SPB Git

spb/wp41_uqo

Public
51.5 KB · 210 lines
1{"event": "train", "step": 20, "loss": -6.12130880355835, "grad_norm": 1.216960072517395, "lr": 3.149999999999999e-05, "ex_per_s": 90.21529772728479, "data_frac": 0.867320454100762, "mem_mb": 91.215104, "nll": -6.12130880355835, "elapsed_s": 14.2}2{"event": "train", "step": 40, "loss": -5.900662422180176, "grad_norm": 0.2182283103466034, "lr": 6.149999999999999e-05, "ex_per_s": 160.17310731302499, "data_frac": 0.7738490659752049, "mem_mb": 91.215104, "nll": -5.900662422180176, "elapsed_s": 16.0}3{"event": "train", "step": 60, "loss": -5.9788103103637695, "grad_norm": 0.5164604783058167, "lr": 9.149999999999999e-05, "ex_per_s": 216.1678323119956, "data_frac": 0.6980814766704329, "mem_mb": 91.215104, "nll": -5.9788103103637695, "elapsed_s": 17.8}4{"event": "train", "step": 80, "loss": -6.077509880065918, "grad_norm": 0.7894268035888672, "lr": 0.0001215, "ex_per_s": 263.3435052350635, "data_frac": 0.6384161293662042, "mem_mb": 91.215104, "nll": -6.077509880065918, "elapsed_s": 19.5}5{"event": "train", "step": 100, "loss": -5.9509172439575195, "grad_norm": 0.2042604386806488, "lr": 0.0001515, "ex_per_s": 303.1157572985438, "data_frac": 0.5883285788423765, "mem_mb": 91.215104, "nll": -5.9509172439575195, "elapsed_s": 21.2}6{"event": "train", "step": 120, "loss": -5.911503314971924, "grad_norm": 0.7511385679244995, "lr": 0.00018149999999999997, "ex_per_s": 337.1068382877, "data_frac": 0.5463020409014148, "mem_mb": 91.215104, "nll": -5.911503314971924, "elapsed_s": 22.8}7{"event": "train", "step": 140, "loss": -5.8316497802734375, "grad_norm": 0.5569416880607605, "lr": 0.00021149999999999996, "ex_per_s": 366.48806932980676, "data_frac": 0.5095367102064066, "mem_mb": 91.215104, "nll": -5.8316497802734375, "elapsed_s": 24.5}8{"event": "train", "step": 160, "loss": -5.736629009246826, "grad_norm": 0.7330977320671082, "lr": 0.0002415, "ex_per_s": 392.09100465909904, "data_frac": 0.47739594122194307, "mem_mb": 91.215104, "nll": -5.736629009246826, "elapsed_s": 26.2}9{"event": "train", "step": 180, "loss": -5.975992202758789, "grad_norm": 0.9773105382919312, "lr": 0.0002715, "ex_per_s": 414.668049042764, "data_frac": 0.4492119631515624, "mem_mb": 91.215104, "nll": -5.975992202758789, "elapsed_s": 27.8}10{"event": "train", "step": 200, "loss": -5.971673011779785, "grad_norm": 0.5980992317199707, "lr": 0.0003, "ex_per_s": 434.6586171096504, "data_frac": 0.4241576852997292, "mem_mb": 91.215104, "nll": -5.971673011779785, "elapsed_s": 29.5}11{"event": "train", "step": 220, "loss": -6.029297828674316, "grad_norm": 1.1882438659667969, "lr": 0.00029998154617398326, "ex_per_s": 452.6849689754217, "data_frac": 0.40192805999298054, "mem_mb": 91.215104, "nll": -6.029297828674316, "elapsed_s": 31.1}12{"event": "train", "step": 240, "loss": -5.915030002593994, "grad_norm": 0.22184117138385773, "lr": 0.000299926189741025, "ex_per_s": 468.9083139594664, "data_frac": 0.38196239620006767, "mem_mb": 91.215104, "nll": -5.915030002593994, "elapsed_s": 32.8}13{"event": "train", "step": 260, "loss": -5.888173580169678, "grad_norm": 0.12297594547271729, "lr": 0.0002998339458350211, "ex_per_s": 483.58142218929544, "data_frac": 0.363931583569444, "mem_mb": 91.215104, "nll": -5.888173580169678, "elapsed_s": 34.5}14{"event": "train", "step": 280, "loss": -5.944266319274902, "grad_norm": 0.13406462967395782, "lr": 0.0002997048396745345, "ex_per_s": 496.8871484916034, "data_frac": 0.34754078571339325, "mem_mb": 91.215104, "nll": -5.944266319274902, "elapsed_s": 36.1}15{"event": "train", "step": 300, "loss": -5.968670845031738, "grad_norm": 0.17011503875255585, "lr": 0.0002995389065559004, "ex_per_s": 509.02807726230395, "data_frac": 0.3325872198301561, "mem_mb": 91.215104, "nll": -5.968670845031738, "elapsed_s": 37.8}16{"event": "train", "step": 320, "loss": -5.965337753295898, "grad_norm": 0.6389368772506714, "lr": 0.0002993361918435765, "ex_per_s": 520.1765762825228, "data_frac": 0.3189043084301191, "mem_mb": 91.215104, "nll": -5.965337753295898, "elapsed_s": 39.4}17{"event": "train", "step": 340, "loss": -5.835088729858398, "grad_norm": 0.3528596758842468, "lr": 0.00029909675095774114, "ex_per_s": 530.4178145470677, "data_frac": 0.30632204267852325, "mem_mb": 91.215104, "nll": -5.835088729858398, "elapsed_s": 41.1}18{"event": "train", "step": 360, "loss": -5.99469518661499, "grad_norm": 0.30039846897125244, "lr": 0.00029882064935914164, "ex_per_s": 539.8818510795322, "data_frac": 0.2947216781788147, "mem_mb": 91.215104, "nll": -5.99469518661499, "elapsed_s": 42.7}19{"event": "train", "step": 380, "loss": -6.005059719085693, "grad_norm": 1.156945824623108, "lr": 0.00029850796253119824, "ex_per_s": 548.6216868745911, "data_frac": 0.28397673962407544, "mem_mb": 91.215104, "nll": -6.005059719085693, "elapsed_s": 44.4}20{"event": "train", "step": 400, "loss": -5.995846748352051, "grad_norm": 1.3440799713134766, "lr": 0.0002981587759593675, "ex_per_s": 556.6971233510234, "data_frac": 0.27398993245443215, "mem_mb": 91.215104, "nll": -5.995846748352051, "elapsed_s": 46.0}21{"event": "train", "step": 420, "loss": -5.9649224281311035, "grad_norm": 0.5912901759147644, "lr": 0.0002977731851077716, "ex_per_s": 564.2208796927771, "data_frac": 0.2647222941423304, "mem_mb": 91.215104, "nll": -5.9649224281311035, "elapsed_s": 47.7}22{"event": "train", "step": 440, "loss": -5.93986701965332, "grad_norm": 0.54170161485672, "lr": 0.0002973512953930992, "ex_per_s": 571.2457551443947, "data_frac": 0.256060106949337, "mem_mb": 91.215104, "nll": -5.93986701965332, "elapsed_s": 49.3}23{"event": "train", "step": 460, "loss": -6.101871967315674, "grad_norm": 0.6077643036842346, "lr": 0.0002968932221557859, "ex_per_s": 577.8282447287402, "data_frac": 0.24796810328705438, "mem_mb": 91.215104, "nll": -6.101871967315674, "elapsed_s": 51.0}24{"event": "train", "step": 480, "loss": -5.878057479858398, "grad_norm": 0.3371906876564026, "lr": 0.0002963990906284808, "ex_per_s": 583.9955442862921, "data_frac": 0.24038454892391373, "mem_mb": 91.215104, "nll": -5.878057479858398, "elapsed_s": 52.6}25{"event": "train", "step": 500, "loss": -5.925525188446045, "grad_norm": 0.5448274612426758, "lr": 0.00029586903590180956, "ex_per_s": 589.8001819218526, "data_frac": 0.233268386981472, "mem_mb": 91.215104, "nll": -5.925525188446045, "elapsed_s": 54.3}26{"event": "val", "step": 500, "val_nll": -6.4401472527837145, "best": -6.4401472527837145, "bad": 0, "mem_mb": 91.215104, "elapsed_s": 119.1}27{"event": "train", "step": 520, "loss": -6.040589332580566, "grad_norm": 0.2938709855079651, "lr": 0.00029530320288744205, "ex_per_s": 