SPB Git

spb/wp41_uqo

Public
51.4 KB · 210 lines
1{"event": "train", "step": 20, "loss": -6.112734317779541, "grad_norm": 1.3628172874450684, "lr": 3.149999999999999e-05, "ex_per_s": 98.04042599804818, "data_frac": 0.8422772600889366, "mem_mb": 62.975488, "nll": -6.112734317779541, "elapsed_s": 13.1}2{"event": "train", "step": 40, "loss": -5.899176597595215, "grad_norm": 0.22842998802661896, "lr": 6.149999999999999e-05, "ex_per_s": 173.828986590742, "data_frac": 0.7474828804463525, "mem_mb": 62.975488, "nll": -5.899176597595215, "elapsed_s": 14.8}3{"event": "train", "step": 60, "loss": -5.982251167297363, "grad_norm": 0.5177673101425171, "lr": 9.149999999999999e-05, "ex_per_s": 234.26093238371865, "data_frac": 0.6722298180188314, "mem_mb": 62.975488, "nll": -5.982251167297363, "elapsed_s": 16.4}4{"event": "train", "step": 80, "loss": -6.077022075653076, "grad_norm": 0.9368844032287598, "lr": 0.0001215, "ex_per_s": 283.5272687381166, "data_frac": 0.6108418194010097, "mem_mb": 62.975488, "nll": -6.077022075653076, "elapsed_s": 18.1}5{"event": "train", "step": 100, "loss": -5.951207160949707, "grad_norm": 0.2255427986383438, "lr": 0.0001515, "ex_per_s": 324.53376949875695, "data_frac": 0.5599082807842036, "mem_mb": 62.975488, "nll": -5.951207160949707, "elapsed_s": 19.8}6{"event": "train", "step": 120, "loss": -5.911109924316406, "grad_norm": 0.8060075640678406, "lr": 0.00018149999999999997, "ex_per_s": 359.1939929840418, "data_frac": 0.516927875493722, "mem_mb": 62.975488, "nll": -5.911109924316406, "elapsed_s": 21.4}7{"event": "train", "step": 140, "loss": -5.830454349517822, "grad_norm": 0.6625245213508606, "lr": 0.00021149999999999996, "ex_per_s": 388.81367506426795, "data_frac": 0.4800748448163165, "mem_mb": 62.975488, "nll": -5.830454349517822, "elapsed_s": 23.1}8{"event": "train", "step": 160, "loss": -5.73636531829834, "grad_norm": 0.7255787253379822, "lr": 0.0002415, "ex_per_s": 414.5158930946589, "data_frac": 0.4482511426828036, "mem_mb": 62.975488, "nll": -5.73636531829834, "elapsed_s": 24.7}9{"event": "train", "step": 180, "loss": -5.975525379180908, "grad_norm": 1.0272362232208252, "lr": 0.0002715, "ex_per_s": 437.0310872737885, "data_frac": 0.4204878929878357, "mem_mb": 62.975488, "nll": -5.975525379180908, "elapsed_s": 26.4}10{"event": "train", "step": 200, "loss": -5.972040176391602, "grad_norm": 0.483538955450058, "lr": 0.0003, "ex_per_s": 456.84897196178974, "data_frac": 0.3959781412004453, "mem_mb": 62.975488, "nll": -5.972040176391602, "elapsed_s": 28.1}11{"event": "train", "step": 220, "loss": -6.0312819480896, "grad_norm": 1.1153539419174194, "lr": 0.00029998154617398326, "ex_per_s": 474.4042727890345, "data_frac": 0.37418714534232494, "mem_mb": 62.975488, "nll": -6.0312819480896, "elapsed_s": 29.7}12{"event": "train", "step": 240, "loss": -5.914257049560547, "grad_norm": 0.16576826572418213, "lr": 0.000299926189741025, "ex_per_s": 490.1285527401282, "data_frac": 0.3547341255453478, "mem_mb": 62.975488, "nll": -5.914257049560547, "elapsed_s": 31.4}13{"event": "train", "step": 260, "loss": -5.887704372406006, "grad_norm": 0.16190165281295776, "lr": 0.0002998339458350211, "ex_per_s": 504.29623423983463, "data_frac": 0.33723322259631683, "mem_mb": 62.975488, "nll": -5.887704372406006, "elapsed_s": 33.0}14{"event": "train", "step": 280, "loss": -5.944139003753662, "grad_norm": 0.13190089166164398, "lr": 0.0002997048396745345, "ex_per_s": 517.0835217482049, "data_frac": 0.321394796063458, "mem_mb": 62.975488, "nll": -5.944139003753662, "elapsed_s": 34.7}15{"event": "train", "step": 300, "loss": -5.968145370483398, "grad_norm": 0.2178327739238739, "lr": 0.0002995389065559004, "ex_per_s": 528.7053851240971, "data_frac": 0.3070035275262114, "mem_mb": 62.975488, "nll": -5.968145370483398, "elapsed_s": 36.4}16{"event": "train", "step": 320, "loss": -5.9657745361328125, "grad_norm": 0.6869058012962341, "lr": 0.0002993361918435765, "ex_per_s": 539.3318122720972, "data_frac": 0.29387894354779726, "mem_mb": 62.975488, "nll": -5.9657745361328125, "elapsed_s": 38.0}17{"event": "train", "step": 340, "loss": -5.834891319274902, "grad_norm": 0.3371953070163727, "lr": 0.00029909675095774114, "ex_per_s": 549.0727886300491, "data_frac": 0.28185374364866506, "mem_mb": 62.975488, "nll": -5.834891319274902, "elapsed_s": 39.7}18{"event": "train", "step": 360, "loss": -5.995591163635254, "grad_norm": 0.4202992618083954, "lr": 0.00029882064935914164, "ex_per_s": 558.0226295685143, "data_frac": 0.27078700721368876, "mem_mb": 62.975488, "nll": -5.995591163635254, "elapsed_s": 41.3}19{"event": "train", "step": 380, "loss": -6.005503177642822, "grad_norm": 1.1860206127166748, "lr": 0.00029850796253119824, "ex_per_s": 566.2581796462491, "data_frac": 0.2605705708743058, "mem_mb": 62.975488, "nll": -6.005503177642822, "elapsed_s": 43.0}20{"event": "train", "step": 400, "loss": -5.9963884353637695, "grad_norm": 1.3682140111923218, "lr": 0.0002981587759593675, "ex_per_s": 573.9015232753194, "data_frac": 0.2511220628538634, "mem_mb": 62.975488, "nll": -5.9963884353637695, "elapsed_s": 44.7}21{"event": "train", "step": 420, "loss": -5.96451473236084, "grad_norm": 0.6703805923461914, "lr": 0.0002977731851077716, "ex_per_s": 580.9860905324715, "data_frac": 0.24234218107519792, "mem_mb": 62.975488, "nll": -5.96451473236084, "elapsed_s": 46.3}22{"event": "train", "step": 440, "loss": -5.938429832458496, "grad_norm": 0.7174604535102844, "lr": 0.0002973512953930992, "ex_per_s": 587.5528904974395, "data_frac": 0.2341729730576749, "mem_mb": 62.975488, "nll": -5.938429832458496, "elapsed_s": 48.0}23{"event": "train", "step": 460, "loss": -6.10232400894165, "grad_norm": 0.6422565579414368, "lr": 0.0002968932221557859, "ex_per_s": 593.6689298322731, "data_frac": 0.22654124526899624, "mem_mb": 62.975488, "nll": -6.10232400894165, "elapsed_s": 49.6}24{"event": "train", "step": 480, "loss": -5.877755641937256, "grad_norm": 0.31036075949668884, "lr": 0.0002963990906284808, "ex_per_s": 599.3984723275087, "data_frac": 0.2194167358612384, "mem_mb": 62.975488, "nll": -5.877755641937256, "elapsed_s": 51.3}25{"event": "train", "step": 500, "loss": -5.924736976623535, "grad_norm": 0.5752387046813965, "lr": 0.00029586903590180956, "ex_per_s": 604.7880540666782, "data_frac": 0.2127320990046963, "mem_mb": 62.975488, "nll": -5.924736976623535, "elapsed_s": 53.0}26{"event": "val", "step": 500, "val_nll": -6.4395459099059345, "best": -6.4395459099059345, "bad": 0, "mem_mb": 62.975488, "elapsed_s": 115.0}27{"event": "train", "step": 520, "loss": -6.039616584777832, "grad_norm": 0.33539721369743347, "lr": 0.00029530320288744205, "ex_per_s": 