1{"event": "train", "step": 20, "loss": -6.110433101654053, "grad_norm": 1.9482828378677368, "lr": 3.149999999999999e-05, "ex_per_s": 85.77476938973098, "data_frac": 0.812731502772781, "mem_mb": 122.998784, "nll": -6.110433101654053, "elapsed_s": 14.9}2{"event": "train", "step": 40, "loss": -5.898983955383301, "grad_norm": 0.20129452645778656, "lr": 6.149999999999999e-05, "ex_per_s": 146.89166377201812, "data_frac": 0.698842969160122, "mem_mb": 122.998784, "nll": -5.898983955383301, "elapsed_s": 17.4}3{"event": "train", "step": 60, "loss": -5.9795613288879395, "grad_norm": 0.6071668863296509, "lr": 9.149999999999999e-05, "ex_per_s": 195.45675921105752, "data_frac": 0.6204923323390553, "mem_mb": 122.998784, "nll": -5.9795613288879395, "elapsed_s": 19.7}4{"event": "train", "step": 80, "loss": -6.071864604949951, "grad_norm": 1.2804334163665771, "lr": 0.0001215, "ex_per_s": 239.25269522859423, "data_frac": 0.5703320432677784, "mem_mb": 122.998784, "nll": -6.071864604949951, "elapsed_s": 21.4}5{"event": "train", "step": 100, "loss": -5.949375152587891, "grad_norm": 0.43503233790397644, "lr": 0.0001515, "ex_per_s": 279.4978553403356, "data_frac": 0.5334044376085495, "mem_mb": 122.998784, "nll": -5.949375152587891, "elapsed_s": 22.9}6{"event": "train", "step": 120, "loss": -5.909389495849609, "grad_norm": 0.9850177764892578, "lr": 0.00018149999999999997, "ex_per_s": 315.09439815851414, "data_frac": 0.5015428176075455, "mem_mb": 122.998784, "nll": -5.909389495849609, "elapsed_s": 24.4}7{"event": "train", "step": 140, "loss": -5.832067489624023, "grad_norm": 0.47693780064582825, "lr": 0.00021149999999999996, "ex_per_s": 347.25471959775814, "data_frac": 0.4742627510755634, "mem_mb": 122.998784, "nll": -5.832067489624023, "elapsed_s": 25.8}8{"event": "train", "step": 160, "loss": -5.733274459838867, "grad_norm": 1.0508215427398682, "lr": 0.0002415, "ex_per_s": 376.40784362446055, "data_frac": 0.45028469006797883, "mem_mb": 122.998784, "nll": -5.733274459838867, "elapsed_s": 27.2}9{"event": "train", "step": 180, "loss": -5.973729610443115, "grad_norm": 1.2135798931121826, "lr": 0.0002715, "ex_per_s": 403.1612713518985, "data_frac": 0.4291388007498016, "mem_mb": 122.998784, "nll": -5.973729610443115, "elapsed_s": 28.6}10{"event": "train", "step": 200, "loss": -5.969871520996094, "grad_norm": 0.8803087472915649, "lr": 0.0003, "ex_per_s": 427.5224956982847, "data_frac": 0.40997832640536563, "mem_mb": 122.998784, "nll": -5.969871520996094, "elapsed_s": 30.0}11{"event": "train", "step": 220, "loss": -6.0281901359558105, "grad_norm": 1.3255536556243896, "lr": 0.00029998154617398326, "ex_per_s": 449.70821616018566, "data_frac": 0.39245662936962944, "mem_mb": 122.998784, "nll": -6.0281901359558105, "elapsed_s": 31.3}12{"event": "train", "step": 240, "loss": -5.91505241394043, "grad_norm": 0.38614538311958313, "lr": 0.000299926189741025, "ex_per_s": 469.4789189571121, "data_frac": 0.37596754931173026, "mem_mb": 122.998784, "nll": -5.91505241394043, "elapsed_s": 32.7}13{"event": "train", "step": 260, "loss": -5.888112545013428, "grad_norm": 0.17299678921699524, "lr": 0.0002998339458350211, "ex_per_s": 487.61175715156315, "data_frac": 0.36082899929187423, "mem_mb": 122.998784, "nll": -5.888112545013428, "elapsed_s": 34.1}14{"event": "train", "step": 280, "loss": -5.944365978240967, "grad_norm": 0.14229217171669006, "lr": 0.0002997048396745345, "ex_per_s": 504.91106239222, "data_frac": 0.34728937057739295, "mem_mb": 122.998784, "nll": -5.944365978240967, "elapsed_s": 35.5}15{"event": "train", "step": 300, "loss": -5.968596935272217, "grad_norm": 0.19384369254112244, "lr": 0.0002995389065559004, "ex_per_s": 520.986418166801, "data_frac": 0.3347977876399036, "mem_mb": 122.998784, "nll": -5.968596935272217, "elapsed_s": 36.9}16{"event": "train", "step": 320, "loss": -5.966061592102051, "grad_norm": 0.6423148512840271, "lr": 0.0002993361918435765, "ex_per_s": 535.785330552501, "data_frac": 0.3231163490593534, "mem_mb": 122.998784, "nll": -5.966061592102051, "elapsed_s": 38.2}17{"event": "train", "step": 340, "loss": -5.834932804107666, "grad_norm": 0.47950026392936707, "lr": 0.00029909675095774114, "ex_per_s": 549.5513735358971, "data_frac": 0.3122367744410234, "mem_mb": 122.998784, "nll": -5.834932804107666, "elapsed_s": 39.6}18{"event": "train", "step": 360, "loss": -5.995471954345703, "grad_norm": 0.23120278120040894, "lr": 0.00029882064935914164, "ex_per_s": 562.5026090104305, "data_frac": 0.3021467463265484, "mem_mb": 122.998784, "nll": -5.995471954345703, "elapsed_s": 41.0}19{"event": "train", "step": 380, "loss": -6.005291938781738, "grad_norm": 1.1866704225540161, "lr": 0.00029850796253119824, "ex_per_s": 574.4524887505712, "data_frac": 0.2926293354306257, "mem_mb": 122.998784, "nll": -6.005291938781738, "elapsed_s": 42.4}20{"event": "train", "step": 400, "loss": -5.994449615478516, "grad_norm": 1.478450894355774, "lr": 0.0002981587759593675, "ex_per_s": 585.6940103586181, "data_frac": 0.2837162390378724, "mem_mb": 122.998784, "nll": -5.994449615478516, "elapsed_s": 43.7}21{"event": "train", "step": 420, "loss": -5.964601039886475, "grad_norm": 0.49710139632225037, "lr": 0.0002977731851077716, "ex_per_s": 596.2535680489775, "data_frac": 0.27535413758557314, "mem_mb": 122.998784, "nll": -5.964601039886475, "elapsed_s": 45.1}22{"event": "train", "step": 440, "loss": -5.9393391609191895, "grad_norm": 0.7307088971138, "lr": 0.0002973512953930992, "ex_per_s": 606.2600817218892, "data_frac": 0.2675148337810192, "mem_mb": 122.998784, "nll": -5.9393391609191895, "elapsed_s": 46.5}23{"event": "train", "step": 460, "loss": -6.101770401000977, "grad_norm": 0.6445439457893372, "lr": 0.0002968932221557859, "ex_per_s": 615.6074624259908, "data_frac": 0.26009343061189766, "mem_mb": 122.998784, "nll": -6.101770401000977, "elapsed_s": 47.8}24{"event": "train", "step": 480, "loss": -5.877471446990967, "grad_norm": 0.21102793514728546, "lr": 0.0002963990906284808, "ex_per_s": 624.5827727678442, "data_frac": 0.25314022589641594, "mem_mb": 122.998784, "nll": -5.877471446990967, "elapsed_s": 49.2}25{"event": "train", "step": 500, "loss": -5.923840045928955, "grad_norm": 0.5386155247688293, "lr": 0.00029586903590180956, "ex_per_s": 632.8733396678008, "data_frac": 0.24671298143788606, "mem_mb": 122.998784, "nll": -5.923840045928955, "elapsed_s": 50.6}26{"event": "val", "step": 500, "val_nll": -6.44022385308593, "best": -6.44022385308593, "bad": 0, "mem_mb": 122.998784, "elapsed_s": 114.2}27{"event": "train", "step": 520, "loss": -6.040554046630859, "grad_norm": 0.26826512813568115, "lr": 0.00029530320288744205, "ex_per_s": 287.773038844767, "data_frac": 