SPB Git

spb/wp41_uqo

Public
51.5 KB · 210 lines
1{"event": "train", "step": 20, "loss": -6.120069980621338, "grad_norm": 0.36424320936203003, "lr": 3.149999999999999e-05, "ex_per_s": 90.88979480005007, "data_frac": 0.8674609737622887, "mem_mb": 91.210752, "nll": -6.120069980621338, "elapsed_s": 14.1}2{"event": "train", "step": 40, "loss": -5.888597011566162, "grad_norm": 0.9745411276817322, "lr": 6.149999999999999e-05, "ex_per_s": 161.43814206354955, "data_frac": 0.7721614985494814, "mem_mb": 91.210752, "nll": -5.888597011566162, "elapsed_s": 15.9}3{"event": "train", "step": 60, "loss": -5.980766296386719, "grad_norm": 0.21469330787658691, "lr": 9.149999999999999e-05, "ex_per_s": 218.26731409124181, "data_frac": 0.6971888853134675, "mem_mb": 91.210752, "nll": -5.980766296386719, "elapsed_s": 17.6}4{"event": "train", "step": 80, "loss": -6.080490589141846, "grad_norm": 0.3367573618888855, "lr": 0.0001215, "ex_per_s": 265.6979073572453, "data_frac": 0.6370256608229468, "mem_mb": 91.210752, "nll": -6.080490589141846, "elapsed_s": 19.3}5{"event": "train", "step": 100, "loss": -5.950717449188232, "grad_norm": 0.15280283987522125, "lr": 0.0001515, "ex_per_s": 305.7616730344553, "data_frac": 0.5869905432113688, "mem_mb": 91.210752, "nll": -5.950717449188232, "elapsed_s": 21.0}6{"event": "train", "step": 120, "loss": -5.911173343658447, "grad_norm": 0.8404181003570557, "lr": 0.00018149999999999997, "ex_per_s": 339.9249397807238, "data_frac": 0.5443237004511241, "mem_mb": 91.210752, "nll": -5.911173343658447, "elapsed_s": 22.6}7{"event": "train", "step": 140, "loss": -5.831188678741455, "grad_norm": 0.7273851633071899, "lr": 0.00021149999999999996, "ex_per_s": 369.47222541168134, "data_frac": 0.5075956301188668, "mem_mb": 91.210752, "nll": -5.831188678741455, "elapsed_s": 24.3}8{"event": "train", "step": 160, "loss": -5.737343788146973, "grad_norm": 0.5886886119842529, "lr": 0.0002415, "ex_per_s": 395.1682175826207, "data_frac": 0.47548942416449325, "mem_mb": 91.210752, "nll": -5.737343788146973, "elapsed_s": 26.0}9{"event": "train", "step": 180, "loss": -5.976075649261475, "grad_norm": 1.0061029195785522, "lr": 0.0002715, "ex_per_s": 417.909998442299, "data_frac": 0.44738694148398134, "mem_mb": 91.210752, "nll": -5.976075649261475, "elapsed_s": 27.6}10{"event": "train", "step": 200, "loss": -5.972075939178467, "grad_norm": 0.4434223175048828, "lr": 0.0003, "ex_per_s": 438.06913988440124, "data_frac": 0.42246531286040767, "mem_mb": 91.210752, "nll": -5.972075939178467, "elapsed_s": 29.3}11{"event": "train", "step": 220, "loss": -6.031540393829346, "grad_norm": 1.1362534761428833, "lr": 0.00029998154617398326, "ex_per_s": 456.10932548697417, "data_frac": 0.40026557707122884, "mem_mb": 91.210752, "nll": -6.031540393829346, "elapsed_s": 30.9}12{"event": "train", "step": 240, "loss": -5.9152140617370605, "grad_norm": 0.20935063064098358, "lr": 0.000299926189741025, "ex_per_s": 472.278454359436, "data_frac": 0.380285128458042, "mem_mb": 91.210752, "nll": -5.9152140617370605, "elapsed_s": 32.6}13{"event": "train", "step": 260, "loss": -5.888271808624268, "grad_norm": 0.14142268896102905, "lr": 0.0002998339458350211, "ex_per_s": 486.91628960358804, "data_frac": 0.36226111069197514, "mem_mb": 91.210752, "nll": -5.888271808624268, "elapsed_s": 34.2}14{"event": "train", "step": 280, "loss": -5.94512939453125, "grad_norm": 0.11715613305568695, "lr": 0.0002997048396745345, "ex_per_s": 500.1984316679085, "data_frac": 0.34588823376657307, "mem_mb": 91.210752, "nll": -5.94512939453125, "elapsed_s": 35.9}15{"event": "train", "step": 300, "loss": -5.968127250671387, "grad_norm": 0.23904262483119965, "lr": 0.0002995389065559004, "ex_per_s": 512.2218539020234, "data_frac": 0.3309215335753968, "mem_mb": 91.210752, "nll": -5.968127250671387, "elapsed_s": 37.5}16{"event": "train", "step": 320, "loss": -5.967316627502441, "grad_norm": 0.3344629108905792, "lr": 0.0002993361918435765, "ex_per_s": 523.3143427468451, "data_frac": 0.3172590189466456, "mem_mb": 91.210752, "nll": -5.967316627502441, "elapsed_s": 39.2}17{"event": "train", "step": 340, "loss": -5.834796905517578, "grad_norm": 0.3146268427371979, "lr": 0.00029909675095774114, "ex_per_s": 533.516705028818, "data_frac": 0.3047156948901245, "mem_mb": 91.210752, "nll": -5.834796905517578, "elapsed_s": 40.8}18{"event": "train", "step": 360, "loss": -5.996510982513428, "grad_norm": 0.24481038749217987, "lr": 0.00029882064935914164, "ex_per_s": 542.9132255331718, "data_frac": 0.2931374127755614, "mem_mb": 91.210752, "nll": -5.996510982513428, "elapsed_s": 42.5}19{"event": "train", "step": 380, "loss": -6.005453586578369, "grad_norm": 1.1869938373565674, "lr": 0.00029850796253119824, "ex_per_s": 551.580573037889, "data_frac": 0.2824092462611414, "mem_mb": 91.210752, "nll": -6.005453586578369, "elapsed_s": 44.1}20{"event": "train", "step": 400, "loss": -5.994329452514648, "grad_norm": 1.48933744430542, "lr": 0.0002981587759593675, "ex_per_s": 559.6394425791304, "data_frac": 0.2724695800485245, "mem_mb": 91.210752, "nll": -5.994329452514648, "elapsed_s": 45.8}21{"event": "train", "step": 420, "loss": -5.964642524719238, "grad_norm": 0.5800729393959045, "lr": 0.0002977731851077716, "ex_per_s": 567.1350076288115, "data_frac": 0.26322517473185675, "mem_mb": 91.210752, "nll": -5.964642524719238, "elapsed_s": 47.4}22{"event": "train", "step": 440, "loss": -5.938342571258545, "grad_norm": 0.763841450214386, "lr": 0.0002973512953930992, "ex_per_s": 574.1406745267701, "data_frac": 0.25460748331162264, "mem_mb": 91.210752, "nll": -5.938342571258545, "elapsed_s": 49.1}23{"event": "train", "step": 460, "loss": -6.101180076599121, "grad_norm": 0.5908114910125732, "lr": 0.0002968932221557859, "ex_per_s": 580.6661900463192, "data_frac": 0.24653862592340176, "mem_mb": 91.210752, "nll": -6.101180076599121, "elapsed_s": 50.7}24{"event": "train", "step": 480, "loss": -5.877658843994141, "grad_norm": 0.26609519124031067, "lr": 0.0002963990906284808, "ex_per_s": 586.7861042254364, "data_frac": 0.23898800053127658, "mem_mb": 91.210752, "nll": -5.877658843994141, "elapsed_s": 52.4}25{"event": "train", "step": 500, "loss": -5.925559997558594, "grad_norm": 0.5425939559936523, "lr": 0.00029586903590180956, "ex_per_s": 592.5265151716153, "data_frac": 0.2318928720767281, "mem_mb": 91.210752, "nll": -5.925559997558594, "elapsed_s": 54.0}26{"event": "val", "step": 500, "val_nll": -6.4398352735964535, "best": -6.4398352735964535, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 118.2}27{"event": "train", "step": 520, "loss": -6.04081392288208, "grad_norm": 0.2562132775783539, "lr": 0.00029530320288744205, "ex_per_s": 