275.04756887743076, "data_frac": 0.10469991071395812, "mem_mb": 91.215104, "nll": -6.040589332580566, "elapsed_s": 121.0}28{"event": "train", "step": 540, "loss": -5.879671096801758, "grad_norm": 0.5306766033172607, "lr": 0.0002947017462784745, "ex_per_s": 281.7380109693555, "data_frac": 0.1033710500632615, "mem_mb": 91.215104, "nll": -5.879671096801758, "elapsed_s": 122.7}29{"event": "train", "step": 560, "loss": -5.972919940948486, "grad_norm": 0.23542259633541107, "lr": 0.00029406483050713824, "ex_per_s": 288.2494172504453, "data_frac": 0.1020686561745317, "mem_mb": 91.215104, "nll": -5.972919940948486, "elapsed_s": 124.4}30{"event": "train", "step": 580, "loss": -5.925448894500732, "grad_norm": 0.11525410413742065, "lr": 0.0002933926296998457, "ex_per_s": 294.5822454642295, "data_frac": 0.10080415882051703, "mem_mb": 91.215104, "nll": -5.925448894500732, "elapsed_s": 126.1}31{"event": "train", "step": 600, "loss": -5.974948883056641, "grad_norm": 0.6361027956008911, "lr": 0.0002926853276295856, "ex_per_s": 300.7496730395515, "data_frac": 0.09957856090735112, "mem_mb": 91.215104, "nll": -5.974948883056641, "elapsed_s": 127.7}32{"event": "train", "step": 620, "loss": -5.953248023986816, "grad_norm": 0.3822346329689026, "lr": 0.00029194311766568174, "ex_per_s": 306.7716073036357, "data_frac": 0.09838184715846922, "mem_mb": 91.215104, "nll": -5.953248023986816, "elapsed_s": 129.4}33{"event": "train", "step": 640, "loss": -6.066324710845947, "grad_norm": 0.39144250750541687, "lr": 0.0002911662027209272, "ex_per_s": 312.6376688119555, "data_frac": 0.09721209496037142, "mem_mb": 91.215104, "nll": -6.066324710845947, "elapsed_s": 131.1}34{"event": "train", "step": 660, "loss": -6.008216381072998, "grad_norm": 1.0173797607421875, "lr": 0.00029035479519611045, "ex_per_s": 318.3557240701155, "data_frac": 0.09607152473619295, "mem_mb": 91.215104, "nll": -6.008216381072998, "elapsed_s": 132.7}35{"event": "train", "step": 680, "loss": -6.007107734680176, "grad_norm": 0.12109087407588959, "lr": 0.0002895091169219469, "ex_per_s": 323.92992047936605, "data_frac": 0.0949573079074645, "mem_mb": 91.215104, "nll": -6.007107734680176, "elapsed_s": 134.4}36{"event": "train", "step": 700, "loss": -5.984099388122559, "grad_norm": 1.0346158742904663, "lr": 0.00028862939909843273, "ex_per_s": 329.365685006289, "data_frac": 0.09387267423242836, "mem_mb": 91.215104, "nll": -5.984099388122559, "elapsed_s": 136.1}37{"event": "train", "step": 720, "loss": -6.062838554382324, "grad_norm": 0.5053321123123169, "lr": 0.0002877158822316369, "ex_per_s": 334.67296407626037, "data_frac": 0.09281366419329386, "mem_mb": 91.215104, "nll": -6.062838554382324, "elapsed_s": 137.7}38{"event": "train", "step": 740, "loss": -6.0340166091918945, "grad_norm": 0.22796742618083954, "lr": 0.00028676881606794945, "ex_per_s": 339.855136702999, "data_frac": 0.09178259279270089, "mem_mb": 91.215104, "nll": -6.0340166091918945, "elapsed_s": 139.4}39{"event": "train", "step": 760, "loss": -5.9542036056518555, "grad_norm": 1.3064792156219482, "lr": 0.000285788459525803, "ex_per_s": 344.91350448518784, "data_frac": 0.09077533881542016, "mem_mb": 91.215104, "nll": -5.9542036056518555, "elapsed_s": 141.1}40{"event": "train", "step": 780, "loss": -5.897231578826904, "grad_norm": 0.30758216977119446, "lr": 0.00028477508062488754, "ex_per_s": 349.85655703848255, "data_frac": 0.08979466611204615, "mem_mb": 91.215104, "nll": -5.897231578826904, "elapsed_s": 142.7}41{"event": "train", "step": 800, "loss": -5.817168712615967, "grad_norm": 0.15203383564949036, "lr": 0.000283728956412876, "ex_per_s": 354.68265173458786, "data_frac": 0.08883300526851357, "mem_mb": 91.215104, "nll": -5.817168712615967, "elapsed_s": 144.4}42{"event": "train", "step": 820, "loss": -5.934083938598633, "grad_norm": 0.7080979943275452, "lr": 0.00028265037288968236, "ex_per_s": 359.39320023924296, "data_frac": 0.08789347110861015, "mem_mb": 91.215104, "nll": -5.934083938598633, "elapsed_s": 146.1}43{"event": "train", "step": 840, "loss": -5.808615684509277, "grad_norm": 0.49091434478759766, "lr": 0.0002815396249292725, "ex_per_s": 363.99203165232785, "data_frac": 0.08698078416407562, "mem_mb": 91.215104, "nll": -5.808615684509277, "elapsed_s": 147.7}44{"event": "train", "step": 860, "loss": -5.915010452270508, "grad_norm": 1.5851658582687378, "lr": 0.00028039701619904774, "ex_per_s": 368.4863929569694, "data_frac": 0.08608334343984238, "mem_mb": 91.215104, "nll": -5.915010452270508, "elapsed_s": 149.4}45{"event": "train", "step": 880, "loss": -5.872539043426514, "grad_norm": 0.3804089426994324, "lr": 0.00027922285907682625, "ex_per_s": 372.88248908098706, "data_frac": 0.08520400472260274, "mem_mb": 91.215104, "nll": -5.872539043426514, "elapsed_s": 151.1}46{"event": "train", "step": 900, "loss": -5.987321376800537, "grad_norm": 0.47866714000701904, "lr": 0.00027801747456544134, "ex_per_s": 377.185685709864, "data_frac": 0.08434713577222458, "mem_mb": 91.215104, "nll": -5.987321376800537, "elapsed_s": 152.8}47{"event": "train", "step": 920, "loss": -6.027098655700684, "grad_norm": 0.5395747423171997, "lr": 0.00027678119220498265, "ex_per_s": 381.39690237181145, "data_frac": 0.08350638438218815, "mem_mb": 91.215104, "nll": -6.027098655700684, "elapsed_s": 154.4}48{"event": "train", "step": 940, "loss": -6.082517147064209, "grad_norm": 0.6361061930656433, "lr": 0.0002755143499827035, "ex_per_s": 385.5137441049608, "data_frac": 0.08268701157549004, "mem_mb": 91.215104, "nll": -6.082517147064209, "elapsed_s": 156.1}49{"event": "train", "step": 960, "loss": -5.967758655548096, "grad_norm": 0.20414111018180847, "lr": 0.00027421729424061787, "ex_per_s": 389.55091824585, "data_frac": 0.08188237020322685, "mem_mb": 91.215104, "nll": -5.967758655548096, "elapsed_s": 157.8}50{"event": "train", "step": 980, "loss": -5.927812576293945, "grad_norm": 0.29191315174102783, "lr": 0.00027289037958081483, "ex_per_s": 393.5084481992346, "data_frac": 0.08109739789601339, "mem_mb": 91.215104, "nll": -5.927812576293945, "elapsed_s": 159.4}51{"event": "train", "step": 1000, "loss": -6.073009967803955, "grad_norm": 0.30325040221214294, "lr": 0.0002715339687685131, "ex_per_s": 397.38448956205576, "data_frac": 0.08032547234766159, "mem_mb": 91.215104, "nll": -6.073009967803955, "elapsed_s": 161.1}52{"event": "val", "step": 1000, "val_nll": -6.447027181473679, "best": -6.447027181473679, "bad": 0, "mem_mb": 91.215104, "elapsed_s": 224.6}53{"event": "train", "step": 1020, "loss": -6.0304436683654785, "grad_norm": 0.1824396252632141, "lr": 0.0002701484326328852, "ex_per_s": 288.41026914330774, "data_frac": 0.05720510863303138, "mem_mb": 