SPB Git

spb/wp41_uqo

Public
51.5 KB · 210 lines
1{"event": "train", "step": 20, "loss": -6.114568710327148, "grad_norm": 0.8954913020133972, "lr": 3.149999999999999e-05, "ex_per_s": 90.8697681210973, "data_frac": 0.8688237453453167, "mem_mb": 91.948032, "nll": -6.114568710327148, "elapsed_s": 14.1}2{"event": "train", "step": 40, "loss": -5.8996710777282715, "grad_norm": 0.3089335262775421, "lr": 6.149999999999999e-05, "ex_per_s": 161.47302658833974, "data_frac": 0.7762240056900295, "mem_mb": 91.948032, "nll": -5.8996710777282715, "elapsed_s": 15.9}3{"event": "train", "step": 60, "loss": -5.974310874938965, "grad_norm": 0.5597636699676514, "lr": 9.149999999999999e-05, "ex_per_s": 218.37608658395575, "data_frac": 0.701940489258357, "mem_mb": 91.948032, "nll": -5.974310874938965, "elapsed_s": 17.6}4{"event": "train", "step": 80, "loss": -6.073030948638916, "grad_norm": 0.5797526836395264, "lr": 0.0001215, "ex_per_s": 265.7182050279959, "data_frac": 0.6412390370298925, "mem_mb": 91.948032, "nll": -6.073030948638916, "elapsed_s": 19.3}5{"event": "train", "step": 100, "loss": -5.9457478523254395, "grad_norm": 0.31914326548576355, "lr": 0.0001515, "ex_per_s": 305.6000216578499, "data_frac": 0.5905429445597864, "mem_mb": 91.948032, "nll": -5.9457478523254395, "elapsed_s": 21.0}6{"event": "train", "step": 120, "loss": -5.896589756011963, "grad_norm": 1.359634280204773, "lr": 0.00018149999999999997, "ex_per_s": 339.50171466402884, "data_frac": 0.54723557971757, "mem_mb": 91.948032, "nll": -5.896589756011963, "elapsed_s": 22.7}7{"event": "train", "step": 140, "loss": -5.830938816070557, "grad_norm": 0.2081683874130249, "lr": 0.00021149999999999996, "ex_per_s": 368.82594133046416, "data_frac": 0.5100515446059785, "mem_mb": 91.948032, "nll": -5.830938816070557, "elapsed_s": 24.3}8{"event": "train", "step": 160, "loss": -5.730869293212891, "grad_norm": 0.910971462726593, "lr": 0.0002415, "ex_per_s": 394.3786117309881, "data_frac": 0.4776597323359987, "mem_mb": 91.948032, "nll": -5.730869293212891, "elapsed_s": 26.0}9{"event": "train", "step": 180, "loss": -5.975837230682373, "grad_norm": 0.9762269258499146, "lr": 0.0002715, "ex_per_s": 416.90687570966924, "data_frac": 0.4492318639134499, "mem_mb": 91.948032, "nll": -5.975837230682373, "elapsed_s": 27.7}10{"event": "train", "step": 200, "loss": -5.962082386016846, "grad_norm": 1.225264549255371, "lr": 0.0003, "ex_per_s": 436.85578903648354, "data_frac": 0.4240425986846535, "mem_mb": 91.948032, "nll": -5.962082386016846, "elapsed_s": 29.3}11{"event": "train", "step": 220, "loss": -6.032444953918457, "grad_norm": 0.9658782482147217, "lr": 0.00029998154617398326, "ex_per_s": 454.7145783991766, "data_frac": 0.40158026323855966, "mem_mb": 91.948032, "nll": -6.032444953918457, "elapsed_s": 31.0}12{"event": "train", "step": 240, "loss": -5.913348197937012, "grad_norm": 0.6255849003791809, "lr": 0.000299926189741025, "ex_per_s": 470.7254744735194, "data_frac": 0.3813948542639274, "mem_mb": 91.948032, "nll": -5.913348197937012, "elapsed_s": 32.7}13{"event": "train", "step": 260, "loss": -5.886681079864502, "grad_norm": 0.1683369278907776, "lr": 0.0002998339458350211, "ex_per_s": 485.19538638076045, "data_frac": 0.3631681621058202, "mem_mb": 91.948032, "nll": -5.886681079864502, "elapsed_s": 34.3}14{"event": "train", "step": 280, "loss": -5.943408012390137, "grad_norm": 0.14963608980178833, "lr": 0.0002997048396745345, "ex_per_s": 498.32545248792877, "data_frac": 0.34675220539159457, "mem_mb": 91.948032, "nll": -5.943408012390137, "elapsed_s": 36.0}15{"event": "train", "step": 300, "loss": -5.967341423034668, "grad_norm": 0.2688431143760681, "lr": 0.0002995389065559004, "ex_per_s": 510.1872546639888, "data_frac": 0.33160899471833344, "mem_mb": 91.948032, "nll": -5.967341423034668, "elapsed_s": 37.7}16{"event": "train", "step": 320, "loss": -5.966516017913818, "grad_norm": 0.604865312576294, "lr": 0.0002993361918435765, "ex_per_s": 521.0498551515867, "data_frac": 0.3177699024984555, "mem_mb": 91.948032, "nll": -5.966516017913818, "elapsed_s": 39.3}17{"event": "train", "step": 340, "loss": -5.833719253540039, "grad_norm": 0.17973990738391876, "lr": 0.00029909675095774114, "ex_per_s": 531.0935761085359, "data_frac": 0.30509342453386423, "mem_mb": 91.948032, "nll": -5.833719253540039, "elapsed_s": 41.0}18{"event": "train", "step": 360, "loss": -5.996518135070801, "grad_norm": 0.16102775931358337, "lr": 0.00029882064935914164, "ex_per_s": 540.3226824081622, "data_frac": 0.29341778886482794, "mem_mb": 91.948032, "nll": -5.996518135070801, "elapsed_s": 42.7}19{"event": "train", "step": 380, "loss": -6.006462574005127, "grad_norm": 1.204532265663147, "lr": 0.00029850796253119824, "ex_per_s": 548.8859562493351, "data_frac": 0.28263117654322467, "mem_mb": 91.948032, "nll": -6.006462574005127, "elapsed_s": 44.4}20{"event": "train", "step": 400, "loss": -5.9933671951293945, "grad_norm": 1.5197361707687378, "lr": 0.0002981587759593675, "ex_per_s": 556.8432073168328, "data_frac": 0.27261809833290634, "mem_mb": 91.948032, "nll": -5.9933671951293945, "elapsed_s": 46.0}21{"event": "train", "step": 420, "loss": -5.962832450866699, "grad_norm": 0.6785749197006226, "lr": 0.0002977731851077716, "ex_per_s": 564.2357408951195, "data_frac": 0.2632948665448291, "mem_mb": 91.948032, "nll": -5.962832450866699, "elapsed_s": 47.7}22{"event": "train", "step": 440, "loss": -5.939571380615234, "grad_norm": 0.4964001774787903, "lr": 0.0002973512953930992, "ex_per_s": 571.1264232640493, "data_frac": 0.2546000379846389, "mem_mb": 91.948032, "nll": -5.939571380615234, "elapsed_s": 49.3}23{"event": "train", "step": 460, "loss": -6.09814977645874, "grad_norm": 0.7545742392539978, "lr": 0.0002968932221557859, "ex_per_s": 577.5786745615817, "data_frac": 0.24648165170509492, "mem_mb": 91.948032, "nll": -6.09814977645874, "elapsed_s": 51.0}24{"event": "train", "step": 480, "loss": -5.876066207885742, "grad_norm": 0.16249093413352966, "lr": 0.0002963990906284808, "ex_per_s": 583.57814572382, "data_frac": 0.23885950234893918, "mem_mb": 91.948032, "nll": -5.876066207885742, "elapsed_s": 52.7}25{"event": "train", "step": 500, "loss": -5.925200462341309, "grad_norm": 0.27171337604522705, "lr": 0.00029586903590180956, "ex_per_s": 589.2145501880088, "data_frac": 0.23170834293899836, "mem_mb": 91.948032, "nll": -5.925200462341309, "elapsed_s": 54.4}26{"event": "val", "step": 500, "val_nll": -6.439585366268128, "best": -6.439585366268128, "bad": 0, "mem_mb": 91.948032, "elapsed_s": 118.3}27{"event": "train", "step": 520, "loss": -6.038392543792725, "grad_norm": 0.23031805455684662, "lr": 0.00029530320288744205, "ex_per_s": 