285.1907422715841, "data_frac": 0.09654722059746, "mem_mb": 62.975488, "nll": -6.039616584777832, "elapsed_s": 116.7}28{"event": "train", "step": 540, "loss": -5.87980318069458, "grad_norm": 0.5404672026634216, "lr": 0.0002947017462784745, "ex_per_s": 291.98146651651246, "data_frac": 0.09527667624627305, "mem_mb": 62.975488, "nll": -5.87980318069458, "elapsed_s": 118.4}29{"event": "train", "step": 560, "loss": -5.972707748413086, "grad_norm": 0.172672301530838, "lr": 0.00029406483050713824, "ex_per_s": 298.5934018505646, "data_frac": 0.0940408314058063, "mem_mb": 62.975488, "nll": -5.972707748413086, "elapsed_s": 120.1}30{"event": "train", "step": 580, "loss": -5.925030708312988, "grad_norm": 0.10883055627346039, "lr": 0.0002933926296998457, "ex_per_s": 305.02463389423315, "data_frac": 0.09284093781400904, "mem_mb": 62.975488, "nll": -5.925030708312988, "elapsed_s": 121.7}31{"event": "train", "step": 600, "loss": -5.973895072937012, "grad_norm": 0.681222677230835, "lr": 0.0002926853276295856, "ex_per_s": 311.2853717312161, "data_frac": 0.09167439802178687, "mem_mb": 62.975488, "nll": -5.973895072937012, "elapsed_s": 123.4}32{"event": "train", "step": 620, "loss": -5.952948093414307, "grad_norm": 0.44103699922561646, "lr": 0.00029194311766568174, "ex_per_s": 317.3774722407099, "data_frac": 0.09053813769279753, "mem_mb": 62.975488, "nll": -5.952948093414307, "elapsed_s": 125.1}33{"event": "train", "step": 640, "loss": -6.06510591506958, "grad_norm": 0.4743615686893463, "lr": 0.0002911662027209272, "ex_per_s": 323.31373827011225, "data_frac": 0.08943288556304207, "mem_mb": 62.975488, "nll": -6.06510591506958, "elapsed_s": 126.7}34{"event": "train", "step": 660, "loss": -6.006404876708984, "grad_norm": 1.0698765516281128, "lr": 0.00029035479519611045, "ex_per_s": 329.09534273235147, "data_frac": 0.08835587833274618, "mem_mb": 62.975488, "nll": -6.006404876708984, "elapsed_s": 128.4}35{"event": "train", "step": 680, "loss": -6.007498264312744, "grad_norm": 0.10287109017372131, "lr": 0.0002895091169219469, "ex_per_s": 334.73496653956187, "data_frac": 0.08730736150544412, "mem_mb": 62.975488, "nll": -6.007498264312744, "elapsed_s": 130.1}36{"event": "train", "step": 700, "loss": -5.987902641296387, "grad_norm": 0.6530570387840271, "lr": 0.00028862939909843273, "ex_per_s": 340.22661589555923, "data_frac": 0.08629063856526042, "mem_mb": 62.975488, "nll": -5.987902641296387, "elapsed_s": 131.7}37{"event": "train", "step": 720, "loss": -6.063712120056152, "grad_norm": 0.3255457282066345, "lr": 0.0002877158822316369, "ex_per_s": 345.59052743554435, "data_frac": 0.08529575502241227, "mem_mb": 62.975488, "nll": -6.063712120056152, "elapsed_s": 133.4}38{"event": "train", "step": 740, "loss": -6.033452987670898, "grad_norm": 0.2814691364765167, "lr": 0.00028676881606794945, "ex_per_s": 350.8206201998592, "data_frac": 0.08432524491688917, "mem_mb": 62.975488, "nll": -6.033452987670898, "elapsed_s": 135.0}39{"event": "train", "step": 760, "loss": -5.957125186920166, "grad_norm": 1.0631072521209717, "lr": 0.000285788459525803, "ex_per_s": 355.9241920962645, "data_frac": 0.08337929612101672, "mem_mb": 62.975488, "nll": -5.957125186920166, "elapsed_s": 136.7}40{"event": "train", "step": 780, "loss": -5.895751953125, "grad_norm": 0.14292125403881073, "lr": 0.00028477508062488754, "ex_per_s": 360.9040144358258, "data_frac": 0.08246017532793427, "mem_mb": 62.975488, "nll": -5.895751953125, "elapsed_s": 138.4}41{"event": "train", "step": 800, "loss": -5.816133499145508, "grad_norm": 0.1872439682483673, "lr": 0.000283728956412876, "ex_per_s": 365.7691907193776, "data_frac": 0.08155879619893373, "mem_mb": 62.975488, "nll": -5.816133499145508, "elapsed_s": 140.0}42{"event": "train", "step": 820, "loss": -5.933557033538818, "grad_norm": 0.6406973600387573, "lr": 0.00028265037288968236, "ex_per_s": 370.52386121389463, "data_frac": 0.08067763884122849, "mem_mb": 62.975488, "nll": -5.933557033538818, "elapsed_s": 141.7}43{"event": "train", "step": 840, "loss": -5.808857440948486, "grad_norm": 0.4880909323692322, "lr": 0.0002815396249292725, "ex_per_s": 375.17060655415025, "data_frac": 0.07981895690584603, "mem_mb": 62.975488, "nll": -5.808857440948486, "elapsed_s": 143.3}44{"event": "train", "step": 860, "loss": -5.915006160736084, "grad_norm": 1.4499086141586304, "lr": 0.00028039701619904774, "ex_per_s": 379.7110298869712, "data_frac": 0.07897988045064165, "mem_mb": 62.975488, "nll": -5.915006160736084, "elapsed_s": 145.0}45{"event": "train", "step": 880, "loss": -5.871411323547363, "grad_norm": 0.19515961408615112, "lr": 0.00027922285907682625, "ex_per_s": 384.15298853428317, "data_frac": 0.07816017599100818, "mem_mb": 62.975488, "nll": -5.871411323547363, "elapsed_s": 146.7}46{"event": "train", "step": 900, "loss": -5.988950729370117, "grad_norm": 0.24830244481563568, "lr": 0.00027801747456544134, "ex_per_s": 388.4933979969443, "data_frac": 0.07735874440531276, "mem_mb": 62.975488, "nll": -5.988950729370117, "elapsed_s": 148.3}47{"event": "train", "step": 920, "loss": -6.025753021240234, "grad_norm": 0.5479169487953186, "lr": 0.00027678119220498265, "ex_per_s": 392.7363285286234, "data_frac": 0.07657373736235162, "mem_mb": 62.975488, "nll": -6.025753021240234, "elapsed_s": 150.0}48{"event": "train", "step": 940, "loss": -6.08470344543457, "grad_norm": 0.38988247513771057, "lr": 0.0002755143499827035, "ex_per_s": 396.87652646108734, "data_frac": 0.07580967057243354, "mem_mb": 62.975488, "nll": -6.08470344543457, "elapsed_s": 151.6}49{"event": "train", "step": 960, "loss": -5.968710422515869, "grad_norm": 0.10873299837112427, "lr": 0.00027421729424061787, "ex_per_s": 400.93715418615415, "data_frac": 0.07505983258560396, "mem_mb": 62.975488, "nll": -5.968710422515869, "elapsed_s": 153.3}50{"event": "train", "step": 980, "loss": -5.92686128616333, "grad_norm": 0.2604270279407501, "lr": 0.00027289037958081483, "ex_per_s": 404.91410536921603, "data_frac": 0.07432678268706189, "mem_mb": 62.975488, "nll": -5.92686128616333, "elapsed_s": 154.9}51{"event": "train", "step": 1000, "loss": -6.072515487670898, "grad_norm": 0.36324992775917053, "lr": 0.0002715339687685131, "ex_per_s": 408.80178606824154, "data_frac": 0.07360892847473496, "mem_mb": 62.975488, "nll": -6.072515487670898, "elapsed_s": 156.6}52{"event": "val", "step": 1000, "val_nll": -6.446911307609957, "best": -6.446911307609957, "bad": 0, "mem_mb": 62.975488, "elapsed_s": 218.6}53{"event": "train", "step": 1020, "loss": -6.031801223754883, "grad_norm": 0.20013369619846344, "lr": 0.0002701484326328852, "ex_per_s": 296.3299134423317, "data_frac": 0.05235918833176961, "mem_mb": 62.975488, "nll": -6.031801223754883, "elapsed_s": 