91.215104, "nll": -6.0304436683654785, "elapsed_s": 226.4}54{"event": "train", "step": 1040, "loss": -5.97283935546875, "grad_norm": 0.16534101963043213, "lr": 0.00026873414996567596, "ex_per_s": 291.8998855578533, "data_frac": 0.056837499299299016, "mem_mb": 91.215104, "nll": -5.97283935546875, "elapsed_s": 228.1}55{"event": "train", "step": 1060, "loss": -6.183597087860107, "grad_norm": 0.239704892039299, "lr": 0.0002672915074176452, "ex_per_s": 295.34594189704507, "data_frac": 0.05647432530982879, "mem_mb": 91.215104, "nll": -6.183597087860107, "elapsed_s": 229.7}56{"event": "train", "step": 1080, "loss": -5.815810203552246, "grad_norm": 0.403895765542984, "lr": 0.00026582089939286075, "ex_per_s": 298.7408965267191, "data_frac": 0.05611331332273867, "mem_mb": 91.215104, "nll": -5.815810203552246, "elapsed_s": 231.4}57{"event": "train", "step": 1100, "loss": -6.054096221923828, "grad_norm": 0.12985016405582428, "lr": 0.0002643227279408728, "ex_per_s": 302.0869911170492, "data_frac": 0.05575864481402415, "mem_mb": 91.215104, "nll": -6.054096221923828, "elapsed_s": 233.1}58{"event": "train", "step": 1120, "loss": -6.053192138671875, "grad_norm": 0.3413015305995941, "lr": 0.000262797402646797, "ex_per_s": 305.38661004563824, "data_frac": 0.05540883190794203, "mem_mb": 91.215104, "nll": -6.053192138671875, "elapsed_s": 234.8}59{"event": "train", "step": 1140, "loss": -6.019399642944336, "grad_norm": 0.2820770740509033, "lr": 0.0002612453405193386, "ex_per_s": 308.6379865770537, "data_frac": 0.05506360347704356, "mem_mb": 91.215104, "nll": -6.019399642944336, "elapsed_s": 236.4}60{"event": "train", "step": 1160, "loss": -5.97030782699585, "grad_norm": 0.5139480233192444, "lr": 0.0002596669658767859, "ex_per_s": 311.84524820390317, "data_frac": 0.05472622406600133, "mem_mb": 91.215104, "nll": -5.97030782699585, "elapsed_s": 238.1}61{"event": "train", "step": 1180, "loss": -5.957934856414795, "grad_norm": 0.12332708388566971, "lr": 0.00025806271023100637, "ex_per_s": 315.00975569662955, "data_frac": 0.05438901886691945, "mem_mb": 91.215104, "nll": -5.957934856414795, "elapsed_s": 239.8}62{"event": "train", "step": 1200, "loss": -5.943552494049072, "grad_norm": 0.23218068480491638, "lr": 0.00025643301216947504, "ex_per_s": 318.1308503023744, "data_frac": 0.05405840860519645, "mem_mb": 91.215104, "nll": -5.943552494049072, "elapsed_s": 241.5}63{"event": "train", "step": 1220, "loss": -6.032486915588379, "grad_norm": 0.6266744136810303, "lr": 0.0002547783172353695, "ex_per_s": 321.211558782192, "data_frac": 0.05373186902619069, "mem_mb": 91.215104, "nll": -6.032486915588379, "elapsed_s": 243.1}64{"event": "train", "step": 1240, "loss": -6.060224533081055, "grad_norm": 0.07179857045412064, "lr": 0.00025309907780576277, "ex_per_s": 324.247983422278, "data_frac": 0.05341040106494013, "mem_mb": 91.215104, "nll": -6.060224533081055, "elapsed_s": 244.8}65{"event": "train", "step": 1260, "loss": -6.052621364593506, "grad_norm": 0.1764248013496399, "lr": 0.0002513957529679477, "ex_per_s": 327.2423377216855, "data_frac": 0.053093030302045194, "mem_mb": 91.215104, "nll": -6.052621364593506, "elapsed_s": 246.5}66{"event": "train", "step": 1280, "loss": -6.043973445892334, "grad_norm": 0.5582030415534973, "lr": 0.00024966880839392746, "ex_per_s": 330.19732285792355, "data_frac": 0.052783569299156584, "mem_mb": 91.215104, "nll": -6.043973445892334, "elapsed_s": 248.1}67{"event": "train", "step": 1300, "loss": -5.8696064949035645, "grad_norm": 0.7024678587913513, "lr": 0.0002479187162131052, "ex_per_s": 333.11702881721084, "data_frac": 0.05247619321575219, "mem_mb": 91.215104, "nll": -5.8696064949035645, "elapsed_s": 249.8}68{"event": "train", "step": 1320, "loss": -5.830576419830322, "grad_norm": 0.19732947647571564, "lr": 0.00024614595488320833, "ex_per_s": 336.00001779686414, "data_frac": 0.052173002660226286, "mem_mb": 91.215104, "nll": -5.830576419830322, "elapsed_s": 251.5}69{"event": "train", "step": 1340, "loss": -5.951085567474365, "grad_norm": 0.3023446798324585, "lr": 0.00024435100905948387, "ex_per_s": 338.8475704834313, "data_frac": 0.05187295605881162, "mem_mb": 91.215104, "nll": -5.951085567474365, "elapsed_s": 253.1}70{"event": "train", "step": 1360, "loss": -6.039588928222656, "grad_norm": 0.6909679174423218, "lr": 0.0002425343694621974, "ex_per_s": 341.6483034214531, "data_frac": 0.0515788024515089, "mem_mb": 91.215104, "nll": -6.039588928222656, "elapsed_s": 254.8}71{"event": "train", "step": 1380, "loss": -6.069843769073486, "grad_norm": 0.24118199944496155, "lr": 0.00024069653274247583, "ex_per_s": 344.4189547846463, "data_frac": 0.05128724639009252, "mem_mb": 91.215104, "nll": -6.069843769073486, "elapsed_s": 256.5}72{"event": "train", "step": 1400, "loss": -5.797152519226074, "grad_norm": 0.1537817418575287, "lr": 0.0002388380013465276, "ex_per_s": 347.1551637403924, "data_frac": 0.050999069570427324, "mem_mb": 91.215104, "nll": -5.797152519226074, "elapsed_s": 258.1}73{"event": "train", "step": 1420, "loss": -5.988410472869873, "grad_norm": 0.8317583203315735, "lr": 0.00023695928337827914, "ex_per_s": 349.85578209009924, "data_frac": 0.050715054469797564, "mem_mb": 91.215104, "nll": -5.988410472869873, "elapsed_s": 259.8}74{"event": "train", "step": 1440, "loss": -5.743403911590576, "grad_norm": 0.2988632917404175, "lr": 0.00023506089246046403, "ex_per_s": 352.52280807104285, "data_frac": 0.05043437898422547, "mem_mb": 91.215104, "nll": -5.743403911590576, "elapsed_s": 261.5}75{"event": "train", "step": 1460, "loss": -5.803734302520752, "grad_norm": 0.36675208806991577, "lr": 0.00023314334759420388, "ex_per_s": 355.15792330839827, "data_frac": 0.050157090705452705, "mem_mb": 91.215104, "nll": -5.803734302520752, "elapsed_s": 263.1}76{"event": "train", "step": 1480, "loss": -6.047390460968018, "grad_norm": 0.1032465398311615, "lr": 0.0002312071730171186, "ex_per_s": 357.7576979106161, "data_frac": 0.04988353779944231, "mem_mb": 91.215104, "nll": -6.047390460968018, "elapsed_s": 264.8}77{"event": "train", "step": 1500, "loss": -5.958870887756348, "grad_norm": 0.25100523233413696, "lr": 0.0002292528980600049, "ex_per_s": 360.32647312322376, "data_frac": 0.049613813032831466, "mem_mb": 91.215104, "nll": -5.958870887756348, "elapsed_s": 266.5}78{"event": "val", "step": 1500, "val_nll": -6.4477003382175075, "best": -6.4477003382175075, "bad": 0, "mem_mb": 91.215104, "elapsed_s": 330.1}79{"event": "train", "step": 1520, "loss": -5.9513020515441895, "grad_norm": 0.5979319214820862, "lr": 0.00022728105700212285, "ex_per_s": 293.1401893157403, "data_frac": 0.03986617670022247, "mem_mb": 91.215104, "nll": -5.9513020515441895, "elapsed_s": 