277.1596117645077, "data_frac": 0.10490089732292701, "mem_mb": 91.948032, "nll": -6.038392543792725, "elapsed_s": 120.1}28{"event": "train", "step": 540, "loss": -5.877813816070557, "grad_norm": 0.5785009860992432, "lr": 0.0002947017462784745, "ex_per_s": 283.86063309409064, "data_frac": 0.10354596649261591, "mem_mb": 91.948032, "nll": -5.877813816070557, "elapsed_s": 121.8}29{"event": "train", "step": 560, "loss": -5.972142696380615, "grad_norm": 0.10391000658273697, "lr": 0.00029406483050713824, "ex_per_s": 290.3725802085437, "data_frac": 0.10222836483299559, "mem_mb": 91.948032, "nll": -5.972142696380615, "elapsed_s": 123.5}30{"event": "train", "step": 580, "loss": -5.9249091148376465, "grad_norm": 0.12703943252563477, "lr": 0.0002933926296998457, "ex_per_s": 296.59449139204054, "data_frac": 0.101569351827238, "mem_mb": 91.948032, "nll": -5.9249091148376465, "elapsed_s": 125.2}31{"event": "train", "step": 600, "loss": -5.974193572998047, "grad_norm": 0.6641250252723694, "lr": 0.0002926853276295856, "ex_per_s": 302.7754906716156, "data_frac": 0.10031405819006028, "mem_mb": 91.948032, "nll": -5.974193572998047, "elapsed_s": 126.9}32{"event": "train", "step": 620, "loss": -5.95150089263916, "grad_norm": 0.3683467507362366, "lr": 0.00029194311766568174, "ex_per_s": 308.78937481236517, "data_frac": 0.09908464470542971, "mem_mb": 91.948032, "nll": -5.95150089263916, "elapsed_s": 128.5}33{"event": "train", "step": 640, "loss": -6.064396381378174, "grad_norm": 0.4986158013343811, "lr": 0.0002911662027209272, "ex_per_s": 314.66238217541763, "data_frac": 0.09789410084368386, "mem_mb": 91.948032, "nll": -6.064396381378174, "elapsed_s": 130.2}34{"event": "train", "step": 660, "loss": -6.00988245010376, "grad_norm": 1.1576873064041138, "lr": 0.00029035479519611045, "ex_per_s": 320.3865443598062, "data_frac": 0.09673597237861095, "mem_mb": 91.948032, "nll": -6.00988245010376, "elapsed_s": 131.9}35{"event": "train", "step": 680, "loss": -6.004580974578857, "grad_norm": 0.3126148581504822, "lr": 0.0002895091169219469, "ex_per_s": 325.96317999026434, "data_frac": 0.09560371122792582, "mem_mb": 91.948032, "nll": -6.004580974578857, "elapsed_s": 133.6}36{"event": "train", "step": 700, "loss": -5.984941482543945, "grad_norm": 1.178809404373169, "lr": 0.00028862939909843273, "ex_per_s": 331.4056276914961, "data_frac": 0.09449945547607008, "mem_mb": 91.948032, "nll": -5.984941482543945, "elapsed_s": 135.2}37{"event": "train", "step": 720, "loss": -6.062106609344482, "grad_norm": 0.3115517795085907, "lr": 0.0002877158822316369, "ex_per_s": 336.71534230405234, "data_frac": 0.09342079117312428, "mem_mb": 91.948032, "nll": -6.062106609344482, "elapsed_s": 136.9}38{"event": "train", "step": 740, "loss": -6.03580904006958, "grad_norm": 0.1256667971611023, "lr": 0.00028676881606794945, "ex_per_s": 341.9006045345279, "data_frac": 0.09236983142565765, "mem_mb": 91.948032, "nll": -6.03580904006958, "elapsed_s": 138.6}39{"event": "train", "step": 760, "loss": -5.9622392654418945, "grad_norm": 0.5633066892623901, "lr": 0.000285788459525803, "ex_per_s": 346.9531088901064, "data_frac": 0.09134323969289637, "mem_mb": 91.948032, "nll": -5.9622392654418945, "elapsed_s": 140.2}40{"event": "train", "step": 780, "loss": -5.891730785369873, "grad_norm": 0.5876479744911194, "lr": 0.00028477508062488754, "ex_per_s": 351.89486390047125, "data_frac": 0.09034327931510841, "mem_mb": 91.948032, "nll": -5.891730785369873, "elapsed_s": 141.9}41{"event": "train", "step": 800, "loss": -5.816739082336426, "grad_norm": 0.12639594078063965, "lr": 0.000283728956412876, "ex_per_s": 356.71948397720985, "data_frac": 0.08937286775430145, "mem_mb": 91.948032, "nll": -5.816739082336426, "elapsed_s": 143.6}42{"event": "train", "step": 820, "loss": -5.935604572296143, "grad_norm": 0.45413610339164734, "lr": 0.00028265037288968236, "ex_per_s": 361.44598087958184, "data_frac": 0.08842081143507591, "mem_mb": 91.948032, "nll": -5.935604572296143, "elapsed_s": 145.2}43{"event": "train", "step": 840, "loss": -5.808259010314941, "grad_norm": 0.5485562682151794, "lr": 0.0002815396249292725, "ex_per_s": 366.062573236612, "data_frac": 0.08748788491885295, "mem_mb": 91.948032, "nll": -5.808259010314941, "elapsed_s": 146.9}44{"event": "train", "step": 860, "loss": -5.912571907043457, "grad_norm": 1.7309210300445557, "lr": 0.00028039701619904774, "ex_per_s": 370.5758665963113, "data_frac": 0.08657781628345772, "mem_mb": 91.948032, "nll": -5.912571907043457, "elapsed_s": 148.6}45{"event": "train", "step": 880, "loss": -5.867587566375732, "grad_norm": 0.550331175327301, "lr": 0.00027922285907682625, "ex_per_s": 374.9924509305565, "data_frac": 0.08568915552371428, "mem_mb": 91.948032, "nll": -5.867587566375732, "elapsed_s": 150.2}46{"event": "train", "step": 900, "loss": -5.984101295471191, "grad_norm": 0.4785093665122986, "lr": 0.00027801747456544134, "ex_per_s": 379.30794091875015, "data_frac": 0.08482143678608, "mem_mb": 91.948032, "nll": -5.984101295471191, "elapsed_s": 151.9}47{"event": "train", "step": 920, "loss": -6.026276111602783, "grad_norm": 0.43799492716789246, "lr": 0.00027678119220498265, "ex_per_s": 383.52982437668913, "data_frac": 0.08397048093144922, "mem_mb": 91.948032, "nll": -6.026276111602783, "elapsed_s": 153.6}48{"event": "train", "step": 940, "loss": -6.08331298828125, "grad_norm": 0.9071180820465088, "lr": 0.0002755143499827035, "ex_per_s": 387.6623017615878, "data_frac": 0.08313834787938094, "mem_mb": 91.948032, "nll": -6.08331298828125, "elapsed_s": 155.2}49{"event": "train", "step": 960, "loss": -5.963986396789551, "grad_norm": 0.2857986390590668, "lr": 0.00027421729424061787, "ex_per_s": 391.7065747087263, "data_frac": 0.08232211226002836, "mem_mb": 91.948032, "nll": -5.963986396789551, "elapsed_s": 156.9}50{"event": "train", "step": 980, "loss": -5.927439212799072, "grad_norm": 0.38125017285346985, "lr": 0.00027289037958081483, "ex_per_s": 395.65794705500883, "data_frac": 0.08152450903057454, "mem_mb": 91.948032, "nll": -5.927439212799072, "elapsed_s": 158.6}51{"event": "train", "step": 1000, "loss": -6.072062969207764, "grad_norm": 0.21991685032844543, "lr": 0.0002715339687685131, "ex_per_s": 399.5297287008356, "data_frac": 0.08074257878566336, "mem_mb": 91.948032, "nll": -6.072062969207764, "elapsed_s": 160.2}52{"event": "val", "step": 1000, "val_nll": -6.446348534925756, "best": -6.446348534925756, "bad": 0, "mem_mb": 91.948032, "elapsed_s": 223.7}53{"event": "train", "step": 1020, "loss": -6.028501033782959, "grad_norm": 0.23558659851551056, "lr": 0.0002701484326328852, "ex_per_s": 289.59082800728766, "data_frac": 0.05742425255787626, "mem_mb": 