277.5509028453889, "data_frac": 0.10455562054576124, "mem_mb": 91.210752, "nll": -6.04081392288208, "elapsed_s": 119.9}28{"event": "train", "step": 540, "loss": -5.878698348999023, "grad_norm": 0.5782443284988403, "lr": 0.0002947017462784745, "ex_per_s": 284.2609143770652, "data_frac": 0.10322523219973158, "mem_mb": 91.210752, "nll": -5.878698348999023, "elapsed_s": 121.6}29{"event": "train", "step": 560, "loss": -5.972837924957275, "grad_norm": 0.088307224214077, "lr": 0.00029406483050713824, "ex_per_s": 290.78989996126455, "data_frac": 0.1020367930823151, "mem_mb": 91.210752, "nll": -5.972837924957275, "elapsed_s": 123.3}30{"event": "train", "step": 580, "loss": -5.924700736999512, "grad_norm": 0.0924358144402504, "lr": 0.0002933926296998457, "ex_per_s": 297.14711231604434, "data_frac": 0.10076924736187466, "mem_mb": 91.210752, "nll": -5.924700736999512, "elapsed_s": 125.0}31{"event": "train", "step": 600, "loss": -5.972741603851318, "grad_norm": 0.7153385877609253, "lr": 0.0002926853276295856, "ex_per_s": 303.33633923846384, "data_frac": 0.09953474984470286, "mem_mb": 91.210752, "nll": -5.972741603851318, "elapsed_s": 126.6}32{"event": "train", "step": 620, "loss": -5.9525837898254395, "grad_norm": 0.37996459007263184, "lr": 0.00029194311766568174, "ex_per_s": 309.35531331161053, "data_frac": 0.09833392843142988, "mem_mb": 91.210752, "nll": -5.9525837898254395, "elapsed_s": 128.3}33{"event": "train", "step": 640, "loss": -6.065352439880371, "grad_norm": 0.47623395919799805, "lr": 0.0002911662027209272, "ex_per_s": 315.23556430311345, "data_frac": 0.09716557582342798, "mem_mb": 91.210752, "nll": -6.065352439880371, "elapsed_s": 130.0}34{"event": "train", "step": 660, "loss": -6.008050918579102, "grad_norm": 1.029016137123108, "lr": 0.00029035479519611045, "ex_per_s": 320.9701197417821, "data_frac": 0.09602656496887427, "mem_mb": 91.210752, "nll": -6.008050918579102, "elapsed_s": 131.6}35{"event": "train", "step": 680, "loss": -6.00679874420166, "grad_norm": 0.11741286516189575, "lr": 0.0002895091169219469, "ex_per_s": 326.559578270249, "data_frac": 0.09491671794754802, "mem_mb": 91.210752, "nll": -6.00679874420166, "elapsed_s": 133.3}36{"event": "train", "step": 700, "loss": -5.9874138832092285, "grad_norm": 0.6614100337028503, "lr": 0.00028862939909843273, "ex_per_s": 332.0141305477142, "data_frac": 0.09383447319895978, "mem_mb": 91.210752, "nll": -5.9874138832092285, "elapsed_s": 135.0}37{"event": "train", "step": 720, "loss": -6.0635786056518555, "grad_norm": 0.22732430696487427, "lr": 0.0002877158822316369, "ex_per_s": 337.3219899035666, "data_frac": 0.09277899444902617, "mem_mb": 91.210752, "nll": -6.0635786056518555, "elapsed_s": 136.6}38{"event": "train", "step": 740, "loss": -6.033602714538574, "grad_norm": 0.2537577748298645, "lr": 0.00028676881606794945, "ex_per_s": 342.49126943310887, "data_frac": 0.09174855168721947, "mem_mb": 91.210752, "nll": -6.033602714538574, "elapsed_s": 138.3}39{"event": "train", "step": 760, "loss": -5.959403991699219, "grad_norm": 0.8005170226097107, "lr": 0.000285788459525803, "ex_per_s": 347.55033307351476, "data_frac": 0.09073881757434611, "mem_mb": 91.210752, "nll": -5.959403991699219, "elapsed_s": 140.0}40{"event": "train", "step": 780, "loss": -5.894862651824951, "grad_norm": 0.2809053957462311, "lr": 0.00028477508062488754, "ex_per_s": 352.49396164123357, "data_frac": 0.08975409850830836, "mem_mb": 91.210752, "nll": -5.894862651824951, "elapsed_s": 141.7}41{"event": "train", "step": 800, "loss": -5.815484523773193, "grad_norm": 0.27126285433769226, "lr": 0.000283728956412876, "ex_per_s": 357.3202905936088, "data_frac": 0.08879475003161331, "mem_mb": 91.210752, "nll": -5.815484523773193, "elapsed_s": 143.3}42{"event": "train", "step": 820, "loss": -5.93386173248291, "grad_norm": 0.25691142678260803, "lr": 0.00028265037288968236, "ex_per_s": 362.0354327597382, "data_frac": 0.08785649319410391, "mem_mb": 91.210752, "nll": -5.93386173248291, "elapsed_s": 145.0}43{"event": "train", "step": 840, "loss": -5.807285308837891, "grad_norm": 0.545007050037384, "lr": 0.0002815396249292725, "ex_per_s": 366.6403792625416, "data_frac": 0.08694125145760122, "mem_mb": 91.210752, "nll": -5.807285308837891, "elapsed_s": 146.7}44{"event": "train", "step": 860, "loss": -5.912852764129639, "grad_norm": 1.5723093748092651, "lr": 0.00028039701619904774, "ex_per_s": 371.14647135954425, "data_frac": 0.08604603447905008, "mem_mb": 91.210752, "nll": -5.912852764129639, "elapsed_s": 148.3}45{"event": "train", "step": 880, "loss": -5.868540287017822, "grad_norm": 0.6701368689537048, "lr": 0.00027922285907682625, "ex_per_s": 375.5583449709214, "data_frac": 0.0851705259636538, "mem_mb": 91.210752, "nll": -5.868540287017822, "elapsed_s": 150.0}46{"event": "train", "step": 900, "loss": -5.986633777618408, "grad_norm": 0.3292490541934967, "lr": 0.00027801747456544134, "ex_per_s": 379.8693867831399, "data_frac": 0.08431369987614805, "mem_mb": 91.210752, "nll": -5.986633777618408, "elapsed_s": 151.7}47{"event": "train", "step": 920, "loss": -6.024877071380615, "grad_norm": 0.32513466477394104, "lr": 0.00027678119220498265, "ex_per_s": 384.08989537373014, "data_frac": 0.08347551448845397, "mem_mb": 91.210752, "nll": -6.024877071380615, "elapsed_s": 153.3}48{"event": "train", "step": 940, "loss": -6.080681800842285, "grad_norm": 1.0707793235778809, "lr": 0.0002755143499827035, "ex_per_s": 388.21862055681237, "data_frac": 0.08265635059980436, "mem_mb": 91.210752, "nll": -6.080681800842285, "elapsed_s": 155.0}49{"event": "train", "step": 960, "loss": -5.966760635375977, "grad_norm": 0.2681189775466919, "lr": 0.00027421729424061787, "ex_per_s": 392.26237673940005, "data_frac": 0.08185536816504851, "mem_mb": 91.210752, "nll": -5.966760635375977, "elapsed_s": 156.7}50{"event": "train", "step": 980, "loss": -5.9257683753967285, "grad_norm": 0.3327540457248688, "lr": 0.00027289037958081483, "ex_per_s": 396.2143508763002, "data_frac": 0.08107403301496038, "mem_mb": 91.210752, "nll": -5.9257683753967285, "elapsed_s": 158.3}51{"event": "train", "step": 1000, "loss": -6.072744846343994, "grad_norm": 0.21865315735340118, "lr": 0.0002715339687685131, "ex_per_s": 400.09646059102016, "data_frac": 0.08030771670792215, "mem_mb": 91.210752, "nll": -6.072744846343994, "elapsed_s": 160.0}52{"event": "val", "step": 1000, "val_nll": -6.44661291971826, "best": -6.44661291971826, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 223.4}53{"event": "train", "step": 1020, "loss": -6.031286239624023, "grad_norm": 0.16486471891403198, "lr": 0.0002701484326328852, "ex_per_s": 289.8886014119289, "data_frac": 0.05710208048306793, "mem_mb": 