220.3}54{"event": "train", "step": 1040, "loss": -5.97294282913208, "grad_norm": 0.197512686252594, "lr": 0.00026873414996567596, "ex_per_s": 299.8711234915973, "data_frac": 0.05201408890176293, "mem_mb": 62.975488, "nll": -5.97294282913208, "elapsed_s": 222.0}55{"event": "train", "step": 1060, "loss": -6.183176517486572, "grad_norm": 0.17014306783676147, "lr": 0.0002672915074176452, "ex_per_s": 303.35764494896193, "data_frac": 0.051672928089342485, "mem_mb": 62.975488, "nll": -6.183176517486572, "elapsed_s": 223.7}56{"event": "train", "step": 1080, "loss": -5.814983367919922, "grad_norm": 0.5223254561424255, "lr": 0.00026582089939286075, "ex_per_s": 306.7954036915853, "data_frac": 0.05133798688928293, "mem_mb": 62.975488, "nll": -5.814983367919922, "elapsed_s": 225.3}57{"event": "train", "step": 1100, "loss": -6.053898811340332, "grad_norm": 0.15549694001674652, "lr": 0.0002643227279408728, "ex_per_s": 310.17998120856345, "data_frac": 0.05100709907293762, "mem_mb": 62.975488, "nll": -6.053898811340332, "elapsed_s": 227.0}58{"event": "train", "step": 1120, "loss": -6.051617622375488, "grad_norm": 0.43184995651245117, "lr": 0.000262797402646797, "ex_per_s": 313.5153082290957, "data_frac": 0.05068320078523309, "mem_mb": 62.975488, "nll": -6.051617622375488, "elapsed_s": 228.7}59{"event": "train", "step": 1140, "loss": -6.018906593322754, "grad_norm": 0.3241274952888489, "lr": 0.0002612453405193386, "ex_per_s": 316.80307902156414, "data_frac": 0.05036217381312443, "mem_mb": 62.975488, "nll": -6.018906593322754, "elapsed_s": 230.3}60{"event": "train", "step": 1160, "loss": -5.970714092254639, "grad_norm": 0.6012495756149292, "lr": 0.0002596669658767859, "ex_per_s": 320.03987118448055, "data_frac": 0.05004919023681718, "mem_mb": 62.975488, "nll": -5.970714092254639, "elapsed_s": 232.0}61{"event": "train", "step": 1180, "loss": -5.95796012878418, "grad_norm": 0.10148540884256363, "lr": 0.00025806271023100637, "ex_per_s": 323.23242302974626, "data_frac": 0.04973613410526335, "mem_mb": 62.975488, "nll": -5.95796012878418, "elapsed_s": 233.7}62{"event": "train", "step": 1200, "loss": -5.944573402404785, "grad_norm": 0.2257121354341507, "lr": 0.00025643301216947504, "ex_per_s": 326.3800568925831, "data_frac": 0.04942793349959023, "mem_mb": 62.975488, "nll": -5.944573402404785, "elapsed_s": 235.4}63{"event": "train", "step": 1220, "loss": -6.033927917480469, "grad_norm": 0.5133039355278015, "lr": 0.0002547783172353695, "ex_per_s": 329.48551948539745, "data_frac": 0.04912510848084747, "mem_mb": 62.975488, "nll": -6.033927917480469, "elapsed_s": 237.0}64{"event": "train", "step": 1240, "loss": -6.0593414306640625, "grad_norm": 0.07659000903367996, "lr": 0.00025309907780576277, "ex_per_s": 332.54589446524767, "data_frac": 0.048826374223822556, "mem_mb": 62.975488, "nll": -6.0593414306640625, "elapsed_s": 238.7}65{"event": "train", "step": 1260, "loss": -6.053786754608154, "grad_norm": 0.192353293299675, "lr": 0.0002513957529679477, "ex_per_s": 335.56575259793453, "data_frac": 0.04853201740060741, "mem_mb": 62.975488, "nll": -6.053786754608154, "elapsed_s": 240.4}66{"event": "train", "step": 1280, "loss": -6.043435573577881, "grad_norm": 0.6000996828079224, "lr": 0.00024966880839392746, "ex_per_s": 338.5447598849411, "data_frac": 0.04824179159637363, "mem_mb": 62.975488, "nll": -6.043435573577881, "elapsed_s": 242.0}67{"event": "train", "step": 1300, "loss": -5.868824481964111, "grad_norm": 0.5854211449623108, "lr": 0.0002479187162131052, "ex_per_s": 341.4845965838924, "data_frac": 0.04795589882504351, "mem_mb": 62.975488, "nll": -5.868824481964111, "elapsed_s": 243.7}68{"event": "train", "step": 1320, "loss": -5.8291401863098145, "grad_norm": 0.18564985692501068, "lr": 0.00024614595488320833, "ex_per_s": 344.38340703980805, "data_frac": 0.047673319044124786, "mem_mb": 62.975488, "nll": -5.8291401863098145, "elapsed_s": 245.4}69{"event": "train", "step": 1340, "loss": -5.95127010345459, "grad_norm": 0.37007787823677063, "lr": 0.00024435100905948387, "ex_per_s": 347.2409107868546, "data_frac": 0.04739433634866822, "mem_mb": 62.975488, "nll": -5.95127010345459, "elapsed_s": 247.0}70{"event": "train", "step": 1360, "loss": -6.039646625518799, "grad_norm": 0.6833720803260803, "lr": 0.0002425343694621974, "ex_per_s": 350.06129330097116, "data_frac": 0.04711953514022373, "mem_mb": 62.975488, "nll": -6.039646625518799, "elapsed_s": 248.7}71{"event": "train", "step": 1380, "loss": -6.070425510406494, "grad_norm": 0.3258994221687317, "lr": 0.00024069653274247583, "ex_per_s": 352.8442670631858, "data_frac": 0.04684856191314029, "mem_mb": 62.975488, "nll": -6.070425510406494, "elapsed_s": 250.4}72{"event": "train", "step": 1400, "loss": -5.7964348793029785, "grad_norm": 0.13125009834766388, "lr": 0.0002388380013465276, "ex_per_s": 355.58811028293934, "data_frac": 0.04658434488215258, "mem_mb": 62.975488, "nll": -5.7964348793029785, "elapsed_s": 252.0}73{"event": "train", "step": 1420, "loss": -5.988722801208496, "grad_norm": 0.7782344222068787, "lr": 0.00023695928337827914, "ex_per_s": 358.2996611797573, "data_frac": 0.046320420188457584, "mem_mb": 62.975488, "nll": -5.988722801208496, "elapsed_s": 253.7}74{"event": "train", "step": 1440, "loss": -5.741990566253662, "grad_norm": 0.28812554478645325, "lr": 0.00023506089246046403, "ex_per_s": 360.9765475827486, "data_frac": 0.046059072758799245, "mem_mb": 62.975488, "nll": -5.741990566253662, "elapsed_s": 255.4}75{"event": "train", "step": 1460, "loss": -5.803056716918945, "grad_norm": 0.4762240946292877, "lr": 0.00023314334759420388, "ex_per_s": 363.6201624756163, "data_frac": 0.04580136384938356, "mem_mb": 62.975488, "nll": -5.803056716918945, "elapsed_s": 257.0}76{"event": "train", "step": 1480, "loss": -6.047018051147461, "grad_norm": 0.09358112514019012, "lr": 0.0002312071730171186, "ex_per_s": 366.22275840550117, "data_frac": 0.04554888039050912, "mem_mb": 62.975488, "nll": -6.047018051147461, "elapsed_s": 258.7}77{"event": "train", "step": 1500, "loss": -5.958409786224365, "grad_norm": 0.30205774307250977, "lr": 0.0002292528980600049, "ex_per_s": 368.7999004404766, "data_frac": 0.045299562084699874, "mem_mb": 62.975488, "nll": -5.958409786224365, "elapsed_s": 260.3}78{"event": "val", "step": 1500, "val_nll": -6.446680243353936, "best": -6.446911307609957, "bad": 1, "mem_mb": 62.975488, "elapsed_s": 322.4}79{"event": "train", "step": 1520, "loss": -5.9507856369018555, "grad_norm": 0.5848132371902466, "lr": 0.00022728105700212285, "ex_per_s": 300.19568111929783, "data_frac": 0.03642069864117683, "mem_mb": 62.975488, "nll": -5.9507856369018555, "elapsed_s": 