91.948032, "nll": -6.028501033782959, "elapsed_s": 225.5}54{"event": "train", "step": 1040, "loss": -5.971804141998291, "grad_norm": 0.2588319480419159, "lr": 0.00026873414996567596, "ex_per_s": 293.0849688197527, "data_frac": 0.057046155711432976, "mem_mb": 91.948032, "nll": -5.971804141998291, "elapsed_s": 227.1}55{"event": "train", "step": 1060, "loss": -6.184239387512207, "grad_norm": 0.2932659983634949, "lr": 0.0002672915074176452, "ex_per_s": 296.5254429706092, "data_frac": 0.05667280510379495, "mem_mb": 91.948032, "nll": -6.184239387512207, "elapsed_s": 228.8}56{"event": "train", "step": 1080, "loss": -5.813992023468018, "grad_norm": 0.5792493224143982, "lr": 0.00026582089939286075, "ex_per_s": 299.88987230770186, "data_frac": 0.05630074296045311, "mem_mb": 91.948032, "nll": -5.813992023468018, "elapsed_s": 230.5}57{"event": "train", "step": 1100, "loss": -6.053476810455322, "grad_norm": 0.17538726329803467, "lr": 0.0002643227279408728, "ex_per_s": 303.2364889677272, "data_frac": 0.055938807975898765, "mem_mb": 91.948032, "nll": -6.053476810455322, "elapsed_s": 232.2}58{"event": "train", "step": 1120, "loss": -6.052066326141357, "grad_norm": 0.32630103826522827, "lr": 0.000262797402646797, "ex_per_s": 306.53833591909967, "data_frac": 0.055584098498372016, "mem_mb": 91.948032, "nll": -6.052066326141357, "elapsed_s": 233.9}59{"event": "train", "step": 1140, "loss": -6.022304058074951, "grad_norm": 0.3433881998062134, "lr": 0.0002612453405193386, "ex_per_s": 309.79659224600755, "data_frac": 0.05523491759557708, "mem_mb": 91.948032, "nll": -6.022304058074951, "elapsed_s": 235.6}60{"event": "train", "step": 1160, "loss": -5.970058917999268, "grad_norm": 0.46782079339027405, "lr": 0.0002596669658767859, "ex_per_s": 313.0068009192088, "data_frac": 0.05488783401385843, "mem_mb": 91.948032, "nll": -5.970058917999268, "elapsed_s": 237.2}61{"event": "train", "step": 1180, "loss": -5.957174301147461, "grad_norm": 0.1890617161989212, "lr": 0.00025806271023100637, "ex_per_s": 316.17558297541837, "data_frac": 0.054548451065104774, "mem_mb": 91.948032, "nll": -5.957174301147461, "elapsed_s": 238.9}62{"event": "train", "step": 1200, "loss": -5.946429252624512, "grad_norm": 0.21117177605628967, "lr": 0.00025643301216947504, "ex_per_s": 319.30011142655115, "data_frac": 0.054213779279813946, "mem_mb": 91.948032, "nll": -5.946429252624512, "elapsed_s": 240.6}63{"event": "train", "step": 1220, "loss": -6.033935546875, "grad_norm": 0.6250803470611572, "lr": 0.0002547783172353695, "ex_per_s": 322.38200541721164, "data_frac": 0.053881403963808035, "mem_mb": 91.948032, "nll": -6.033935546875, "elapsed_s": 242.2}64{"event": "train", "step": 1240, "loss": -6.059999465942383, "grad_norm": 0.08376622945070267, "lr": 0.00025309907780576277, "ex_per_s": 325.41368548034137, "data_frac": 0.05355586177862712, "mem_mb": 91.948032, "nll": -6.059999465942383, "elapsed_s": 243.9}65{"event": "train", "step": 1260, "loss": -6.056064128875732, "grad_norm": 0.2920980155467987, "lr": 0.0002513957529679477, "ex_per_s": 328.4127192088969, "data_frac": 0.05323481999348971, "mem_mb": 91.948032, "nll": -6.056064128875732, "elapsed_s": 245.6}66{"event": "train", "step": 1280, "loss": -6.046092987060547, "grad_norm": 0.6420236229896545, "lr": 0.00024966880839392746, "ex_per_s": 331.37011633605147, "data_frac": 0.052917500533140094, "mem_mb": 91.948032, "nll": -6.046092987060547, "elapsed_s": 247.3}67{"event": "train", "step": 1300, "loss": -5.865380764007568, "grad_norm": 0.8256497979164124, "lr": 0.0002479187162131052, "ex_per_s": 334.28668399466846, "data_frac": 0.052603370446419946, "mem_mb": 91.948032, "nll": -5.865380764007568, "elapsed_s": 248.9}68{"event": "train", "step": 1320, "loss": -5.824046611785889, "grad_norm": 0.2934054434299469, "lr": 0.00024614595488320833, "ex_per_s": 337.1683128906336, "data_frac": 0.0522947132401474, "mem_mb": 91.948032, "nll": -5.824046611785889, "elapsed_s": 250.6}69{"event": "train", "step": 1340, "loss": -5.951751708984375, "grad_norm": 0.19316071271896362, "lr": 0.00024435100905948387, "ex_per_s": 340.0148552996967, "data_frac": 0.051992289846599225, "mem_mb": 91.948032, "nll": -5.951751708984375, "elapsed_s": 252.3}70{"event": "train", "step": 1360, "loss": -6.038275718688965, "grad_norm": 0.694021999835968, "lr": 0.0002425343694621974, "ex_per_s": 342.823960242038, "data_frac": 0.051692295108353666, "mem_mb": 91.948032, "nll": -6.038275718688965, "elapsed_s": 253.9}71{"event": "train", "step": 1380, "loss": -6.067817687988281, "grad_norm": 0.1304801106452942, "lr": 0.00024069653274247583, "ex_per_s": 345.5956920994909, "data_frac": 0.05139615168911057, "mem_mb": 91.948032, "nll": -6.067817687988281, "elapsed_s": 255.6}72{"event": "train", "step": 1400, "loss": -5.795437335968018, "grad_norm": 0.27846094965934753, "lr": 0.0002388380013465276, "ex_per_s": 348.33300197283876, "data_frac": 0.051104370310209306, "mem_mb": 91.948032, "nll": -5.795437335968018, "elapsed_s": 257.3}73{"event": "train", "step": 1420, "loss": -5.990711212158203, "grad_norm": 0.7294551134109497, "lr": 0.00023695928337827914, "ex_per_s": 351.0358603785843, "data_frac": 0.05081786981110184, "mem_mb": 91.948032, "nll": -5.990711212158203, "elapsed_s": 258.9}74{"event": "train", "step": 1440, "loss": -5.744258880615234, "grad_norm": 0.14301711320877075, "lr": 0.00023506089246046403, "ex_per_s": 353.70252005288035, "data_frac": 0.050534034480068116, "mem_mb": 91.948032, "nll": -5.744258880615234, "elapsed_s": 260.6}75{"event": "train", "step": 1460, "loss": -5.809722423553467, "grad_norm": 0.353193998336792, "lr": 0.00023314334759420388, "ex_per_s": 356.3398693504559, "data_frac": 0.050253887626146194, "mem_mb": 91.948032, "nll": -5.809722423553467, "elapsed_s": 262.3}76{"event": "train", "step": 1480, "loss": -6.047080039978027, "grad_norm": 0.1772068589925766, "lr": 0.0002312071730171186, "ex_per_s": 358.94278161525216, "data_frac": 0.04997871200007039, "mem_mb": 91.948032, "nll": -6.047080039978027, "elapsed_s": 263.9}77{"event": "train", "step": 1500, "loss": -5.962352752685547, "grad_norm": 0.6932603716850281, "lr": 0.0002292528980600049, "ex_per_s": 361.5218062024965, "data_frac": 0.049706794710773, "mem_mb": 91.948032, "nll": -5.962352752685547, "elapsed_s": 265.6}78{"event": "val", "step": 1500, "val_nll": -6.445987889542723, "best": -6.446348534925756, "bad": 1, "mem_mb": 91.948032, "elapsed_s": 328.9}79{"event": "train", "step": 1520, "loss": -5.951105117797852, "grad_norm": 0.624967098236084, "lr": 0.00022728105700212285, "ex_per_s": 294.2349142526732, "data_frac": 0.03995452327073143, "mem_mb": 91.948032, "nll": -5.951105117797852, "elapsed_s": 