91.210752, "nll": -6.031286239624023, "elapsed_s": 225.2}54{"event": "train", "step": 1040, "loss": -5.9719343185424805, "grad_norm": 0.3425344228744507, "lr": 0.00026873414996567596, "ex_per_s": 293.39373580146236, "data_frac": 0.05673370886048714, "mem_mb": 91.210752, "nll": -5.9719343185424805, "elapsed_s": 226.9}55{"event": "train", "step": 1060, "loss": -6.18073844909668, "grad_norm": 0.1776650995016098, "lr": 0.0002672915074176452, "ex_per_s": 296.84975708140445, "data_frac": 0.05637319950771705, "mem_mb": 91.210752, "nll": -6.18073844909668, "elapsed_s": 228.6}56{"event": "train", "step": 1080, "loss": -5.814178943634033, "grad_norm": 0.4762097895145416, "lr": 0.00026582089939286075, "ex_per_s": 300.2520500891749, "data_frac": 0.05601823909174326, "mem_mb": 91.210752, "nll": -5.814178943634033, "elapsed_s": 230.3}57{"event": "train", "step": 1100, "loss": -6.053707599639893, "grad_norm": 0.11912494152784348, "lr": 0.0002643227279408728, "ex_per_s": 303.60674298420184, "data_frac": 0.05566771838908222, "mem_mb": 91.210752, "nll": -6.053707599639893, "elapsed_s": 231.9}58{"event": "train", "step": 1120, "loss": -6.051762104034424, "grad_norm": 0.2870774269104004, "lr": 0.000262797402646797, "ex_per_s": 306.91471962137166, "data_frac": 0.055321348752941535, "mem_mb": 91.210752, "nll": -6.051762104034424, "elapsed_s": 233.6}59{"event": "train", "step": 1140, "loss": -6.019619941711426, "grad_norm": 0.21239204704761505, "lr": 0.0002612453405193386, "ex_per_s": 310.17404547513644, "data_frac": 0.054979889789317275, "mem_mb": 91.210752, "nll": -6.019619941711426, "elapsed_s": 235.3}60{"event": "train", "step": 1160, "loss": -5.969160556793213, "grad_norm": 0.7282019257545471, "lr": 0.0002596669658767859, "ex_per_s": 313.3866133061394, "data_frac": 0.0546445827314099, "mem_mb": 91.210752, "nll": -5.969160556793213, "elapsed_s": 236.9}61{"event": "train", "step": 1180, "loss": -5.958034038543701, "grad_norm": 0.1252554953098297, "lr": 0.00025806271023100637, "ex_per_s": 316.5524898040203, "data_frac": 0.054317030783497665, "mem_mb": 91.210752, "nll": -5.958034038543701, "elapsed_s": 238.6}62{"event": "train", "step": 1200, "loss": -5.945321083068848, "grad_norm": 0.25687989592552185, "lr": 0.00025643301216947504, "ex_per_s": 319.68162645294086, "data_frac": 0.053990115873987186, "mem_mb": 91.210752, "nll": -5.945321083068848, "elapsed_s": 240.3}63{"event": "train", "step": 1220, "loss": -6.033962726593018, "grad_norm": 0.5134910941123962, "lr": 0.0002547783172353695, "ex_per_s": 322.77196831584587, "data_frac": 0.05366736102076471, "mem_mb": 91.210752, "nll": -6.033962726593018, "elapsed_s": 241.9}64{"event": "train", "step": 1240, "loss": -6.059638023376465, "grad_norm": 0.06685866415500641, "lr": 0.00025309907780576277, "ex_per_s": 325.82277222126413, "data_frac": 0.05334960625381823, "mem_mb": 91.210752, "nll": -6.059638023376465, "elapsed_s": 243.6}65{"event": "train", "step": 1260, "loss": -6.054544925689697, "grad_norm": 0.19677866995334625, "lr": 0.0002513957529679477, "ex_per_s": 328.83077390709036, "data_frac": 0.05303621630318158, "mem_mb": 91.210752, "nll": -6.054544925689697, "elapsed_s": 245.3}66{"event": "train", "step": 1280, "loss": -6.0430827140808105, "grad_norm": 0.5033031702041626, "lr": 0.00024966880839392746, "ex_per_s": 331.7954003851886, "data_frac": 0.052726421852926815, "mem_mb": 91.210752, "nll": -6.0430827140808105, "elapsed_s": 246.9}67{"event": "train", "step": 1300, "loss": -5.867796897888184, "grad_norm": 0.7240657210350037, "lr": 0.0002479187162131052, "ex_per_s": 334.7263623841849, "data_frac": 0.05242098935644481, "mem_mb": 91.210752, "nll": -5.867796897888184, "elapsed_s": 248.6}68{"event": "train", "step": 1320, "loss": -5.8280229568481445, "grad_norm": 0.2152756005525589, "lr": 0.00024614595488320833, "ex_per_s": 337.6191096416125, "data_frac": 0.05212031208026861, "mem_mb": 91.210752, "nll": -5.8280229568481445, "elapsed_s": 250.3}69{"event": "train", "step": 1340, "loss": -5.9495086669921875, "grad_norm": 0.4887429177761078, "lr": 0.00024435100905948387, "ex_per_s": 340.4711168241129, "data_frac": 0.05182223748857386, "mem_mb": 91.210752, "nll": -5.9495086669921875, "elapsed_s": 251.9}70{"event": "train", "step": 1360, "loss": -6.037879467010498, "grad_norm": 0.39390358328819275, "lr": 0.0002425343694621974, "ex_per_s": 343.28531842466623, "data_frac": 0.051529277520180994, "mem_mb": 91.210752, "nll": -6.037879467010498, "elapsed_s": 253.6}71{"event": "train", "step": 1380, "loss": -6.066823482513428, "grad_norm": 0.21028296649456024, "lr": 0.00024069653274247583, "ex_per_s": 346.0625608336384, "data_frac": 0.05124121583176193, "mem_mb": 91.210752, "nll": -6.066823482513428, "elapsed_s": 255.3}72{"event": "train", "step": 1400, "loss": -5.795958995819092, "grad_norm": 0.31144019961357117, "lr": 0.0002388380013465276, "ex_per_s": 348.7960731507029, "data_frac": 0.05095455834611864, "mem_mb": 91.210752, "nll": -5.795958995819092, "elapsed_s": 256.9}73{"event": "train", "step": 1420, "loss": -5.989187717437744, "grad_norm": 0.7897518873214722, "lr": 0.00023695928337827914, "ex_per_s": 351.485495061227, "data_frac": 0.05067471007551574, "mem_mb": 91.210752, "nll": -5.989187717437744, "elapsed_s": 258.6}74{"event": "train", "step": 1440, "loss": -5.742162227630615, "grad_norm": 0.22511489689350128, "lr": 0.00023506089246046403, "ex_per_s": 354.1494701900075, "data_frac": 0.05039586800300268, "mem_mb": 91.210752, "nll": -5.742162227630615, "elapsed_s": 260.3}75{"event": "train", "step": 1460, "loss": -5.798532009124756, "grad_norm": 0.37716561555862427, "lr": 0.00023314334759420388, "ex_per_s": 356.77953327498363, "data_frac": 0.05011948688587864, "mem_mb": 91.210752, "nll": -5.798532009124756, "elapsed_s": 261.9}76{"event": "train", "step": 1480, "loss": -6.047031879425049, "grad_norm": 0.1628723293542862, "lr": 0.0002312071730171186, "ex_per_s": 359.3764682262541, "data_frac": 0.049847230628529825, "mem_mb": 91.210752, "nll": -6.047031879425049, "elapsed_s": 263.6}77{"event": "train", "step": 1500, "loss": -5.9575605392456055, "grad_norm": 0.42324820160865784, "lr": 0.0002292528980600049, "ex_per_s": 361.943406118988, "data_frac": 0.04957900134817707, "mem_mb": 91.210752, "nll": -5.9575605392456055, "elapsed_s": 265.3}78{"event": "val", "step": 1500, "val_nll": -6.4462253332324355, "best": -6.44661291971826, "bad": 1, "mem_mb": 91.210752, "elapsed_s": 328.6}79{"event": "train", "step": 1520, "loss": -5.949007987976074, "grad_norm": 0.5437551736831665, "lr": 0.00022728105700212285, "ex_per_s": 294.49908328148445, "data_frac": 0.039846425265353654, "mem_mb": 91.210752, "nll": -5.949007987976074, "elapsed_s": 