331.9}80{"event": "train", "step": 1540, "loss": -5.879368782043457, "grad_norm": 0.10964525490999222, "lr": 0.00022529218892512886, "ex_per_s": 295.5065384684792, "data_frac": 0.03969914934822153, "mem_mb": 91.215104, "nll": -5.879368782043457, "elapsed_s": 333.6}81{"event": "train", "step": 1560, "loss": -5.9830641746521, "grad_norm": 0.3881894648075104, "lr": 0.00022328683756569652, "ex_per_s": 297.8514332887254, "data_frac": 0.0395338474864762, "mem_mb": 91.215104, "nll": -5.9830641746521, "elapsed_s": 335.2}82{"event": "train", "step": 1580, "loss": -5.982606410980225, "grad_norm": 0.07147280871868134, "lr": 0.0002212655511668642, "ex_per_s": 300.1712122503152, "data_frac": 0.03937066132496763, "mem_mb": 91.215104, "nll": -5.982606410980225, "elapsed_s": 336.9}83{"event": "train", "step": 1600, "loss": -5.956474781036377, "grad_norm": 0.19703662395477295, "lr": 0.00021922888232815089, "ex_per_s": 302.4702874501538, "data_frac": 0.03920846273828501, "mem_mb": 91.215104, "nll": -5.956474781036377, "elapsed_s": 338.6}84{"event": "train", "step": 1620, "loss": -5.88133430480957, "grad_norm": 0.13996350765228271, "lr": 0.00021717738785448102, "ex_per_s": 304.7455389545542, "data_frac": 0.039048671954738504, "mem_mb": 91.215104, "nll": -5.88133430480957, "elapsed_s": 340.3}85{"event": "train", "step": 1640, "loss": -6.0767412185668945, "grad_norm": 0.15043489634990692, "lr": 0.00021511162860395977, "ex_per_s": 307.00122635593647, "data_frac": 0.03888995171043059, "mem_mb": 91.215104, "nll": -6.0767412185668945, "elapsed_s": 341.9}86{"event": "train", "step": 1660, "loss": -6.0420379638671875, "grad_norm": 0.7123115062713623, "lr": 0.0002130321693345397, "ex_per_s": 309.2340132385466, "data_frac": 0.03873269421646796, "mem_mb": 91.215104, "nll": -6.0420379638671875, "elapsed_s": 343.6}87{"event": "train", "step": 1680, "loss": -5.865314960479736, "grad_norm": 0.17836101353168488, "lr": 0.0002109395785496218, "ex_per_s": 311.44529571225166, "data_frac": 0.03857737318895883, "mem_mb": 91.215104, "nll": -5.865314960479736, "elapsed_s": 345.3}88{"event": "train", "step": 1700, "loss": -5.926872730255127, "grad_norm": 0.5828495621681213, "lr": 0.00020883442834263225, "ex_per_s": 313.63558502115757, "data_frac": 0.03842412843107331, "mem_mb": 91.215104, "nll": -5.926872730255127, "elapsed_s": 346.9}89{"event": "train", "step": 1720, "loss": -5.887933254241943, "grad_norm": 0.7826312780380249, "lr": 0.00020671729424061788, "ex_per_s": 315.8065253164275, "data_frac": 0.03827319553261655, "mem_mb": 91.215104, "nll": -5.887933254241943, "elapsed_s": 348.6}90{"event": "train", "step": 1740, "loss": -6.273977756500244, "grad_norm": 0.28695547580718994, "lr": 0.00020458875504690275, "ex_per_s": 317.95638919180976, "data_frac": 0.038123104186388965, "mem_mb": 91.215104, "nll": -6.273977756500244, "elapsed_s": 350.3}91{"event": "train", "step": 1760, "loss": -5.9323835372924805, "grad_norm": 0.1828981190919876, "lr": 0.000202449392682849, "ex_per_s": 320.0889338152773, "data_frac": 0.037973012195067955, "mem_mb": 91.215104, "nll": -5.9323835372924805, "elapsed_s": 351.9}92{"event": "train", "step": 1780, "loss": -5.871510028839111, "grad_norm": 0.4624090790748596, "lr": 0.00020029979202876563, "ex_per_s": 322.2013265828262, "data_frac": 0.03782529134108226, "mem_mb": 91.215104, "nll": -5.871510028839111, "elapsed_s": 353.6}93{"event": "train", "step": 1800, "loss": -6.048696994781494, "grad_norm": 1.1054823398590088, "lr": 0.00019814054076400787, "ex_per_s": 324.29559621136985, "data_frac": 0.03767814141162565, "mem_mb": 91.215104, "nll": -6.048696994781494, "elapsed_s": 355.3}94{"event": "train", "step": 1820, "loss": -5.9351606369018555, "grad_norm": 0.8482238054275513, "lr": 0.00019597222920631147, "ex_per_s": 326.3721590319129, "data_frac": 0.03753347438371799, "mem_mb": 91.215104, "nll": -5.9351606369018555, "elapsed_s": 356.9}95{"event": "train", "step": 1840, "loss": -6.07248592376709, "grad_norm": 0.3304218649864197, "lr": 0.0001937954501504058, "ex_per_s": 328.42874807418553, "data_frac": 0.03739037280925427, "mem_mb": 91.215104, "nll": -6.07248592376709, "elapsed_s": 358.6}96{"event": "train", "step": 1860, "loss": -6.052064418792725, "grad_norm": 0.18450504541397095, "lr": 0.0001916107987059498, "ex_per_s": 330.46589490972843, "data_frac": 0.037248447786227976, "mem_mb": 91.215104, "nll": -6.052064418792725, "elapsed_s": 360.3}97{"event": "train", "step": 1880, "loss": -5.949106216430664, "grad_norm": 0.6017156839370728, "lr": 0.00018941887213483484, "ex_per_s": 332.4847033537508, "data_frac": 0.03710828454483193, "mem_mb": 91.215104, "nll": -5.949106216430664, "elapsed_s": 361.9}98{"event": "train", "step": 1900, "loss": -5.984969615936279, "grad_norm": 0.22655457258224487, "lr": 0.00018722026968789906, "ex_per_s": 334.4856787682205, "data_frac": 0.03696887836717822, "mem_mb": 91.215104, "nll": -5.984969615936279, "elapsed_s": 363.6}99{"event": "train", "step": 1920, "loss": -5.964358329772949, "grad_norm": 0.41785281896591187, "lr": 0.00018501559244109864, "ex_per_s": 336.46876881854973, "data_frac": 0.03683119249521182, "mem_mb": 91.215104, "nll": -5.964358329772949, "elapsed_s": 365.2}100{"event": "train", "step": 1940, "loss": -6.050399303436279, "grad_norm": 0.3159654438495636, "lr": 0.00018280544313117932, "ex_per_s": 338.4303895984325, "data_frac": 0.03669515802252464, "mem_mb": 91.215104, "nll": -6.050399303436279, "elapsed_s": 366.9}101{"event": "train", "step": 1960, "loss": -5.9387526512146, "grad_norm": 0.07015792280435562, "lr": 0.00018059042599089415, "ex_per_s": 340.37406760060725, "data_frac": 0.03656181533616253, "mem_mb": 91.215104, "nll": -5.9387526512146, "elapsed_s": 368.6}102{"event": "train", "step": 1980, "loss": -5.966446399688721, "grad_norm": 0.42941656708717346, "lr": 0.00017837114658381247, "ex_per_s": 342.3018960026013, "data_frac": 0.03642854483622673, "mem_mb": 91.215104, "nll": -5.966446399688721, "elapsed_s": 370.2}103{"event": "train", "step": 2000, "loss": -5.916291236877441, "grad_norm": 0.4670546352863312, "lr": 0.00017614821163876486, "ex_per_s": 344.21451900683695, "data_frac": 0.036294932298218834, "mem_mb": 91.215104, "nll": -5.916291236877441, "elapsed_s": 371.9}104{"event": "val", "step": 2000, "val_nll": -6.4491966507070115, "best": -6.4491966507070115, "bad": 0, "mem_mb": 91.215104, "elapsed_s": 435.3}105{"event": "train", "step": 2020, "loss": -5.927700042724609, "grad_norm": 0.5770193934440613, "lr": 0.00017392222888396941, "ex_per_s": 295.83357090446776, "data_frac": 0.030911128246102534, "mem_mb": 91.215104, "nll": -5.927700042724609, "elapsed_s": 