0.10798051289531656, "mem_mb": 122.998784, "nll": -6.040554046630859, "elapsed_s": 115.7}28{"event": "train", "step": 540, "loss": -5.880362033843994, "grad_norm": 0.47141316533088684, "lr": 0.0002947017462784745, "ex_per_s": 295.33468334489856, "data_frac": 0.10682286588649083, "mem_mb": 122.998784, "nll": -5.880362033843994, "elapsed_s": 117.0}29{"event": "train", "step": 560, "loss": -5.971432685852051, "grad_norm": 0.110936738550663, "lr": 0.00029406483050713824, "ex_per_s": 302.7327254528038, "data_frac": 0.105695302525427, "mem_mb": 122.998784, "nll": -5.971432685852051, "elapsed_s": 118.4}30{"event": "train", "step": 580, "loss": -5.9249162673950195, "grad_norm": 0.1171557679772377, "lr": 0.0002933926296998457, "ex_per_s": 309.966581776894, "data_frac": 0.10459661859563477, "mem_mb": 122.998784, "nll": -5.9249162673950195, "elapsed_s": 119.8}31{"event": "train", "step": 600, "loss": -5.97295618057251, "grad_norm": 0.7895901203155518, "lr": 0.0002926853276295856, "ex_per_s": 317.0413414722907, "data_frac": 0.10352211425100485, "mem_mb": 122.998784, "nll": -5.97295618057251, "elapsed_s": 121.1}32{"event": "train", "step": 620, "loss": -5.95200252532959, "grad_norm": 0.33668315410614014, "lr": 0.00029194311766568174, "ex_per_s": 323.94948225011444, "data_frac": 0.10247072089463088, "mem_mb": 122.998784, "nll": -5.95200252532959, "elapsed_s": 122.5}33{"event": "train", "step": 640, "loss": -6.065615653991699, "grad_norm": 0.473548948764801, "lr": 0.0002911662027209272, "ex_per_s": 330.67430640228133, "data_frac": 0.1014328657947993, "mem_mb": 122.998784, "nll": -6.065615653991699, "elapsed_s": 123.9}34{"event": "train", "step": 660, "loss": -6.007021903991699, "grad_norm": 1.0704600811004639, "lr": 0.00029035479519611045, "ex_per_s": 337.2846802039992, "data_frac": 0.10042930704381804, "mem_mb": 122.998784, "nll": -6.007021903991699, "elapsed_s": 125.3}35{"event": "train", "step": 680, "loss": -6.006908416748047, "grad_norm": 0.09999515116214752, "lr": 0.0002895091169219469, "ex_per_s": 343.7603924315471, "data_frac": 0.09944490854425586, "mem_mb": 122.998784, "nll": -6.006908416748047, "elapsed_s": 126.6}36{"event": "train", "step": 700, "loss": -5.988533020019531, "grad_norm": 0.5482091307640076, "lr": 0.00028862939909843273, "ex_per_s": 350.0845973848498, "data_frac": 0.09848383944933697, "mem_mb": 122.998784, "nll": -5.988533020019531, "elapsed_s": 128.0}37{"event": "train", "step": 720, "loss": -6.063944339752197, "grad_norm": 0.23534362018108368, "lr": 0.0002877158822316369, "ex_per_s": 356.28241266255196, "data_frac": 0.09754063461121294, "mem_mb": 122.998784, "nll": -6.063944339752197, "elapsed_s": 129.4}38{"event": "train", "step": 740, "loss": -6.03377628326416, "grad_norm": 0.2820655405521393, "lr": 0.00028676881606794945, "ex_per_s": 362.3518809243701, "data_frac": 0.09661891706807424, "mem_mb": 122.998784, "nll": -6.03377628326416, "elapsed_s": 130.7}39{"event": "train", "step": 760, "loss": -5.958082675933838, "grad_norm": 0.9461580514907837, "lr": 0.000285788459525803, "ex_per_s": 368.28747460074277, "data_frac": 0.09571175697617879, "mem_mb": 122.998784, "nll": -5.958082675933838, "elapsed_s": 132.1}40{"event": "train", "step": 780, "loss": -5.894946098327637, "grad_norm": 0.3145083487033844, "lr": 0.00028477508062488754, "ex_per_s": 374.12408765091027, "data_frac": 0.09483119741952516, "mem_mb": 122.998784, "nll": -5.894946098327637, "elapsed_s": 133.5}41{"event": "train", "step": 800, "loss": -5.815495014190674, "grad_norm": 0.2539801001548767, "lr": 0.000283728956412876, "ex_per_s": 379.75110723091825, "data_frac": 0.09394825159189958, "mem_mb": 122.998784, "nll": -5.815495014190674, "elapsed_s": 134.8}42{"event": "train", "step": 820, "loss": -5.934131145477295, "grad_norm": 0.27588072419166565, "lr": 0.00028265037288968236, "ex_per_s": 385.339565075148, "data_frac": 0.09309664708229728, "mem_mb": 122.998784, "nll": -5.934131145477295, "elapsed_s": 136.2}43{"event": "train", "step": 840, "loss": -5.808046340942383, "grad_norm": 0.5301992893218994, "lr": 0.0002815396249292725, "ex_per_s": 390.81347037433596, "data_frac": 0.0922622914114332, "mem_mb": 122.998784, "nll": -5.808046340942383, "elapsed_s": 137.6}44{"event": "train", "step": 860, "loss": -5.916110038757324, "grad_norm": 1.430367350578308, "lr": 0.00028039701619904774, "ex_per_s": 396.18515997835954, "data_frac": 0.09144598979144841, "mem_mb": 122.998784, "nll": -5.916110038757324, "elapsed_s": 138.9}45{"event": "train", "step": 880, "loss": -5.869335651397705, "grad_norm": 0.5202000737190247, "lr": 0.00027922285907682625, "ex_per_s": 401.4567002424285, "data_frac": 0.0906461326245068, "mem_mb": 122.998784, "nll": -5.869335651397705, "elapsed_s": 140.3}46{"event": "train", "step": 900, "loss": -5.986691474914551, "grad_norm": 0.46550554037094116, "lr": 0.00027801747456544134, "ex_per_s": 406.625748816249, "data_frac": 0.08986296386381952, "mem_mb": 122.998784, "nll": -5.986691474914551, "elapsed_s": 141.7}47{"event": "train", "step": 920, "loss": -6.025026321411133, "grad_norm": 0.3383634090423584, "lr": 0.00027678119220498265, "ex_per_s": 411.6899211903674, "data_frac": 0.08909226772667865, "mem_mb": 122.998784, "nll": -6.025026321411133, "elapsed_s": 143.0}48{"event": "train", "step": 940, "loss": -6.080915927886963, "grad_norm": 1.1434645652770996, "lr": 0.0002755143499827035, "ex_per_s": 416.66591928164115, "data_frac": 0.08833735646625475, "mem_mb": 122.998784, "nll": -6.080915927886963, "elapsed_s": 144.4}49{"event": "train", "step": 960, "loss": -5.968276023864746, "grad_norm": 0.30710524320602417, "lr": 0.00027421729424061787, "ex_per_s": 421.5514149490798, "data_frac": 0.0875981955021391, "mem_mb": 122.998784, "nll": -5.968276023864746, "elapsed_s": 145.8}50{"event": "train", "step": 980, "loss": -5.925605297088623, "grad_norm": 0.30890342593193054, "lr": 0.00027289037958081483, "ex_per_s": 426.3520473895018, "data_frac": 0.08687552659553494, "mem_mb": 122.998784, "nll": -5.925605297088623, "elapsed_s": 147.1}51{"event": "train", "step": 1000, "loss": -6.0715718269348145, "grad_norm": 0.1896844357252121, "lr": 0.0002715339687685131, "ex_per_s": 431.0605503761673, "data_frac": 0.0861623982082955, "mem_mb": 122.998784, "nll": -6.0715718269348145, "elapsed_s": 148.5}52{"event": "val", "step": 1000, "val_nll": -6.446381143079613, "best": -6.446381143079613, "bad": 0, "mem_mb": 122.998784, "elapsed_s": 211.4}53{"event": "train", "step": 1020, "loss": -6.030033588409424, "grad_norm": 0.14524635672569275, "lr": 0.0002701484326328852, "ex_per_s": 306.75942836557095, "data_frac": 0.06017377930631278, "mem_mb": 122.998784, "nll": -6.030033588409424, "elapsed_s": 