.07408304512500763, 0.3704969882965088, -0.14236298203468323, -0.13941875100135803, -0.25196874141693115, 0.19473262131214142, -0.21140211820602417, -0.09095360338687897, 0.09032346308231354, 0.30762791633605957, -0.014147154986858368, -0.016491485759615898, -0.09004611521959305, 0.3420785963535309, -0.19772456586360931, 0.3473963141441345, 0.06613952666521072, 0.11991210281848907, 0.03165199235081673, -0.18702729046344757, 0.10462108999490738, 0.020478149875998497, 0.10501891374588013, 0.19834381341934204, -0.1874205470085144, 0.4752040505409241, 0.09016667306423187, -0.0594591423869133, 0.09390716254711151, 0.03295445442199707, 0.1509227156639099, -0.23204569518566132, -0.30256161093711853, 0.012772440910339355, 0.38707464933395386, 0.1158938854932785, -0.3680190443992615, -0.373274028301239, -0.37219515442848206, -0.06634137779474258, 0.1566101759672165, 0.005696229636669159, 0.21773910522460938, 0.1688700020313263, -0.022963881492614746, -0.2444078028202057, 0.2332589030265808, -0.0959550216794014, 0.3129541873931885, 0.33390384912490845, -0.0884549617767334, 0.03661441057920456, -0.34952226281166077, 0.038188837468624115, 0.1847696602344513, 0.21045924723148346, -0.003719043917953968, 0.1817546784877777, 0.4397560656070709, -0.03354679420590401, -0.4896514415740967, -0.41225647926330566, -0.2240450382232666, 0.048277124762535095, -0.24939961731433868, -0.012751483358442783, -0.06169266998767853, -0.14514479041099548, -0.055051274597644806, 0.09441997855901718, 0.0958615392446518, -0.36516016721725464, 0.28364133834838867, 0.3310319483280182, -0.25196075439453125, 0.19968518614768982, -0.011269407346844673, 0.6269261240959167, -0.040750518441200256, 0.12832379341125488, -0.2383573055267334, -0.05031697824597359, -0.2411792278289795, 0.26183512806892395, -0.24112819135189056, 0.031092854216694832, 0.4830784201622009, -0.18815001845359802, 0.26334232091903687, 0.051257871091365814, -0.6950429081916809, 0.019227750599384308, -0.12177248299121857, -0.06724271178245544, 0.10748785734176636, -0.31135642528533936, -0.10114424675703049, -0.0694996640086174, 0.04360108822584152, -0.6364909410476685, -0.34482109546661377, -0.013514799065887928, -0.23367322981357574, 0.16604368388652802, -0.4094182252883911, -0.43483468890190125, -0.104634590446949, -0.37184298038482666, -0.12796847522258759, 0.09901575744152069, 0.18223023414611816, 0.3901268243789673, -0.0058502196334302425, 0.17498679459095, 0.061040233820676804, 0.05882640928030014, -0.31342655420303345, 0.17934037744998932, 0.42067214846611023, -0.40306001901626587, -0.5246384143829346, 0.008748900145292282, -0.09246602654457092, 0.3225846290588379, 0.09169071912765503, -0.11638527363538742, -0.39315280318260193, 0.17101828753948212, 0.03368833288550377, -0.13248781859874725, 0.2708769142627716, 0.4044116139411926, -0.03216135874390602, -0.21041570603847504, -0.1765938401222229, -0.2458331286907196, 0.15085195004940033, 0.4518738090991974, 0.13134139776229858, -0.24895256757736206, 0.27934443950653076, 0.08622346818447113, 0.11332763731479645, 0.10454657673835754, -0.16755010187625885, 0.2823275327682495, -0.027697758749127388, 0.20288461446762085, -0.001618117094039917, -0.00043096207082271576, 0.07328081876039505, 0.07297275215387344, 0.2253543883562088, 0.5441485047340393, 0.14013338088989258, -0.4167070686817169, -0.35925528407096863, -0.30449584126472473, -0.03829622268676758, -0.08819417655467987, -0.057566821575164795, 0.2055095136165619, 0.3595203757286072, -0.028068874031305313, -0.19928357005119324, -0.41280996799468994, -0.20246300101280212, -0.012766246683895588, 0.2373708039522171, 0.004086844623088837, -0.06508514285087585, 0.09634464979171753, 0.038656361401081085, -0.4979262948036194, 0.38196176290512085, 0.11513083428144455, 0.0731695294380188, 0.03802059218287468, -0.08453947305679321, 0.0638473629951477, -0.10432162880897522, 0.18705996870994568, -0.11142290383577347, 0.0054567791521549225, -0.05644497647881508, -0.10295884311199188, -0.1770627200603485, 0.05641874298453331, -0.378465473651886, -0.19693318009376526, 0.07580872625112534, -0.009619325399398804, -0.0009151995182037354, -0.17845577001571655, 0.660025954246521, -0.08007253706455231, -0.05716114491224289, -0.19711533188819885, -0.012537848204374313, -0.4126783013343811, -0.0750909373164177, -0.003672763705253601, -0.1563308835029602, 0.2179161012172699, -0.12022322416305542, -0.08209224045276642, 0.17336639761924744, -0.017646897584199905, 0.11866284906864166, 0.28284525871276855, -0.07113154232501984, 0.1216844990849495, -0.01521703228354454, -0.025802522897720337, 0.07832600176334381, 0.1132291704416275, 0.11903879046440125, 0.18738582730293274, -0.06267866492271423, 0.05007655546069145, -0.05787062644958496, 0.29155233502388, 0.18741901218891144, 0.09294836223125458, 0.022943638265132904, -0.3587072491645813, -0.09099902212619781, -0.08881594985723495, -0.0064687058329582214, 0.28515785932540894, 0.23027527332305908, -0.2992844581604004, 0.018360627815127373, 0.4084852337837219, 0.11278808116912842, -0.2255142480134964, 0.13825823366641998, 0.03543659299612045, -0.43261680006980896, 0.5216683149337769, -0.047607384622097015, 0.6814639568328857, 0.2664852738380432, 0.06667885184288025, 0.4089907109737396, 0.01980314403772354, 0.6191628575325012, -0.12821854650974274, 0.0904495045542717, 0.008613530546426773, -0.07493537664413452, -0.02831854857504368, 0.2580915689468384, 0.28022441267967224, -0.14534053206443787, -0.3026861846446991, 0.35803529620170593, 0.4741479754447937, -0.028228703886270523, 0.0660800188779831, 0.1906587779521942, 0.2675008475780487, -0.35727664828300476, -0.17591629922389984, 0.19538575410842896, -0.05740012973546982, 0.1983063966035843, -0.009514261037111282, -0.07632473856210709, -0.03352481126785278, -0.34462064504623413, -0.09483597427606583, -0.0046200137585401535, 0.40876203775405884, 0.06272788345813751, -0.013134617358446121, 0.00018253177404403687, 0.35457631945610046, -0.25106653571128845, 0.20681530237197876, 0.1270361840724945, -0.23320049047470093, 0.13416142761707306, 0.030415289103984833, 0.033475086092948914, -0.057564765214920044, 0.06580778956413269, 0.17728278040885925, -0.21683084964752197, -0.09247618913650513, 0.13309326767921448, 0.2801990211009979, -0.2276497483253479, 0.07511956989765167, -0.05992458760738373, -0.09679773449897766, 0.09913437068462372, 0.2709742784500122, 0.5199607014656067, 0.04232681170105934, -0.2932434380054474, 0.24941076338291168, 0.21730060875415802, -0.2473541647195816, 0.1259586215019226, 0.1490776687860489, -0.14991825819015503, -0.126140296459198, 0.4295508861541748, 0.3287908434867859, -0.17596818506717682, 0.1623610258102417, 0.028476331382989883, -0.06598136574029922, -0.2654573321342468, -0.24508410692214966, -0.0956837460398674, -0.575057327747345, -0.0652821883559227, 0.11142811924219131, -0.1890368014574051, 0.20219171047210693, 0.7269651293754578, -0.03920883685350418, 0.20746362209320068, -0.10957221686840057, -0.2568226456642151, -0.27607429027557373, 0.05067761242389679, -0.4521557688713074, 0.2244728058576584, 0.14675357937812805, 0.05748944357037544, -0.057873405516147614, 0.08658529818058014, -0.4668600857257843, -0.22374434769153595, 0.07354412227869034, -0.011187952011823654, 0.3666561245918274, 0.08408299833536148, 0.33175262808799744, 0.07002590596675873, 0.2591492533683777, 0.01709246262907982, -0.18132854998111725, -0.27145418524742126, -0.12523697316646576, 0.07885979861021042, 0.021880485117435455, 0.17113512754440308, 0.09806933254003525, -0.056750908493995667, -0.2730424106121063, -0.0597003772854805, 0.1882500797510147, -0.08539941161870956, -0.1864279955625534, 0.37596821784973145, -0.20979566872119904, -0.13131308555603027, 0.37401872873306274, 0.29977643489837646, 0.20191779732704163, 0.013176128268241882, -0.10522716492414474, 0.4919646680355072, -0.06271235644817352, -0.10722265392541885, 0.04128659516572952, 0.1994611769914627, -0.21136488020420074, 0.10680319368839264, 0.08347342163324356, 0.0005004703998565674, -0.25967687368392944, -0.20306336879730225, 0.2125796377658844, -0.29589134454727173, 0.015859704464673996, 0.06989350914955139, 0.0752127394080162, 0.3328959047794342, -0.43360841274261475, 0.15516597032546997, -0.06994199752807617, -0.15005750954151154, 0.3807324469089508, 0.1659955382347107, 0.2593381404876709, -0.1449124664068222, 0.6386486291885376, -0.011185507290065289, -0.019398774951696396, 0.06352224200963974, 0.2700665295124054, 0.007332455366849899, 0.08318677544593811, 0.0074922144412994385, 0.06729623675346375, 0.002586677670478821, 0.26146405935287476, -0.024469688534736633, 0.043507687747478485, -0.026396160945296288, 0.1955518126487732, 0.39803504943847656, -0.004165023565292358, -0.21858030557632446, -0.04063562676310539, 0.2692480981349945, 0.1335301250219345, 0.22033315896987915, 0.12551990151405334, 0.013996191322803497, -0.0264744870364666, 0.0733947902917862, -0.6518140435218811, 0.10224491357803345, 0.046333711594343185, -0.2674894332885742, -0.19240398705005646, -0.060146886855363846, -0.5283677577972412, -0.03477142006158829, 0.1602238118648529, -0.09570488333702087, 0.11657209694385529, -0.05166510492563248, 0.07509457319974899, -0.17832216620445251, -0.06448644399642944, 0.1521918773651123, 0.2097455859184265, -0.07242782413959503, 0.26005953550338745, -0.3143095374107361, -0.13356737792491913, 0.11316905915737152, 0.33179306983947754, 0.25361499190330505, -0.07983915507793427, 0.25138118863105774, 0.18931564688682556, 0.07819020748138428, -0.06341112405061722, 0.015542060136795044, 0.1274963617324829, 0.1751863658428192, 0.15679259598255157, 0.34565529227256775, 0.19967436790466309, -0.23538140952587128, 0.10904023051261902, 0.018087133765220642, -0.1851758062839508, -0.18787136673927307, 0.2947615683078766, 0.2506910562515259, 0.12793797254562378, -0.15492777526378632, 0.056655243039131165, -0.22269272804260254, -0.31515467166900635, -0.05655486509203911, -0.04162641614675522, -0.06579113751649857, -0.269204705953598, 0.1421038955450058, -0.10145178437232971, 0.24289833009243011, 0.20207802951335907, 0.3052961528301239, 0.07524070888757706, -0.4313358664512634, -0.16718514263629913, 0.21204820275306702, 0.18553729355335236, -0.23350785672664642, -0.04269712418317795, 0.16694118082523346, 0.14643627405166626, -0.2412625402212143, 0.10201048851013184, -0.1333187371492386, 0.42490410804748535, -0.0015241876244544983, -0.39028072357177734, 0.1590537130832672, 0.14354091882705688, -0.0917150229215622, -0.03733240067958832, -0.259224534034729, -0.3049335777759552, -0.1383424997329712, 0.13726061582565308, -0.2919718027114868, 0.09674203395843506, 0.006600707769393921, 0.13372191786766052, 0.419193297624588, 0.029713384807109833, 0.2779366374015808, -0.33394160866737366, -0.006887739524245262, -0.3844677209854126, -0.25012096762657166, -0.06227702647447586, 0.046573176980018616, 0.096226766705513, 0.17797496914863586, 0.020027371123433113, 0.1866796761751175, 0.04949438199400902, 0.39270493388175964, 0.03392908722162247, -0.23147547245025635, -0.04233120381832123, 0.5437836647033691, -0.3518795371055603, -0.20718622207641602, -0.07825567573308945, -0.00487276166677475, 0.039167795330286026, -0.13390235602855682, -0.05519881844520569, -0.30653655529022217, 0.3257245421409607, -0.3351879119873047, -0.058628469705581665, -0.07195790857076645, 0.3476722240447998, 0.1555512547492981, -0.006516776978969574, -0.3395659625530243, 0.16316473484039307, 0.14248691499233246, 0.16235291957855225, -0.03403221070766449, 0.39861780405044556, -0.1481122374534607, -0.012054719030857086, -0.21766544878482819, -0.16810856759548187, -0.06432831287384033, -0.08481381088495255, 0.13156446814537048, -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).
[ -0.22634360194206238, -0.2859199047088623, -0.11841721832752228, -0.4919188618659973, -0.45303940773010254, -0.25085964798927307, 0.21630574762821198, 0.22147011756896973, -0.035638511180877686, 0.3164646625518799, 0.07013271749019623, 0.4571438431739807, 0.032457053661346436, 0.353317528963089, -0.16244831681251526, 0.4214378893375397, 0.11718563735485077, 0.2206689417362213, -0.030111968517303467, -0.25561654567718506, -0.22714577615261078, 0.1309107542037964, -0.16859979927539825, -0.1252732276916504, 0.16298648715019226, 0.2887749671936035, -0.2544236481189728, -0.09574774652719498, -0.04146188497543335, -0.18337297439575195, 0.4028764069080353, 0.34786319732666016, 0.17024505138397217, 0.06462868303060532, -0.00011004399857483804, -0.13638831675052643, 0.4567965865135193, -0.06395504623651505, -0.2753702700138092, -0.1536128669977188, -0.35569217801094055, -0.3650529980659485, 0.07075216621160507, -0.07159383594989777, 0.21727675199508667, 0.16272687911987305, 0.14945903420448303, -0.031100904569029808, 0.04020855575799942, 0.16627220809459686, 0.269146591424942, 0.2860799729824066, -0.03167672082781792, -0.33285099267959595, 0.21256232261657715, 0.47362667322158813, -0.3600451350212097, -0.07488913834095001, 0.043695442378520966, -0.09709266573190689, -0.14182782173156738, 0.3820023536682129, 0.22545596957206726, -0.17204217612743378, -0.2601481080055237, -0.030674681067466736, 0.21465083956718445, 0.1305542290210724, 0.2605668902397156, 0.0569026842713356, 0.35295259952545166, 0.1386500895023346, -0.26492398977279663, 0.2500227987766266, -0.05883818492293358, -0.42920055985450745, 0.2250400185585022, -0.16231562197208405, 0.021105274558067322, 0.004746139049530029, 0.35412001609802246, -0.18666288256645203, -0.5334497690200806, 0.15618228912353516, -0.21667085587978363, 0.17143064737319946, -0.2124955803155899, -0.14087381958961487, 0.09422009438276291, 0.10761164128780365, -0.09298375993967056, 0.11195191740989685, -0.20708155632019043, 0.18042734265327454, -0.34744328260421753, -0.10264956206083298, 0.30851954221725464, -0.06392960995435715, 0.42677047848701477, 0.15892182290554047, 0.19507107138633728, 0.19056737422943115, -0.01569845899939537, 0.15070189535617828, 0.01156392227858305, 0.36866244673728943, 0.5664448142051697, 0.0640154629945755, 0.4030507206916809, 0.14254896342754364, 0.40475839376449585, 0.016548022627830505, 0.0781300961971283, 0.02118225395679474, -0.002247307449579239, -0.16156670451164246, 0.11037556827068329, -0.5184801816940308, -0.18017205595970154, -0.03196031600236893, -0.10153421759605408, 0.09271793812513351, 0.004587896168231964, 0.2791184186935425, 0.2888188064098358, 0.1953786313533783, 0.32147789001464844, -0.02211962640285492, 0.016755154356360435, -0.2821887135505676, -0.2884853184223175, -0.08126683533191681, -0.2485470026731491, 0.2028690129518509, 0.19906418025493622, -0.3987305164337158, 0.2090894877910614, -0.364455908536911, 0.4094005823135376, -0.06413433700799942, -0.22893166542053223, 0.3046197295188904, -0.035716377198696136, 0.26344993710517883, 0.038811665028333664, -0.017299553379416466, -0.09945590794086456, 0.3986690044403076, -0.16598179936408997, 0.11357759684324265, -0.21756243705749512, -0.48887720704078674, -0.2699550688266754, 0.26509127020835876, -0.1804889291524887, -0.2526893615722656, -0.24294298887252808, 0.41736164689064026, -0.2470448613166809, -0.007607344537973404, 0.01368434727191925, 0.1410551220178604, -0.32141146063804626, -0.20329102873802185, -0.08240259438753128, 0.19734466075897217, -0.04359922558069229, -0.08860630542039871, -0.6579610705375671, -0.09929913282394409, 0.015475288033485413, -0.18410447239875793, -0.018393991515040398, -0.30888569355010986, -0.24293749034404755, -0.09758506715297699, 0.4091992974281311, -0.540909469127655, -0.08889669179916382, 0.30070286989212036, -0.05289766564965248, -0.21085941791534424, 0.1997181475162506, -0.30009984970092773, -0.15993160009384155, -0.2848050892353058, -0.3112843334674835, 0.05204365402460098, -0.03445470333099365, -0.05105540156364441, -0.0711141973733902, -0.34644249081611633, -0.34897786378860474, 0.2585710883140564, 0.36376211047172546, 0.03293449059128761, 0.3567889630794525, -0.41883379220962524, 0.15049877762794495, -0.05560467764735222, 0.13320334255695343, -0.025527622550725937, 0.6704121828079224, 0.10371123254299164, 0.19556771218776703, 0.13810691237449646, -0.3721381425857544, 0.09697645157575607, 0.18270781636238098, 0.1522466391324997, 0.20965853333473206, -0.36246728897094727, -0.15044835209846497, 0.14429157972335815, -0.006205279380083084, -0.19019728899002075, 0.18903614580631256, 0.16746029257774353, -0.06673913449048996, -0.03944755345582962, -0.04032126069068909, 0.4120458960533142, -0.333312064409256, -0.08935071527957916, 0.16916675865650177, 0.06708654761314392, -0.05909055471420288, -0.007216337136924267, -0.052231565117836, 0.09110017120838165, -0.003276742994785309, 0.008714205585420132, -0.1518043875694275, 0.3978510797023773, -0.014699926599860191, 0.137334942817688, 0.07569354772567749, 0.20276620984077454, 0.12991942465305328, -0.11296959221363068, 0.015640564262866974, 0.031186847016215324, -0.0605439729988575, 0.10886550694704056, -0.21376872062683105, 0.36783480644226074, -0.04400141164660454, 0.07868464291095734, -0.003053363412618637, -0.08954405039548874, 0.026837507262825966, -0.0728670060634613, -0.5226487517356873, -0.21800456941127777, 0.24215564131736755, 0.10695752501487732, -0.42331528663635254, -0.23976372182369232, -0.22046196460723877, -0.08050710707902908, 0.1762455701828003, -0.10011544823646545, 0.15004447102546692, 0.3162190020084381, 0.10816863924264908, -0.1911516785621643, 0.27675729990005493, -0.1665319800376892, 0.406342089176178, 0.31514111161231995, -0.1061272919178009, -0.048021040856838226, -0.30542588233947754, -0.05322825536131859, 0.13648022711277008, 0.2142147272825241, 0.12367867678403854, 0.1257350742816925, 0.41697460412979126, -0.14952263236045837, -0.5176544189453125, -0.4135934114456177, -0.06272319704294205, -0.16659285128116608, -0.19666188955307007, -0.16656161844730377, -0.06399566680192947, 0.09422243386507034, -0.10740857571363449, 0.038636185228824615, 0.08334244787693024, -0.23619933426380157, 0.2636566162109375, 0.22466324269771576, -0.061354316771030426, 0.23576238751411438, 0.025401555001735687, 0.6541306972503662, -0.02252654731273651, 0.13648198544979095, -0.15872688591480255, 0.06660034507513046, -0.420367956161499, 0.19410818815231323, -0.5060151219367981, -0.22600547969341278, 0.4339211881160736, -0.007591546513140202, 0.3097063899040222, -0.06009943038225174, -0.6933549642562866, 0.07365789264440536, -0.19269134104251862, -0.1378568857908249, 0.06659695506095886, -0.28835171461105347, -0.11153594404459, 0.04072219133377075, 0.08749237656593323, -0.39441999793052673, -0.3563591539859772, -0.08151979744434357, -0.23026928305625916, 0.27909958362579346, -0.34588751196861267, -0.4851737320423126, 0.23486736416816711, -0.33899614214897156, -0.11091116070747375, -0.08392272144556046, 0.12878653407096863, 0.5446902513504028, 0.13814018666744232, 0.08742569386959076, 0.1251380741596222, -0.014643311500549316, -0.41065657138824463, 0.23893660306930542, 0.340101957321167, -0.3349621295928955, -0.40006378293037415, 0.10895256698131561, 0.09359128773212433, 0.1544482409954071, 0.20974069833755493, -0.186634361743927, -0.5588259696960449, 0.18018372356891632, 0.15571671724319458, -0.2045297622680664, 0.15855006873607635, 0.38408079743385315, -0.01926819048821926, -0.19568870961666107, -0.18685048818588257, -0.3143167197704315, 0.2257436215877533, 0.3081617057323456, 0.1298706978559494, -0.16425472497940063, 0.2678801119327545, 0.16116443276405334, 0.1489793360233307, 0.08203333616256714, 0.019120145589113235, 0.0505388081073761, 0.11084745079278946, 0.35109445452690125, -0.025693602859973907, 0.05527877062559128, 0.13333112001419067, 0.08192263543605804, 0.2069118618965149, 0.46736425161361694, 0.1389227658510208, -0.2309902012348175, -0.2874261736869812, -0.2523173689842224, 0.16300898790359497, 0.03166082873940468, -0.16662462055683136, 0.18828527629375458, 0.43356117606163025, -0.12905935943126678, -0.09658840298652649, -0.32698288559913635, 0.03286070004105568, 0.061074819415807724, 0.42140090465545654, 0.09682787209749222, -0.17435342073440552, -0.004774656146764755, 0.14035241305828094, -0.32818493247032166, 0.3598346710205078, 0.10829274356365204, 0.051345594227313995, -0.003542773425579071, -0.20926478505134583, 0.21159037947654724, -0.029142964631319046, 0.43104124069213867, -0.180899515748024, 0.013701669871807098, -0.11118042469024658, -0.07789909094572067, -0.0930248275399208, -0.16937604546546936, -0.117866650223732, -0.30604276061058044, -0.03530868515372276, -0.0355609655380249, -0.0528147891163826, -0.02926049381494522, 0.5762631893157959, -0.2663783133029938, 0.07449094951152802, -0.2874302566051483, -0.2082001268863678, -0.4295231103897095, -0.10088328272104263, 0.004545547068119049, -0.05962265282869339, -0.09537079930305481, -0.22082705795764923, -0.13561594486236572, 0.2697882056236267, -0.22745123505592346, -0.14196427166461945, 0.12684498727321625, -0.2542559802532196, -0.0007774727419018745, -0.03305472806096077, 0.10001487284898758, -0.11615099012851715, 0.3154967725276947, 0.016632303595542908, 0.28698280453681946, -0.3732559382915497, 0.10811797529459, -0.1803286075592041, 0.21710634231567383, 0.2719831168651581, 0.0013405452482402325, 0.14543338119983673, -0.44116175174713135, -0.23803676664829254, -0.013946308754384518, -0.08445215970277786, 0.21031159162521362, -0.004942900966852903, -0.20481325685977936, -0.046241022646427155, 0.3996998369693756, 0.05420348420739174, -0.2666933834552765, 0.2724321484565735, 0.14732803404331207, -0.2883746027946472, 0.44838619232177734, 0.011259645223617554, 0.6613934636116028, 0.2898107171058655, -0.2178923785686493, 0.29652997851371765, -0.21648341417312622, 0.567108690738678, -0.18582533299922943, 0.13965463638305664, 0.09834583103656769, -0.14164145290851593, 0.02411883883178234, 0.26291313767433167, 0.36447304487228394, -0.04198654741048813, -0.32919690012931824, 0.24835877120494843, 0.21675215661525726, 0.30089035630226135, 0.21273399889469147, 0.11631877720355988, 0.0853671133518219, -0.3070460855960846, 0.0013402141630649567, 0.18750104308128357, 0.08409202843904495, 0.27029120922088623, -0.04117569699883461, 0.09877137094736099, 0.028510134667158127, -0.3716997504234314, -0.22798886895179749, 0.03163043409585953, 0.4609610438346863, 0.11891809105873108, 0.01270783506333828, 0.06544384360313416, 0.23905551433563232, -0.32361117005348206, 0.2711412012577057, 0.1031317487359047, -0.0883096233010292, 0.2994316518306732, 0.31932589411735535, 0.01049517560750246, -0.17676059901714325, -0.1510268598794937, 0.024320952594280243, -0.15881362557411194, -0.2019999623298645, 0.22722762823104858, 0.17419648170471191, -0.23866330087184906, -0.20172880589962006, 0.033909138292074203, -0.05447171628475189, 0.14230918884277344, 0.5334226489067078, 0.43454355001449585, 0.10620944201946259, -0.1701265126466751, 0.20466430485248566, 0.126929372549057, -0.11475132405757904, 0.0031180456280708313, 0.15291021764278412, -0.19069086015224457, -0.18325138092041016, 0.5322153568267822, 0.2791718542575836, -0.13204531371593475, -0.06292682886123657, -0.11571274697780609, 0.01629609242081642, -0.21505886316299438, -0.1890406608581543, 0.14092497527599335, -0.3937106430530548, -0.12774957716464996, 0.1681395024061203, -0.28275278210639954, 0.24938490986824036, 0.8095275163650513, 0.07201125472784042, 0.20189455151557922, -0.22826357185840607, -0.12816625833511353, -0.0254083052277565, -0.07955542206764221, -0.47719642519950867, 0.1812090128660202, -0.020505765452980995, -0.05071515217423439, -0.08789937943220139, -0.012811316177248955, -0.3956336975097656, -0.24205580353736877, 0.19528831541538239, -0.09695497155189514, 0.4021480083465576, 0.21221643686294556, 0.4604622423648834, 0.18588757514953613, 0.21235516667366028, 0.015105679631233215, -0.2088434398174286, -0.1787453293800354, -0.013656483963131905, 0.10726441442966461, 0.08703909069299698, 0.2987450957298279, -0.1990329772233963, -0.182963564991951, -0.29235193133354187, 0.022287528961896896, 0.3350929617881775, 0.10991951078176498, -0.19674436748027802, 0.08827021718025208, -0.45101431012153625, 0.010249175131320953, 0.3422631025314331, 0.18725672364234924, 0.17638683319091797, 0.0917046070098877, -0.06608221679925919, 0.3762291669845581, -0.12612779438495636, -0.026595328003168106, 0.12284565716981888, 0.3449653685092926, -0.25990837812423706, 0.1292843222618103, 0.008576061576604843, 0.08216619491577148, -0.19888152182102203, -0.17354610562324524, 0.01640097051858902, -0.3018588125705719, 0.006897013634443283, -0.05663786828517914, -0.16726350784301758, 0.30348318815231323, -0.23948132991790771, 0.06966474652290344, -0.2762695550918579, -0.16598938405513763, 0.2079390287399292, 0.2772292494773865, 0.25617948174476624, -0.09345092624425888, 0.7590211629867554, 0.10688691586256027, 0.14143744111061096, -0.07472974807024002, 0.22079774737358093, 0.16938209533691406, 0.04263456538319588, 0.16868597269058228, 0.275814026594162, -0.08789417147636414, 0.3604969382286072, -0.07755021750926971, 0.3008502721786499, 0.03680393472313881, 0.2739916145801544, 0.5524515509605408, 0.05506028234958649, -0.31329697370529175, 0.10778632014989853, 0.46236613392829895, 0.05116477236151695, 0.058433495461940765, 0.18636423349380493, -0.22566846013069153, -0.02091851457953453, 0.18213818967342377, -0.7633129954338074, -0.1767711490392685, 0.08755525201559067, -0.2771150469779968, -0.19742590188980103, -0.10568825155496597, -0.47327953577041626, -0.0067762937396764755, 0.2994886040687561, 0.05078279227018356, -0.01612088829278946, -0.20815585553646088, 0.00819290429353714, -0.060377806425094604, -0.27781009674072266, 0.07377645373344421, 0.25696396827697754, 0.0064140744507312775, 0.2032153308391571, -0.41788995265960693, -0.13610747456550598, 0.37700480222702026, 0.13934074342250824, 0.13725166022777557, -0.08849247545003891, 0.5053783059120178, 0.11966408789157867, 0.10259053111076355, -0.09160061180591583, -0.009294308722019196, 0.18863511085510254, 0.12328091263771057, 0.10457594692707062, 0.18037049472332, 0.1920611411333084, -0.18656572699546814, 0.15649476647377014, -0.03523563966155052, -0.24298474192619324, -0.2443149983882904, 0.18434317409992218, 0.10635501146316528, 0.15036237239837646, -0.08765771985054016, 0.22046926617622375, 0.0323292538523674, -0.3462274670600891, -0.06211291626095772, 0.020574815571308136, -0.127640962600708, -0.14604659378528595, 0.11986339092254639, -0.06510604918003082, -0.12488344311714172, 0.1699475198984146, -0.1149890199303627, 0.06258704513311386, -0.37308019399642944, -0.004313208162784576, 0.3659430742263794, 0.18773038685321808, 0.07066258043050766, 0.054785750806331635, 0.14213617146015167, 0.14274626970291138, -0.2538003921508789, 0.12222763895988464, -0.4326280951499939, 0.5907477140426636, -0.08773459494113922, -0.182802215218544, 0.14077596366405487, 0.4150146245956421, -0.11227633059024811, -0.12778548896312714, -0.29606902599334717, -0.2116578370332718, 0.06730365008115768, 0.10888493806123734, -0.33292514085769653, -0.1272968351840973, 0.18393906950950623, 0.03788965940475464, 0.4441293179988861, 0.04809124022722244, 0.21306069195270538, -0.12853814661502838, -0.07115915417671204, -0.5014457702636719, -0.13466623425483704, -0.02391596883535385, 0.09259101003408432, -0.1338445246219635, -0.02983829379081726, 0.04316052049398422, 0.1195140928030014, 0.02203383296728134, 0.3425883650779724, 0.04612986743450165, -0.0907575711607933, -0.09505608677864075, 0.4244759976863861, -0.5147942304611206, -0.3080165684223175, -0.106026791036129, -0.10277212411165237, 0.06682396680116653, -0.12948061525821686, -0.04089657962322235, -0.2425185889005661, 0.3495897054672241, -0.22609196603298187, -0.08503811061382294, -0.08611143380403519, 0.434344083070755, 0.20386970043182373, 0.08677305281162262, -0.22049090266227722, 0.09851088374853134, 0.2275274693965912, 0.19771936535835266, -0.2760642170906067, 0.15084731578826904, -0.04482172802090645, 0.017772745341062546, -0.2207435965538025, 0.023583587259054184, -0.1709718108177185, 0.01504305936396122, 0.1533428579568863, -0.3238043785095215 ]
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)
[ -0.16037391126155853, -0.1905597746372223, -0.04093010723590851, -0.5251523852348328, -0.5301496982574463, -0.20913472771644592, 0.07916980981826782, 0.30271077156066895, -0.08486784994602203, 0.299315482378006, 0.09133685380220413, 0.5883416533470154, -0.1273888498544693, 0.3007548153400421, 0.14482657611370087, 0.19229458272457123, 0.04938075318932533, 0.35735592246055603, 0.016091279685497284, -0.14633303880691528, -0.2847519516944885, 0.2668963074684143, -0.15900404751300812, 0.022066358476877213, 0.1762133240699768, 0.1823170781135559, -0.3217965066432953, -0.026325665414333344, -0.06103866547346115, -0.2656816840171814, 0.43385595083236694, 0.29688435792922974, 0.24196362495422363, -0.15443021059036255, -0.00011306988017167896, -0.24875320494174957, 0.5088641047477722, -0.06220303475856781, -0.4297455847263336, -0.30758827924728394, -0.33396410942077637, -0.518495500087738, 0.1690250188112259, -0.09877884387969971, 0.18584349751472473, 0.15064656734466553, 0.36876288056373596, -0.03254234418272972, -0.0012722685933113098, 0.10815970599651337, 0.22739464044570923, 0.28239163756370544, -0.08409285545349121, -0.2982408404350281, 0.11302663385868073, 0.3750301003456116, -0.3086332678794861, -0.16344043612480164, -0.04228101670742035, -0.2713524103164673, -0.052446700632572174, 0.3673299252986908, 0.20168179273605347, -0.22526933252811432, -0.1958562433719635, -0.047848042100667953, 0.14524710178375244, 0.04428553581237793, 0.07859809696674347, 0.09422263503074646, 0.34997427463531494, 0.046612564474344254, -0.26323071122169495, 0.04236665368080139, 0.07684116065502167, -0.5963772535324097, 0.2815065383911133, -0.06955289095640182, -0.17123259603977203, -0.07540741562843323, 0.265972375869751, -0.27415749430656433, -0.42526862025260925, 0.38042956590652466, -0.06080397218465805, 0.26074498891830444, -0.06658745557069778, -0.06787668913602829, 0.23503394424915314, 0.08089528977870941, -0.10468436777591705, 0.02762921340763569, -0.14414864778518677, 0.2545181214809418, -0.41611653566360474, -0.14212238788604736, 0.4012792706489563, 0.04141942784190178, 0.2769642770290375, 0.14983642101287842, 0.27908506989479065, 0.012850144878029823, -0.05144141614437103, 0.18539756536483765, 0.08551662415266037, 0.31466540694236755, 0.6022413372993469, 0.0287075936794281, 0.2610603868961334, 0.09194577485322952, 0.5128618478775024, 0.10024430602788925, 0.0753757506608963, 0.02059401199221611, -0.16388960182666779, -0.12788313627243042, 0.11135821044445038, -0.3812928795814514, -0.1226559653878212, -0.09454679489135742, -0.09384286403656006, 0.06300080567598343, 0.02480902150273323, 0.19082999229431152, 0.2591687738895416, 0.2565681040287018, 0.3442854881286621, -0.07276070863008499, -0.1324227899312973, -0.38469749689102173, -0.24910198152065277, -0.05333347246050835, -0.39312514662742615, 0.27025628089904785, 0.3587169647216797, -0.23135316371917725, 0.28058433532714844, -0.4026279151439667, 0.19323456287384033, 0.01676630973815918, -0.21883098781108856, 0.31003624200820923, -0.16564753651618958, 0.07989118993282318, 0.04688231274485588, 0.013723818585276604, -0.30024245381355286, 0.2614317536354065, -0.25514549016952515, 0.09527446329593658, -0.2514302134513855, -0.5217938423156738, -0.2898748219013214, 0.2131016105413437, -0.21021588146686554, -0.3397212326526642, -0.16480210423469543, 0.48887547850608826, -0.22434589266777039, -0.2618383467197418, -0.01977597177028656, 0.07405805587768555, -0.37112361192703247, -0.1326882541179657, -0.08441707491874695, 0.1536417007446289, 0.032275158911943436, -0.23951095342636108, -0.6589329242706299, 0.023701321333646774, 0.05040479078888893, -0.10603713989257812, -0.10561701655387878, -0.2657807171344757, -0.14965106546878815, -0.059897467494010925, 0.7229763269424438, -0.4373762905597687, -0.11550042033195496, 0.2689363956451416, -0.06760762631893158, -0.20526154339313507, 0.1048460379242897, -0.13712191581726074, -0.23001375794410706, -0.15859819948673248, -0.17310106754302979, 0.2563074827194214, -0.05131041258573532, -0.10487190634012222, -0.08494409918785095, -0.3723757565021515, -0.08402266353368759, 0.2475232034921646, 0.2598855793476105, 0.010238341987133026, 0.26540127396583557, -0.17162618041038513, 0.20318901538848877, -0.0694110244512558, 0.14847148954868317, 0.03821396082639694, 0.4409765899181366, 0.04296783357858658, 0.09988922625780106, 0.1025068536400795, -0.5050915479660034, 0.07856355607509613, -0.04806385189294815, 0.2747189998626709, 0.32303059101104736, -0.21497683227062225, -0.0498177707195282, -0.05138123407959938, 0.06278397887945175, -0.242958664894104, 0.12956157326698303, 0.2916696071624756, -0.0453818254172802, -0.027404244989156723, -0.08810260891914368, 0.4771674871444702, -0.2147001028060913, -0.17016050219535828, -0.22500795125961304, 0.0794714018702507, -0.1317957043647766, -0.03842912241816521, 0.0011595971882343292, 0.30664271116256714, -0.005750447511672974, 0.0314609594643116, -0.08589667081832886, 0.3410400450229645, -0.170721635222435, 0.2327120304107666, 0.0020121224224567413, 0.051782846450805664, 0.15532881021499634, -0.17103667557239532, 0.07149931788444519, -0.03874899446964264, -0.05914821848273277, 0.1741674244403839, -0.23587444424629211, 0.40878480672836304, 0.055551592260599136, -0.004943102598190308, 0.015664152801036835, -0.025185253471136093, 0.024581311270594597, -0.2130439132452011, -0.5066878795623779, -0.1608453392982483, 0.3907111585140228, 0.09037633240222931, -0.4192774295806885, -0.21827377378940582, -0.4219490885734558, -0.053912483155727386, 0.2659192383289337, 0.010921310633420944, 0.06860508769750595, 0.34612029790878296, 0.19898223876953125, -0.2583036720752716, 0.28767603635787964, -0.03375620394945145, 0.5213517546653748, 0.3559092879295349, -0.14374244213104248, 0.055378712713718414, -0.497071772813797, -0.03433407098054886, 0.19374391436576843, 0.22271722555160522, 0.16498364508152008, 0.034512847661972046, 0.47993749380111694, 0.02471921592950821, -0.4151760935783386, -0.4975740611553192, 0.03008299507200718, -0.066778264939785, -0.1709928810596466, -0.07678590714931488, -0.23566105961799622, -0.24855297803878784, -0.13437846302986145, 0.15273742377758026, -0.0066424161195755005, -0.282269686460495, 0.25305429100990295, 0.2228183001279831, -0.09684845805168152, 0.27046480774879456, -0.08819729089736938, 0.7165156006813049, -0.1732388734817505, 0.13250194489955902, -0.21116551756858826, -0.019733045250177383, -0.43862247467041016, 0.1435931771993637, -0.36732134222984314, -0.12818720936775208, 0.43063753843307495, 0.03370526432991028, 0.2416907399892807, 0.10172414034605026, -0.6566219925880432, 0.07370659708976746, -0.16742011904716492, -0.08497529476881027, 0.07531517744064331, -0.3813067674636841, -0.17965354025363922, -0.09716552495956421, 0.09173286706209183, -0.34687790274620056, -0.3567788600921631, -0.06084062159061432, -0.17782963812351227, 0.19364377856254578, -0.38614848256111145, -0.5043299794197083, -0.0028921980410814285, -0.3854024410247803, -0.06033317744731903, -0.03287291154265404, 0.26456350088119507, 0.7287460565567017, -0.02927185222506523, 0.06091072037816048, -0.018814094364643097, 0.05397900193929672, -0.32225850224494934, 0.41111987829208374, 0.2417624145746231, -0.36633071303367615, -0.32564133405685425, -0.09408034384250641, -0.00901089422404766, 0.27595460414886475, 0.1377246379852295, -0.20072084665298462, -0.3760529160499573, 0.17288640141487122, 0.09563268721103668, -0.18954989314079285, 0.16957513988018036, 0.49473267793655396, 0.03568524867296219, -0.14348937571048737, -0.1199135035276413, -0.3978126049041748, 0.2254878580570221, 0.5513763427734375, 0.3134695291519165, -0.09973718971014023, 0.23797953128814697, 0.12890547513961792, 0.05465463176369667, 0.18605999648571014, -0.03154749795794487, 0.21331404149532318, 0.07510698586702347, 0.30275553464889526, -0.06819289177656174, -0.04883962124586105, 0.1862042099237442, 0.11914383620023727, 0.32899224758148193, 0.4975132346153259, 0.10180947184562683, -0.49265363812446594, -0.3609597980976105, -0.24205844104290009, 0.12460757046937943, 0.003417808562517166, -0.06890664994716644, 0.060732077807188034, 0.4087863266468048, 0.02603987231850624, -0.26158949732780457, -0.413523405790329, -0.17632880806922913, -0.10904942452907562, 0.438190221786499, 0.032993145287036896, 0.01286526769399643, -0.08266288042068481, 0.026466118171811104, -0.42439061403274536, 0.2321767956018448, 0.06501637399196625, 0.2848566770553589, -0.05420083925127983, -0.13267439603805542, 0.11208420991897583, 0.08531028777360916, 0.2600693106651306, -0.23703260719776154, -0.12139749526977539, -0.011055062524974346, 0.12416991591453552, -0.19625326991081238, -0.06323409080505371, -0.16072770953178406, -0.17356686294078827, 0.07299992442131042, 0.016527269035577774, 0.012415766716003418, -0.18408209085464478, 0.5894104242324829, -0.1919885128736496, 0.02772824838757515, -0.3377680480480194, -0.15445546805858612, -0.2920680642127991, -0.09544730186462402, -0.088752381503582, -0.1869194507598877, 0.1422540843486786, -0.20191369950771332, -0.005332905799150467, 0.22868721187114716, -0.0948631688952446, -0.005437694489955902, 0.22884415090084076, -0.24832817912101746, 0.13451328873634338, -0.0950876772403717, 0.0174885131418705, 0.09789498895406723, 0.039771780371665955, -0.045927368104457855, 0.23973794281482697, -0.09043021500110626, 0.1836397647857666, 0.04171430692076683, 0.34280651807785034, 0.29112493991851807, 0.07612721621990204, 0.15024366974830627, -0.38553011417388916, -0.25055748224258423, -0.04244660586118698, -0.08079680800437927, 0.24254879355430603, -0.06657206267118454, -0.33067649602890015, -0.09095704555511475, 0.4912468492984772, 0.06609167158603668, -0.3306334912776947, 0.32408303022384644, 0.08614321053028107, -0.3378436267375946, 0.4092395305633545, -0.128919780254364, 0.7491973638534546, 0.39504677057266235, 0.03774014860391617, 0.32539910078048706, -0.03980815410614014, 0.8740342855453491, -0.03195716440677643, 0.276810884475708, 0.07144159078598022, 0.05260826647281647, 0.03373285382986069, 0.2498621940612793, 0.3760501444339752, -0.015048772096633911, -0.3114895820617676, 0.37499985098838806, 0.5791000127792358, 0.0683809369802475, 0.19329120218753815, 0.35138118267059326, 0.1492745578289032, -0.4552721381187439, -0.16807512938976288, 0.12077014148235321, -0.10806265473365784, 0.38529083132743835, -0.01413146685808897, -0.006577968597412109, 0.01694067195057869, -0.3202211260795593, -0.21057552099227905, 0.04029100388288498, 0.39293041825294495, 