437.0}106{"event": "train", "step": 2040, "loss": -5.874471187591553, "grad_norm": 0.4356057047843933, "lr": 0.000171693806880885, "ex_per_s": 297.62212444606115, "data_frac": 0.030818295648068535, "mem_mb": 91.215104, "nll": -5.874471187591553, "elapsed_s": 438.7}107{"event": "train", "step": 2060, "loss": -6.144719123840332, "grad_norm": 0.3447990417480469, "lr": 0.0001694635548578366, "ex_per_s": 299.397437670697, "data_frac": 0.03072635416811397, "mem_mb": 91.215104, "nll": -6.144719123840332, "elapsed_s": 440.4}108{"event": "train", "step": 2080, "loss": -5.963033199310303, "grad_norm": 0.3682815134525299, "lr": 0.0001672320825434584, "ex_per_s": 301.15958248857163, "data_frac": 0.030634920626016613, "mem_mb": 91.215104, "nll": -5.963033199310303, "elapsed_s": 442.1}109{"event": "train", "step": 2100, "loss": -5.899506092071533, "grad_norm": 1.0498464107513428, "lr": 0.000165, "ex_per_s": 302.9076989227332, "data_frac": 0.030543974911120646, "mem_mb": 91.215104, "nll": -5.899506092071533, "elapsed_s": 443.7}110{"event": "train", "step": 2120, "loss": -6.150656223297119, "grad_norm": 0.4923146963119507, "lr": 0.0001627679174565416, "ex_per_s": 304.64289082246637, "data_frac": 0.030453403479536383, "mem_mb": 91.215104, "nll": -6.150656223297119, "elapsed_s": 445.4}111{"event": "train", "step": 2140, "loss": -5.9782633781433105, "grad_norm": 0.0980057641863823, "lr": 0.00016053644514216336, "ex_per_s": 306.36577334974316, "data_frac": 0.030363592117890855, "mem_mb": 91.215104, "nll": -5.9782633781433105, "elapsed_s": 447.1}112{"event": "train", "step": 2160, "loss": -5.927781581878662, "grad_norm": 0.36891186237335205, "lr": 0.00015830619311911498, "ex_per_s": 308.0732151478528, "data_frac": 0.03027411581791975, "mem_mb": 91.215104, "nll": -5.927781581878662, "elapsed_s": 448.8}113{"event": "train", "step": 2180, "loss": -6.0838727951049805, "grad_norm": 0.34346652030944824, "lr": 0.00015607777111603058, "ex_per_s": 309.7689655159673, "data_frac": 0.030188353680090294, "mem_mb": 91.215104, "nll": -6.0838727951049805, "elapsed_s": 450.4}114{"event": "train", "step": 2200, "loss": -5.963127613067627, "grad_norm": 0.5841734409332275, "lr": 0.00015385178836123514, "ex_per_s": 311.4552285563746, "data_frac": 0.03010152815015165, "mem_mb": 91.215104, "nll": -5.963127613067627, "elapsed_s": 452.1}115{"event": "train", "step": 2220, "loss": -5.828350067138672, "grad_norm": 0.5744604468345642, "lr": 0.00015162885341618752, "ex_per_s": 313.12946408301974, "data_frac": 0.030014674299824944, "mem_mb": 91.215104, "nll": -5.828350067138672, "elapsed_s": 453.8}116{"event": "train", "step": 2240, "loss": -5.963306427001953, "grad_norm": 0.6331019401550293, "lr": 0.00014940957400910585, "ex_per_s": 314.7932390929224, "data_frac": 0.029928685972532804, "mem_mb": 91.215104, "nll": -5.963306427001953, "elapsed_s": 455.5}117{"event": "train", "step": 2260, "loss": -5.73758602142334, "grad_norm": 0.7794532179832458, "lr": 0.00014719455686882068, "ex_per_s": 316.44265387691917, "data_frac": 0.02984404565438445, "mem_mb": 91.215104, "nll": -5.73758602142334, "elapsed_s": 457.1}118{"event": "train", "step": 2280, "loss": -6.0114569664001465, "grad_norm": 0.551724910736084, "lr": 0.0001449844075589013, "ex_per_s": 318.08229997767785, "data_frac": 0.029760017245704803, "mem_mb": 91.215104, "nll": -6.0114569664001465, "elapsed_s": 458.8}119{"event": "train", "step": 2300, "loss": -6.002670764923096, "grad_norm": 0.17329198122024536, "lr": 0.0001427797303121009, "ex_per_s": 319.711699357137, "data_frac": 0.02967655507201811, "mem_mb": 91.215104, "nll": -6.002670764923096, "elapsed_s": 460.5}120{"event": "train", "step": 2320, "loss": -6.070153713226318, "grad_norm": 0.33174678683280945, "lr": 0.00014058112786516518, "ex_per_s": 321.329953628346, "data_frac": 0.029593477557245305, "mem_mb": 91.215104, "nll": -6.070153713226318, "elapsed_s": 462.1}121{"event": "train", "step": 2340, "loss": -5.851592063903809, "grad_norm": 0.3826586902141571, "lr": 0.00013838920129405016, "ex_per_s": 322.9368699451218, "data_frac": 0.02951127353239029, "mem_mb": 91.215104, "nll": -5.851592063903809, "elapsed_s": 463.8}122{"event": "train", "step": 2360, "loss": -5.930376052856445, "grad_norm": 0.09487227350473404, "lr": 0.0001362045498495942, "ex_per_s": 324.5324185041799, "data_frac": 0.029429449029854365, "mem_mb": 91.215104, "nll": -5.930376052856445, "elapsed_s": 465.5}123{"event": "train", "step": 2380, "loss": -5.932293891906738, "grad_norm": 0.20008136332035065, "lr": 0.00013402777079368853, "ex_per_s": 326.11751021346146, "data_frac": 0.029348140194755866, "mem_mb": 91.215104, "nll": -5.932293891906738, "elapsed_s": 467.1}124{"event": "train", "step": 2400, "loss": -5.9557905197143555, "grad_norm": 0.9805769920349121, "lr": 0.00013185945923599213, "ex_per_s": 327.68821553930684, "data_frac": 0.02926737296923238, "mem_mb": 91.215104, "nll": -5.9557905197143555, "elapsed_s": 468.8}125{"event": "train", "step": 2420, "loss": -5.995728969573975, "grad_norm": 0.4514530599117279, "lr": 0.00012970020797123432, "ex_per_s": 329.2478517690522, "data_frac": 0.029188963615806928, "mem_mb": 91.215104, "nll": -5.995728969573975, "elapsed_s": 470.4}126{"event": "train", "step": 2440, "loss": -6.098369598388672, "grad_norm": 0.4209986925125122, "lr": 0.000127550607317151, "ex_per_s": 330.7997740780251, "data_frac": 0.02911015783956376, "mem_mb": 91.215104, "nll": -6.098369598388672, "elapsed_s": 472.1}127{"event": "train", "step": 2460, "loss": -6.1350860595703125, "grad_norm": 0.16000571846961975, "lr": 0.00012541124495309727, "ex_per_s": 332.3399316468288, "data_frac": 0.029030648337831165, "mem_mb": 91.215104, "nll": -6.1350860595703125, "elapsed_s": 473.8}128{"event": "train", "step": 2480, "loss": -5.947309970855713, "grad_norm": 0.16889125108718872, "lr": 0.00012328270575938212, "ex_per_s": 333.87233781694124, "data_frac": 0.028952763474746818, "mem_mb": 91.215104, "nll": -5.947309970855713, "elapsed_s": 475.4}129{"event": "train", "step": 2500, "loss": -5.757325649261475, "grad_norm": 0.6998139023780823, "lr": 0.00012116557165736775, "ex_per_s": 335.39392299244565, "data_frac": 0.028876125396725724, "mem_mb": 91.215104, "nll": -5.757325649261475, "elapsed_s": 477.1}130{"event": "val", "step": 2500, "val_nll": -6.450486796058327, "best": -6.450486796058327, "bad": 0, "mem_mb": 91.215104, "elapsed_s": 540.4}131{"event": "train", "step": 2520, "loss": -5.959629058837891, "grad_norm": 0.7660537362098694, "lr": 0.00011906042145037817, "ex_per_s": 297.4976002601066, "data_frac": 0.0254307606195846, "mem_mb": 91.215104, "nll": -5.959629058837891, "elapsed_s": 