324.1}80{"event": "train", "step": 1540, "loss": -5.879823684692383, "grad_norm": 0.09802502393722534, "lr": 0.00022529218892512886, "ex_per_s": 302.5864346015158, "data_frac": 0.03626710638174305, "mem_mb": 62.975488, "nll": -5.879823684692383, "elapsed_s": 325.8}81{"event": "train", "step": 1560, "loss": -5.983138084411621, "grad_norm": 0.36058902740478516, "lr": 0.00022328683756569652, "ex_per_s": 304.9534342709791, "data_frac": 0.03611471018962402, "mem_mb": 62.975488, "nll": -5.983138084411621, "elapsed_s": 327.4}82{"event": "train", "step": 1580, "loss": -5.981291770935059, "grad_norm": 0.07129236310720444, "lr": 0.0002212655511668642, "ex_per_s": 307.29648027657646, "data_frac": 0.0359639299615821, "mem_mb": 62.975488, "nll": -5.981291770935059, "elapsed_s": 329.1}83{"event": "train", "step": 1600, "loss": -5.957001209259033, "grad_norm": 0.22638897597789764, "lr": 0.00021922888232815089, "ex_per_s": 309.6158024737466, "data_frac": 0.035814262823879936, "mem_mb": 62.975488, "nll": -5.957001209259033, "elapsed_s": 330.8}84{"event": "train", "step": 1620, "loss": -5.881519317626953, "grad_norm": 0.08632777631282806, "lr": 0.00021717738785448102, "ex_per_s": 311.9102889633801, "data_frac": 0.03566816380800474, "mem_mb": 62.975488, "nll": -5.881519317626953, "elapsed_s": 332.4}85{"event": "train", "step": 1640, "loss": -6.076923370361328, "grad_norm": 0.11934700608253479, "lr": 0.00021511162860395977, "ex_per_s": 314.18270704216764, "data_frac": 0.03552246191749198, "mem_mb": 62.975488, "nll": -6.076923370361328, "elapsed_s": 334.1}86{"event": "train", "step": 1660, "loss": -6.04171895980835, "grad_norm": 0.6585996150970459, "lr": 0.0002130321693345397, "ex_per_s": 316.4310494542095, "data_frac": 0.03537712880224468, "mem_mb": 62.975488, "nll": -6.04171895980835, "elapsed_s": 335.8}87{"event": "train", "step": 1680, "loss": -5.8646769523620605, "grad_norm": 0.155347540974617, "lr": 0.0002109395785496218, "ex_per_s": 318.6613588243107, "data_frac": 0.03523432704646835, "mem_mb": 62.975488, "nll": -5.8646769523620605, "elapsed_s": 337.5}88{"event": "train", "step": 1700, "loss": -5.927036762237549, "grad_norm": 0.5905569195747375, "lr": 0.00020883442834263225, "ex_per_s": 320.8680993229987, "data_frac": 0.0350917332155673, "mem_mb": 62.975488, "nll": -5.927036762237549, "elapsed_s": 339.1}89{"event": "train", "step": 1720, "loss": -5.888317108154297, "grad_norm": 0.8120751976966858, "lr": 0.00020671729424061788, "ex_per_s": 323.053229747005, "data_frac": 0.03495127445237788, "mem_mb": 62.975488, "nll": -5.888317108154297, "elapsed_s": 340.8}90{"event": "train", "step": 1740, "loss": -6.274004936218262, "grad_norm": 0.34550929069519043, "lr": 0.00020458875504690275, "ex_per_s": 325.21772118331427, "data_frac": 0.034812927695522564, "mem_mb": 62.975488, "nll": -6.274004936218262, "elapsed_s": 342.5}91{"event": "train", "step": 1760, "loss": -5.931365489959717, "grad_norm": 0.3549400269985199, "lr": 0.000202449392682849, "ex_per_s": 327.361635070022, "data_frac": 0.03467555863903685, "mem_mb": 62.975488, "nll": -5.931365489959717, "elapsed_s": 344.1}92{"event": "train", "step": 1780, "loss": -5.871510982513428, "grad_norm": 0.540438175201416, "lr": 0.00020029979202876563, "ex_per_s": 329.4855904172358, "data_frac": 0.034538959292090506, "mem_mb": 62.975488, "nll": -5.871510982513428, "elapsed_s": 345.8}93{"event": "train", "step": 1800, "loss": -6.047980785369873, "grad_norm": 1.030329942703247, "lr": 0.00019814054076400787, "ex_per_s": 331.59096107701293, "data_frac": 0.03440419798166464, "mem_mb": 62.975488, "nll": -6.047980785369873, "elapsed_s": 347.5}94{"event": "train", "step": 1820, "loss": -5.935070514678955, "grad_norm": 0.8612695932388306, "lr": 0.00019597222920631147, "ex_per_s": 333.67303828795406, "data_frac": 0.034271060996230854, "mem_mb": 62.975488, "nll": -5.935070514678955, "elapsed_s": 349.1}95{"event": "train", "step": 1840, "loss": -6.0724592208862305, "grad_norm": 0.3367674946784973, "lr": 0.0001937954501504058, "ex_per_s": 335.7392709307063, "data_frac": 0.03413901751290413, "mem_mb": 62.975488, "nll": -6.0724592208862305, "elapsed_s": 350.8}96{"event": "train", "step": 1860, "loss": -6.050938606262207, "grad_norm": 0.21909049153327942, "lr": 0.0001916107987059498, "ex_per_s": 337.78210030907616, "data_frac": 0.03401187510309183, "mem_mb": 62.975488, "nll": -6.050938606262207, "elapsed_s": 352.5}97{"event": "train", "step": 1880, "loss": -5.949497699737549, "grad_norm": 0.43559423089027405, "lr": 0.00018941887213483484, "ex_per_s": 339.8106174489008, "data_frac": 0.03388178964088489, "mem_mb": 62.975488, "nll": -5.949497699737549, "elapsed_s": 354.1}98{"event": "train", "step": 1900, "loss": -5.981949806213379, "grad_norm": 0.2535601556301117, "lr": 0.00018722026968789906, "ex_per_s": 341.8213150145473, "data_frac": 0.033753139521472124, "mem_mb": 62.975488, "nll": -5.981949806213379, "elapsed_s": 355.8}99{"event": "train", "step": 1920, "loss": -5.963447570800781, "grad_norm": 0.4147723615169525, "lr": 0.00018501559244109864, "ex_per_s": 343.81215992739317, "data_frac": 0.03362627735495172, "mem_mb": 62.975488, "nll": -5.963447570800781, "elapsed_s": 357.4}100{"event": "train", "step": 1940, "loss": -6.049401760101318, "grad_norm": 0.22874698042869568, "lr": 0.00018280544313117932, "ex_per_s": 345.7815258491124, "data_frac": 0.033500863085029024, "mem_mb": 62.975488, "nll": -6.049401760101318, "elapsed_s": 359.1}101{"event": "train", "step": 1960, "loss": -5.939452171325684, "grad_norm": 0.0760868489742279, "lr": 0.00018059042599089415, "ex_per_s": 347.7314095195021, "data_frac": 0.033375669311104625, "mem_mb": 62.975488, "nll": -5.939452171325684, "elapsed_s": 360.8}102{"event": "train", "step": 1980, "loss": -5.964908123016357, "grad_norm": 0.5904769897460938, "lr": 0.00017837114658381247, "ex_per_s": 349.66549524233284, "data_frac": 0.03325124817992298, "mem_mb": 62.975488, "nll": -5.964908123016357, "elapsed_s": 362.4}103{"event": "train", "step": 2000, "loss": -5.9158806800842285, "grad_norm": 0.3984600007534027, "lr": 0.00017614821163876486, "ex_per_s": 351.5821249295198, "data_frac": 0.033128067576988586, "mem_mb": 62.975488, "nll": -5.9158806800842285, "elapsed_s": 364.1}104{"event": "val", "step": 2000, "val_nll": -6.448608993111561, "best": -6.448608993111561, "bad": 0, "mem_mb": 62.975488, "elapsed_s": 426.3}105{"event": "train", "step": 2020, "loss": -5.927382469177246, "grad_norm": 0.5840640068054199, "lr": 0.00017392222888396941, "ex_per_s": 302.0801735390752, "data_frac": 0.028207675680625047, "mem_mb": 62.975488, "nll": -5.927382469177246, "elapsed_s": 