330.7}80{"event": "train", "step": 1540, "loss": -5.880123138427734, "grad_norm": 0.24063517153263092, "lr": 0.00022529218892512886, "ex_per_s": 296.60264045805684, "data_frac": 0.03978447773874328, "mem_mb": 91.948032, "nll": -5.880123138427734, "elapsed_s": 332.3}81{"event": "train", "step": 1560, "loss": -5.987843036651611, "grad_norm": 0.14150190353393555, "lr": 0.00022328683756569652, "ex_per_s": 298.94604281903815, "data_frac": 0.039616506679703886, "mem_mb": 91.948032, "nll": -5.987843036651611, "elapsed_s": 334.0}82{"event": "train", "step": 1580, "loss": -5.982827663421631, "grad_norm": 0.17644762992858887, "lr": 0.0002212655511668642, "ex_per_s": 301.26598369011526, "data_frac": 0.039449160127491036, "mem_mb": 91.948032, "nll": -5.982827663421631, "elapsed_s": 335.7}83{"event": "train", "step": 1600, "loss": -5.959133625030518, "grad_norm": 0.10829127579927444, "lr": 0.00021922888232815089, "ex_per_s": 303.5628307766297, "data_frac": 0.03928347521386161, "mem_mb": 91.948032, "nll": -5.959133625030518, "elapsed_s": 337.4}84{"event": "train", "step": 1620, "loss": -5.878755569458008, "grad_norm": 0.19909828901290894, "lr": 0.00021717738785448102, "ex_per_s": 305.83915002018045, "data_frac": 0.03911951274461745, "mem_mb": 91.948032, "nll": -5.878755569458008, "elapsed_s": 339.0}85{"event": "train", "step": 1640, "loss": -6.081871032714844, "grad_norm": 0.1201581060886383, "lr": 0.00021511162860395977, "ex_per_s": 308.0906552428067, "data_frac": 0.03896031612380188, "mem_mb": 91.948032, "nll": -6.081871032714844, "elapsed_s": 340.7}86{"event": "train", "step": 1660, "loss": -6.044800758361816, "grad_norm": 0.6925144791603088, "lr": 0.0002130321693345397, "ex_per_s": 310.322290913885, "data_frac": 0.03879904433297596, "mem_mb": 91.948032, "nll": -6.044800758361816, "elapsed_s": 342.4}87{"event": "train", "step": 1680, "loss": -5.868305206298828, "grad_norm": 0.16533441841602325, "lr": 0.0002109395785496218, "ex_per_s": 312.5286282006942, "data_frac": 0.038641410053974395, "mem_mb": 91.948032, "nll": -5.868305206298828, "elapsed_s": 344.1}88{"event": "train", "step": 1700, "loss": -5.9301018714904785, "grad_norm": 0.4428260326385498, "lr": 0.00020883442834263225, "ex_per_s": 314.7206882059636, "data_frac": 0.03848440792898302, "mem_mb": 91.948032, "nll": -5.9301018714904785, "elapsed_s": 345.7}89{"event": "train", "step": 1720, "loss": -5.884887218475342, "grad_norm": 0.9557400345802307, "lr": 0.00020671729424061788, "ex_per_s": 316.89702078759325, "data_frac": 0.038329509607394584, "mem_mb": 91.948032, "nll": -5.884887218475342, "elapsed_s": 347.4}90{"event": "train", "step": 1740, "loss": -6.2759928703308105, "grad_norm": 0.349621444940567, "lr": 0.00020458875504690275, "ex_per_s": 319.0468521529989, "data_frac": 0.03817575573634889, "mem_mb": 91.948032, "nll": -6.2759928703308105, "elapsed_s": 349.1}91{"event": "train", "step": 1760, "loss": -5.933294773101807, "grad_norm": 0.2620752155780792, "lr": 0.000202449392682849, "ex_per_s": 321.17967819457414, "data_frac": 0.03802345759640717, "mem_mb": 91.948032, "nll": -5.933294773101807, "elapsed_s": 350.8}92{"event": "train", "step": 1780, "loss": -5.870974540710449, "grad_norm": 0.4922235906124115, "lr": 0.00020029979202876563, "ex_per_s": 323.2882521334459, "data_frac": 0.03787200537356553, "mem_mb": 91.948032, "nll": -5.870974540710449, "elapsed_s": 352.4}93{"event": "train", "step": 1800, "loss": -6.054152488708496, "grad_norm": 0.8082236647605896, "lr": 0.00019814054076400787, "ex_per_s": 325.3787235833411, "data_frac": 0.03772162859664469, "mem_mb": 91.948032, "nll": -6.054152488708496, "elapsed_s": 354.1}94{"event": "train", "step": 1820, "loss": -5.9386091232299805, "grad_norm": 0.5874931812286377, "lr": 0.00019597222920631147, "ex_per_s": 327.4508204005412, "data_frac": 0.03757286915119504, "mem_mb": 91.948032, "nll": -5.9386091232299805, "elapsed_s": 355.8}95{"event": "train", "step": 1840, "loss": -6.074554443359375, "grad_norm": 0.38001736998558044, "lr": 0.0001937954501504058, "ex_per_s": 329.5015006503529, "data_frac": 0.03742567749119788, "mem_mb": 91.948032, "nll": -6.074554443359375, "elapsed_s": 357.4}96{"event": "train", "step": 1860, "loss": -6.050271987915039, "grad_norm": 0.312745600938797, "lr": 0.0001916107987059498, "ex_per_s": 331.5343379878604, "data_frac": 0.037280466803477366, "mem_mb": 91.948032, "nll": -6.050271987915039, "elapsed_s": 359.1}97{"event": "train", "step": 1880, "loss": -5.946768283843994, "grad_norm": 0.4819527566432953, "lr": 0.00018941887213483484, "ex_per_s": 333.55011964161207, "data_frac": 0.03713617890522809, "mem_mb": 91.948032, "nll": -5.946768283843994, "elapsed_s": 360.8}98{"event": "train", "step": 1900, "loss": -5.982052326202393, "grad_norm": 0.13743415474891663, "lr": 0.00018722026968789906, "ex_per_s": 335.54820352894507, "data_frac": 0.036993220911014646, "mem_mb": 91.948032, "nll": -5.982052326202393, "elapsed_s": 362.4}99{"event": "train", "step": 1920, "loss": -5.9679694175720215, "grad_norm": 0.33721476793289185, "lr": 0.00018501559244109864, "ex_per_s": 337.5231876855683, "data_frac": 0.03685250362500961, "mem_mb": 91.948032, "nll": -5.9679694175720215, "elapsed_s": 364.1}100{"event": "train", "step": 1940, "loss": -6.051275730133057, "grad_norm": 0.4332176446914673, "lr": 0.00018280544313117932, "ex_per_s": 339.4822624028315, "data_frac": 0.03671205787342078, "mem_mb": 91.948032, "nll": -6.051275730133057, "elapsed_s": 365.8}101{"event": "train", "step": 1960, "loss": -5.941980838775635, "grad_norm": 0.1281217634677887, "lr": 0.00018059042599089415, "ex_per_s": 341.42556080435617, "data_frac": 0.036574410782746834, "mem_mb": 91.948032, "nll": -5.941980838775635, "elapsed_s": 367.4}102{"event": "train", "step": 1980, "loss": -5.967869281768799, "grad_norm": 0.7233005166053772, "lr": 0.00017837114658381247, "ex_per_s": 343.35271455872817, "data_frac": 0.03643688479576107, "mem_mb": 91.948032, "nll": -5.967869281768799, "elapsed_s": 369.1}103{"event": "train", "step": 2000, "loss": -5.916905879974365, "grad_norm": 0.3420059084892273, "lr": 0.00017614821163876486, "ex_per_s": 345.26284524689663, "data_frac": 0.03630264704497054, "mem_mb": 91.948032, "nll": -5.916905879974365, "elapsed_s": 370.8}104{"event": "val", "step": 2000, "val_nll": -6.449990747742488, "best": -6.449990747742488, "bad": 0, "mem_mb": 91.948032, "elapsed_s": 434.2}105{"event": "train", "step": 2020, "loss": -5.929187297821045, "grad_norm": 0.43268483877182007, "lr": 0.00017392222888396941, "ex_per_s": 296.61795733512565, "data_frac": 0.030903198815699244, "mem_mb": 91.948032, "nll": -5.929187297821045, "elapsed_s": 