0.05890291929244995, 0.07368309795856476, 0.028256915509700775, 0.4237452745437622, -0.23016300797462463, 0.18266725540161133, 0.1609453707933426, -0.2040279060602188, 0.16447216272354126, 0.13954228162765503, -0.01857990026473999, -0.07313825190067291, -0.14397455751895905, 0.08395542949438095, -0.17628920078277588, -0.2379300743341446, 0.10621485114097595, 0.19553440809249878, -0.20474018156528473, -0.11868377774953842, -0.0075542256236076355, -0.038433343172073364, 0.245245561003685, 0.29218775033950806, 0.5004429817199707, 0.10781100392341614, -0.23766829073429108, 0.17232924699783325, 0.09804810583591461, -0.10802941769361496, 0.02397996559739113, 0.20026244223117828, -0.15120664238929749, -0.11280268430709839, 0.5301057696342468, 0.21330490708351135, -0.18939775228500366, 0.10306406766176224, -0.14518506824970245, 0.07328271120786667, -0.19599440693855286, -0.15487869083881378, -0.03298511356115341, -0.5144609212875366, -0.19930748641490936, 0.1846107691526413, -0.2377409040927887, 0.3092068135738373, 0.7213867902755737, 0.01655513234436512, 0.24946650862693787, -0.1088578999042511, -0.18847283720970154, -0.05346190929412842, 0.02773410826921463, -0.5827196836471558, 0.16098614037036896, 0.13868549466133118, 0.05973051115870476, -0.03966548293828964, 0.07866834104061127, -0.36422815918922424, -0.2929225564002991, 0.12223701924085617, 0.04758665710687637, 0.5287963151931763, 0.1722342073917389, 0.22847604751586914, 0.08470907807350159, 0.13575135171413422, -0.1003018170595169, -0.18807870149612427, -0.15801215171813965, -0.13643504679203033, 0.11905047297477722, 0.17177554965019226, 0.17103181779384613, -0.17962628602981567, -0.19826346635818481, -0.20716488361358643, -0.024118181318044662, 0.1777467578649521, -0.0709180161356926, -0.19365088641643524, 0.3347185254096985, -0.46812567114830017, -0.045132651925086975, 0.3891853094100952, 0.3572053015232086, 0.16981005668640137, 0.04913054406642914, -0.02882903628051281, 0.5353574752807617, -0.036852143704891205, -0.05114138126373291, -0.08615133166313171, 0.24315854907035828, -0.32708412408828735, 0.24757322669029236, -0.09059052914381027, -0.007488466799259186, -0.22181108593940735, -0.22819679975509644, -0.030339930206537247, -0.3078686594963074, 0.035737134516239166, -0.04483149200677872, 0.06395252794027328, 0.23791752755641937, -0.3148873746395111, 0.05522292107343674, -0.035948533564805984, -0.02913190796971321, 0.18411898612976074, 0.21877333521842957, 0.321713924407959, -0.11545760929584503, 0.6415478587150574, -0.04418723285198212, 0.014910969883203506, -0.18163026869297028, 0.2836940884590149, 0.19051030278205872, 0.168939471244812, 0.08880265802145004, 0.18636415898799896, 0.008465366438031197, 0.333029180765152, -0.06245899945497513, 0.22023767232894897, -0.048584818840026855, 0.3439480662345886, 0.5547138452529907, 0.07117709517478943, -0.35833030939102173, -0.0543023981153965, 0.38938528299331665, 0.14579077064990997, 0.12391214072704315, 0.13414651155471802, 0.007062092423439026, -0.07100224494934082, 0.18138790130615234, -0.6420041918754578, 0.05338055640459061, 0.031960733234882355, -0.35199853777885437, -0.3825598955154419, -0.09260974079370499, -0.5104484558105469, -0.062152255326509476, 0.14173433184623718, -0.007492981851100922, 0.0329425185918808, -0.07868874073028564, 0.023906812071800232, -0.20594541728496552, -0.05458962172269821, 0.22706550359725952, 0.1682695746421814, -0.023408660665154457, 0.3689236640930176, -0.5310978293418884, -0.21128693222999573, 0.1906963288784027, 0.24770760536193848, 0.20929521322250366, -0.03900149464607239, 0.4058581292629242, 0.11202459782361984, 0.14556202292442322, -0.070763498544693, -0.05981936678290367, 0.16103318333625793, 0.3544895052909851, 0.04199705272912979, 0.261322021484375, 0.09808698296546936, -0.09979218989610672, 0.14007270336151123, 0.04343094676733017, -0.29908275604248047, -0.21188074350357056, 0.3249717056751251, 0.1753530353307724, 0.18041372299194336, -0.02329987660050392, 0.375754714012146, -0.13017168641090393, -0.23831558227539062, 0.08112317323684692, 0.10263880342245102, -0.10894875228404999, -0.24176271259784698, 0.09944389015436172, -0.07116594910621643, 0.007066130638122559, 0.18419763445854187, -0.03055773675441742, 0.021134626120328903, -0.41406774520874023, -0.12865155935287476, 0.2505955696105957, 0.11373455822467804, -0.12676304578781128, 0.09774207323789597, 0.1597735583782196, 0.04668580740690231, -0.28360629081726074, -0.12935732305049896, -0.0501638762652874, 0.5626107454299927, -0.019448615610599518, -0.2112591415643692, 0.15099968016147614, 0.3584880828857422, -0.02639116160571575, -0.12816977500915527, -0.18867045640945435, -0.28754135966300964, 0.043898917734622955, -0.008159462362527847, -0.3241661787033081, 0.011525779962539673, 0.04369834437966347, 0.23319722712039948, 0.3404276669025421, 0.10100193321704865, 0.2775700092315674, -0.15534351766109467, -0.08501657843589783, -0.5489779710769653, -0.18481047451496124, -0.05637962371110916, 0.06647180765867233, -0.0956384539604187, 0.052344925701618195, -0.03479737415909767, 0.3438432514667511, -0.01012194063514471, 0.31706640124320984, -0.02081414684653282, -0.09819796681404114, -0.032694943249225616, 0.4011123776435852, -0.47488024830818176, -0.1552659571170807, -0.021511338651180267, -0.017494864761829376, 0.0346456840634346, -0.06058492511510849, -0.0785987451672554, -0.23899737000465393, 0.35251232981681824, -0.1608169674873352, -0.04476703703403473, -0.020304527133703232, 0.4243362843990326, 0.25574547052383423, 0.07693658024072647, -0.259465754032135, -0.01638060063123703, 0.21026991307735443, 0.1324426829814911, -0.1359894573688507, 0.2562759220600128, -0.15202341973781586, -0.025489356368780136, -0.2026941031217575, -0.10972937941551208, -0.11596860736608505, -0.15834075212478638, 0.11542041599750519, -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, 0.2715889811515808, -0.02540157362818718, -0.1736123263835907, 0.11133774369955063, -0.032490555197000504, 0.3496929109096527, 0.3046949505805969, 0.09639129042625427, 0.13277660310268402, 0.1502353847026825, 0.18139126896858215, -0.29633480310440063, 0.05446108430624008, 0.16967593133449554, 0.030123520642518997, -0.0018446296453475952, 0.23788097500801086, -0.10852649807929993, 0.26449400186538696, -0.1570073366165161, 0.02594122849404812, -0.08346939086914062, 0.034720588475465775, -0.007924170233309269, 0.14469146728515625, -0.15397818386554718, 0.01550845056772232, -0.37364181876182556, -0.9120110869407654, 0.4226647913455963, 0.08653847128152847, 0.18527039885520935, 0.4016510844230652, -0.00012305592827033252, -0.07805018872022629, 0.5848661065101624, -0.08126456290483475, -0.6614242792129517, -0.24752645194530487, 0.17996779084205627, -0.44016483426094055, 0.33093389868736267, -0.030891790986061096, 0.05642986297607422, -0.1428367644548416, 0.22567245364189148, 0.036392658948898315, 0.06255191564559937, 0.3939168453216553, 0.14157582819461823, 0.278430312871933, 0.1039668470621109, -0.22894807159900665, -0.11398161202669144, 0.06613485515117645, -0.13955555856227875, 0.5098291039466858, 0.33754754066467285, 0.08086451143026352, -0.02614521235227585, -0.1172662228345871, 0.07583940029144287, 0.48263734579086304, 0.1798248291015625, -0.02010524645447731, 0.45073410868644714, -0.026417814195156097, -0.19099846482276917, 0.000986546277999878, 0.8557760715484619, -0.557236909866333, -0.28589072823524475, 0.020645715296268463, 0.09727933257818222, -0.1694420576095581, 0.2176305055618286, -0.20579713582992554, -0.3246438503265381, 0.017004769295454025, -0.19486898183822632, -0.33706381916999817, -0.0902523398399353, 0.31063973903656006, -0.21952788531780243, 0.18331260979175568, 0.06803271919488907, 0.21227160096168518, 0.055997855961322784, -0.09097349643707275, 0.23839086294174194, -0.02267947793006897, -0.0061225928366184235, 0.25748273730278015, -0.08425585925579071, -0.24596454203128815, -0.31582510471343994, -0.20292413234710693, 0.1599837839603424, 0.05829679220914841, -0.4827140271663666, -0.1835443377494812, 0.04289053753018379, 0.21460548043251038, 0.11195938289165497, 0.31773316860198975, 0.34393370151519775, 0.284592866897583, 0.19506846368312836, 0.17445611953735352, 0.2424221932888031, 0.15372461080551147, -0.07111602276563644, -0.39698195457458496, 0.106655552983284, 0.08801091462373734, 0.24009372293949127, -0.28065356612205505, -0.1707746535539627, -0.26196417212486267, -0.3381372094154358, 0.009782221168279648, -0.07732273638248444, 0.2922813296318054, -0.09014897793531418, -0.23386314511299133, 0.049078308045864105, 0.44676899909973145, -0.19828662276268005, -0.1562919318675995, -0.027289625257253647, 0.10530099272727966, -0.21463525295257568, 0.049913570284843445, 0.3252117931842804, 0.014612821862101555, 0.4042748212814331, -0.24093686044216156, -0.1593516618013382, -0.046743668615818024, 0.10865804553031921, 0.05923198536038399, -0.2635221779346466, 0.30100688338279724, -0.07218009978532791, 0.0834098532795906, 0.048662979155778885, -0.27629342675209045, -0.21490874886512756, 0.2309892475605011, 0.06791456043720245, -0.495361328125, -0.023342696949839592, 0.11333990842103958, -0.5093955397605896, -0.059858813881874084, -0.06913221627473831, -0.16433435678482056, 0.10568083077669144, -0.39293548464775085, -0.03081464022397995, -0.43483680486679077, -0.37226438522338867, -0.11169140040874481, -0.03351198881864548, 0.29697728157043457, -0.3369503319263458, -0.2700841724872589, -0.4456767737865448, -0.38237226009368896, 0.21316760778427124, -0.2579703629016876, -0.07269936800003052, 0.20810380578041077, -0.41918885707855225, -0.02921367436647415, 0.5602774620056152, -0.5683714151382446, -0.3613015115261078, 0.4470440149307251, -0.41898679733276367, -0.25901299715042114, 0.18650424480438232, -0.0008598624845035374, 0.06650566309690475, -0.3295419216156006, -0.1430531144142151, 0.06461173295974731, 0.042373839765787125, -0.13622698187828064, -0.11024289578199387, -0.2676573097705841, 0.19914427399635315, 0.2632276713848114, -0.09678244590759277, 0.13125520944595337, 0.14763687551021576, 0.10275794565677643, 0.32669585943222046, 0.14259028434753418, 0.1718102991580963, -0.08835853636264801, 0.23017264902591705, -0.2256695181131363, -0.1406850665807724, -0.10932890325784683, -0.6099806427955627, 0.23831871151924133, -0.06623587012290955, 0.20779454708099365, -0.14170894026756287, 0.023358382284641266, -0.6258837580680847, 0.01587006263434887, -0.059171970933675766, -0.22863852977752686, 0.006851989775896072, 0.1909550577402115, 0.11202706396579742, 0.018214717507362366, 0.03710780665278435, 0.0626564770936966, -0.2197256088256836, 0.1759641319513321, 0.010265050455927849, -0.0012474823743104935, -0.1033817008137703, 0.03828231245279312, 0.23613706231117249, 0.22019842267036438, 0.08650315552949905, -0.3108551800251007, -0.0424867607653141, 0.28622928261756897, 0.2748509347438812, 0.2039646953344345, 0.1929638385772705, -0.07467357814311981, 0.13896968960762024, 0.153835266828537, 0.17406843602657318, 0.21679812669754028, 0.0869944840669632, -0.2102239429950714, 0.08055438101291656, 0.11069171875715256, 0.19249221682548523, -0.02169358730316162, -0.17491629719734192, 0.026053832843899727, 0.4049431383609772, 0.12913748621940613, -0.02623719535768032, -0.38228338956832886, 0.100460484623909, 0.29537445306777954, 0.6366948485374451, -0.047104500234127045, 0.36636239290237427, -0.10841915756464005, -0.19691750407218933, -0.34861764311790466, 0.008201133459806442, 0.1602071225643158, -0.15804868936538696, -0.09225586801767349, -0.005740595981478691, 0.2688911557197571, 0.8805612921714783, -0.02649056911468506, 0.17164528369903564, 0.14539217948913574, -0.11143910884857178, -0.2071063071489334, 0.18835212290287018, 0.07700619101524353, 0.0482562780380249, 0.14204102754592896, -0.014179175719618797, -0.38311082124710083, -0.08877219259738922, -0.053082115948200226, 0.45894408226013184, 0.4843933582305908, -0.3677518665790558, -0.19629822671413422, -0.24734243750572205, -0.3671186566352844, -0.41309159994125366, -0.32595205307006836, -0.15261602401733398, -0.3070302903652191, -0.12076395750045776, 0.276625394821167, 0.13867108523845673, 0.22709307074546814, -0.2515861988067627, 0.18580688536167145, -0.2386370152235031, -0.31146034598350525, 0.09593987464904785, -0.17135804891586304, -0.34025344252586365, 0.030635088682174683, 0.2242274135351181, 0.031321022659540176, 0.41515225172042847, -0.14686699211597443, -0.2182319313287735, -0.25129908323287964, -0.24454772472381592, 0.12566012144088745, 0.08897501230239868, 0.5211694836616516, 0.45991751551628113, -0.027609139680862427, 0.45716768503189087, 0.11719471216201782, 0.03460448607802391, -0.2661333382129669, -0.46694308519363403, -0.1576690822839737, -0.01980411261320114, 0.08315401524305344, -0.03829817846417427, -0.4752780497074127, -0.5595124959945679, -0.07798320055007935, 0.017075933516025543, 0.10753892362117767, 0.02640506625175476, 0.044131919741630554, -0.09966827183961868, 0.2051914483308792, 0.02571728453040123, -0.11663936078548431, -0.14416952431201935, 0.0730145275592804, 0.020061299204826355, -0.1757686883211136, -0.17963053286075592, 0.11319698393344879, -0.06287625432014465, 0.08196168392896652, -0.15483397245407104, -0.4178168773651123, -0.3486687242984772, 0.07465767115354538, 0.18385706841945648, 0.28053760528564453, -0.2723371088504791, 0.2895950376987457, 0.24983292818069458, 0.14815521240234375, -0.12449968606233597, -0.3925282061100006, 0.1562221199274063, 0.019437197595834732, 0.25050416588783264, 0.14687776565551758, 0.48256438970565796, -0.19213886559009552, 0.750655472278595, 0.11803838610649109, -0.16311019659042358, 0.02059563621878624, -0.2556125521659851, 0.04147198796272278, -0.053950242698192596, -0.12413864582777023, 0.31781071424484253, -0.13975435495376587, -0.24665479362010956, 0.195745587348938, -0.22311368584632874, -0.05246236175298691, -0.09552216529846191, 0.22133150696754456, -0.06543809175491333, -0.025436192750930786, -0.053477056324481964, -0.20936018228530884, 0.05761290341615677, 0.037889156490564346, -0.07096286118030548, -0.194407656788826, -0.21401989459991455, 0.004863161593675613, 0.3135368824005127, 0.004929304122924805, 0.1602725237607956, -0.300514817237854, 0.01857445202767849, -0.44620150327682495, 0.30397462844848633, 0.2712736427783966, 0.5984058380126953, 0.08295075595378876, 0.07675129175186157, 0.12854966521263123, 0.10277162492275238, 0.21404898166656494, -0.028820769861340523, 0.12243359535932541, 0.2516896724700928, -0.13396108150482178, -0.08915548026561737, -0.24068519473075867, -0.06588228046894073, 0.212519109249115, -0.029730867594480515, 0.09589413553476334, -0.3500882387161255, -0.08926252275705338, 0.26529330015182495, 0.14095553755760193, -0.08590082079172134, -0.3551112711429596, -0.12330791354179382, -0.0726567730307579, -0.15032236278057098, -0.2127329707145691, -0.17580492794513702, 0.10044097900390625, -0.12927871942520142, 0.28205567598342896, 0.0149540388956666, -0.09018874168395996, -0.08661633729934692, 0.11567696928977966, 0.14001023769378662, -0.1844521313905716, 0.3376268446445465, 0.17730818688869476, 0.04688061773777008, 0.4023391008377075, 0.3133348226547241, -0.10161547362804413, -0.2746429443359375, 0.026888955384492874, -0.006902131251990795, 0.3866167962551117, 0.12779274582862854, 0.1136166974902153, 0.0418584905564785, -0.2697530686855316, -0.14895150065422058, -0.298460453748703, 0.1984386295080185, 0.38250574469566345, -0.09436509013175964, -0.05138221010565758, -0.3670404255390167, 0.6012394428253174, 0.2057901918888092, -0.15702910721302032, 0.4772726893424988, -0.25194334983825684, -0.2803328335285187, -0.055179085582494736, 0.07219109684228897, 0.8358827233314514, 0.12051726132631302, 0.16542378067970276, 0.37753602862358093, -0.26247894763946533, 0.3626871705055237, 0.0903792679309845, 0.16148532927036285, -0.5544987916946411, -0.08480050414800644, 0.024787774309515953, -0.23059780895709991, 0.24379867315292358, -0.1052333265542984, -0.23620760440826416, 0.37085607647895813, 0.04794023931026459, 0.2831416130065918, 0.2545764744281769, 0.3559746742248535, -0.11178473383188248, -0.0445249006152153, -0.10924947261810303, 0.10165834426879883, -0.16587114334106445, 0.40507522225379944, -0.2347889542579651, 0.08953969180583954, -0.0864686518907547, -0.012371841818094254, -0.2658649981021881, -0.11775972694158554, -0.14358939230442047, 0.06133344769477844, 0.07406353950500488, -0.1460016369819641, 0.10794614255428314, 0.19950461387634277, 0.1334606111049652, -0.16214729845523834, -0.16512370109558105, 0.15120500326156616, -0.15446165204048157, 0.10339847207069397, 0.33099159598350525, 0.05736219137907028, 0.23977355659008026, -0.3374669849872589, -0.1727195680141449, 0.13349249958992004, -0.23813071846961975, -0.1332811564207077, -0.09552429616451263, -0.1865772157907486, 0.10564965009689331, -0.3507993221282959, -0.32004013657569885, -0.14020805060863495, 0.0729285180568695, -0.1441505253314972, 0.0400049202144146, 0.27288639545440674, -0.3790874183177948, -0.006065584719181061, -0.12708011269569397, -0.30062317848205566, 0.01657910645008087, 0.4735439121723175, -0.061930589377880096, -0.1693362593650818, 0.7641564607620239, -0.17778785526752472, 0.06867421418428421, -0.1367506980895996, 0.06282451003789902, 0.0116067910566926, -0.13863560557365417, 0.0257415808737278, -0.04889092221856117, -0.18072542548179626, 0.10184190422296524, 0.5352264642715454, 0.33291196823120117, 0.00482177734375, 0.00259973481297493, -0.4617112874984741, -0.44225797057151794, -0.09019428491592407, -0.1543557196855545, 0.1846054494380951, -0.07652467489242554, 0.13444074988365173, 0.2172488421201706, 0.10431493073701859, -0.20098242163658142, -0.17669543623924255, 0.0075926464051008224, 0.24287089705467224, -0.1265757530927658, 0.0439223013818264, -0.02672361396253109, 0.0873129814863205, 0.031429849565029144, -0.027036983519792557, -0.3348514139652252, -0.07218397408723831, -0.06499525904655457, 0.2020931839942932, 0.09678614139556885, -0.028630653396248817, 0.07220377027988434, -0.15105386078357697, -0.02545313909649849, 0.1781972348690033, -0.026853755116462708, 0.08564923703670502, -0.09723049402236938, 0.15172505378723145, 0.32388049364089966, -0.07167687267065048, -0.05003277212381363, 0.21732322871685028, 0.05624870955944061, 0.1620158553123474, -0.06300629675388336, 0.2864287495613098, -0.13696002960205078, 0.0546017587184906, 0.20251110196113586, 0.10919035971164703, 0.00827055424451828, 0.25383976101875305, 0.11799740791320801, -0.1682087481021881, 0.14139583706855774, -0.04990189149975777, 0.355796217918396, 0.7263289093971252, -0.2195245325565338, -0.12481633573770523, 0.4329032003879547, 0.05636553466320038, -0.25470447540283203, -0.04399612173438072, 0.30986154079437256, 0.13776184618473053, 0.0462355799973011, 0.21163588762283325, 0.0015202611684799194, -0.03864777460694313, 0.16678792238235474, 0.13611534237861633, 0.1286683976650238, -0.2262921780347824, -0.059038933366537094, 0.4483218193054199, -0.2203175276517868, 0.16248925030231476, 0.3900028169155121, 0.06479001045227051, 0.21330344676971436, 0.2186937928199768, 0.03703268989920616, 0.6142652630805969, 0.05644655600190163, 0.01788436621427536, 0.08402562141418457, -0.04976402968168259, -0.0006030946969985962, 0.27663666009902954, -0.030039355158805847, -0.08608447760343552, 0.32754379510879517, 0.554908037185669, -0.35028696060180664, -0.15907345712184906, -0.26408180594444275, -0.08660198748111725, 0.031998004764318466, -0.1442386656999588, -0.285094678401947, 0.08155801147222519, -0.21613916754722595, -0.13049781322479248, -0.118877112865448, -0.309390127658844, 0.16270625591278076, 0.017042867839336395, -0.3983369469642639, -0.24440379440784454, -0.2909781038761139, 0.1936606466770172, -0.10940318554639816, -0.269906222820282, 0.22087472677230835, -0.013040043413639069, 0.0328560397028923, 0.29115331172943115, 0.32487064599990845, 0.46994632482528687, 0.1625407487154007, -0.31814834475517273, -0.2325119972229004, -0.1214059591293335, 0.09721105545759201, 0.0350799523293972, 0.3806981146335602, -0.04988767206668854, -0.10703162848949432, 0.46314841508865356, 0.0726584866642952, 0.03713392838835716, 0.21864308416843414, 0.43211278319358826, -0.4861971139907837, -0.17150580883026123, 0.7934107780456543, -0.07552888989448547, -0.14116674661636353, -0.12957748770713806, 0.25525903701782227, -0.26042264699935913, 0.1343466192483902, 0.09955894947052002, -0.07344657927751541, 0.1301887035369873, 0.025473061949014664, 0.03947754204273224, -0.1064056009054184, 0.3546757698059082, 0.31101417541503906, -0.04381667822599411, -0.3694727420806885, -0.2513991594314575, -0.7986485362052917, 0.22746261954307556, -0.11356881260871887, -0.3236978054046631, -0.06485424935817719, -0.2603072226047516, 0.1893695592880249, 0.04811112582683563, -0.14149154722690582, 0.19518303871154785, 0.054870352149009705, 0.13517674803733826, -0.1899448037147522, -0.31790873408317566, -0.2887229323387146, -0.023948533460497856, -0.10388312488794327, -0.38692671060562134, 0.11374451965093613, -0.33049675822257996, -0.07958605885505676, 0.11841166019439697, -0.00978144258260727, 0.4779796600341797, -0.005791578441858292, 0.3654896914958954, -0.11248210072517395, 0.7752768993377686, 0.18795090913772583, 0.06328277289867401, -0.1649438440799713, 0.12217646837234497, -0.1206473559141159, 0.13483703136444092, -0.0703357458114624, 0.10441069304943085, -0.11896272748708725, 0.5212554335594177, -0.44372111558914185, 0.4605461657047272, -0.028103414922952652, 0.07408122718334198, -0.3021467626094818, 0.045647986233234406, -0.25546106696128845, 0.12183534353971481, 0.3444596230983734, 0.14114543795585632, -0.09161031246185303, 0.3550700843334198, -0.21494290232658386, -0.19039209187030792, 0.3962547779083252, -0.10348934680223465, -0.0939866304397583, 0.021515648812055588, 0.26462745666503906, -0.05172332376241684, -0.14009112119674683, -0.4393866956233978, 0.19632798433303833, 0.2907601296901703, -0.09371230006217957, 0.11838948726654053, 0.2602345049381256, 0.2837517559528351, 0.13402226567268372, 0.01832754909992218, 0.9827132225036621, -0.05974341928958893, 0.17604415118694305, -0.10538431257009506, -0.2278595268726349 ]
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
[ -0.19621720910072327, 0.4000372290611267, -0.031468600034713745, -0.16441306471824646, 0.04428334906697273, -0.13329291343688965, 0.27948498725891113, 0.37617409229278564, 0.009489316493272781, 0.08292682468891144, 0.2064318209886551, 0.15806886553764343, -0.300568163394928, 0.06804143637418747, 0.13692401349544525, 0.001200944185256958, -0.056462906301021576, 0.16989606618881226, -0.09795263409614563, 0.21250595152378082, -0.12722066044807434, 0.07482471317052841, 0.02045775204896927, 0.10806221514940262, -0.027999823912978172, 0.08875592052936554, -0.21589478850364685, 0.03439011424779892, -0.38094663619995117, -0.785807728767395, 0.38182637095451355, 0.07057259231805801, 0.18520070612430573, 0.3587058484554291, -0.00012268955470062792, -0.08625639975070953, 0.5880786776542664, -0.056077003479003906, -0.7038208842277527, -0.2316201776266098, 0.24540674686431885, -0.4804558753967285, 0.3018532991409302, -0.021508552134037018, 0.1885167360305786, -0.13163866102695465, 0.244207963347435, 0.10435053706169128, 0.16175124049186707, 0.32244858145713806, 0.11873665452003479, 0.2776326537132263, 0.05146857723593712, -0.17972375452518463, -0.10094422847032547, 0.08600423485040665, -0.089267797768116, 0.5028809905052185, 0.36486244201660156, 0.031120793893933296, -0.11236346513032913, -0.13111886382102966, 0.09226366877555847, 0.49313563108444214, 0.1250217854976654, -0.022293921560049057, 0.5011737942695618, 0.019112227484583855, -0.21289393305778503, 0.041048698127269745, 0.7716661095619202, -0.5064082145690918, -0.37548553943634033, -0.07860944420099258, 0.0564955398440361, -0.13751456141471863, 0.2607557475566864, -0.1600874811410904, -0.3946448266506195, 0.019819922745227814, -0.1663755178451538, -0.28037229180336, -0.06712237000465393, 0.29351305961608887, -0.2830924987792969, 0.2682918310165405, 0.07774843275547028, 0.17557892203330994, 0.0895024761557579, -0.1008504182100296, 0.25734415650367737, -0.061836645007133484, 0.024405064061284065, 0.23488165438175201, -0.030216123908758163, -0.2238750010728836, -0.26101967692375183, -0.22214099764823914, 0.19221365451812744, 0.08718404173851013, -0.4626269042491913, -0.1906760036945343, -0.031207039952278137, 0.14959892630577087, 0.1416134089231491, 0.4171294867992401, 0.33842408657073975, 0.2389010190963745, 0.24328817427158356, 0.16279183328151703, 0.2434389591217041, 0.2634248733520508, -0.059813324362039566, -0.3704584240913391, -0.032496556639671326, 0.09223347902297974, 0.14825278520584106, -0.32870444655418396, -0.15333986282348633, -0.2270713895559311, -0.38531753420829773, -0.046314116567373276, -0.05492560938000679, 0.23764824867248535, -0.10888989269733429, -0.14659425616264343, 0.04901201277971268, 0.41204380989074707, -0.20630092918872833, -0.15113496780395508, -0.02948884107172489, 0.0445263497531414, -0.19153976440429688, -0.012614157050848007, 0.30864521861076355, 0.055757902562618256, 0.37803590297698975, -0.36712178587913513, -0.1176607757806778, -0.03431224077939987, 0.11314693093299866, 0.1852354258298874, -0.26896554231643677, 0.2899591028690338, -0.15554548799991608, 0.05203224718570709, 0.07850468158721924, -0.2927016317844391, -0.22106757760047913, 0.16356539726257324, 0.12039398401975632, -0.45691055059432983, 0.03789440914988518, 0.0930616483092308, -0.6060149073600769, -0.0646645650267601, 0.012726657092571259, -0.11572961509227753, 0.09138727188110352, -0.3326219618320465, 0.026575684547424316, -0.39136946201324463, -0.2052660435438156, -0.06727050244808197, -0.0997258797287941, 0.2391044795513153, -0.2667432725429535, -0.33723169565200806, -0.44814231991767883, -0.32592347264289856, 0.12920187413692474, -0.22480536997318268, -0.07712345570325851, 0.21498161554336548, -0.37946203351020813, -0.00851573795080185, 0.5497831702232361, -0.521664023399353, -0.3586077094078064, 0.4454956650733948, -0.42921698093414307, -0.2467309534549713, 0.234197199344635, -0.048981454223394394, -0.002735354006290436, -0.36214715242385864, -0.08949658274650574, 0.08227436989545822, 0.04962074011564255, -0.17818878591060638, -0.18848390877246857, -0.34391722083091736, 0.11360957473516464, 0.22009776532649994, -0.03988174721598625, 0.08876357972621918, 0.09676428884267807, 0.05409690737724304, 0.35139429569244385, 0.11439789831638336, 0.1550447791814804, -0.14401833713054657, 0.17716747522354126, -0.23787619173526764, -0.17119497060775757, -0.0517728291451931, -0.5992653965950012, 0.2523059844970703, -0.11843126267194748, 0.21302318572998047, -0.046452537178993225, -0.011181812733411789, -0.5071437358856201, 0.0033057983964681625, 0.019839856773614883, -0.2721046805381775, 0.002070598304271698, 0.2044222056865692, 0.12961789965629578, -0.04795825853943825, -0.010605555027723312, 0.03222370520234108, -0.21072933077812195, 0.19253289699554443, -0.009809723123908043, -0.020285815000534058, -0.046190306544303894, 0.06240805611014366, 0.24959459900856018, 0.2555881440639496, 0.05471888929605484, -0.3409825265407562, 0.007953019812703133, 0.2820030152797699, 0.20832470059394836, 0.18946516513824463, 0.18919211626052856, -0.07011069357395172, 0.17034898698329926, 0.26392462849617004, 0.16457989811897278, 0.2629548907279968, -0.015182099305093288, -0.2123050093650818, -0.08616406470537186, 0.24468600749969482, 0.1680406779050827, 0.04419178515672684, -0.1283629685640335, 0.06571394205093384, 0.3238115608692169, 0.11381283402442932, -0.13485538959503174, -0.22805827856063843, 0.19604414701461792, 0.28523752093315125, 0.6019864678382874, -0.050456494092941284, 0.3259407579898834, -0.04792296141386032, -0.2781184911727905, -0.3706555962562561, 0.0179365836083889, 0.11389971524477005, -0.09621275216341019, -0.05627509206533432, 0.09414380043745041, 0.36090919375419617, 0.8490374088287354, -0.0392586886882782, 0.270916223526001, 0.08667962998151779, -0.18442897498607635, -0.231657937169075, 0.15094077587127686, 0.11443627625703812, -0.026465723291039467, 0.15560941398143768, 0.0009208349511027336, -0.401189923286438, -0.08902250230312347, -0.00013074278831481934, 0.4331216812133789, 0.39993196725845337, -0.3471737802028656, -0.10980457067489624, -0.3021012246608734, -0.44454970955848694, -0.39858970046043396, -0.36305147409439087, -0.2080814093351364, -0.3813641369342804, -0.061455756425857544, 0.22124451398849487, -0.00008041411638259888, 0.23552905023097992, -0.19494938850402832, 0.28329282999038696, -0.30940207839012146, -0.3343961834907532, 0.1164737343788147, -0.2011948674917221, -0.40045812726020813, 0.03807687386870384, 0.28132164478302, 0.0704876109957695, 0.4426933228969574, -0.1078203022480011, -0.29736101627349854, -0.21326030790805817, -0.31885039806365967, 0.13285496830940247, 0.06095752865076065, 0.5699266791343689, 0.4761795103549957, -0.014883913099765778, 0.4204102158546448, 0.07882925122976303, 0.042044125497341156, -0.2335488647222519, -0.4335296154022217, -0.2341788411140442, 0.09182915836572647, 0.0969143807888031, -0.13966424763202667, -0.39729323983192444, -0.52669358253479, -0.08559001982212067, 0.1703435778617859, 0.12703530490398407, 0.07320810854434967, 0.0939830020070076, -0.11142731457948685, 0.2511827349662781, 0.08683265745639801, -0.18744917213916779, -0.11685249209403992, 0.07108868658542633, -0.009083796292543411, -0.1429181545972824, -0.1871635913848877, 0.0666741281747818, -0.11608918011188507, 0.08681826293468475, -0.1572064608335495, -0.48865634202957153, -0.44232651591300964, 0.1542486846446991, 0.23099470138549805, 0.39259210228919983, -0.20678259432315826, 0.2817327082157135, 0.201801598072052, 0.12314780056476593, -0.10951663553714752, -0.34012529253959656, 0.1602247953414917, 0.09377153217792511, 0.3110189139842987, 0.13703428208827972, 0.34237098693847656, -0.11723089218139648, 0.7626632452011108, 0.11879229545593262, -0.0941152274608612, -0.03337521851062775, -0.1344081610441208, 0.015389334410429, -0.017840825021266937, -0.04059571772813797, 0.33505961298942566, -0.20015400648117065, -0.3104262351989746, 0.20721706748008728, -0.11684448271989822, -0.03705424815416336, -0.13537587225437164, 0.2891801595687866, -0.017412874847650528, -0.06730526685714722, 0.0038466546684503555, -0.21916741132736206, 0.008076868951320648, 0.09175167977809906, -0.02824871614575386, -0.19920854270458221, -0.15857353806495667, 0.031044337898492813, 0.3056078553199768, 0.009298950433731079, 0.17336422204971313, -0.25022339820861816, 0.006671939045190811, -0.3804589807987213, 0.32095175981521606, 0.3083265721797943, 0.4791111648082733, 0.08994166553020477, 0.04050246626138687, 0.128177210688591, 0.023768838495016098, 0.2255096435546875, 0.07952557504177094, 0.1641315370798111, 0.25303009152412415, -0.13103324174880981, -0.04519321769475937, -0.24799959361553192, -0.06108873710036278, 0.0637972354888916, 0.004301479086279869, 0.09554530680179596, -0.40971314907073975, -0.09873314946889877, 0.1882554143667221, 0.08911511301994324, -0.07824192941188812, -0.23516100645065308, -0.0749972015619278, -0.08631077408790588, -0.11132027953863144, -0.21148717403411865, -0.22183217108249664, 0.01701824925839901, -0.21502210199832916, 0.4063695967197418, -0.0038536395877599716, -0.08811713010072708, -0.06015479564666748, 0.09694615006446838, 0.08987809717655182, -0.28205639123916626, 0.29240572452545166, 0.2651258111000061, 0.07484520971775055, 0.3037792146205902, 0.31530141830444336, -0.09081141650676727, -0.2188895046710968, 0.025368649512529373, 0.055203232914209366, 0.4211719334125519, 0.35064882040023804, 0.10434022545814514, -0.05036770552396774, -0.23994183540344238, -0.20737259089946747, -0.23432552814483643, 0.19697564840316772, 0.2857738137245178, -0.030911825597286224, 0.0785093754529953, -0.31849727034568787, 0.7194925546646118, 0.22033126652240753, -0.2501452565193176, 0.4313562512397766, -0.2180648297071457, -0.26280924677848816, -0.06578311324119568, 0.12758301198482513, 0.9033367037773132, 0.14794975519180298, 0.23772436380386353, 0.4199579656124115, -0.26912355422973633, 0.4643268585205078, 0.07282888889312744, 0.15510977804660797, -0.49473056197166443, -0.00016525853425264359, 0.01968093402683735, -0.1823749989271164, 0.20952045917510986, -0.09575746953487396, -0.2247014343738556, 0.3749465048313141, 0.1465231329202652, 0.28106725215911865, 0.3701074421405792, 0.3088003396987915, -0.2517704367637634, -0.0336604006588459, -0.1858089119195938, 0.09027861803770065, -0.1498020440340042, 0.40192827582359314, -0.28164350986480713, -0.00031447969377040863, -0.14423677325248718, -0.044248081743717194, -0.2529454231262207, -0.11223725229501724, -0.173439621925354, 0.06214238703250885, 0.15585285425186157, -0.13206005096435547, -0.06616057455539703, 0.18333186209201813, 0.08269128203392029, -0.21432547271251678, -0.14978264272212982, 0.1254417449235916, -0.14331692457199097, 0.16847188770771027, 0.3818264901638031, 0.04532884433865547, 0.3266820013523102, -0.3169657289981842, -0.05816731974482536, 0.06710837036371231, -0.2803245484828949, -0.12472721934318542, -0.13294482231140137, -0.17246964573860168, 0.14165642857551575, -0.29994386434555054, -0.1425621658563614, -0.10251238942146301, -0.009897671639919281, -0.115902841091156, 0.02239544875919819, 0.28106242418289185, -0.3552970886230469, -0.061010830104351044, -0.11896087229251862, -0.22990930080413818, 0.009612835943698883, 0.4747704863548279, -0.06723793596029282, -0.1376439929008484, 0.6765257120132446, -0.2103765457868576, 0.06588243693113327, -0.12991788983345032, 0.062012724578380585, -0.10107389092445374, -0.11184273660182953, 0.07678307592868805, -0.16967424750328064, -0.3008666932582855, 0.04371436685323715, 0.5850268006324768, 0.4418943524360657, 0.012407097965478897, -0.07040699571371078, -0.46289244294166565, -0.41166603565216064, -0.0003568902611732483, -0.14149731397628784, 0.20699253678321838, -0.2121647298336029, 0.22073912620544434, 0.16328157484531403, 0.046032145619392395, -0.18724983930587769, -0.2017056792974472, 0.049464620649814606, 0.20594938099384308, -0.1384763866662979, 0.04342859610915184, -0.015562606044113636, 0.14493200182914734, -0.03392532095313072, 0.06349072605371475, -0.36153310537338257, -0.07694461941719055, -0.10113637149333954, 0.2054293155670166, 0.11184283345937729, -0.03894909471273422, 0.15560282766819, -0.11734654009342194, 0.03972204774618149, 0.2143976390361786, -0.06621687114238739, -0.048358578234910965, -0.09268885850906372, 0.1126721203327179, 0.34582823514938354, 0.09795041382312775, -0.006950933486223221, 0.30149030685424805, 0.10003826022148132, 0.17635124921798706, -0.11613377183675766, 0.2669965922832489, -0.2339322417974472, 0.08855469524860382, 0.20248861610889435, 0.12920896708965302, -0.019432008266448975, 0.2600127160549164, 0.13960126042366028, -0.1141728013753891, 0.08041916787624359, -0.1625291109085083, 0.3586396276950836, 0.8415080308914185, -0.17826029658317566, -0.13109177350997925, 0.4703673720359802, 0.03272813931107521, -0.19339068233966827, 0.027116749435663223, 0.22767898440361023, 0.20254075527191162, 0.039652761071920395, 0.26613757014274597, 0.010690603405237198, -0.13747431337833405, 0.19611197710037231, 0.1407437026500702, 0.09474238008260727, -0.17869940400123596, -0.07266853749752045, 0.5383134484291077, -0.2553038001060486, 0.16207416355609894, 0.3502659797668457, 0.00047488510608673096, 0.21568593382835388, 0.08506079018115997, 0.025451581925153732, 0.5338096022605896, -0.012156344950199127, 0.058235470205545425, 0.11353041231632233, 0.019126318395137787, -0.04424290731549263, 0.2593919038772583, 0.05144317448139191, -0.011776052415370941, 0.3216552138328552, 0.6301924586296082, -0.43068036437034607, -0.17472785711288452, -0.27786174416542053, -0.0494084469974041, 0.08228203654289246, -0.14998453855514526, -0.3964492380619049, 0.0435982421040535, -0.2213098108768463, -0.07336217910051346, -0.14795054495334625, -0.3185235857963562, 0.2299106866121292, 0.00329657644033432, -0.35559141635894775, -0.3461935818195343, -0.39676228165626526, 0.1408468633890152, -0.03580810874700546, -0.27528682351112366, 0.2508656084537506, 0.029407210648059845, 0.018531780689954758, 0.2829647660255432, 0.28391745686531067, 0.4441567063331604, 0.08628561347723007, -0.43873557448387146, -0.2540844678878784, -0.19133098423480988, 0.13583232462406158, -0.07762150466442108, 0.399572491645813, -0.059817150235176086, -0.16802312433719635, 0.49920526146888733, 0.05886781960725784, 0.043532341718673706, 0.26643142104148865, 0.3724864423274994, -0.5206148624420166, -0.16898134350776672, 0.82081538438797, -0.06794869899749756, -0.16675175726413727, -0.027103736996650696, 0.17767956852912903, -0.23198482394218445, 0.08259600400924683, -0.03570432588458061, -0.04749995470046997, 0.18776077032089233, 0.11940518021583557, 0.04107680171728134, -0.1780635267496109, 0.3992721140384674, 0.33888137340545654, -0.07150008529424667, -0.3425288796424866, -0.21393299102783203, -0.8131759166717529, 0.08850471675395966, -0.10057628154754639, -0.3737140893936157, -0.06355031579732895, -0.23741291463375092, 0.25356799364089966, 0.10478387773036957, -0.257435142993927, 0.17715543508529663, -0.0015824269503355026, 0.1259436309337616, -0.2627962529659271, -0.36598822474479675, -0.2140856832265854, -0.025764482095837593, -0.02974064275622368, -0.402548223733902, 0.07842466235160828, -0.3133713901042938, -0.09926487505435944, 0.22371746599674225, -0.03939526528120041, 0.49409544467926025, 0.0730830579996109, 0.25377756357192993, -0.012197377160191536, 0.6868562698364258, 0.24270299077033997, -0.03130260854959488, -0.24734267592430115, 0.17876696586608887, -0.13330836594104767, 0.1253570318222046, -0.10701526701450348, 0.13508643209934235, -0.17025792598724365, 0.6143170595169067, -0.414183109998703, 0.4506014883518219, 0.0749511867761612, 0.0781891867518425, -0.2869768440723419, 0.047970786690711975, -0.3413870632648468, 0.14387987554073334, 0.3280654847621918, 0.30497345328330994, -0.11956609785556793, 0.34439802169799805, -0.26570069789886475, -0.15193809568881989, 0.4117482304573059, -0.24894636869430542, -0.05838599056005478, 0.05723504349589348, 0.21902164816856384, 0.0009547099471092224, -0.15867112576961517, -0.44349995255470276, 0.12104444950819016, 0.2671409547328949, -0.13255095481872559, 0.0848160833120346, 0.1458553820848465, 0.2846495807170868, 0.026441995054483414, -0.035740792751312256, 0.9360103607177734, -0.07236693054437637, 0.15447020530700684, -0.018570510670542717, -0.24318110942840576 ]
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?