212.8}54{"event": "train", "step": 1040, "loss": -5.9717841148376465, "grad_norm": 0.297324001789093, "lr": 0.00026873414996567596, "ex_per_s": 310.75337040573856, "data_frac": 0.05984409139527968, "mem_mb": 122.998784, "nll": -5.9717841148376465, "elapsed_s": 214.2}55{"event": "train", "step": 1060, "loss": -6.181902885437012, "grad_norm": 0.17286348342895508, "lr": 0.0002672915074176452, "ex_per_s": 314.6870626219501, "data_frac": 0.059519012912933815, "mem_mb": 122.998784, "nll": -6.181902885437012, "elapsed_s": 215.6}56{"event": "train", "step": 1080, "loss": -5.816494464874268, "grad_norm": 0.323133260011673, "lr": 0.00026582089939286075, "ex_per_s": 318.5995750509506, "data_frac": 0.059200298707653025, "mem_mb": 122.998784, "nll": -5.816494464874268, "elapsed_s": 217.0}57{"event": "train", "step": 1100, "loss": -6.052651405334473, "grad_norm": 0.09884874522686005, "lr": 0.0002643227279408728, "ex_per_s": 322.45501109738547, "data_frac": 0.05888639202239936, "mem_mb": 122.998784, "nll": -6.052651405334473, "elapsed_s": 218.3}58{"event": "train", "step": 1120, "loss": -6.052364349365234, "grad_norm": 0.39899924397468567, "lr": 0.000262797402646797, "ex_per_s": 326.2713860625268, "data_frac": 0.05857687203442324, "mem_mb": 122.998784, "nll": -6.052364349365234, "elapsed_s": 219.7}59{"event": "train", "step": 1140, "loss": -6.020354270935059, "grad_norm": 0.2538900375366211, "lr": 0.0002612453405193386, "ex_per_s": 330.0430374382825, "data_frac": 0.058270272844945875, "mem_mb": 122.998784, "nll": -6.020354270935059, "elapsed_s": 221.1}60{"event": "train", "step": 1160, "loss": -5.968044757843018, "grad_norm": 0.6192050576210022, "lr": 0.0002596669658767859, "ex_per_s": 333.7620915727639, "data_frac": 0.05796721675908713, "mem_mb": 122.998784, "nll": -5.968044757843018, "elapsed_s": 222.5}61{"event": "train", "step": 1180, "loss": -5.95754337310791, "grad_norm": 0.11487113684415817, "lr": 0.00025806271023100637, "ex_per_s": 337.4365789029812, "data_frac": 0.05766764508737119, "mem_mb": 122.998784, "nll": -5.95754337310791, "elapsed_s": 223.8}62{"event": "train", "step": 1200, "loss": -5.947057247161865, "grad_norm": 0.17878015339374542, "lr": 0.00025643301216947504, "ex_per_s": 341.07252943162547, "data_frac": 0.05737300277179486, "mem_mb": 122.998784, "nll": -5.947057247161865, "elapsed_s": 225.2}63{"event": "train", "step": 1220, "loss": -6.033038139343262, "grad_norm": 0.3514738976955414, "lr": 0.0002547783172353695, "ex_per_s": 344.669334692542, "data_frac": 0.05708219883226069, "mem_mb": 122.998784, "nll": -6.033038139343262, "elapsed_s": 226.6}64{"event": "train", "step": 1240, "loss": -6.059147357940674, "grad_norm": 0.10496452450752258, "lr": 0.00025309907780576277, "ex_per_s": 348.2198638940162, "data_frac": 0.056796193069989, "mem_mb": 122.998784, "nll": -6.059147357940674, "elapsed_s": 227.9}65{"event": "train", "step": 1260, "loss": -6.054773330688477, "grad_norm": 0.2866310179233551, "lr": 0.0002513957529679477, "ex_per_s": 351.73028398755974, "data_frac": 0.056512229002650835, "mem_mb": 122.998784, "nll": -6.054773330688477, "elapsed_s": 229.3}66{"event": "train", "step": 1280, "loss": -6.045374393463135, "grad_norm": 0.6539835929870605, "lr": 0.00024966880839392746, "ex_per_s": 355.1960305172653, "data_frac": 0.05623189942109088, "mem_mb": 122.998784, "nll": -6.045374393463135, "elapsed_s": 230.7}67{"event": "train", "step": 1300, "loss": -5.868870735168457, "grad_norm": 0.5940892696380615, "lr": 0.0002479187162131052, "ex_per_s": 358.62587254175395, "data_frac": 0.05595486934157029, "mem_mb": 122.998784, "nll": -5.868870735168457, "elapsed_s": 232.0}68{"event": "train", "step": 1320, "loss": -5.82714319229126, "grad_norm": 0.1721428781747818, "lr": 0.00024614595488320833, "ex_per_s": 362.0118540893097, "data_frac": 0.05568108937166399, "mem_mb": 122.998784, "nll": -5.82714319229126, "elapsed_s": 233.4}69{"event": "train", "step": 1340, "loss": -5.951812744140625, "grad_norm": 0.5069289803504944, "lr": 0.00024435100905948387, "ex_per_s": 365.36031798238304, "data_frac": 0.055409620006601754, "mem_mb": 122.998784, "nll": -5.951812744140625, "elapsed_s": 234.7}70{"event": "train", "step": 1360, "loss": -6.041578769683838, "grad_norm": 0.40070629119873047, "lr": 0.0002425343694621974, "ex_per_s": 368.6323404394837, "data_frac": 0.055138061827114324, "mem_mb": 122.998784, "nll": -6.041578769683838, "elapsed_s": 236.1}71{"event": "train", "step": 1380, "loss": -6.067385673522949, "grad_norm": 0.1060183122754097, "lr": 0.00024069653274247583, "ex_per_s": 371.89913126964217, "data_frac": 0.05487346170327353, "mem_mb": 122.998784, "nll": -6.067385673522949, "elapsed_s": 237.5}72{"event": "train", "step": 1400, "loss": -5.796326160430908, "grad_norm": 0.07116777449846268, "lr": 0.0002388380013465276, "ex_per_s": 375.1263498800833, "data_frac": 0.05461284729938549, "mem_mb": 122.998784, "nll": -5.796326160430908, "elapsed_s": 238.9}73{"event": "train", "step": 1420, "loss": -5.989113807678223, "grad_norm": 0.7632351517677307, "lr": 0.00023695928337827914, "ex_per_s": 378.32096643797115, "data_frac": 0.05435443107042899, "mem_mb": 122.998784, "nll": -5.989113807678223, "elapsed_s": 240.2}74{"event": "train", "step": 1440, "loss": -5.745086193084717, "grad_norm": 0.15515998005867004, "lr": 0.00023506089246046403, "ex_per_s": 381.47585509806174, "data_frac": 0.054098612680817615, "mem_mb": 122.998784, "nll": -5.745086193084717, "elapsed_s": 241.6}75{"event": "train", "step": 1460, "loss": -5.800703048706055, "grad_norm": 0.28848379850387573, "lr": 0.00023314334759420388, "ex_per_s": 384.5963782460596, "data_frac": 0.05384420126788877, "mem_mb": 122.998784, "nll": -5.800703048706055, "elapsed_s": 243.0}76{"event": "train", "step": 1480, "loss": -6.0456061363220215, "grad_norm": 0.22621232271194458, "lr": 0.0002312071730171186, "ex_per_s": 387.66422165233627, "data_frac": 0.05359259813935194, "mem_mb": 122.998784, "nll": -6.0456061363220215, "elapsed_s": 244.4}77{"event": "train", "step": 1500, "loss": -5.958396911621094, "grad_norm": 0.5174848437309265, "lr": 0.0002292528980600049, "ex_per_s": 390.7158186339633, "data_frac": 0.05334612474958968, "mem_mb": 122.998784, "nll": -5.958396911621094, "elapsed_s": 245.7}78{"event": "val", "step": 1500, "val_nll": -6.446817688011299, "best": -6.446817688011299, "bad": 0, "mem_mb": 122.998784, "elapsed_s": 308.6}79{"event": "train", "step": 1520, "loss": -5.949045181274414, "grad_norm": 0.690746009349823, "lr": 0.00022728105700212285, "ex_per_s": 313.8228555574555, "data_frac": 0.04232460882743632, "mem_mb": 122.998784, "nll": -5.949045181274414, "elapsed_s": 