435.9}106{"event": "train", "step": 2040, "loss": -5.873630523681641, "grad_norm": 0.2231506109237671, "lr": 0.000171693806880885, "ex_per_s": 298.4080179541199, "data_frac": 0.03080908074546427, "mem_mb": 91.948032, "nll": -5.873630523681641, "elapsed_s": 437.6}107{"event": "train", "step": 2060, "loss": -6.145679473876953, "grad_norm": 0.15180470049381256, "lr": 0.0001694635548578366, "ex_per_s": 300.1850637743754, "data_frac": 0.030715353359698255, "mem_mb": 91.948032, "nll": -6.145679473876953, "elapsed_s": 439.2}108{"event": "train", "step": 2080, "loss": -5.966085910797119, "grad_norm": 0.3131261169910431, "lr": 0.0001672320825434584, "ex_per_s": 301.94713596915665, "data_frac": 0.030621394432191616, "mem_mb": 91.948032, "nll": -5.966085910797119, "elapsed_s": 440.9}109{"event": "train", "step": 2100, "loss": -5.9080424308776855, "grad_norm": 1.124796748161316, "lr": 0.000165, "ex_per_s": 303.69728334635505, "data_frac": 0.030528611889949587, "mem_mb": 91.948032, "nll": -5.9080424308776855, "elapsed_s": 442.6}110{"event": "train", "step": 2120, "loss": -6.151602745056152, "grad_norm": 0.542772650718689, "lr": 0.0001627679174565416, "ex_per_s": 305.4327366124995, "data_frac": 0.030436718078994854, "mem_mb": 91.948032, "nll": -6.151602745056152, "elapsed_s": 444.3}111{"event": "train", "step": 2140, "loss": -5.981777191162109, "grad_norm": 0.15379057824611664, "lr": 0.00016053644514216336, "ex_per_s": 307.1546965731857, "data_frac": 0.030347887994547317, "mem_mb": 91.948032, "nll": -5.981777191162109, "elapsed_s": 445.9}112{"event": "train", "step": 2160, "loss": -5.926473140716553, "grad_norm": 0.5294952988624573, "lr": 0.00015830619311911498, "ex_per_s": 308.8657418019444, "data_frac": 0.03025720586785487, "mem_mb": 91.948032, "nll": -5.926473140716553, "elapsed_s": 447.6}113{"event": "train", "step": 2180, "loss": -6.0829081535339355, "grad_norm": 0.2427883744239807, "lr": 0.00015607777111603058, "ex_per_s": 310.5648697045539, "data_frac": 0.030167129875054988, "mem_mb": 91.948032, "nll": -6.0829081535339355, "elapsed_s": 449.3}114{"event": "train", "step": 2200, "loss": -5.961925506591797, "grad_norm": 0.4604598879814148, "lr": 0.00015385178836123514, "ex_per_s": 312.2506452313981, "data_frac": 0.03007759434149455, "mem_mb": 91.948032, "nll": -5.961925506591797, "elapsed_s": 451.0}115{"event": "train", "step": 2220, "loss": -5.827727794647217, "grad_norm": 0.7004027366638184, "lr": 0.00015162885341618752, "ex_per_s": 313.9234633021921, "data_frac": 0.02999022479095265, "mem_mb": 91.948032, "nll": -5.827727794647217, "elapsed_s": 452.6}116{"event": "train", "step": 2240, "loss": -5.962932586669922, "grad_norm": 0.6406000256538391, "lr": 0.00014940957400910585, "ex_per_s": 315.58492875680014, "data_frac": 0.02990295514250746, "mem_mb": 91.948032, "nll": -5.962932586669922, "elapsed_s": 454.3}117{"event": "train", "step": 2260, "loss": -5.738107204437256, "grad_norm": 0.8575554490089417, "lr": 0.00014719455686882068, "ex_per_s": 317.23612441995573, "data_frac": 0.029815919122350657, "mem_mb": 91.948032, "nll": -5.738107204437256, "elapsed_s": 456.0}118{"event": "train", "step": 2280, "loss": -6.012794017791748, "grad_norm": 0.33007824420928955, "lr": 0.0001449844075589013, "ex_per_s": 318.87666188302484, "data_frac": 0.02973024163561879, "mem_mb": 91.948032, "nll": -6.012794017791748, "elapsed_s": 457.6}119{"event": "train", "step": 2300, "loss": -6.004396915435791, "grad_norm": 0.10653365403413773, "lr": 0.0001427797303121009, "ex_per_s": 320.50420120151335, "data_frac": 0.029645008443857708, "mem_mb": 91.948032, "nll": -6.004396915435791, "elapsed_s": 459.3}120{"event": "train", "step": 2320, "loss": -6.06927490234375, "grad_norm": 0.4123423397541046, "lr": 0.00014058112786516518, "ex_per_s": 322.120155180158, "data_frac": 0.029560799795922896, "mem_mb": 91.948032, "nll": -6.06927490234375, "elapsed_s": 461.0}121{"event": "train", "step": 2340, "loss": -5.855721473693848, "grad_norm": 0.34750133752822876, "lr": 0.00013838920129405016, "ex_per_s": 323.7265308135143, "data_frac": 0.029476317489882588, "mem_mb": 91.948032, "nll": -5.855721473693848, "elapsed_s": 462.7}122{"event": "train", "step": 2360, "loss": -5.928738117218018, "grad_norm": 0.2551330327987671, "lr": 0.0001362045498495942, "ex_per_s": 325.3192335698259, "data_frac": 0.02939357638063458, "mem_mb": 91.948032, "nll": -5.928738117218018, "elapsed_s": 464.3}123{"event": "train", "step": 2380, "loss": -5.9308295249938965, "grad_norm": 0.10052561014890671, "lr": 0.00013402777079368853, "ex_per_s": 326.9005281767776, "data_frac": 0.029312007072553, "mem_mb": 91.948032, "nll": -5.9308295249938965, "elapsed_s": 466.0}124{"event": "train", "step": 2400, "loss": -5.958132743835449, "grad_norm": 0.5935574173927307, "lr": 0.00013185945923599213, "ex_per_s": 328.4731516933894, "data_frac": 0.02923007018563365, "mem_mb": 91.948032, "nll": -5.958132743835449, "elapsed_s": 467.7}125{"event": "train", "step": 2420, "loss": -6.000174522399902, "grad_norm": 0.5294894576072693, "lr": 0.00012970020797123432, "ex_per_s": 330.0335947907117, "data_frac": 0.029149189891659513, "mem_mb": 91.948032, "nll": -6.000174522399902, "elapsed_s": 469.3}126{"event": "train", "step": 2440, "loss": -6.096232891082764, "grad_norm": 0.3721625804901123, "lr": 0.000127550607317151, "ex_per_s": 331.5842406695748, "data_frac": 0.029068296529510436, "mem_mb": 91.948032, "nll": -6.096232891082764, "elapsed_s": 471.0}127{"event": "train", "step": 2460, "loss": -6.136342525482178, "grad_norm": 0.11310028284788132, "lr": 0.00012541124495309727, "ex_per_s": 333.11993758546043, "data_frac": 0.028987411249081465, "mem_mb": 91.948032, "nll": -6.136342525482178, "elapsed_s": 472.7}128{"event": "train", "step": 2480, "loss": -5.947104454040527, "grad_norm": 0.11823279410600662, "lr": 0.00012328270575938212, "ex_per_s": 334.6446953738477, "data_frac": 0.028907095567792463, "mem_mb": 91.948032, "nll": -5.947104454040527, "elapsed_s": 474.3}129{"event": "train", "step": 2500, "loss": -5.755773067474365, "grad_norm": 0.5295006036758423, "lr": 0.00012116557165736775, "ex_per_s": 336.15959862808216, "data_frac": 0.028827717862074158, "mem_mb": 91.948032, "nll": -5.755773067474365, "elapsed_s": 476.0}130{"event": "val", "step": 2500, "val_nll": -6.4500738139905165, "best": -6.4500738139905165, "bad": 0, "mem_mb": 91.948032, "elapsed_s": 539.2}131{"event": "train", "step": 2520, "loss": -5.95928955078125, "grad_norm": 0.8842701315879822, "lr": 0.00011906042145037817, "ex_per_s": 298.19843770693296, "data_frac": 0.02538935173323619, "mem_mb": 91.948032, "nll": -5.95928955078125, "elapsed_s": 540.9}132{"event": "train", "step": 