542.2}132{"event": "train", "step": 2540, "loss": -6.031249046325684, "grad_norm": 0.45644015073776245, "lr": 0.00011696783066546026, "ex_per_s": 298.93536971700496, "data_frac": 0.025372536860007813, "mem_mb": 91.215104, "nll": -6.031249046325684, "elapsed_s": 543.8}133{"event": "train", "step": 2560, "loss": -5.948279857635498, "grad_norm": 0.2101321816444397, "lr": 0.00011488837139604017, "ex_per_s": 300.36339906012176, "data_frac": 0.02531431082534162, "mem_mb": 91.215104, "nll": -5.948279857635498, "elapsed_s": 545.5}134{"event": "train", "step": 2580, "loss": -5.9629316329956055, "grad_norm": 0.2286670207977295, "lr": 0.00011282261214551898, "ex_per_s": 301.78168473638397, "data_frac": 0.025256936525988528, "mem_mb": 91.215104, "nll": -5.9629316329956055, "elapsed_s": 547.2}135{"event": "train", "step": 2600, "loss": -5.850475788116455, "grad_norm": 0.34373316168785095, "lr": 0.00011077111767184915, "ex_per_s": 303.192630249835, "data_frac": 0.025200059809644956, "mem_mb": 91.215104, "nll": -5.850475788116455, "elapsed_s": 548.9}136{"event": "train", "step": 2620, "loss": -5.891404151916504, "grad_norm": 0.9027217626571655, "lr": 0.00010873444883313583, "ex_per_s": 304.5929143853556, "data_frac": 0.025145536663195054, "mem_mb": 91.215104, "nll": -5.891404151916504, "elapsed_s": 550.5}137{"event": "train", "step": 2640, "loss": -5.988446235656738, "grad_norm": 0.4525202214717865, "lr": 0.00010671316243430345, "ex_per_s": 305.987152558296, "data_frac": 0.02508955078458143, "mem_mb": 91.215104, "nll": -5.988446235656738, "elapsed_s": 552.2}138{"event": "train", "step": 2660, "loss": -5.823091983795166, "grad_norm": 0.09269967675209045, "lr": 0.00010470781107487112, "ex_per_s": 307.37288606966416, "data_frac": 0.02503375294070117, "mem_mb": 91.215104, "nll": -5.823091983795166, "elapsed_s": 553.9}139{"event": "train", "step": 2680, "loss": -5.9598212242126465, "grad_norm": 0.3574829399585724, "lr": 0.00010271894299787715, "ex_per_s": 308.7530831284515, "data_frac": 0.024978244128501345, "mem_mb": 91.215104, "nll": -5.9598212242126465, "elapsed_s": 555.6}140{"event": "train", "step": 2700, "loss": -6.132906913757324, "grad_norm": 0.7921686768531799, "lr": 0.00010074710193999506, "ex_per_s": 310.12413987973315, "data_frac": 0.02492307787694202, "mem_mb": 91.215104, "nll": -6.132906913757324, "elapsed_s": 557.2}141{"event": "train", "step": 2720, "loss": -5.882750988006592, "grad_norm": 0.5461236834526062, "lr": 9.879282698288144e-05, "ex_per_s": 311.48731895286176, "data_frac": 0.024868351440792396, "mem_mb": 91.215104, "nll": -5.882750988006592, "elapsed_s": 558.9}142{"event": "train", "step": 2740, "loss": -6.1264543533325195, "grad_norm": 0.09225895255804062, "lr": 9.685665240579613e-05, "ex_per_s": 312.8424725392287, "data_frac": 0.024813485025479195, "mem_mb": 91.215104, "nll": -6.1264543533325195, "elapsed_s": 560.6}143{"event": "train", "step": 2760, "loss": -5.960803985595703, "grad_norm": 0.2384318858385086, "lr": 9.4939107539536e-05, "ex_per_s": 314.189969021018, "data_frac": 0.024758822442630557, "mem_mb": 91.215104, "nll": -5.960803985595703, "elapsed_s": 562.3}144{"event": "train", "step": 2780, "loss": -6.012152671813965, "grad_norm": 0.8248691558837891, "lr": 9.304071662172087e-05, "ex_per_s": 315.52986684317835, "data_frac": 0.02470509073639497, "mem_mb": 91.215104, "nll": -6.012152671813965, "elapsed_s": 563.9}145{"event": "train", "step": 2800, "loss": -5.980536460876465, "grad_norm": 0.8521714210510254, "lr": 9.116199865347241e-05, "ex_per_s": 316.8619227667329, "data_frac": 0.02465163806374485, "mem_mb": 91.215104, "nll": -5.980536460876465, "elapsed_s": 565.6}146{"event": "train", "step": 2820, "loss": -6.008166790008545, "grad_norm": 0.34530705213546753, "lr": 8.930346725752419e-05, "ex_per_s": 318.1859266788397, "data_frac": 0.024598888759669167, "mem_mb": 91.215104, "nll": -6.008166790008545, "elapsed_s": 567.3}147{"event": "train", "step": 2840, "loss": -5.879316329956055, "grad_norm": 0.11155584454536438, "lr": 8.746563053780262e-05, "ex_per_s": 319.5021429941837, "data_frac": 0.024545930211744078, "mem_mb": 91.215104, "nll": -5.879316329956055, "elapsed_s": 568.9}148{"event": "train", "step": 2860, "loss": -6.049804210662842, "grad_norm": 0.19793064892292023, "lr": 8.564899094051614e-05, "ex_per_s": 320.80714958621115, "data_frac": 0.024495030068158428, "mem_mb": 91.215104, "nll": -6.049804210662842, "elapsed_s": 570.6}149{"event": "train", "step": 2880, "loss": -5.772400856018066, "grad_norm": 0.3590930700302124, "lr": 8.385404511679161e-05, "ex_per_s": 322.10695588122303, "data_frac": 0.0244427760344617, "mem_mb": 91.215104, "nll": -5.772400856018066, "elapsed_s": 572.3}150{"event": "train", "step": 2900, "loss": -5.795255184173584, "grad_norm": 0.5785411596298218, "lr": 8.208128378689484e-05, "ex_per_s": 323.3993948765856, "data_frac": 0.024390755746686935, "mem_mb": 91.215104, "nll": -5.795255184173584, "elapsed_s": 573.9}151{"event": "train", "step": 2920, "loss": -5.970411777496338, "grad_norm": 0.7491505742073059, "lr": 8.03311916060725e-05, "ex_per_s": 324.68450251314357, "data_frac": 0.024338974299415894, "mem_mb": 91.215104, "nll": -5.970411777496338, "elapsed_s": 575.6}152{"event": "train", "step": 2940, "loss": -5.9031662940979, "grad_norm": 0.40215733647346497, "lr": 7.860424703205225e-05, "ex_per_s": 325.96120915673475, "data_frac": 0.024287651407732546, "mem_mb": 91.215104, "nll": -5.9031662940979, "elapsed_s": 577.3}153{"event": "train", "step": 2960, "loss": -5.939562797546387, "grad_norm": 0.47129929065704346, "lr": 7.690092219423722e-05, "ex_per_s": 327.2316344756275, "data_frac": 0.024237248160125378, "mem_mb": 91.215104, "nll": -5.939562797546387, "elapsed_s": 579.0}154{"event": "train", "step": 2980, "loss": -5.894694805145264, "grad_norm": 0.4889563024044037, "lr": 7.522168276463048e-05, "ex_per_s": 328.4943117697068, "data_frac": 0.02418664325620632, "mem_mb": 91.215104, "nll": -5.894694805145264, "elapsed_s": 580.6}155{"event": "train", "step": 3000, "loss": -5.995248317718506, "grad_norm": 0.11850406974554062, "lr": 7.356698783052496e-05, "ex_per_s": 329.75098053113027, "data_frac": 0.024136612910767238, "mem_mb": 91.215104, "nll": -5.995248317718506, "elapsed_s": 582.3}156{"event": "val", "step": 3000, "val_nll": -6.4488843529272835, "best": -6.450486796058327, "bad": 1, "mem_mb": 91.215104, "elapsed_s": 645.4}157{"event": "train", "step": 3020, "loss": -5.961102485656738, "grad_norm": 0.9697180390357971, "lr": 7.193728976899362e-05, "ex_per_s": 298.6779697027662, "data_frac": 0.021734776931900127, "mem_mb": 91.215104, "nll": -5.961102485656738, "elapsed_s": 