330.4}80{"event": "train", "step": 1540, "loss": -5.879621505737305, "grad_norm": 0.23518329858779907, "lr": 0.00022529218892512886, "ex_per_s": 296.8706219111185, "data_frac": 0.03968163835932199, "mem_mb": 91.210752, "nll": -5.879621505737305, "elapsed_s": 332.0}81{"event": "train", "step": 1560, "loss": -5.98321533203125, "grad_norm": 0.2751246988773346, "lr": 0.00022328683756569652, "ex_per_s": 299.2184895231456, "data_frac": 0.039518192302577554, "mem_mb": 91.210752, "nll": -5.98321533203125, "elapsed_s": 333.7}82{"event": "train", "step": 1580, "loss": -5.9809675216674805, "grad_norm": 0.0954427644610405, "lr": 0.0002212655511668642, "ex_per_s": 301.54392659965816, "data_frac": 0.03935628883714587, "mem_mb": 91.210752, "nll": -5.9809675216674805, "elapsed_s": 335.4}83{"event": "train", "step": 1600, "loss": -5.95818567276001, "grad_norm": 0.26962560415267944, "lr": 0.00021922888232815089, "ex_per_s": 303.8468319138065, "data_frac": 0.039196186052133444, "mem_mb": 91.210752, "nll": -5.95818567276001, "elapsed_s": 337.1}84{"event": "train", "step": 1620, "loss": -5.880678176879883, "grad_norm": 0.17133653163909912, "lr": 0.00021717738785448102, "ex_per_s": 306.1241527185006, "data_frac": 0.039038603603849704, "mem_mb": 91.210752, "nll": -5.880678176879883, "elapsed_s": 338.7}85{"event": "train", "step": 1640, "loss": -6.076656341552734, "grad_norm": 0.15402498841285706, "lr": 0.00021511162860395977, "ex_per_s": 308.37900885580757, "data_frac": 0.03888187869847526, "mem_mb": 91.210752, "nll": -6.076656341552734, "elapsed_s": 340.4}86{"event": "train", "step": 1660, "loss": -6.041697978973389, "grad_norm": 0.7559201717376709, "lr": 0.0002130321693345397, "ex_per_s": 310.61528043156517, "data_frac": 0.03872718888437109, "mem_mb": 91.210752, "nll": -6.041697978973389, "elapsed_s": 342.1}87{"event": "train", "step": 1680, "loss": -5.865788459777832, "grad_norm": 0.17714549601078033, "lr": 0.0002109395785496218, "ex_per_s": 312.8287018671997, "data_frac": 0.03857328484572117, "mem_mb": 91.210752, "nll": -5.865788459777832, "elapsed_s": 343.7}88{"event": "train", "step": 1700, "loss": -5.92788028717041, "grad_norm": 0.5239881873130798, "lr": 0.00020883442834263225, "ex_per_s": 315.0208370826053, "data_frac": 0.0384209105252021, "mem_mb": 91.210752, "nll": -5.92788028717041, "elapsed_s": 345.4}89{"event": "train", "step": 1720, "loss": -5.889191150665283, "grad_norm": 0.68222576379776, "lr": 0.00020671729424061788, "ex_per_s": 317.192840524902, "data_frac": 0.038270258111383, "mem_mb": 91.210752, "nll": -5.889191150665283, "elapsed_s": 347.1}90{"event": "train", "step": 1740, "loss": -6.275003910064697, "grad_norm": 0.32420048117637634, "lr": 0.00020458875504690275, "ex_per_s": 319.3443046069082, "data_frac": 0.03812136696883874, "mem_mb": 91.210752, "nll": -6.275003910064697, "elapsed_s": 348.8}91{"event": "train", "step": 1760, "loss": -5.934019565582275, "grad_norm": 0.2429039627313614, "lr": 0.000202449392682849, "ex_per_s": 321.4737673275212, "data_frac": 0.037973763987557196, "mem_mb": 91.210752, "nll": -5.934019565582275, "elapsed_s": 350.4}92{"event": "train", "step": 1780, "loss": -5.86998987197876, "grad_norm": 0.49571388959884644, "lr": 0.00020029979202876563, "ex_per_s": 323.58620585148367, "data_frac": 0.037827774959328875, "mem_mb": 91.210752, "nll": -5.86998987197876, "elapsed_s": 352.1}93{"event": "train", "step": 1800, "loss": -6.047395706176758, "grad_norm": 1.1008087396621704, "lr": 0.00019814054076400787, "ex_per_s": 325.67795893069757, "data_frac": 0.03768250219329241, "mem_mb": 91.210752, "nll": -6.047395706176758, "elapsed_s": 353.8}94{"event": "train", "step": 1820, "loss": -5.933111190795898, "grad_norm": 0.9901772141456604, "lr": 0.00019597222920631147, "ex_per_s": 327.74906114552095, "data_frac": 0.03753894801614725, "mem_mb": 91.210752, "nll": -5.933111190795898, "elapsed_s": 355.4}95{"event": "train", "step": 1840, "loss": -6.0723652839660645, "grad_norm": 0.41734182834625244, "lr": 0.0001937954501504058, "ex_per_s": 329.80105817344213, "data_frac": 0.03739713468902408, "mem_mb": 91.210752, "nll": -6.0723652839660645, "elapsed_s": 357.1}96{"event": "train", "step": 1860, "loss": -6.052231311798096, "grad_norm": 0.12866413593292236, "lr": 0.0001916107987059498, "ex_per_s": 331.83662023868067, "data_frac": 0.0372588274810841, "mem_mb": 91.210752, "nll": -6.052231311798096, "elapsed_s": 358.8}97{"event": "train", "step": 1880, "loss": -5.950647354125977, "grad_norm": 0.2850516140460968, "lr": 0.00018941887213483484, "ex_per_s": 333.8561998770593, "data_frac": 0.037121074528637284, "mem_mb": 91.210752, "nll": -5.950647354125977, "elapsed_s": 360.4}98{"event": "train", "step": 1900, "loss": -5.979011058807373, "grad_norm": 0.2849499583244324, "lr": 0.00018722026968789906, "ex_per_s": 335.85660986088857, "data_frac": 0.036983779451184785, "mem_mb": 91.210752, "nll": -5.979011058807373, "elapsed_s": 362.1}99{"event": "train", "step": 1920, "loss": -5.963472366333008, "grad_norm": 0.4363224506378174, "lr": 0.00018501559244109864, "ex_per_s": 337.83930152591375, "data_frac": 0.036847829807172904, "mem_mb": 91.210752, "nll": -5.963472366333008, "elapsed_s": 363.8}100{"event": "train", "step": 1940, "loss": -6.049674034118652, "grad_norm": 0.29839521646499634, "lr": 0.00018280544313117932, "ex_per_s": 339.80469337654876, "data_frac": 0.03671228091680137, "mem_mb": 91.210752, "nll": -6.049674034118652, "elapsed_s": 365.4}101{"event": "train", "step": 1960, "loss": -5.939127445220947, "grad_norm": 0.05706757679581642, "lr": 0.00018059042599089415, "ex_per_s": 341.7511754243266, "data_frac": 0.03657840592545589, "mem_mb": 91.210752, "nll": -5.939127445220947, "elapsed_s": 367.1}102{"event": "train", "step": 1980, "loss": -5.963512420654297, "grad_norm": 0.7613920569419861, "lr": 0.00017837114658381247, "ex_per_s": 343.6795419477707, "data_frac": 0.03644545492808853, "mem_mb": 91.210752, "nll": -5.963512420654297, "elapsed_s": 368.8}103{"event": "train", "step": 2000, "loss": -5.915711402893066, "grad_norm": 0.37094271183013916, "lr": 0.00017614821163876486, "ex_per_s": 345.5926327199205, "data_frac": 0.03631428392324587, "mem_mb": 91.210752, "nll": -5.915711402893066, "elapsed_s": 370.4}104{"event": "val", "step": 2000, "val_nll": -6.448200591967574, "best": -6.448200591967574, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 433.9}105{"event": "train", "step": 2020, "loss": -5.929598331451416, "grad_norm": 0.49822893738746643, "lr": 0.00017392222888396941, "ex_per_s": 296.81256057158504, "data_frac": 0.030908156182554453, "mem_mb": 91.210752, "nll": -5.929598331451416, "elapsed_s": 