310.0}80{"event": "train", "step": 1540, "loss": -5.878180980682373, "grad_norm": 0.08919516205787659, "lr": 0.00022529218892512886, "ex_per_s": 316.5559827632069, "data_frac": 0.042179098526940094, "mem_mb": 122.998784, "nll": -5.878180980682373, "elapsed_s": 311.4}81{"event": "train", "step": 1560, "loss": -5.982547760009766, "grad_norm": 0.3776604235172272, "lr": 0.00022328683756569652, "ex_per_s": 319.26607680349986, "data_frac": 0.042034660924440255, "mem_mb": 122.998784, "nll": -5.982547760009766, "elapsed_s": 312.7}82{"event": "train", "step": 1580, "loss": -5.980978012084961, "grad_norm": 0.10232552140951157, "lr": 0.0002212655511668642, "ex_per_s": 321.95162088379084, "data_frac": 0.041891402099924424, "mem_mb": 122.998784, "nll": -5.980978012084961, "elapsed_s": 314.1}83{"event": "train", "step": 1600, "loss": -5.955664157867432, "grad_norm": 0.10776771605014801, "lr": 0.00021922888232815089, "ex_per_s": 324.60009556994885, "data_frac": 0.0417486159668292, "mem_mb": 122.998784, "nll": -5.955664157867432, "elapsed_s": 315.5}84{"event": "train", "step": 1620, "loss": -5.879905700683594, "grad_norm": 0.20575720071792603, "lr": 0.00021717738785448102, "ex_per_s": 327.22381498307925, "data_frac": 0.04160639405065995, "mem_mb": 122.998784, "nll": -5.879905700683594, "elapsed_s": 316.9}85{"event": "train", "step": 1640, "loss": -6.079751968383789, "grad_norm": 0.38650062680244446, "lr": 0.00021511162860395977, "ex_per_s": 329.94361960241065, "data_frac": 0.041480127326953366, "mem_mb": 122.998784, "nll": -6.079751968383789, "elapsed_s": 318.1}86{"event": "train", "step": 1660, "loss": -6.039597034454346, "grad_norm": 0.7509287595748901, "lr": 0.0002130321693345397, "ex_per_s": 332.5390923558743, "data_frac": 0.04134193893314993, "mem_mb": 122.998784, "nll": -6.039597034454346, "elapsed_s": 319.5}87{"event": "train", "step": 1680, "loss": -5.866462707519531, "grad_norm": 0.1452990621328354, "lr": 0.0002109395785496218, "ex_per_s": 335.1150543212988, "data_frac": 0.0412058185910202, "mem_mb": 122.998784, "nll": -5.866462707519531, "elapsed_s": 320.9}88{"event": "train", "step": 1700, "loss": -5.928055286407471, "grad_norm": 0.41136598587036133, "lr": 0.00020883442834263225, "ex_per_s": 337.66528503730655, "data_frac": 0.04107025475932843, "mem_mb": 122.998784, "nll": -5.928055286407471, "elapsed_s": 322.2}89{"event": "train", "step": 1720, "loss": -5.887325763702393, "grad_norm": 1.0006409883499146, "lr": 0.00020671729424061788, "ex_per_s": 340.1928886943626, "data_frac": 0.04093519803330987, "mem_mb": 122.998784, "nll": -5.887325763702393, "elapsed_s": 323.6}90{"event": "train", "step": 1740, "loss": -6.2744035720825195, "grad_norm": 0.39533907175064087, "lr": 0.00020458875504690275, "ex_per_s": 342.70040572001585, "data_frac": 0.040801063468854114, "mem_mb": 122.998784, "nll": -6.2744035720825195, "elapsed_s": 325.0}91{"event": "train", "step": 1760, "loss": -5.929758548736572, "grad_norm": 0.21467886865139008, "lr": 0.000202449392682849, "ex_per_s": 345.1840930322206, "data_frac": 0.040668447755137796, "mem_mb": 122.998784, "nll": -5.929758548736572, "elapsed_s": 326.3}92{"event": "train", "step": 1780, "loss": -5.87169885635376, "grad_norm": 0.49215641617774963, "lr": 0.00020029979202876563, "ex_per_s": 347.6483013091188, "data_frac": 0.04053735402910136, "mem_mb": 122.998784, "nll": -5.87169885635376, "elapsed_s": 327.7}93{"event": "train", "step": 1800, "loss": -6.050236701965332, "grad_norm": 0.8600218892097473, "lr": 0.00019814054076400787, "ex_per_s": 350.0936877948609, "data_frac": 0.040407400740607564, "mem_mb": 122.998784, "nll": -6.050236701965332, "elapsed_s": 329.1}94{"event": "train", "step": 1820, "loss": -5.936669826507568, "grad_norm": 0.47294068336486816, "lr": 0.00019597222920631147, "ex_per_s": 352.51951875783914, "data_frac": 0.04027852587821599, "mem_mb": 122.998784, "nll": -5.936669826507568, "elapsed_s": 330.4}95{"event": "train", "step": 1840, "loss": -6.073741436004639, "grad_norm": 0.4995134174823761, "lr": 0.0001937954501504058, "ex_per_s": 354.92367669392115, "data_frac": 0.04015030318904641, "mem_mb": 122.998784, "nll": -6.073741436004639, "elapsed_s": 331.8}96{"event": "train", "step": 1860, "loss": -6.0512309074401855, "grad_norm": 0.18128614127635956, "lr": 0.0001916107987059498, "ex_per_s": 357.3145429857316, "data_frac": 0.04002402994045562, "mem_mb": 122.998784, "nll": -6.0512309074401855, "elapsed_s": 333.2}97{"event": "train", "step": 1880, "loss": -5.947964668273926, "grad_norm": 0.7085979580879211, "lr": 0.00018941887213483484, "ex_per_s": 359.6684696850504, "data_frac": 0.03989733141926899, "mem_mb": 122.998784, "nll": -5.947964668273926, "elapsed_s": 334.5}98{"event": "train", "step": 1900, "loss": -5.983514308929443, "grad_norm": 0.16910776495933533, "lr": 0.00018722026968789906, "ex_per_s": 361.98016794649266, "data_frac": 0.039770255196974184, "mem_mb": 122.998784, "nll": -5.983514308929443, "elapsed_s": 335.9}99{"event": "train", "step": 1920, "loss": -5.964282512664795, "grad_norm": 0.3231184780597687, "lr": 0.00018501559244109864, "ex_per_s": 364.31429506884695, "data_frac": 0.03964708779400878, "mem_mb": 122.998784, "nll": -5.964282512664795, "elapsed_s": 337.3}100{"event": "train", "step": 1940, "loss": -6.049917221069336, "grad_norm": 0.318715363740921, "lr": 0.00018280544313117932, "ex_per_s": 366.61075963083886, "data_frac": 0.03952408745620519, "mem_mb": 122.998784, "nll": -6.049917221069336, "elapsed_s": 338.7}101{"event": "train", "step": 1960, "loss": -5.9396772384643555, "grad_norm": 0.24915988743305206, "lr": 0.00018059042599089415, "ex_per_s": 368.89895054081427, "data_frac": 0.03940129446218978, "mem_mb": 122.998784, "nll": -5.9396772384643555, "elapsed_s": 340.1}102{"event": "train", "step": 1980, "loss": -5.963989734649658, "grad_norm": 0.8476231098175049, "lr": 0.00017837114658381247, "ex_per_s": 371.1777673922459, "data_frac": 0.039280961571387545, "mem_mb": 122.998784, "nll": -5.963989734649658, "elapsed_s": 341.4}103{"event": "train", "step": 2000, "loss": -5.914310932159424, "grad_norm": 0.48780950903892517, "lr": 0.00017614821163876486, "ex_per_s": 373.4321204791486, "data_frac": 0.03916085090675612, "mem_mb": 122.998784, "nll": -5.914310932159424, "elapsed_s": 342.8}104{"event": "val", "step": 2000, "val_nll": -6.4492783828584, "best": -6.4492783828584, "bad": 0, "mem_mb": 122.998784, "elapsed_s": 405.6}105{"event": "train", "step": 2020, "loss": -5.9262213706970215, "grad_norm": 0.568572461605072, "lr": 0.00017392222888396941, "ex_per_s": 317.65545031345897, "data_frac": 0.03301267563999575, "mem_mb": 122.998784, "nll": -5.9262213706970215, "elapsed_s": 