647.2}158{"event": "train", "step": 3040, "loss": -5.860126972198486, "grad_norm": 0.7739183306694031, "lr": 7.033303412321404e-05, "ex_per_s": 299.87999051315876, "data_frac": 0.021695701440778495, "mem_mb": 91.215104, "nll": -5.860126972198486, "elapsed_s": 648.8}159{"event": "train", "step": 3060, "loss": -6.07365608215332, "grad_norm": 0.22459697723388672, "lr": 6.875465948066139e-05, "ex_per_s": 301.0745774516306, "data_frac": 0.021657904610964205, "mem_mb": 91.215104, "nll": -6.07365608215332, "elapsed_s": 650.5}160{"event": "train", "step": 3080, "loss": -5.971156597137451, "grad_norm": 0.2524579167366028, "lr": 6.720259735320299e-05, "ex_per_s": 302.2636679051358, "data_frac": 0.021619831963361577, "mem_mb": 91.215104, "nll": -5.971156597137451, "elapsed_s": 652.2}161{"event": "train", "step": 3100, "loss": -6.016089916229248, "grad_norm": 0.32837510108947754, "lr": 6.567727205912722e-05, "ex_per_s": 303.4463427762516, "data_frac": 0.02158143906756419, "mem_mb": 91.215104, "nll": -6.016089916229248, "elapsed_s": 653.9}162{"event": "train", "step": 3120, "loss": -6.091784477233887, "grad_norm": 0.7085905075073242, "lr": 6.417910060713926e-05, "ex_per_s": 304.6237035709868, "data_frac": 0.021543093796286973, "mem_mb": 91.215104, "nll": -6.091784477233887, "elapsed_s": 655.5}163{"event": "train", "step": 3140, "loss": -6.010141372680664, "grad_norm": 0.4651021659374237, "lr": 6.270849258235485e-05, "ex_per_s": 305.7952599198208, "data_frac": 0.021504623696723027, "mem_mb": 91.215104, "nll": -6.010141372680664, "elapsed_s": 657.2}164{"event": "train", "step": 3160, "loss": -6.021839618682861, "grad_norm": 0.12677766382694244, "lr": 6.126585003432406e-05, "ex_per_s": 306.96098924459017, "data_frac": 0.021466724315405614, "mem_mb": 91.215104, "nll": -6.021839618682861, "elapsed_s": 658.9}165{"event": "train", "step": 3180, "loss": -6.095400333404541, "grad_norm": 0.19351884722709656, "lr": 5.985156736711483e-05, "ex_per_s": 308.1195146144445, "data_frac": 0.021428914399601963, "mem_mb": 91.215104, "nll": -6.095400333404541, "elapsed_s": 660.6}166{"event": "train", "step": 3200, "loss": -6.002824783325195, "grad_norm": 0.3955107033252716, "lr": 5.846603123148688e-05, "ex_per_s": 309.2740642753095, "data_frac": 0.021391678524486, "mem_mb": 91.215104, "nll": -6.002824783325195, "elapsed_s": 662.2}167{"event": "train", "step": 3220, "loss": -5.916471481323242, "grad_norm": 0.15857668220996857, "lr": 5.710962041918516e-05, "ex_per_s": 310.4226401148819, "data_frac": 0.021354152062033546, "mem_mb": 91.215104, "nll": -5.916471481323242, "elapsed_s": 663.9}168{"event": "train", "step": 3240, "loss": -5.8036065101623535, "grad_norm": 0.37809523940086365, "lr": 5.578270575938211e-05, "ex_per_s": 311.5662277107046, "data_frac": 0.021317056216476518, "mem_mb": 91.215104, "nll": -5.8036065101623535, "elapsed_s": 665.6}169{"event": "train", "step": 3260, "loss": -5.962217330932617, "grad_norm": 0.43212854862213135, "lr": 5.448565001729653e-05, "ex_per_s": 312.70371679820533, "data_frac": 0.021279952154386265, "mem_mb": 91.215104, "nll": -5.962217330932617, "elapsed_s": 667.3}170{"event": "train", "step": 3280, "loss": -6.014777660369873, "grad_norm": 0.36320826411247253, "lr": 5.3218807795017296e-05, "ex_per_s": 313.83475680425624, "data_frac": 0.021243729566516427, "mem_mb": 91.215104, "nll": -6.014777660369873, "elapsed_s": 668.9}171{"event": "train", "step": 3300, "loss": -5.80853796005249, "grad_norm": 0.1892099678516388, "lr": 5.198252543455864e-05, "ex_per_s": 314.96080464437773, "data_frac": 0.021208129433797906, "mem_mb": 91.215104, "nll": -5.80853796005249, "elapsed_s": 670.6}172{"event": "train", "step": 3320, "loss": -5.98966646194458, "grad_norm": 0.3030041456222534, "lr": 5.07771409231737e-05, "ex_per_s": 316.08237553835113, "data_frac": 0.021172845402160398, "mem_mb": 91.215104, "nll": -5.98966646194458, "elapsed_s": 672.3}173{"event": "train", "step": 3340, "loss": -5.932128429412842, "grad_norm": 0.31679683923721313, "lr": 4.960298380095221e-05, "ex_per_s": 317.1985620845264, "data_frac": 0.021136143851589677, "mem_mb": 91.215104, "nll": -5.932128429412842, "elapsed_s": 673.9}174{"event": "train", "step": 3360, "loss": -5.90730094909668, "grad_norm": 0.3605494797229767, "lr": 4.846037507072751e-05, "ex_per_s": 318.31152034637654, "data_frac": 0.021100658145939113, "mem_mb": 91.215104, "nll": -5.90730094909668, "elapsed_s": 675.6}175{"event": "train", "step": 3380, "loss": -6.123069763183594, "grad_norm": 0.5651793479919434, "lr": 4.7349627110317556e-05, "ex_per_s": 319.41892973556264, "data_frac": 0.021064623698726866, "mem_mb": 91.215104, "nll": -6.123069763183594, "elapsed_s": 677.3}176{"event": "train", "step": 3400, "loss": -5.884279251098633, "grad_norm": 0.33114516735076904, "lr": 4.6271043587123986e-05, "ex_per_s": 320.5205219795414, "data_frac": 0.021028644540896133, "mem_mb": 91.215104, "nll": -5.884279251098633, "elapsed_s": 678.9}177{"event": "train", "step": 3420, "loss": -5.837557315826416, "grad_norm": 0.23839141428470612, "lr": 4.52249193751124e-05, "ex_per_s": 321.6165770669124, "data_frac": 0.02099332924326728, "mem_mb": 91.215104, "nll": -5.837557315826416, "elapsed_s": 680.6}178{"event": "train", "step": 3440, "loss": -5.91670560836792, "grad_norm": 0.13230085372924805, "lr": 4.4211540474196926e-05, "ex_per_s": 322.70731364067177, "data_frac": 0.020958424992943894, "mem_mb": 91.215104, "nll": -5.91670560836792, "elapsed_s": 682.3}179{"event": "train", "step": 3460, "loss": -6.058538913726807, "grad_norm": 0.22063493728637695, "lr": 4.3231183932050546e-05, "ex_per_s": 323.7918325599667, "data_frac": 0.020923567626522428, "mem_mb": 91.215104, "nll": -6.058538913726807, "elapsed_s": 683.9}180{"event": "train", "step": 3480, "loss": -5.916459560394287, "grad_norm": 0.07499156892299652, "lr": 4.228411776836306e-05, "ex_per_s": 324.87022710115366, "data_frac": 0.02088907096385751, "mem_mb": 91.215104, "nll": -5.916459560394287, "elapsed_s": 685.6}181{"event": "train", "step": 3500, "loss": -6.03853702545166, "grad_norm": 0.12891988456249237, "lr": 4.137060090156723e-05, "ex_per_s": 325.9451297634804, "data_frac": 0.020855181678175693, "mem_mb": 91.215104, "nll": -6.03853702545166, "elapsed_s": 687.3}182{"event": "val", "step": 3500, "val_nll": -6.450911094056047, "best": -6.450911094056047, "bad": 0, "mem_mb": 91.215104, "elapsed_s": 750.4}183{"event": "train", "step": 3520, "loss": -5.939018726348877, "grad_norm": 0.45455610752105713, "lr": 4.049088307805306e-05, "ex_per_s": 299.55226262386924, "data_frac": 0.01907388780827778, "mem_mb": 91.215104, "nll": -5.939018726348877, "elapsed_s": 