2540, "loss": -6.033793926239014, "grad_norm": 0.27164581418037415, "lr": 0.00011696783066546026, "ex_per_s": 299.63645297973056, "data_frac": 0.02532945560806979, "mem_mb": 91.948032, "nll": -6.033793926239014, "elapsed_s": 542.6}133{"event": "train", "step": 2560, "loss": -5.948520660400391, "grad_norm": 0.15622012317180634, "lr": 0.00011488837139604017, "ex_per_s": 301.0657292138652, "data_frac": 0.02527025699327992, "mem_mb": 91.948032, "nll": -5.948520660400391, "elapsed_s": 544.2}134{"event": "train", "step": 2580, "loss": -5.970556735992432, "grad_norm": 0.36877864599227905, "lr": 0.00011282261214551898, "ex_per_s": 302.4836651047529, "data_frac": 0.02521350381264276, "mem_mb": 91.948032, "nll": -5.970556735992432, "elapsed_s": 545.9}135{"event": "train", "step": 2600, "loss": -5.851034164428711, "grad_norm": 0.1486823707818985, "lr": 0.00011077111767184915, "ex_per_s": 303.8952560869882, "data_frac": 0.025154928663053373, "mem_mb": 91.948032, "nll": -5.851034164428711, "elapsed_s": 547.6}136{"event": "train", "step": 2620, "loss": -5.890315055847168, "grad_norm": 1.1503969430923462, "lr": 0.00010873444883313583, "ex_per_s": 305.29729810897896, "data_frac": 0.025097126719556627, "mem_mb": 91.948032, "nll": -5.890315055847168, "elapsed_s": 549.3}137{"event": "train", "step": 2640, "loss": -5.987398147583008, "grad_norm": 0.33095428347587585, "lr": 0.00010671316243430345, "ex_per_s": 306.68999290883454, "data_frac": 0.025039341462304573, "mem_mb": 91.948032, "nll": -5.987398147583008, "elapsed_s": 551.0}138{"event": "train", "step": 2660, "loss": -5.821589946746826, "grad_norm": 0.0990462452173233, "lr": 0.00010470781107487112, "ex_per_s": 308.07604698031713, "data_frac": 0.024981538120448007, "mem_mb": 91.948032, "nll": -5.821589946746826, "elapsed_s": 552.6}139{"event": "train", "step": 2680, "loss": -5.959189414978027, "grad_norm": 0.32004880905151367, "lr": 0.00010271894299787715, "ex_per_s": 309.4532681073981, "data_frac": 0.024923897209036344, "mem_mb": 91.948032, "nll": -5.959189414978027, "elapsed_s": 554.3}140{"event": "train", "step": 2700, "loss": -6.135731220245361, "grad_norm": 0.5906044840812683, "lr": 0.00010074710193999506, "ex_per_s": 310.8223558030503, "data_frac": 0.02486693457600441, "mem_mb": 91.948032, "nll": -6.135731220245361, "elapsed_s": 556.0}141{"event": "train", "step": 2720, "loss": -5.880568504333496, "grad_norm": 0.6047245264053345, "lr": 9.879282698288144e-05, "ex_per_s": 312.1830500498533, "data_frac": 0.024810152565794285, "mem_mb": 91.948032, "nll": -5.880568504333496, "elapsed_s": 557.7}142{"event": "train", "step": 2740, "loss": -6.12882137298584, "grad_norm": 0.09843790531158447, "lr": 9.685665240579613e-05, "ex_per_s": 313.5359112987255, "data_frac": 0.024753243230165606, "mem_mb": 91.948032, "nll": -6.12882137298584, "elapsed_s": 559.3}143{"event": "train", "step": 2760, "loss": -5.962754249572754, "grad_norm": 0.22729913890361786, "lr": 9.4939107539536e-05, "ex_per_s": 314.88180384082034, "data_frac": 0.024697249128137172, "mem_mb": 91.948032, "nll": -5.962754249572754, "elapsed_s": 561.0}144{"event": "train", "step": 2780, "loss": -6.013645172119141, "grad_norm": 0.9309803247451782, "lr": 9.304071662172087e-05, "ex_per_s": 316.2196645303833, "data_frac": 0.024641970986614237, "mem_mb": 91.948032, "nll": -6.013645172119141, "elapsed_s": 562.7}145{"event": "train", "step": 2800, "loss": -5.980140209197998, "grad_norm": 0.6617916822433472, "lr": 9.116199865347241e-05, "ex_per_s": 317.55036486632025, "data_frac": 0.024587403568900112, "mem_mb": 91.948032, "nll": -5.980140209197998, "elapsed_s": 564.4}146{"event": "train", "step": 2820, "loss": -6.007987976074219, "grad_norm": 0.4184640944004059, "lr": 8.930346725752419e-05, "ex_per_s": 318.86976392968415, "data_frac": 0.02453506010731683, "mem_mb": 91.948032, "nll": -6.007987976074219, "elapsed_s": 566.0}147{"event": "train", "step": 2840, "loss": -5.883373260498047, "grad_norm": 0.20720839500427246, "lr": 8.746563053780262e-05, "ex_per_s": 320.1858266534677, "data_frac": 0.0244808749534041, "mem_mb": 91.948032, "nll": -5.883373260498047, "elapsed_s": 567.7}148{"event": "train", "step": 2860, "loss": -6.053739547729492, "grad_norm": 0.08719151467084885, "lr": 8.564899094051614e-05, "ex_per_s": 321.49428634800796, "data_frac": 0.024427314600994704, "mem_mb": 91.948032, "nll": -6.053739547729492, "elapsed_s": 569.4}149{"event": "train", "step": 2880, "loss": -5.773558139801025, "grad_norm": 0.3197888433933258, "lr": 8.385404511679161e-05, "ex_per_s": 322.79609676346763, "data_frac": 0.024373765626427093, "mem_mb": 91.948032, "nll": -5.773558139801025, "elapsed_s": 571.1}150{"event": "train", "step": 2900, "loss": -5.797021865844727, "grad_norm": 0.566809892654419, "lr": 8.208128378689484e-05, "ex_per_s": 324.0901841785524, "data_frac": 0.024321460899830907, "mem_mb": 91.948032, "nll": -5.797021865844727, "elapsed_s": 572.7}151{"event": "train", "step": 2920, "loss": -5.971386909484863, "grad_norm": 0.6449095010757446, "lr": 8.03311916060725e-05, "ex_per_s": 325.3788636289074, "data_frac": 0.02426974877348726, "mem_mb": 91.948032, "nll": -5.971386909484863, "elapsed_s": 574.4}152{"event": "train", "step": 2940, "loss": -5.900071144104004, "grad_norm": 0.5531981587409973, "lr": 7.860424703205225e-05, "ex_per_s": 326.6604282711779, "data_frac": 0.024217858760072344, "mem_mb": 91.948032, "nll": -5.900071144104004, "elapsed_s": 576.1}153{"event": "train", "step": 2960, "loss": -5.940980911254883, "grad_norm": 0.2742380201816559, "lr": 7.690092219423722e-05, "ex_per_s": 327.9336012528712, "data_frac": 0.02416580392670162, "mem_mb": 91.948032, "nll": -5.940980911254883, "elapsed_s": 577.7}154{"event": "train", "step": 2980, "loss": -5.89277982711792, "grad_norm": 0.4898792803287506, "lr": 7.522168276463048e-05, "ex_per_s": 329.2010495093839, "data_frac": 0.024114591907867106, "mem_mb": 91.948032, "nll": -5.89277982711792, "elapsed_s": 579.4}155{"event": "train", "step": 3000, "loss": -5.997138023376465, "grad_norm": 0.2165524810552597, "lr": 7.356698783052496e-05, "ex_per_s": 330.45821774322104, "data_frac": 0.02406408119908665, "mem_mb": 91.948032, "nll": -5.997138023376465, "elapsed_s": 581.1}156{"event": "val", "step": 3000, "val_nll": -6.450102870222967, "best": -6.450102870222967, "bad": 0, "mem_mb": 91.948032, "elapsed_s": 644.1}157{"event": "train", "step": 3020, "loss": -5.961453437805176, "grad_norm": 0.9912112951278687, "lr": 7.193728976899362e-05, "ex_per_s": 299.2929188509756, "data_frac": 0.021667232145973037, "mem_mb": 91.948032, "nll": -5.961453437805176, "elapsed_s": 645.8}158{"event": "train", "step": 