435.6}106{"event": "train", "step": 2040, "loss": -5.8742594718933105, "grad_norm": 0.27631962299346924, "lr": 0.000171693806880885, "ex_per_s": 298.6073077813814, "data_frac": 0.030817158258009705, "mem_mb": 91.210752, "nll": -5.8742594718933105, "elapsed_s": 437.3}107{"event": "train", "step": 2060, "loss": -6.145603656768799, "grad_norm": 0.27360278367996216, "lr": 0.0001694635548578366, "ex_per_s": 300.3864360677603, "data_frac": 0.030726922727333794, "mem_mb": 91.210752, "nll": -6.145603656768799, "elapsed_s": 438.9}108{"event": "train", "step": 2080, "loss": -5.963015079498291, "grad_norm": 0.6466829776763916, "lr": 0.0001672320825434584, "ex_per_s": 302.15257431824523, "data_frac": 0.030639991970705677, "mem_mb": 91.210752, "nll": -5.963015079498291, "elapsed_s": 440.6}109{"event": "train", "step": 2100, "loss": -5.899890422821045, "grad_norm": 0.8096009492874146, "lr": 0.000165, "ex_per_s": 303.90571352542344, "data_frac": 0.03055122172311537, "mem_mb": 91.210752, "nll": -5.899890422821045, "elapsed_s": 442.3}110{"event": "train", "step": 2120, "loss": -6.151511192321777, "grad_norm": 0.4180460274219513, "lr": 0.0001627679174565416, "ex_per_s": 305.64820694162023, "data_frac": 0.030462910952027515, "mem_mb": 91.210752, "nll": -6.151511192321777, "elapsed_s": 444.0}111{"event": "train", "step": 2140, "loss": -5.98084831237793, "grad_norm": 0.14250099658966064, "lr": 0.00016053644514216336, "ex_per_s": 307.37596977209785, "data_frac": 0.030374699971841414, "mem_mb": 91.210752, "nll": -5.98084831237793, "elapsed_s": 445.6}112{"event": "train", "step": 2160, "loss": -5.924901962280273, "grad_norm": 0.5160773992538452, "lr": 0.00015830619311911498, "ex_per_s": 309.09008735277826, "data_frac": 0.030288387951796548, "mem_mb": 91.210752, "nll": -5.924901962280273, "elapsed_s": 447.3}113{"event": "train", "step": 2180, "loss": -6.082257270812988, "grad_norm": 0.19706229865550995, "lr": 0.00015607777111603058, "ex_per_s": 310.79331823949195, "data_frac": 0.03020332613185344, "mem_mb": 91.210752, "nll": -6.082257270812988, "elapsed_s": 449.0}114{"event": "train", "step": 2200, "loss": -5.962789535522461, "grad_norm": 0.595615029335022, "lr": 0.00015385178836123514, "ex_per_s": 312.48474277078685, "data_frac": 0.03011829314601005, "mem_mb": 91.210752, "nll": -5.962789535522461, "elapsed_s": 450.6}115{"event": "train", "step": 2220, "loss": -5.824714183807373, "grad_norm": 0.8048741221427917, "lr": 0.00015162885341618752, "ex_per_s": 314.1632849077003, "data_frac": 0.03003383675281762, "mem_mb": 91.210752, "nll": -5.824714183807373, "elapsed_s": 452.3}116{"event": "train", "step": 2240, "loss": -5.960598468780518, "grad_norm": 0.786283016204834, "lr": 0.00014940957400910585, "ex_per_s": 315.8311548390084, "data_frac": 0.029950291621526882, "mem_mb": 91.210752, "nll": -5.960598468780518, "elapsed_s": 454.0}117{"event": "train", "step": 2260, "loss": -5.739709854125977, "grad_norm": 0.692767858505249, "lr": 0.00014719455686882068, "ex_per_s": 317.48695096617513, "data_frac": 0.029866896686993993, "mem_mb": 91.210752, "nll": -5.739709854125977, "elapsed_s": 455.6}118{"event": "train", "step": 2280, "loss": -6.010705471038818, "grad_norm": 0.47674041986465454, "lr": 0.0001449844075589013, "ex_per_s": 319.13232668053183, "data_frac": 0.029784352593357953, "mem_mb": 91.210752, "nll": -6.010705471038818, "elapsed_s": 457.3}119{"event": "train", "step": 2300, "loss": -6.00281286239624, "grad_norm": 0.10559073835611343, "lr": 0.0001427797303121009, "ex_per_s": 320.76608783016155, "data_frac": 0.02970173989277334, "mem_mb": 91.210752, "nll": -6.00281286239624, "elapsed_s": 458.9}120{"event": "train", "step": 2320, "loss": -6.068641662597656, "grad_norm": 0.4743320643901825, "lr": 0.00014058112786516518, "ex_per_s": 322.3887694536237, "data_frac": 0.029622948250210016, "mem_mb": 91.210752, "nll": -6.068641662597656, "elapsed_s": 460.6}121{"event": "train", "step": 2340, "loss": -5.849883079528809, "grad_norm": 0.36671528220176697, "lr": 0.00013838920129405016, "ex_per_s": 324.00389743069115, "data_frac": 0.029542346269771848, "mem_mb": 91.210752, "nll": -5.849883079528809, "elapsed_s": 462.3}122{"event": "train", "step": 2360, "loss": -5.929271697998047, "grad_norm": 0.09555365145206451, "lr": 0.0001362045498495942, "ex_per_s": 325.6083534990483, "data_frac": 0.02946301876106387, "mem_mb": 91.210752, "nll": -5.929271697998047, "elapsed_s": 463.9}123{"event": "train", "step": 2380, "loss": -5.9323554039001465, "grad_norm": 0.10872185975313187, "lr": 0.00013402777079368853, "ex_per_s": 327.20297695549834, "data_frac": 0.029383870384010496, "mem_mb": 91.210752, "nll": -5.9323554039001465, "elapsed_s": 465.6}124{"event": "train", "step": 2400, "loss": -5.953352451324463, "grad_norm": 1.0205931663513184, "lr": 0.00013185945923599213, "ex_per_s": 328.7858284772794, "data_frac": 0.029305274796503435, "mem_mb": 91.210752, "nll": -5.953352451324463, "elapsed_s": 467.2}125{"event": "train", "step": 2420, "loss": -5.998089790344238, "grad_norm": 0.15693260729312897, "lr": 0.00012970020797123432, "ex_per_s": 330.35718441808064, "data_frac": 0.0292274735004563, "mem_mb": 91.210752, "nll": -5.998089790344238, "elapsed_s": 468.9}126{"event": "train", "step": 2440, "loss": -6.0979180335998535, "grad_norm": 0.5580943822860718, "lr": 0.000127550607317151, "ex_per_s": 331.9165050769943, "data_frac": 0.02915024518067564, "mem_mb": 91.210752, "nll": -6.0979180335998535, "elapsed_s": 470.5}127{"event": "train", "step": 2460, "loss": -6.13224458694458, "grad_norm": 0.0740194171667099, "lr": 0.00012541124495309727, "ex_per_s": 333.4654437267783, "data_frac": 0.029073762274749504, "mem_mb": 91.210752, "nll": -6.13224458694458, "elapsed_s": 472.2}128{"event": "train", "step": 2480, "loss": -5.942627906799316, "grad_norm": 0.19153273105621338, "lr": 0.00012328270575938212, "ex_per_s": 335.0038335277876, "data_frac": 0.02899796090776193, "mem_mb": 91.210752, "nll": -5.942627906799316, "elapsed_s": 473.8}129{"event": "train", "step": 2500, "loss": -5.751960277557373, "grad_norm": 0.657241940498352, "lr": 0.00012116557165736775, "ex_per_s": 336.53024593089884, "data_frac": 0.028922216338763348, "mem_mb": 91.210752, "nll": -5.751960277557373, "elapsed_s": 475.5}130{"event": "val", "step": 2500, "val_nll": -6.449132441905125, "best": -6.449132441905125, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 538.8}131{"event": "train", "step": 2520, "loss": -5.962214469909668, "grad_norm": 0.3913393020629883, "lr": 0.00011906042145037817, "ex_per_s": 298.36861186741385, "data_frac": 0.025462854448235026, "mem_mb": 91.210752, "nll": -5.962214469909668, "elapsed_s": 