752.1}184{"event": "train", "step": 3540, "loss": -6.136294364929199, "grad_norm": 0.6263049840927124, "lr": 3.964520480388956e-05, "ex_per_s": 300.5860373395576, "data_frac": 0.01904607315312516, "mem_mb": 91.215104, "nll": -6.136294364929199, "elapsed_s": 753.8}185{"event": "train", "step": 3560, "loss": -6.139619827270508, "grad_norm": 0.08266972750425339, "lr": 3.88337972790728e-05, "ex_per_s": 301.61529271458585, "data_frac": 0.019018700178495624, "mem_mb": 91.215104, "nll": -6.139619827270508, "elapsed_s": 755.4}186{"event": "train", "step": 3580, "loss": -5.903184413909912, "grad_norm": 0.2471899390220642, "lr": 3.805688233431824e-05, "ex_per_s": 302.6393163408922, "data_frac": 0.018991241270859948, "mem_mb": 91.215104, "nll": -5.903184413909912, "elapsed_s": 757.1}187{"event": "train", "step": 3600, "loss": -6.098885536193848, "grad_norm": 0.4571618139743805, "lr": 3.7314672370414325e-05, "ex_per_s": 303.6581351214057, "data_frac": 0.0189639522892677, "mem_mb": 91.215104, "nll": -6.098885536193848, "elapsed_s": 758.8}188{"event": "train", "step": 3620, "loss": -5.910143852233887, "grad_norm": 0.9950910210609436, "lr": 3.660737030015427e-05, "ex_per_s": 304.6744099962628, "data_frac": 0.018936864605800675, "mem_mb": 91.215104, "nll": -5.910143852233887, "elapsed_s": 760.5}189{"event": "train", "step": 3640, "loss": -6.053521633148193, "grad_norm": 0.510341227054596, "lr": 3.5935169492861716e-05, "ex_per_s": 305.68571424887864, "data_frac": 0.018909815891590954, "mem_mb": 91.215104, "nll": -6.053521633148193, "elapsed_s": 762.1}190{"event": "train", "step": 3660, "loss": -5.83250617980957, "grad_norm": 0.4992491900920868, "lr": 3.52982537215255e-05, "ex_per_s": 306.6934562431782, "data_frac": 0.01888294411232872, "mem_mb": 91.215104, "nll": -5.83250617980957, "elapsed_s": 763.8}191{"event": "train", "step": 3680, "loss": -5.992978096008301, "grad_norm": 0.10519673675298691, "lr": 3.469679711255794e-05, "ex_per_s": 307.69587850739947, "data_frac": 0.018856076871018958, "mem_mb": 91.215104, "nll": -5.992978096008301, "elapsed_s": 765.5}192{"event": "train", "step": 3700, "loss": -6.066473960876465, "grad_norm": 0.8506190180778503, "lr": 3.41309640981904e-05, "ex_per_s": 308.6939851430093, "data_frac": 0.018829485711798287, "mem_mb": 91.215104, "nll": -6.066473960876465, "elapsed_s": 767.1}193{"event": "train", "step": 3720, "loss": -5.948521614074707, "grad_norm": 0.5990975499153137, "lr": 3.3600909371519225e-05, "ex_per_s": 309.68816661392316, "data_frac": 0.01880316349549808, "mem_mb": 91.215104, "nll": -5.948521614074707, "elapsed_s": 768.8}194{"event": "train", "step": 3740, "loss": -6.191071033477783, "grad_norm": 0.20834407210350037, "lr": 3.31067778442141e-05, "ex_per_s": 310.6793576199438, "data_frac": 0.018776550039661673, "mem_mb": 91.215104, "nll": -6.191071033477783, "elapsed_s": 770.5}195{"event": "train", "step": 3760, "loss": -6.147716045379639, "grad_norm": 0.2828250229358673, "lr": 3.2648704606900735e-05, "ex_per_s": 311.66631459084977, "data_frac": 0.018751733533077557, "mem_mb": 91.215104, "nll": -6.147716045379639, "elapsed_s": 772.2}196{"event": "train", "step": 3780, "loss": -5.963770866394043, "grad_norm": 0.368939608335495, "lr": 3.222681489222838e-05, "ex_per_s": 312.6522659939616, "data_frac": 0.01872564457065758, "mem_mb": 91.215104, "nll": -5.963770866394043, "elapsed_s": 773.8}197{"event": "train", "step": 3800, "loss": -6.099841117858887, "grad_norm": 0.09215747565031052, "lr": 3.1841224040632484e-05, "ex_per_s": 313.6343278758557, "data_frac": 0.01869961434708237, "mem_mb": 91.215104, "nll": -6.099841117858887, "elapsed_s": 775.5}198{"event": "train", "step": 3820, "loss": -5.97218132019043, "grad_norm": 0.1657426804304123, "lr": 3.149203746880175e-05, "ex_per_s": 314.6123807352985, "data_frac": 0.01867389444034316, "mem_mb": 91.215104, "nll": -5.97218132019043, "elapsed_s": 777.1}199{"event": "train", "step": 3840, "loss": -5.874682426452637, "grad_norm": 0.6334619522094727, "lr": 3.117935064085835e-05, "ex_per_s": 315.58567921020443, "data_frac": 0.018648310722421846, "mem_mb": 91.215104, "nll": -5.874682426452637, "elapsed_s": 778.8}200{"event": "train", "step": 3860, "loss": -6.013962745666504, "grad_norm": 0.39144638180732727, "lr": 3.090324904225885e-05, "ex_per_s": 316.5533231574144, "data_frac": 0.018622743586942977, "mem_mb": 91.215104, "nll": -6.013962745666504, "elapsed_s": 780.4}201{"event": "train", "step": 3880, "loss": -5.892791271209717, "grad_norm": 0.07039698958396912, "lr": 3.0663808156423456e-05, "ex_per_s": 317.5189755179148, "data_frac": 0.018597516894859743, "mem_mb": 91.215104, "nll": -5.892791271209717, "elapsed_s": 782.1}202{"event": "train", "step": 3900, "loss": -6.1823601722717285, "grad_norm": 0.6564260125160217, "lr": 3.046109344409957e-05, "ex_per_s": 318.4803193473154, "data_frac": 0.018572226335453978, "mem_mb": 91.215104, "nll": -6.1823601722717285, "elapsed_s": 783.8}203{"event": "train", "step": 3920, "loss": -6.038522720336914, "grad_norm": 0.19618947803974152, "lr": 3.0295160325465435e-05, "ex_per_s": 319.43753410811496, "data_frac": 0.018547012157529186, "mem_mb": 91.215104, "nll": -6.038522720336914, "elapsed_s": 785.4}204{"event": "train", "step": 3940, "loss": -5.923311710357666, "grad_norm": 0.13536901772022247, "lr": 3.016605416497886e-05, "ex_per_s": 320.38991367588716, "data_frac": 0.0185222896306796, "mem_mb": 91.215104, "nll": -5.923311710357666, "elapsed_s": 787.1}205{"event": "train", "step": 3960, "loss": -6.070324897766113, "grad_norm": 0.1668504923582077, "lr": 3.0073810258974985e-05, "ex_per_s": 321.3373476087486, "data_frac": 0.018498199274834936, "mem_mb": 91.215104, "nll": -6.070324897766113, "elapsed_s": 788.7}206{"event": "train", "step": 3980, "loss": -5.8040771484375, "grad_norm": 0.16629590094089508, "lr": 3.0018453826016683e-05, "ex_per_s": 322.2816866578597, "data_frac": 0.01847367920255224, "mem_mb": 91.215104, "nll": -5.8040771484375, "elapsed_s": 790.4}207{"event": "train", "step": 4000, "loss": -6.05793571472168, "grad_norm": 0.7715352773666382, "lr": 2.9999999999999997e-05, "ex_per_s": 323.2200138976062, "data_frac": 0.018449350135243223, "mem_mb": 91.215104, "nll": -6.05793571472168, "elapsed_s": 792.1}208{"event": "val", "step": 4000, "val_nll": -6.450383087696146, "best": -6.450911094056047, "bad": 1, "mem_mb": 91.215104, "elapsed_s": 855.2}209{"event": "done", "steps": 4000, "best_val": -6.450911094056047, "wall_s": 855.1630790233612, "ex_per_s": 299.3581064004362, "peak_mem_mb": 91.215104, "n_params": 3456584, "device": "mps", "data_time_frac": 0.017087316017982346, "elapsed_s": 855.2}210
Files · main
Menu

Theme