3040, "loss": -5.8605828285217285, "grad_norm": 0.7164281010627747, "lr": 7.033303412321404e-05, "ex_per_s": 300.49331761716337, "data_frac": 0.02162687697889341, "mem_mb": 91.948032, "nll": -5.8605828285217285, "elapsed_s": 647.5}159{"event": "train", "step": 3060, "loss": -6.070032596588135, "grad_norm": 0.2553955912590027, "lr": 6.875465948066139e-05, "ex_per_s": 301.68801697375125, "data_frac": 0.02158603689804281, "mem_mb": 91.948032, "nll": -6.070032596588135, "elapsed_s": 649.2}160{"event": "train", "step": 3080, "loss": -5.969526290893555, "grad_norm": 0.24710893630981445, "lr": 6.720259735320299e-05, "ex_per_s": 302.8765338664485, "data_frac": 0.021545665880571975, "mem_mb": 91.948032, "nll": -5.969526290893555, "elapsed_s": 650.9}161{"event": "train", "step": 3100, "loss": -6.015357494354248, "grad_norm": 0.2585790157318115, "lr": 6.567727205912722e-05, "ex_per_s": 304.05926075743406, "data_frac": 0.021505829290179357, "mem_mb": 91.948032, "nll": -6.015357494354248, "elapsed_s": 652.5}162{"event": "train", "step": 3120, "loss": -6.092260360717773, "grad_norm": 0.6582404375076294, "lr": 6.417910060713926e-05, "ex_per_s": 305.2368909933527, "data_frac": 0.021465697202597916, "mem_mb": 91.948032, "nll": -6.092260360717773, "elapsed_s": 654.2}163{"event": "train", "step": 3140, "loss": -6.0110602378845215, "grad_norm": 0.12471671402454376, "lr": 6.270849258235485e-05, "ex_per_s": 306.40796557345675, "data_frac": 0.02142617692050555, "mem_mb": 91.948032, "nll": -6.0110602378845215, "elapsed_s": 655.9}164{"event": "train", "step": 3160, "loss": -6.02394962310791, "grad_norm": 0.14500701427459717, "lr": 6.126585003432406e-05, "ex_per_s": 307.5739243640572, "data_frac": 0.0213868729763994, "mem_mb": 91.948032, "nll": -6.02394962310791, "elapsed_s": 657.6}165{"event": "train", "step": 3180, "loss": -6.095670223236084, "grad_norm": 0.15215188264846802, "lr": 5.985156736711483e-05, "ex_per_s": 308.73453036560056, "data_frac": 0.02134750214380185, "mem_mb": 91.948032, "nll": -6.095670223236084, "elapsed_s": 659.3}166{"event": "train", "step": 3200, "loss": -6.002058982849121, "grad_norm": 0.30492720007896423, "lr": 5.846603123148688e-05, "ex_per_s": 309.8895732457196, "data_frac": 0.021308411736064507, "mem_mb": 91.948032, "nll": -6.002058982849121, "elapsed_s": 660.9}167{"event": "train", "step": 3220, "loss": -5.919881343841553, "grad_norm": 0.09206683188676834, "lr": 5.710962041918516e-05, "ex_per_s": 311.03855472950795, "data_frac": 0.02126955187557946, "mem_mb": 91.948032, "nll": -5.919881343841553, "elapsed_s": 662.6}168{"event": "train", "step": 3240, "loss": -5.799683570861816, "grad_norm": 0.3701797425746918, "lr": 5.578270575938211e-05, "ex_per_s": 312.1811805180326, "data_frac": 0.02123080497765669, "mem_mb": 91.948032, "nll": -5.799683570861816, "elapsed_s": 664.3}169{"event": "train", "step": 3260, "loss": -5.963250637054443, "grad_norm": 0.4359613060951233, "lr": 5.448565001729653e-05, "ex_per_s": 313.31783027670235, "data_frac": 0.021193207571420927, "mem_mb": 91.948032, "nll": -5.963250637054443, "elapsed_s": 665.9}170{"event": "train", "step": 3280, "loss": -6.0186872482299805, "grad_norm": 0.39252445101737976, "lr": 5.3218807795017296e-05, "ex_per_s": 314.4494570806242, "data_frac": 0.021155831285560926, "mem_mb": 91.948032, "nll": -6.0186872482299805, "elapsed_s": 667.6}171{"event": "train", "step": 3300, "loss": -5.8029303550720215, "grad_norm": 0.2731097936630249, "lr": 5.198252543455864e-05, "ex_per_s": 315.5762497699815, "data_frac": 0.02111820646999918, "mem_mb": 91.948032, "nll": -5.8029303550720215, "elapsed_s": 669.3}172{"event": "train", "step": 3320, "loss": -5.996059417724609, "grad_norm": 0.3443026542663574, "lr": 5.07771409231737e-05, "ex_per_s": 316.69717519185116, "data_frac": 0.021080569952996845, "mem_mb": 91.948032, "nll": -5.996059417724609, "elapsed_s": 671.0}173{"event": "train", "step": 3340, "loss": -5.936288356781006, "grad_norm": 0.23905493319034576, "lr": 4.960298380095221e-05, "ex_per_s": 317.8133078804331, "data_frac": 0.021043375293805712, "mem_mb": 91.948032, "nll": -5.936288356781006, "elapsed_s": 672.6}174{"event": "train", "step": 3360, "loss": -5.90895414352417, "grad_norm": 0.4398115873336792, "lr": 4.846037507072751e-05, "ex_per_s": 318.9257952882428, "data_frac": 0.02100614071534085, "mem_mb": 91.948032, "nll": -5.90895414352417, "elapsed_s": 674.3}175{"event": "train", "step": 3380, "loss": -6.121904373168945, "grad_norm": 0.5573644638061523, "lr": 4.7349627110317556e-05, "ex_per_s": 320.0330045980612, "data_frac": 0.020969056687407853, "mem_mb": 91.948032, "nll": -6.121904373168945, "elapsed_s": 676.0}176{"event": "train", "step": 3400, "loss": -5.8813910484313965, "grad_norm": 0.42699578404426575, "lr": 4.6271043587123986e-05, "ex_per_s": 321.1341437201412, "data_frac": 0.020932883674727894, "mem_mb": 91.948032, "nll": -5.8813910484313965, "elapsed_s": 677.6}177{"event": "train", "step": 3420, "loss": -5.83729362487793, "grad_norm": 0.2634403705596924, "lr": 4.52249193751124e-05, "ex_per_s": 322.2318080916584, "data_frac": 0.020896499173833023, "mem_mb": 91.948032, "nll": -5.83729362487793, "elapsed_s": 679.3}178{"event": "train", "step": 3440, "loss": -5.91678524017334, "grad_norm": 0.1849982738494873, "lr": 4.4211540474196926e-05, "ex_per_s": 323.3231780458644, "data_frac": 0.020860593894311368, "mem_mb": 91.948032, "nll": -5.91678524017334, "elapsed_s": 681.0}179{"event": "train", "step": 3460, "loss": -6.058093070983887, "grad_norm": 0.11190175265073776, "lr": 4.3231183932050546e-05, "ex_per_s": 324.41117778671827, "data_frac": 0.020824513809130963, "mem_mb": 91.948032, "nll": -6.058093070983887, "elapsed_s": 682.6}180{"event": "train", "step": 3480, "loss": -5.916826248168945, "grad_norm": 0.08897316455841064, "lr": 4.228411776836306e-05, "ex_per_s": 325.4933522025635, "data_frac": 0.020789157769865126, "mem_mb": 91.948032, "nll": -5.916826248168945, "elapsed_s": 684.3}181{"event": "train", "step": 3500, "loss": -6.039107799530029, "grad_norm": 0.12224172055721283, "lr": 4.137060090156723e-05, "ex_per_s": 326.56877218799104, "data_frac": 0.020753957019832122, "mem_mb": 91.948032, "nll": -6.039107799530029, "elapsed_s": 686.0}182{"event": "val", "step": 3500, "val_nll": -6.450933618499165, "best": -6.450933618499165, "bad": 0, "mem_mb": 91.948032, "elapsed_s": 749.1}183{"event": "train", "step": 3520, "loss": -5.9413957595825195, "grad_norm": 0.38463830947875977, "lr": 4.049088307805306e-05, "ex_per_s": 300.0702551527161, "data_frac": 0.018975322865389947, "mem_mb": 91.948032, "nll": -5.9413957595825195, "elapsed_s": 750.8}184{"event": "train", "step": 