[ -0.10087041556835175, 0.5024332404136658, -0.039797212928533554, -0.08978384733200073, 0.0968024805188179, -0.12240689992904663, 0.2872871160507202, 0.31355321407318115, -0.054385315626859665, 0.11574174463748932, 0.04618164151906967, 0.16493384540081024, -0.2744549512863159, -0.005963634699583054, 0.165956512093544, -0.0370168462395668, -0.07802139967679977, 0.25511327385902405, -0.09114491939544678, 0.2149970978498459, -0.23069195449352264, 0.12446141242980957, -0.0709371566772461, 0.07457991689443588, 0.1365300565958023, 0.19969812035560608, -0.16580642759799957, 0.05431877821683884, -0.31603482365608215, -0.8764681220054626, 0.3479219675064087, 0.18099507689476013, 0.24238255620002747, 0.23878884315490723, -0.00012480560690164566, -0.153802752494812, 0.5954884886741638, -0.12650670111179352, -0.6724078059196472, -0.1728314608335495, 0.24162229895591736, -0.4436880946159363, 0.2643413245677948, -0.0716199055314064, 0.2790188491344452, -0.16837634146213531, 0.18214735388755798, 0.11028318107128143, 0.07405327260494232, 0.38633590936660767, 0.11946217715740204, 0.24075785279273987, 0.08585745096206665, -0.12019345164299011, -0.05844981223344803, 0.024083450436592102, -0.0008228309452533722, 0.6254268884658813, 0.40664759278297424, 0.021436626091599464, 0.021082788705825806, -0.010083932429552078, -0.06172134727239609, 0.45013153553009033, 0.24951234459877014, -0.04781989008188248, 0.49971404671669006, -0.07006866484880447, -0.1631966531276703, -0.013614587485790253, 0.6971592307090759, -0.5866325497627258, -0.36203816533088684, -0.05475027114152908, 0.15433429181575775, -0.20426039397716522, 0.1349426805973053, -0.28096291422843933, -0.43949681520462036, 0.058052726089954376, -0.10192982852458954, -0.5488121509552002, 0.017848558723926544, 0.2542361915111542, -0.3011972904205322, 0.28678974509239197, 0.11398311704397202, 0.1454489380121231, 0.24242769181728363, -0.05278812348842621, 0.21367669105529785, -0.022138364613056183, -0.08245567977428436, 0.16982369124889374, 0.0019519105553627014, -0.304499089717865, -0.31498029828071594, -0.21552741527557373, 0.09691528230905533, 0.13492155075073242, -0.37371471524238586, -0.14380395412445068, -0.05633974075317383, 0.1341736614704132, 0.09189127385616302, 0.41653361916542053, 0.3381829857826233, 0.22063644230365753, 0.10820474475622177, 0.2282303422689438, 0.23546725511550903, 0.12068722397089005, 0.03867356479167938, -0.3611081838607788, 0.019834887236356735, 0.1914500594139099, 0.1347879022359848, -0.19446946680545807, -0.08906851708889008, -0.24023905396461487, -0.4562418460845947, -0.057777341455221176, -0.17793463170528412, 0.20956367254257202, -0.08753730356693268, -0.08982730656862259, 0.10732926428318024, 0.40889471769332886, -0.12385638058185577, -0.2714741826057434, 0.003222732339054346, 0.19675253331661224, -0.098024383187294, 0.1111707016825676, 0.21871055662631989, 0.02494201436638832, 0.33839648962020874, -0.37400129437446594, -0.034397415816783905, -0.0379936546087265, 0.10087799280881882, -0.0139267323538661, -0.17526453733444214, 0.2979564368724823, -0.13329315185546875, 0.03545448184013367, 0.020310809835791588, -0.35186827182769775, -0.2085862159729004, 0.24709346890449524, 0.22873249650001526, -0.4571879804134369, -0.09473897516727448, 0.05991044268012047, -0.5005056262016296, -0.02239135652780533, -0.05310019105672836, -0.07679609954357147, -0.07516729086637497, -0.5091546773910522, 0.15669900178909302, -0.36302250623703003, -0.28936347365379333, -0.05024613067507744, -0.059419915080070496, 0.37651246786117554, -0.2985680401325226, -0.2998124957084656, -0.540572464466095, -0.3079770803451538, 0.1497212052345276, -0.24499936401844025, -0.16400101780891418, 0.2662373483181, -0.34657225012779236, 0.04316511005163193, 0.667220413684845, -0.6826150417327881, -0.38173288106918335, 0.3879488408565521, -0.29687464237213135, -0.2016325145959854, 0.13723118603229523, -0.08174246549606323, 0.07141735404729843, -0.3140597343444824, -0.20285391807556152, -0.005927072837948799, 0.0915985181927681, -0.21299782395362854, -0.13809290528297424, -0.4739532172679901, 0.13925617933273315, 0.22118902206420898, 0.0007849894464015961, 0.24563631415367126, 0.03358955681324005, 0.14707604050636292, 0.3634043335914612, 0.1003376767039299, 0.13167636096477509, -0.15881599485874176, 0.249946728348732, -0.18818019330501556, -0.1837327778339386, -0.08811596035957336, -0.5891331434249878, 0.21658015251159668, -0.1606370508670807, 0.22971299290657043, -0.07063737511634827, -0.0018130950629711151, -0.4160648286342621, -0.03989424556493759, 0.09383711963891983, -0.2353096604347229, -0.025886792689561844, 0.1969701051712036, 0.19552312791347504, 0.004466298967599869, -0.0157301127910614, 0.06437630206346512, -0.16206413507461548, 0.2334323674440384, 0.07400716841220856, -0.05165005475282669, -0.022029435262084007, 0.14289583265781403, 0.20402976870536804, 0.2436189353466034, 0.030902357771992683, -0.25026312470436096, -0.03667755797505379, 0.3134893476963043, 0.14573079347610474, 0.27442294359207153, 0.10608696937561035, -0.04028916358947754, 0.1848420351743698, 0.2803106904029846, 0.16047827899456024, 0.2589093744754791, 0.11254415661096573, -0.12400823831558228, 0.08150599151849747, 0.11401208490133286, 0.3063763678073883, 0.036905985325574875, -0.10141032934188843, 0.048838041722774506, 0.2526004910469055, 0.06619587540626526, -0.12181852757930756, -0.3277985155582428, -0.02463017776608467, 0.26366302371025085, 0.6258196830749512, -0.09830470383167267, 0.30670344829559326, -0.10810399800539017, -0.27656808495521545, -0.3729349374771118, 0.02712627872824669, 0.1354561299085617, -0.24313884973526, -0.05934799462556839, 0.0059099141508340836, 0.4128139019012451, 0.889143168926239, -0.08765040338039398, 0.21271073818206787, 0.2576933801174164, 0.09946372359991074, -0.14763511717319489, 0.3102174699306488, 0.27317506074905396, 0.03670463711023331, 0.15607155859470367, 0.017295630648732185, -0.2442476898431778, -0.032427191734313965, -0.09209543466567993, 0.33091625571250916, 0.35919851064682007, -0.30033421516418457, -0.10634095221757889, -0.30636581778526306, -0.5244472622871399, -0.36918240785598755, -0.3438991606235504, -0.20262132585048676, -0.3367350101470947, -0.16702450811862946, 0.2561098039150238, -0.016800876706838608, 0.2510792315006256, -0.1769595891237259, 0.23437507450580597, -0.35818105936050415, -0.261892169713974, 0.01408899575471878, -0.06981229037046432, -0.28657788038253784, 0.005503062158823013, 0.23285901546478271, 0.16187742352485657, 0.24962705373764038, -0.2545627951622009, -0.21234312653541565, -0.34649068117141724, -0.18361705541610718, 0.19877390563488007, 0.24772752821445465, 0.4137888252735138, 0.5279136896133423, -0.07021409273147583, 0.35673272609710693, 0.059449680149555206, -0.0660829097032547, -0.1920318752527237, -0.37169942259788513, -0.30342361330986023, -0.0005277288728393614, 0.27848020195961, -0.08362515270709991, -0.3137316405773163, -0.6029024720191956, -0.14122089743614197, 0.03423180431127548, 0.07534775137901306, 0.12522128224372864, 0.0305909626185894, -0.241026371717453, 0.22648760676383972, -0.0675986036658287, -0.23341895639896393, -0.20004770159721375, 0.15993042290210724, -0.009144742041826248, -0.16310185194015503, -0.0454622246325016, 0.07400737702846527, -0.10867740213871002, 0.0565042719244957, -0.23117852210998535, -0.4390268325805664, -0.4993737041950226, 0.2032100111246109, -0.011472251266241074, 0.14936907589435577, -0.18084731698036194, 0.22669407725334167, 0.24462680518627167, 0.1441558301448822, 0.007774624973535538, -0.40508028864860535, 0.12961594760417938, 0.1733417510986328, 0.2596013844013214, 0.2025211751461029, 0.4707232415676117, -0.18957823514938354, 0.6809959411621094, 0.08019046485424042, -0.06540736556053162, -0.006915055215358734, -0.19380617141723633, 0.12983432412147522, 0.09109852463006973, -0.07360170036554337, 0.21883246302604675, -0.19008871912956238, -0.27239489555358887, 0.26072990894317627, -0.36413365602493286, -0.09074007719755173, -0.15291041135787964, 0.21480898559093475, -0.1826370358467102, -0.1637749969959259, 0.07783818989992142, -0.20450931787490845, 0.0020859837532043457, 0.1484215408563614, 0.00008292868733406067, -0.21906617283821106, -0.05846768245100975, 0.10795803368091583, 0.2922644317150116, 0.0909213125705719, 0.11736852675676346, -0.2829023003578186, -0.0073561519384384155, -0.4521166682243347, 0.357546865940094, 0.3472751975059509, 0.4865119755268097, 0.21581445634365082, 0.03920448571443558, 0.008247800171375275, 0.040773775428533554, 0.21377846598625183, -0.029948057606816292, 0.01300492137670517, 0.25568151473999023, -0.1364879012107849, 0.061477094888687134, -0.22550590336322784, -0.09265458583831787, 0.07190170884132385, 0.0817243754863739, -0.0707302838563919, -0.3118754029273987, -0.15123586356639862, 0.33285269141197205, 0.15579362213611603, -0.11175557971000671, -0.24236750602722168, 0.008510123938322067, -0.15314172208309174, -0.19700324535369873, -0.19007036089897156, -0.24555917084217072, 0.1381574273109436, -0.12386288493871689, 0.44703003764152527, 0.013527256436645985, -0.09884018450975418, 0.08706212043762207, 0.024769417941570282, 0.1871049702167511, -0.15448376536369324, 0.13126252591609955, 0.17856888473033905, 0.002818576991558075, 0.30415141582489014, 0.28680574893951416, -0.02494533360004425, -0.2621816098690033, 0.02490977756679058, 0.11324077099561691, 0.3784663677215576, 0.3422499895095825, 0.19741113483905792, 0.06382715702056885, -0.2843545079231262, -0.1820514053106308, -0.14466552436351776, 0.1674746423959732, 0.3507557511329651, -0.10339272767305374, 0.05127161368727684, -0.2381136119365692, 0.6501338481903076, 0.1520877480506897, -0.21758562326431274, 0.2924673557281494, -0.1957998424768448, -0.251296728849411, -0.12040594965219498, 0.14373937249183655, 0.9772177934646606, 0.1787053644657135, 0.1921570748090744, 0.3799651563167572, -0.27018001675605774, 0.5282027125358582, 0.12167224287986755, 0.11431455612182617, -0.49746766686439514, -0.036809418350458145, 0.012771746143698692, -0.21958325803279877, 0.22603967785835266, -0.15513795614242554, -0.23718351125717163, 0.3542986512184143, -0.010540299117565155, 0.28239768743515015, 0.3365882933139801, 0.3203301429748535, -0.22276359796524048, -0.026248697191476822, -0.1613606959581375, 0.04255671054124832, -0.049717940390110016, 0.41778963804244995, -0.29927900433540344, 0.009875815361738205, -0.0863073468208313, -0.04552706331014633, -0.3256157636642456, -0.20060059428215027, -0.29272350668907166, 0.017192630097270012, 0.0827985405921936, -0.13996833562850952, -0.010616284795105457, 0.17796529829502106, 0.08981610089540482, -0.21600724756717682, -0.09981639683246613, 0.0924350768327713, -0.14039334654808044, 0.2634114623069763, 0.3869931399822235, 0.09648358076810837, 0.2686980366706848, -0.34610795974731445, -0.09964456409215927, 0.30464762449264526, -0.23491992056369781, -0.2469954490661621, -0.04180097579956055, -0.20667655766010284, 0.13844750821590424, -0.19164037704467773, -0.32677707076072693, -0.025653012096881866, 0.00028639286756515503, -0.1344083845615387, -0.003131527453660965, 0.16473454236984253, -0.36389032006263733, -0.008035466074943542, -0.20352201163768768, -0.2534613609313965, 0.049440596252679825, 0.4804389774799347, -0.02904576063156128, -0.22069184482097626, 0.665513277053833, -0.26307475566864014, 0.117669478058815, -0.1220516636967659, -0.0074471160769462585, -0.0930064469575882, -0.10929860174655914, 0.07604210078716278, -0.17611928284168243, -0.17633569240570068, -0.010683082975447178, 0.5820418000221252, 0.3484700322151184, 0.08524657785892487, -0.15745815634727478, -0.4427753686904907, -0.4029707908630371, 0.017590399831533432, -0.06714573502540588, 0.1379370242357254, -0.18047964572906494, 0.18105773627758026, 0.1291486918926239, 0.08300850540399551, -0.17618054151535034, -0.10881167650222778, 0.02771732024848461, 0.1015767902135849, -0.1334146112203598, 0.17750445008277893, 0.005330667831003666, 0.14194746315479279, -0.062301814556121826, 0.17821498215198517, -0.29202935099601746, -0.052992284297943115, -0.1140306144952774, 0.22683238983154297, 0.023889118805527687, 0.06089300662279129, 0.1892106533050537, -0.047966498881578445, 0.015474952757358551, 0.24556544423103333, -0.10414773225784302, -0.03355497866868973, -0.05473878234624863, 0.21973399817943573, 0.3419259786605835, 0.04210352897644043, -0.1363525092601776, 0.2392880618572235, 0.2469775229692459, 0.21957457065582275, -0.1397828459739685, 0.3792558014392853, -0.11718574166297913, 0.06506052613258362, 0.28737008571624756, 0.04528255760669708, -0.07998436689376831, 0.26757439970970154, 0.09384791553020477, 0.03071209415793419, 0.11529344320297241, -0.07018141448497772, 0.2299191802740097, 0.8423088192939758, -0.24265402555465698, -0.116752989590168, 0.40150654315948486, 0.009781595319509506, -0.233236163854599, -0.08031406253576279, 0.21382126212120056, 0.06446291506290436, 0.009682219475507736, 0.29438072443008423, 0.04901884123682976, -0.017301306128501892, 0.3186420798301697, 0.09393900632858276, 0.08764500916004181, -0.14573413133621216, -0.07335054874420166, 0.5390334725379944, -0.26027172803878784, 0.21228526532649994, 0.3923051059246063, 0.13994263112545013, 0.38456282019615173, 0.15995630621910095, -0.06687822192907333, 0.564166247844696, -0.06525686383247375, 0.007685359567403793, 0.20442351698875427, -0.08264769613742828, -0.08976000547409058, 0.3919522166252136, 0.028775639832019806, 0.022416558116674423, 0.3010081648826599, 0.3995402157306671, -0.3488949239253998, -0.1671561896800995, -0.20334208011627197, 0.03066154569387436, 0.12445712834596634, -0.11075709760189056, -0.23745091259479523, 0.005827784538269043, -0.2603166699409485, -0.05811135843396187, -0.18110890686511993, -0.3672564625740051, 0.25372642278671265, -0.05136583000421524, -0.3961370587348938, -0.1772400140762329, -0.5898517370223999, 0.02973219007253647, -0.10691618174314499, -0.26929712295532227, 0.05167829245328903, 0.03170248121023178, 0.07564826309680939, 0.22591061890125275, 0.4408133029937744, 0.3618394434452057, 0.10118646174669266, -0.32470765709877014, -0.3104070723056793, -0.20572011172771454, 0.16892816126346588, -0.0008588489145040512, 0.26479315757751465, 0.06462579220533371, -0.18335506319999695, 0.40197765827178955, 0.034111350774765015, 0.05196739733219147, 0.18164271116256714, 0.41097551584243774, -0.4785667359828949, -0.2713505029678345, 0.9588333368301392, 0.018587686121463776, -0.0753745585680008, -0.10514387488365173, 0.20924463868141174, -0.24594122171401978, 0.15358968079090118, -0.05407014861702919, -0.025078708305954933, 0.16526871919631958, 0.20228441059589386, 0.01158355176448822, -0.04843902215361595, 0.2834644913673401, 0.2929685711860657, -0.0652443915605545, -0.2998434901237488, -0.32373642921447754, -0.8680163621902466, 0.1610240489244461, 0.09295070171356201, -0.2784692049026489, -0.019421810284256935, -0.2899569869041443, 0.17587065696716309, 0.18767693638801575, -0.2499033361673355, 0.25636425614356995, -0.007288940716534853, -0.03545430302619934, -0.28572243452072144, -0.2891830801963806, -0.24124287068843842, -0.023864595219492912, -0.022986827418208122, -0.39517971873283386, 0.19543111324310303, -0.43772879242897034, -0.1360103338956833, 0.16049259901046753, -0.051106780767440796, 0.4187009334564209, 0.07762956619262695, 0.4448578953742981, -0.03417609632015228, 0.505246639251709, 0.15348967909812927, -0.05015362799167633, -0.047820840030908585, 0.25087106227874756, -0.14112074673175812, 0.1136057898402214, -0.07329361140727997, 0.09428730607032776, -0.17538873851299286, 0.6784104108810425, -0.41788214445114136, 0.40138810873031616, 0.01165064051747322, 0.12361008673906326, -0.4917759299278259, -0.002498175948858261, -0.1737731397151947, 0.14623795449733734, 0.37973955273628235, 0.25357934832572937, -0.17763745784759521, 0.36763089895248413, -0.4226214289665222, 0.02618400938808918, 0.504698634147644, -0.16471484303474426, -0.07719218730926514, -0.0422571636736393, 0.31683629751205444, -0.046242401003837585, -0.2065516859292984, -0.5365791916847229, 0.06525742262601852, 0.24328738451004028, -0.06352643668651581, 0.14099594950675964, 0.20620779693126678, 0.17869821190834045, 0.0790720134973526, -0.038183897733688354, 0.8223003149032593, -0.002348240464925766, 0.11937950551509857, -0.19420163333415985, -0.2611883580684662 ]
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
[ 0.0540458969771862, 0.07422429323196411, 0.009660083800554276, 0.17164751887321472, 0.018456868827342987, -0.1827668398618698, 0.40755602717399597, 0.1807367354631424, 0.031773824244737625, 0.24988536536693573, 0.038060203194618225, 0.06830566376447678, -0.06527296453714371, -0.2527969479560852, 0.016633106395602226, 0.15957820415496826, 0.1347910314798355, 0.04466550052165985, -0.2199346125125885, -0.0055113062262535095, -0.35158419609069824, 0.20710384845733643, -0.18735049664974213, 0.1459728181362152, -0.09561737626791, 0.2706403136253357, -0.22023658454418182, 0.253609836101532, -0.19941851496696472, -0.5675516128540039, 0.14673244953155518, 0.0295448936522007, 0.16571646928787231, 0.574688732624054, -0.00013265380403026938, -0.07714223116636276, 0.5401036143302917, -0.12277983129024506, -0.6687530875205994, -0.23048795759677887, 0.2879287898540497, -0.15296036005020142, 0.2021130472421646, -0.08125125616788864, 0.27061575651168823, -0.17419052124023438, 0.20436622202396393, -0.06535617262125015, 0.12489980459213257, 0.352989137172699, 0.019502006471157074, 0.1126968264579773, -0.010982226580381393, 0.07280859351158142, 0.09343883395195007, 0.2516893744468689, -0.09757190942764282, 0.46646958589553833, 0.21144595742225647, 0.0007744794711470604, -0.04591620713472366, -0.015806565061211586, 0.019108260050415993, 0.46090826392173767, 0.3064592480659485, 0.18696486949920654, 0.4650615453720093, 0.10373464971780777, -0.012513544410467148, -0.037865325808525085, 0.6569064259529114, -0.483895868062973, -0.4344106912612915, -0.0811818316578865, -0.12728391587734222, -0.20948468148708344, 0.38311195373535156, -0.10318616777658463, -0.3200289011001587, 0.09504377841949463, 0.014491195790469646, -0.2616751194000244, -0.03538689389824867, 0.21442347764968872, -0.3736017048358917, 0.39959248900413513, 0.03471030667424202, 0.19331267476081848, 0.0728292390704155, -0.049595415592193604, 0.30319857597351074, -0.34249186515808105, 0.04692394286394119, 0.16503408551216125, 0.09963254630565643, -0.0047329626977443695, -0.16642418503761292, -0.06243463605642319, -0.14533305168151855, 0.03446204215288162, -0.4331532418727875, -0.13929203152656555, -0.18535861372947693, 0.2131970226764679, 0.18500934541225433, 0.44877105951309204, 0.25296205282211304, 0.49305203557014465, 0.17895863950252533, 0.13217853009700775, 0.13476116955280304, 0.1885266900062561, -0.007200427819043398, -0.4462215304374695, -0.00336444191634655, 0.10703611373901367, 0.2182113230228424, -0.2600003182888031, -0.26317456364631653, -0.2860173285007477, -0.46615487337112427, -0.036148957908153534, -0.057705942541360855, 0.2854338586330414, -0.21941198408603668, -0.11122133582830429, 0.06850851327180862, 0.38588768243789673, -0.19950073957443237, 0.058217186480760574, 0.0005526789464056492, 0.039048463106155396, -0.2579006552696228, 0.03631364181637764, 0.09758567810058594, 0.1127357929944992, 0.21362905204296112, -0.3273884952068329, 0.1367974430322647, -0.14557063579559326, 0.07292064279317856, -0.0004481300711631775, 0.11817510426044464, 0.35788652300834656, -0.14977261424064636, 0.1903308928012848, 0.10121174156665802, -0.45273327827453613, -0.25621846318244934, 0.29938071966171265, -0.10430816560983658, -0.4040021300315857, 0.09910072386264801, -0.005106193944811821, -0.4261919856071472, 0.060874685645103455, 0.07882974296808243, -0.3964563310146332, 0.19351822137832642, -0.5260316729545593, 0.06719474494457245, -0.27525579929351807, -0.2804780900478363, -0.12022753059864044, -0.09893935173749924, 0.48066312074661255, -0.34747314453125, -0.2877002954483032, -0.3806164562702179, -0.043282147496938705, 0.1235184296965599, -0.0717243105173111, -0.12091965973377228, 0.043958328664302826, -0.5625421404838562, 0.018757790327072144, 0.43581622838974, -0.5859134197235107, -0.3239840567111969, 0.3623734712600708, -0.5026475191116333, -0.19662752747535706, 0.4976199269294739, -0.11810924112796783, -0.019310494884848595, -0.4362892508506775, 0.00029174238443374634, 0.04905955493450165, 0.06989674270153046, -0.04112561047077179, -0.17711324989795685, -0.2051975578069687, -0.02500303089618683, 0.08768302202224731, -0.17620697617530823, 0.10934769362211227, 0.04550113528966904, -0.10576526820659637, 0.4786716401576996, 0.0737261027097702, 0.09647504985332489, -0.04458694905042648, 0.20204539597034454, 0.050491005182266235, -0.04118292033672333, 0.25568586587905884, -0.5695757269859314, 0.14952585101127625, -0.15539118647575378, 0.12621267139911652, -0.2437577098608017, -0.08564536273479462, -0.3687710762023926, 0.06484142690896988, -0.14587223529815674, -0.15639926493167877, -0.10608261823654175, 0.5268374681472778, -0.0165015310049057, -0.06277436017990112, 0.04309337958693504, 0.08777257055044174, -0.1442553848028183, 0.24223849177360535, -0.11042909324169159, 0.0053618308156728745, -0.051639944314956665, 0.08585318177938461, -0.07348939776420593, 0.030897952616214752, 0.22429129481315613, -0.40818122029304504, 0.03320857882499695, 0.3498995900154114, 0.27478379011154175, 0.46276575326919556, -0.022554911673069, -0.1348695009946823, 0.1559963971376419, 0.24798241257667542, 0.30912429094314575, 0.34455084800720215, 0.08198750764131546, -0.28991881012916565, -0.01919369399547577, 0.27006644010543823, 0.1689300537109375, 0.31897249817848206, -0.031788356602191925, -0.049406748265028, 0.04129443317651749, 0.13392701745033264, -0.06262287497520447, -0.2984601855278015, 0.18931813538074493, 0.16936108469963074, 0.5996588468551636, 0.0712875947356224, 0.32798266410827637, 0.02567467838525772, -0.17272032797336578, -0.27904778718948364, -0.08578654378652573, -0.00003851577639579773, -0.18876498937606812, -0.08370187878608704, 0.09133266657590866, 0.40567946434020996, 0.9681351780891418, -0.09375648945569992, 0.33288437128067017, 0.03304961323738098, -0.10807809978723526, -0.2747822701931, 0.28881046175956726, 0.22763341665267944, -0.07494223117828369, 0.17336967587471008, -0.031398676335811615, -0.318402498960495, -0.0401051789522171, -0.050305068492889404, 0.5000061988830566, 0.14260096848011017, -0.5107771754264832, 0.22430628538131714, -0.25457626581192017, -0.3097939193248749, -0.6035181283950806, -0.31199488043785095, -0.06053781881928444, -0.5021187663078308, -0.09551560878753662, 0.3961228132247925, -0.030482858419418335, 0.20940709114074707, -0.07476457953453064, 0.08459112048149109, -0.44329190254211426, -0.44742313027381897, -0.014093589037656784, 0.008263982832431793, -0.48457759618759155, -0.1461125612258911, 0.3326222896575928, -0.08755472302436829, 0.22779244184494019, -0.23387663066387177, -0.20161934196949005, -0.23572483658790588, -0.29723623394966125, 0.12632709741592407, 0.15794748067855835, 0.4473174214363098, 0.23237921297550201, 0.00020796805620193481, 0.46073251962661743, 0.00311282929033041, 0.12738218903541565, -0.20606091618537903, -0.3891749083995819, -0.1575377732515335, 0.12728087604045868, 0.13422878086566925, -0.1593405157327652, -0.39147821068763733, -0.4792371988296509, -0.03614705055952072, -0.09529043734073639, 0.05918179824948311, 0.12565350532531738, 0.2484217882156372, -0.1625194549560547, 0.14527693390846252, 0.06382455676794052, -0.09572599828243256, -0.03304025158286095, 0.0895199328660965, 0.038432203233242035, -0.23889976739883423, -0.0015531666576862335, 0.044273559004068375, 0.00772886723279953, 0.056634943932294846, -0.09045366942882538, -0.4622616171836853, -0.47036102414131165, 0.19328805804252625, 0.055600088089704514, 0.18048320710659027, -0.10260474681854248, 0.45524048805236816, 0.2974397540092468, 0.13758741319179535, -0.09040047228336334, -0.3450206518173218, 0.024569302797317505, 0.08051788806915283, 0.3078870177268982, 0.1523379385471344, 0.3881366550922394, -0.19256743788719177, 0.8443016409873962, 0.12629902362823486, 0.035268448293209076, 0.08385501801967621, -0.15377211570739746, 0.1894451081752777, -0.08218657225370407, -0.031549304723739624, 0.15041430294513702, -0.08360810577869415, -0.5522128939628601, 0.36745840311050415, -0.05712199956178665, -0.23132175207138062, -0.06647630035877228, 0.2273532748222351, -0.14406490325927734, -0.15885165333747864, 0.0742008164525032, -0.19066937267780304, -0.01243416965007782, 0.16560128331184387, 0.002517789602279663, -0.2638314366340637, -0.2158578336238861, 0.01242510974407196, 0.3022843599319458, 0.02733498066663742, 0.0839228555560112, -0.330544114112854, -0.1284155398607254, -0.14755147695541382, 0.3084336519241333, 0.3300758898258209, 0.4101949632167816, -0.01594208925962448, -0.1134246289730072, 0.28132885694503784, 0.007648559287190437, 0.3506704568862915, 0.0673462450504303, 0.21243363618850708, 0.19533106684684753, -0.0024002231657505035, -0.11793342232704163, -0.18610519170761108, -0.07736092805862427, 0.14718542993068695, 0.10939393937587738, 0.18208011984825134, -0.23847326636314392, -0.26433059573173523, 0.20997938513755798, 0.19632458686828613, -0.15524880588054657, -0.2132083773612976, -0.2094298154115677, -0.12904754281044006, -0.1241973415017128, -0.3455800712108612, -0.13917407393455505, -0.06293387711048126, -0.09446351230144501, 0.3878284990787506, 0.01809648983180523, -0.18357983231544495, 0.05521734431385994, 0.004763578996062279, 0.30929049849510193, -0.1802648901939392, 0.15318980813026428, 0.3662013113498688, 0.10744713246822357, 0.406025767326355, 0.38970839977264404, -0.18061383068561554, -0.12755854427814484, 0.02575552463531494, 0.09579350054264069, 0.39173397421836853, 0.6246479749679565, 0.046177927404642105, -0.1896105408668518, -0.2906065285205841, -0.2193542867898941, -0.2398289144039154, 0.09601430594921112, 0.40891286730766296, 0.009106417186558247, 0.10071636736392975, -0.2854565680027008, 0.5554723739624023, 0.20811237394809723, -0.35643666982650757, 0.5225512981414795, -0.003554403781890869, -0.2918553054332733, 0.11549738794565201, -0.05118612200021744, 1.1163864135742188, 0.29423439502716064, 0.42396584153175354, 0.36086010932922363, -0.20886582136154175, 0.09465119987726212, 0.1167902946472168, 0.1752214878797531, -0.4606742858886719, -0.021532047539949417, -0.019033282995224, -0.33305877447128296, 0.23628032207489014, -0.03547031432390213, -0.09988817572593689, 0.4064427316188812, -0.04007875546813011, 0.3567257523536682, 0.193649023771286, 0.15737178921699524, -0.2964199185371399, 0.028264716267585754, 0.026144342496991158, -0.04065537452697754, 0.20728260278701782, 0.45894935727119446, -0.16293072700500488, 0.03178037330508232, -0.26329657435417175, -0.07733353972434998, -0.030731793493032455, -0.0730976015329361, -0.13402125239372253, 0.07461818307638168, 0.022244608029723167, -0.25191012024879456, -0.2052023857831955, 0.3793686628341675, 0.19289639592170715, -0.22925756871700287, 0.022599468007683754, 0.17318694293498993, -0.17188304662704468, 0.365487277507782, 0.3521057367324829, -0.04134276881814003, 0.30463165044784546, -0.13484054803848267, -0.13243813812732697, 0.10160700976848602, -0.2835327088832855, -0.19923004508018494, -0.2421746850013733, -0.09257858991622925, 0.22591666877269745, -0.3246315121650696, -0.14834462106227875, -0.14957788586616516, -0.08646188676357269, -0.10077975690364838, -0.05219167098402977, 0.14122480154037476, -0.28765591979026794, 0.010074004530906677, -0.2608496844768524, -0.42889732122421265, -0.0010194424539804459, 0.5187014937400818, -0.12068293243646622, -0.05585869401693344, 0.8927491903305054, -0.2553896903991699, 0.07004066556692123, 0.03783545643091202, 0.16071438789367676, -0.22212950885295868, -0.26517605781555176, 0.06714945286512375, -0.18132168054580688, -0.24391548335552216, 0.02168789505958557, 0.56891930103302, 0.32607346773147583, 0.08179120719432831, -0.3220890760421753, -0.4693026840686798, -0.24841606616973877, 0.006860598921775818, -0.13581523299217224, 0.18076582252979279, -0.2004815638065338, 0.02838112786412239, 0.18213112652301788, -0.11743354797363281, -0.10768835991621017, -0.019980747252702713, 0.11945166438817978, 0.22964121401309967, -0.15153944492340088, 0.05161178857088089, 0.06271803379058838, 0.03996072709560394, -0.08371775597333908, 0.10385216027498245, -0.2190840244293213, 0.056734517216682434, -0.0864664763212204, 0.2645622193813324, 0.028700631111860275, 0.004550819750875235, 0.006185074336826801, -0.32587987184524536, 0.13018319010734558, 0.06811779737472534, -0.02315872721374035, 0.02889157086610794, -0.050951115787029266, 0.04266997426748276, 0.30082887411117554, 0.06145787984132767, 0.029404588043689728, 0.3885266184806824, 0.05780847370624542, 0.25832945108413696, -0.18535099923610687, 0.21913087368011475, -0.2305954396724701, 0.020766451954841614, 0.040415238589048386, 0.10006482899188995, -0.018763378262519836, 0.2092517614364624, 0.13397377729415894, -0.18022271990776062, 0.11828559637069702, -0.10023938119411469, 0.36394810676574707, 0.9179773926734924, -0.14689356088638306, -0.23009686172008514, 0.4129929840564728, -0.09078706800937653, -0.057011187076568604, -0.12697485089302063, 0.07886934280395508, 0.1358589082956314, -0.058446865528821945, 0.3512890040874481, -0.052806586027145386, -0.11743712425231934, 0.34743165969848633, 0.21134522557258606, 0.20100340247154236, -0.23741672933101654, -0.10164037346839905, 0.4046119749546051, -0.30559778213500977, 0.4028899371623993, 0.4748879075050354, 0.0623086653649807, 0.20034976303577423, 0.3251499533653259, 0.08612170815467834, 0.5218544006347656, -0.14537331461906433, -0.02890695258975029, 0.12090800702571869, -0.10828767716884613, -0.029849331825971603, 0.2552216053009033, -0.005976662039756775, 0.04436124861240387, 0.35601699352264404, 0.5558950304985046, -0.47650131583213806, -0.10528735816478729, -0.23334404826164246, -0.13780389726161957, 0.01676420494914055, -0.2931421101093292, -0.21884988248348236, -0.03580223396420479, -0.1372559666633606, -0.03259801119565964, -0.19874674081802368, -0.19986771047115326, 0.2898292541503906, 0.056157853454351425, -0.2658582329750061, -0.2286197692155838, -0.36382436752319336, 0.019897585734725, 0.039980411529541016, -0.34758293628692627, 0.25455909967422485, -0.00025400519371032715, 0.13587933778762817, 0.2975616157054901, 0.13145563006401062, 0.48757925629615784, 0.1934380829334259, -0.2696226239204407, -0.2617153227329254, -0.2879706621170044, -0.0018292609602212906, -0.07789882272481918, 0.35549309849739075, -0.029869720339775085, -0.09149475395679474, 0.35833123326301575, -0.01570925861597061, -0.005666815675795078, 0.22060003876686096, 0.40760788321495056, -0.2764076292514801, -0.14980629086494446, 0.809942901134491, -0.24034637212753296, -0.2101065218448639, -0.07573121786117554, 0.033688705414533615, -0.4592142701148987, 0.12718133628368378, 0.011043690145015717, -0.14476530253887177, 0.1875208616256714, 0.11137866973876953, -0.03916380926966667, -0.09055313467979431, 0.3421502709388733, 0.4022057056427002, -0.1118353083729744, -0.33004412055015564, -0.19953064620494843, -0.8534044623374939, 0.2650274932384491, -0.05009061470627785, -0.14365550875663757, -0.20968382060527802, -0.07952213287353516, 0.21403393149375916, -0.017104536294937134, -0.11020126938819885, 0.11080672591924667, 0.05849342793226242, 0.183257058262825, -0.3073982000350952, -0.3775629997253418, -0.18099546432495117, 0.03965005278587341, -0.008477963507175446, -0.5915560722351074, 0.19940391182899475, -0.41751718521118164, -0.15232011675834656, 0.16196760535240173, -0.07171527296304703, 0.36899691820144653, 0.20901601016521454, 0.36452004313468933, -0.10860452055931091, 0.6161583662033081, 0.17797735333442688, -0.14680546522140503, -0.26281970739364624, 0.057859353721141815, -0.09725827723741531, 0.018361754715442657, -0.0892745703458786, 0.17369011044502258, -0.3242899477481842, 0.4118088483810425, -0.3782767951488495, 0.22013422846794128, 0.0850958377122879, -0.003485518041998148, -0.22224657237529755, 0.18447397649288177, -0.35888946056365967, 0.17099478840827942, 0.32443973422050476, 0.3272290527820587, 0.04074651747941971, 0.3295930325984955, -0.30356234312057495, -0.24247659742832184, 0.6405372619628906, -0.0990561991930008, -0.1643480807542801, 0.13678687810897827, 0.2945443391799927, -0.011159442365169525, -0.27710604667663574, -0.5036942958831787, 0.1726938784122467, 0.24136310815811157, 0.06933782994747162, 0.024126015603542328, 0.05193420872092247, 0.2529219090938568, 0.01719309762120247, -0.06566262990236282, 0.9180787801742554, 0.031080707907676697, 0.07061383873224258, 0.09902819991111755, -0.20347744226455688 ]
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?
[ -0.07890299707651138, 0.2855145335197449, 0.0002486761659383774, 0.052755214273929596, 0.05063414201140404, -0.260754257440567, 0.21438196301460266, 0.2759755849838257, -0.05403422564268112, 0.08017334342002869, 0.20722761750221252, 0.22357384860515594, -0.2214677333831787, -0.034581124782562256, 0.12442823499441147, 0.05447620525956154, -0.10658464580774307, 0.0847872644662857, -0.22504177689552307, 0.15503334999084473, -0.24932248890399933, 0.175332173705101, 0.0034216642379760742, 0.16723057627677917, 0.015413494780659676, 0.14610643684864044, -0.22377105057239532, 0.20182129740715027, -0.4194304645061493, -0.7930228114128113, 0.3121379017829895, 0.030213307589292526, 0.23527833819389343, 0.3840058743953705, -0.00012361556582618505, -0.09185232222080231, 0.5898000001907349, -0.15598922967910767, -0.7334057092666626, -0.42152416706085205, 0.2433733344078064, -0.5072743892669678, 0.2773005962371826, -0.1551036238670349, 0.18571290373802185, -0.06349457055330276, 0.2981802523136139, 0.20960953831672668, 0.18837349116802216, 0.4003972113132477, 0.07241328805685043, 0.30033084750175476, 0.051999133080244064, -0.1126798614859581, 0.00852164626121521, 0.1745273917913437, -0.05216128006577492, 0.587381899356842, 0.20252278447151184, 0.07431738078594208, -0.15627047419548035, -0.07906942069530487, 0.10531939566135406, 0.5026991367340088, 0.15227723121643066, 0.06298090517520905, 0.39901190996170044, 0.04652874916791916, -0.26758745312690735, 0.08663026243448257, 0.6675558090209961, -0.5247460007667542, -0.35870662331581116, -0.0709836483001709, -0.07901205867528915, -0.057431042194366455, 0.38079071044921875, -0.1414184868335724, -0.41265830397605896, 0.09209894388914108, -0.10644978284835815, -0.17128974199295044, -0.03199546039104462, 0.39670461416244507, -0.2983643412590027, 0.4051721394062042, -0.01846560835838318, 0.16227881610393524, 0.1562568098306656, 0.05911780148744583, 0.256651908159256, -0.17506058514118195, 0.17345485091209412, 0.18063046038150787, 0.10605965554714203, -0.20942501723766327, -0.2895841598510742, -0.27422627806663513, 0.1452917456626892, 0.11145221441984177, -0.30826395750045776, -0.19240990281105042, -0.06261324882507324, 0.22528690099716187, 0.23071277141571045, 0.3241159915924072, 0.1796366274356842, 0.2692381739616394, 0.2589324116706848, 0.30539003014564514, 0.2700367271900177, 0.16195189952850342, 0.007208423689007759, -0.33093106746673584, -0.121018186211586, 0.15665867924690247, 0.2048185169696808, -0.21207661926746368, -0.19729706645011902, -0.2485964596271515, -0.4048672914505005, -0.1265595555305481, -0.06877486407756805, 0.31161001324653625, -0.19009706377983093, -0.10306080430746078, 0.09242895245552063, 0.3832482695579529, -0.12803812325000763, -0.07070959359407425, 0.028685245662927628, 0.043144356459379196, -0.24823197722434998, 0.02863708883523941, 0.2963487207889557, 0.2202124446630478, 0.4854157269001007, -0.29047438502311707, -0.16111861169338226, -0.058454759418964386, 0.12964698672294617, 0.04346011206507683, -0.04721284657716751, 0.21914930641651154, -0.07588900625705719, 0.18278975784778595, 0.09156713634729385, -0.3037327527999878, -0.2998288571834564, 0.17252540588378906, -0.07229041308164597, -0.4023594856262207, 0.10723474621772766, 0.06545607000589371, -0.5473224520683289, 0.02173425815999508, 0.09589454531669617, -0.16957339644432068, -0.05753979831933975, -0.3571425974369049, 0.04796421900391579, -0.23688243329524994, -0.25407958030700684, -0.1595519483089447, -0.013156723231077194, 0.3331226110458374, -0.16873471438884735, -0.31161749362945557, -0.4004577696323395, -0.29786086082458496, 0.2116445004940033, -0.19554108381271362, -0.1178334653377533, 0.272870808839798, -0.4967009425163269, 0.06526096165180206, 0.6163495182991028, -0.4722350835800171, -0.29971882700920105, 0.5806673169136047, -0.45252561569213867, -0.06820160150527954, 0.263814240694046, -0.036954089999198914, -0.22122503817081451, -0.2977391183376312, -0.08449353277683258, 0.13490734994411469, 0.026060912758111954, -0.15002301335334778, -0.1978701651096344, -0.3536168336868286, 0.14594607055187225, 0.23883675038814545, -0.08798076212406158, 0.09268943965435028, 0.05308478698134422, -0.07039500772953033, 0.45289844274520874, 0.10668683052062988, 0.06376808136701584, -0.06113768368959427, 0.12078891694545746, -0.19681991636753082, -0.10716584324836731, 0.02618805319070816, -0.5845916271209717, 0.12513163685798645, -0.31481361389160156, 0.19763019680976868, -0.1885025054216385, -0.07421603798866272, -0.5055443644523621, -0.018278643488883972, 0.02284902334213257, -0.2564987242221832, 0.009734272956848145, 0.35500970482826233, 0.09080711007118225, -0.04917621612548828, 0.03225105628371239, 0.015585926361382008, -0.20264923572540283, 0.2942972183227539, -0.010932495817542076, 0.06413185596466064, -0.08776205778121948, 0.11537980288267136, 0.24681350588798523, 0.2504459619522095, 0.16802050173282623, -0.40396788716316223, 0.030690042302012444, 0.3176622688770294, 0.19105228781700134, 0.18399843573570251, 0.2012675404548645, -0.06688788533210754, 0.16839499771595, 0.191099613904953, 0.08350713551044464, 0.46942803263664246, 0.04755901172757149, -0.12958696484565735, 0.04580263793468475, 0.20797201991081238, 0.15497848391532898, -0.0021299943327903748, -0.07428234815597534, 0.07928677648305893, 0.3448399603366852, 0.022174391895532608, -0.10134498029947281, -0.2545163631439209, 0.10405221581459045, 0.3349344730377197, 0.6334259510040283, -0.006708128377795219, 0.2735132575035095, -0.10974115878343582, -0.14043673872947693, -0.3109172284603119, 0.046973202377557755, 0.14044451713562012, -0.1851264238357544, -0.10904297232627869, 0.0009038746356964111, 0.375546395778656, 0.8084438443183899, -0.00867752730846405, 0.337849497795105, 0.09863558411598206, -0.16344335675239563, -0.23178473114967346, 0.24697157740592957, 0.1726224720478058, -0.09168840199708939, 0.125870019197464, -0.15991605818271637, -0.3018524646759033, -0.14503926038742065, 0.0199350044131279, 0.42337411642074585, 0.3181326985359192, -0.4547545611858368, -0.0568389967083931, -0.2974061369895935, -0.5187103152275085, -0.5051658749580383, -0.31656911969184875, -0.20915921032428741, -0.35626786947250366, -0.056083157658576965, 0.28348276019096375, -0.13092832267284393, 0.166825070977211, -0.11503647267818451, 0.2502066493034363, -0.32945573329925537, -0.3059857487678528, 0.05414070188999176, -0.09253225475549698, -0.4306950271129608, 0.0013972930610179901, 0.40533092617988586, 0.09085674583911896, 0.25072017312049866, -0.23175708949565887, -0.2034362554550171, -0.2620789706707001, -0.27703648805618286, 0.13477784395217896, 0.18046344816684723, 0.5565597414970398, 0.4462127685546875, -0.0005804598331451416, 0.4154548645019531, 0.023446351289749146, 0.021319406107068062, -0.17712323367595673, -0.3594783842563629, -0.16657453775405884, 0.11461669206619263, 0.11637789756059647, -0.17482349276542664, -0.4352656900882721, -0.5285282731056213, -0.05853002518415451, 0.02924032136797905, 0.08637484908103943, 0.1540939211845398, 0.09188532829284668, -0.17155218124389648, 0.18889687955379486, 0.04282163083553314, -0.14498649537563324, -0.1551208794116974, -0.07493049651384354, 0.02633287012577057, -0.14079639315605164, -0.1739194393157959, 0.008674915879964828, -0.2095598429441452, 0.05793796479701996, -0.16223236918449402, -0.40622490644454956, -0.5565923452377319, 0.14096695184707642, 0.18127119541168213, 0.25342342257499695, -0.15074804425239563, 0.37731078267097473, 0.18244123458862305, 0.09180489927530289, -0.13267090916633606, -0.3363349437713623, 0.2459818720817566, 0.10700096935033798, 0.40160325169563293, 0.2060067057609558, 0.2957024872303009, -0.02632203698158264, 0.7256442308425903, 0.14918823540210724, -0.07229571044445038, 0.09985023736953735, -0.1860622614622116, 0.10383379459381104, 0.15628525614738464, 0.00590037927031517, 0.18452025949954987, -0.15817192196846008, -0.2670908570289612, 0.3186189830303192, -0.19768214225769043, -0.04666414111852646, -0.11971660703420639, 0.2846112847328186, -0.25960278511047363, -0.12736406922340393, 0.12735623121261597, -0.05860152840614319, -0.04352828860282898, 0.09806585311889648, -0.04892734810709953, -0.16124717891216278, -0.061358604580163956, 0.12111495435237885, 0.35151779651641846, 0.033287689089775085, 0.14570960402488708, -0.3123936951160431, -0.04856228455901146, -0.33045876026153564, 0.28733155131340027, 0.382264643907547, 0.42916834354400635, 0.05898071452975273, 0.013043813407421112, 0.0952792763710022, 0.0012861217837780714, 0.2868995666503906, 0.019052179530262947, 0.15850667655467987, 0.13101911544799805, -0.15485472977161407, 0.02473199926316738, -0.15523315966129303, -0.09564242511987686, 0.09846919029951096, 0.007519437000155449, 0.15079307556152344, -0.46832722425460815, -0.13879436254501343, 0.19019170105457306, 0.1147201880812645, -0.058022391051054, -0.22207745909690857, -0.01593119278550148, -0.12120790779590607, -0.047307536005973816, -0.22213134169578552, -0.14633366465568542, -0.009872816503047943, -0.10142961889505386, 0.29026487469673157, 0.07743420451879501, -0.1481543332338333, -0.014838173985481262, 0.08856996893882751, 0.2045021653175354, -0.3280985951423645, 0.11342334747314453, 0.27185356616973877, 0.19918057322502136, 0.22384952008724213, 0.40062448382377625, -0.1000058501958847, -0.21022135019302368, 0.03399192914366722, 0.05974150821566582, 0.3214837610721588, 0.4006612002849579, 0.10434781759977341, -0.039059631526470184, -0.2513597905635834, -0.15662208199501038, -0.33567067980766296, 0.22357992827892303, 0.4120948612689972, -0.10309460014104843, 0.1445939540863037, -0.22911985218524933, 0.6646549701690674, 0.13159412145614624, -0.2383853942155838, 0.5089074373245239, -0.28388911485671997, -0.2806604504585266, -0.077364481985569, 0.046777043491601944, 1.0281771421432495, 0.24669449031352997, 0.25193649530410767, 0.4927613437175751, -0.26039767265319824, 0.2928636074066162, 0.06944653391838074, 0.09858127683401108, -0.43218085169792175, 0.01908116042613983, -0.06404997408390045, -0.20193004608154297, 0.25413867831230164, 0.0323689728975296, -0.23341050744056702, 0.33858874440193176, 0.1327558159828186, 0.38942986726760864, 0.24295952916145325, 0.4002091884613037, -0.24319452047348022, -0.02153216302394867, -0.2314874529838562, 0.04449514299631119, -0.02915450930595398, 0.357897013425827, -0.20181342959403992, -0.08174533396959305, -0.20169222354888916, -0.13223759829998016, -0.0011773854494094849, -0.08184608072042465, -0.3922838568687439, 0.028829768300056458, 0.0667119100689888, -0.10534249246120453, -0.23339538276195526, 0.2026793509721756, 0.10940498113632202, -0.2765062749385834, -0.131254643201828, 0.14214517176151276, -0.31567680835723877, 0.2580554187297821, 0.32302427291870117, -0.006119425408542156, 0.24911458790302277, -0.3315946161746979, -0.026277735829353333, 0.07291926443576813, -0.3236357271671295, -0.19239988923072815, 0.07747150957584381, -0.21757884323596954, 0.07555195689201355, -0.3103991746902466, -0.2757740318775177, -0.1316889226436615, -0.08827870339155197, -0.11784468591213226, -0.004206649959087372, 0.22242960333824158, -0.35297122597694397, -0.07808303087949753, -0.06195707991719246, -0.26087620854377747, -0.005462471395730972, 0.41550642251968384, -0.2048795223236084, -0.048417191952466965, 0.6307448148727417, -0.41792699694633484, 0.09337036311626434, -0.07985340058803558, 0.13359007239341736, -0.15759526193141937, -0.11624114215373993, 0.05989890545606613, -0.08217970281839371, -0.31458503007888794, -0.05027610808610916, 0.5699489712715149, 0.42291873693466187, 0.1599046289920807, -0.18029290437698364, -0.4978507161140442, -0.30219781398773193, 0.04070744663476944, -0.19161346554756165, 0.1929028332233429, -0.08632348477840424, 0.21488791704177856, 0.23817583918571472, 0.0008389707654714584, -0.17753931879997253, -0.11061951518058777, -0.044823408126831055, 0.18705295026302338, -0.17645959556102753, 0.028117746114730835, 0.08777680993080139, 0.11093795299530029, -0.05144146829843521, 0.07127989083528519, -0.26637977361679077, -0.03337724506855011, -0.10417591035366058, 0.22602683305740356, 0.15439248085021973, -0.07422837615013123, 0.1739831566810608, -0.14891168475151062, 0.02848183363676071, 0.15573972463607788, -0.13240166008472443, -0.0511799156665802, -0.030449509620666504, 0.11647063493728638, 0.23405784368515015, 0.0443299263715744, -0.20257171988487244, 0.2824553847312927, -0.0093793123960495, 0.15508587658405304, -0.09460094571113586, 0.20901909470558167, -0.32232046127319336, 0.03160156309604645, 0.17144279181957245, 0.0984681248664856, -0.06700043380260468, 0.2769375145435333, 0.13658586144447327, 0.0017016790807247162, 0.0024577639997005463, -0.04743047058582306, 0.3074738681316376, 0.7473387718200684, -0.20278300344944, -0.1288805902004242, 0.46390342712402344, -0.003537174314260483, -0.211854487657547, 0.06120787560939789, 0.1913791447877884, 0.004370968788862228, 0.0706610381603241, 0.2556614279747009, 0.045390624552965164, 0.0798020213842392, 0.2536444067955017, 0.05678654462099075, 0.15905046463012695, -0.06254468113183975, -0.05567086860537529, 0.5163021683692932, -0.24758440256118774, 0.34521734714508057, 0.34482505917549133, -0.026771780103445053, 0.31721267104148865, 0.043735288083553314, -0.022251777350902557, 0.548308789730072, -0.005632694810628891, 0.007812529802322388, 0.12488171458244324, 0.007367581129074097, -0.10260365903377533, 0.20615416765213013, 0.0983358845114708, 0.05027972534298897, 0.25428175926208496, 0.6443103551864624, -0.5301637053489685, -0.19425921142101288, -0.2695655822753906, -0.09844449162483215, 0.08751742541790009, -0.11784860491752625, -0.45354101061820984, -0.008798956871032715, -0.325763463973999, -0.14106620848178864, -0.13782058656215668, -0.3443898558616638, 0.25596433877944946, 0.003879956901073456, -0.3412185311317444, -0.25168558955192566, -0.43340232968330383, 0.13744866847991943, -0.04456020146608353, -0.3266587257385254, 0.2011287808418274, 0.033438317477703094, 0.031245622783899307, 0.27481305599212646, 0.29906928539276123, 0.4600580930709839, 0.07712925970554352, -0.2869519293308258, -0.3478582799434662, -0.1940525621175766, 0.07553841173648834, -0.07693585008382797, 0.2965201437473297, 0.006350502371788025, -0.16871435940265656, 0.5234032273292542, 0.04897032678127289, 0.015593406744301319, 0.2165277898311615, 0.4239833354949951, -0.4900081157684326, -0.1489172875881195, 0.7677968740463257, -0.07951385527849197, -0.08629504591226578, -0.05311248451471329, -0.0124601349234581, -0.2715527415275574, 0.06507033109664917, -0.10145239531993866, -0.13405360281467438, 0.22263294458389282, 0.12485482543706894, 0.026110291481018066, -0.059313204139471054, 0.29371005296707153, 0.36874908208847046, -0.12910723686218262, -0.4022064507007599, -0.2841159701347351, -0.9730969667434692, 0.07774315774440765, 0.019219333305954933, -0.3939654529094696, -0.07160516828298569, -0.276656836271286, 0.21714946627616882, 0.11051171272993088, -0.2178172916173935, 0.2072967290878296, 0.24071791768074036, 0.04598340019583702, -0.26249340176582336, -0.3137997090816498, -0.24278898537158966, 0.03638816624879837, -0.04189561679959297, -0.4255923628807068, 0.11773188412189484, -0.22743497788906097, -0.14005209505558014, 0.2893140912055969, -0.04074516519904137, 0.34043124318122864, 0.0762927234172821, 0.28897255659103394, -0.07201776653528214, 0.6170159578323364, 0.10390684753656387, 0.010898256674408913, -0.11753702163696289, 0.21040001511573792, -0.11691687256097794, 0.0672914907336235, -0.07970216870307922, 0.20908327400684357, -0.18039096891880035, 0.6635023355484009, -0.5068444013595581, 0.4374556839466095, 0.04022591561079025, 0.08135279268026352, -0.26135605573654175, 0.10217372328042984, -0.3644084334373474, 0.13158951699733734, 0.35818034410476685, 0.28291043639183044, -0.12177640199661255, 0.30615195631980896, -0.1378706991672516, -0.19781450927257538, 0.3982648551464081, -0.12612102925777435, -0.041531506925821304, 0.04977952316403389, 0.2727915346622467, -0.09283065050840378, -0.061108775436878204, -0.44523483514785767, 0.05422306805849075, 0.27372869849205017, 0.02921084128320217, 0.11292874813079834, 0.09593437612056732, 0.15679357945919037, 0.11953762173652649, -0.06224054843187332, 0.8743886947631836, 0.04461134970188141, 0.017599323764443398, 0.0011314190924167633, -0.24424110352993011 ]
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.