407.0}106{"event": "train", "step": 2040, "loss": -5.874476432800293, "grad_norm": 0.144888773560524, "lr": 0.000171693806880885, "ex_per_s": 319.72407741414105, "data_frac": 0.0329332824873911, "mem_mb": 122.998784, "nll": -5.874476432800293, "elapsed_s": 408.4}107{"event": "train", "step": 2060, "loss": -6.141441345214844, "grad_norm": 0.27485692501068115, "lr": 0.0001694635548578366, "ex_per_s": 321.77945444908005, "data_frac": 0.032853489270896256, "mem_mb": 122.998784, "nll": -6.141441345214844, "elapsed_s": 409.7}108{"event": "train", "step": 2080, "loss": -5.963988304138184, "grad_norm": 0.31237179040908813, "lr": 0.0001672320825434584, "ex_per_s": 323.82332515600183, "data_frac": 0.032774606937097064, "mem_mb": 122.998784, "nll": -5.963988304138184, "elapsed_s": 411.1}109{"event": "train", "step": 2100, "loss": -5.898028373718262, "grad_norm": 1.1269603967666626, "lr": 0.000165, "ex_per_s": 325.854475084396, "data_frac": 0.03269656250288266, "mem_mb": 122.998784, "nll": -5.898028373718262, "elapsed_s": 412.5}110{"event": "train", "step": 2120, "loss": -6.152252197265625, "grad_norm": 0.4341965913772583, "lr": 0.0001627679174565416, "ex_per_s": 327.8671243450703, "data_frac": 0.032618184167146035, "mem_mb": 122.998784, "nll": -6.152252197265625, "elapsed_s": 413.8}111{"event": "train", "step": 2140, "loss": -5.97750186920166, "grad_norm": 0.1042579859495163, "lr": 0.00016053644514216336, "ex_per_s": 329.86910225206776, "data_frac": 0.032541201310209725, "mem_mb": 122.998784, "nll": -5.97750186920166, "elapsed_s": 415.2}112{"event": "train", "step": 2160, "loss": -5.926516056060791, "grad_norm": 0.5131976008415222, "lr": 0.00015830619311911498, "ex_per_s": 331.84560138180547, "data_frac": 0.03246391042935491, "mem_mb": 122.998784, "nll": -5.926516056060791, "elapsed_s": 416.6}113{"event": "train", "step": 2180, "loss": -6.084172248840332, "grad_norm": 0.3232581615447998, "lr": 0.00015607777111603058, "ex_per_s": 333.8163493359732, "data_frac": 0.032386567116467195, "mem_mb": 122.998784, "nll": -6.084172248840332, "elapsed_s": 418.0}114{"event": "train", "step": 2200, "loss": -5.966341018676758, "grad_norm": 0.5541177988052368, "lr": 0.00015385178836123514, "ex_per_s": 335.7802892166785, "data_frac": 0.03231049479986777, "mem_mb": 122.998784, "nll": -5.966341018676758, "elapsed_s": 419.3}115{"event": "train", "step": 2220, "loss": -5.825324535369873, "grad_norm": 0.5964010953903198, "lr": 0.00015162885341618752, "ex_per_s": 337.73269953171996, "data_frac": 0.03223540470237104, "mem_mb": 122.998784, "nll": -5.825324535369873, "elapsed_s": 420.7}116{"event": "train", "step": 2240, "loss": -5.960657119750977, "grad_norm": 0.710385799407959, "lr": 0.00014940957400910585, "ex_per_s": 339.67145421871106, "data_frac": 0.03216038674959756, "mem_mb": 122.998784, "nll": -5.960657119750977, "elapsed_s": 422.1}117{"event": "train", "step": 2260, "loss": -5.7387495040893555, "grad_norm": 0.7334255576133728, "lr": 0.00014719455686882068, "ex_per_s": 341.5980619047376, "data_frac": 0.032086018344739474, "mem_mb": 122.998784, "nll": -5.7387495040893555, "elapsed_s": 423.4}118{"event": "train", "step": 2280, "loss": -6.009469509124756, "grad_norm": 0.5012499690055847, "lr": 0.0001449844075589013, "ex_per_s": 343.51018271614333, "data_frac": 0.03201164582964041, "mem_mb": 122.998784, "nll": -6.009469509124756, "elapsed_s": 424.8}119{"event": "train", "step": 2300, "loss": -6.000481605529785, "grad_norm": 0.11448080837726593, "lr": 0.0001427797303121009, "ex_per_s": 345.40668439610374, "data_frac": 0.03193893358960215, "mem_mb": 122.998784, "nll": -6.000481605529785, "elapsed_s": 426.2}120{"event": "train", "step": 2320, "loss": -6.071065902709961, "grad_norm": 0.44215261936187744, "lr": 0.00014058112786516518, "ex_per_s": 347.29610230688814, "data_frac": 0.03186555102455135, "mem_mb": 122.998784, "nll": -6.071065902709961, "elapsed_s": 427.6}121{"event": "train", "step": 2340, "loss": -5.85146951675415, "grad_norm": 0.36102259159088135, "lr": 0.00013838920129405016, "ex_per_s": 349.1794183399664, "data_frac": 0.03179370551882147, "mem_mb": 122.998784, "nll": -5.85146951675415, "elapsed_s": 428.9}122{"event": "train", "step": 2360, "loss": -5.929818630218506, "grad_norm": 0.07300731539726257, "lr": 0.0001362045498495942, "ex_per_s": 351.0462671229338, "data_frac": 0.03172302119954024, "mem_mb": 122.998784, "nll": -5.929818630218506, "elapsed_s": 430.3}123{"event": "train", "step": 2380, "loss": -5.932983875274658, "grad_norm": 0.10422924160957336, "lr": 0.00013402777079368853, "ex_per_s": 352.9020760681868, "data_frac": 0.03165177615731044, "mem_mb": 122.998784, "nll": -5.932983875274658, "elapsed_s": 431.6}124{"event": "train", "step": 2400, "loss": -5.956429958343506, "grad_norm": 0.8430474400520325, "lr": 0.00013185945923599213, "ex_per_s": 354.74475847867325, "data_frac": 0.03158075702604614, "mem_mb": 122.998784, "nll": -5.956429958343506, "elapsed_s": 433.0}125{"event": "train", "step": 2420, "loss": -5.997149467468262, "grad_norm": 0.3919523060321808, "lr": 0.00012970020797123432, "ex_per_s": 356.57702581345336, "data_frac": 0.031510188215062886, "mem_mb": 122.998784, "nll": -5.997149467468262, "elapsed_s": 434.4}126{"event": "train", "step": 2440, "loss": -6.099442005157471, "grad_norm": 0.43480098247528076, "lr": 0.000127550607317151, "ex_per_s": 358.3962135749994, "data_frac": 0.03144000692748584, "mem_mb": 122.998784, "nll": -6.099442005157471, "elapsed_s": 435.7}127{"event": "train", "step": 2460, "loss": -6.1333723068237305, "grad_norm": 0.11131129413843155, "lr": 0.00012541124495309727, "ex_per_s": 360.1846250554889, "data_frac": 0.03136934085049955, "mem_mb": 122.998784, "nll": -6.1333723068237305, "elapsed_s": 437.1}128{"event": "train", "step": 2480, "loss": -5.945925235748291, "grad_norm": 0.1050860658288002, "lr": 0.00012328270575938212, "ex_per_s": 361.9838089821486, "data_frac": 0.03130029820928757, "mem_mb": 122.998784, "nll": -5.945925235748291, "elapsed_s": 438.5}129{"event": "train", "step": 2500, "loss": -5.755810260772705, "grad_norm": 0.710875391960144, "lr": 0.00012116557165736775, "ex_per_s": 363.7726577177013, "data_frac": 0.031231621201480802, "mem_mb": 122.998784, "nll": -5.755810260772705, "elapsed_s": 439.9}130{"event": "val", "step": 2500, "val_nll": -6.449994033240759, "best": -6.449994033240759, "bad": 0, "mem_mb": 122.998784, "elapsed_s": 502.7}131{"event": "train", "step": 2520, "loss": -5.961863994598389, "grad_norm": 0.8790099620819092, "lr": 0.00011906042145037817, "ex_per_s": 319.9437502194017, "data_frac": 0.027275916519523505, "mem_mb": 122.998784, "nll": -5.961863994598389, "elapsed_s": 504.1}132{"event": "train", "step": 