3540, "loss": -6.139245986938477, "grad_norm": 0.5055034160614014, "lr": 3.964520480388956e-05, "ex_per_s": 301.1026113601539, "data_frac": 0.018947034299472976, "mem_mb": 91.948032, "nll": -6.139245986938477, "elapsed_s": 752.5}185{"event": "train", "step": 3560, "loss": -6.142125129699707, "grad_norm": 0.21315120160579681, "lr": 3.88337972790728e-05, "ex_per_s": 302.13125166940335, "data_frac": 0.018918513794276264, "mem_mb": 91.948032, "nll": -6.142125129699707, "elapsed_s": 754.2}186{"event": "train", "step": 3580, "loss": -5.906026840209961, "grad_norm": 0.16623954474925995, "lr": 3.805688233431824e-05, "ex_per_s": 303.1548305834114, "data_frac": 0.018890159361406474, "mem_mb": 91.948032, "nll": -5.906026840209961, "elapsed_s": 755.8}187{"event": "train", "step": 3600, "loss": -6.099078178405762, "grad_norm": 0.22672276198863983, "lr": 3.7314672370414325e-05, "ex_per_s": 304.1735153521383, "data_frac": 0.018861718335593625, "mem_mb": 91.948032, "nll": -6.099078178405762, "elapsed_s": 757.5}188{"event": "train", "step": 3620, "loss": -5.910958766937256, "grad_norm": 0.9805033802986145, "lr": 3.660737030015427e-05, "ex_per_s": 305.18822404124086, "data_frac": 0.018833349911978867, "mem_mb": 91.948032, "nll": -5.910958766937256, "elapsed_s": 759.2}189{"event": "train", "step": 3640, "loss": -6.055211067199707, "grad_norm": 0.4925552010536194, "lr": 3.5935169492861716e-05, "ex_per_s": 306.19869164367634, "data_frac": 0.018804960359273277, "mem_mb": 91.948032, "nll": -6.055211067199707, "elapsed_s": 760.9}190{"event": "train", "step": 3660, "loss": -5.841465473175049, "grad_norm": 0.5924697518348694, "lr": 3.52982537215255e-05, "ex_per_s": 307.20386733224046, "data_frac": 0.01877691518357768, "mem_mb": 91.948032, "nll": -5.841465473175049, "elapsed_s": 762.5}191{"event": "train", "step": 3680, "loss": -5.994565010070801, "grad_norm": 0.15795479714870453, "lr": 3.469679711255794e-05, "ex_per_s": 308.20525864749334, "data_frac": 0.018749265007852515, "mem_mb": 91.948032, "nll": -5.994565010070801, "elapsed_s": 764.2}192{"event": "train", "step": 3700, "loss": -6.068084716796875, "grad_norm": 0.8386463522911072, "lr": 3.41309640981904e-05, "ex_per_s": 309.20273366656545, "data_frac": 0.018721642931054036, "mem_mb": 91.948032, "nll": -6.068084716796875, "elapsed_s": 765.9}193{"event": "train", "step": 3720, "loss": -5.948995590209961, "grad_norm": 0.6138898134231567, "lr": 3.3600909371519225e-05, "ex_per_s": 310.19471382209724, "data_frac": 0.018695512951922736, "mem_mb": 91.948032, "nll": -5.948995590209961, "elapsed_s": 767.6}194{"event": "train", "step": 3740, "loss": -6.189114570617676, "grad_norm": 0.1650201380252838, "lr": 3.31067778442141e-05, "ex_per_s": 311.1848099955434, "data_frac": 0.018667884432772047, "mem_mb": 91.948032, "nll": -6.189114570617676, "elapsed_s": 769.2}195{"event": "train", "step": 3760, "loss": -6.1484246253967285, "grad_norm": 0.23271554708480835, "lr": 3.2648704606900735e-05, "ex_per_s": 312.170973633148, "data_frac": 0.01864062953152697, "mem_mb": 91.948032, "nll": -6.1484246253967285, "elapsed_s": 770.9}196{"event": "train", "step": 3780, "loss": -5.964338302612305, "grad_norm": 0.3947637379169464, "lr": 3.222681489222838e-05, "ex_per_s": 313.1511405189989, "data_frac": 0.018613809263669335, "mem_mb": 91.948032, "nll": -5.964338302612305, "elapsed_s": 772.6}197{"event": "train", "step": 3800, "loss": -6.099814414978027, "grad_norm": 0.12338342517614365, "lr": 3.1841224040632484e-05, "ex_per_s": 314.1279252704734, "data_frac": 0.01858646383963014, "mem_mb": 91.948032, "nll": -6.099814414978027, "elapsed_s": 774.3}198{"event": "train", "step": 3820, "loss": -5.969507217407227, "grad_norm": 0.19360361993312836, "lr": 3.149203746880175e-05, "ex_per_s": 315.1011808460545, "data_frac": 0.018559534306269172, "mem_mb": 91.948032, "nll": -5.969507217407227, "elapsed_s": 775.9}199{"event": "train", "step": 3840, "loss": -5.874866962432861, "grad_norm": 0.6047665476799011, "lr": 3.117935064085835e-05, "ex_per_s": 316.07070700037445, "data_frac": 0.018532725943427253, "mem_mb": 91.948032, "nll": -5.874866962432861, "elapsed_s": 777.6}200{"event": "train", "step": 3860, "loss": -6.016587734222412, "grad_norm": 0.17599019408226013, "lr": 3.090324904225885e-05, "ex_per_s": 317.0354075956646, "data_frac": 0.018505912872350577, "mem_mb": 91.948032, "nll": -6.016587734222412, "elapsed_s": 779.3}201{"event": "train", "step": 3880, "loss": -5.891308307647705, "grad_norm": 0.14324964582920074, "lr": 3.0663808156423456e-05, "ex_per_s": 317.99681637770703, "data_frac": 0.01847955485134065, "mem_mb": 91.948032, "nll": -5.891308307647705, "elapsed_s": 780.9}202{"event": "train", "step": 3900, "loss": -6.1838178634643555, "grad_norm": 0.6464358568191528, "lr": 3.046109344409957e-05, "ex_per_s": 318.9536028164335, "data_frac": 0.01845333431183711, "mem_mb": 91.948032, "nll": -6.1838178634643555, "elapsed_s": 782.6}203{"event": "train", "step": 3920, "loss": -6.040684700012207, "grad_norm": 0.13889116048812866, "lr": 3.0295160325465435e-05, "ex_per_s": 319.90632046469943, "data_frac": 0.01842744664568358, "mem_mb": 91.948032, "nll": -6.040684700012207, "elapsed_s": 784.3}204{"event": "train", "step": 3940, "loss": -5.922958850860596, "grad_norm": 0.13983269035816193, "lr": 3.016605416497886e-05, "ex_per_s": 320.85540454722, "data_frac": 0.018401842208733154, "mem_mb": 91.948032, "nll": -5.922958850860596, "elapsed_s": 785.9}205{"event": "train", "step": 3960, "loss": -6.069833755493164, "grad_norm": 0.19217656552791595, "lr": 3.0073810258974985e-05, "ex_per_s": 321.79891890300047, "data_frac": 0.018377177744723614, "mem_mb": 91.948032, "nll": -6.069833755493164, "elapsed_s": 787.6}206{"event": "train", "step": 3980, "loss": -5.809405326843262, "grad_norm": 0.2465561479330063, "lr": 3.0018453826016683e-05, "ex_per_s": 322.74067441158246, "data_frac": 0.018351749690582266, "mem_mb": 91.948032, "nll": -5.809405326843262, "elapsed_s": 789.3}207{"event": "train", "step": 4000, "loss": -6.059482097625732, "grad_norm": 0.6459692716598511, "lr": 2.9999999999999997e-05, "ex_per_s": 323.6783904720467, "data_frac": 0.018326435617190048, "mem_mb": 91.948032, "nll": -6.059482097625732, "elapsed_s": 791.0}208{"event": "val", "step": 4000, "val_nll": -6.451585810243266, "best": -6.451585810243266, "bad": 0, "mem_mb": 91.948032, "elapsed_s": 853.9}209{"event": "done", "steps": 4000, "best_val": -6.451585810243266, "wall_s": 853.8766720294952, "ex_per_s": 299.8091040378687, "peak_mem_mb": 91.948032, "n_params": 3501699, "device": "mps", "data_time_frac": 0.016974973938125592, "elapsed_s": 853.9}210
Files · main
Menu

Theme