[ -0.16123457252979279, 0.25518926978111267, 0.016922496259212494, -0.0688912570476532, -0.01873302087187767, -0.07994215935468674, 0.4035909175872803, 0.33871880173683167, 0.07540589570999146, 0.048679523169994354, 0.19280530512332916, 0.1107184961438179, -0.29939645528793335, 0.10281576961278915, 0.10059244930744171, -0.11661408841609955, 0.022554785013198853, 0.22912092506885529, -0.17938730120658875, 0.22091637551784515, -0.25348174571990967, 0.10033519566059113, -0.11082915961742401, 0.17837145924568176, -0.011252150870859623, 0.1582915186882019, -0.28702640533447266, 0.0593031570315361, -0.33930763602256775, -0.7205919027328491, 0.35620036721229553, 0.07208052277565002, 0.2591177821159363, 0.4941641688346863, -0.00012244142999406904, -0.07432939857244492, 0.5197003483772278, -0.1447192281484604, -0.675597071647644, -0.23255501687526703, 0.24015584588050842, -0.35562556982040405, 0.33595699071884155, 0.001460447907447815, 0.05675724148750305, -0.22579844295978546, 0.24793411791324615, 0.1120186448097229, 0.22972828149795532, 0.36378130316734314, 0.13691160082817078, 0.29282712936401367, 0.05487312749028206, -0.12325570732355118, -0.14173871278762817, -0.011339962482452393, -0.11851313710212708, 0.544145405292511, 0.34197700023651123, 0.03315939009189606, 0.007737316191196442, -0.05453301966190338, 0.08724568039178848, 0.39128631353378296, 0.19234415888786316, 0.00016621872782707214, 0.5235511064529419, -0.04455576092004776, -0.11648385971784592, 0.0907764807343483, 0.7365921139717102, -0.5787572264671326, -0.38571593165397644, -0.00001623481512069702, 0.051615335047245026, -0.11142754554748535, 0.17219947278499603, -0.1005590409040451, -0.29932621121406555, 0.05937647074460983, -0.10987194627523422, -0.22395355999469757, -0.10024036467075348, 0.2735373079776764, -0.21386848390102386, 0.23304477334022522, 0.06023814529180527, 0.22848206758499146, 0.0010030008852481842, -0.10714861750602722, 0.19523392617702484, -0.130006343126297, -0.05614924803376198, 0.2684307396411896, -0.09157226979732513, -0.18151730298995972, -0.344387024641037, -0.1629757583141327, 0.13957548141479492, -0.1011957973241806, -0.31825655698776245, -0.11622599512338638, 0.0414256751537323, 0.18302571773529053, 0.15644404292106628, 0.36657193303108215, 0.2711389660835266, 0.2031317502260208, 0.12128384411334991, 0.1413872390985489, 0.2718666195869446, 0.20314240455627441, -0.06695925444364548, -0.3533778488636017, -0.012509308755397797, 0.09310374408960342, 0.20516282320022583, -0.279924601316452, -0.16181987524032593, -0.21086351573467255, -0.3621448278427124, 0.011534607037901878, -0.12015796452760696, 0.3297590911388397, -0.19222243130207062, -0.18938562273979187, 0.07900408655405045, 0.5592828989028931, -0.20093844830989838, -0.21602337062358856, -0.01442256011068821, 0.02494082972407341, -0.21503762900829315, -0.06536278128623962, 0.2374853938817978, 0.009567760862410069, 0.40282773971557617, -0.255793035030365, -0.2903883457183838, -0.006051894277334213, 0.2314986139535904, 0.020160995423793793, -0.1914072036743164, 0.2558795213699341, -0.05008906126022339, -0.02477305382490158, 0.10018885135650635, -0.33269721269607544, -0.22785735130310059, 0.22125381231307983, 0.10084739327430725, -0.46670836210250854, -0.2118019014596939, 0.09065298736095428, -0.4503854513168335, 0.021674536168575287, 0.017006002366542816, -0.046402547508478165, 0.09855165332555771, -0.36174729466438293, -0.08257303386926651, -0.4229154586791992, -0.2915908098220825, -0.11434758454561234, -0.057407885789871216, 0.32801368832588196, -0.43154019117355347, -0.27291274070739746, -0.40979209542274475, -0.29973480105400085, 0.10333655774593353, -0.26148858666419983, -0.07541268318891525, 0.2706596255302429, -0.39036864042282104, -0.10745108127593994, 0.5073695778846741, -0.5006064176559448, -0.2522049844264984, 0.44504252076148987, -0.3928949236869812, -0.19235235452651978, 0.17964544892311096, 0.03588748350739479, 0.002271756064146757, -0.2983287572860718, -0.12099172174930573, 0.07886011153459549, 0.08520586043596268, -0.13364507257938385, -0.1600973904132843, -0.3697941303253174, 0.139165997505188, 0.2701592743396759, -0.08963312208652496, 0.1255897730588913, 0.021427154541015625, -0.011060118675231934, 0.33645498752593994, 0.08014397323131561, 0.1863797903060913, -0.08931335061788559, 0.11018542945384979, -0.1883334070444107, -0.09345072507858276, -0.011489389464259148, -0.6761329174041748, 0.2105301022529602, -0.20468048751354218, 0.3002508282661438, -0.01447094976902008, 0.03403789550065994, -0.602074146270752, 0.01895187236368656, -0.08776955306529999, -0.2432880401611328, 0.04227837175130844, 0.19854815304279327, 0.13480232656002045, 0.05542029067873955, 0.0055914223194122314, 0.13680914044380188, -0.15025188028812408, 0.25498688220977783, -0.07851333916187286, -0.028988558799028397, -0.11474830657243729, 0.09138835966587067, 0.30210480093955994, 0.28191596269607544, 0.08846971392631531, -0.3099205493927002, 0.05243707820773125, 0.28284236788749695, 0.2346717268228531, 0.14354732632637024, 0.2016160488128662, -0.11179046332836151, 0.08561429381370544, 0.22559982538223267, 0.2614986002445221, 0.3159290552139282, 0.10255780816078186, -0.21182101964950562, -0.001661062240600586, 0.24940182268619537, 0.2455359399318695, -0.004479527473449707, -0.22424300014972687, 0.058573052287101746, 0.29595980048179626, 0.10568228363990784, -0.10605470091104507, -0.31116706132888794, 0.09343528002500534, 0.3330496847629547, 0.5936193466186523, -0.09914626181125641, 0.19831421971321106, -0.03524620085954666, -0.26504409313201904, -0.320637971162796, 0.11222005635499954, 0.11663520336151123, -0.283305287361145, -0.10186264663934708, 0.032512299716472626, 0.2716321349143982, 0.8015956282615662, -0.05509653687477112, 0.05959898978471756, 0.11370687186717987, -0.153585746884346, -0.1946409046649933, 0.19420088827610016, 0.10261137783527374, -0.07959931343793869, 0.11133971810340881, -0.04581393301486969, -0.3477388322353363, -0.03733554854989052, -0.00987880676984787, 0.41767418384552, 0.40672385692596436, -0.40734031796455383, -0.05698326975107193, -0.15354426205158234, -0.37526407837867737, -0.38578373193740845, -0.3960627317428589, -0.20939934253692627, -0.3791486918926239, -0.06100854277610779, 0.3446577489376068, 0.018216382712125778, 0.25030481815338135, -0.09401370584964752, 0.26372215151786804, -0.20558635890483856, -0.3096761107444763, 0.1600995659828186, -0.13870495557785034, -0.4587654173374176, 0.03016543760895729, 0.24290655553340912, 0.16651377081871033, 0.36580029129981995, -0.26891186833381653, -0.21224729716777802, -0.20119880139827728, -0.31963756680488586, 0.1471567153930664, -0.010631486773490906, 0.6500743627548218, 0.4507486820220947, 0.05853548273444176, 0.29792624711990356, 0.11121876537799835, 0.003967395052313805, -0.1363392025232315, -0.3817053437232971, -0.08096225559711456, -0.021139590069651604, 0.06345357745885849, -0.1135386973619461, -0.4393893778324127, -0.5979088544845581, -0.02806536667048931, 0.07417380809783936, 0.11462894082069397, 0.12124425172805786, 0.13008537888526917, -0.01536463387310505, 0.18101954460144043, 0.03190884366631508, -0.19629228115081787, -0.1938023418188095, 0.05310021713376045, 0.08232904970645905, -0.17546875774860382, -0.28174200654029846, 0.16572795808315277, -0.21765828132629395, 0.08913887292146683, -0.26760947704315186, -0.4637167453765869, -0.46481841802597046, 0.24480928480625153, 0.163632333278656, 0.30453333258628845, -0.20452965795993805, 0.29216694831848145, 0.251864492893219, 0.17327265441417694, -0.12148859351873398, -0.3563101589679718, 0.20366646349430084, 0.015652041882276535, 0.32238906621932983, 0.19924810528755188, 0.39701831340789795, -0.27665087580680847, 0.7669869661331177, 0.18207097053527832, -0.08122356235980988, -0.03907804191112518, -0.2118774652481079, 0.09913470596075058, -0.08708595484495163, -0.05764743685722351, 0.23883286118507385, -0.1382332444190979, -0.2020779699087143, 0.3131054639816284, -0.17024582624435425, -0.009244050830602646, -0.12770086526870728, 0.23863375186920166, -0.11140106618404388, -0.06311244517564774, 0.059580087661743164, -0.16654354333877563, 0.03780529275536537, 0.098680779337883, -0.014956295490264893, -0.26176801323890686, -0.19446444511413574, -0.05254671350121498, 0.3353535830974579, 0.0021536946296691895, 0.11820797622203827, -0.3188006281852722, 0.01123667974025011, -0.42654624581336975, 0.27860593795776367, 0.2302422970533371, 0.5973356366157532, 0.12704850733280182, 0.08186542987823486, 0.10294012725353241, 0.10305597633123398, 0.21146516501903534, -0.007445132359862328, 0.099692702293396, 0.2119990736246109, -0.0997568815946579, -0.21547281742095947, -0.24166740477085114, -0.08823574334383011, 0.15256533026695251, -0.04120660200715065, 0.12435625493526459, -0.4224529266357422, -0.06276498734951019, 0.23153817653656006, 0.19613614678382874, -0.18541806936264038, -0.3014693558216095, -0.19508326053619385, -0.013291332870721817, -0.10634040832519531, -0.22760498523712158, -0.2369612753391266, 0.07546576857566833, -0.19839833676815033, 0.434601366519928, 0.026548190042376518, -0.0950339064002037, -0.0094674751162529, 0.22040186822414398, 0.21856418251991272, -0.22556573152542114, 0.34710660576820374, 0.212654709815979, 0.04422229155898094, 0.1937830001115799, 0.2733420133590698, 0.04669993370771408, -0.26991817355155945, 0.07171868532896042, -0.055965546518564224, 0.3882956802845001, 0.2309386134147644, 0.07861830294132233, -0.00005412846803665161, -0.2286573350429535, -0.15552125871181488, -0.2668573260307312, 0.11353760957717896, 0.36508840322494507, -0.012089261785149574, -0.057566914707422256, -0.3237791657447815, 0.5772803425788879, 0.1279294192790985, -0.2223547101020813, 0.3620973527431488, -0.2282121330499649, -0.2341749221086502, 0.1770775318145752, 0.05815240740776062, 0.8921768665313721, 0.16323108971118927, 0.26067525148391724, 0.4901677072048187, -0.2221510261297226, 0.35116297006607056, 0.21474815905094147, 0.11311942338943481, -0.6522517204284668, -0.002711131237447262, 0.03967949002981186, -0.14634232223033905, 0.21416574716567993, -0.14346690475940704, -0.33664727210998535, 0.3871508240699768, 0.004389531910419464, 0.2776857614517212, 0.28608784079551697, 0.3106580376625061, -0.178711399435997, -0.027079682797193527, -0.11676041781902313, 0.08298251032829285, -0.08146439492702484, 0.3729132413864136, -0.23302605748176575, -0.07518521696329117, -0.26928502321243286, -0.062462709844112396, -0.19180415570735931, -0.13464801013469696, -0.16934475302696228, -0.052009984850883484, 0.11992661654949188, -0.11077143251895905, -0.04196767508983612, 0.23126064240932465, 0.24265559017658234, -0.14473292231559753, -0.2273571789264679, 0.05623307079076767, -0.2652631402015686, 0.07546389847993851, 0.32904568314552307, 0.06114065647125244, 0.25031226873397827, -0.2712119221687317, -0.1735934019088745, 0.10917004197835922, -0.24852509796619415, -0.06863466650247574, -0.14945855736732483, -0.20029298961162567, 0.1652698963880539, -0.3002743422985077, -0.3457500636577606, -0.11714127659797668, -0.049097299575805664, -0.09219112992286682, 0.02859160676598549, 0.22841797769069672, -0.34068039059638977, -0.005799546837806702, -0.11406514048576355, -0.2673192024230957, -0.028432749211788177, 0.4541555643081665, -0.029472291469573975, -0.01376478374004364, 0.8067978620529175, -0.09696400165557861, 0.1509828418493271, -0.12489811331033707, 0.0640556737780571, -0.00214083562605083, -0.20873409509658813, 0.007497620768845081, -0.07276029884815216, -0.3054572343826294, 0.06333999335765839, 0.6002620458602905, 0.3502771854400635, -0.006405100226402283, -0.015813302248716354, -0.5503168106079102, -0.47054189443588257, -0.01743742823600769, -0.13372236490249634, 0.10882799327373505, -0.09000854194164276, 0.1982867568731308, 0.259976327419281, 0.030203042551875114, -0.20928725600242615, -0.13878828287124634, -0.02385549061000347, 0.19493086636066437, -0.1560315042734146, 0.01562574878334999, 0.04091256856918335, 0.10926033556461334, -0.015660665929317474, 0.01699753664433956, -0.39186587929725647, -0.07138168066740036, -0.10062699019908905, 0.20227810740470886, 0.11084682494401932, -0.11428558081388474, 0.06330285221338272, -0.12285292148590088, -0.11143916845321655, 0.19947606325149536, -0.0942072719335556, 0.017175450921058655, -0.1065148115158081, 0.18531452119350433, 0.3287245035171509, 0.051338911056518555, -0.09047839790582657, 0.2512051463127136, 0.10262133926153183, 0.189700186252594, -0.11579453200101852, 0.2186831831932068, -0.25777652859687805, 0.04440464824438095, 0.12394805252552032, 0.13644638657569885, -0.06430709362030029, 0.31939175724983215, 0.1831655204296112, -0.14177630841732025, 0.14575521647930145, -0.12332576513290405, 0.3411395251750946, 0.7627294659614563, -0.2596036195755005, -0.1603023260831833, 0.4350288510322571, 0.04208820313215256, -0.23856458067893982, 0.005389163736253977, 0.20815473794937134, 0.10267128795385361, 0.05342177674174309, 0.26197218894958496, 0.0547143928706646, -0.08130946010351181, 0.17981886863708496, 0.1486179530620575, 0.1108655333518982, -0.19002793729305267, -0.08867540955543518, 0.504672646522522, -0.2913787364959717, 0.2332158386707306, 0.37284061312675476, 0.042408380657434464, 0.1642451137304306, 0.16576366126537323, 0.0522635318338871, 0.47337597608566284, 0.055275097489356995, 0.05170278996229172, 0.11520883440971375, -0.026024559512734413, -0.004513751715421677, 0.3991509974002838, -0.009431816637516022, 0.008123233914375305, 0.3110029995441437, 0.5946427583694458, -0.4088001847267151, -0.15509100258350372, -0.3245684802532196, -0.10484825074672699, 0.035217925906181335, -0.13292965292930603, -0.38027244806289673, 0.07117969542741776, -0.253785103559494, -0.15523608028888702, -0.1535770148038864, -0.32617002725601196, 0.1191234141588211, 0.0025212764739990234, -0.3347843885421753, -0.23041340708732605, -0.46209514141082764, 0.296138197183609, -0.10243219882249832, -0.21056632697582245, 0.2456563264131546, 0.06732206046581268, 0.13905887305736542, 0.35602322220802307, 0.38687512278556824, 0.4162227213382721, 0.15023067593574524, -0.4203754961490631, -0.17785893380641937, -0.15420058369636536, 0.10862676799297333, 0.07844686508178711, 0.3482341170310974, -0.01889950782060623, -0.18846863508224487, 0.551433801651001, 0.06358034908771515, 0.041565146297216415, 0.2617347538471222, 0.4324660897254944, -0.4136831760406494, -0.1912732571363449, 0.7280192375183105, -0.14776593446731567, -0.12769809365272522, -0.11568330228328705, 0.17403200268745422, -0.3391324281692505, 0.1881857067346573, 0.05504055321216583, -0.011934895068407059, 0.20172029733657837, 0.004270358011126518, 0.0377386100590229, -0.1704934686422348, 0.42981845140457153, 0.39635366201400757, -0.030322395265102386, -0.4015501141548157, -0.241623654961586, -0.8262504935264587, 0.19436654448509216, -0.09572447836399078, -0.317832887172699, -0.037830013781785965, -0.23111912608146667, 0.19117669761180878, 0.14636509120464325, -0.2504756450653076, 0.2390633225440979, -0.015760596841573715, 0.1646002233028412, -0.351351261138916, -0.3869234621524811, -0.2707552909851074, 0.11316990852355957, -0.08452510088682175, -0.4501946270465851, 0.1749820113182068, -0.2466265857219696, -0.09763594716787338, 0.16377025842666626, -0.01951543614268303, 0.4524601101875305, 0.10025601834058762, 0.3120701313018799, -0.04562416672706604, 0.5762510895729065, 0.17790579795837402, -0.040750354528427124, -0.23221498727798462, 0.19575996696949005, -0.1450602412223816, 0.09771563112735748, 0.049104947596788406, 0.15912657976150513, -0.12989236414432526, 0.5070652365684509, -0.3636961281299591, 0.4363595247268677, 0.08669717609882355, 0.1201665997505188, -0.4500044286251068, 0.17375850677490234, -0.22239556908607483, 0.1384895145893097, 0.24804355204105377, 0.26967653632164, -0.10547012090682983, 0.4141574501991272, -0.18960265815258026, -0.05892527848482132, 0.4289872944355011, -0.20631742477416992, -0.10361835360527039, 0.06802307814359665, 0.19691064953804016, -0.11535202711820602, -0.17309705913066864, -0.4120732247829437, 0.2724257707595825, 0.24500282108783722, -0.06220118701457977, 0.1468648612499237, 0.16532815992832184, 0.261756956577301, 0.09281069040298462, -0.012294737622141838, 0.8619529008865356, -0.04305990785360336, 0.12368963658809662, -0.07380132377147675, -0.23250676691532135 ]
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, 0.211175799369812, -0.009480161592364311, -0.13179269433021545, 0.07344302535057068, -0.14275367558002472, 0.4176613688468933, 0.2573600709438324, 0.04997287318110466, 0.11905574053525925, 0.22333268821239471, 0.06963523477315903, -0.2887466251850128, 0.08405555784702301, 0.09769432246685028, -0.007682241499423981, 0.06393394619226456, 0.14133603870868683, -0.10444194078445435, 0.20766906440258026, -0.25531724095344543, -0.005777670070528984, -0.05395089089870453, 0.10496244579553604, -0.021724596619606018, 0.12302412837743759, -0.209518700838089, 0.10340224206447601, -0.43610069155693054, -0.8527143001556396, 0.3802427053451538, 0.10694992542266846, 0.288836807012558, 0.3949592411518097, -0.00012843438889831305, -0.05747811868786812, 0.5988147258758545, -0.1096535176038742, -0.6252241134643555, -0.1496017426252365, 0.27855581045150757, -0.3912251591682434, 0.3590851426124573, -0.03522179648280144, 0.1500791609287262, -0.1874672919511795, 0.21873152256011963, 0.041837818920612335, 0.1443508267402649, 0.2882784307003021, 0.0764792263507843, 0.2813306748867035, 0.06819677352905273, -0.21769098937511444, -0.11556913703680038, 0.05953776836395264, -0.08263536542654037, 0.5680218935012817, 0.4420577883720398, 0.16450203955173492, -0.04790712147951126, -0.1347491294145584, 0.09925621747970581, 0.45589518547058105, 0.18909728527069092, -0.04519214481115341, 0.6421708464622498, 0.01721707358956337, -0.14175671339035034, -0.0005506984889507294, 0.9479653239250183, -0.5009499192237854, -0.39741554856300354, 0.05029571056365967, -0.011666812933981419, -0.121365986764431, 0.23012828826904297, -0.192775696516037, -0.29002612829208374, 0.04725087806582451, -0.17219513654708862, -0.30624648928642273, -0.12384654581546783, 0.30030035972595215, -0.30258822441101074, 0.30529946088790894, 0.10334379225969315, 0.24936047196388245, 0.018844015896320343, -0.11828610301017761, 0.28435900807380676, -0.03635101020336151, -0.014641761779785156, 0.18634501099586487, 0.07852449268102646, -0.28582820296287537, -0.3459250032901764, -0.2041810154914856, 0.11312181502580643, -0.041234441101551056, -0.4233127534389496, -0.18879029154777527, -0.03495577722787857, 0.17621392011642456, 0.2227984070777893, 0.2793654501438141, 0.32268935441970825, 0.19643908739089966, 0.24403151869773865, 0.07674850523471832, 0.3281916081905365, 0.2670007646083832, -0.07985208928585052, -0.4030311107635498, 0.016208577901124954, 0.059761494398117065, 0.21140821278095245, -0.35215792059898376, -0.08716558665037155, -0.26098182797431946, -0.33429986238479614, -0.04781266674399376, -0.1376747190952301, 0.25438007712364197, -0.11698386818170547, -0.09509874135255814, 0.030406370759010315, 0.42387014627456665, -0.15298610925674438, -0.1591300517320633, -0.010798684321343899, 0.014280443079769611, -0.16778892278671265, -0.01822957955300808, 0.21012519299983978, 0.08893676102161407, 0.3157946467399597, -0.27740445733070374, -0.20177383720874786, 0.003569364547729492, -0.015546571463346481, 0.09739150106906891, -0.19562101364135742, 0.2981414198875427, -0.2475421279668808, -0.021107912063598633, 0.043806929141283035, -0.29938891530036926, -0.20705442130565643, 0.2779179811477661, 0.09650816023349762, -0.45383507013320923, -0.017878426238894463, 0.04868265241384506, -0.5866076350212097, -0.006349720060825348, -0.018574371933937073, -0.11635701358318329, 0.11945286393165588, -0.382311075925827, 0.01369631290435791, -0.44516515731811523, -0.3795369267463684, -0.10745582729578018, -0.15624310076236725, 0.2226879596710205, -0.3308177590370178, -0.2761637568473816, -0.3116797208786011, -0.34300097823143005, 0.15970644354820251, -0.2424597293138504, -0.09673304855823517, 0.09309795498847961, -0.35627588629722595, -0.21318884193897247, 0.51652991771698, -0.6141237020492554, -0.31686344742774963, 0.47944414615631104, -0.3933299779891968, -0.31927359104156494, 0.24876073002815247, -0.020787766203284264, 0.09857186675071716, -0.4370945990085602, -0.12158043682575226, -0.015840131789445877, -0.01884741149842739, -0.16953249275684357, -0.21318617463111877, -0.3432087004184723, 0.15771746635437012, 0.2769617438316345, -0.06715589761734009, 0.1459978222846985, 0.1615191102027893, 0.1352694183588028, 0.3999654948711395, 0.19973193109035492, 0.15013937652111053, -0.17016687989234924, 0.24233378469944, -0.2819077968597412, -0.03480280563235283, -0.06494960188865662, -0.6823077201843262, 0.20384737849235535, -0.08313953876495361, 0.16828688979148865, 0.0175626203417778, -0.010356884449720383, -0.5097718834877014, -0.004371738061308861, 0.0005457699298858643, -0.18655158579349518, -0.07908060401678085, 0.1180780753493309, 0.17097052931785583, -0.06011899560689926, -0.029049022123217583, 0.008818209171295166, -0.2176758199930191, 0.19490456581115723, 0.06572453677654266, 0.021833261474967003, -0.04672509431838989, 0.05511993542313576, 0.20559047162532806, 0.27329277992248535, 0.057880252599716187, -0.34605303406715393, -0.007602706551551819, 0.27380260825157166, 0.37539637088775635, 0.24678641557693481, 0.1457490473985672, 0.03405618667602539, 0.20076186954975128, 0.24317732453346252, 0.27019351720809937, 0.24104556441307068, 0.013897025026381016, -0.18850594758987427, 0.04307830333709717, 0.19351914525032043, 0.16598354279994965, -0.03977649658918381, -0.16236159205436707, -0.028106730431318283, 0.27386921644210815, 0.16490115225315094, -0.16081154346466064, -0.32471802830696106, 0.016357697546482086, 0.3211113512516022, 0.6261170506477356, 0.021359853446483612, 0.22093909978866577, -0.10340329259634018, -0.24081888794898987, -0.36838605999946594, -0.03585381060838699, 0.1324320137500763, -0.14667385816574097, -0.13123926520347595, 0.03457293286919594, 0.2909606397151947, 0.813957154750824, -0.05824141949415207, 0.14619754254817963, 0.09321149438619614, -0.0868021696805954, -0.21096459031105042, 0.13557134568691254, 0.13352090120315552, 0.08008913695812225, 0.13520997762680054, 0.024450553581118584, -0.33166074752807617, -0.012322034686803818, 0.05064729228615761, 0.42544808983802795, 0.3760313391685486, -0.3850039541721344, -0.14506055414676666, -0.1898254007101059, -0.41637134552001953, -0.46484336256980896, -0.26002487540245056, -0.24017611145973206, -0.32249775528907776, -0.11344344913959503, 0.32590705156326294, 0.017790652811527252, 0.20503288507461548, -0.1692851036787033, 0.2984495162963867, -0.29495880007743835, -0.2706928849220276, 0.12766197323799133, -0.20652709901332855, -0.3765334486961365, -0.00597800500690937, 0.22037088871002197, -0.059067513793706894, 0.42291051149368286, -0.15566517412662506, -0.30039364099502563, -0.15991686284542084, -0.3217743933200836, 0.143232062458992, 0.16493967175483704, 0.5009920597076416, 0.4258585274219513, -0.019074521958827972, 0.46502506732940674, 0.18073001503944397, 0.029476545751094818, -0.1994137018918991, -0.4520101249217987, -0.2094363123178482, -0.013931268826127052, 0.09201016277074814, -0.09755668044090271, -0.3627033829689026, -0.6268526315689087, -0.049434974789619446, 0.09246359020471573, 0.046491969376802444, 0.08684596419334412, 0.07008558511734009, -0.17430393397808075, 0.1262962818145752, 0.06489109992980957, -0.14631466567516327, -0.13924884796142578, 0.12885414063930511, -0.06379900872707367, -0.1144397035241127, -0.25293752551078796, 0.08709757030010223, -0.09722965955734253, -0.02210501953959465, -0.06927597522735596, -0.41635191440582275, -0.4980792999267578, 0.18607714772224426, 0.12020492553710938, 0.3934658169746399, -0.3054565191268921, 0.3518294095993042, 0.22900041937828064, 0.1961493343114853, -0.07248476892709732, -0.34929347038269043, 0.21882480382919312, 0.11768500506877899, 0.21085339784622192, 0.3140135109424591, 0.39684075117111206, -0.17006197571754456, 0.6948559880256653, 0.1315314918756485, -0.17439702153205872, -0.08181397616863251, -0.18463508784770966, 0.026518724858760834, -0.07165741175413132, 0.025269342586398125, 0.2892470061779022, -0.12184956669807434, -0.3842000961303711, 0.23318777978420258, -0.17590048909187317, -0.04469316452741623, -0.18192744255065918, 0.30313539505004883, 0.04550124704837799, -0.023568280041217804, -0.017659490928053856, -0.18742920458316803, 0.062108974903821945, 0.06367619335651398, -0.038225844502449036, -0.16917134821414948, -0.19289445877075195, 0.006161367520689964, 0.3494342565536499, 0.0028035715222358704, 0.20179885625839233, -0.23703844845294952, 0.06796187162399292, -0.3249555230140686, 0.3246842324733734, 0.3511451184749603, 0.6161120533943176, 0.12353888154029846, 0.006404511630535126, 0.12947885692119598, 0.045251764357089996, 0.16851752996444702, 0.015866238623857498, 0.20254437625408173, 0.1941901445388794, -0.15339314937591553, -0.10331499576568604, -0.2794736623764038, -0.05562499165534973, 0.10736370086669922, -0.014489255845546722, 0.20647253096103668, -0.37307024002075195, -0.16325289011001587, 0.21614541113376617, 0.13574643433094025, -0.10049250721931458, -0.33303171396255493, -0.08526661247015, -0.01718934066593647, -0.032943207770586014, -0.28790518641471863, -0.2201475203037262, -0.0035184796433895826, -0.17048873007297516, 0.3609248995780945, -0.00987154059112072, -0.15480631589889526, -0.045074060559272766, 0.12206505239009857, 0.1488848328590393, -0.28252148628234863, 0.30649590492248535, 0.2818530797958374, 0.10912124812602997, 0.42920392751693726, 0.2410028576850891, -0.07917197048664093, -0.2065410017967224, 0.06244774907827377, -0.008030018769204617, 0.38697096705436707, 0.366956889629364, 0.1010068953037262, -0.04046178609132767, -0.19147416949272156, -0.18607087433338165, -0.35990071296691895, 0.10493271052837372, 0.3145606517791748, -0.07839662581682205, -0.08095967769622803, -0.37498563528060913, 0.6471115946769714, 0.18893469870090485, -0.2820153832435608, 0.4349672198295593, -0.29321223497390747, -0.24159453809261322, 0.016749652102589607, 0.11472064256668091, 0.9402281045913696, 0.10115319490432739, 0.1793755739927292, 0.4074016213417053, -0.2569122314453125, 0.2777276933193207, 0.20907984673976898, 0.09308325499296188, -0.5795201659202576, 0.03690945729613304, 0.006617432460188866, -0.18128754198551178, 0.22857549786567688, -0.10842613875865936, -0.14125262200832367, 0.3231203854084015, 0.06759403645992279, 0.34466981887817383, 0.29964393377304077, 0.3267166018486023, -0.13545875251293182, 0.038733746856451035, -0.1763441562652588, 0.014816254377365112, -0.1632358729839325, 0.3315982520580292, -0.2451387345790863, 0.02070743218064308, -0.20245015621185303, -0.0634014755487442, -0.26311010122299194, -0.20407173037528992, -0.14579352736473083, 0.02936004474759102, 0.05805596709251404, -0.17235895991325378, -0.003578571602702141, 0.2685673236846924, 0.18531253933906555, -0.28354403376579285, -0.11798523366451263, 0.0657288059592247, -0.14783445000648499, 0.12417230755090714, 0.3541322350502014, 0.09130000323057175, 0.324359267950058, -0.3461558222770691, -0.04170437902212143, 0.07845711708068848, -0.3147277235984802, -0.11034661531448364, -0.1682370901107788, -0.1697041094303131, 0.17790326476097107, -0.35208582878112793, -0.2943044900894165, -0.14130601286888123, 0.028502169996500015, -0.052879318594932556, -0.021204594522714615, 0.2687152326107025, -0.24451056122779846, -0.011231012642383575, -0.13838420808315277, -0.24387352168560028, 0.014475924894213676, 0.44725728034973145, 0.03498488664627075, -0.10985312610864639, 0.6268347501754761, -0.200073704123497, 0.07830161601305008, -0.08096441626548767, 0.04233304411172867, -0.03040461242198944, -0.13849696516990662, 0.07945126295089722, -0.06528531759977341, -0.3092535138130188, 0.03099972754716873, 0.6317120790481567, 0.4538189768791199, -0.022200830280780792, -0.018440190702676773, -0.4137721061706543, -0.4684459865093231, 0.060229018330574036, -0.13376890122890472, 0.15528979897499084, -0.14459756016731262, 0.1882135421037674, 0.1062481701374054, -0.04616036266088486, -0.1483946144580841, -0.15582576394081116, 0.17581810057163239, 0.22977690398693085, -0.18760833144187927, 0.04707951843738556, -0.02007327973842621, 0.042338963598012924, -0.06394980102777481, -0.0045035220682621, -0.373477041721344, -0.0004291422665119171, -0.06305216997861862, 0.2435455471277237, 0.10599242150783539, -0.10039091110229492, 0.10395446419715881, -0.19135868549346924, -0.013094846159219742, 0.2529705762863159, -0.04260454326868057, 0.09640418738126755, -0.08035597205162048, 0.1069931611418724, 0.3446801006793976, 0.15482018887996674, -0.00897163338959217, 0.3556568920612335, 0.08667542040348053, 0.302620530128479, -0.0783776044845581, 0.318191260099411, -0.25813016295433044, 0.07906076312065125, 0.18990476429462433, 0.15776759386062622, -0.01704178750514984, 0.20191439986228943, 0.11414919793605804, -0.17145690321922302, 0.14508678019046783, -0.04361787438392639, 0.4042547643184662, 0.8169977068901062, -0.23315949738025665, -0.20942068099975586, 0.39549365639686584, -0.006465056911110878, -0.20162171125411987, 0.04672809690237045, 0.27429765462875366, 0.16639955341815948, -0.024561770260334015, 0.31011277437210083, 0.09837274253368378, -0.016525067389011383, 0.13899202644824982, 0.20030370354652405, 0.1752067506313324, -0.25456076860427856, -0.014316165819764137, 0.43618109822273254, -0.19119298458099365, 0.2348250150680542, 0.3925989866256714, 0.10050824284553528, 0.19480380415916443, 0.09433528780937195, 0.08113208413124084, 0.5245576500892639, 0.12521156668663025, 0.10351605713367462, 0.1761600524187088, -0.013916222378611565, 0.026047738268971443, 0.40117451548576355, -0.0035628527402877808, -0.04608459770679474, 0.40405765175819397, 0.607949435710907, -0.38512474298477173, -0.2014325112104416, -0.2968820035457611, -0.19066695868968964, 0.051356736570596695, -0.13128788769245148, -0.2509273886680603, -0.02136513963341713, -0.18068495392799377, -0.06044846773147583, -0.24239742755889893, -0.2871664762496948, 0.23147481679916382, -0.004475004971027374, -0.27352219820022583, -0.281095415353775, -0.36880767345428467, 0.13310948014259338, -0.02977117896080017, -0.24174059927463531, 0.2890284061431885, -0.010914258658885956, 0.04701438546180725, 0.39378583431243896, 0.2517334520816803, 0.4028513729572296, 0.13960056006908417, -0.3574991822242737, -0.28657081723213196, -0.2061855047941208, 0.05763343721628189, 0.04515703022480011, 0.3485932946205139, -0.0030630454421043396, -0.13039466738700867, 0.47411972284317017, 0.029031751677393913, 0.106221504509449, 0.2274838089942932, 0.3550854027271271, -0.46599534153938293, -0.10022158920764923, 0.8348633050918579, -0.08329667896032333, -0.2012389898300171, 0.04727889969944954, 0.17994128167629242, -0.201237291097641, 0.04935801029205322, -0.06658066809177399, 0.008687353692948818, 0.16352641582489014, 0.14323116838932037, -0.0007180701941251755, -0.09857748448848724, 0.2886528968811035, 0.3661203682422638, 0.025453779846429825, -0.2843555212020874, -0.2156105786561966, -0.7466064095497131, 0.1942780315876007, -0.056634511798620224, -0.3099117577075958, -0.021589912474155426, -0.25138479471206665, 0.2322976291179657, 0.061979860067367554, -0.17731793224811554, 0.18057352304458618, 0.09646039456129074, 0.10784228146076202, -0.19800600409507751, -0.27597445249557495, -0.2155531495809555, 0.022782862186431885, -0.10384839028120041, -0.42491912841796875, 0.11890013515949249, -0.3571050465106964, -0.15924254059791565, 0.08887829631567001, 0.017661772668361664, 0.4647807478904724, 0.005589578300714493, 0.2521131634712219, -0.0838739424943924, 0.7441830635070801, 0.2783360183238983, -0.06739955395460129, -0.16776920855045319, 0.12654736638069153, -0.1436951458454132, 0.1740298867225647, -0.12162190675735474, 0.0844416692852974, -0.1829201877117157, 0.5333899855613708, -0.41235408186912537, 0.4977741241455078, -0.12797129154205322, 0.0690179243683815, -0.3403317928314209, 0.059882499277591705, -0.26378893852233887, 0.031390152871608734, 0.30836620926856995, 0.1954003870487213, -0.03780286759138107, 0.30247578024864197, -0.22645694017410278, -0.12371501326560974, 0.4578656256198883, -0.17120201885700226, -0.03230812028050423, 0.08505763113498688, 0.2551974654197693, 0.05695349723100662, -0.14033180475234985, -0.48256686329841614, 0.1895676553249359, 0.24742929637432098, -0.09605929255485535, 0.12113316357135773, 0.16393177211284637, 0.2560332119464874, 0.014003757387399673, -0.023637229576706886, 1.0229697227478027, -0.111495740711689, 0.08293845504522324, -0.1492394059896469, -0.23327721655368805 ]
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 ?