540.6}132{"event": "train", "step": 2540, "loss": -6.030779838562012, "grad_norm": 0.4552779793739319, "lr": 0.00011696783066546026, "ex_per_s": 299.8086688045799, "data_frac": 0.025406771037757427, "mem_mb": 91.210752, "nll": -6.030779838562012, "elapsed_s": 542.3}133{"event": "train", "step": 2560, "loss": -5.949045181274414, "grad_norm": 0.24085958302021027, "lr": 0.00011488837139604017, "ex_per_s": 301.2411083332522, "data_frac": 0.025350946941538423, "mem_mb": 91.210752, "nll": -5.949045181274414, "elapsed_s": 543.9}134{"event": "train", "step": 2580, "loss": -5.9643235206604, "grad_norm": 0.21490180492401123, "lr": 0.00011282261214551898, "ex_per_s": 302.66325073670095, "data_frac": 0.02529565999124431, "mem_mb": 91.210752, "nll": -5.9643235206604, "elapsed_s": 545.6}135{"event": "train", "step": 2600, "loss": -5.851477146148682, "grad_norm": 0.21817806363105774, "lr": 0.00011077111767184915, "ex_per_s": 304.07796270175345, "data_frac": 0.02524023829831243, "mem_mb": 91.210752, "nll": -5.851477146148682, "elapsed_s": 547.3}136{"event": "train", "step": 2620, "loss": -5.890172004699707, "grad_norm": 0.9029527902603149, "lr": 0.00010873444883313583, "ex_per_s": 305.48400961297057, "data_frac": 0.02518511781859109, "mem_mb": 91.210752, "nll": -5.890172004699707, "elapsed_s": 548.9}137{"event": "train", "step": 2640, "loss": -5.984892845153809, "grad_norm": 0.10841244459152222, "lr": 0.00010671316243430345, "ex_per_s": 306.8814296923322, "data_frac": 0.025130469240220887, "mem_mb": 91.210752, "nll": -5.984892845153809, "elapsed_s": 550.6}138{"event": "train", "step": 2660, "loss": -5.821751594543457, "grad_norm": 0.08009308576583862, "lr": 0.00010470781107487112, "ex_per_s": 308.2713316209426, "data_frac": 0.02507581487723828, "mem_mb": 91.210752, "nll": -5.821751594543457, "elapsed_s": 552.3}139{"event": "train", "step": 2680, "loss": -5.958373546600342, "grad_norm": 0.40573960542678833, "lr": 0.00010271894299787715, "ex_per_s": 309.6520176859161, "data_frac": 0.025021344162485543, "mem_mb": 91.210752, "nll": -5.958373546600342, "elapsed_s": 554.0}140{"event": "train", "step": 2700, "loss": -6.132401943206787, "grad_norm": 0.8206536173820496, "lr": 0.00010074710193999506, "ex_per_s": 311.0248026528537, "data_frac": 0.02496793880479601, "mem_mb": 91.210752, "nll": -6.132401943206787, "elapsed_s": 555.6}141{"event": "train", "step": 2720, "loss": -5.882719039916992, "grad_norm": 0.25992679595947266, "lr": 9.879282698288144e-05, "ex_per_s": 312.3885994775572, "data_frac": 0.024914483509883345, "mem_mb": 91.210752, "nll": -5.882719039916992, "elapsed_s": 557.3}142{"event": "train", "step": 2740, "loss": -6.124802112579346, "grad_norm": 0.1368250548839569, "lr": 9.685665240579613e-05, "ex_per_s": 313.745466346556, "data_frac": 0.02486139552171857, "mem_mb": 91.210752, "nll": -6.124802112579346, "elapsed_s": 559.0}143{"event": "train", "step": 2760, "loss": -5.962599277496338, "grad_norm": 0.23092204332351685, "lr": 9.4939107539536e-05, "ex_per_s": 315.093518395051, "data_frac": 0.02480939081292178, "mem_mb": 91.210752, "nll": -5.962599277496338, "elapsed_s": 560.6}144{"event": "train", "step": 2780, "loss": -6.014966011047363, "grad_norm": 0.9323200583457947, "lr": 9.304071662172087e-05, "ex_per_s": 316.4366855759404, "data_frac": 0.024757464186078302, "mem_mb": 91.210752, "nll": -6.014966011047363, "elapsed_s": 562.3}145{"event": "train", "step": 2800, "loss": -5.979398250579834, "grad_norm": 0.7600387334823608, "lr": 9.116199865347241e-05, "ex_per_s": 317.77290840067667, "data_frac": 0.024705739834805365, "mem_mb": 91.210752, "nll": -5.979398250579834, "elapsed_s": 564.0}146{"event": "train", "step": 2820, "loss": -6.009033203125, "grad_norm": 0.06505793333053589, "lr": 8.930346725752419e-05, "ex_per_s": 319.10079264238243, "data_frac": 0.024654188070486527, "mem_mb": 91.210752, "nll": -6.009033203125, "elapsed_s": 565.6}147{"event": "train", "step": 2840, "loss": -5.87723970413208, "grad_norm": 0.09920528531074524, "lr": 8.746563053780262e-05, "ex_per_s": 320.4224397551746, "data_frac": 0.02460309365407193, "mem_mb": 91.210752, "nll": -5.87723970413208, "elapsed_s": 567.3}148{"event": "train", "step": 2860, "loss": -6.049408435821533, "grad_norm": 0.16015097498893738, "lr": 8.564899094051614e-05, "ex_per_s": 321.73362394097336, "data_frac": 0.024553163442426692, "mem_mb": 91.210752, "nll": -6.049408435821533, "elapsed_s": 569.0}149{"event": "train", "step": 2880, "loss": -5.772556781768799, "grad_norm": 0.23239129781723022, "lr": 8.385404511679161e-05, "ex_per_s": 323.0394880568601, "data_frac": 0.024502944848474376, "mem_mb": 91.210752, "nll": -5.772556781768799, "elapsed_s": 570.6}150{"event": "train", "step": 2900, "loss": -5.797028541564941, "grad_norm": 0.5727108120918274, "lr": 8.208128378689484e-05, "ex_per_s": 324.3389335416719, "data_frac": 0.024453059968253164, "mem_mb": 91.210752, "nll": -5.797028541564941, "elapsed_s": 572.3}151{"event": "train", "step": 2920, "loss": -5.969655990600586, "grad_norm": 0.6025851368904114, "lr": 8.03311916060725e-05, "ex_per_s": 325.63030572632397, "data_frac": 0.024403847239662928, "mem_mb": 91.210752, "nll": -5.969655990600586, "elapsed_s": 573.9}152{"event": "train", "step": 2940, "loss": -5.902832984924316, "grad_norm": 0.3537255525588989, "lr": 7.860424703205225e-05, "ex_per_s": 326.9139225432558, "data_frac": 0.024354871857519337, "mem_mb": 91.210752, "nll": -5.902832984924316, "elapsed_s": 575.6}153{"event": "train", "step": 2960, "loss": -5.935858726501465, "grad_norm": 0.3392787575721741, "lr": 7.690092219423722e-05, "ex_per_s": 328.19076490594205, "data_frac": 0.024306128396775085, "mem_mb": 91.210752, "nll": -5.935858726501465, "elapsed_s": 577.3}154{"event": "train", "step": 2980, "loss": -5.894184589385986, "grad_norm": 0.37363800406455994, "lr": 7.522168276463048e-05, "ex_per_s": 329.460955028663, "data_frac": 0.024257385239369943, "mem_mb": 91.210752, "nll": -5.894184589385986, "elapsed_s": 578.9}155{"event": "train", "step": 3000, "loss": -5.994454860687256, "grad_norm": 0.10308095812797546, "lr": 7.356698783052496e-05, "ex_per_s": 330.72108398673834, "data_frac": 0.024208885549979832, "mem_mb": 91.210752, "nll": -5.994454860687256, "elapsed_s": 580.6}156{"event": "val", "step": 3000, "val_nll": -6.448212869054721, "best": -6.449132441905125, "bad": 1, "mem_mb": 91.210752, "elapsed_s": 643.5}157{"event": "train", "step": 3020, "loss": -5.961949825286865, "grad_norm": 0.8533371686935425, "lr": 7.193728976899362e-05, "ex_per_s": 299.5569580522303, "data_frac": 0.02180151665838792, "mem_mb": 91.210752, "nll": -5.961949825286865, "elapsed_s": 645.3}158{"event": "train", "step": 