2540, "loss": -6.03085470199585, "grad_norm": 0.4609578549861908, "lr": 0.00011696783066546026, "ex_per_s": 321.6080497337152, "data_frac": 0.02722682236622011, "mem_mb": 122.998784, "nll": -6.03085470199585, "elapsed_s": 505.5}133{"event": "train", "step": 2560, "loss": -5.949787616729736, "grad_norm": 0.1311957687139511, "lr": 0.00011488837139604017, "ex_per_s": 323.26828272935643, "data_frac": 0.02717816600415608, "mem_mb": 122.998784, "nll": -5.949787616729736, "elapsed_s": 506.8}134{"event": "train", "step": 2580, "loss": -5.959117889404297, "grad_norm": 0.2478628158569336, "lr": 0.00011282261214551898, "ex_per_s": 324.9168316499413, "data_frac": 0.027129325194693597, "mem_mb": 122.998784, "nll": -5.959117889404297, "elapsed_s": 508.2}135{"event": "train", "step": 2600, "loss": -5.8481574058532715, "grad_norm": 0.1580209732055664, "lr": 0.00011077111767184915, "ex_per_s": 326.55636565176354, "data_frac": 0.02708071762431824, "mem_mb": 122.998784, "nll": -5.8481574058532715, "elapsed_s": 509.6}136{"event": "train", "step": 2620, "loss": -5.8895745277404785, "grad_norm": 0.8732475638389587, "lr": 0.00010873444883313583, "ex_per_s": 328.18482631392425, "data_frac": 0.027032829267746333, "mem_mb": 122.998784, "nll": -5.8895745277404785, "elapsed_s": 511.0}137{"event": "train", "step": 2640, "loss": -5.986438751220703, "grad_norm": 0.4925490915775299, "lr": 0.00010671316243430345, "ex_per_s": 329.80823643604407, "data_frac": 0.026985134695328355, "mem_mb": 122.998784, "nll": -5.986438751220703, "elapsed_s": 512.3}138{"event": "train", "step": 2660, "loss": -5.819965839385986, "grad_norm": 0.0937548503279686, "lr": 0.00010470781107487112, "ex_per_s": 331.4210341330677, "data_frac": 0.02693794617683604, "mem_mb": 122.998784, "nll": -5.819965839385986, "elapsed_s": 513.7}139{"event": "train", "step": 2680, "loss": -5.959562301635742, "grad_norm": 0.3820561170578003, "lr": 0.00010271894299787715, "ex_per_s": 333.02708954271975, "data_frac": 0.026890889879902017, "mem_mb": 122.998784, "nll": -5.959562301635742, "elapsed_s": 515.1}140{"event": "train", "step": 2700, "loss": -6.135176181793213, "grad_norm": 0.707733154296875, "lr": 0.00010074710193999506, "ex_per_s": 334.616505732214, "data_frac": 0.026843863538208505, "mem_mb": 122.998784, "nll": -6.135176181793213, "elapsed_s": 516.4}141{"event": "train", "step": 2720, "loss": -5.8815083503723145, "grad_norm": 0.43106263875961304, "lr": 9.879282698288144e-05, "ex_per_s": 336.1901377010434, "data_frac": 0.026797116683790333, "mem_mb": 122.998784, "nll": -5.8815083503723145, "elapsed_s": 517.8}142{"event": "train", "step": 2740, "loss": -6.127623081207275, "grad_norm": 0.0777747631072998, "lr": 9.685665240579613e-05, "ex_per_s": 337.76820186658824, "data_frac": 0.026751341040238984, "mem_mb": 122.998784, "nll": -6.127623081207275, "elapsed_s": 519.2}143{"event": "train", "step": 2760, "loss": -5.96030855178833, "grad_norm": 0.21186193823814392, "lr": 9.4939107539536e-05, "ex_per_s": 339.3330641811362, "data_frac": 0.026704686231677124, "mem_mb": 122.998784, "nll": -5.96030855178833, "elapsed_s": 520.6}144{"event": "train", "step": 2780, "loss": -6.01186990737915, "grad_norm": 0.9069233536720276, "lr": 9.304071662172087e-05, "ex_per_s": 340.89777871697135, "data_frac": 0.026659671119139864, "mem_mb": 122.998784, "nll": -6.01186990737915, "elapsed_s": 521.9}145{"event": "train", "step": 2800, "loss": -5.981179714202881, "grad_norm": 0.8079415559768677, "lr": 9.116199865347241e-05, "ex_per_s": 342.4534332888748, "data_frac": 0.026615145410359194, "mem_mb": 122.998784, "nll": -5.981179714202881, "elapsed_s": 523.3}146{"event": "train", "step": 2820, "loss": -6.008249759674072, "grad_norm": 0.25687292218208313, "lr": 8.930346725752419e-05, "ex_per_s": 344.00202640313535, "data_frac": 0.026569932149381293, "mem_mb": 122.998784, "nll": -6.008249759674072, "elapsed_s": 524.7}147{"event": "train", "step": 2840, "loss": -5.878984451293945, "grad_norm": 0.1270655393600464, "lr": 8.746563053780262e-05, "ex_per_s": 345.5446694194885, "data_frac": 0.02652519050982625, "mem_mb": 122.998784, "nll": -5.878984451293945, "elapsed_s": 526.0}148{"event": "train", "step": 2860, "loss": -6.047909259796143, "grad_norm": 0.38326582312583923, "lr": 8.564899094051614e-05, "ex_per_s": 347.07707141085115, "data_frac": 0.026479971782848995, "mem_mb": 122.998784, "nll": -6.047909259796143, "elapsed_s": 527.4}149{"event": "train", "step": 2880, "loss": -5.770879745483398, "grad_norm": 0.4135778248310089, "lr": 8.385404511679161e-05, "ex_per_s": 348.6024746973539, "data_frac": 0.026435585941653822, "mem_mb": 122.998784, "nll": -5.770879745483398, "elapsed_s": 528.8}150{"event": "train", "step": 2900, "loss": -5.795256614685059, "grad_norm": 0.6063058376312256, "lr": 8.208128378689484e-05, "ex_per_s": 350.121397090796, "data_frac": 0.026391307364307382, "mem_mb": 122.998784, "nll": -5.795256614685059, "elapsed_s": 530.1}151{"event": "train", "step": 2920, "loss": -5.971447467803955, "grad_norm": 0.7554762363433838, "lr": 8.03311916060725e-05, "ex_per_s": 351.62893537609057, "data_frac": 0.02634703186514644, "mem_mb": 122.998784, "nll": -5.971447467803955, "elapsed_s": 531.5}152{"event": "train", "step": 2940, "loss": -5.903770923614502, "grad_norm": 0.5900241732597351, "lr": 7.860424703205225e-05, "ex_per_s": 353.1287654991494, "data_frac": 0.02630311024307326, "mem_mb": 122.998784, "nll": -5.903770923614502, "elapsed_s": 532.9}153{"event": "train", "step": 2960, "loss": -5.9393310546875, "grad_norm": 0.3303596079349518, "lr": 7.690092219423722e-05, "ex_per_s": 354.62251232544486, "data_frac": 0.026259462873600183, "mem_mb": 122.998784, "nll": -5.9393310546875, "elapsed_s": 534.2}154{"event": "train", "step": 2980, "loss": -5.894739151000977, "grad_norm": 0.4595528542995453, "lr": 7.522168276463048e-05, "ex_per_s": 356.1089585434196, "data_frac": 0.02621571832638608, "mem_mb": 122.998784, "nll": -5.894739151000977, "elapsed_s": 535.6}155{"event": "train", "step": 3000, "loss": -5.994903564453125, "grad_norm": 0.16125330328941345, "lr": 7.356698783052496e-05, "ex_per_s": 357.580908433726, "data_frac": 0.02617259749047671, "mem_mb": 122.998784, "nll": -5.994903564453125, "elapsed_s": 537.0}156{"event": "val", "step": 3000, "val_nll": -6.447894408494198, "best": -6.449994033240759, "bad": 1, "mem_mb": 122.998784, "elapsed_s": 599.7}157{"event": "train", "step": 3020, "loss": -5.961067199707031, "grad_norm": 0.8925390243530273, "lr": 7.193728976899362e-05, "ex_per_s": 321.54129164098106, "data_frac": 0.02339987290426963, "mem_mb": 122.998784, "nll": -5.961067199707031, "elapsed_s": 601.1}158{"event": "train", "step": 