[ -0.07840734720230103, 0.21769899129867554, -0.00799448974430561, -0.21871642768383026, -0.016750335693359375, -0.15013828873634338, 0.45485737919807434, 0.3005519211292267, 0.01723051816225052, 0.06836661696434021, 0.2465081661939621, 0.07947704941034317, -0.3443646728992462, 0.09023884683847427, 0.06199255958199501, 0.004056885838508606, 0.11398136615753174, 0.08718730509281158, -0.06385979056358337, 0.15320876240730286, -0.23857025802135468, -0.046687252819538116, 0.058539554476737976, 0.18519851565361023, 0.010334712453186512, 0.11389532685279846, -0.24757279455661774, 0.08724653720855713, -0.4137381315231323, -0.8389006853103638, 0.3315029442310333, 0.16535544395446777, 0.38253968954086304, 0.4212987422943115, -0.0001269610074814409, -0.09459542483091354, 0.6248884201049805, -0.02055453322827816, -0.602401077747345, -0.16927851736545563, 0.3652231693267822, -0.3112027049064636, 0.4110460877418518, 0.024779781699180603, 0.16198405623435974, -0.2027106136083603, 0.1706407368183136, 0.16206806898117065, 0.131315216422081, 0.13344086706638336, 0.06785380095243454, 0.22151798009872437, 0.05243736132979393, -0.14496354758739471, -0.19909533858299255, 0.0519704595208168, -0.06580491364002228, 0.6167840361595154, 0.5382672548294067, 0.177215576171875, -0.02148120105266571, -0.002134070498868823, 0.14147163927555084, 0.328044056892395, 0.20711109042167664, -0.07337615638971329, 0.7348245978355408, 0.06136088818311691, -0.13098597526550293, 0.00815608724951744, 0.9527407288551331, -0.4113694429397583, -0.5182245373725891, 0.055669642984867096, 0.04883920028805733, -0.1310153603553772, 0.18289068341255188, -0.16959063708782196, -0.1973831057548523, 0.10677914321422577, -0.18128953874111176, -0.3202578127384186, -0.22140301764011383, 0.28122878074645996, -0.2812548875808716, 0.22530825436115265, 0.0833771601319313, 0.24231508374214172, -0.10925690829753876, -0.17242580652236938, 0.26023194193840027, -0.09966812282800674, -0.018706846982240677, 0.19757528603076935, 0.1821487843990326, -0.3591311275959015, -0.3890572488307953, -0.11086901277303696, 0.14589783549308777, -0.10912869870662689, -0.39681699872016907, -0.2765677273273468, -0.05124089494347572, 0.09083911031484604, 0.25132492184638977, 0.2249232530593872, 0.41865625977516174, 0.1173015758395195, 0.23796439170837402, 0.028204115107655525, 0.38567882776260376, 0.27554240822792053, -0.03827491030097008, -0.3110353946685791, 0.03038856014609337, 0.0722004845738411, 0.18404839932918549, -0.40609803795814514, 0.04705127328634262, -0.23230332136154175, -0.34174269437789917, -0.13881908357143402, -0.1190888062119484, 0.18631941080093384, -0.1159299835562706, 0.0003313906490802765, -0.007226672023534775, 0.3783921003341675, -0.12641994655132294, -0.18677881360054016, -0.02923940122127533, -0.057438816875219345, -0.13046739995479584, -0.0203388798981905, 0.19088779389858246, 0.03255531191825867, 0.28977516293525696, -0.2891445755958557, -0.18417240679264069, 0.020452052354812622, -0.08764714002609253, 0.12490960955619812, -0.2652386426925659, 0.26830825209617615, -0.40122637152671814, -0.01537303440272808, -0.07192283868789673, -0.36129653453826904, -0.15017588436603546, 0.2882692515850067, 0.07951914519071579, -0.43441006541252136, -0.015946442261338234, 0.052783165127038956, -0.48552602529525757, -0.012248285114765167, -0.02927067130804062, -0.05822480842471123, 0.14872422814369202, -0.32598841190338135, 0.043539706617593765, -0.4276154041290283, -0.32640188932418823, -0.1250613033771515, -0.20648102462291718, 0.23732170462608337, -0.3779642879962921, -0.24657811224460602, -0.21628716588020325, -0.3262249529361725, 0.051763907074928284, -0.20617172122001648, -0.09009283035993576, 0.034393928945064545, -0.14357246458530426, -0.24087359011173248, 0.525596022605896, -0.6463609337806702, -0.2875416874885559, 0.5019087791442871, -0.2739056348800659, -0.2755510210990906, 0.252553254365921, -0.024601487442851067, 0.1583230346441269, -0.5408313870429993, -0.15352262556552887, -0.10466593503952026, -0.038840390741825104, -0.17798306047916412, -0.33155345916748047, -0.42146816849708557, 0.20058287680149078, 0.3345435857772827, -0.14667843282222748, 0.10944339632987976, 0.09240778535604477, 0.07713993638753891, 0.5037010312080383, 0.1834126114845276, 0.05109693109989166, -0.2177705317735672, 0.3656325042247772, -0.23294924199581146, -0.07816824316978455, -0.014533840119838715, -0.7206133008003235, 0.18981388211250305, -0.10229069739580154, 0.09577057510614395, 0.13708694279193878, 0.005196623504161835, -0.43599677085876465, -0.023731933906674385, -0.013888668268918991, -0.2627043128013611, -0.0815754309296608, -0.06165136396884918, 0.05206908658146858, -0.029485389590263367, -0.010522857308387756, 0.02294955775141716, -0.20240578055381775, 0.22217687964439392, 0.08579118549823761, -0.0024010632187128067, -0.09218499064445496, 0.04781786724925041, 0.15576091408729553, 0.293927937746048, 0.07502377033233643, -0.30252036452293396, -0.020342204719781876, 0.3204720914363861, 0.380601704120636, 0.2452130913734436, 0.11209166049957275, 0.10057678818702698, 0.29977232217788696, 0.3446667790412903, 0.324491411447525, 0.1484437733888626, -0.04383333399891853, -0.09661421179771423, -0.002046257257461548, 0.19905754923820496, 0.19929979741573334, -0.15083476901054382, -0.24920322000980377, -0.08458079397678375, 0.18606244027614594, 0.2040824294090271, -0.22944794595241547, -0.298494815826416, -0.02239767089486122, 0.3367256820201874, 0.5842093229293823, 0.010334759950637817, 0.18908604979515076, -0.035991378128528595, -0.22549137473106384, -0.381384015083313, -0.11914383620023727, 0.1591705083847046, -0.15317359566688538, -0.17685851454734802, 0.03739115968346596, 0.2746085524559021, 0.6558879613876343, -0.05080951005220413, 0.06169295310974121, 0.07968907058238983, -0.09127530455589294, -0.15820492804050446, 0.13297709822654724, 0.19518598914146423, 0.07974904775619507, 0.07448263466358185, 0.02701413445174694, -0.22329097986221313, 0.05465976148843765, 0.0741833746433258, 0.29648903012275696, 0.30123257637023926, -0.30336418747901917, -0.1087808758020401, -0.14369496703147888, -0.3550179898738861, -0.44845256209373474, -0.2716017961502075, -0.36121875047683716, -0.2839103639125824, -0.04440955072641373, 0.3055323660373688, -0.008049050346016884, 0.16161353886127472, -0.0884672999382019, 0.35138940811157227, -0.30625659227371216, -0.08021794259548187, 0.13091756403446198, -0.16871540248394012, -0.4021384119987488, -0.0075457170605659485, 0.1693287491798401, -0.1414065659046173, 0.39903172850608826, -0.07358181476593018, -0.31047070026397705, -0.07312048971652985, -0.4179686903953552, 0.07629488408565521, 0.1368345022201538, 0.5680723786354065, 0.4437888562679291, -0.015858083963394165, 0.4559708833694458, 0.21411621570587158, 0.02653995342552662, -0.18789558112621307, -0.37476012110710144, -0.2456979751586914, 0.002451459178701043, 0.062213171273469925, -0.0628884807229042, -0.34115564823150635, -0.674794614315033, 0.01595253497362137, 0.14174193143844604, 0.04154038056731224, 0.09176705777645111, 0.09137057512998581, -0.2075149267911911, 0.009894395247101784, -0.033491384238004684, -0.12606926262378693, -0.09700173139572144, 0.1426059752702713, -0.11858950555324554, -0.11486150324344635, -0.21240977942943573, 0.09414660930633545, 0.05003800243139267, -0.04971389099955559, -0.1521892100572586, -0.38123035430908203, -0.5387279987335205, 0.27524423599243164, 0.0959395170211792, 0.44321736693382263, -0.37508341670036316, 0.3168640434741974, 0.25746065378189087, 0.2106017917394638, -0.030558090656995773, -0.3124099373817444, 0.22272086143493652, 0.20246106386184692, 0.23872721195220947, 0.3416236340999603, 0.4278688430786133, -0.17156076431274414, 0.698760449886322, 0.26078981161117554, -0.21202413737773895, -0.23117151856422424, -0.0054813530296087265, -0.12854982912540436, -0.0936509296298027, 0.03923318535089493, 0.26029449701309204, -0.11454686522483826, -0.36592668294906616, 0.28301018476486206, -0.07185623794794083, 0.000910244882106781, -0.17153982818126678, 0.23263579607009888, 0.048266418278217316, -0.025847740471363068, -0.01699184440076351, -0.10419665277004242, 0.09018269926309586, 0.036169275641441345, -0.061963461339473724, -0.16355231404304504, -0.24948880076408386, 0.0705663412809372, 0.45116740465164185, -0.0071663036942481995, 0.23893849551677704, -0.1418859213590622, 0.14389027655124664, -0.3231486976146698, 0.29190781712532043, 0.3646889626979828, 0.5911163091659546, 0.20548398792743683, -0.06498757004737854, 0.16462990641593933, 0.03949234262108803, 0.16430622339248657, 0.06811744719743729, 0.22538653016090393, 0.08400434255599976, -0.23833215236663818, -0.08051209151744843, -0.3215267062187195, 0.0013930052518844604, 0.043014269322156906, 0.0670982152223587, 0.207816019654274, -0.3719228506088257, -0.22546428442001343, 0.1916746348142624, 0.1168142780661583, -0.1499803066253662, -0.32392820715904236, -0.04555001109838486, 0.13843384385108948, 0.07787726074457169, -0.3222935199737549, -0.31121325492858887, -0.0988617017865181, -0.23056131601333618, 0.3555451035499573, -0.003292117267847061, -0.1536957025527954, -0.06030814349651337, 0.1760215163230896, 0.12314103543758392, -0.35025733709335327, 0.32314881682395935, 0.30365249514579773, 0.1975913792848587, 0.41991445422172546, 0.11987721174955368, -0.12025178968906403, -0.10920153558254242, -0.038991253823041916, -0.09179171919822693, 0.33053553104400635, 0.44662657380104065, 0.02940857782959938, -0.07550718635320663, -0.15196441113948822, -0.21442179381847382, -0.369579017162323, -0.03050500713288784, 0.28523194789886475, -0.049796491861343384, -0.06767596304416656, -0.34444305300712585, 0.6544525623321533, 0.1329580396413803, -0.3539952039718628, 0.38949456810951233, -0.26032763719558716, -0.20492435991764069, 0.014469064772129059, 0.12303577363491058, 0.9271784424781799, 0.13575266301631927, 0.22614456713199615, 0.35611456632614136, -0.26288020610809326, 0.19106149673461914, 0.2073080688714981, -0.021173972636461258, -0.5951223969459534, 0.07956109941005707, -0.01045474037528038, -0.11937639117240906, 0.25975117087364197, -0.06818681210279465, -0.0672571063041687, 0.3064366281032562, 0.11517393589019775, 0.3037260174751282, 0.27472925186157227, 0.31663256883621216, -0.0745704397559166, 0.12467688322067261, -0.14576467871665955, -0.002733757719397545, -0.19871364533901215, 0.22155605256557465, -0.2173815667629242, -0.05912107229232788, -0.13456155359745026, -0.13533854484558105, -0.19719448685646057, -0.27885568141937256, -0.1417296677827835, 0.002587556838989258, 0.09939530491828918, -0.2422436624765396, 0.09887169301509857, 0.20783917605876923, 0.21243155002593994, -0.2769320011138916, -0.10583136975765228, -0.006890475749969482, -0.09415439516305923, 0.19183050096035004, 0.23260439932346344, 0.11455319821834564, 0.4135674238204956, -0.3656522035598755, -0.029833689332008362, -0.0021504461765289307, -0.30152446031570435, -0.07882071286439896, -0.24475958943367004, -0.22926904261112213, 0.20691938698291779, -0.34996190667152405, -0.24335049092769623, -0.07171539962291718, 0.060952719300985336, -0.04963530972599983, -0.016589514911174774, 0.2987672984600067, -0.16764627397060394, 0.0024614334106445312, -0.06528228521347046, -0.14836618304252625, 0.0035221297293901443, 0.4236816167831421, 0.03534163162112236, -0.1286199986934662, 0.5450732707977295, -0.17769336700439453, -0.012218818068504333, -0.06524503976106644, 0.0422201007604599, -0.0569339320063591, -0.2640582025051117, 0.024849386885762215, -0.08269265294075012, -0.254199355840683, 0.026647554710507393, 0.6445192098617554, 0.5140349864959717, -0.08816101402044296, 0.008511319756507874, -0.3115549087524414, -0.4337587356567383, 0.11887909471988678, -0.14966128766536713, 0.1279405653476715, -0.2016347050666809, 0.19356271624565125, 0.08962829411029816, -0.08079346269369125, -0.1644304394721985, -0.16892458498477936, 0.28728964924812317, 0.16615432500839233, -0.20746590197086334, 0.05575067922472954, -0.05666795000433922, 0.00035731587558984756, -0.07731951028108597, -0.09144656360149384, -0.4320649802684784, -0.009923234581947327, -0.0681835412979126, 0.22385188937187195, 0.09701959043741226, -0.0993722602725029, 0.07501251995563507, -0.1820572167634964, 0.032962217926979065, 0.33374613523483276, -0.029270030558109283, 0.05774620175361633, 0.040878526866436005, 0.2147265374660492, 0.3124197721481323, 0.2434011697769165, -0.026759691536426544, 0.41367027163505554, 0.052882254123687744, 0.27785611152648926, -0.031204234808683395, 0.30528852343559265, -0.2380792200565338, 0.03763595223426819, 0.07074977457523346, 0.19237913191318512, -0.1881691962480545, 0.32301491498947144, 0.044290460646152496, -0.19937767088413239, 0.17659614980220795, 0.008234193548560143, 0.3879280090332031, 0.7847013473510742, -0.19644686579704285, -0.300943523645401, 0.45363563299179077, 0.027808213606476784, -0.12452957779169083, 0.08317085355520248, 0.262359619140625, 0.19259516894817352, -0.04848910868167877, 0.33698800206184387, 0.14231334626674652, 0.037863247096538544, 0.10244549810886383, 0.23270544409751892, 0.11558371782302856, -0.3071414828300476, -0.014103030785918236, 0.4850953221321106, -0.19111378490924835, 0.25854936242103577, 0.42612993717193604, 0.2196153700351715, 0.042595185339450836, 0.01736105978488922, 0.08998870104551315, 0.40315714478492737, 0.20004752278327942, 0.04172372817993164, 0.21260783076286316, 0.07285183668136597, -0.05200717970728874, 0.3586235046386719, 0.010092176496982574, -0.07322878390550613, 0.43289124965667725, 0.6263275742530823, -0.24576957523822784, -0.2735118865966797, -0.265232652425766, -0.2573288381099701, 0.11760283261537552, -0.15073952078819275, -0.3253525197505951, -0.031386084854602814, -0.172324538230896, -0.06642265617847443, -0.3928256630897522, -0.23233743011951447, 0.24641136825084686, 0.060927681624889374, -0.2492004781961441, -0.35219913721084595, -0.4377782344818115, 0.042293112725019455, 0.06692017614841461, -0.14867763221263885, 0.24843816459178925, 0.003967605531215668, 0.09395244717597961, 0.4052654802799225, 0.24692629277706146, 0.24045298993587494, 0.09698637574911118, -0.39798590540885925, -0.33229953050613403, -0.2425733357667923, 0.0760229080915451, 0.16924098134040833, 0.2883455455303192, 0.02029649168252945, -0.1690838783979416, 0.43928295373916626, 0.004951445385813713, 0.1724262833595276, 0.36718422174453735, 0.30420124530792236, -0.41586968302726746, -0.11310699582099915, 0.866989016532898, -0.01036369800567627, -0.2329448163509369, 0.14805153012275696, 0.19929929077625275, -0.20762670040130615, 0.12630227208137512, -0.041104625910520554, 0.05794176459312439, 0.128401517868042, 0.15669363737106323, 0.012644585222005844, -0.03198806196451187, 0.18832483887672424, 0.3991011083126068, -0.01953880488872528, -0.2369244396686554, -0.12392812967300415, -0.5654717087745667, 0.07696305215358734, 0.051341086626052856, -0.21406476199626923, -0.018772482872009277, -0.24497383832931519, 0.20464445650577545, 0.06571180373430252, -0.11546401679515839, 0.022159021347761154, -0.05080560967326164, 0.14864704012870789, -0.17918580770492554, -0.19621114432811737, -0.13138964772224426, 0.1199856549501419, -0.1198875829577446, -0.30400794744491577, 0.05137854814529419, -0.3177834749221802, -0.19125229120254517, 0.018191250041127205, 0.047366246581077576, 0.4646492600440979, 0.03857523575425148, 0.13066107034683228, -0.06797149032354355, 0.5171793699264526, 0.3847469091415405, -0.17758965492248535, -0.1369655728340149, 0.10956788063049316, -0.12803296744823456, 0.18178817629814148, -0.13699199259281158, 0.1513938158750534, -0.122855544090271, 0.4253520369529724, -0.3331281542778015, 0.41835877299308777, -0.22756367921829224, 0.0001626563025638461, -0.4170818626880646, 0.10477407276630402, -0.11555361747741699, -0.11772534996271133, 0.2918640971183777, 0.27572381496429443, -0.07171913981437683, 0.2615319788455963, -0.2811012268066406, -0.07436523586511612, 0.4302865266799927, -0.22998549044132233, 0.04756632819771767, 0.09626506268978119, 0.2981060743331909, 0.11599423736333847, -0.05385667085647583, -0.531058132648468, 0.12829163670539856, 0.23650754988193512, -0.14698582887649536, 0.022672098129987717, 0.2222898006439209, 0.3116585314273834, -0.06905553489923477, -0.03913304582238197, 1.0794569253921509, -0.17351484298706055, 0.11268402636051178, -0.15644319355487823, -0.19749954342842102 ]
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
[ -0.14118072390556335, 0.2805648446083069, -0.010229086503386497, -0.1734384149312973, 0.09962055087089539, -0.05622601509094238, 0.3951389193534851, 0.35677868127822876, 0.1325157731771469, 0.1633101850748062, 0.2544320225715637, 0.11689873039722443, -0.26364362239837646, -0.032296013087034225, 0.08826718479394913, -0.022581331431865692, 0.024319849908351898, 0.153996542096138, -0.06924255192279816, 0.253808856010437, -0.15772663056850433, 0.02764403447508812, -0.07621055841445923, 0.0034688711166381836, -0.03206753730773926, 0.14956483244895935, -0.12435545027256012, -0.02923256903886795, -0.39765340089797974, -0.8835684061050415, 0.38515710830688477, 0.08030331134796143, 0.15719950199127197, 0.4587191045284271, -0.0001239301636815071, -0.09218358993530273, 0.6603642106056213, -0.09545937180519104, -0.7126159071922302, -0.11028550565242767, 0.21409562230110168, -0.44317543506622314, 0.31489816308021545, -0.12270475178956985, 0.07477682828903198, -0.12571555376052856, 0.24232164025306702, 0.0029502809047698975, 0.04234407842159271, 0.2779586613178253, 0.11370958387851715, 0.20985466241836548, 0.14470228552818298, -0.20508523285388947, -0.06326273828744888, 0.07268209010362625, -0.1255190372467041, 0.5561835765838623, 0.29881545901298523, 0.08979629725217819, -0.011444158852100372, -0.08872745931148529, 0.08542583882808685, 0.47456541657447815, 0.2024761140346527, -0.0256972573697567, 0.5517069697380066, 0.010823531076312065, -0.16354912519454956, -0.0021503716707229614, 0.9606097936630249, -0.5900260806083679, -0.2856663763523102, 0.055983081459999084, 0.06181573495268822, -0.044158145785331726, 0.22932526469230652, -0.1933574676513672, -0.3710883855819702, 0.005593400448560715, -0.113864466547966, -0.34605908393859863, -0.13907605409622192, 0.3863377571105957, -0.2816322147846222, 0.3920343220233917, 0.12043490260839462, 0.2018035352230072, 0.029666583985090256, -0.12271088361740112, 0.1766301840543747, -0.037354256957769394, 0.05294102430343628, 0.2122683823108673, 0.053161948919296265, -0.29137104749679565, -0.36121830344200134, -0.23289954662322998, 0.11769244819879532, -0.0717003345489502, -0.5114047527313232, -0.21405303478240967, 0.060136254876852036, 0.21360686421394348, 0.11943654716014862, 0.2886669933795929, 0.3164389431476593, 0.26326677203178406, 0.2690865695476532, 0.1590999960899353, 0.23070356249809265, 0.22681641578674316, -0.001236175769008696, -0.39059746265411377, 0.00787830725312233, 0.043517276644706726, 0.23724471032619476, -0.27922511100769043, -0.001813456416130066, -0.223365917801857, -0.35482579469680786, -0.03768952563405037, -0.15496297180652618, 0.27120161056518555, -0.21224386990070343, -0.21906012296676636, 0.03549148142337799, 0.3669431209564209, -0.19033747911453247, -0.0829746425151825, -0.030062638223171234, 0.10269162058830261, -0.20946572721004486, 0.04133981466293335, 0.2736015319824219, 0.11987332999706268, 0.3815648853778839, -0.22116583585739136, -0.17527373135089874, -0.06266403198242188, 0.04889075458049774, 0.16503886878490448, -0.2584547996520996, 0.3138117790222168, -0.17547893524169922, 0.09148240834474564, 0.0757020115852356, -0.3986794352531433, -0.2606750428676605, 0.190580815076828, 0.08797687292098999, -0.39845409989356995, -0.07537218183279037, 0.09867958724498749, -0.5322721004486084, -0.04728294909000397, 0.03677947074174881, -0.10876646637916565, 0.16949871182441711, -0.43949902057647705, 0.039980269968509674, -0.3296203017234802, -0.32755377888679504, -0.12069296091794968, -0.04477449879050255, 0.22451508045196533, -0.3344626724720001, -0.3448535203933716, -0.3996001183986664, -0.262083500623703, 0.19108641147613525, -0.2722560465335846, -0.05161494389176369, 0.1798931062221527, -0.3446755111217499, -0.09361154586076736, 0.5451646447181702, -0.5609612464904785, -0.3666335344314575, 0.4103245735168457, -0.39714157581329346, -0.30935925245285034, 0.2160005271434784, -0.0019480688497424126, 0.072139672935009, -0.2954326272010803, -0.13503113389015198, 0.026698965579271317, 0.05478984862565994, -0.12957745790481567, -0.17708122730255127, -0.3129168748855591, 0.15713399648666382, 0.24047555029392242, -0.057767730206251144, 0.12762778997421265, 0.17676837742328644, 0.22123213112354279, 0.33956870436668396, 0.12165810167789459, 0.19460859894752502, -0.08479996770620346, 0.20399177074432373, -0.17558155953884125, -0.07350810617208481, -0.09873545169830322, -0.6286861896514893, 0.21739527583122253, -0.040479544550180435, 0.28844982385635376, -0.07102851569652557, -0.051276326179504395, -0.6030595898628235, -0.028186338022351265, 0.007144272327423096, -0.254384845495224, 0.0009690672159194946, 0.2109040468931198, 0.12575265765190125, 0.07182292640209198, 0.08742757141590118, -0.05625335872173309, -0.20948581397533417, 0.18995748460292816, -0.039936479181051254, -0.016076302155852318, -0.11674629896879196, 0.11367613822221756, 0.12232330441474915, 0.20108270645141602, 0.03238675370812416, -0.26659008860588074, -0.011400045827031136, 0.18859511613845825, 0.3017377555370331, 0.26696479320526123, 0.2012650966644287, -0.028640851378440857, 0.18482038378715515, 0.14330661296844482, 0.23561516404151917, 0.3190940022468567, 0.07314888387918472, -0.2795489430427551, 0.0007633194327354431, 0.13875505328178406, 0.21099036931991577, 0.0036713331937789917, -0.22517900168895721, -0.008010979741811752, 0.3341837525367737, 0.1293625384569168, -0.1294441968202591, -0.3318285048007965, 0.09353769570589066, 0.3465370535850525, 0.5930149555206299, -0.03065631352365017, 0.3187482953071594, -0.17704996466636658, -0.1961137056350708, -0.33191588521003723, 0.010572269558906555, 0.1472766101360321, -0.1484430730342865, -0.13274461030960083, 0.004782436415553093, 0.2607828974723816, 0.7473019957542419, -0.017134200781583786, 0.21024386584758759, 0.08281274884939194, -0.19435252249240875, -0.18114571273326874, 0.19616247713565826, 0.14606767892837524, 0.05849679559469223, 0.11732535809278488, 0.008353392593562603, -0.3501940071582794, -0.0421718992292881, -0.01063714548945427, 0.46164852380752563, 0.4212026000022888, -0.3261987566947937, -0.1254185289144516, -0.2086644470691681, -0.4111090302467346, -0.4504467248916626, -0.4019869863986969, -0.3150010406970978, -0.3542323410511017, -0.10665440559387207, 0.2702239751815796, 0.08679600059986115, 0.2829837501049042, -0.224715918302536, 0.30379197001457214, -0.25691190361976624, -0.35868990421295166, 0.07040300965309143, -0.2922021448612213, -0.3351595401763916, 0.011352062225341797, 0.233516663312912, 0.0032299216836690903, 0.4592945873737335, -0.12466802448034286, -0.3249701261520386, -0.22680646181106567, -0.36631810665130615, 0.10803449153900146, 0.06434579193592072, 0.5244491100311279, 0.41656604409217834, 0.04030180349946022, 0.47716838121414185, 0.04909319430589676, 0.034976787865161896, -0.18719081580638885, -0.424094557762146, -0.23610278964042664, -0.012634423561394215, 0.1769871711730957, -0.09271982312202454, -0.37856370210647583, -0.5373615622520447, -0.10313151776790619, 0.0880431979894638, 0.06720203161239624, 0.006694704294204712, 0.1528788059949875, -0.1076124906539917, 0.1514066904783249, 0.10712626576423645, -0.055705778300762177, -0.13396570086479187, 0.12558957934379578, 0.02072177082300186, -0.16886909306049347, -0.10974130034446716, 0.10617506504058838, -0.11159314215183258, 0.03736857697367668, -0.13643242418766022, -0.45776915550231934, -0.4050000309944153, 0.10826603323221207, 0.1651182919740677, 0.40959522128105164, -0.252712219953537, 0.2943410277366638, 0.23262062668800354, 0.14786069095134735, -0.09710405021905899, -0.3855840563774109, 0.1825857162475586, 0.03584716096520424, 0.25486335158348083, 0.23821192979812622, 0.42762523889541626, -0.18805336952209473, 0.8638371825218201, 0.12577833235263824, -0.0933363139629364, -0.06046298146247864, -0.2205599844455719, 0.003717418760061264, -0.05921005457639694, -0.044465601444244385, 0.32291167974472046, -0.11662009358406067, -0.32144469022750854, 0.27593040466308594, -0.20463067293167114, -0.15403355658054352, -0.13479188084602356, 0.28695449233055115, -0.006108589470386505, -0.0675116553902626, -0.0446653738617897, -0.14884768426418304, 0.04530714452266693, 0.054063793271780014, -0.049502961337566376, -0.22893069684505463, -0.25206342339515686, 0.04696158319711685, 0.31666630506515503, -0.15776339173316956, 0.1885860413312912, -0.24663028120994568, 0.031642619520425797, -0.44746971130371094, 0.2941854000091553, 0.33318576216697693, 0.5381502509117126, 0.10278929769992828, -0.015300169587135315, 0.11338121443986893, 0.1049145832657814, 0.20852705836296082, -0.017759324982762337, 0.1344524323940277, 0.24128973484039307, -0.09606370329856873, -0.12461607903242111, -0.25974878668785095, -0.008720126003026962, 0.1897033452987671, -0.004294808954000473, 0.13426710665225983, -0.4184403419494629, -0.07045483589172363, 0.2634961009025574, 0.19597165286540985, -0.09762413799762726, -0.3244107961654663, -0.08676270395517349, -0.028297312557697296, -0.08013050258159637, -0.23945486545562744, -0.16850070655345917, 0.10118240863084793, -0.07583743333816528, 0.3622826933860779, 0.036848463118076324, -0.06702408194541931, -0.0823664516210556, 0.15544454753398895, 0.146712988615036, -0.20712384581565857, 0.2648325562477112, 0.22875167429447174, 0.027473947033286095, 0.3402649164199829, 0.241371288895607, -0.11826840043067932, -0.2586958706378937, -0.047578878700733185, 0.007216406054794788, 0.35540905594825745, 0.23109319806098938, 0.1883850246667862, -0.019396789371967316, -0.1601121723651886, -0.14804227650165558, -0.3491378724575043, 0.19839347898960114, 0.32603761553764343, -0.09454551339149475, -0.08157745748758316, -0.3744667172431946, 0.6699296832084656, 0.2106671929359436, -0.2216612696647644, 0.5757378339767456, -0.21588154137134552, -0.361310750246048, -0.02096259407699108, 0.09729389101266861, 0.9390982389450073, 0.12304747104644775, 0.22184543311595917, 0.3271251916885376, -0.2036057412624359, 0.40682482719421387, 0.1139771044254303, 0.1631506383419037, -0.5522783994674683, 0.035590000450611115, 0.041692376136779785, -0.18077494204044342, 0.18681372702121735, -0.12864047288894653, -0.2766346335411072, 0.34582558274269104, 0.09094040095806122, 0.3162321150302887, 0.28049272298812866, 0.3736043870449066, -0.06282483041286469, -0.047387201339006424, -0.16777822375297546, 0.06752441823482513, -0.17312860488891602, 0.3932380676269531, -0.250163733959198, 0.09912212193012238, -0.1918768584728241, -0.03845309466123581, -0.31208086013793945, -0.19377604126930237, -0.12246156483888626, 0.12016215175390244, 0.05442022532224655, -0.1686798483133316, -0.02505229040980339, 0.22702550888061523, 0.12529388070106506, -0.16949120163917542, -0.17709821462631226, 0.14197885990142822, -0.1813591718673706, 0.06414823234081268, 0.30314627289772034, 0.05139261856675148, 0.3021053671836853, -0.3474116623401642, -0.09088625013828278, 0.12196370959281921, -0.2841305136680603, -0.20523595809936523, -0.1263769567012787, -0.1708373725414276, 0.08427345752716064, -0.3267577886581421, -0.282860666513443, -0.13302059471607208, 0.04689626768231392, -0.14486834406852722, 0.02616175264120102, 0.22552156448364258, -0.3177244961261749, 0.0012358054518699646, -0.03404851630330086, -0.3448730707168579, 0.018304821103811264, 0.44678616523742676, -0.08538536727428436, -0.07896117866039276, 0.7237139940261841, -0.1690613180398941, 0.038177311420440674, -0.11864368617534637, 0.023047514259815216, -0.18729455769062042, -0.15366312861442566, 0.049600694328546524, -0.05159308761358261, -0.2729284167289734, 0.01061205193400383, 0.5785098671913147, 0.4408789277076721, 0.04504683241248131, -0.02072543278336525, -0.4588911235332489, -0.41634103655815125, 0.007794685661792755, -0.18109852075576782, 0.2008616030216217, -0.08202992379665375, 0.18932278454303741, 0.15709927678108215, 0.01595485582947731, -0.187468022108078, -0.16336193680763245, 0.040294744074344635, 0.24360188841819763, -0.13370975852012634, 0.0225283894687891, -0.07007408887147903, 0.08745723962783813, 0.0018108636140823364, -0.0883050337433815, -0.28574225306510925, -0.05251786857843399, -0.13653428852558136, 0.2203616350889206, 0.0687347948551178, -0.002946962136775255, 0.0809541642665863, -0.19381718337535858, -0.005588049069046974, 0.1685536950826645, -0.05514583736658096, 0.10774029046297073, -0.09697437286376953, 0.19977815449237823, 0.33400067687034607, 0.048022352159023285, -0.019415859133005142, 0.2950708270072937, 0.1199374794960022, 0.2520233392715454, -0.04795244708657265, 0.3378598690032959, -0.19110234081745148, 0.06257529556751251, 0.17886307835578918, 0.2028523087501526, -0.0861780047416687, 0.17287001013755798, 0.0845031663775444, -0.10913452506065369, 0.10387429594993591, 0.04550378397107124, 0.3643854856491089, 0.7686900496482849, -0.26971736550331116, -0.1473635882139206, 0.433666467666626, 0.04313832148909569, -0.2070346474647522, -0.06151576340198517, 0.3559982180595398, 0.0967506617307663, 0.05956525728106499, 0.24292254447937012, -0.003486037254333496, -0.019756466150283813, 0.15557114779949188, 0.16882425546646118, 0.08332838118076324, -0.22388117015361786, -0.014182325452566147, 0.38112443685531616, -0.204329714179039, 0.16130280494689941, 0.3652931749820709, 0.12268346548080444, 0.2558835744857788, 0.11702345311641693, 0.12785866856575012, 0.4928240478038788, 0.06652931869029999, 0.03491068631410599, 0.07122965157032013, 0.07629594206809998, 0.061420440673828125, 0.2696564197540283, -0.09182748198509216, -0.021174706518650055, 0.37101098895072937, 0.551271378993988, -0.3462454676628113, -0.1316280961036682, -0.2795216143131256, -0.14258629083633423, 0.04645552113652229, -0.18091359734535217, -0.24025443196296692, 0.013755857944488525, -0.13747644424438477, -0.11026807129383087, -0.15492680668830872, -0.3587164282798767, 0.24380084872245789, 0.008351609110832214, -0.36951738595962524, -0.30901604890823364, -0.2120947539806366, 0.17734560370445251, -0.006553530693054199, -0.3145363926887512, 0.18164990842342377, -0.06015767157077789, 0.00930163636803627, 0.28603386878967285, 0.296482115983963, 0.45625102519989014, 0.08334749191999435, -0.4143468141555786, -0.2541944980621338, -0.1805916279554367, 0.06383518129587173, -0.02755061164498329, 0.34688523411750793, 0.037008341401815414, -0.1990613043308258, 0.4957922399044037, 0.06062592566013336, 0.0358571894466877, 0.13405011594295502, 0.4426795244216919, -0.5077688694000244, -0.12595930695533752, 0.6570250988006592, -0.06691066920757294, -0.17114922404289246, -0.014399982988834381, 0.2682860195636749, -0.31312260031700134, 0.09172575920820236, -0.0007790941745042801, -0.04264107346534729, 0.14234831929206848, 0.07496239244937897, 0.025010138750076294, -0.17087215185165405, 0.40392714738845825, 0.31939420104026794, 0.01959727704524994, -0.26244738698005676, -0.259154736995697, -0.7384799122810364, 0.19107267260551453, -0.11629621684551239, -0.34262752532958984, -0.07228414714336395, -0.19995804131031036, 0.24219250679016113, 0.04681190848350525, -0.1436450034379959, 0.10192464292049408, 0.029758833348751068, 0.14397859573364258, -0.18503157794475555, -0.3121945559978485, -0.20069928467273712, -0.02207382582128048, -0.06889434158802032, -0.40647274255752563, 0.11685904860496521, -0.36691635847091675, -0.09841718524694443, 0.07443760335445404, 0.0026448816061019897, 0.5323455333709717, 0.07270558178424835, 0.2942667007446289, -0.10212308168411255, 0.8140294551849365, 0.17402300238609314, -0.04890239238739014, -0.17330361902713776, 0.11475576460361481, -0.10631570219993591, 0.16286811232566833, -0.12736479938030243, 0.1119912713766098, -0.20888233184814453, 0.4868563711643219, -0.38907647132873535, 0.601815402507782, -0.017861688509583473, 0.08193861693143845, -0.2825641930103302, 0.011469272896647453, -0.25492197275161743, 0.08918756246566772, 0.2956137955188751, 0.195259690284729, -0.07665424048900604, 0.354806125164032, -0.15306319296360016, -0.2579878866672516, 0.41756850481033325, -0.16063252091407776, -0.11916562914848328, 0.03131815791130066, 0.22145994007587433, -0.012549281120300293, -0.09062881022691727, -0.43135032057762146, 0.1975601315498352, 0.3064979016780853, -0.06527065485715866, 0.03529031202197075, 0.19647520780563354, 0.22549034655094147, 0.09795331954956055, -0.005873385816812515, 0.9912326335906982, -0.10930561274290085, 0.11685295403003693, -0.1647682934999466, -0.2565416693687439 ]
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.