3040, "loss": -5.859830856323242, "grad_norm": 0.814677357673645, "lr": 7.033303412321404e-05, "ex_per_s": 300.76203959989766, "data_frac": 0.021763520666212205, "mem_mb": 91.210752, "nll": -5.859830856323242, "elapsed_s": 646.9}159{"event": "train", "step": 3060, "loss": -6.0737714767456055, "grad_norm": 0.20591221749782562, "lr": 6.875465948066139e-05, "ex_per_s": 301.96098023182077, "data_frac": 0.021725967524517078, "mem_mb": 91.210752, "nll": -6.0737714767456055, "elapsed_s": 648.6}160{"event": "train", "step": 3080, "loss": -5.969391345977783, "grad_norm": 0.215475395321846, "lr": 6.720259735320299e-05, "ex_per_s": 303.1517335667469, "data_frac": 0.021688236080797414, "mem_mb": 91.210752, "nll": -5.969391345977783, "elapsed_s": 650.3}161{"event": "train", "step": 3100, "loss": -6.011258125305176, "grad_norm": 0.38323384523391724, "lr": 6.567727205912722e-05, "ex_per_s": 304.3374927487603, "data_frac": 0.021650665473230066, "mem_mb": 91.210752, "nll": -6.011258125305176, "elapsed_s": 652.0}162{"event": "train", "step": 3120, "loss": -6.0930585861206055, "grad_norm": 0.4449201822280884, "lr": 6.417910060713926e-05, "ex_per_s": 305.5178983066157, "data_frac": 0.021613695925600522, "mem_mb": 91.210752, "nll": -6.0930585861206055, "elapsed_s": 653.6}163{"event": "train", "step": 3140, "loss": -6.008031368255615, "grad_norm": 0.36498942971229553, "lr": 6.270849258235485e-05, "ex_per_s": 306.69150177208223, "data_frac": 0.021576928308127, "mem_mb": 91.210752, "nll": -6.008031368255615, "elapsed_s": 655.3}164{"event": "train", "step": 3160, "loss": -6.024993896484375, "grad_norm": 0.10445784032344818, "lr": 6.126585003432406e-05, "ex_per_s": 307.859856760044, "data_frac": 0.021540159287388318, "mem_mb": 91.210752, "nll": -6.024993896484375, "elapsed_s": 657.0}165{"event": "train", "step": 3180, "loss": -6.097517013549805, "grad_norm": 0.10034821182489395, "lr": 5.985156736711483e-05, "ex_per_s": 309.0198821681983, "data_frac": 0.021503587532823144, "mem_mb": 91.210752, "nll": -6.097517013549805, "elapsed_s": 658.6}166{"event": "train", "step": 3200, "loss": -6.001983165740967, "grad_norm": 0.41901201009750366, "lr": 5.846603123148688e-05, "ex_per_s": 310.17621141835104, "data_frac": 0.021467312451234326, "mem_mb": 91.210752, "nll": -6.001983165740967, "elapsed_s": 660.3}167{"event": "train", "step": 3220, "loss": -5.913397312164307, "grad_norm": 0.294695645570755, "lr": 5.710962041918516e-05, "ex_per_s": 311.3250399013421, "data_frac": 0.021431869072481343, "mem_mb": 91.210752, "nll": -5.913397312164307, "elapsed_s": 662.0}168{"event": "train", "step": 3240, "loss": -5.801403522491455, "grad_norm": 0.3009123206138611, "lr": 5.578270575938211e-05, "ex_per_s": 312.4720369784467, "data_frac": 0.021396177170233913, "mem_mb": 91.210752, "nll": -5.801403522491455, "elapsed_s": 663.7}169{"event": "train", "step": 3260, "loss": -5.960986137390137, "grad_norm": 0.32249948382377625, "lr": 5.448565001729653e-05, "ex_per_s": 313.6122470907717, "data_frac": 0.021361240088305065, "mem_mb": 91.210752, "nll": -5.960986137390137, "elapsed_s": 665.3}170{"event": "train", "step": 3280, "loss": -6.014379501342773, "grad_norm": 0.4219035804271698, "lr": 5.3218807795017296e-05, "ex_per_s": 314.74956418440235, "data_frac": 0.02132592415135685, "mem_mb": 91.210752, "nll": -6.014379501342773, "elapsed_s": 667.0}171{"event": "train", "step": 3300, "loss": -5.804994583129883, "grad_norm": 0.1873633861541748, "lr": 5.198252543455864e-05, "ex_per_s": 315.88050614690536, "data_frac": 0.021290921249722605, "mem_mb": 91.210752, "nll": -5.804994583129883, "elapsed_s": 668.7}172{"event": "train", "step": 3320, "loss": -5.986661911010742, "grad_norm": 0.24588103592395782, "lr": 5.07771409231737e-05, "ex_per_s": 317.00665093759375, "data_frac": 0.021255862186208948, "mem_mb": 91.210752, "nll": -5.986661911010742, "elapsed_s": 670.3}173{"event": "train", "step": 3340, "loss": -5.931816577911377, "grad_norm": 0.29511553049087524, "lr": 4.960298380095221e-05, "ex_per_s": 318.1268159854256, "data_frac": 0.021221072250864028, "mem_mb": 91.210752, "nll": -5.931816577911377, "elapsed_s": 672.0}174{"event": "train", "step": 3360, "loss": -5.908273220062256, "grad_norm": 0.2392471730709076, "lr": 4.846037507072751e-05, "ex_per_s": 319.24212775278255, "data_frac": 0.021186461700453997, "mem_mb": 91.210752, "nll": -5.908273220062256, "elapsed_s": 673.6}175{"event": "train", "step": 3380, "loss": -6.124768257141113, "grad_norm": 0.41153132915496826, "lr": 4.7349627110317556e-05, "ex_per_s": 320.34967353385355, "data_frac": 0.02115218514702641, "mem_mb": 91.210752, "nll": -6.124768257141113, "elapsed_s": 675.3}176{"event": "train", "step": 3400, "loss": -5.8806352615356445, "grad_norm": 0.47771432995796204, "lr": 4.6271043587123986e-05, "ex_per_s": 321.45055070162766, "data_frac": 0.021118102923265592, "mem_mb": 91.210752, "nll": -5.8806352615356445, "elapsed_s": 677.0}177{"event": "train", "step": 3420, "loss": -5.838315010070801, "grad_norm": 0.1278006136417389, "lr": 4.52249193751124e-05, "ex_per_s": 322.5469466633883, "data_frac": 0.021083607264906216, "mem_mb": 91.210752, "nll": -5.838315010070801, "elapsed_s": 678.6}178{"event": "train", "step": 3440, "loss": -5.9176716804504395, "grad_norm": 0.10846627503633499, "lr": 4.4211540474196926e-05, "ex_per_s": 323.63797059471614, "data_frac": 0.02104933731524033, "mem_mb": 91.210752, "nll": -5.9176716804504395, "elapsed_s": 680.3}179{"event": "train", "step": 3460, "loss": -6.059169292449951, "grad_norm": 0.1935606747865677, "lr": 4.3231183932050546e-05, "ex_per_s": 324.7209941012376, "data_frac": 0.02101588757380928, "mem_mb": 91.210752, "nll": -6.059169292449951, "elapsed_s": 682.0}180{"event": "train", "step": 3480, "loss": -5.9176459312438965, "grad_norm": 0.06862011551856995, "lr": 4.228411776836306e-05, "ex_per_s": 325.80089377814613, "data_frac": 0.020982155144109844, "mem_mb": 91.210752, "nll": -5.9176459312438965, "elapsed_s": 683.7}181{"event": "train", "step": 3500, "loss": -6.037683963775635, "grad_norm": 0.19733794033527374, "lr": 4.137060090156723e-05, "ex_per_s": 326.874735561381, "data_frac": 0.020949569040579052, "mem_mb": 91.210752, "nll": -6.037683963775635, "elapsed_s": 685.3}182{"event": "val", "step": 3500, "val_nll": -6.449357611608497, "best": -6.449357611608497, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 748.4}183{"event": "train", "step": 3520, "loss": -5.939156532287598, "grad_norm": 0.4007114768028259, "lr": 4.049088307805306e-05, "ex_per_s": 300.34878165972674, "data_frac": 0.019156337122232768, "mem_mb": 91.210752, "nll": -5.939156532287598, "elapsed_s": 750.1}184{"event": "train", "step": 