3040, "loss": -5.859010696411133, "grad_norm": 0.784035861492157, "lr": 7.033303412321404e-05, "ex_per_s": 322.93665904470197, "data_frac": 0.023367362501235632, "mem_mb": 122.998784, "nll": -5.859010696411133, "elapsed_s": 602.5}159{"event": "train", "step": 3060, "loss": -6.071619033813477, "grad_norm": 0.18126346170902252, "lr": 6.875465948066139e-05, "ex_per_s": 324.32551874710384, "data_frac": 0.02333487833753284, "mem_mb": 122.998784, "nll": -6.071619033813477, "elapsed_s": 603.9}160{"event": "train", "step": 3080, "loss": -5.970253944396973, "grad_norm": 0.34150463342666626, "lr": 6.720259735320299e-05, "ex_per_s": 325.7080574826776, "data_frac": 0.023303154750513172, "mem_mb": 122.998784, "nll": -5.970253944396973, "elapsed_s": 605.2}161{"event": "train", "step": 3100, "loss": -6.015607833862305, "grad_norm": 0.33053314685821533, "lr": 6.567727205912722e-05, "ex_per_s": 327.0765280155901, "data_frac": 0.023271165453799578, "mem_mb": 122.998784, "nll": -6.015607833862305, "elapsed_s": 606.6}162{"event": "train", "step": 3120, "loss": -6.089792251586914, "grad_norm": 0.747597873210907, "lr": 6.417910060713926e-05, "ex_per_s": 328.44482483640286, "data_frac": 0.023239360193542887, "mem_mb": 122.998784, "nll": -6.089792251586914, "elapsed_s": 608.0}163{"event": "train", "step": 3140, "loss": -6.0100932121276855, "grad_norm": 0.38754138350486755, "lr": 6.270849258235485e-05, "ex_per_s": 329.8076210315514, "data_frac": 0.02320788672533994, "mem_mb": 122.998784, "nll": -6.0100932121276855, "elapsed_s": 609.3}164{"event": "train", "step": 3160, "loss": -6.021377086639404, "grad_norm": 0.1710870862007141, "lr": 6.126585003432406e-05, "ex_per_s": 331.16565181270363, "data_frac": 0.023177029080394554, "mem_mb": 122.998784, "nll": -6.021377086639404, "elapsed_s": 610.7}165{"event": "train", "step": 3180, "loss": -6.095218181610107, "grad_norm": 0.11017276346683502, "lr": 5.985156736711483e-05, "ex_per_s": 332.51572883705035, "data_frac": 0.023145760460786625, "mem_mb": 122.998784, "nll": -6.095218181610107, "elapsed_s": 612.1}166{"event": "train", "step": 3200, "loss": -6.001607418060303, "grad_norm": 0.33603256940841675, "lr": 5.846603123148688e-05, "ex_per_s": 333.8632514952405, "data_frac": 0.02311458102170624, "mem_mb": 122.998784, "nll": -6.001607418060303, "elapsed_s": 613.4}167{"event": "train", "step": 3220, "loss": -5.915058612823486, "grad_norm": 0.1176900789141655, "lr": 5.710962041918516e-05, "ex_per_s": 335.20316671682303, "data_frac": 0.023083418323762313, "mem_mb": 122.998784, "nll": -5.915058612823486, "elapsed_s": 614.8}168{"event": "train", "step": 3240, "loss": -5.802231788635254, "grad_norm": 0.3798986077308655, "lr": 5.578270575938211e-05, "ex_per_s": 336.537444335299, "data_frac": 0.023052843143415283, "mem_mb": 122.998784, "nll": -5.802231788635254, "elapsed_s": 616.2}169{"event": "train", "step": 3260, "loss": -5.959538459777832, "grad_norm": 0.43990013003349304, "lr": 5.448565001729653e-05, "ex_per_s": 337.8537228795522, "data_frac": 0.023022084740782604, "mem_mb": 122.998784, "nll": -5.959538459777832, "elapsed_s": 617.6}170{"event": "train", "step": 3280, "loss": -6.015836715698242, "grad_norm": 0.32004016637802124, "lr": 5.3218807795017296e-05, "ex_per_s": 339.175180875376, "data_frac": 0.02299165860229216, "mem_mb": 122.998784, "nll": -6.015836715698242, "elapsed_s": 618.9}171{"event": "train", "step": 3300, "loss": -5.8040080070495605, "grad_norm": 0.24969184398651123, "lr": 5.198252543455864e-05, "ex_per_s": 340.49183767919095, "data_frac": 0.022961333283461192, "mem_mb": 122.998784, "nll": -5.8040080070495605, "elapsed_s": 620.3}172{"event": "train", "step": 3320, "loss": -5.989705562591553, "grad_norm": 0.3674097955226898, "lr": 5.07771409231737e-05, "ex_per_s": 341.80236862851416, "data_frac": 0.022931424512709263, "mem_mb": 122.998784, "nll": -5.989705562591553, "elapsed_s": 621.7}173{"event": "train", "step": 3340, "loss": -5.931643486022949, "grad_norm": 0.29789379239082336, "lr": 4.960298380095221e-05, "ex_per_s": 343.10679293753304, "data_frac": 0.022901415307333278, "mem_mb": 122.998784, "nll": -5.931643486022949, "elapsed_s": 623.0}174{"event": "train", "step": 3360, "loss": -5.907341957092285, "grad_norm": 0.366054892539978, "lr": 4.846037507072751e-05, "ex_per_s": 344.4047556765105, "data_frac": 0.022871529802835872, "mem_mb": 122.998784, "nll": -5.907341957092285, "elapsed_s": 624.4}175{"event": "train", "step": 3380, "loss": -6.1231689453125, "grad_norm": 0.3759686052799225, "lr": 4.7349627110317556e-05, "ex_per_s": 345.69937043361494, "data_frac": 0.022842148802176724, "mem_mb": 122.998784, "nll": -6.1231689453125, "elapsed_s": 625.8}176{"event": "train", "step": 3400, "loss": -5.883131504058838, "grad_norm": 0.5659611225128174, "lr": 4.6271043587123986e-05, "ex_per_s": 346.9871135615081, "data_frac": 0.022812789141338257, "mem_mb": 122.998784, "nll": -5.883131504058838, "elapsed_s": 627.1}177{"event": "train", "step": 3420, "loss": -5.8352885246276855, "grad_norm": 0.25451725721359253, "lr": 4.52249193751124e-05, "ex_per_s": 348.2697129264822, "data_frac": 0.02278313559915749, "mem_mb": 122.998784, "nll": -5.8352885246276855, "elapsed_s": 628.5}178{"event": "train", "step": 3440, "loss": -5.9163103103637695, "grad_norm": 0.09925639629364014, "lr": 4.4211540474196926e-05, "ex_per_s": 349.54615934727707, "data_frac": 0.022754016252832143, "mem_mb": 122.998784, "nll": -5.9163103103637695, "elapsed_s": 629.9}179{"event": "train", "step": 3460, "loss": -6.057917594909668, "grad_norm": 0.12103645503520966, "lr": 4.3231183932050546e-05, "ex_per_s": 350.81740873642497, "data_frac": 0.022724680729813037, "mem_mb": 122.998784, "nll": -6.057917594909668, "elapsed_s": 631.2}180{"event": "train", "step": 3480, "loss": -5.918457984924316, "grad_norm": 0.05323360115289688, "lr": 4.228411776836306e-05, "ex_per_s": 352.083036032681, "data_frac": 0.022695255075463473, "mem_mb": 122.998784, "nll": -5.918457984924316, "elapsed_s": 632.6}181{"event": "train", "step": 3500, "loss": -6.037656784057617, "grad_norm": 0.07289581745862961, "lr": 4.137060090156723e-05, "ex_per_s": 353.3436093324443, "data_frac": 0.022666077983856753, "mem_mb": 122.998784, "nll": -6.037656784057617, "elapsed_s": 634.0}182{"event": "val", "step": 3500, "val_nll": -6.45034669881108, "best": -6.45034669881108, "bad": 0, "mem_mb": 122.998784, "elapsed_s": 697.0}183{"event": "train", "step": 3520, "loss": -5.938324928283691, "grad_norm": 0.45730358362197876, "lr": 4.049088307805306e-05, "ex_per_s": 322.560222274545, "data_frac": 0.02059276938884729, "mem_mb": 122.998784, "nll": -5.938324928283691, "elapsed_s": 