[ -0.0867203027009964, 0.2737462520599365, 0.011248860508203506, -0.14076009392738342, 0.06271111220121384, -0.048166945576667786, 0.3938751518726349, 0.3804904520511627, 0.08933477103710175, 0.20729276537895203, 0.2584836483001709, 0.14728564023971558, -0.23994089663028717, -0.015442601405084133, 0.04751161113381386, -0.006765063852071762, -0.058515287935733795, 0.21216976642608643, -0.029032625257968903, 0.2738305926322937, -0.08489841967821121, -0.011124121025204659, 0.07523150742053986, 0.09017886966466904, -0.044721562415361404, 0.11217676103115082, -0.12271841615438461, -0.06276867538690567, -0.26378533244132996, -0.9248581528663635, 0.34500613808631897, 0.12437264621257782, 0.1424262821674347, 0.5053983926773071, -0.0001223480503540486, -0.053185708820819855, 0.6260417103767395, -0.10737155377864838, -0.6992904543876648, -0.15020211040973663, 0.230825275182724, -0.46198561787605286, 0.28589895367622375, -0.18482646346092224, 0.034542784094810486, -0.13203290104866028, 0.23104481399059296, 0.05033749341964722, 0.0913199782371521, 0.14613573253154755, 0.10933606326580048, 0.14259855449199677, 0.11538048088550568, -0.16234828531742096, -0.10149707645177841, 0.027807284146547318, -0.09675844758749008, 0.5679475665092468, 0.2544049024581909, 0.13260623812675476, 0.016347140073776245, -0.010725745931267738, 0.06912654638290405, 0.45414209365844727, 0.13088610768318176, -0.05509103834629059, 0.5264000296592712, 0.0546017661690712, -0.12745758891105652, 0.13458862900733948, 0.9771101474761963, -0.5941919088363647, -0.34010767936706543, 0.00994681566953659, 0.14042051136493683, -0.003393854945898056, 0.2467772364616394, -0.1763128787279129, -0.36520838737487793, 0.008558828383684158, -0.09733168035745621, -0.29024937748908997, -0.1943594217300415, 0.3475720286369324, -0.25344792008399963, 0.42434707283973694, 0.13284409046173096, 0.2170562744140625, -0.050475139170885086, -0.23413114249706268, 0.14442618191242218, -0.04758576303720474, 0.06462261080741882, 0.1204758733510971, 0.06505173444747925, -0.31929999589920044, -0.4705291986465454, -0.3281729519367218, 0.11237967759370804, -0.17740869522094727, -0.40035638213157654, -0.24110457301139832, 0.06570759415626526, 0.20268884301185608, 0.12156444787979126, 0.19925321638584137, 0.3851895034313202, 0.22244952619075775, 0.2539423704147339, 0.14471569657325745, 0.17460191249847412, 0.20782190561294556, 0.0239003524184227, -0.4101085364818573, 0.0025973469018936157, 0.030733253806829453, 0.28490954637527466, -0.2973403334617615, 0.017245011404156685, -0.1251339316368103, -0.3631380498409271, -0.026332607492804527, -0.2165975570678711, 0.35842543840408325, -0.25124818086624146, -0.2243998944759369, 0.08188818395137787, 0.3014507293701172, -0.1021781712770462, -0.07199285924434662, -0.023294907063245773, 0.02457548677921295, -0.12462329864501953, 0.11025374382734299, 0.24170337617397308, 0.08670319616794586, 0.4142286777496338, -0.22951820492744446, -0.1562180072069168, -0.1250024139881134, -0.003739631734788418, 0.21193437278270721, -0.23602980375289917, 0.31187009811401367, -0.17207065224647522, 0.09038291126489639, 0.025926532223820686, -0.4385862946510315, -0.30044984817504883, 0.19515305757522583, 0.0872914046049118, -0.34400737285614014, -0.11515498161315918, 0.09680038690567017, -0.5595490336418152, -0.08119669556617737, 0.1679888665676117, -0.12055999040603638, 0.13395191729068756, -0.46944481134414673, 0.03654436767101288, -0.2592833638191223, -0.30312782526016235, -0.15861903131008148, -0.0653032585978508, 0.27007025480270386, -0.31470486521720886, -0.35894307494163513, -0.33123230934143066, -0.20738773047924042, 0.09957849979400635, -0.3180846571922302, -0.036983706057071686, 0.13217496871948242, -0.23593921959400177, -0.07427404820919037, 0.5720571875572205, -0.5332987904548645, -0.3166050910949707, 0.4164281487464905, -0.3450377583503723, -0.24903108179569244, 0.25984981656074524, 0.024157822132110596, 0.16989871859550476, -0.2648543417453766, -0.12308821082115173, -0.050398506224155426, 0.12207365781068802, -0.10380485653877258, -0.19102084636688232, -0.4137361943721771, 0.09981665760278702, 0.2717949151992798, -0.108523428440094, 0.10879955440759659, 0.12809619307518005, 0.19002844393253326, 0.40333476662635803, 0.08722396194934845, 0.12643882632255554, -0.08443883806467056, 0.19085730612277985, -0.09954707324504852, -0.15517735481262207, -0.037928398698568344, -0.650165319442749, 0.2827853560447693, -0.022536560893058777, 0.22074221074581146, 0.029344789683818817, -0.11142091453075409, -0.5947408080101013, -0.07118986546993256, -0.014745701104402542, -0.2298116683959961, 0.038247525691986084, 0.24519820511341095, 0.04038124158978462, 0.028742395341396332, 0.1323024034500122, -0.0691981315612793, -0.1727847009897232, 0.15835630893707275, -0.057500746101140976, -0.035947319120168686, -0.10883574932813644, 0.10054506361484528, 0.14131920039653778, 0.19206145405769348, 0.08820653706789017, -0.20974351465702057, -0.008258426561951637, 0.19893258810043335, 0.2664552927017212, 0.29947489500045776, 0.17117848992347717, -0.03975299000740051, 0.11774808168411255, 0.16865256428718567, 0.22219626605510712, 0.280743271112442, 0.0006242794916033745, -0.26477840542793274, 0.012452129274606705, 0.18017485737800598, 0.2809547483921051, -0.021284624934196472, -0.3174169659614563, -0.028278619050979614, 0.3193451464176178, 0.06371940672397614, -0.1611047089099884, -0.28946101665496826, 0.04568493738770485, 0.3787929117679596, 0.5581892728805542, -0.052863121032714844, 0.1500057578086853, -0.24770140647888184, -0.24770912528038025, -0.3083847463130951, 0.01466410607099533, 0.0906410962343216, -0.17407086491584778, -0.025309447199106216, 0.058844588696956635, 0.2664116322994232, 0.6372910737991333, 0.046900514513254166, 0.1413993239402771, 0.08129169046878815, -0.23515871167182922, -0.1235392615199089, 0.1703161597251892, 0.18290171027183533, 0.12640784680843353, 0.14134618639945984, -0.009156955406069756, -0.28563162684440613, 0.027415350079536438, -0.033082716166973114, 0.4702059328556061, 0.3248135447502136, -0.33750948309898376, -0.10117167234420776, -0.23019549250602722, -0.3934479355812073, -0.5377248525619507, -0.4357340335845947, -0.4151223599910736, -0.30199292302131653, -0.03800155967473984, 0.29891565442085266, 0.021401599049568176, 0.32440927624702454, -0.21723322570323944, 0.29080918431282043, -0.27205806970596313, -0.4005089998245239, 0.0599130317568779, -0.2860786020755768, -0.3161654472351074, 0.008332610130310059, 0.24271626770496368, -0.040144819766283035, 0.47745466232299805, -0.0952480286359787, -0.3994389772415161, -0.24949181079864502, -0.3282950818538666, 0.08962157368659973, 0.08339346945285797, 0.5332790017127991, 0.4108935594558716, 0.11481627076864243, 0.5304829478263855, -0.03753073886036873, 0.017749542370438576, -0.07441209256649017, -0.35614117980003357, -0.2321394830942154, -0.01416950486600399, 0.21951957046985626, -0.05585123598575592, -0.33744117617607117, -0.5183432102203369, -0.11137236654758453, 0.08031556010246277, 0.058061737567186356, 0.07833410054445267, 0.2125878483057022, -0.09202147275209427, 0.10448282212018967, 0.06396199762821198, 0.012659653089940548, -0.13171161711215973, 0.1249200850725174, 0.04393646866083145, -0.16999350488185883, -0.03391966596245766, 0.12768365442752838, -0.06895244121551514, 0.04899900034070015, -0.1415833681821823, -0.4356604814529419, -0.4000314176082611, 0.16557317972183228, 0.17159394919872284, 0.3695302903652191, -0.2987920939922333, 0.25390851497650146, 0.21395723521709442, 0.12991255521774292, -0.10910617560148239, -0.42785999178886414, 0.22798621654510498, 0.10129096359014511, 0.328249990940094, 0.277874618768692, 0.3687107563018799, -0.19837872684001923, 0.8586010336875916, 0.20692913234233856, -0.2106233388185501, -0.1409544050693512, -0.24552223086357117, -0.06579752266407013, 0.03593466803431511, 0.013525234535336494, 0.21882010996341705, -0.02735571563243866, -0.31299543380737305, 0.30828821659088135, -0.23085924983024597, -0.1324729472398758, -0.12156108021736145, 0.20832255482673645, -0.02576594240963459, -0.11315509676933289, 0.060084156692028046, -0.14215581119060516, 0.060876816511154175, 0.008019208908081055, -0.06661482900381088, -0.242212176322937, -0.25388821959495544, 0.02873433381319046, 0.3385995626449585, -0.18707025051116943, 0.199899360537529, -0.15251682698726654, 0.018473247066140175, -0.40777039527893066, 0.3384018540382385, 0.3403618633747101, 0.5474084615707397, 0.12012188136577606, -0.03538349270820618, 0.130252867937088, 0.1420363336801529, 0.2074132114648819, -0.09370404481887817, 0.14432397484779358, 0.2877647876739502, -0.07581368833780289, -0.13675853610038757, -0.3179689347743988, 0.03660054877400398, 0.18235085904598236, 0.03332754597067833, 0.12281167507171631, -0.47096461057662964, -0.09846272319555283, 0.28636598587036133, 0.19446681439876556, -0.16313233971595764, -0.3334006369113922, -0.10241515934467316, 0.04879126697778702, -0.0787477195262909, -0.2163064181804657, -0.24037285149097443, 0.044299788773059845, -0.038856104016304016, 0.3086199164390564, 0.008862460032105446, -0.035333748906850815, -0.061219364404678345, 0.2663484513759613, 0.203140527009964, -0.20016586780548096, 0.28932374715805054, 0.26163649559020996, 0.08416691422462463, 0.3153352439403534, 0.2017228603363037, -0.08268649876117706, -0.2403506636619568, -0.08250287920236588, 0.01741049438714981, 0.3056638836860657, 0.30834639072418213, 0.24035514891147614, -0.008557572960853577, -0.1876189410686493, -0.14486680924892426, -0.2892586886882782, 0.16906335949897766, 0.2935923933982849, -0.1301058530807495, -0.10033605247735977, -0.32827988266944885, 0.753041684627533, 0.1722811609506607, -0.29610389471054077, 0.5668507218360901, -0.23936660587787628, -0.4109605550765991, -0.013560649938881397, 0.10263343155384064, 0.983655571937561, 0.07682211697101593, 0.21855401992797852, 0.22350868582725525, -0.19695599377155304, 0.39025282859802246, 0.16065627336502075, 0.061364591121673584, -0.5651345252990723, 0.08804209530353546, 0.02915951795876026, -0.09895415604114532, 0.24448162317276, -0.0641174241900444, -0.3517036437988281, 0.340796560049057, 0.08218178153038025, 0.29725372791290283, 0.251806378364563, 0.37033355236053467, -0.04620489478111267, -0.03925669938325882, -0.24055305123329163, 0.051032960414886475, -0.22437436878681183, 0.37193751335144043, -0.2197386622428894, 0.0341653898358345, -0.18881729245185852, -0.07785946130752563, -0.32790809869766235, -0.23958012461662292, -0.1501384824514389, 0.0694781243801117, 0.01919705979526043, -0.18754321336746216, -0.027795296162366867, 0.23903633654117584, 0.06307213008403778, -0.12049311399459839, -0.19020836055278778, 0.11415368318557739, -0.20278063416481018, 0.08895593881607056, 0.23756279051303864, -0.017683597281575203, 0.3286873996257782, -0.344938188791275, -0.11815755814313889, 0.153084397315979, -0.18728309869766235, -0.2327251136302948, -0.0833955630660057, -0.19458208978176117, 0.04582098871469498, -0.3402763903141022, -0.24017846584320068, -0.07632221281528473, 0.027845080941915512, -0.10413603484630585, 0.03824275732040405, 0.1613779217004776, -0.3257570266723633, 0.017180494964122772, 0.05420565605163574, -0.3329341411590576, 0.046963922679424286, 0.3687138855457306, -0.08637317270040512, -0.0008773207664489746, 0.6735666990280151, -0.24754710495471954, 0.0873531699180603, -0.1080072820186615, 0.00033754855394363403, -0.3327271342277527, -0.13322949409484863, -0.041136953979730606, -0.04393347352743149, -0.2875954210758209, -0.06571599096059799, 0.5596664547920227, 0.5117817521095276, 0.10100822895765305, 0.06241864338517189, -0.38301175832748413, -0.3422272801399231, 0.07540889829397202, -0.12961728870868683, 0.26565831899642944, -0.07393591105937958, 0.2888074219226837, 0.09227896481752396, 0.04482793062925339, -0.2038612961769104, -0.14279797673225403, 0.017405079677700996, 0.2378128170967102, -0.1625673472881317, 0.12307857722043991, -0.11569193005561829, 0.02843397483229637, -0.014393167570233345, -0.10343128442764282, -0.26048529148101807, -0.046801865100860596, -0.1658414751291275, 0.22512242197990417, 0.030804991722106934, 0.08343673497438431, 0.012772510759532452, -0.18992915749549866, 0.09936897456645966, 0.17254619300365448, 0.004947792738676071, 0.11314799636602402, -0.06754785776138306, 0.27282944321632385, 0.3285064697265625, 0.11845237016677856, -0.12467221170663834, 0.27501192688941956, 0.10709872841835022, 0.21082846820354462, -0.006238457281142473, 0.3608352243900299, -0.2627809941768646, 0.03662126511335373, 0.18331493437290192, 0.1857268363237381, -0.20792989432811737, 0.13544869422912598, 0.016381077468395233, -0.043131835758686066, 0.033261869102716446, 0.14998728036880493, 0.3344946801662445, 0.7781230807304382, -0.31009289622306824, -0.17035159468650818, 0.45299553871154785, 0.06668003648519516, -0.19272948801517487, -0.05105960741639137, 0.33120492100715637, 0.04032453894615173, 0.07683902978897095, 0.26023223996162415, 0.011644836515188217, -0.06465494632720947, 0.19975247979164124, 0.1704651117324829, -0.010798591189086437, -0.25926119089126587, 0.08984926342964172, 0.3426205813884735, -0.1821492463350296, 0.1657041758298874, 0.3916597068309784, 0.1482623666524887, 0.27847620844841003, 0.0073833465576171875, 0.1359972506761551, 0.42005884647369385, 0.041430309414863586, -0.07995542883872986, 0.09235894680023193, 0.11966158449649811, 0.10311749577522278, 0.2714880704879761, -0.11963316798210144, 0.03871629387140274, 0.40465420484542847, 0.5331206917762756, -0.2811599373817444, -0.1410217136144638, -0.3413615822792053, -0.15611927211284637, 0.06328190863132477, -0.2039298415184021, -0.31007635593414307, 0.02806532382965088, -0.12455867975950241, -0.12496000528335571, -0.17175202071666718, -0.3785262703895569, 0.26906606554985046, 0.069513238966465, -0.37244659662246704, -0.33372125029563904, -0.2165621817111969, 0.2483433187007904, 0.09707961976528168, -0.35574105381965637, 0.15877225995063782, -0.02369759976863861, -0.032223887741565704, 0.2971591055393219, 0.337347149848938, 0.4456102252006531, -0.01678197830915451, -0.3285940885543823, -0.2467661201953888, -0.16939690709114075, 0.09490392357110977, 0.032078564167022705, 0.20311376452445984, 0.010795369744300842, -0.30086663365364075, 0.550868570804596, 0.07143069058656693, 0.05135621130466461, 0.12386082112789154, 0.39140841364860535, -0.437709778547287, -0.17570669949054718, 0.5832920670509338, -0.017454370856285095, -0.18201804161071777, 0.0052442774176597595, 0.24485048651695251, -0.32897526025772095, 0.1818324476480484, 0.021824555471539497, 0.04144666716456413, 0.14532455801963806, 0.08675605058670044, 0.021519936621189117, -0.11364661157131195, 0.41381996870040894, 0.2832196354866028, -0.067356176674366, -0.26395896077156067, -0.22315652668476105, -0.7593231201171875, 0.18017226457595825, -0.13770221173763275, -0.37075382471084595, 0.00038871075958013535, -0.2127862572669983, 0.24636933207511902, 0.0333915539085865, -0.1169736236333847, -0.01768765226006508, -0.004545869305729866, 0.18568389117717743, -0.23927201330661774, -0.3589065968990326, -0.13322223722934723, 0.01474012155085802, 0.013599071651697159, -0.3658691942691803, 0.11531181633472443, -0.28083860874176025, -0.1155008003115654, 0.061025869101285934, 0.012041762471199036, 0.5453343391418457, 0.053665194660425186, 0.24777232110500336, -0.16881844401359558, 0.7727642059326172, 0.15961363911628723, -0.06454945355653763, -0.1531020998954773, 0.09544989466667175, -0.13104097545146942, 0.1424974501132965, -0.08216860890388489, 0.1293344348669052, -0.16671285033226013, 0.4938800632953644, -0.36660730838775635, 0.5406874418258667, -0.09922477602958679, 0.08082855492830276, -0.22181880474090576, -0.03709619492292404, -0.1514166295528412, 0.11986333131790161, 0.22769096493721008, 0.27160707116127014, -0.10734821856021881, 0.35704654455184937, -0.23412734270095825, -0.35361191630363464, 0.4814947545528412, -0.10037000477313995, -0.15508422255516052, 0.021687045693397522, 0.3206314742565155, -0.09507077187299728, 0.07444199919700623, -0.42972448468208313, 0.2376844584941864, 0.3241073191165924, -0.04060295969247818, 0.008704844862222672, 0.2114335149526596, 0.22685380280017853, 0.08315175771713257, 0.0017669722437858582, 1.0600345134735107, -0.1092364564538002, 0.02344811148941517, -0.2541804015636444, -0.3114486336708069 ]
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, 0.21056944131851196, 0.005356155335903168, -0.15212193131446838, 0.1695309579372406, -0.12297330796718597, 0.34444764256477356, 0.3772251605987549, 0.013979699462652206, 0.22968517243862152, 0.1677790880203247, 0.015225032344460487, -0.334241658449173, 0.031307391822338104, 0.022818464785814285, -0.1002444326877594, -0.0594150647521019, 0.197006493806839, -0.07303541898727417, 0.34329521656036377, -0.06011242792010307, -0.07316155731678009, 0.0019713863730430603, 0.1934303343296051, 0.010817679576575756, 0.14580464363098145, -0.1153508797287941, -0.018754221498966217, -0.3885170817375183, -0.8583570718765259, 0.13438120484352112, 0.09185026586055756, 0.2320997565984726, 0.43688473105430603, -0.00012612353020813316, -0.07396548241376877, 0.5866605639457703, -0.024042073637247086, -0.7674411535263062, -0.16982929408550262, 0.34464287757873535, -0.3756412863731384, 0.25355127453804016, -0.06578004360198975, 0.11371976882219315, -0.1838429570198059, 0.18135371804237366, 0.07539583742618561, 0.1361459195613861, 0.245559960603714, 0.09919436275959015, 0.31069257855415344, 0.028131205588579178, -0.11110033094882965, -0.05956900864839554, 0.12354503571987152, -0.14380034804344177, 0.5142217874526978, 0.48955070972442627, 0.24067121744155884, -0.07860211282968521, -0.07082480192184448, 0.06261495500802994, 0.4359807074069977, 0.13713976740837097, -0.0326455682516098, 0.5384781360626221, 0.044237054884433746, -0.19004222750663757, 0.16044630110263824, 0.8945168852806091, -0.5086084604263306, -0.4655519425868988, -0.06761332601308823, 0.017939716577529907, -0.10367359966039658, 0.10765813291072845, -0.19388851523399353, -0.3774493634700775, 0.20891021192073822, -0.1669583022594452, -0.33218520879745483, -0.09904003143310547, 0.28893008828163147, -0.17620722949504852, 0.23123739659786224, 0.1270466297864914, 0.25155389308929443, -0.02958333119750023, -0.0673801377415657, 0.1549283266067505, -0.052094221115112305, 0.015269063413143158, 0.1340530812740326, 0.08804330229759216, -0.3319090008735657, -0.4149754047393799, -0.29810580611228943, 0.1995212733745575, 0.025635112076997757, -0.5338376760482788, -0.18990269303321838, -0.06855496019124985, 0.17624177038669586, 0.08776327967643738, 0.2744106650352478, 0.36686965823173523, 0.22054314613342285, 0.311817467212677, -0.010775187984108925, 0.28586411476135254, 0.21933385729789734, -0.00553248543292284, -0.27972203493118286, -0.002387784421443939, 0.07539418339729309, 0.10619138181209564, -0.21187756955623627, -0.02138431742787361, -0.25974953174591064, -0.42282959818840027, -0.0277206152677536, -0.09297081083059311, 0.23292586207389832, -0.17843417823314667, -0.1466093361377716, -0.01997271366417408, 0.3341640830039978, -0.10098528116941452, -0.1949630081653595, -0.00010901223868131638, -0.020320232957601547, -0.24842840433120728, -0.06086083501577377, 0.22468288242816925, -0.13178770244121552, 0.3598708212375641, -0.35544416308403015, -0.14415745437145233, -0.03814074397087097, -0.022520100697875023, 0.03089410811662674, -0.17065325379371643, 0.28824383020401, -0.23571887612342834, 0.017072374001145363, 0.05582606792449951, -0.30029064416885376, -0.2188141644001007, 0.21825045347213745, -0.0032299980521202087, -0.43608370423316956, -0.13643638789653778, 0.06006927415728569, -0.5637033581733704, 0.02172400988638401, 0.04728544503450394, -0.16956348717212677, 0.18883562088012695, -0.2967650294303894, 0.14172445237636566, -0.250382661819458, -0.36074426770210266, -0.15156853199005127, -0.1783817857503891, 0.22806385159492493, -0.31079187989234924, -0.2898619771003723, -0.22972798347473145, -0.38468137383461, 0.07798102498054504, -0.13051730394363403, -0.07232106477022171, 0.20862683653831482, -0.33610525727272034, -0.035919029265642166, 0.6512778997421265, -0.5617484450340271, -0.2883245646953583, 0.6501520872116089, -0.39960044622421265, -0.2595352530479431, 0.2737455368041992, -0.07269967347383499, 0.14991018176078796, -0.44446316361427307, -0.1325370967388153, 0.004221666604280472, -0.04450642317533493, -0.1131410151720047, -0.08715605735778809, -0.42463457584381104, 0.2098807692527771, 0.20373685657978058, -0.06479039788246155, 0.2234140932559967, 0.1218448057770729, 0.10671255737543106, 0.4471711218357086, 0.03521332889795303, 0.16927526891231537, -0.21021631360054016, 0.20478932559490204, -0.1228678822517395, -0.1074662134051323, -0.13562634587287903, -0.5853860378265381, 0.24203673005104065, -0.04733653366565704, 0.15349692106246948, -0.006374247372150421, -0.06405392289161682, -0.4417177438735962, 0.02850235067307949, 0.10575544834136963, -0.21586723625659943, -0.05539102852344513, 0.14584729075431824, 0.025066480040550232, -0.03469604253768921, -0.02631925232708454, -0.00754067488014698, -0.2913230061531067, 0.16251416504383087, 0.20891226828098297, -0.14848344027996063, -0.08332210779190063, 0.09957940131425858, 0.15043026208877563, 0.2522900700569153, -0.06979330629110336, -0.3053167164325714, 0.02069862186908722, 0.32338932156562805, 0.25860595703125, 0.14943833649158478, 0.07985370606184006, 0.05053447186946869, 0.18621467053890228, 0.26682206988334656, 0.14637459814548492, 0.20671021938323975, 0.024877086281776428, -0.2785708010196686, 0.023999277502298355, 0.15529467165470123, 0.15474027395248413, -0.024672679603099823, -0.2123958319425583, -0.042884498834609985, 0.2966698110103607, 0.08589234948158264, -0.15950703620910645, -0.28861576318740845, 0.10071870684623718, 0.2132439762353897, 0.6271629333496094, 0.013069907203316689, 0.23440656065940857, -0.16070425510406494, -0.19119104743003845, -0.4197957515716553, 0.047629404813051224, 0.2504350244998932, -0.08639683574438095, -0.06156587973237038, -0.03870204836130142, 0.24559788405895233, 0.8113179206848145, 0.0071622407995164394, 0.2236732691526413, 0.05936606228351593, -0.06221941113471985, -0.2031267285346985, 0.21539631485939026, 0.18741638958454132, -0.019955070689320564, 0.14958512783050537, -0.009534548968076706, -0.213071808218956, 0.011151015758514404, -0.02994939126074314, 0.37450703978538513, 0.3891151249408722, -0.45985034108161926, -0.071262426674366, -0.2434227615594864, -0.4637845754623413, -0.35133445262908936, -0.32140475511550903, -0.2830204963684082, -0.28866180777549744, -0.07606277614831924, 0.32052579522132874, 0.0429517887532711, 0.22180591523647308, -0.27010664343833923, 0.2553025782108307, -0.2692341208457947, -0.2900591194629669, 0.039597876369953156, -0.13836529850959778, -0.3783036172389984, -0.010134518146514893, 0.2034052461385727, -0.12987132370471954, 0.49237048625946045, -0.04010806977748871, -0.2758660316467285, -0.21922940015792847, -0.34642720222473145, 0.1785959154367447, -0.008019987493753433, 0.5385330319404602, 0.5283278822898865, 0.011747114360332489, 0.49219995737075806, 0.04509781673550606, -0.10305843502283096, -0.2824671268463135, -0.44600310921669006, -0.1996392160654068, 0.04542342200875282, 0.12376132607460022, -0.22519946098327637, -0.24200621247291565, -0.5144346952438354, -0.10970619320869446, 0.14452898502349854, -0.03499829024076462, 0.12034574151039124, 0.27052778005599976, -0.18616634607315063, 0.13806453347206116, -0.006836146581918001, -0.06195448338985443, -0.15893393754959106, 0.07213608920574188, -0.09341348707675934, -0.15120358765125275, -0.1503472477197647, 0.05138719454407692, 0.009144652634859085, -0.026123594492673874, -0.03587005287408829, -0.34018993377685547, -0.46588650345802307, 0.15529721975326538, 0.13913819193840027, 0.34824758768081665, -0.22888325154781342, 0.19969868659973145, 0.2338806539773941, 0.1612386405467987, -0.06107719987630844, -0.34528347849845886, 0.14796283841133118, 0.04752747714519501, 0.24498242139816284, 0.26928919553756714, 0.4195297360420227, -0.01958618126809597, 0.7755740284919739, 0.14332546293735504, -0.0958905965089798, -0.07831813395023346, -0.2671540081501007, 0.006282620131969452, -0.05424752086400986, -0.005567440763115883, 0.31085702776908875, -0.10074695944786072, -0.28282248973846436, 0.3185688853263855, -0.18947568535804749, 0.05149201303720474, -0.09044620394706726, 0.2656155824661255, -0.06809309124946594, -0.0441906563937664, 0.05007345229387283, -0.28195029497146606, 0.039340317249298096, -0.012657798826694489, -0.03752116113901138, -0.13320261240005493, -0.21860477328300476, 0.09952671825885773, 0.3623552918434143, -0.018679574131965637, 0.29034313559532166, -0.11583904922008514, 0.15687815845012665, -0.34239599108695984, 0.2570706009864807, 0.4411430060863495, 0.6185430288314819, 0.08319763839244843, 0.09134873002767563, 0.1832665205001831, 0.12001954764127731, 0.2273235023021698, -0.03647734224796295, 0.23378071188926697, 0.06326091289520264, -0.23603403568267822, -0.017865262925624847, -0.2264620065689087, 0.00059470534324646, 0.09605272859334946, -0.07571941614151001, 0.17361825704574585, -0.4774007499217987, -0.23207788169384003, 0.21772129833698273, 0.10532347857952118, -0.07665837556123734, -0.38931038975715637, -0.06505332887172699, 0.03828480839729309, -0.05286567658185959, -0.31292566657066345, -0.136131152510643, -0.0052222381345927715, -0.08902165293693542, 0.43324971199035645, -0.10226196050643921, -0.14556947350502014, -0.005133286118507385, 0.10432887077331543, 0.044709838926792145, -0.19614487886428833, 0.20612138509750366, 0.19784824550151825, 0.1746518313884735, 0.21705137193202972, 0.18835315108299255, -0.0840260237455368, -0.1561623066663742, -0.03743549808859825, -0.05730261281132698, 0.3986186385154724, 0.3663225769996643, 0.10062596201896667, 0.04277164861559868, -0.2101757824420929, -0.15302953124046326, -0.34225931763648987, 0.22055724263191223, 0.28233063220977783, -0.16399061679840088, -0.05102991312742233, -0.28409022092819214, 0.6471732258796692, 0.14356431365013123, -0.2456749975681305, 0.4512062668800354, -0.1504860818386078, -0.30368852615356445, -0.025458380579948425, 0.12860321998596191, 0.985715925693512, 0.043363362550735474, 0.22869481146335602, 0.38514605164527893, -0.33894407749176025, 0.42819705605506897, 0.14748515188694, 0.02666272222995758, -0.5198193192481995, -0.05517926067113876, -0.05943042039871216, -0.165473073720932, 0.274152547121048, -0.15859314799308777, -0.20023588836193085, 0.34560856223106384, 0.08795057237148285, 0.3736129701137543, 0.4137502610683441, 0.39923718571662903, -0.10544490814208984, -0.05788303539156914, -0.16087783873081207, 0.02206481248140335, -0.18493792414665222, 0.28433409333229065, -0.21571719646453857, -0.020652487874031067, -0.18660970032215118, -0.028540097177028656, -0.24675114452838898, -0.23737698793411255, -0.1255301535129547, 0.017985105514526367, 0.1404607594013214, -0.10921992361545563, 0.05769843980669975, 0.16871511936187744, 0.1828477382659912, -0.09496639668941498, -0.13535992801189423, 0.025371313095092773, -0.23170378804206848, 0.2188091278076172, 0.3044604957103729, 0.028277065604925156, 0.3999037444591522, -0.28152593970298767, -0.07076147198677063, 0.14869144558906555, -0.2799605131149292, -0.2452802211046219, -0.17162904143333435, -0.21622012555599213, 0.16090381145477295, -0.35440143942832947, -0.2652537226676941, -0.1846257746219635, -0.020434550940990448, -0.03614324703812599, -0.03854873403906822, 0.23290929198265076, -0.3300037682056427, -0.01323217898607254, -0.037422265857458115, -0.3219470977783203, 0.03071577660739422, 0.48213672637939453, -0.04637731611728668, -0.17012344300746918, 0.6307938098907471, -0.24155570566654205, 0.028976164758205414, -0.0897454023361206, 0.03222358971834183, -0.18571768701076508, -0.1862180531024933, 0.03852279856801033, -0.07007022202014923, -0.3210486173629761, -0.03493773564696312, 0.6218963861465454, 0.5002497434616089, 0.026229266077280045, -0.07093337178230286, -0.4239475727081299, -0.4768606126308441, -0.019827544689178467, -0.21596606075763702, 0.14384189248085022, -0.20705530047416687, 0.3047216832637787, 0.09505714476108551, 0.03632984682917595, -0.1691185086965561, -0.2332279086112976, 0.12708936631679535, 0.26113492250442505, -0.20231159031391144, 0.03588312864303589, -0.010400895029306412, 0.04883423447608948, -0.07755549252033234, 0.054522521793842316, -0.34435224533081055, -0.022740554064512253, -0.22121775150299072, 0.2403482049703598, 0.12901584804058075, 0.015883587300777435, 0.1009754091501236, -0.12922483682632446, -0.03020528331398964, 0.19713294506072998, -0.03883189707994461, 0.1435362994670868, -0.08869205415248871, 0.2655784785747528, 0.31918010115623474, 0.1678486168384552, -0.0894201248884201, 0.2904230058193207, 0.08135998994112015, 0.2633010745048523, -0.03736348822712898, 0.3029985725879669, -0.22787117958068848, 0.036965735256671906, 0.10371330380439758, 0.246509850025177, -0.018953166902065277, 0.24331048130989075, 0.11851654946804047, -0.08313368260860443, 0.09952300786972046, 0.0027661691419780254, 0.38336580991744995, 0.7761162519454956, -0.18420851230621338, -0.3148205578327179, 0.4969845116138458, 0.017806198447942734, -0.10703211277723312, 0.006450509186834097, 0.24761880934238434, 0.16800172626972198, 0.02530844137072563, 0.3604945242404938, 0.03009425476193428, -0.035909440368413925, 0.09229389578104019, 0.1062341257929802, 0.08866747468709946, -0.26417475938796997, -0.2379094362258911, 0.5210915207862854, -0.2259112149477005, 0.24124941229820251, 0.3535771667957306, 0.13580672442913055, 0.29599878191947937, 0.03766085207462311, 0.07566314935684204, 0.5019505620002747, -0.010013192892074585, 0.04312687739729881, 0.14046424627304077, 0.13297908008098602, 0.014500580728054047, 0.2870316505432129, 0.07500987499952316, 0.013744734227657318, 0.33756130933761597, 0.5905369520187378, -0.20217715203762054, -0.1644720584154129, -0.1727122813463211, -0.16748780012130737, 0.0769597664475441, -0.228070467710495, -0.2735540568828583, 0.1288490891456604, -0.21248170733451843, -0.1030297726392746, -0.23020565509796143, -0.29267215728759766, 0.21165421605110168, 0.05275091901421547, -0.33862531185150146, -0.27144676446914673, -0.4014511704444885, 0.1491457223892212, 0.05067802220582962, -0.34592893719673157, 0.12932947278022766, -0.07099097967147827, 0.11562046408653259, 0.3956110179424286, 0.2540735602378845, 0.3397336006164551, 0.06733053922653198, -0.3097277283668518, -0.31341180205345154, -0.32432109117507935, 0.15070955455303192, 0.008740007877349854, 0.39771607518196106, 0.017714716494083405, -0.11327928304672241, 0.6163249015808105, 0.004212956875562668, 0.10033030062913895, 0.2640596628189087, 0.36592113971710205, -0.5369967222213745, -0.11764606833457947, 0.7859679460525513, 0.07999040186405182, -0.1935425102710724, 0.07940411567687988, 0.22868478298187256, -0.19130200147628784, 0.17380353808403015, 0.046440206468105316, -0.11265183240175247, 0.1843630075454712, 0.11147186905145645, 0.01518373191356659, -0.05229577049612999, 0.34844377636909485, 0.4192717373371124, -0.0645558089017868, -0.2744944393634796, -0.015578880906105042, -0.6457434296607971, 0.181339830160141, -0.05328163132071495, -0.25216221809387207, 0.0043504745699465275, -0.242372527718544, 0.17777614295482635, 0.11598408967256546, -0.2468876987695694, 0.13935711979866028, -0.08880764245986938, 0.08666262775659561, -0.1472114473581314, -0.3679315149784088, -0.17883417010307312, 0.0805131047964096, -0.07546988129615784, -0.4362681806087494, 0.20022113621234894, -0.4206122159957886, -0.12240853160619736, 0.07628065347671509, 0.015164688229560852, 0.5335344672203064, -0.019309580326080322, 0.2926962673664093, -0.05495215207338333, 0.5333094596862793, 0.19510090351104736, 0.006029875949025154, -0.18519791960716248, 0.18206799030303955, -0.1992652863264084, 0.13084818422794342, -0.04895910248160362, 0.06954202055931091, -0.14772485196590424, 0.5281546115875244, -0.4214254915714264, 0.5098816752433777, -0.05077594146132469, 0.16153916716575623, -0.32901185750961304, 0.10240151733160019, -0.24033862352371216, 0.08635570108890533, 0.2888404428958893, 0.21862655878067017, -0.11910434067249298, 0.3717103898525238, -0.2540442943572998, -0.20215848088264465, 0.4779549837112427, -0.1859750598669052, -0.10019859671592712, 0.15425102412700653, 0.3032086491584778, 0.19870346784591675, -0.18652045726776123, -0.5294916033744812, 0.15762919187545776, 0.3263663649559021, -0.10436972230672836, -0.04502715915441513, 0.22184592485427856, 0.20797038078308105, 0.06086595728993416, 0.007538933306932449, 0.9585251808166504, -0.13304033875465393, 0.08896137028932571, -0.06882606446743011, -0.24809643626213074 ]
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?
[ 0.05760304257273674, 0.24249693751335144, -0.000598544254899025, -0.016872867941856384, -0.026223141700029373, -0.10918831825256348, 0.3703441321849823, 0.27449801564216614, 0.0017391927540302277, -0.07336340844631195, 0.19066615402698517, 0.1067383736371994, -0.2590656578540802, 0.05271521955728531, 0.06369110941886902, -0.17159409821033478, 0.06112443655729294, 0.205256849527359, -0.09032115340232849, 0.19936902821063995, -0.10669591277837753, -0.01393563486635685, -0.021723750978708267, 0.2025269865989685, -0.09185992181301117, -0.044857464730739594, -0.290224552154541, 0.06356648355722427, -0.35639968514442444, -0.9518452882766724, 0.2559972405433655, 0.16132852435112, 0.21488924324512482, 0.39023557305336, -0.00012299601803533733, -0.13422223925590515, 0.5136099457740784, -0.030447958037257195, -0.7506924867630005, -0.1976187378168106, 0.3274133801460266, -0.4366213083267212, 0.3908313810825348, -0.0833529680967331, 0.20270678400993347, -0.2667604982852936, 0.31416061520576477, 0.2079792022705078, 0.1440405696630478, 0.2145276963710785, 0.07082146406173706, 0.2529842257499695, 0.02971867099404335, -0.09841020405292511, -0.04017750918865204, 0.11929966509342194, -0.034477971494197845, 0.5099808573722839, 0.39377060532569885, 0.12621459364891052, -0.16321870684623718, 0.11727796494960785, 0.14007452130317688, 0.3983493447303772, 0.2259792685508728, -0.01795342192053795, 0.5039622783660889, -0.047598786652088165, -0.20237979292869568, 0.05471815541386604, 0.8494493961334229, -0.5009220838546753, -0.4326480031013489, -0.14169380068778992, 0.032289765775203705, -0.16478854417800903, 0.2154712677001953, -0.10920235514640808, -0.15507255494594574, 0.05712485313415527, -0.24445389211177826, -0.11133310943841934, -0.15746980905532837, 0.336891770362854, -0.22511881589889526, 0.2915149927139282, -0.010365025140345097, 0.19610467553138733, -0.030724529176950455, -0.1835590898990631, 0.057907577604055405, -0.2011939138174057, 0.07630202174186707, 0.17462009191513062, 0.016123972833156586, -0.2886023223400116, -0.29430362582206726, -0.18726423382759094, 0.12454863637685776, 0.014857377856969833, -0.4481185972690582, -0.17962482571601868, -0.03736473247408867, 0.2117084562778473, 0.25023096799850464, 0.3405122458934784, 0.3932395279407501, 0.19038574397563934, 0.15328194200992584, 0.08018219470977783, 0.15144771337509155, 0.1508202850818634, -0.041883792728185654, -0.22569634020328522, -0.011925308033823967, 0.02554197423160076, 0.16613106429576874, -0.2824716866016388, -0.05167902633547783, -0.12572945654392242, -0.5409555435180664, -0.07574271410703659, -0.1315053552389145, 0.23289451003074646, -0.17989923059940338, -0.0450725182890892, 0.03212364763021469, 0.4197337031364441, -0.1784875988960266, -0.15913546085357666, -0.015688810497522354, 0.0746147632598877, -0.2581358551979065, -0.04566466808319092, 0.174321711063385, -0.09063209593296051, 0.4508668780326843, -0.2195093035697937, -0.33825039863586426, -0.179396390914917, 0.06953290104866028, 0.013142002746462822, -0.09206866472959518, 0.2377602905035019, -0.195138081908226, 0.06764184683561325, 0.0803680494427681, -0.41097766160964966, -0.2782211899757385, 0.010887682437896729, 0.05627662315964699, -0.34333714842796326, -0.027284765616059303, 0.06280625611543655, -0.4833716154098511, -0.010563388466835022, 0.12420543283224106, 0.0902782678604126, 0.14917361736297607, -0.3166505694389343, 0.10074393451213837, -0.41337746381759644, -0.2347722351551056, -0.09195572882890701, -0.1889968365430832, 0.3042604923248291, -0.2840292155742645, -0.27550458908081055, -0.1664571464061737, -0.31274059414863586, 0.09089353680610657, -0.18867769837379456, -0.0761040449142456, 0.2745271623134613, -0.1902463138103485, 0.03018895536661148, 0.5996158719062805, -0.5389538407325745, -0.2087562531232834, 0.5325261354446411, -0.2751666009426117, -0.052935075014829636, 0.28938454389572144, -0.073978491127491, 0.110174261033535, -0.35946279764175415, -0.1081964522600174, 0.05834294110536575, 0.003969097509980202, -0.19246141612529755, -0.21571467816829681, -0.26801130175590515, 0.18028435111045837, 0.28513103723526, -0.1920585185289383, 0.018702417612075806, 0.06548072397708893, 0.18505845963954926, 0.3273560106754303, 0.09518390893936157, 0.12832678854465485, -0.18868356943130493, -0.05513500049710274, -0.07846146821975708, -0.12363823503255844, -0.01617143303155899, -0.7373564839363098, 0.1525452435016632, -0.10349711030721664, 0.20631973445415497, 0.04400281608104706, 0.0035772137343883514, -0.4655626118183136, -0.049690138548612595, -0.07461187243461609, -0.30068257451057434, 0.030343875288963318, 0.23752397298812866, 0.07605691254138947, -0.003967924043536186, 0.00890287384390831, -0.012331390753388405, -0.3098595142364502, 0.24812154471874237, -0.04932260885834694, -0.18346983194351196, -0.0836695060133934, 0.14995940029621124, 0.205399289727211, 0.25154605507850647, 0.033978402614593506, -0.2759888768196106, 0.10466857999563217, 0.32568368315696716, 0.23714834451675415, 0.24148528277873993, 0.17153821885585785, 0.06900224834680557, 0.23371590673923492, 0.3315989673137665, 0.3070366680622101, 0.16558153927326202, 0.08676372468471527, -0.26936161518096924, -0.09193149954080582, 0.1549011915922165, 0.2427971512079239, 0.004586212337017059, -0.16184183955192566, -0.06739960610866547, 0.33853983879089355, 0.1212911605834961, -0.2309657782316208, -0.2655969262123108, 0.012635204941034317, 0.29515546560287476, 0.6402701735496521, 0.06318031251430511, 0.08673670887947083, -0.03449440747499466, -0.11137320101261139, -0.46820640563964844, -0.023292236030101776, 0.09991331398487091, -0.2714596688747406, -0.16550683975219727, -0.019511381164193153, 0.2567764222621918, 0.7403609752655029, 0.014357052743434906, 0.28377461433410645, 0.043051786720752716, -0.26060429215431213, -0.2131253033876419, 0.1901426613330841, 0.29233241081237793, -0.07247414439916611, 0.14260801672935486, 0.05944399535655975, -0.22141100466251373, -0.011801742017269135, -0.14425168931484222, 0.3677428364753723, 0.3290640711784363, -0.4422619044780731, 0.011568421497941017, -0.22433897852897644, -0.4463372230529785, -0.5092363357543945, -0.37744784355163574, -0.2606242895126343, -0.348259836435318, -0.05424051731824875, 0.2230663299560547, -0.09885327517986298, 0.20850633084774017, -0.04445458948612213, 0.3514876663684845, -0.33391475677490234, -0.061890989542007446, 0.18121905624866486, -0.2874465882778168, -0.43042632937431335, -0.02127774804830551, 0.37744420766830444, 0.010301250964403152, 0.3699018955230713, -0.16296160221099854, -0.3067917227745056, -0.1718650460243225, -0.35774680972099304, 0.17577336728572845, -0.05133640766143799, 0.4902479350566864, 0.4374166429042816, 0.05680596083402634, 0.42201197147369385, -0.07975700497627258, -0.005827383138239384, -0.061605677008628845, -0.35395562648773193, -0.19850434362888336, 0.00805377122014761, 0.011529458686709404, -0.22299200296401978, -0.46764835715293884, -0.5795467495918274, -0.06716534495353699, 0.23849734663963318, 0.02374003455042839, 0.1362578123807907, 0.3584892749786377, -0.2352583259344101, 0.23094291985034943, -0.11682120710611343, 0.06175294518470764, -0.1307680606842041, 0.04842299595475197, 0.033469658344984055, -0.1410067081451416, -0.14650937914848328, 0.08306850492954254, -0.06929121911525726, -0.056239258497953415, -0.12516795098781586, -0.4092954397201538, -0.46038907766342163, 0.19253312051296234, -0.019878249615430832, 0.3854581415653229, -0.11975400149822235, 0.19413024187088013, 0.22859154641628265, 0.15268848836421967, -0.0922197699546814, -0.25398755073547363, 0.3085111975669861, 0.0481681227684021, 0.2682611346244812, 0.2667809724807739, 0.30996257066726685, -0.08249562233686447, 0.8548855185508728, 0.23433223366737366, -0.2229716032743454, -0.17823828756809235, -0.028071874752640724, -0.07432746887207031, 0.008990399539470673, 0.039054155349731445, 0.24598310887813568, -0.19373923540115356, -0.2576615810394287, 0.4539285898208618, -0.11471707373857498, -0.07941103726625443, -0.012890700250864029, 0.28912875056266785, -0.10928496718406677, -0.005734983831644058, 0.19463348388671875, -0.15763550996780396, 0.023471180349588394, 0.010373320430517197, -0.033706650137901306, -0.15833380818367004, -0.06846459209918976, -0.0331968329846859, 0.34193724393844604, -0.11140764504671097, 0.21972520649433136, -0.27268800139427185, -0.06407055258750916, -0.2533634603023529, 0.23277431726455688, 0.39788496494293213, 0.4647171199321747, 0.15090231597423553, -0.0316687636077404, 0.21203121542930603, 0.20424798130989075, 0.08792264014482498, -0.03190042823553085, 0.1063772439956665, 0.18112264573574066, -0.22042010724544525, -0.03037141263484955, -0.30280011892318726, -0.2254151701927185, 0.04275547340512276, 0.01996736414730549, 0.17918384075164795, -0.4684464931488037, -0.24010691046714783, 0.3211110830307007, 0.06214383244514465, -0.05001574009656906, -0.2363814115524292, 0.02350929006934166, 0.005604702979326248, -0.04983554035425186, -0.17327040433883667, -0.19891439378261566, -0.013889316469430923, -0.11765550822019577, 0.42405927181243896, -0.03836795687675476, 0.045314691960811615, -0.021976172924041748, 0.24746011197566986, 0.09659837186336517, -0.1492275893688202, 0.142380952835083, 0.2692519426345825, 0.14296641945838928, 0.2716633379459381, 0.1835295855998993, -0.13022004067897797, -0.22181808948516846, 0.051952674984931946, -0.06875784695148468, 0.3754238188266754, 0.441513329744339, 0.20746377110481262, 0.031254470348358154, -0.24363237619400024, -0.17298300564289093, -0.3376981317996979, 0.2777041494846344, 0.3622957766056061, -0.1226685494184494, -0.018863897770643234, -0.3984578251838684, 0.7639327049255371, 0.11514200270175934, -0.31864988803863525, 0.3197759687900543, -0.1712692826986313, -0.2754104435443878, 0.06674183160066605, 0.04492731764912605, 0.9632916450500488, 0.18758024275302887, 0.2697294056415558, 0.33561018109321594, -0.30502283573150635, 0.5088148713111877, 0.07729946076869965, 0.08329827338457108, -0.526681125164032, -0.004951021634042263, 0.011898981407284737, -0.08526362478733063, 0.2526697516441345, -0.09175126254558563, -0.2902084290981293, 0.3714599907398224, 0.15412487089633942, 0.3369047939777374, 0.26662394404411316, 0.3904864192008972, -0.10265128314495087, -0.08184851706027985, -0.3148295283317566, 0.03274140506982803, -0.18126249313354492, 0.29423069953918457, -0.2555635869503021, -0.05453521013259888, -0.19459013640880585, -0.0489114373922348, -0.16635634005069733, -0.232101172208786, -0.17980781197547913, 0.044044747948646545, 0.09738385677337646, -0.2445479780435562, 0.007083317264914513, 0.2463042289018631, 0.13479098677635193, -0.3061772584915161, -0.0829959437251091, -0.0007392577826976776, -0.20389783382415771, 0.19736739993095398, 0.2699483036994934, -0.0520637109875679, 0.4594918489456177, -0.2614249587059021, 0.004677951335906982, 0.06200576201081276, -0.2999454140663147, -0.011491157114505768, -0.07748061418533325, -0.1792587786912918, 0.1897275447845459, -0.29735347628593445, -0.275772362947464, 0.034674856811761856, 0.015729404985904694, -0.016935501247644424, -0.013131961226463318, 0.1991761028766632, -0.2707010507583618, 0.01630931720137596, -0.09080842137336731, -0.23815381526947021, -0.012713782489299774, 0.34332647919654846, -0.08797552436590195, -0.07147163152694702, 0.7205212116241455, -0.1423330307006836, -0.013048091903328896, -0.08775792270898819, 0.07892707735300064, -0.07721299678087234, -0.21021470427513123, -0.024668309837579727, -0.09397958219051361, -0.3097341060638428, -0.11627878248691559, 0.6664244532585144, 0.5130177736282349, 0.08737722039222717, -0.012764431536197662, -0.39167970418930054, -0.3361733555793762, 0.1102447584271431, -0.2133309692144394, 0.20463064312934875, -0.15378166735172272, 0.23010988533496857, 0.2059989869594574, 0.09173562377691269, -0.19896355271339417, -0.21286660432815552, -0.05241968482732773, 0.3219778537750244, -0.20810826122760773, 0.008621208369731903, -0.022832272574305534, 0.07475611567497253, -0.07202239334583282, -0.03039177507162094, -0.3601388931274414, -0.0361163504421711, -0.16067591309547424, 0.23085974156856537, 0.16787391901016235, -0.10099488496780396, 0.051194820553064346, -0.05666758865118027, 0.08302770555019379, 0.15836556255817413, -0.0659610852599144, -0.06922027468681335, -0.046982504427433014, 0.31562042236328125, 0.22112752497196198, 0.2653515338897705, -0.061261486262083054, 0.1533360630273819, 0.07793214172124863, 0.14180797338485718, 0.0833163633942604, 0.18333101272583008, -0.32301101088523865, 0.01377502828836441, 0.07649018615484238, 0.29494187235832214, -0.11261296272277832, 0.2398705780506134, 0.11625219881534576, -0.03209961950778961, 0.14109624922275543, 0.08694121241569519, 0.2844572961330414, 0.8438327312469482, -0.1847461611032486, -0.2881440818309784, 0.4106006324291229, 0.017603598535060883, -0.04218664765357971, 0.08298292011022568, 0.2660800516605377, 0.11460378766059875, 0.022273149341344833, 0.2854962646961212, 0.055850084871053696, -0.0035331547260284424, 0.03952415660023689, 0.16859549283981323, 0.19243678450584412, -0.21285240352153778, -0.06870892643928528, 0.4142073392868042, -0.2447899430990219, 0.2616272270679474, 0.3799400329589844, 0.05843599885702133, 0.21717175841331482, -0.09376569092273712, 0.0666029304265976, 0.4720230996608734, 0.10343413054943085, 0.07247675955295563, 0.14047127962112427, 0.1095653623342514, -0.05472439154982567, 0.4614467918872833, 0.027293212711811066, 0.019859664142131805, 0.2861618995666504, 0.6132856011390686, -0.33437731862068176, -0.3217318058013916, -0.17343229055404663, -0.07680444419384003, 0.10721223801374435, -0.18119587004184723, -0.52765953540802, 0.014906935393810272, -0.25306904315948486, -0.07244852930307388, -0.2701844871044159, -0.35571712255477905, 0.31993842124938965, 0.0411810539662838, -0.3105320334434509, -0.4160935878753662, -0.360491544008255, 0.08546432852745056, 0.1281217634677887, -0.3214928209781647, 0.2644851803779602, 0.23103877902030945, 0.0690561830997467, 0.4039289951324463, 0.33244723081588745, 0.40230700373649597, 0.09167840331792831, -0.3861663341522217, -0.3036430776119232, -0.2539188861846924, 0.10707499086856842, 0.12506237626075745, 0.2783576250076294, -0.042769819498062134, -0.21499799191951752, 0.5813987255096436, 0.01551702618598938, 0.07906574010848999, 0.29227203130722046, 0.4092690944671631, -0.5363081693649292, -0.24184051156044006, 0.6709061861038208, -0.0029571689665317535, -0.2478887438774109, 0.10846426337957382, 0.15607410669326782, -0.36080628633499146, 0.08438092470169067, 0.0010098125785589218, 0.07150506973266602, 0.19846251606941223, 0.09360723942518234, 0.026088673621416092, -0.009924501180648804, 0.49392151832580566, 0.49144279956817627, -0.10784625262022018, -0.3457840085029602, -0.10265505313873291, -0.708646297454834, 0.02320411056280136, -0.02492237463593483, -0.3069426715373993, -0.0286514051258564, -0.1313231736421585, 0.1805087924003601, 0.0959317684173584, -0.25933027267456055, 0.05915506184101105, -0.03458626940846443, 0.07855816930532455, -0.31683024764060974, -0.2308543175458908, -0.1634584665298462, 0.011753806844353676, -0.11217958480119705, -0.2473425567150116, 0.11470585316419601, -0.390979140996933, -0.14848636090755463, 0.22859202325344086, 0.048662856221199036, 0.47063031792640686, 0.04398009553551674, 0.16804176568984985, -0.01494508795440197, 0.5717011094093323, 0.27138984203338623, -0.12703824043273926, -0.22806525230407715, 0.21541136503219604, -0.1466309130191803, 0.17723961174488068, -0.12875758111476898, 0.36341142654418945, -0.1993665248155594, 0.5129457116127014, -0.20585551857948303, 0.4702882766723633, -0.04468383640050888, 0.057025883346796036, -0.3319069445133209, 0.0854540541768074, -0.22928383946418762, 0.04520866274833679, 0.33929333090782166, 0.3558689057826996, -0.06925636529922485, 0.29817691445350647, -0.26029518246650696, -0.22530728578567505, 0.320079505443573, -0.3188404142856598, -0.11840410530567169, 0.1429828405380249, 0.2893074154853821, 0.14046311378479004, 0.024536065757274628, -0.6277332305908203, 0.07381164282560349, 0.23111511766910553, -0.17052239179611206, -0.09328325092792511, 0.17623235285282135, 0.23813951015472412, -0.06259158998727798, -0.03592406585812569, 0.9138442277908325, -0.07508876919746399, -0.0677710548043251, -0.10770955681800842, -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?