428.0}106{"event": "train", "step": 2040, "loss": -5.873814582824707, "grad_norm": 0.2995949983596802, "lr": 0.000171693806880885, "ex_per_s": 303.8801918827276, "data_frac": 0.028127600021813116, "mem_mb": 62.975488, "nll": -5.873814582824707, "elapsed_s": 429.7}107{"event": "train", "step": 2060, "loss": -6.143423080444336, "grad_norm": 0.34642481803894043, "lr": 0.0001694635548578366, "ex_per_s": 305.67085018031753, "data_frac": 0.02804893147104485, "mem_mb": 62.975488, "nll": -6.143423080444336, "elapsed_s": 431.4}108{"event": "train", "step": 2080, "loss": -5.962284088134766, "grad_norm": 0.4696488082408905, "lr": 0.0001672320825434584, "ex_per_s": 307.44605528388774, "data_frac": 0.027967293499185895, "mem_mb": 62.975488, "nll": -5.962284088134766, "elapsed_s": 433.0}109{"event": "train", "step": 2100, "loss": -5.899939060211182, "grad_norm": 0.9049482941627502, "lr": 0.000165, "ex_per_s": 309.21069079059475, "data_frac": 0.027884880761394042, "mem_mb": 62.975488, "nll": -5.899939060211182, "elapsed_s": 434.7}110{"event": "train", "step": 2120, "loss": -6.152348041534424, "grad_norm": 0.4764515161514282, "lr": 0.0001627679174565416, "ex_per_s": 310.9605483791432, "data_frac": 0.027803205606195595, "mem_mb": 62.975488, "nll": -6.152348041534424, "elapsed_s": 436.4}111{"event": "train", "step": 2140, "loss": -5.977404594421387, "grad_norm": 0.11701152473688126, "lr": 0.00016053644514216336, "ex_per_s": 312.6961549734219, "data_frac": 0.027720909984895174, "mem_mb": 62.975488, "nll": -5.977404594421387, "elapsed_s": 438.0}112{"event": "train", "step": 2160, "loss": -5.927570819854736, "grad_norm": 0.4986119270324707, "lr": 0.00015830619311911498, "ex_per_s": 314.41897437196906, "data_frac": 0.027641279360351656, "mem_mb": 62.975488, "nll": -5.927570819854736, "elapsed_s": 439.7}113{"event": "train", "step": 2180, "loss": -6.083738327026367, "grad_norm": 0.2802984416484833, "lr": 0.00015607777111603058, "ex_per_s": 316.12595545902167, "data_frac": 0.0275670636209528, "mem_mb": 62.975488, "nll": -6.083738327026367, "elapsed_s": 441.4}114{"event": "train", "step": 2200, "loss": -5.962801933288574, "grad_norm": 0.5513305068016052, "lr": 0.00015385178836123514, "ex_per_s": 317.8227601761607, "data_frac": 0.027496533157955885, "mem_mb": 62.975488, "nll": -5.962801933288574, "elapsed_s": 443.1}115{"event": "train", "step": 2220, "loss": -5.825422763824463, "grad_norm": 0.8283140659332275, "lr": 0.00015162885341618752, "ex_per_s": 319.5094185229669, "data_frac": 0.027417239824374183, "mem_mb": 62.975488, "nll": -5.825422763824463, "elapsed_s": 444.7}116{"event": "train", "step": 2240, "loss": -5.962882041931152, "grad_norm": 0.7582634687423706, "lr": 0.00014940957400910585, "ex_per_s": 321.1840832126448, "data_frac": 0.0273385597843705, "mem_mb": 62.975488, "nll": -5.962882041931152, "elapsed_s": 446.4}117{"event": "train", "step": 2260, "loss": -5.738546848297119, "grad_norm": 0.8257561326026917, "lr": 0.00014719455686882068, "ex_per_s": 322.8448279915691, "data_frac": 0.027260330685833608, "mem_mb": 62.975488, "nll": -5.738546848297119, "elapsed_s": 448.1}118{"event": "train", "step": 2280, "loss": -6.011847019195557, "grad_norm": 0.5475426912307739, "lr": 0.0001449844075589013, "ex_per_s": 324.4860914228267, "data_frac": 0.02719152147135468, "mem_mb": 62.975488, "nll": -6.011847019195557, "elapsed_s": 449.7}119{"event": "train", "step": 2300, "loss": -6.001817226409912, "grad_norm": 0.15369388461112976, "lr": 0.0001427797303121009, "ex_per_s": 326.11668919848796, "data_frac": 0.02712685376219777, "mem_mb": 62.975488, "nll": -6.001817226409912, "elapsed_s": 451.4}120{"event": "train", "step": 2320, "loss": -6.07060432434082, "grad_norm": 0.3306375741958618, "lr": 0.00014058112786516518, "ex_per_s": 327.7354002268954, "data_frac": 0.027060593533063473, "mem_mb": 62.975488, "nll": -6.07060432434082, "elapsed_s": 453.1}121{"event": "train", "step": 2340, "loss": -5.852264881134033, "grad_norm": 0.32329216599464417, "lr": 0.00013838920129405016, "ex_per_s": 329.35164954265366, "data_frac": 0.026984810655404857, "mem_mb": 62.975488, "nll": -5.852264881134033, "elapsed_s": 454.8}122{"event": "train", "step": 2360, "loss": -5.929887771606445, "grad_norm": 0.08320047706365585, "lr": 0.0001362045498495942, "ex_per_s": 330.95618016587275, "data_frac": 0.02690961071215786, "mem_mb": 62.975488, "nll": -5.929887771606445, "elapsed_s": 456.4}123{"event": "train", "step": 2380, "loss": -5.930965423583984, "grad_norm": 0.14077050983905792, "lr": 0.00013402777079368853, "ex_per_s": 332.5491734275869, "data_frac": 0.026835679181778076, "mem_mb": 62.975488, "nll": -5.930965423583984, "elapsed_s": 458.1}124{"event": "train", "step": 2400, "loss": -5.957098007202148, "grad_norm": 0.9138509631156921, "lr": 0.00013185945923599213, "ex_per_s": 334.1307667393262, "data_frac": 0.026762230816298672, "mem_mb": 62.975488, "nll": -5.957098007202148, "elapsed_s": 459.7}125{"event": "train", "step": 2420, "loss": -5.997183799743652, "grad_norm": 0.27584362030029297, "lr": 0.00012970020797123432, "ex_per_s": 335.7009324613584, "data_frac": 0.026689163422268264, "mem_mb": 62.975488, "nll": -5.997183799743652, "elapsed_s": 461.4}126{"event": "train", "step": 2440, "loss": -6.099389553070068, "grad_norm": 0.49315932393074036, "lr": 0.000127550607317151, "ex_per_s": 337.2600902795528, "data_frac": 0.026616819715622998, "mem_mb": 62.975488, "nll": -6.099389553070068, "elapsed_s": 463.1}127{"event": "train", "step": 2460, "loss": -6.135371208190918, "grad_norm": 0.14774130284786224, "lr": 0.00012541124495309727, "ex_per_s": 338.81004946663745, "data_frac": 0.026545007908546133, "mem_mb": 62.975488, "nll": -6.135371208190918, "elapsed_s": 464.7}128{"event": "train", "step": 2480, "loss": -5.945044994354248, "grad_norm": 0.13607066869735718, "lr": 0.00012328270575938212, "ex_per_s": 340.3499007159993, "data_frac": 0.02647370300421631, "mem_mb": 62.975488, "nll": -5.945044994354248, "elapsed_s": 466.4}129{"event": "train", "step": 2500, "loss": -5.759143352508545, "grad_norm": 0.5679115056991577, "lr": 0.00012116557165736775, "ex_per_s": 341.8762832915531, "data_frac": 0.026402444790484075, "mem_mb": 62.975488, "nll": -5.759143352508545, "elapsed_s": 468.0}130{"event": "val", "step": 2500, "val_nll": -6.450061465410914, "best": -6.450061465410914, "bad": 0, "mem_mb": 62.975488, "elapsed_s": 530.3}131{"event": "train", "step": 2520, "loss": -5.95911169052124, "grad_norm": 0.7094666361808777, "lr": 0.00011906042145037817, "ex_per_s": 303.15568210781123, "data_frac": 0.02324697325991217, "mem_mb": 62.975488, "nll": -5.95911169052124, "elapsed_s": 532.0}132{"event": "train", "step": 