3540, "loss": -6.1357245445251465, "grad_norm": 0.5887722969055176, "lr": 3.964520480388956e-05, "ex_per_s": 301.3835913727602, "data_frac": 0.019129453618071328, "mem_mb": 91.210752, "nll": -6.1357245445251465, "elapsed_s": 751.8}185{"event": "train", "step": 3560, "loss": -6.138377666473389, "grad_norm": 0.2110646516084671, "lr": 3.88337972790728e-05, "ex_per_s": 302.41309864111486, "data_frac": 0.019102724935368263, "mem_mb": 91.210752, "nll": -6.138377666473389, "elapsed_s": 753.5}186{"event": "train", "step": 3580, "loss": -5.9038166999816895, "grad_norm": 0.17634493112564087, "lr": 3.805688233431824e-05, "ex_per_s": 303.43826525958417, "data_frac": 0.019076182728704243, "mem_mb": 91.210752, "nll": -5.9038166999816895, "elapsed_s": 755.1}187{"event": "train", "step": 3600, "loss": -6.09845495223999, "grad_norm": 0.32665014266967773, "lr": 3.7314672370414325e-05, "ex_per_s": 304.45870654547235, "data_frac": 0.019049918467141234, "mem_mb": 91.210752, "nll": -6.09845495223999, "elapsed_s": 756.8}188{"event": "train", "step": 3620, "loss": -5.91001033782959, "grad_norm": 1.1012402772903442, "lr": 3.660737030015427e-05, "ex_per_s": 305.4747224530384, "data_frac": 0.019023462785029192, "mem_mb": 91.210752, "nll": -5.91001033782959, "elapsed_s": 758.5}189{"event": "train", "step": 3640, "loss": -6.054691314697266, "grad_norm": 0.3974912762641907, "lr": 3.5935169492861716e-05, "ex_per_s": 306.4862621313659, "data_frac": 0.018997095470675122, "mem_mb": 91.210752, "nll": -6.054691314697266, "elapsed_s": 760.1}190{"event": "train", "step": 3660, "loss": -5.831389904022217, "grad_norm": 0.5044687390327454, "lr": 3.52982537215255e-05, "ex_per_s": 307.49180868775403, "data_frac": 0.018972175257571406, "mem_mb": 91.210752, "nll": -5.831389904022217, "elapsed_s": 761.8}191{"event": "train", "step": 3680, "loss": -5.992786884307861, "grad_norm": 0.11707262694835663, "lr": 3.469679711255794e-05, "ex_per_s": 308.4951899276708, "data_frac": 0.01894649354289652, "mem_mb": 91.210752, "nll": -5.992786884307861, "elapsed_s": 763.5}192{"event": "train", "step": 3700, "loss": -6.065377712249756, "grad_norm": 0.832677960395813, "lr": 3.41309640981904e-05, "ex_per_s": 309.4946317523685, "data_frac": 0.018920337346171472, "mem_mb": 91.210752, "nll": -6.065377712249756, "elapsed_s": 765.2}193{"event": "train", "step": 3720, "loss": -5.947065353393555, "grad_norm": 0.6132579445838928, "lr": 3.3600909371519225e-05, "ex_per_s": 310.48968063056054, "data_frac": 0.018894820103615092, "mem_mb": 91.210752, "nll": -5.947065353393555, "elapsed_s": 766.8}194{"event": "train", "step": 3740, "loss": -6.190441608428955, "grad_norm": 0.0996050238609314, "lr": 3.31067778442141e-05, "ex_per_s": 311.4809826518084, "data_frac": 0.01886929314319714, "mem_mb": 91.210752, "nll": -6.190441608428955, "elapsed_s": 768.5}195{"event": "train", "step": 3760, "loss": -6.1484575271606445, "grad_norm": 0.380490779876709, "lr": 3.2648704606900735e-05, "ex_per_s": 312.4675848683724, "data_frac": 0.018844038585676223, "mem_mb": 91.210752, "nll": -6.1484575271606445, "elapsed_s": 770.2}196{"event": "train", "step": 3780, "loss": -5.964290142059326, "grad_norm": 0.3559935688972473, "lr": 3.222681489222838e-05, "ex_per_s": 313.44990574903835, "data_frac": 0.018819010938400143, "mem_mb": 91.210752, "nll": -5.964290142059326, "elapsed_s": 771.8}197{"event": "train", "step": 3800, "loss": -6.097623348236084, "grad_norm": 0.10870803147554398, "lr": 3.1841224040632484e-05, "ex_per_s": 314.42774176081383, "data_frac": 0.018793954177340505, "mem_mb": 91.210752, "nll": -6.097623348236084, "elapsed_s": 773.5}198{"event": "train", "step": 3820, "loss": -5.969527244567871, "grad_norm": 0.1228601410984993, "lr": 3.149203746880175e-05, "ex_per_s": 315.40275904839456, "data_frac": 0.018768923967454814, "mem_mb": 91.210752, "nll": -5.969527244567871, "elapsed_s": 775.2}199{"event": "train", "step": 3840, "loss": -5.872769832611084, "grad_norm": 0.6533001661300659, "lr": 3.117935064085835e-05, "ex_per_s": 316.37289185291775, "data_frac": 0.018744436430762974, "mem_mb": 91.210752, "nll": -5.872769832611084, "elapsed_s": 776.8}200{"event": "train", "step": 3860, "loss": -6.0154571533203125, "grad_norm": 0.33838358521461487, "lr": 3.090324904225885e-05, "ex_per_s": 317.3393442742233, "data_frac": 0.018720074205108433, "mem_mb": 91.210752, "nll": -6.0154571533203125, "elapsed_s": 778.5}201{"event": "train", "step": 3880, "loss": -5.890719413757324, "grad_norm": 0.17937645316123962, "lr": 3.0663808156423456e-05, "ex_per_s": 318.3005054790942, "data_frac": 0.018696044200126244, "mem_mb": 91.210752, "nll": -5.890719413757324, "elapsed_s": 780.2}202{"event": "train", "step": 3900, "loss": -6.180943965911865, "grad_norm": 0.6863453388214111, "lr": 3.046109344409957e-05, "ex_per_s": 319.25796055107423, "data_frac": 0.01867240020385714, "mem_mb": 91.210752, "nll": -6.180943965911865, "elapsed_s": 781.9}203{"event": "train", "step": 3920, "loss": -6.037132740020752, "grad_norm": 0.21645264327526093, "lr": 3.0295160325465435e-05, "ex_per_s": 320.21372462964644, "data_frac": 0.018648121882351293, "mem_mb": 91.210752, "nll": -6.037132740020752, "elapsed_s": 783.5}204{"event": "train", "step": 3940, "loss": -5.919177532196045, "grad_norm": 0.3821324408054352, "lr": 3.016605416497886e-05, "ex_per_s": 321.16638544026614, "data_frac": 0.01862384622767353, "mem_mb": 91.210752, "nll": -5.919177532196045, "elapsed_s": 785.2}205{"event": "train", "step": 3960, "loss": -6.069334506988525, "grad_norm": 0.15899649262428284, "lr": 3.0073810258974985e-05, "ex_per_s": 322.1135877569718, "data_frac": 0.018600513544261454, "mem_mb": 91.210752, "nll": -6.069334506988525, "elapsed_s": 786.8}206{"event": "train", "step": 3980, "loss": -5.802150726318359, "grad_norm": 0.123723104596138, "lr": 3.0018453826016683e-05, "ex_per_s": 323.05718836732154, "data_frac": 0.01857678888243268, "mem_mb": 91.210752, "nll": -5.802150726318359, "elapsed_s": 788.5}207{"event": "train", "step": 4000, "loss": -6.055910110473633, "grad_norm": 0.865119993686676, "lr": 2.9999999999999997e-05, "ex_per_s": 323.99667512276415, "data_frac": 0.01855320957451825, "mem_mb": 91.210752, "nll": -6.055910110473633, "elapsed_s": 790.2}208{"event": "val", "step": 4000, "val_nll": -6.449095020952024, "best": -6.449357611608497, "bad": 1, "mem_mb": 91.210752, "elapsed_s": 853.1}209{"event": "done", "steps": 4000, "best_val": -6.449357611608497, "wall_s": 853.1240651607513, "ex_per_s": 300.07358888857834, "peak_mem_mb": 91.210752, "n_params": 3455619, "device": "mps", "data_time_frac": 0.017183288008490002, "elapsed_s": 853.1}210
Files · main
Menu

Theme