[ -0.1263233721256256, -0.22070911526679993, -0.10063229501247406, 0.27801111340522766, 0.24547500908374786, -0.20522364974021912, 0.13721121847629547, 0.512859046459198, 0.18828602135181427, -0.06233552098274231, 0.1278189718723297, 0.38913494348526, 0.0037763379514217377, -0.16061760485172272, -0.010743602178990841, 0.024212224408984184, 0.09608031809329987, 0.23505264520645142, -0.1455627828836441, -0.16482312977313995, 0.0818943977355957, 0.2129845768213272, -0.042410045862197876, 0.11652722954750061, -0.39034977555274963, 0.04128977656364441, 0.2123567909002304, -0.230890691280365, 0.06845765560865402, -0.2570657432079315, 0.16785027086734772, -0.09324291348457336, 0.025834325700998306, 0.16631808876991272, -0.00010973814642056823, 0.006797894835472107, -0.015387743711471558, -0.13856983184814453, -0.007274414878338575, 0.04019510746002197, 0.11813420802354813, -0.21172915399074554, -0.10918302088975906, -0.31213220953941345, -0.33201220631599426, -0.037661660462617874, 0.017809096723794937, -0.3556599020957947, 0.3128577172756195, 0.14745266735553741, 0.2607105076313019, 0.22490960359573364, -0.2538115978240967, 0.21113674342632294, 0.1451275497674942, -0.02932754158973694, -0.1741972714662552, -0.04230468347668648, -0.04397884011268616, -0.3344639539718628, -0.37676432728767395, 0.525456964969635, -0.07658185809850693, 0.17014150321483612, 0.21466577053070068, 0.2250560224056244, 0.1726061850786209, 0.035937413573265076, 0.14293654263019562, -0.11396264284849167, 0.011095292866230011, -0.2269170582294464, -0.02711626887321472, -0.3733709454536438, -0.41864001750946045, -0.07852613925933838, 0.137624591588974, 0.18104572594165802, -0.13639286160469055, -0.06371629238128662, -0.43193674087524414, 0.23928946256637573, 0.06524854898452759, -0.012788370251655579, 0.11952775716781616, 0.2951611876487732, 0.043805740773677826, 0.19522324204444885, 0.1294461488723755, -0.06973283737897873, -0.02255835384130478, -0.21451729536056519, -0.058972302824258804, 0.22640188038349152, -0.4359855055809021, 0.07208243757486343, 0.4315193295478821, -0.1404145359992981, -0.05324290320277214, 0.0306372232735157, 0.39193007349967957, 0.07921387255191803, 0.11681866645812988, -0.008491560816764832, -0.10416118800640106, 0.5898669958114624, 0.044322285801172256, 0.17118452489376068, -0.01497051864862442, -0.19827789068222046, -0.3560173213481903, 0.004987429827451706, 0.2880176305770874, -0.2866014540195465, 0.4284125566482544, 0.40236005187034607, -0.0422595851123333, -0.12152935564517975, -0.030101587995886803, 0.09402371197938919, -0.33085376024246216, 0.055669233202934265, 0.05872948467731476, 0.31395789980888367, -0.22545060515403748, 0.1492653489112854, -0.1737677901983261, -0.10335496068000793, -0.38496947288513184, 0.12735126912593842, -0.33784639835357666, -0.03332383185625076, -0.4962179958820343, 0.0015164613723754883, 0.28538399934768677, -0.002860570326447487, 0.28535300493240356, -0.03844557702541351, -0.13722029328346252, -0.301989883184433, 0.10317452996969223, -0.39851242303848267, 0.5702308416366577, -0.1072293221950531, -0.1970168650150299, 0.24102140963077545, 0.13037806749343872, 0.08526398241519928, -0.29768770933151245, -0.1983770728111267, -0.16980770230293274, -0.2090688943862915, 0.2598598897457123, 0.2422453612089157, -0.07942710071802139, 0.1413964033126831, 0.09248155355453491, 0.07767830789089203, 0.3432585597038269, -0.2328546792268753, 0.10284028947353363, -0.2014889419078827, -0.2819805443286896, -0.20435944199562073, 0.16256943345069885, 0.2623099088668823, 0.0667768269777298, -0.13737235963344574, 0.40179744362831116, 0.23718246817588806, 0.14729131758213043, 0.4138064980506897, -0.196533665060997, 0.30949509143829346, -0.4334923028945923, 0.34861892461776733, 0.5064778923988342, -0.32673120498657227, -0.49634435772895813, 0.09944827109575272, -0.1563153862953186, 0.1002877876162529, 0.10244903713464737, 0.1513729840517044, 0.02908817119896412, 0.014419601298868656, 0.12853626906871796, 0.01870240643620491, -0.0005144402384757996, 0.20842039585113525, -0.3169277310371399, -0.021933740004897118, 0.5500645041465759, -0.25246739387512207, 0.028472337871789932, -0.1312367022037506, -0.05504260212182999, 0.021932099014520645, 0.006768524646759033, -0.11218959093093872, 0.12185925245285034, 0.1146051287651062, -0.05687848851084709, -0.15818428993225098, -0.10643843561410904, -0.02469290792942047, -0.2882624864578247, 0.23830124735832214, -0.14981965720653534, -0.09996873140335083, 0.06172499060630798, -0.09704941511154175, -0.1620030403137207, -0.1292746216058731, -0.1809171587228775, -0.27306482195854187, 0.20130175352096558, 0.23052407801151276, 0.3202403485774994, -0.07290694117546082, 0.22520966827869415, 0.3680819571018219, 0.04836329072713852, -0.21607185900211334, -0.4300349950790405, 0.08568014949560165, -0.17773526906967163, -0.21882861852645874, -0.074846550822258, 0.22306931018829346, 0.2090265452861786, -0.0479268915951252, -0.14242854714393616, 0.1482759267091751, -0.22208523750305176, 0.41891855001449585, -0.13702836632728577, 0.2694522738456726, -0.06965650618076324, 0.029306214302778244, 0.016667230054736137, 0.2540055215358734, 0.05882029980421066, -0.08980119228363037, -0.07270593196153641, 0.47423404455184937, 0.18140165507793427, 0.17011570930480957, -0.04807114973664284, -0.2162683755159378, 0.05198418349027634, -0.1783173829317093, -0.12091565877199173, -0.11881639063358307, 0.1364068239927292, -0.0805719643831253, 0.14888811111450195, 0.15232901275157928, -0.10698775947093964, 0.3120076358318329, 0.7600834369659424, 0.19691896438598633, -0.011016584932804108, 0.015740428119897842, -0.05721340328454971, -0.2177915871143341, 0.33385711908340454, -0.058683205395936966, 0.31038856506347656, 0.10473139584064484, 0.12716472148895264, -0.04620931297540665, -0.21964949369430542, -0.1544356644153595, 0.12464860081672668, -0.21436886489391327, -0.19197627902030945, 0.3080342710018158, 0.28208714723587036, -0.1583673059940338, -0.3537829518318176, 0.09401586651802063, -0.01276421919465065, 0.02619149163365364, -0.1580010950565338, 0.21723711490631104, -0.166529580950737, -0.2780844271183014, -0.2864159643650055, 0.13615699112415314, -0.1222032904624939, -0.09216521680355072, -0.04126402735710144, 0.42542994022369385, -0.03665407374501228, 0.19515199959278107, -0.3149715065956116, 0.0895179957151413, 0.2266039401292801, -0.26632726192474365, -0.07168534398078918, -0.1902872771024704, -0.39492014050483704, 0.042231734842061996, -0.09523066878318787, -0.0929960310459137, 0.3783302307128906, -0.047316424548625946, -0.06796006858348846, -0.2881466746330261, -0.17254027724266052, 0.10717801749706268, -0.09261122345924377, -0.15536727011203766, 0.16724826395511627, -0.03133828938007355, -0.10115892440080643, -0.29874566197395325, 0.2937973141670227, -0.30531302094459534, -0.2991430461406708, -0.05198727175593376, 0.09890619665384293, 0.030108986422419548, -0.2803470194339752, -0.4286248981952667, 0.27357742190361023, -0.31724435091018677, 0.07546759396791458, -0.14892880618572235, 0.18622717261314392, 0.33093392848968506, 0.008953814394772053, -0.014561107382178307, -0.06404063105583191, 0.16439421474933624, -0.4674096703529358, -0.14095471799373627, 0.32257843017578125, 0.01876021921634674, -0.21630443632602692, -0.3245909512042999, -0.2757244110107422, 0.34772399067878723, 0.19238810241222382, -0.5803955793380737, -0.21442262828350067, -0.048900216817855835, 0.2480318248271942, -0.05125897377729416, -0.04649518430233002, 0.5745418071746826, 0.01972545124590397, -0.17946316301822662, -0.16727802157402039, -0.33183929324150085, 0.35973066091537476, 0.4192623794078827, 0.24948757886886597, -0.011158477514982224, 0.17781516909599304, 0.20092011988162994, 0.6707470417022705, 0.2019203156232834, -0.2590579092502594, 0.33173102140426636, 0.10723146051168442, 0.07499773055315018, -0.2034967839717865, -0.3130229413509369, 0.18576018512248993, -0.18160304427146912, 0.01973230391740799, 0.27198368310928345, 0.006250595673918724, -0.2801837921142578, 0.017742767930030823, 0.21881836652755737, -0.2253446877002716, -0.328058123588562, 0.2597455084323883, -0.3424280285835266, 0.24094843864440918, 0.036369793117046356, 0.3157997131347656, -0.5820137858390808, -0.24272391200065613, 0.19462794065475464, -0.3359457850456238, 0.27984461188316345, 0.04449816048145294, -0.5836011171340942, 0.12147033214569092, -0.5380215048789978, 0.18194156885147095, 0.31462806463241577, 0.27240774035453796, -0.09202472865581512, -0.2465253472328186, 0.043347518891096115, -0.14713148772716522, 0.5711420178413391, 0.04097308963537216, 0.40857934951782227, 0.16992636024951935, -0.13024693727493286, -0.19947344064712524, -0.22471317648887634, 0.017234940081834793, 0.28013670444488525, 0.3265002369880676, 0.400847464799881, -0.345244437456131, -0.3632737100124359, -0.2657657265663147, 0.007363845594227314, -0.046181585639715195, 0.007106936536729336, -0.12583458423614502, 0.18938523530960083, 0.01648571901023388, 0.17434579133987427, -0.014980330131947994, 0.11830316483974457, 0.016039708629250526, -0.13654674589633942, -0.03913415968418121, -0.21942690014839172, 0.12824007868766785, 0.22643797099590302, 0.4735948443412781, 0.1503245085477829, 0.03246166184544563, 0.033214349299669266, 0.21515773236751556, -0.5671082735061646, 0.2659573554992676, -0.08884961903095245, -0.004131801426410675, 0.07464389503002167, 0.14443765580654144, -0.030300907790660858, 0.3378967344760895, -0.04252535104751587, 0.1457139402627945, -0.12862969934940338, 0.008957155048847198, -0.5539682507514954, 0.4854547381401062, 0.3874596357345581, 0.13246718049049377, -0.044888898730278015, -0.24685737490653992, 0.17346280813217163, 0.07412064075469971, -0.27653080224990845, 0.13225962221622467, -0.43069255352020264, -0.4319950044155121, 0.18579739332199097, 0.2094864398241043, 0.9232466220855713, 0.33202219009399414, 0.37195777893066406, 0.10070358961820602, 0.10358034074306488, 0.26425158977508545, -0.23411162197589874, 0.40924984216690063, -0.3693644404411316, -0.0463179312646389, 0.024538787081837654, -0.16673341393470764, -0.22738268971443176, -0.010426290333271027, -0.010993778705596924, 0.30259543657302856, 0.32404959201812744, 0.5999362468719482, -0.11212000250816345, 0.16670005023479462, 0.215691938996315, -0.15810950100421906, -0.24923086166381836, 0.20107315480709076, -0.2923040986061096, 0.31524065136909485, -0.11240848153829575, -0.07410834729671478, -0.01650886982679367, -0.35134220123291016, 0.06714004278182983, 0.0609416738152504, -0.3365170359611511, 0.3048418164253235, 0.13346192240715027, -0.27266669273376465, -0.29698777198791504, 0.2550782561302185, 0.11198711395263672, 0.08252803981304169, -0.048599302768707275, -0.035185206681489944, 0.4006449580192566, 0.35045796632766724, -0.006891845725476742, -0.20158877968788147, 0.18641246855258942, 0.06387515366077423, -0.012894652783870697, 0.030242733657360077, -0.05640551447868347, -0.3015826344490051, -0.2525118589401245, 0.14820465445518494, 0.0033729001879692078, -0.4129619598388672, -0.055318914353847504, 0.06144453212618828, -0.0667886808514595, -0.09413722157478333, 0.15088041126728058, -0.026068218052387238, -0.10794125497341156, 0.3151545822620392, 0.11733834445476532, -0.24304209649562836, 0.05235828459262848, 0.47483235597610474, -0.04039648920297623, 0.031818099319934845, 0.4154282510280609, -0.15281841158866882, -0.10380619764328003, -0.25539109110832214, -0.28417226672172546, -0.10206017643213272, -0.44884029030799866, 0.11220822483301163, -0.2194056510925293, -0.40191107988357544, 0.04423205554485321, 0.2808186709880829, 0.014706198126077652, 0.3184961974620819, -0.011618245393037796, -0.12495805323123932, -0.3769727349281311, -0.10379086434841156, -0.10658213496208191, 0.26717838644981384, 0.27951928973197937, 0.2843479812145233, -0.29586678743362427, 0.2492268979549408, -0.38310953974723816, -0.10714222490787506, -0.1871705949306488, 0.4379130005836487, -0.0391375795006752, -0.36558276414871216, -0.009147514589130878, 0.0018548620864748955, 0.12997101247310638, 0.17754510045051575, -0.6386584639549255, -0.29261907935142517, -0.09099219739437103, 0.08186075091362, -0.007588040083646774, -0.1598217487335205, -0.23928231000900269, -0.01721424236893654, 0.18238495290279388, -0.427402526140213, 0.09215688705444336, 0.15121665596961975, 0.04703199863433838, 0.23718711733818054, -0.1610298752784729, 0.1214555948972702, 0.28816717863082886, -0.18652743101119995, -0.1841491460800171, -0.07622881233692169, 0.08048095554113388, 0.25701695680618286, -0.06946776062250137, 0.020479001104831696, -0.2104034423828125, 0.030471812933683395, 0.08927443623542786, 0.14279493689537048, 0.23692923784255981, -0.024752572178840637, -0.21566128730773926, 0.0005505271255970001, 0.18279926478862762, 0.11411947757005692, -0.07855326682329178, -0.32897576689720154, 0.2802233397960663, 0.2783650755882263, -0.11751837283372879, -0.0102371945977211, 0.1677074134349823, -0.09938855469226837, 0.25285133719444275, 0.20715340971946716, 0.10758769512176514, 0.05261163413524628, 0.26961183547973633, 0.12649410963058472, 0.40504008531570435, -0.34832900762557983, 0.08942112326622009, 0.06441552191972733, 0.10164903104305267, 0.03777839243412018, 0.38121557235717773, -0.1532439887523651, 0.08992154896259308, 0.14838308095932007, 0.18833690881729126, 0.20253999531269073, 0.26289018988609314, 0.21354106068611145, -0.1878642439842224, -0.03412104770541191, -0.24108661711215973, 0.1436368077993393, -0.14272664487361908, 0.4156598448753357, 0.15059244632720947, 0.19285538792610168, -0.11485158652067184, -0.4619392156600952, -0.2462674379348755, 0.23536445200443268, -0.3245418667793274, -0.30920952558517456, -0.04867886006832123, -0.08018464595079422, 0.05592931807041168, 0.19456081092357635, -0.07928960025310516, 0.2038593888282776, 0.3428382873535156, -0.00865977257490158, -0.013750068843364716, -0.08435389399528503, -0.170617014169693, 0.11397916823625565, 0.2748645842075348, -0.08173888921737671, 0.446936696767807, 0.010800972580909729, -0.051569364964962006, -0.2344720959663391, 0.4114176034927368, 0.2861287593841553, 0.19494742155075073, -0.1688622683286667, 0.06395319104194641, 0.20064730942249298, 0.03061327524483204, -0.35468870401382446, 0.3286513686180115, 0.15114429593086243, -0.2930281460285187, 0.3680053651332855, 0.12379556894302368, -0.23215582966804504, 0.2764035165309906, -0.1041957288980484, -0.20146708190441132, -0.049888961017131805, 0.3224371671676636, 0.09970720112323761, -0.03993305563926697, -0.1949632167816162, 0.1111203134059906, -0.1445714384317398, -0.28533875942230225, 0.15934628248214722, -0.023898672312498093, 0.4676608443260193, -0.2756762206554413, 0.11720523238182068, -0.06182504817843437, 0.6220859885215759, -0.02922816202044487, 0.12432920932769775, -0.35013818740844727, 0.10264356434345245, -0.3331668972969055, 0.198409765958786, -0.08381709456443787, 0.15374305844306946, -0.39758387207984924, 0.20426522195339203, -0.11549905687570572, -0.02346649393439293, -0.11804220080375671, -0.007886471226811409, 0.14740730822086334, 0.17893224954605103, -0.3869178295135498, 0.1554817110300064, -0.08625976741313934, -0.006267093122005463, 0.29177382588386536, -0.37667521834373474, 0.22199100255966187, 0.042478568851947784, 0.20691052079200745, 0.010985862463712692, -0.02172362059354782, -0.054806359112262726, 0.33657363057136536, 0.6232408881187439, 0.34702926874160767, 0.14255046844482422, -0.10102763026952744, 0.08155438303947449, -0.3474372923374176, 0.014104174450039864, -0.06088797375559807, -0.20964977145195007, 0.09756873548030853, 0.5258306264877319, -0.22562795877456665, 0.13194617629051208, -0.178046315908432, 0.23669393360614777, 0.3822764754295349, -0.1527290940284729, 0.0618000254034996, 0.04822578281164169, -0.2598402798175812, -0.005199543200433254, -0.12586651742458344, 0.45379629731178284, 0.007182497531175613, 0.32383060455322266, -0.3592911958694458, -0.5417625904083252, 0.43290984630584717, -0.45112577080726624, -0.30252525210380554, 0.12327728420495987, 0.23878571391105652, 0.34956857562065125, -0.06000012159347534, -0.5436676144599915, -0.17249035835266113, 0.2691141963005066, -0.030661150813102722, -0.1570448875427246, 0.20768548548221588, -0.14864324033260345, -0.04107460752129555, -0.1167544275522232, 0.17646947503089905, 0.10152687877416611, -0.2914122939109802, 0.24369880557060242, -0.3701038360595703 ]
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.
[ -0.1263233721256256, -0.22070911526679993, -0.10063229501247406, 0.27801111340522766, 0.24547500908374786, -0.20522364974021912, 0.13721121847629547, 0.512859046459198, 0.18828602135181427, -0.06233552098274231, 0.1278189718723297, 0.38913494348526, 0.0037763379514217377, -0.16061760485172272, -0.010743602178990841, 0.024212224408984184, 0.09608031809329987, 0.23505264520645142, -0.1455627828836441, -0.16482312977313995, 0.0818943977355957, 0.2129845768213272, -0.042410045862197876, 0.11652722954750061, -0.39034977555274963, 0.04128977656364441, 0.2123567909002304, -0.230890691280365, 0.06845765560865402, -0.2570657432079315, 0.16785027086734772, -0.09324291348457336, 0.025834325700998306, 0.16631808876991272, -0.00010973814642056823, 0.006797894835472107, -0.015387743711471558, -0.13856983184814453, -0.007274414878338575, 0.04019510746002197, 0.11813420802354813, -0.21172915399074554, -0.10918302088975906, -0.31213220953941345, -0.33201220631599426, -0.037661660462617874, 0.017809096723794937, -0.3556599020957947, 0.3128577172756195, 0.14745266735553741, 0.2607105076313019, 0.22490960359573364, -0.2538115978240967, 0.21113674342632294, 0.1451275497674942, -0.02932754158973694, -0.1741972714662552, -0.04230468347668648, -0.04397884011268616, -0.3344639539718628, -0.37676432728767395, 0.525456964969635, -0.07658185809850693, 0.17014150321483612, 0.21466577053070068, 0.2250560224056244, 0.1726061850786209, 0.035937413573265076, 0.14293654263019562, -0.11396264284849167, 0.011095292866230011, -0.2269170582294464, -0.02711626887321472, -0.3733709454536438, -0.41864001750946045, -0.07852613925933838, 0.137624591588974, 0.18104572594165802, -0.13639286160469055, -0.06371629238128662, -0.43193674087524414, 0.23928946256637573, 0.06524854898452759, -0.012788370251655579, 0.11952775716781616, 0.2951611876487732, 0.043805740773677826, 0.19522324204444885, 0.1294461488723755, -0.06973283737897873, -0.02255835384130478, -0.21451729536056519, -0.058972302824258804, 0.22640188038349152, -0.4359855055809021, 0.07208243757486343, 0.4315193295478821, -0.1404145359992981, -0.05324290320277214, 0.0306372232735157, 0.39193007349967957, 0.07921387255191803, 0.11681866645812988, -0.008491560816764832, -0.10416118800640106, 0.5898669958114624, 0.044322285801172256, 0.17118452489376068, -0.01497051864862442, -0.19827789068222046, -0.3560173213481903, 0.004987429827451706, 0.2880176305770874, -0.2866014540195465, 0.4284125566482544, 0.40236005187034607, -0.0422595851123333, -0.12152935564517975, -0.030101587995886803, 0.09402371197938919, -0.33085376024246216, 0.055669233202934265, 0.05872948467731476, 0.31395789980888367, -0.22545060515403748, 0.1492653489112854, -0.1737677901983261, -0.10335496068000793, -0.38496947288513184, 0.12735126912593842, -0.33784639835357666, -0.03332383185625076, -0.4962179958820343, 0.0015164613723754883, 0.28538399934768677, -0.002860570326447487, 0.28535300493240356, -0.03844557702541351, -0.13722029328346252, -0.301989883184433, 0.10317452996969223, -0.39851242303848267, 0.5702308416366577, -0.1072293221950531, -0.1970168650150299, 0.24102140963077545, 0.13037806749343872, 0.08526398241519928, -0.29768770933151245, -0.1983770728111267, -0.16980770230293274, -0.2090688943862915, 0.2598598897457123, 0.2422453612089157, -0.07942710071802139, 0.1413964033126831, 0.09248155355453491, 0.07767830789089203, 0.3432585597038269, -0.2328546792268753, 0.10284028947353363, -0.2014889419078827, -0.2819805443286896, -0.20435944199562073, 0.16256943345069885, 0.2623099088668823, 0.0667768269777298, -0.13737235963344574, 0.40179744362831116, 0.23718246817588806, 0.14729131758213043, 0.4138064980506897, -0.196533665060997, 0.30949509143829346, -0.4334923028945923, 0.34861892461776733, 0.5064778923988342, -0.32673120498657227, -0.49634435772895813, 0.09944827109575272, -0.1563153862953186, 0.1002877876162529, 0.10244903713464737, 0.1513729840517044, 0.02908817119896412, 0.014419601298868656, 0.12853626906871796, 0.01870240643620491, -0.0005144402384757996, 0.20842039585113525, -0.3169277310371399, -0.021933740004897118, 0.5500645041465759, -0.25246739387512207, 0.028472337871789932, -0.1312367022037506, -0.05504260212182999, 0.021932099014520645, 0.006768524646759033, -0.11218959093093872, 0.12185925245285034, 0.1146051287651062, -0.05687848851084709, -0.15818428993225098, -0.10643843561410904, -0.02469290792942047, -0.2882624864578247, 0.23830124735832214, -0.14981965720653534, -0.09996873140335083, 0.06172499060630798, -0.09704941511154175, -0.1620030403137207, -0.1292746216058731, -0.1809171587228775, -0.27306482195854187, 0.20130175352096558, 0.23052407801151276, 0.3202403485774994, -0.07290694117546082, 0.22520966827869415, 0.3680819571018219, 0.04836329072713852, -0.21607185900211334, -0.4300349950790405, 0.08568014949560165, -0.17773526906967163, -0.21882861852645874, -0.074846550822258, 0.22306931018829346, 0.2090265452861786, -0.0479268915951252, -0.14242854714393616, 0.1482759267091751, -0.22208523750305176, 0.41891855001449585, -0.13702836632728577, 0.2694522738456726, -0.06965650618076324, 0.029306214302778244, 0.016667230054736137, 0.2540055215358734, 0.05882029980421066, -0.08980119228363037, -0.07270593196153641, 0.47423404455184937, 0.18140165507793427, 0.17011570930480957, -0.04807114973664284, -0.2162683755159378, 0.05198418349027634, -0.1783173829317093, -0.12091565877199173, -0.11881639063358307, 0.1364068239927292, -0.0805719643831253, 0.14888811111450195, 0.15232901275157928, -0.10698775947093964, 0.3120076358318329, 0.7600834369659424, 0.19691896438598633, -0.011016584932804108, 0.015740428119897842, -0.05721340328454971, -0.2177915871143341, 0.33385711908340454, -0.058683205395936966, 0.31038856506347656, 0.10473139584064484, 0.12716472148895264, -0.04620931297540665, -0.21964949369430542, -0.1544356644153595, 0.12464860081672668, -0.21436886489391327, -0.19197627902030945, 0.3080342710018158, 0.28208714723587036, -0.1583673059940338, -0.3537829518318176, 0.09401586651802063, -0.01276421919465065, 0.02619149163365364, -0.1580010950565338, 0.21723711490631104, -0.166529580950737, -0.2780844271183014, -0.2864159643650055, 0.13615699112415314, -0.1222032904624939, -0.09216521680355072, -0.04126402735710144, 0.42542994022369385, -0.03665407374501228, 0.19515199959278107, -0.3149715065956116, 0.0895179957151413, 0.2266039401292801, -0.26632726192474365, -0.07168534398078918, -0.1902872771024704, -0.39492014050483704, 0.042231734842061996, -0.09523066878318787, -0.0929960310459137, 0.3783302307128906, -0.047316424548625946, -0.06796006858348846, -0.2881466746330261, -0.17254027724266052, 0.10717801749706268, -0.09261122345924377, -0.15536727011203766, 0.16724826395511627, -0.03133828938007355, -0.10115892440080643, -0.29874566197395325, 0.2937973141670227, -0.30531302094459534, -0.2991430461406708, -0.05198727175593376, 0.09890619665384293, 0.030108986422419548, -0.2803470194339752, -0.4286248981952667, 0.27357742190361023, -0.31724435091018677, 0.07546759396791458, -0.14892880618572235, 0.18622717261314392, 0.33093392848968506, 0.008953814394772053, -0.014561107382178307, -0.06404063105583191, 0.16439421474933624, -0.4674096703529358, -0.14095471799373627, 0.32257843017578125, 0.01876021921634674, -0.21630443632602692, -0.3245909512042999, -0.2757244110107422, 0.34772399067878723, 0.19238810241222382, -0.5803955793380737, -0.21442262828350067, -0.048900216817855835, 0.2480318248271942, -0.05125897377729416, -0.04649518430233002, 0.5745418071746826, 0.01972545124590397, -0.17946316301822662, -0.16727802157402039, -0.33183929324150085, 0.35973066091537476, 0.4192623794078827, 0.24948757886886597, -0.011158477514982224, 0.17781516909599304, 0.20092011988162994, 0.6707470417022705, 0.2019203156232834, -0.2590579092502594, 0.33173102140426636, 0.10723146051168442, 0.07499773055315018, -0.2034967839717865, -0.3130229413509369, 0.18576018512248993, -0.18160304427146912, 0.01973230391740799, 0.27198368310928345, 0.006250595673918724, -0.2801837921142578, 0.017742767930030823, 0.21881836652755737, -0.2253446877002716, -0.328058123588562, 0.2597455084323883, -0.3424280285835266, 0.24094843864440918, 0.036369793117046356, 0.3157997131347656, -0.5820137858390808, -0.24272391200065613, 0.19462794065475464, -0.3359457850456238, 0.27984461188316345, 0.04449816048145294, -0.5836011171340942, 0.12147033214569092, -0.5380215048789978, 0.18194156885147095, 0.31462806463241577, 0.27240774035453796, -0.09202472865581512, -0.2465253472328186, 0.043347518891096115, -0.14713148772716522, 0.5711420178413391, 0.04097308963537216, 0.40857934951782227, 0.16992636024951935, -0.13024693727493286, -0.19947344064712524, -0.22471317648887634, 0.017234940081834793, 0.28013670444488525, 0.3265002369880676, 0.400847464799881, -0.345244437456131, -0.3632737100124359, -0.2657657265663147, 0.007363845594227314, -0.046181585639715195, 0.007106936536729336, -0.12583458423614502, 0.18938523530960083, 0.01648571901023388, 0.17434579133987427, -0.014980330131947994, 0.11830316483974457, 0.016039708629250526, -0.13654674589633942, -0.03913415968418121, -0.21942690014839172, 0.12824007868766785, 0.22643797099590302, 0.4735948443412781, 0.1503245085477829, 0.03246166184544563, 0.033214349299669266, 0.21515773236751556, -0.5671082735061646, 0.2659573554992676, -0.08884961903095245, -0.004131801426410675, 0.07464389503002167, 0.14443765580654144, -0.030300907790660858, 0.3378967344760895, -0.04252535104751587, 0.1457139402627945, -0.12862969934940338, 0.008957155048847198, -0.5539682507514954, 0.4854547381401062, 0.3874596357345581, 0.13246718049049377, -0.044888898730278015, -0.24685737490653992, 0.17346280813217163, 0.07412064075469971, -0.27653080224990845, 0.13225962221622467, -0.43069255352020264, -0.4319950044155121, 0.18579739332199097, 0.2094864398241043, 0.9232466220855713, 0.33202219009399414, 0.37195777893066406, 0.10070358961820602, 0.10358034074306488, 0.26425158977508545, -0.23411162197589874, 0.40924984216690063, -0.3693644404411316, -0.0463179312646389, 0.024538787081837654, -0.16673341393470764, -0.22738268971443176, -0.010426290333271027, -0.010993778705596924, 0.30259543657302856, 0.32404959201812744, 0.5999362468719482, -0.11212000250816345, 0.16670005023479462, 0.215691938996315, -0.15810950100421906, -0.24923086166381836, 0.20107315480709076, -0.2923040986061096, 0.31524065136909485, -0.11240848153829575, -0.07410834729671478, -0.01650886982679367, -0.35134220123291016, 0.06714004278182983, 0.0609416738152504, -0.3365170359611511, 0.3048418164253235, 0.13346192240715027, -0.27266669273376465, -0.29698777198791504, 0.2550782561302185, 0.11198711395263672, 0.08252803981304169, -0.048599302768707275, -0.035185206681489944, 0.4006449580192566, 0.35045796632766724, -0.006891845725476742, -0.20158877968788147, 0.18641246855258942, 0.06387515366077423, -0.012894652783870697, 0.030242733657360077, -0.05640551447868347, -0.3015826344490051, -0.2525118589401245, 0.14820465445518494, 0.0033729001879692078, -0.4129619598388672, -0.055318914353847504, 0.06144453212618828, -0.0667886808514595, -0.09413722157478333, 0.15088041126728058, -0.026068218052387238, -0.10794125497341156, 0.3151545822620392, 0.11733834445476532, -0.24304209649562836, 0.05235828459262848, 0.47483235597610474, -0.04039648920297623, 0.031818099319934845, 0.4154282510280609, -0.15281841158866882, -0.10380619764328003, -0.25539109110832214, -0.28417226672172546, -0.10206017643213272, -0.44884029030799866, 0.11220822483301163, -0.2194056510925293, -0.40191107988357544, 0.04423205554485321, 0.2808186709880829, 0.014706198126077652, 0.3184961974620819, -0.011618245393037796, -0.12495805323123932, -0.3769727349281311, -0.10379086434841156, -0.10658213496208191, 0.26717838644981384, 0.27951928973197937, 0.2843479812145233, -0.29586678743362427, 0.2492268979549408, -0.38310953974723816, -0.10714222490787506, -0.1871705949306488, 0.4379130005836487, -0.0391375795006752, -0.36558276414871216, -0.009147514589130878, 0.0018548620864748955, 0.12997101247310638, 0.17754510045051575, -0.6386584639549255, -0.29261907935142517, -0.09099219739437103, 0.08186075091362, -0.007588040083646774, -0.1598217487335205, -0.23928231000900269, -0.01721424236893654, 0.18238495290279388, -0.427402526140213, 0.09215688705444336, 0.15121665596961975, 0.04703199863433838, 0.23718711733818054, -0.1610298752784729, 0.1214555948972702, 0.28816717863082886, -0.18652743101119995, -0.1841491460800171, -0.07622881233692169, 0.08048095554113388, 0.25701695680618286, -0.06946776062250137, 0.020479001104831696, -0.2104034423828125, 0.030471812933683395, 0.08927443623542786, 0.14279493689537048, 0.23692923784255981, -0.024752572178840637, -0.21566128730773926, 0.0005505271255970001, 0.18279926478862762, 0.11411947757005692, -0.07855326682329178, -0.32897576689720154, 0.2802233397960663, 0.2783650755882263, -0.11751837283372879, -0.0102371945977211, 0.1677074134349823, -0.09938855469226837, 0.25285133719444275, 0.20715340971946716, 0.10758769512176514, 0.05261163413524628, 0.26961183547973633, 0.12649410963058472, 0.40504008531570435, -0.34832900762557983, 0.08942112326622009, 0.06441552191972733, 0.10164903104305267, 0.03777839243412018, 0.38121557235717773, -0.1532439887523651, 0.08992154896259308, 0.14838308095932007, 0.18833690881729126, 0.20253999531269073, 0.26289018988609314, 0.21354106068611145, -0.1878642439842224, -0.03412104770541191, -0.24108661711215973, 0.1436368077993393, -0.14272664487361908, 0.4156598448753357, 0.15059244632720947, 0.19285538792610168, -0.11485158652067184, -0.4619392156600952, -0.2462674379348755, 0.23536445200443268, -0.3245418667793274, -0.30920952558517456, -0.04867886006832123, -0.08018464595079422, 0.05592931807041168, 0.19456081092357635, -0.07928960025310516, 0.2038593888282776, 0.3428382873535156, -0.00865977257490158, -0.013750068843364716, -0.08435389399528503, -0.170617014169693, 0.11397916823625565, 0.2748645842075348, -0.08173888921737671, 0.446936696767807, 0.010800972580909729, -0.051569364964962006, -0.2344720959663391, 0.4114176034927368, 0.2861287593841553, 0.19494742155075073, -0.1688622683286667, 0.06395319104194641, 0.20064730942249298, 0.03061327524483204, -0.35468870401382446, 0.3286513686180115, 0.15114429593086243, -0.2930281460285187, 0.3680053651332855, 0.12379556894302368, -0.23215582966804504, 0.2764035165309906, -0.1041957288980484, -0.20146708190441132, -0.049888961017131805, 0.3224371671676636, 0.09970720112323761, -0.03993305563926697, -0.1949632167816162, 0.1111203134059906, -0.1445714384317398, -0.28533875942230225, 0.15934628248214722, -0.023898672312498093, 0.4676608443260193, -0.2756762206554413, 0.11720523238182068, -0.06182504817843437, 0.6220859885215759, -0.02922816202044487, 0.12432920932769775, -0.35013818740844727, 0.10264356434345245, -0.3331668972969055, 0.198409765958786, -0.08381709456443787, 0.15374305844306946, -0.39758387207984924, 0.20426522195339203, -0.11549905687570572, -0.02346649393439293, -0.11804220080375671, -0.007886471226811409, 0.14740730822086334, 0.17893224954605103, -0.3869178295135498, 0.1554817110300064, -0.08625976741313934, -0.006267093122005463, 0.29177382588386536, -0.37667521834373474, 0.22199100255966187, 0.042478568851947784, 0.20691052079200745, 0.010985862463712692, -0.02172362059354782, -0.054806359112262726, 0.33657363057136536, 0.6232408881187439, 0.34702926874160767, 0.14255046844482422, -0.10102763026952744, 0.08155438303947449, -0.3474372923374176, 0.014104174450039864, -0.06088797375559807, -0.20964977145195007, 0.09756873548030853, 0.5258306264877319, -0.22562795877456665, 0.13194617629051208, -0.178046315908432, 0.23669393360614777, 0.3822764754295349, -0.1527290940284729, 0.0618000254034996, 0.04822578281164169, -0.2598402798175812, -0.005199543200433254, -0.12586651742458344, 0.45379629731178284, 0.007182497531175613, 0.32383060455322266, -0.3592911958694458, -0.5417625904083252, 0.43290984630584717, -0.45112577080726624, -0.30252525210380554, 0.12327728420495987, 0.23878571391105652, 0.34956857562065125, -0.06000012159347534, -0.5436676144599915, -0.17249035835266113, 0.2691141963005066, -0.030661150813102722, -0.1570448875427246, 0.20768548548221588, -0.14864324033260345, -0.04107460752129555, -0.1167544275522232, 0.17646947503089905, 0.10152687877416611, -0.2914122939109802, 0.24369880557060242, -0.3701038360595703 ]
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
[ -0.1263233721256256, -0.22070911526679993, -0.10063229501247406, 0.27801111340522766, 0.24547500908374786, -0.20522364974021912, 0.13721121847629547, 0.512859046459198, 0.18828602135181427, -0.06233552098274231, 0.1278189718723297, 0.38913494348526, 0.0037763379514217377, -0.16061760485172272, -0.010743602178990841, 0.024212224408984184, 0.09608031809329987, 0.23505264520645142, -0.1455627828836441, -0.16482312977313995, 0.0818943977355957, 0.2129845768213272, -0.042410045862197876, 0.11652722954750061, -0.39034977555274963, 0.04128977656364441, 0.2123567909002304, -0.230890691280365, 0.06845765560865402, -0.2570657432079315, 0.16785027086734772, -0.09324291348457336, 0.025834325700998306, 0.16631808876991272, -0.00010973814642056823, 0.006797894835472107, -0.015387743711471558, -0.13856983184814453, -0.007274414878338575, 0.04019510746002197, 0.11813420802354813, -0.21172915399074554, -0.10918302088975906, -0.31213220953941345, -0.33201220631599426, -0.037661660462617874, 0.017809096723794937, -0.3556599020957947, 0.3128577172756195, 0.14745266735553741, 0.2607105076313019, 0.22490960359573364, -0.2538115978240967, 0.21113674342632294, 0.1451275497674942, -0.02932754158973694, -0.1741972714662552, -0.04230468347668648, -0.04397884011268616, -0.3344639539718628, -0.37676432728767395, 0.525456964969635, -0.07658185809850693, 0.17014150321483612, 0.21466577053070068, 0.2250560224056244, 0.1726061850786209, 0.035937413573265076, 0.14293654263019562, -0.11396264284849167, 0.011095292866230011, -0.2269170582294464, -0.02711626887321472, -0.3733709454536438, -0.41864001750946045, -0.07852613925933838, 0.137624591588974, 0.18104572594165802, -0.13639286160469055, -0.06371629238128662, -0.43193674087524414, 0.23928946256637573, 0.06524854898452759, -0.012788370251655579, 0.11952775716781616, 0.2951611876487732, 0.043805740773677826, 0.19522324204444885, 0.1294461488723755, -0.06973283737897873, -0.02255835384130478, -0.21451729536056519, -0.058972302824258804, 0.22640188038349152, -0.4359855055809021, 0.07208243757486343, 0.4315193295478821, -0.1404145359992981, -0.05324290320277214, 0.0306372232735157, 0.39193007349967957, 0.07921387255191803, 0.11681866645812988, -0.008491560816764832, -0.10416118800640106, 0.5898669958114624, 0.044322285801172256, 0.17118452489376068, -0.01497051864862442, -0.19827789068222046, -0.3560173213481903, 0.004987429827451706, 0.2880176305770874, -0.2866014540195465, 0.4284125566482544, 0.40236005187034607, -0.0422595851123333, -0.12152935564517975, -0.030101587995886803, 0.09402371197938919, -0.33085376024246216, 0.055669233202934265, 0.05872948467731476, 0.31395789980888367, -0.22545060515403748, 0.1492653489112854, -0.1737677901983261, -0.10335496068000793, -0.38496947288513184, 0.12735126912593842, -0.33784639835357666, -0.03332383185625076, -0.4962179958820343, 0.0015164613723754883, 0.28538399934768677, -0.002860570326447487, 0.28535300493240356, -0.03844557702541351, -0.13722029328346252, -0.301989883184433, 0.10317452996969223, -0.39851242303848267, 0.5702308416366577, -0.1072293221950531, -0.1970168650150299, 0.24102140963077545, 0.13037806749343872, 0.08526398241519928, -0.29768770933151245, -0.1983770728111267, -0.16980770230293274, -0.2090688943862915, 0.2598598897457123, 0.2422453612089157, -0.07942710071802139, 0.1413964033126831, 0.09248155355453491, 0.07767830789089203, 0.3432585597038269, -0.2328546792268753, 0.10284028947353363, -0.2014889419078827, -0.2819805443286896, -0.20435944199562073, 0.16256943345069885, 0.2623099088668823, 0.0667768269777298, -0.13737235963344574, 0.40179744362831116, 0.23718246817588806, 0.14729131758213043, 0.4138064980506897, -0.196533665060997, 0.30949509143829346, -0.4334923028945923, 0.34861892461776733, 0.5064778923988342, -0.32673120498657227, -0.49634435772895813, 0.09944827109575272, -0.1563153862953186, 0.1002877876162529, 0.10244903713464737, 0.1513729840517044, 0.02908817119896412, 0.014419601298868656, 0.12853626906871796, 0.01870240643620491, -0.0005144402384757996, 0.20842039585113525, -0.3169277310371399, -0.021933740004897118, 0.5500645041465759, -0.25246739387512207, 0.028472337871789932, -0.1312367022037506, -0.05504260212182999, 0.021932099014520645, 0.006768524646759033, -0.11218959093093872, 0.12185925245285034, 0.1146051287651062, -0.05687848851084709, -0.15818428993225098, -0.10643843561410904, -0.02469290792942047, -0.2882624864578247, 0.23830124735832214, -0.14981965720653534, -0.09996873140335083, 0.06172499060630798, -0.09704941511154175, -0.1620030403137207, -0.1292746216058731, -0.1809171587228775, -0.27306482195854187, 0.20130175352096558, 0.23052407801151276, 0.3202403485774994, -0.07290694117546082, 0.22520966827869415, 0.3680819571018219, 0.04836329072713852, -0.21607185900211334, -0.4300349950790405, 0.08568014949560165, -0.17773526906967163, -0.21882861852645874, -0.074846550822258, 0.22306931018829346, 0.2090265452861786, -0.0479268915951252, -0.14242854714393616, 0.1482759267091751, -0.22208523750305176, 0.41891855001449585, -0.13702836632728577, 0.2694522738456726, -0.06965650618076324, 0.029306214302778244, 0.016667230054736137, 0.2540055215358734, 0.05882029980421066, -0.08980119228363037, -0.07270593196153641, 0.47423404455184937, 0.18140165507793427, 0.17011570930480957, -0.04807114973664284, -0.2162683755159378, 0.05198418349027634, -0.1783173829317093, -0.12091565877199173, -0.11881639063358307, 0.1364068239927292, -0.0805719643831253, 0.14888811111450195, 0.15232901275157928, -0.10698775947093964, 0.3120076358318329, 0.7600834369659424, 0.19691896438598633, -0.011016584932804108, 0.015740428119897842, -0.05721340328454971, -0.2177915871143341, 0.33385711908340454, -0.058683205395936966, 0.31038856506347656, 0.10473139584064484, 0.12716472148895264, -0.04620931297540665, -0.21964949369430542, -0.1544356644153595, 0.12464860081672668, -0.21436886489391327, -0.19197627902030945, 0.3080342710018158, 0.28208714723587036, -0.1583673059940338, -0.3537829518318176, 0.09401586651802063, -0.01276421919465065, 0.02619149163365364, -0.1580010950565338, 0.21723711490631104, -0.166529580950737, -0.2780844271183014, -0.2864159643650055, 0.13615699112415314, -0.1222032904624939, -0.09216521680355072, -0.04126402735710144, 0.42542994022369385, -0.03665407374501228, 0.19515199959278107, -0.3149715065956116, 0.0895179957151413, 0.2266039401292801, -0.26632726192474365, -0.07168534398078918, -0.1902872771024704, -0.39492014050483704, 0.042231734842061996, -0.09523066878318787, -0.0929960310459137, 0.3783302307128906, -0.047316424548625946, -0.06796006858348846, -0.2881466746330261, -0.17254027724266052, 0.10717801749706268, -0.09261122345924377, -0.15536727011203766, 0.16724826395511627, -0.03133828938007355, -0.10115892440080643, -0.29874566197395325, 0.2937973141670227, -0.30531302094459534, -0.2991430461406708, -0.05198727175593376, 0.09890619665384293, 0.030108986422419548, -0.2803470194339752, -0.4286248981952667, 0.27357742190361023, -0.31724435091018677, 0.07546759396791458, -0.14892880618572235, 0.18622717261314392, 0.33093392848968506, 0.008953814394772053, -0.014561107382178307, -0.06404063105583191, 0.16439421474933624, -0.4674096703529358, -0.14095471799373627, 0.32257843017578125, 0.01876021921634674, -0.21630443632602692, -0.3245909512042999, -0.2757244110107422, 0.34772399067878723, 0.19238810241222382, -0.5803955793380737, -0.21442262828350067, -0.048900216817855835, 0.2480318248271942, -0.05125897377729416, -0.04649518430233002, 0.5745418071746826, 0.01972545124590397, -0.17946316301822662, -0.16727802157402039, -0.33183929324150085, 0.35973066091537476, 0.4192623794078827, 0.24948757886886597, -0.011158477514982224, 0.17781516909599304, 0.20092011988162994, 0.6707470417022705, 0.2019203156232834, -0.2590579092502594, 0.33173102140426636, 0.10723146051168442, 0.07499773055315018, -0.2034967839717865, -0.3130229413509369, 0.18576018512248993, -0.18160304427146912, 0.01973230391740799, 0.27198368310928345, 0.006250595673918724, -0.2801837921142578, 0.017742767930030823, 0.21881836652755737, -0.2253446877002716, -0.328058123588562, 0.2597455084323883, -0.3424280285835266, 0.24094843864440918, 0.036369793117046356, 