2540, "loss": -6.031495094299316, "grad_norm": 0.46055319905281067, "lr": 0.00011696783066546026, "ex_per_s": 304.60253963843445, "data_frac": 0.02319589335593957, "mem_mb": 62.975488, "nll": -6.031495094299316, "elapsed_s": 533.7}133{"event": "train", "step": 2560, "loss": -5.948669910430908, "grad_norm": 0.1938357949256897, "lr": 0.00011488837139604017, "ex_per_s": 306.043205688183, "data_frac": 0.02314346360315964, "mem_mb": 62.975488, "nll": -5.948669910430908, "elapsed_s": 535.4}134{"event": "train", "step": 2580, "loss": -5.963498115539551, "grad_norm": 0.17277778685092926, "lr": 0.00011282261214551898, "ex_per_s": 307.4743781565143, "data_frac": 0.023090701546745657, "mem_mb": 62.975488, "nll": -5.963498115539551, "elapsed_s": 537.1}135{"event": "train", "step": 2600, "loss": -5.8505778312683105, "grad_norm": 0.36497628688812256, "lr": 0.00011077111767184915, "ex_per_s": 308.89765742660694, "data_frac": 0.023039286657716426, "mem_mb": 62.975488, "nll": -5.8505778312683105, "elapsed_s": 538.7}136{"event": "train", "step": 2620, "loss": -5.890466690063477, "grad_norm": 0.9796863198280334, "lr": 0.00010873444883313583, "ex_per_s": 310.31342612450993, "data_frac": 0.022987820814908105, "mem_mb": 62.975488, "nll": -5.890466690063477, "elapsed_s": 540.4}137{"event": "train", "step": 2640, "loss": -5.988236904144287, "grad_norm": 0.4960867464542389, "lr": 0.00010671316243430345, "ex_per_s": 311.72054758046033, "data_frac": 0.0229368912238403, "mem_mb": 62.975488, "nll": -5.988236904144287, "elapsed_s": 542.1}138{"event": "train", "step": 2660, "loss": -5.820505142211914, "grad_norm": 0.0909847617149353, "lr": 0.00010470781107487112, "ex_per_s": 313.1198586074303, "data_frac": 0.02288617152241294, "mem_mb": 62.975488, "nll": -5.820505142211914, "elapsed_s": 543.7}139{"event": "train", "step": 2680, "loss": -5.959369659423828, "grad_norm": 0.32114461064338684, "lr": 0.00010271894299787715, "ex_per_s": 314.5099656690498, "data_frac": 0.02283543068454518, "mem_mb": 62.975488, "nll": -5.959369659423828, "elapsed_s": 545.4}140{"event": "train", "step": 2700, "loss": -6.133889198303223, "grad_norm": 0.8131086230278015, "lr": 0.00010074710193999506, "ex_per_s": 315.8917811790561, "data_frac": 0.022785389897713667, "mem_mb": 62.975488, "nll": -6.133889198303223, "elapsed_s": 547.1}141{"event": "train", "step": 2720, "loss": -5.882911205291748, "grad_norm": 0.3685641288757324, "lr": 9.879282698288144e-05, "ex_per_s": 317.2645422335585, "data_frac": 0.0227357851708208, "mem_mb": 62.975488, "nll": -5.882911205291748, "elapsed_s": 548.7}142{"event": "train", "step": 2740, "loss": -6.126672744750977, "grad_norm": 0.19870847463607788, "lr": 9.685665240579613e-05, "ex_per_s": 318.62930040343355, "data_frac": 0.022686338504438037, "mem_mb": 62.975488, "nll": -6.126672744750977, "elapsed_s": 550.4}143{"event": "train", "step": 2760, "loss": -5.962531089782715, "grad_norm": 0.20926372706890106, "lr": 9.4939107539536e-05, "ex_per_s": 319.98599196680397, "data_frac": 0.02263727723635008, "mem_mb": 62.975488, "nll": -5.962531089782715, "elapsed_s": 552.1}144{"event": "train", "step": 2780, "loss": -6.014058589935303, "grad_norm": 0.8711726665496826, "lr": 9.304071662172087e-05, "ex_per_s": 321.3338342747095, "data_frac": 0.022589819399705897, "mem_mb": 62.975488, "nll": -6.014058589935303, "elapsed_s": 553.7}145{"event": "train", "step": 2800, "loss": -5.981621265411377, "grad_norm": 0.8087593913078308, "lr": 9.116199865347241e-05, "ex_per_s": 322.6753926662191, "data_frac": 0.02254149731809018, "mem_mb": 62.975488, "nll": -5.981621265411377, "elapsed_s": 555.4}146{"event": "train", "step": 2820, "loss": -6.007476329803467, "grad_norm": 0.3585527539253235, "lr": 8.930346725752419e-05, "ex_per_s": 324.00989019547444, "data_frac": 0.022493070910707846, "mem_mb": 62.975488, "nll": -6.007476329803467, "elapsed_s": 557.1}147{"event": "train", "step": 2840, "loss": -5.8803019523620605, "grad_norm": 0.09727909415960312, "lr": 8.746563053780262e-05, "ex_per_s": 325.3366004923835, "data_frac": 0.022444840461835206, "mem_mb": 62.975488, "nll": -5.8803019523620605, "elapsed_s": 558.7}148{"event": "train", "step": 2860, "loss": -6.049093723297119, "grad_norm": 0.17868046462535858, "lr": 8.564899094051614e-05, "ex_per_s": 326.6547709008837, "data_frac": 0.022398394714057388, "mem_mb": 62.975488, "nll": -6.049093723297119, "elapsed_s": 560.4}149{"event": "train", "step": 2880, "loss": -5.772797584533691, "grad_norm": 0.33841952681541443, "lr": 8.385404511679161e-05, "ex_per_s": 327.9671430672805, "data_frac": 0.022351228139185955, "mem_mb": 62.975488, "nll": -5.772797584533691, "elapsed_s": 562.1}150{"event": "train", "step": 2900, "loss": -5.795140743255615, "grad_norm": 0.6148343682289124, "lr": 8.208128378689484e-05, "ex_per_s": 329.27107849457576, "data_frac": 0.02230476592914878, "mem_mb": 62.975488, "nll": -5.795140743255615, "elapsed_s": 563.7}151{"event": "train", "step": 2920, "loss": -5.970316410064697, "grad_norm": 0.7454568147659302, "lr": 8.03311916060725e-05, "ex_per_s": 330.5669138366921, "data_frac": 0.022258906522894337, "mem_mb": 62.975488, "nll": -5.970316410064697, "elapsed_s": 565.4}152{"event": "train", "step": 2940, "loss": -5.904286861419678, "grad_norm": 0.5246623754501343, "lr": 7.860424703205225e-05, "ex_per_s": 331.8554385084055, "data_frac": 0.022213643172123845, "mem_mb": 62.975488, "nll": -5.904286861419678, "elapsed_s": 567.0}153{"event": "train", "step": 2960, "loss": -5.938860893249512, "grad_norm": 0.3099009692668915, "lr": 7.690092219423722e-05, "ex_per_s": 333.1373630202534, "data_frac": 0.0221690426744889, "mem_mb": 62.975488, "nll": -5.938860893249512, "elapsed_s": 568.7}154{"event": "train", "step": 2980, "loss": -5.895195960998535, "grad_norm": 0.445403128862381, "lr": 7.522168276463048e-05, "ex_per_s": 334.4109554028518, "data_frac": 0.02212364200642189, "mem_mb": 62.975488, "nll": -5.895195960998535, "elapsed_s": 570.4}155{"event": "train", "step": 3000, "loss": -5.995512008666992, "grad_norm": 0.11661339551210403, "lr": 7.356698783052496e-05, "ex_per_s": 335.67774225181, "data_frac": 0.022077751521494438, "mem_mb": 62.975488, "nll": -5.995512008666992, "elapsed_s": 572.0}156{"event": "val", "step": 3000, "val_nll": -6.448553899015936, "best": -6.450061465410914, "bad": 1, "mem_mb": 62.975488, "elapsed_s": 634.3}157{"event": "train", "step": 3020, "loss": -5.961388111114502, "grad_norm": 0.924397349357605, "lr": 7.193728976899362e-05, "ex_per_s": 303.9222407398381, "data_frac": 0.01987409717159427, "mem_mb": 62.975488, "nll": -5.961388111114502, "elapsed_s": 636.0}158{"event": "train", "step": 