698.4}184{"event": "train", "step": 3540, "loss": -6.136664867401123, "grad_norm": 0.6080268025398254, "lr": 3.964520480388956e-05, "ex_per_s": 323.7576423709366, "data_frac": 0.020570182240193566, "mem_mb": 122.998784, "nll": -6.136664867401123, "elapsed_s": 699.8}185{"event": "train", "step": 3560, "loss": -6.139615535736084, "grad_norm": 0.058858174830675125, "lr": 3.88337972790728e-05, "ex_per_s": 324.9505434385888, "data_frac": 0.020548141803312006, "mem_mb": 122.998784, "nll": -6.139615535736084, "elapsed_s": 701.2}186{"event": "train", "step": 3580, "loss": -5.902353286743164, "grad_norm": 0.2221805304288864, "lr": 3.805688233431824e-05, "ex_per_s": 326.1380956832522, "data_frac": 0.020526421157434322, "mem_mb": 122.998784, "nll": -5.902353286743164, "elapsed_s": 702.5}187{"event": "train", "step": 3600, "loss": -6.098299026489258, "grad_norm": 0.41853854060173035, "lr": 3.7314672370414325e-05, "ex_per_s": 327.3218981096386, "data_frac": 0.020504334490801492, "mem_mb": 122.998784, "nll": -6.098299026489258, "elapsed_s": 703.9}188{"event": "train", "step": 3620, "loss": -5.910061359405518, "grad_norm": 1.020456314086914, "lr": 3.660737030015427e-05, "ex_per_s": 328.5020313857078, "data_frac": 0.020482063974594747, "mem_mb": 122.998784, "nll": -5.910061359405518, "elapsed_s": 705.3}189{"event": "train", "step": 3640, "loss": -6.053691387176514, "grad_norm": 0.566350519657135, "lr": 3.5935169492861716e-05, "ex_per_s": 329.6779516190242, "data_frac": 0.020460041757135245, "mem_mb": 122.998784, "nll": -6.053691387176514, "elapsed_s": 706.6}190{"event": "train", "step": 3660, "loss": -5.8307671546936035, "grad_norm": 0.5456658601760864, "lr": 3.52982537215255e-05, "ex_per_s": 330.8482063220539, "data_frac": 0.020438050935397137, "mem_mb": 122.998784, "nll": -5.8307671546936035, "elapsed_s": 708.0}191{"event": "train", "step": 3680, "loss": -5.99201774597168, "grad_norm": 0.20685671269893646, "lr": 3.469679711255794e-05, "ex_per_s": 332.0155731167048, "data_frac": 0.02041656826434047, "mem_mb": 122.998784, "nll": -5.99201774597168, "elapsed_s": 709.4}192{"event": "train", "step": 3700, "loss": -6.066749095916748, "grad_norm": 0.8447211384773254, "lr": 3.41309640981904e-05, "ex_per_s": 333.17740162495187, "data_frac": 0.02039501693092409, "mem_mb": 122.998784, "nll": -6.066749095916748, "elapsed_s": 710.8}193{"event": "train", "step": 3720, "loss": -5.949129104614258, "grad_norm": 0.6105130314826965, "lr": 3.3600909371519225e-05, "ex_per_s": 334.3359197764061, "data_frac": 0.02037361003324863, "mem_mb": 122.998784, "nll": -5.949129104614258, "elapsed_s": 712.1}194{"event": "train", "step": 3740, "loss": -6.191511154174805, "grad_norm": 0.20429512858390808, "lr": 3.31067778442141e-05, "ex_per_s": 335.4895207245655, "data_frac": 0.02035181073666852, "mem_mb": 122.998784, "nll": -6.191511154174805, "elapsed_s": 713.5}195{"event": "train", "step": 3760, "loss": -6.149302005767822, "grad_norm": 0.2055748552083969, "lr": 3.2648704606900735e-05, "ex_per_s": 336.6375059827093, "data_frac": 0.020330800521036144, "mem_mb": 122.998784, "nll": -6.149302005767822, "elapsed_s": 714.9}196{"event": "train", "step": 3780, "loss": -5.963177680969238, "grad_norm": 0.2792591452598572, "lr": 3.222681489222838e-05, "ex_per_s": 337.7804982748155, "data_frac": 0.02030945952741029, "mem_mb": 122.998784, "nll": -5.963177680969238, "elapsed_s": 716.2}197{"event": "train", "step": 3800, "loss": -6.099578380584717, "grad_norm": 0.08008743822574615, "lr": 3.1841224040632484e-05, "ex_per_s": 338.9202397340902, "data_frac": 0.02028796380638761, "mem_mb": 122.998784, "nll": -6.099578380584717, "elapsed_s": 717.6}198{"event": "train", "step": 3820, "loss": -5.971640586853027, "grad_norm": 0.16242456436157227, "lr": 3.149203746880175e-05, "ex_per_s": 340.0479004849959, "data_frac": 0.020266915840537206, "mem_mb": 122.998784, "nll": -5.971640586853027, "elapsed_s": 719.0}199{"event": "train", "step": 3840, "loss": -5.873305320739746, "grad_norm": 0.6954818964004517, "lr": 3.117935064085835e-05, "ex_per_s": 341.18059906301454, "data_frac": 0.020245756957653148, "mem_mb": 122.998784, "nll": -5.873305320739746, "elapsed_s": 720.3}200{"event": "train", "step": 3860, "loss": -6.016756534576416, "grad_norm": 0.2109784483909607, "lr": 3.090324904225885e-05, "ex_per_s": 342.3071039813143, "data_frac": 0.020224684209066382, "mem_mb": 122.998784, "nll": -6.016756534576416, "elapsed_s": 721.7}201{"event": "train", "step": 3880, "loss": -5.890910625457764, "grad_norm": 0.07784958183765411, "lr": 3.0663808156423456e-05, "ex_per_s": 343.43143172075025, "data_frac": 0.02020401257968359, "mem_mb": 122.998784, "nll": -5.890910625457764, "elapsed_s": 723.1}202{"event": "train", "step": 3900, "loss": -6.181256294250488, "grad_norm": 0.7928140163421631, "lr": 3.046109344409957e-05, "ex_per_s": 344.5518466035305, "data_frac": 0.0201833281078576, "mem_mb": 122.998784, "nll": -6.181256294250488, "elapsed_s": 724.4}203{"event": "train", "step": 3920, "loss": -6.038115978240967, "grad_norm": 0.13669168949127197, "lr": 3.0295160325465435e-05, "ex_per_s": 345.6666461461103, "data_frac": 0.02016236841295051, "mem_mb": 122.998784, "nll": -6.038115978240967, "elapsed_s": 725.8}204{"event": "train", "step": 3940, "loss": -5.922160625457764, "grad_norm": 0.22462010383605957, "lr": 3.016605416497886e-05, "ex_per_s": 346.7701913781109, "data_frac": 0.020141908386351888, "mem_mb": 122.998784, "nll": -5.922160625457764, "elapsed_s": 727.2}205{"event": "train", "step": 3960, "loss": -6.0683369636535645, "grad_norm": 0.08295023441314697, "lr": 3.0073810258974985e-05, "ex_per_s": 347.87781125210233, "data_frac": 0.020121031431613718, "mem_mb": 122.998784, "nll": -6.0683369636535645, "elapsed_s": 728.6}206{"event": "train", "step": 3980, "loss": -5.805968761444092, "grad_norm": 0.221418097615242, "lr": 3.0018453826016683e-05, "ex_per_s": 348.98223828788645, "data_frac": 0.020100659796170865, "mem_mb": 122.998784, "nll": -5.805968761444092, "elapsed_s": 729.9}207{"event": "train", "step": 4000, "loss": -6.057643890380859, "grad_norm": 0.8227823972702026, "lr": 2.9999999999999997e-05, "ex_per_s": 350.0826785186805, "data_frac": 0.020080433998051084, "mem_mb": 122.998784, "nll": -6.057643890380859, "elapsed_s": 731.3}208{"event": "val", "step": 4000, "val_nll": -6.449896290569716, "best": -6.45034669881108, "bad": 1, "mem_mb": 122.998784, "elapsed_s": 793.9}209{"event": "done", "steps": 4000, "best_val": -6.45034669881108, "wall_s": 793.9436521530151, "ex_per_s": 322.4410187118187, "peak_mem_mb": 122.998784, "n_params": 5533443, "device": "mps", "data_time_frac": 0.018494932745327285, "elapsed_s": 793.9}210