0.3157997131347656, -0.5820137858390808, -0.24272391200065613, 0.19462794065475464, -0.3359457850456238, 0.27984461188316345, 0.04449816048145294, -0.5836011171340942, 0.12147033214569092, -0.5380215048789978, 0.18194156885147095, 0.31462806463241577, 0.27240774035453796, -0.09202472865581512, -0.2465253472328186, 0.043347518891096115, -0.14713148772716522, 0.5711420178413391, 0.04097308963537216, 0.40857934951782227, 0.16992636024951935, -0.13024693727493286, -0.19947344064712524, -0.22471317648887634, 0.017234940081834793, 0.28013670444488525, 0.3265002369880676, 0.400847464799881, -0.345244437456131, -0.3632737100124359, -0.2657657265663147, 0.007363845594227314, -0.046181585639715195, 0.007106936536729336, -0.12583458423614502, 0.18938523530960083, 0.01648571901023388, 0.17434579133987427, -0.014980330131947994, 0.11830316483974457, 0.016039708629250526, -0.13654674589633942, -0.03913415968418121, -0.21942690014839172, 0.12824007868766785, 0.22643797099590302, 0.4735948443412781, 0.1503245085477829, 0.03246166184544563, 0.033214349299669266, 0.21515773236751556, -0.5671082735061646, 0.2659573554992676, -0.08884961903095245, -0.004131801426410675, 0.07464389503002167, 0.14443765580654144, -0.030300907790660858, 0.3378967344760895, -0.04252535104751587, 0.1457139402627945, -0.12862969934940338, 0.008957155048847198, -0.5539682507514954, 0.4854547381401062, 0.3874596357345581, 0.13246718049049377, -0.044888898730278015, -0.24685737490653992, 0.17346280813217163, 0.07412064075469971, -0.27653080224990845, 0.13225962221622467, -0.43069255352020264, -0.4319950044155121, 0.18579739332199097, 0.2094864398241043, 0.9232466220855713, 0.33202219009399414, 0.37195777893066406, 0.10070358961820602, 0.10358034074306488, 0.26425158977508545, -0.23411162197589874, 0.40924984216690063, -0.3693644404411316, -0.0463179312646389, 0.024538787081837654, -0.16673341393470764, -0.22738268971443176, -0.010426290333271027, -0.010993778705596924, 0.30259543657302856, 0.32404959201812744, 0.5999362468719482, -0.11212000250816345, 0.16670005023479462, 0.215691938996315, -0.15810950100421906, -0.24923086166381836, 0.20107315480709076, -0.2923040986061096, 0.31524065136909485, -0.11240848153829575, -0.07410834729671478, -0.01650886982679367, -0.35134220123291016, 0.06714004278182983, 0.0609416738152504, -0.3365170359611511, 0.3048418164253235, 0.13346192240715027, -0.27266669273376465, -0.29698777198791504, 0.2550782561302185, 0.11198711395263672, 0.08252803981304169, -0.048599302768707275, -0.035185206681489944, 0.4006449580192566, 0.35045796632766724, -0.006891845725476742, -0.20158877968788147, 0.18641246855258942, 0.06387515366077423, -0.012894652783870697, 0.030242733657360077, -0.05640551447868347, -0.3015826344490051, -0.2525118589401245, 0.14820465445518494, 0.0033729001879692078, -0.4129619598388672, -0.055318914353847504, 0.06144453212618828, -0.0667886808514595, -0.09413722157478333, 0.15088041126728058, -0.026068218052387238, -0.10794125497341156, 0.3151545822620392, 0.11733834445476532, -0.24304209649562836, 0.05235828459262848, 0.47483235597610474, -0.04039648920297623, 0.031818099319934845, 0.4154282510280609, -0.15281841158866882, -0.10380619764328003, -0.25539109110832214, -0.28417226672172546, -0.10206017643213272, -0.44884029030799866, 0.11220822483301163, -0.2194056510925293, -0.40191107988357544, 0.04423205554485321, 0.2808186709880829, 0.014706198126077652, 0.3184961974620819, -0.011618245393037796, -0.12495805323123932, -0.3769727349281311, -0.10379086434841156, -0.10658213496208191, 0.26717838644981384, 0.27951928973197937, 0.2843479812145233, -0.29586678743362427, 0.2492268979549408, -0.38310953974723816, -0.10714222490787506, -0.1871705949306488, 0.4379130005836487, -0.0391375795006752, -0.36558276414871216, -0.009147514589130878, 0.0018548620864748955, 0.12997101247310638, 0.17754510045051575, -0.6386584639549255, -0.29261907935142517, -0.09099219739437103, 0.08186075091362, -0.007588040083646774, -0.1598217487335205, -0.23928231000900269, -0.01721424236893654, 0.18238495290279388, -0.427402526140213, 0.09215688705444336, 0.15121665596961975, 0.04703199863433838, 0.23718711733818054, -0.1610298752784729, 0.1214555948972702, 0.28816717863082886, -0.18652743101119995, -0.1841491460800171, -0.07622881233692169, 0.08048095554113388, 0.25701695680618286, -0.06946776062250137, 0.020479001104831696, -0.2104034423828125, 0.030471812933683395, 0.08927443623542786, 0.14279493689537048, 0.23692923784255981, -0.024752572178840637, -0.21566128730773926, 0.0005505271255970001, 0.18279926478862762, 0.11411947757005692, -0.07855326682329178, -0.32897576689720154, 0.2802233397960663, 0.2783650755882263, -0.11751837283372879, -0.0102371945977211, 0.1677074134349823, -0.09938855469226837, 0.25285133719444275, 0.20715340971946716, 0.10758769512176514, 0.05261163413524628, 0.26961183547973633, 0.12649410963058472, 0.40504008531570435, -0.34832900762557983, 0.08942112326622009, 0.06441552191972733, 0.10164903104305267, 0.03777839243412018, 0.38121557235717773, -0.1532439887523651, 0.08992154896259308, 0.14838308095932007, 0.18833690881729126, 0.20253999531269073, 0.26289018988609314, 0.21354106068611145, -0.1878642439842224, -0.03412104770541191, -0.24108661711215973, 0.1436368077993393, -0.14272664487361908, 0.4156598448753357, 0.15059244632720947, 0.19285538792610168, -0.11485158652067184, -0.4619392156600952, -0.2462674379348755, 0.23536445200443268, -0.3245418667793274, -0.30920952558517456, -0.04867886006832123, -0.08018464595079422, 0.05592931807041168, 0.19456081092357635, -0.07928960025310516, 0.2038593888282776, 0.3428382873535156, -0.00865977257490158, -0.013750068843364716, -0.08435389399528503, -0.170617014169693, 0.11397916823625565, 0.2748645842075348, -0.08173888921737671, 0.446936696767807, 0.010800972580909729, -0.051569364964962006, -0.2344720959663391, 0.4114176034927368, 0.2861287593841553, 0.19494742155075073, -0.1688622683286667, 0.06395319104194641, 0.20064730942249298, 0.03061327524483204, -0.35468870401382446, 0.3286513686180115, 0.15114429593086243, -0.2930281460285187, 0.3680053651332855, 0.12379556894302368, -0.23215582966804504, 0.2764035165309906, -0.1041957288980484, -0.20146708190441132, -0.049888961017131805, 0.3224371671676636, 0.09970720112323761, -0.03993305563926697, -0.1949632167816162, 0.1111203134059906, -0.1445714384317398, -0.28533875942230225, 0.15934628248214722, -0.023898672312498093, 0.4676608443260193, -0.2756762206554413, 0.11720523238182068, -0.06182504817843437, 0.6220859885215759, -0.02922816202044487, 0.12432920932769775, -0.35013818740844727, 0.10264356434345245, -0.3331668972969055, 0.198409765958786, -0.08381709456443787, 0.15374305844306946, -0.39758387207984924, 0.20426522195339203, -0.11549905687570572, -0.02346649393439293, -0.11804220080375671, -0.007886471226811409, 0.14740730822086334, 0.17893224954605103, -0.3869178295135498, 0.1554817110300064, -0.08625976741313934, -0.006267093122005463, 0.29177382588386536, -0.37667521834373474, 0.22199100255966187, 0.042478568851947784, 0.20691052079200745, 0.010985862463712692, -0.02172362059354782, -0.054806359112262726, 0.33657363057136536, 0.6232408881187439, 0.34702926874160767, 0.14255046844482422, -0.10102763026952744, 0.08155438303947449, -0.3474372923374176, 0.014104174450039864, -0.06088797375559807, -0.20964977145195007, 0.09756873548030853, 0.5258306264877319, -0.22562795877456665, 0.13194617629051208, -0.178046315908432, 0.23669393360614777, 0.3822764754295349, -0.1527290940284729, 0.0618000254034996, 0.04822578281164169, -0.2598402798175812, -0.005199543200433254, -0.12586651742458344, 0.45379629731178284, 0.007182497531175613, 0.32383060455322266, -0.3592911958694458, -0.5417625904083252, 0.43290984630584717, -0.45112577080726624, -0.30252525210380554, 0.12327728420495987, 0.23878571391105652, 0.34956857562065125, -0.06000012159347534, -0.5436676144599915, -0.17249035835266113, 0.2691141963005066, -0.030661150813102722, -0.1570448875427246, 0.20768548548221588, -0.14864324033260345, -0.04107460752129555, -0.1167544275522232, 0.17646947503089905, 0.10152687877416611, -0.2914122939109802, 0.24369880557060242, -0.3701038360595703 ]
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
[ -0.1263233721256256, -0.22070911526679993, -0.10063229501247406, 0.27801111340522766, 0.24547500908374786, -0.20522364974021912, 0.13721121847629547, 0.512859046459198, 0.18828602135181427, -0.06233552098274231, 0.1278189718723297, 0.38913494348526, 0.0037763379514217377, -0.16061760485172272, -0.010743602178990841, 0.024212224408984184, 0.09608031809329987, 0.23505264520645142, -0.1455627828836441, -0.16482312977313995, 0.0818943977355957, 0.2129845768213272, -0.042410045862197876, 0.11652722954750061, -0.39034977555274963, 0.04128977656364441, 0.2123567909002304, -0.230890691280365, 0.06845765560865402, -0.2570657432079315, 0.16785027086734772, -0.09324291348457336, 0.025834325700998306, 0.16631808876991272, -0.00010973814642056823, 0.006797894835472107, -0.015387743711471558, -0.13856983184814453, -0.007274414878338575, 0.04019510746002197, 0.11813420802354813, -0.21172915399074554, -0.10918302088975906, -0.31213220953941345, -0.33201220631599426, -0.037661660462617874, 0.017809096723794937, -0.3556599020957947, 0.3128577172756195, 0.14745266735553741, 0.2607105076313019, 0.22490960359573364, -0.2538115978240967, 0.21113674342632294, 0.1451275497674942, -0.02932754158973694, -0.1741972714662552, -0.04230468347668648, -0.04397884011268616, -0.3344639539718628, -0.37676432728767395, 0.525456964969635, -0.07658185809850693, 0.17014150321483612, 0.21466577053070068, 0.2250560224056244, 0.1726061850786209, 0.035937413573265076, 0.14293654263019562, -0.11396264284849167, 0.011095292866230011, -0.2269170582294464, -0.02711626887321472, -0.3733709454536438, -0.41864001750946045, -0.07852613925933838, 0.137624591588974, 0.18104572594165802, -0.13639286160469055, -0.06371629238128662, -0.43193674087524414, 0.23928946256637573, 0.06524854898452759, -0.012788370251655579, 0.11952775716781616, 0.2951611876487732, 0.043805740773677826, 0.19522324204444885, 0.1294461488723755, -0.06973283737897873, -0.02255835384130478, -0.21451729536056519, -0.058972302824258804, 0.22640188038349152, -0.4359855055809021, 0.07208243757486343, 0.4315193295478821, -0.1404145359992981, -0.05324290320277214, 0.0306372232735157, 0.39193007349967957, 0.07921387255191803, 0.11681866645812988, -0.008491560816764832, -0.10416118800640106, 0.5898669958114624, 0.044322285801172256, 0.17118452489376068, -0.01497051864862442, -0.19827789068222046, -0.3560173213481903, 0.004987429827451706, 0.2880176305770874, -0.2866014540195465, 0.4284125566482544, 0.40236005187034607, -0.0422595851123333, -0.12152935564517975, -0.030101587995886803, 0.09402371197938919, -0.33085376024246216, 0.055669233202934265, 0.05872948467731476, 0.31395789980888367, -0.22545060515403748, 0.1492653489112854, -0.1737677901983261, -0.10335496068000793, -0.38496947288513184, 0.12735126912593842, -0.33784639835357666, -0.03332383185625076, -0.4962179958820343, 0.0015164613723754883, 0.28538399934768677, -0.002860570326447487, 0.28535300493240356, -0.03844557702541351, -0.13722029328346252, -0.301989883184433, 0.10317452996969223, -0.39851242303848267, 0.5702308416366577, -0.1072293221950531, -0.1970168650150299, 0.24102140963077545, 0.13037806749343872, 0.08526398241519928, -0.29768770933151245, -0.1983770728111267, -0.16980770230293274, -0.2090688943862915, 0.2598598897457123, 0.2422453612089157, -0.07942710071802139, 0.1413964033126831, 0.09248155355453491, 0.07767830789089203, 0.3432585597038269, -0.2328546792268753, 0.10284028947353363, -0.2014889419078827, -0.2819805443286896, -0.20435944199562073, 0.16256943345069885, 0.2623099088668823, 0.0667768269777298, -0.13737235963344574, 0.40179744362831116, 0.23718246817588806, 0.14729131758213043, 0.4138064980506897, -0.196533665060997, 0.30949509143829346, -0.4334923028945923, 0.34861892461776733, 0.5064778923988342, -0.32673120498657227, -0.49634435772895813, 0.09944827109575272, -0.1563153862953186, 0.1002877876162529, 0.10244903713464737, 0.1513729840517044, 0.02908817119896412, 0.014419601298868656, 0.12853626906871796, 0.01870240643620491, -0.0005144402384757996, 0.20842039585113525, -0.3169277310371399, -0.021933740004897118, 0.5500645041465759, -0.25246739387512207, 0.028472337871789932, -0.1312367022037506, -0.05504260212182999, 0.021932099014520645, 0.006768524646759033, -0.11218959093093872, 0.12185925245285034, 0.1146051287651062, -0.05687848851084709, -0.15818428993225098, -0.10643843561410904, -0.02469290792942047, -0.2882624864578247, 0.23830124735832214, -0.14981965720653534, -0.09996873140335083, 0.06172499060630798, -0.09704941511154175, -0.1620030403137207, -0.1292746216058731, -0.1809171587228775, -0.27306482195854187, 0.20130175352096558, 0.23052407801151276, 0.3202403485774994, -0.07290694117546082, 0.22520966827869415, 0.3680819571018219, 0.04836329072713852, -0.21607185900211334, -0.4300349950790405, 0.08568014949560165, -0.17773526906967163, -0.21882861852645874, -0.074846550822258, 0.22306931018829346, 0.2090265452861786, -0.0479268915951252, -0.14242854714393616, 0.1482759267091751, -0.22208523750305176, 0.41891855001449585, -0.13702836632728577, 0.2694522738456726, -0.06965650618076324, 0.029306214302778244, 0.016667230054736137, 0.2540055215358734, 0.05882029980421066, -0.08980119228363037, -0.07270593196153641, 0.47423404455184937, 0.18140165507793427, 0.17011570930480957, -0.04807114973664284, -0.2162683755159378, 0.05198418349027634, -0.1783173829317093, -0.12091565877199173, -0.11881639063358307, 0.1364068239927292, -0.0805719643831253, 0.14888811111450195, 0.15232901275157928, -0.10698775947093964, 0.3120076358318329, 0.7600834369659424, 0.19691896438598633, -0.011016584932804108, 0.015740428119897842, -0.05721340328454971, -0.2177915871143341, 0.33385711908340454, -0.058683205395936966, 0.31038856506347656, 0.10473139584064484, 0.12716472148895264, -0.04620931297540665, -0.21964949369430542, -0.1544356644153595, 0.12464860081672668, -0.21436886489391327, -0.19197627902030945, 0.3080342710018158, 0.28208714723587036, -0.1583673059940338, -0.3537829518318176, 0.09401586651802063, -0.01276421919465065, 0.02619149163365364, -0.1580010950565338, 0.21723711490631104, -0.166529580950737, -0.2780844271183014, -0.2864159643650055, 0.13615699112415314, -0.1222032904624939, -0.09216521680355072, -0.04126402735710144, 0.42542994022369385, -0.03665407374501228, 0.19515199959278107, -0.3149715065956116, 0.0895179957151413, 0.2266039401292801, -0.26632726192474365, -0.07168534398078918, -0.1902872771024704, -0.39492014050483704, 0.042231734842061996, -0.09523066878318787, -0.0929960310459137, 0.3783302307128906, -0.047316424548625946, -0.06796006858348846, -0.2881466746330261, -0.17254027724266052, 0.10717801749706268, -0.09261122345924377, -0.15536727011203766, 0.16724826395511627, -0.03133828938007355, -0.10115892440080643, -0.29874566197395325, 0.2937973141670227, -0.30531302094459534, -0.2991430461406708, -0.05198727175593376, 0.09890619665384293, 0.030108986422419548, -0.2803470194339752, -0.4286248981952667, 0.27357742190361023, -0.31724435091018677, 0.07546759396791458, -0.14892880618572235, 0.18622717261314392, 0.33093392848968506, 0.008953814394772053, -0.014561107382178307, -0.06404063105583191, 0.16439421474933624, -0.4674096703529358, -0.14095471799373627, 0.32257843017578125, 0.01876021921634674, -0.21630443632602692, -0.3245909512042999, -0.2757244110107422, 0.34772399067878723, 0.19238810241222382, -0.5803955793380737, -0.21442262828350067, -0.048900216817855835, 0.2480318248271942, -0.05125897377729416, -0.04649518430233002, 0.5745418071746826, 0.01972545124590397, -0.17946316301822662, -0.16727802157402039, -0.33183929324150085, 0.35973066091537476, 0.4192623794078827, 0.24948757886886597, -0.011158477514982224, 0.17781516909599304, 0.20092011988162994, 0.6707470417022705, 0.2019203156232834, -0.2590579092502594, 0.33173102140426636, 0.10723146051168442, 0.07499773055315018, -0.2034967839717865, -0.3130229413509369, 0.18576018512248993, -0.18160304427146912, 0.01973230391740799, 0.27198368310928345, 0.006250595673918724, -0.2801837921142578, 0.017742767930030823, 0.21881836652755737, -0.2253446877002716, -0.328058123588562, 0.2597455084323883, -0.3424280285835266, 0.24094843864440918, 0.036369793117046356, 0.3157997131347656, -0.5820137858390808, -0.24272391200065613, 0.19462794065475464, -0.3359457850456238, 0.27984461188316345, 0.04449816048145294, -0.5836011171340942, 0.12147033214569092, -0.5380215048789978, 0.18194156885147095, 0.31462806463241577, 0.27240774035453796, -0.09202472865581512, -0.2465253472328186, 0.043347518891096115, -0.14713148772716522, 0.5711420178413391, 0.04097308963537216, 0.40857934951782227, 0.16992636024951935, -0.13024693727493286, -0.19947344064712524, -0.22471317648887634, 0.017234940081834793, 0.28013670444488525, 0.3265002369880676, 0.400847464799881, -0.345244437456131, -0.3632737100124359, -0.2657657265663147, 0.007363845594227314, -0.046181585639715195, 0.007106936536729336, -0.12583458423614502, 0.18938523530960083, 0.01648571901023388, 0.17434579133987427, -0.014980330131947994, 0.11830316483974457, 0.016039708629250526, -0.13654674589633942, -0.03913415968418121, -0.21942690014839172, 0.12824007868766785, 0.22643797099590302, 0.4735948443412781, 0.1503245085477829, 0.03246166184544563, 0.033214349299669266, 0.21515773236751556, -0.5671082735061646, 0.2659573554992676, -0.08884961903095245, -0.004131801426410675, 0.07464389503002167, 0.14443765580654144, -0.030300907790660858, 0.3378967344760895, -0.04252535104751587, 0.1457139402627945, -0.12862969934940338, 0.008957155048847198, -0.5539682507514954, 0.4854547381401062, 0.3874596357345581, 0.13246718049049377, -0.044888898730278015, -0.24685737490653992, 0.17346280813217163, 0.07412064075469971, -0.27653080224990845, 0.13225962221622467, -0.43069255352020264, -0.4319950044155121, 0.18579739332199097, 0.2094864398241043, 0.9232466220855713, 0.33202219009399414, 0.37195777893066406, 0.10070358961820602, 0.10358034074306488, 0.26425158977508545, -0.23411162197589874, 0.40924984216690063, -0.3693644404411316, -0.0463179312646389, 0.024538787081837654, -0.16673341393470764, -0.22738268971443176, -0.010426290333271027, -0.010993778705596924, 0.30259543657302856, 0.32404959201812744, 0.5999362468719482, -0.11212000250816345, 0.16670005023479462, 0.215691938996315, -0.15810950100421906, -0.24923086166381836, 0.20107315480709076, -0.2923040986061096, 0.31524065136909485, -0.11240848153829575, -0.07410834729671478, -0.01650886982679367, -0.35134220123291016, 0.06714004278182983, 0.0609416738152504, -0.3365170359611511, 0.3048418164253235, 0.13346192240715027, -0.27266669273376465, -0.29698777198791504, 0.2550782561302185, 0.11198711395263672, 0.08252803981304169, -0.048599302768707275, -0.035185206681489944, 0.4006449580192566, 0.35045796632766724, -0.006891845725476742, -0.20158877968788147, 0.18641246855258942, 0.06387515366077423, -0.012894652783870697, 0.030242733657360077, -0.05640551447868347, -0.3015826344490051, -0.2525118589401245, 0.14820465445518494, 0.0033729001879692078, -0.4129619598388672, -0.055318914353847504, 0.06144453212618828, -0.0667886808514595, -0.09413722157478333, 0.15088041126728058, -0.026068218052387238, -0.10794125497341156, 0.3151545822620392, 0.11733834445476532, -0.24304209649562836, 0.05235828459262848, 0.47483235597610474, -0.04039648920297623, 0.031818099319934845, 0.4154282510280609, -0.15281841158866882, -0.10380619764328003, -0.25539109110832214, -0.28417226672172546, -0.10206017643213272, -0.44884029030799866, 0.11220822483301163, -0.2194056510925293, -0.40191107988357544, 0.04423205554485321, 0.2808186709880829, 0.014706198126077652, 0.3184961974620819, -0.011618245393037796, -0.12495805323123932, -0.3769727349281311, -0.10379086434841156, -0.10658213496208191, 0.26717838644981384, 0.27951928973197937, 0.2843479812145233, -0.29586678743362427, 0.2492268979549408, -0.38310953974723816, -0.10714222490787506, -0.1871705949306488, 0.4379130005836487, -0.0391375795006752, -0.36558276414871216, -0.009147514589130878, 0.0018548620864748955, 0.12997101247310638, 0.17754510045051575, -0.6386584639549255, -0.29261907935142517, -0.09099219739437103, 0.08186075091362, -0.007588040083646774, -0.1598217487335205, -0.23928231000900269, -0.01721424236893654, 0.18238495290279388, -0.427402526140213, 0.09215688705444336, 0.15121665596961975, 0.04703199863433838, 0.23718711733818054, -0.1610298752784729, 0.1214555948972702, 0.28816717863082886, -0.18652743101119995, -0.1841491460800171, -0.07622881233692169, 0.08048095554113388, 0.25701695680618286, -0.06946776062250137, 0.020479001104831696, -0.2104034423828125, 0.030471812933683395, 0.08927443623542786, 0.14279493689537048, 0.23692923784255981, -0.024752572178840637, -0.21566128730773926, 0.0005505271255970001, 0.18279926478862762, 0.11411947757005692, -0.07855326682329178, -0.32897576689720154, 0.2802233397960663, 0.2783650755882263, -0.11751837283372879, -0.0102371945977211, 0.1677074134349823, -0.09938855469226837, 0.25285133719444275, 0.20715340971946716, 0.10758769512176514, 0.05261163413524628, 0.26961183547973633, 0.12649410963058472, 0.40504008531570435, -0.34832900762557983, 0.08942112326622009, 0.06441552191972733, 0.10164903104305267, 0.03777839243412018, 0.38121557235717773, -0.1532439887523651, 0.08992154896259308, 0.14838308095932007, 0.18833690881729126, 0.20253999531269073, 0.26289018988609314, 0.21354106068611145, -0.1878642439842224, -0.03412104770541191, -0.24108661711215973, 0.1436368077993393, -0.14272664487361908, 0.4156598448753357, 0.15059244632720947, 0.19285538792610168, -0.11485158652067184, -0.4619392156600952, -0.2462674379348755, 0.23536445200443268, -0.3245418667793274, -0.30920952558517456, -0.04867886006832123, -0.08018464595079422, 0.05592931807041168, 0.19456081092357635, -0.07928960025310516, 0.2038593888282776, 0.3428382873535156, -0.00865977257490158, -0.013750068843364716, -0.08435389399528503, -0.170617014169693, 0.11397916823625565, 0.2748645842075348, -0.08173888921737671, 0.446936696767807, 0.010800972580909729, -0.051569364964962006, -0.2344720959663391, 0.4114176034927368, 0.2861287593841553, 0.19494742155075073, -0.1688622683286667, 0.06395319104194641, 0.20064730942249298, 0.03061327524483204, -0.35468870401382446, 0.3286513686180115, 0.15114429593086243, -0.2930281460285187, 0.3680053651332855, 0.12379556894302368, -0.23215582966804504, 0.2764035165309906, -0.1041957288980484, -0.20146708190441132, -0.049888961017131805, 0.3224371671676636, 0.09970720112323761, -0.03993305563926697, -0.1949632167816162, 0.1111203134059906, -0.1445714384317398, -0.28533875942230225, 0.15934628248214722, -0.023898672312498093, 0.4676608443260193, -0.2756762206554413, 0.11720523238182068, -0.06182504817843437, 0.6220859885215759, -0.02922816202044487, 0.12432920932769775, -0.35013818740844727, 0.10264356434345245, -0.3331668972969055, 0.198409765958786, -0.08381709456443787, 0.15374305844306946, -0.39758387207984924, 0.20426522195339203, -0.11549905687570572, -0.02346649393439293, -0.11804220080375671, -0.007886471226811409, 0.14740730822086334, 0.17893224954605103, -0.3869178295135498, 0.1554817110300064, -0.08625976741313934, -0.006267093122005463, 0.29177382588386536, -0.37667521834373474, 0.22199100255966187, 0.042478568851947784, 0.20691052079200745, 0.010985862463712692, -0.02172362059354782, -0.054806359112262726, 0.33657363057136536, 0.6232408881187439, 0.34702926874160767, 0.14255046844482422, -0.10102763026952744, 0.08155438303947449, -0.3474372923374176, 0.014104174450039864, -0.06088797375559807, -0.20964977145195007, 0.09756873548030853, 0.5258306264877319, -0.22562795877456665, 0.13194617629051208, -0.178046315908432, 0.23669393360614777, 0.3822764754295349, -0.1527290940284729, 0.0618000254034996, 0.04822578281164169, -0.2598402798175812, -0.005199543200433254, -0.12586651742458344, 0.45379629731178284, 0.007182497531175613, 0.32383060455322266, -0.3592911958694458, -0.5417625904083252, 0.43290984630584717, -0.45112577080726624, -0.30252525210380554, 0.12327728420495987, 0.23878571391105652, 0.34956857562065125, -0.06000012159347534, -0.5436676144599915, -0.17249035835266113, 0.2691141963005066, -0.030661150813102722, -0.1570448875427246, 0.20768548548221588, -0.14864324033260345, -0.04107460752129555, -0.1167544275522232, 0.17646947503089905, 0.10152687877416611, -0.2914122939109802, 0.24369880557060242, -0.3701038360595703 ]
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).
[ -0.1263233721256256, -0.22070911526679993, -0.10063229501247406, 0.27801111340522766, 0.24547500908374786, -0.20522364974021912, 0.13721121847629547, 0.512859046459198, 0.18828602135181427, -0.06233552098274231, 0.1278189718723297, 0.38913494348526, 0.0037763379514217377, -0.16061760485172272, -0.010743602178990841, 0.024212224408984184, 0.09608031809329987, 0.23505264520645142, -0.1455627828836441, -0.16482312977313995, 0.0818943977355957, 0.2129845768213272, -0.042410045862197876, 0.11652722954750061, -0.39034977555274963, 0.04128977656364441, 0.2123567909002304, -0.230890691280365, 0.06845765560865402, -0.2570657432079315, 0.16785027086734772, -0.09324291348457336, 0.025834325700998306, 0.16631808876991272, -0.00010973814642056823, 0.006797894835472107, -0.015387743711471558, -0.13856983184814453, -0.007274414878338575, 0.04019510746002197, 0.11813420802354813, -0.21172915399074554, -0.10918302088975906, -0.31213220953941345, -0.33201220631599426, -0.037661660462617874, 0.017809096723794937, -0.3556599020957947, 0.3128577172756195, 0.14745266735553741, 0.2607105076313019, 0.22490960359573364, -0.2538115978240967, 0.21113674342632294, 0.1451275497674942, -0.02932754158973694, -0.1741972714662552, -0.04230468347668648, -0.04397884011268616, -0.3344639539718628, -0.37676432728767395, 0.525456964969635, -0.07658185809850693, 0.17014150321483612, 0.21466577053070068, 0.2250560224056244, 0.1726061850786209, 0.035937413573265076, 0.14293654263019562, -0.11396264284849167, 0.011095292866230011, -0.2269170582294464, -0.02711626887321472, -0.3733709454536438, -0.41864001750946045, -0.07852613925933838, 0.137624591588974, 0.18104572594165802, -0.13639286160469055, -0.06371629238128662, -0.43193674087524414, 0.23928946256637573, 0.06524854898452759, -0.012788370251655579, 0.11952775716781616, 0.2951611876487732, 0.043805740773677826, 0.19522324204444885, 0.1294461488723755, -0.06973283737897873, -0.02255835384130478, -0.21451729536056519, -0.058972302824258804, 0.22640188038349152, -0.4359855055809021, 0.07208243757486343, 0.4315193295478821, -0.1404145359992981, -0.05324290320277214, 0.0306372232735157, 0.39193007349967957, 0.07921387255191803, 0.11681866645812988, -0.008491560816764832, -0.10416118800640106, 0.5898669958114624, 0.044322285801172256, 0.17118452489376068, -0.01497051864862442, -0.19827789068222046, -0.3560173213481903, 0.004987429827451706, 0.2880176305770874, -0.2866014540195465, 0.4284125566482544, 0.40236005187034607, -0.0422595851123333, -0.12152935564517975, -0.030101587995886803, 0.09402371197938919, -0.33085376024246216, 0.055669233202934265, 0.05872948467731476, 0.31395789980888367, -0.22545060515403748, 0.1492653489112854, -0.1737677901983261, -0.10335496068000793, -0.38496947288513184, 0.12735126912593842, -0.33784639835357666, -0.03332383185625076, -0.4962179958820343, 0.0015164613723754883, 0.28538399934768677, -0.002860570326447487, 0.28535300493240356, -0.03844557702541351, -0.13722029328346252, -0.301989883184433, 0.10317452996969223, -0.39851242303848267, 0.5702308416366577, -0.1072293221950531, -0.1970168650150299, 0.24102140963077545, 0.13037806749343872, 0.08526398241519928, -0.29768770933151245, -0.1983770728111267, -0.16980770230293274, -0.2090688943862915, 0.2598598897457123, 0.2422453612089157, -0.07942710071802139, 0.1413964033126831, 0.09248155355453491, 0.07767830789089203, 0.3432585597038269, -0.2328546792268753, 0.10284028947353363, -0.2014889419078827, -0.2819805443286896, -0.20435944199562073, 0.16256943345069885, 0.2623099088668823, 0.0667768269777298, -0.13737235963344574, 0.40179744362831116, 0.23718246817588806, 0.14729131758213043, 0.4138064980506897, -0.196533665060997, 0.30949509143829346, -0.4334923028945923, 0.34861892461776733, 0.5064778923988342, -0.32673120498657227, -0.49634435772895813, 0.09944827109575272, -0.1563153862953186, 0.1002877876162529, 0.10244903713464737, 0.1513729840517044, 0.02908817119896412, 0.014419601298868656, 0.12853626906871796, 0.01870240643620491, -0.0005144402384757996, 0.20842039585113525, -0.3169277310371399, -0.021933740004897118, 0.5500645041465759, -0.25246739387512207, 0.028472337871789932, -0.1312367022037506, -0.05504260212182999, 0.021932099014520645, 0.006768524646759033, -0.11218959093093872, 0.12185925245285034, 0.1146051287651062, -0.05687848851084709, -0.15818428993225098, -0.10643843561410904, -0.02469290792942047, -0.2882624864578247, 0.23830124735832214, -0.14981965720653534, -0.09996873140335083, 0.06172499060630798, -0.09704941511154175, -0.1620030403137207, -0.1292746216058731, -0.1809171587228775, -0.27306482195854187, 0.20130175352096558, 0.23052407801151276, 0.3202403485774994, -0.07290694117546082, 0.22520966827869415, 0.3680819571018219, 0.04836329072713852, -0.21607185900211334, -0.4300349950790405, 0.08568014949560165, -0.17773526906967163, -0.21882861852645874, -0.074846550822258, 0.22306931018829346, 0.2090265452861786, -0.0479268915951252, -0.14242854714393616, 0.1482759267091751, -0.22208523750305176, 0.41891855001449585, -0.13702836632728577, 0.2694522738456726, -0.06965650618076324, 0.029306214302778244, 0.016667230054736137, 0.2540055215358734, 0.05882029980421066, -0.08980119228363037, -0.07270593196153641, 0.47423404455184937, 0.18140165507793427, 0.17011570930480957, -0.04807114973664284, -0.2162683755159378, 0.05198418349027634, -0.1783173829317093, -0.12091565877199173, -0.11881639063358307, 0.1364068239927292, -0.0805719643831253, 0.14888811111450195, 0.15232901275157928, -0.10698775947093964, 0.3120076358318329, 0.7600834369659424, 0.19691896438598633, -0.011016584932804108, 0.015740428119897842, -0.05721340328454971, -0.2177915871143341, 0.33385711908340454, -0.058683205395936966, 0.31038856506347656, 0.10473139584064484, 0.12716472148895264, -0.04620931297540665, -0.21964949369430542, -0.1544356644153595, 0.12464860081672668, -0.21436886489391327, -0.19197627902030945, 0.3080342710018158, 0.28208714723587036, -0.1583673059940338, -0.3537829518318176, 0.09401586651802063, -0.01276421919465065, 0.02619149163365364, -0.1580010950565338, 0.21723711490631104, -0.166529580950737, -0.2780844271183014, -0.2864159643650055, 0.13615699112415314, -0.1222032904624939, -0.09216521680355072, -0.04126402735710144, 0.42542994022369385, -0.03665407374501228, 0.19515199959278107, -0.3149715065956116, 0.0895179957151413, 0.2266039401292801, -0.26632726192474365, -0.07168534398078918, -0.1902872771024704, -0.39492014050483704, 0.042231734842061996, -0.09523066878318787, -0.0929960310459137, 0.3783302307128906, -0.047316424548625946, -0.06796006858348846, -0.2881466746330261, -0.17254027724266052, 0.10717801749706268, -0.09261122345924377, -0.15536727011203766, 0.16724826395511627, -0.03133828938007355, -0.10115892440080643, -0.29874566197395325, 0.2937973141670227, -0.30531302094459534, -0.2991430461406708, -0.05198727175593376, 0.09890619665384293, 0.030108986422419548, -0.2803470194339752, -0.4286248981952667, 0.27357742190361023, -0.31724435091018677, 0.07546759396791458, -0.14892880618572235, 0.18622717261314392, 0.33093392848968506, 0.008953814394772053, -0.014561107382178307, -0.06404063105583191, 0.16439421474933624, -0.4674096703529358, -0.14095471799373627, 0.32257843017578125, 0.01876021921634674, -0.21630443632602692, -0.3245909512042999, -0.2757244110107422, 0.34772399067878723, 0.19238810241222382, -0.5803955793380737, -0.21442262828350067, -0.048900216817855835, 0.2480318248271942, -0.05125897377729416, -0.04649518430233002, 0.5745418071746826, 0.01972545124590397, -0.17946316301822662, -0.16727802157402039, -0.33183929324150085, 0.35973066091537476, 0.4192623794078827, 0.24948757886886597, -0.011158477514982224, 0.17781516909599304, 0.20092011988162994, 0.6707470417022705, 0.2019203156232834, -0.2590579092502594, 0.33173102140426636, 0.10723146051168442, 0.07499773055315018, -0.2034967839717865, -0.3130229413509369, 0.18576018512248993, -0.18160304427146912, 0.01973230391740799, 0.27198368310928345, 0.006250595673918724, -0.2801837921142578, 0.017742767930030823, 0.21881836652755737, -0.2253446877002716, -0.328058123588562, 0.2597455084323883, -0.3424280285835266, 0.24094843864440918, 0.036369793117046356, 0.3157997131347656, -0.5820137858390808, -0.24272391200065613, 0.19462794065475464, -0.3359457850456238, 0.27984461188316345, 0.04449816048145294, -0.5836011171340942, 0.12147033214569092, -0.5380215048789978, 0.18194156885147095, 0.31462806463241577, 0.27240774035453796, -0.09202472865581512, -0.2465253472328186, 0.043347518891096115, -0.14713148772716522, 0.5711420178413391, 0.04097308963537216, 0.40857934951782227, 0.16992636024951935, -0.13024693727493286, -0.19947344064712524, -0.22471317648887634, 0.017234940081834793, 0.28013670444488525, 0.3265002369880676, 0.400847464799881, -0.345244437456131, -0.3632737100124359, -0.2657657265663147, 0.007363845594227314, -0.046181585639715195, 0.007106936536729336, -0.12583458423614502, 0.18938523530960083, 0.01648571901023388, 0.17434579133987427, -0.014980330131947994, 0.11830316483974457, 0.016039708629250526, -0.13654674589633942, -0.03913415968418121, -0.21942690014839172, 0.12824007868766785, 0.22643797099590302, 0.4735948443412781, 0.1503245085477829, 0.03246166184544563, 0.033214349299669266, 0.21515773236751556, -0.5671082735061646, 0.2659573554992676, -0.08884961903095245, -0.004131801426410675, 0.07464389503002167, 0.14443765580654144, -0.030300907790660858, 0.3378967344760895, -0.04252535104751587, 0.1457139402627945, -0.12862969934940338, 0.008957155048847198, -0.5539682507514954, 0.4854547381401062, 0.3874596357345581, 0.13246718049049377, -0.044888898730278015, -0.24685737490653992, 0.17346280813217163, 0.07412064075469971, -0.27653080224990845, 0.13225962221622467, -0.43069255352020264, -0.4319950044155121, 0.18579739332199097, 0.2094864398241043, 0.9232466220855713, 0.33202219009399414, 0.37195777893066406, 0.10070358961820602, 0.10358034074306488, 0.26425158977508545, -0.23411162197589874, 0.40924984216690063, -0.3693644404411316, -0.0463179312646389, 0.024538787081837654, -0.16673341393470764, -0.22738268971443176, -0.010426290333271027, -0.010993778705596924, 0.30259543657302856, 0.32404959201812744, 0.5999362468719482, -0.11212000250816345, 0.16670005023479462, 0.215691938996315, -0.15810950100421906, -0.24923086166381836, 0.20107315480709076, -0.2923040986061096, 0.31524065136909485, -0.11240848153829575, -0.07410834729671478, -0.01650886982679367, -0.35134220123291016, 0.06714004278182983, 0.0609416738152504, -0.3365170359611511, 0.3048418164253235, 0.13346192240715027, -0.27266669273376465, -0.29698777198791504, 0.2550782561302185, 0.11198711395263672, 0.08252803981304169, -0.048599302768707275, -0.035185206681489944, 0.4006449580192566, 0.35045796632766724, -0.006891845725476742, -0.20158877968788147, 0.18641246855258942, 0.06387515366077423, -0.012894652783870697, 0.030242733657360077, -0.05640551447868347, -0.3015826344490051, -0.2525118589401245, 0.14820465445518494, 0.0033729001879692078, -0.4129619598388672, -0.055318914353847504, 0.06144453212618828, -0.0667886808514595, -0.09413722157478333, 0.15088041126728058, -0.026068218052387238, -0.10794125497341156, 0.3151545822620392, 0.11733834445476532, -0.24304209649562836, 0.05235828459262848, 0.47483235597610474, -0.04039648920297623, 0.031818099319934845, 0.4154282510280609, -0.15281841158866882, -0.10380619764328003, -0.25539109110832214, -0.28417226672172546, -0.10206017643213272, -0.44884029030799866, 0.11220822483301163, -0.2194056510925293, -0.40191107988357544, 0.04423205554485321, 0.2808186709880829, 0.014706198126077652, 0.3184961974620819, -0.011618245393037796, -0.12495805323123932, -0.3769727349281311, -0.10379086434841156, -0.10658213496208191, 0.26717838644981384, 0.27951928973197937, 0.2843479812145233, -0.29586678743362427, 0.2492268979549408, -0.38310953974723816, -0.10714222490787506, -0.1871705949306488, 0.4379130005836487, -0.0391375795006752, -0.36558276414871216, -0.009147514589130878, 0.0018548620864748955, 0.12997101247310638, 0.17754510045051575, -0.6386584639549255, -0.29261907935142517, -0.09099219739437103, 0.08186075091362, -0.007588040083646774, -0.1598217487335205, -0.23928231000900269, -0.01721424236893654, 0.18238495290279388, -0.427402526140213, 0.09215688705444336, 0.15121665596961975, 0.04703199863433838, 0.23718711733818054, -0.1610298752784729, 0.1214555948972702, 0.28816717863082886, -0.18652743101119995, -0.1841491460800171, -0.07622881233692169, 0.08048095554113388, 0.25701695680618286, -0.06946776062250137, 0.020479001104831696, -0.2104034423828125, 0.030471812933683395, 0.08927443623542786, 0.14279493689537048, 0.23692923784255981, -0.024752572178840637, -0.21566128730773926, 0.0005505271255970001, 0.18279926478862762, 0.11411947757005692, -0.07855326682329178, -0.32897576689720154, 0.2802233397960663, 0.2783650755882263, -0.11751837283372879, -0.0102371945977211, 0.1677074134349823, -0.09938855469226837, 0.25285133719444275, 0.20715340971946716, 0.10758769512176514, 0.05261163413524628, 0.26961183547973633, 0.12649410963058472, 0.40504008531570435, -0.34832900762557983, 0.08942112326622009, 0.06441552191972733, 0.10164903104305267, 0.03777839243412018, 0.38121557235717773, -0.1532439887523651, 0.08992154896259308, 0.14838308095932007, 0.18833690881729126, 0.20253999531269073, 0.26289018988609314, 0.21354106068611145, -0.1878642439842224, -0.03412104770541191, -0.24108661711215973, 0.1436368077993393, -0.14272664487361908, 0.4156598448753357, 0.15059244632720947, 0.19285538792610168, -0.11485158652067184, -0.4619392156600952, -0.2462674379348755, 0.23536445200443268, -0.3245418667793274, -0.30920952558517456, -0.04867886006832123, -0.08018464595079422, 0.05592931807041168, 0.19456081092357635, -0.07928960025310516, 0.2038593888282776, 0.3428382873535156, -0.00865977257490158, -0.013750068843364716, -0.08435389399528503, -0.170617014169693, 0.11397916823625565, 0.2748645842075348, -0.08173888921737671, 0.446936696767807, 0.010800972580909729, -0.051569364964962006, -0.2344720959663391, 0.4114176034927368, 0.2861287593841553, 0.19494742155075073, -0.1688622683286667, 0.06395319104194641, 0.20064730942249298, 0.03061327524483204, -0.35468870401382446, 0.3286513686180115, 0.15114429593086243, -0.2930281460285187, 0.3680053651332855, 0.12379556894302368, -0.23215582966804504, 0.2764035165309906, -0.1041957288980484, -0.20146708190441132, -0.049888961017131805, 0.3224371671676636, 0.09970720112323761, -0.03993305563926697, -0.1949632167816162, 0.1111203134059906, -0.1445714384317398, -0.28533875942230225, 0.15934628248214722, -0.023898672312498093, 0.4676608443260193, -0.2756762206554413, 0.11720523238182068, -0.06182504817843437, 0.6220859885215759, -0.02922816202044487, 0.12432920932769775, -0.35013818740844727, 0.10264356434345245, -0.3331668972969055, 0.198409765958786, -0.08381709456443787, 0.15374305844306946, -0.39758387207984924, 0.20426522195339203, -0.11549905687570572, -0.02346649393439293, -0.11804220080375671, -0.007886471226811409, 0.14740730822086334, 0.17893224954605103, -0.3869178295135498, 0.1554817110300064, -0.08625976741313934, -0.006267093122005463, 0.29177382588386536, -0.37667521834373474, 0.22199100255966187, 0.042478568851947784, 0.20691052079200745, 0.010985862463712692, -0.02172362059354782, -0.054806359112262726, 0.33657363057136536, 0.6232408881187439, 0.34702926874160767, 0.14255046844482422, -0.10102763026952744, 0.08155438303947449, -0.3474372923374176, 0.014104174450039864, -0.06088797375559807, -0.20964977145195007, 0.09756873548030853, 0.5258306264877319, -0.22562795877456665, 0.13194617629051208, -0.178046315908432, 0.23669393360614777, 0.3822764754295349, -0.1527290940284729, 0.0618000254034996, 0.04822578281164169, -0.2598402798175812, -0.005199543200433254, -0.12586651742458344, 0.45379629731178284, 0.007182497531175613, 0.32383060455322266, -0.3592911958694458, -0.5417625904083252, 0.43290984630584717, -0.45112577080726624, -0.30252525210380554, 0.12327728420495987, 0.23878571391105652, 0.34956857562065125, -0.06000012159347534, -0.5436676144599915, -0.17249035835266113, 0.2691141963005066, -0.030661150813102722, -0.1570448875427246, 0.20768548548221588, -0.14864324033260345, -0.04107460752129555, -0.1167544275522232, 0.17646947503089905, 0.10152687877416611, -0.2914122939109802, 0.24369880557060242, -0.3701038360595703 ]