3040, "loss": -5.858161449432373, "grad_norm": 0.8244529962539673, "lr": 7.033303412321404e-05, "ex_per_s": 305.1322676658147, "data_frac": 0.019839910191708773, "mem_mb": 62.975488, "nll": -5.858161449432373, "elapsed_s": 637.7}159{"event": "train", "step": 3060, "loss": -6.071932792663574, "grad_norm": 0.28709253668785095, "lr": 6.875465948066139e-05, "ex_per_s": 306.336194486772, "data_frac": 0.019804761464914902, "mem_mb": 62.975488, "nll": -6.071932792663574, "elapsed_s": 639.3}160{"event": "train", "step": 3080, "loss": -5.970698833465576, "grad_norm": 0.24104391038417816, "lr": 6.720259735320299e-05, "ex_per_s": 307.5346788876798, "data_frac": 0.01976956117649423, "mem_mb": 62.975488, "nll": -5.970698833465576, "elapsed_s": 641.0}161{"event": "train", "step": 3100, "loss": -6.016872406005859, "grad_norm": 0.27783533930778503, "lr": 6.567727205912722e-05, "ex_per_s": 308.7265867361288, "data_frac": 0.019735149972592857, "mem_mb": 62.975488, "nll": -6.016872406005859, "elapsed_s": 642.7}162{"event": "train", "step": 3120, "loss": -6.091909885406494, "grad_norm": 0.5370648503303528, "lr": 6.417910060713926e-05, "ex_per_s": 309.9128656890685, "data_frac": 0.019700823454994975, "mem_mb": 62.975488, "nll": -6.091909885406494, "elapsed_s": 644.4}163{"event": "train", "step": 3140, "loss": -6.009003639221191, "grad_norm": 0.3426913321018219, "lr": 6.270849258235485e-05, "ex_per_s": 311.09264905943013, "data_frac": 0.019666470052222506, "mem_mb": 62.975488, "nll": -6.009003639221191, "elapsed_s": 646.0}164{"event": "train", "step": 3160, "loss": -6.021994590759277, "grad_norm": 0.12237885594367981, "lr": 6.126585003432406e-05, "ex_per_s": 312.26637387385983, "data_frac": 0.01963224792595946, "mem_mb": 62.975488, "nll": -6.021994590759277, "elapsed_s": 647.7}165{"event": "train", "step": 3180, "loss": -6.09563684463501, "grad_norm": 0.14587832987308502, "lr": 5.985156736711483e-05, "ex_per_s": 313.4332197169019, "data_frac": 0.019598159823619322, "mem_mb": 62.975488, "nll": -6.09563684463501, "elapsed_s": 649.4}166{"event": "train", "step": 3200, "loss": -6.002330780029297, "grad_norm": 0.3112635016441345, "lr": 5.846603123148688e-05, "ex_per_s": 314.5942000477231, "data_frac": 0.019565445305447227, "mem_mb": 62.975488, "nll": -6.002330780029297, "elapsed_s": 651.0}167{"event": "train", "step": 3220, "loss": -5.915475845336914, "grad_norm": 0.17776775360107422, "lr": 5.710962041918516e-05, "ex_per_s": 315.74985010942606, "data_frac": 0.019531531119712234, "mem_mb": 62.975488, "nll": -5.915475845336914, "elapsed_s": 652.7}168{"event": "train", "step": 3240, "loss": -5.801457405090332, "grad_norm": 0.49123379588127136, "lr": 5.578270575938211e-05, "ex_per_s": 316.89846197004766, "data_frac": 0.019499583681546723, "mem_mb": 62.975488, "nll": -5.801457405090332, "elapsed_s": 654.4}169{"event": "train", "step": 3260, "loss": -5.960771083831787, "grad_norm": 0.524750828742981, "lr": 5.448565001729653e-05, "ex_per_s": 318.0430169145838, "data_frac": 0.019466443528559432, "mem_mb": 62.975488, "nll": -5.960771083831787, "elapsed_s": 656.1}170{"event": "train", "step": 3280, "loss": -6.016517639160156, "grad_norm": 0.3465154469013214, "lr": 5.3218807795017296e-05, "ex_per_s": 319.18186203642455, "data_frac": 0.01943362070027745, "mem_mb": 62.975488, "nll": -6.016517639160156, "elapsed_s": 657.7}171{"event": "train", "step": 3300, "loss": -5.806588172912598, "grad_norm": 0.15143848955631256, "lr": 5.198252543455864e-05, "ex_per_s": 320.3144089356399, "data_frac": 0.01940056529681855, "mem_mb": 62.975488, "nll": -5.806588172912598, "elapsed_s": 659.4}172{"event": "train", "step": 3320, "loss": -5.989216327667236, "grad_norm": 0.2907240092754364, "lr": 5.07771409231737e-05, "ex_per_s": 321.44130749724417, "data_frac": 0.01936767510172663, "mem_mb": 62.975488, "nll": -5.989216327667236, "elapsed_s": 661.1}173{"event": "train", "step": 3340, "loss": -5.932101249694824, "grad_norm": 0.286105215549469, "lr": 4.960298380095221e-05, "ex_per_s": 322.5634620936124, "data_frac": 0.01933523789358007, "mem_mb": 62.975488, "nll": -5.932101249694824, "elapsed_s": 662.7}174{"event": "train", "step": 3360, "loss": -5.9081268310546875, "grad_norm": 0.2813342213630676, "lr": 4.846037507072751e-05, "ex_per_s": 323.68006457432546, "data_frac": 0.019302416014307308, "mem_mb": 62.975488, "nll": -5.9081268310546875, "elapsed_s": 664.4}175{"event": "train", "step": 3380, "loss": -6.122231483459473, "grad_norm": 0.4649679958820343, "lr": 4.7349627110317556e-05, "ex_per_s": 324.79174200742773, "data_frac": 0.019269877750155633, "mem_mb": 62.975488, "nll": -6.122231483459473, "elapsed_s": 666.1}176{"event": "train", "step": 3400, "loss": -5.883323669433594, "grad_norm": 0.3584122955799103, "lr": 4.6271043587123986e-05, "ex_per_s": 325.89777835060005, "data_frac": 0.019237805170155674, "mem_mb": 62.975488, "nll": -5.883323669433594, "elapsed_s": 667.7}177{"event": "train", "step": 3420, "loss": -5.837618827819824, "grad_norm": 0.29298651218414307, "lr": 4.52249193751124e-05, "ex_per_s": 327.00005337722297, "data_frac": 0.019206006456945663, "mem_mb": 62.975488, "nll": -5.837618827819824, "elapsed_s": 669.4}178{"event": "train", "step": 3440, "loss": -5.916536331176758, "grad_norm": 0.14620330929756165, "lr": 4.4211540474196926e-05, "ex_per_s": 328.09657600915375, "data_frac": 0.01917445241763894, "mem_mb": 62.975488, "nll": -5.916536331176758, "elapsed_s": 671.1}179{"event": "train", "step": 3460, "loss": -6.05958890914917, "grad_norm": 0.14066673815250397, "lr": 4.3231183932050546e-05, "ex_per_s": 329.1889063305394, "data_frac": 0.019143145017257106, "mem_mb": 62.975488, "nll": -6.05958890914917, "elapsed_s": 672.7}180{"event": "train", "step": 3480, "loss": -5.917280197143555, "grad_norm": 0.06323210895061493, "lr": 4.228411776836306e-05, "ex_per_s": 330.2754437935999, "data_frac": 0.01911305505079715, "mem_mb": 62.975488, "nll": -5.917280197143555, "elapsed_s": 674.4}181{"event": "train", "step": 3500, "loss": -6.038781642913818, "grad_norm": 0.08257296681404114, "lr": 4.137060090156723e-05, "ex_per_s": 331.3593360819161, "data_frac": 0.01908191633401858, "mem_mb": 62.975488, "nll": -6.038781642913818, "elapsed_s": 676.0}182{"event": "val", "step": 3500, "val_nll": -6.450709887063427, "best": -6.450709887063427, "bad": 0, "mem_mb": 62.975488, "elapsed_s": 738.4}183{"event": "train", "step": 3520, "loss": -5.94032621383667, "grad_norm": 0.46814173460006714, "lr": 4.049088307805306e-05, "ex_per_s": 304.41434196492753, "data_frac": 0.01744561490940905, "mem_mb": 62.975488, "nll": -5.94032621383667, "elapsed_s": 740.1}184{"event": "train", "step": 3540, "loss": -6.1358466148376465, "grad_norm": 0.6212455034255981, "lr": 3.964520480388956e-05, "ex_per_s": 305.4552981931815, "data_frac": 0.01742156949398362, "mem_mb": 62.975488, "nll": -6.1358466148376465, "elapsed_s": 741.8}185{"event": "train", "step": 3560, "loss": -6.138979434967041, "grad_norm": 0.07811778038740158, "lr": 3.88337972790728e-05, "ex_per_s": 306.4905092379817, "data_frac": 0.017396931623307948, "mem_mb": 62.975488, "nll": -6.138979434967041, "elapsed_s": 743.4}186{"event": "train", "step": 3580, "loss": -5.902897357940674, "grad_norm": 0.2616788148880005, "lr": 3.805688233431824e-05, "ex_per_s": 307.5212791537497, "data_frac": 0.017371915236942065, "mem_mb": 62.975488, "nll": -5.902897357940674, "elapsed_s": 745.1}187{"event": "train", "step": 3600, "loss": -6.100764274597168, "grad_norm": 0.41903388500213623, "lr": 3.7314672370414325e-05, "ex_per_s": 308.5474303382385, "data_frac": 0.017347104922699117, "mem_mb": 62.975488, "nll": -6.100764274597168, "elapsed_s": 746.8}188{"event": "train", "step": 3620, "loss": -5.9105610847473145, "grad_norm": 0.9625649452209473, "lr": 3.660737030015427e-05, "ex_per_s": 309.5688087432627, "data_frac": 0.017322643144260817, "mem_mb": 62.975488, "nll": -5.9105610847473145, "elapsed_s": 748.4}189{"event": "train", "step": 3640, "loss": -6.053289890289307, "grad_norm": 0.5041558146476746, "lr": 3.5935169492861716e-05, "ex_per_s": 310.5855057831974, "data_frac": 0.01729843536077831, "mem_mb": 62.975488, "nll": -6.053289890289307, "elapsed_s": 750.1}190{"event": "train", "step": 3660, "loss": -5.830911636352539, "grad_norm": 0.5979053974151611, "lr": 3.52982537215255e-05, "ex_per_s": 311.59806549504174, "data_frac": 0.017274106571323347, "mem_mb": 62.975488, "nll": -5.830911636352539, "elapsed_s": 751.8}191{"event": "train", "step": 3680, "loss": -5.992390155792236, "grad_norm": 0.1085156798362732, "lr": 3.469679711255794e-05, "ex_per_s": 312.605806571519, "data_frac": 0.017249940321298084, "mem_mb": 62.975488, "nll": -5.992390155792236, "elapsed_s": 753.5}192{"event": "train", "step": 3700, "loss": -6.068389892578125, "grad_norm": 0.8185369372367859, "lr": 3.41309640981904e-05, "ex_per_s": 313.60817429861066, "data_frac": 0.017226811656773457, "mem_mb": 62.975488, "nll": -6.068389892578125, "elapsed_s": 755.1}193{"event": "train", "step": 3720, "loss": -5.9491682052612305, "grad_norm": 0.6196960806846619, "lr": 3.3600909371519225e-05, "ex_per_s": 314.60599150692593, "data_frac": 0.01720296539791034, "mem_mb": 62.975488, "nll": -5.9491682052612305, "elapsed_s": 756.8}194{"event": "train", "step": 3740, "loss": -6.190184593200684, "grad_norm": 0.2136981040239334, "lr": 3.31067778442141e-05, "ex_per_s": 315.60119693631106, "data_frac": 0.017179083977512256, "mem_mb": 62.975488, "nll": -6.190184593200684, "elapsed_s": 758.5}195{"event": "train", "step": 3760, "loss": -6.147538185119629, "grad_norm": 0.18072570860385895, "lr": 3.2648704606900735e-05, "ex_per_s": 316.5915567709223, "data_frac": 0.017155096398056898, "mem_mb": 62.975488, "nll": -6.147538185119629, "elapsed_s": 760.1}196{"event": "train", "step": 3780, "loss": -5.964047431945801, "grad_norm": 0.34271475672721863, "lr": 3.222681489222838e-05, "ex_per_s": 317.5775365875816, "data_frac": 0.01713165243597053, "mem_mb": 62.975488, "nll": -5.964047431945801, "elapsed_s": 761.8}197{"event": "train", "step": 3800, "loss": -6.099583625793457, "grad_norm": 0.09709912538528442, "lr": 3.1841224040632484e-05, "ex_per_s": 318.55896792352524, "data_frac": 0.017108392869621757, "mem_mb": 62.975488, "nll": -6.099583625793457, "elapsed_s": 763.5}198{"event": "train", "step": 3820, "loss": -5.9716620445251465, "grad_norm": 0.14700378477573395, "lr": 3.149203746880175e-05, "ex_per_s": 319.537436937726, "data_frac": 0.01708498106287876, "mem_mb": 62.975488, "nll": -5.9716620445251465, "elapsed_s": 765.1}199{"event": "train", "step": 3840, "loss": -5.875292778015137, "grad_norm": 0.6468054056167603, "lr": 3.117935064085835e-05, "ex_per_s": 320.5117562744875, "data_frac": 0.017061863708153553, "mem_mb": 62.975488, "nll": -5.875292778015137, "elapsed_s": 766.8}200{"event": "train", "step": 3860, "loss": -6.013393402099609, "grad_norm": 0.44664889574050903, "lr": 3.090324904225885e-05, "ex_per_s": 321.48201168173557, "data_frac": 0.017038674885284676, "mem_mb": 62.975488, "nll": -6.013393402099609, "elapsed_s": 768.5}201{"event": "train", "step": 3880, "loss": -5.891435623168945, "grad_norm": 0.10579986870288849, "lr": 3.0663808156423456e-05, "ex_per_s": 322.44740827671666, "data_frac": 0.017015658286114208, "mem_mb": 62.975488, "nll": -5.891435623168945, "elapsed_s": 770.2}202{"event": "train", "step": 3900, "loss": -6.182982444763184, "grad_norm": 0.7727223634719849, "lr": 3.046109344409957e-05, "ex_per_s": 323.4095703126334, "data_frac": 0.016992682538598355, "mem_mb": 62.975488, "nll": -6.182982444763184, "elapsed_s": 771.8}203{"event": "train", "step": 3920, "loss": -6.038151264190674, "grad_norm": 0.19042260944843292, "lr": 3.0295160325465435e-05, "ex_per_s": 324.3681710185235, "data_frac": 0.016969877747019673, "mem_mb": 62.975488, "nll": -6.038151264190674, "elapsed_s": 773.5}204{"event": "train", "step": 3940, "loss": -5.9236555099487305, "grad_norm": 0.11390967667102814, "lr": 3.016605416497886e-05, "ex_per_s": 325.32251043690496, "data_frac": 0.016948326860726864, "mem_mb": 62.975488, "nll": -5.9236555099487305, "elapsed_s": 775.2}205{"event": "train", "step": 3960, "loss": -6.069462776184082, "grad_norm": 0.11679738014936447, "lr": 3.0073810258974985e-05, "ex_per_s": 326.27505025900246, "data_frac": 0.016926132857922372, "mem_mb": 62.975488, "nll": -6.069462776184082, "elapsed_s": 776.8}206{"event": "train", "step": 3980, "loss": -5.802906513214111, "grad_norm": 0.1844089925289154, "lr": 3.0018453826016683e-05, "ex_per_s": 327.2218003981505, "data_frac": 0.01690420742265346, "mem_mb": 62.975488, "nll": -5.802906513214111, "elapsed_s": 778.5}207{"event": "train", "step": 4000, "loss": -6.058515548706055, "grad_norm": 0.734529435634613, "lr": 2.9999999999999997e-05, "ex_per_s": 328.16325203522456, "data_frac": 0.016882186086508356, "mem_mb": 62.975488, "nll": -6.058515548706055, "elapsed_s": 780.1}208{"event": "val", "step": 4000, "val_nll": -6.450159950636393, "best": -6.450709887063427, "bad": 1, "mem_mb": 62.975488, "elapsed_s": 842.5}209{"event": "done", "steps": 4000, "best_val": -6.450709887063427, "wall_s": 842.46826004982, "ex_per_s": 303.86901458443225, "peak_mem_mb": 62.975488, "n_params": 3455619, "device": "mps", "data_time_frac": 0.01563238180485749, "elapsed_s": 842.5}210
Files · main
Menu

Theme