SPB Git

spb/wp41_uqo

Public
51.5 KB · 210 lines
1{"event": "train", "step": 20, "loss": -6.120003700256348, "grad_norm": 0.3687268793582916, "lr": 3.149999999999999e-05, "ex_per_s": 91.32313348297372, "data_frac": 0.867917463909163, "mem_mb": 91.210752, "nll": -6.120003700256348, "elapsed_s": 14.1}2{"event": "train", "step": 40, "loss": -5.888540744781494, "grad_norm": 0.97832852602005, "lr": 6.149999999999999e-05, "ex_per_s": 162.11890409594662, "data_frac": 0.7757495812393413, "mem_mb": 91.210752, "nll": -5.888540744781494, "elapsed_s": 15.8}3{"event": "train", "step": 60, "loss": -5.980739593505859, "grad_norm": 0.21500058472156525, "lr": 9.149999999999999e-05, "ex_per_s": 219.14567470393558, "data_frac": 0.7003999166039334, "mem_mb": 91.210752, "nll": -5.980739593505859, "elapsed_s": 17.6}4{"event": "train", "step": 80, "loss": -6.080530166625977, "grad_norm": 0.3451729416847229, "lr": 0.0001215, "ex_per_s": 266.53354260502545, "data_frac": 0.6394094668892633, "mem_mb": 91.210752, "nll": -6.080530166625977, "elapsed_s": 19.3}5{"event": "train", "step": 100, "loss": -5.950662136077881, "grad_norm": 0.1547926813364029, "lr": 0.0001515, "ex_per_s": 306.5350113581115, "data_frac": 0.5888320818368696, "mem_mb": 91.210752, "nll": -5.950662136077881, "elapsed_s": 20.9}6{"event": "train", "step": 120, "loss": -5.911513805389404, "grad_norm": 0.7984780073165894, "lr": 0.00018149999999999997, "ex_per_s": 340.609985237108, "data_frac": 0.5457650708892892, "mem_mb": 91.210752, "nll": -5.911513805389404, "elapsed_s": 22.6}7{"event": "train", "step": 140, "loss": -5.831068992614746, "grad_norm": 0.7362173199653625, "lr": 0.00021149999999999996, "ex_per_s": 369.9868373519128, "data_frac": 0.5086098592406509, "mem_mb": 91.210752, "nll": -5.831068992614746, "elapsed_s": 24.3}8{"event": "train", "step": 160, "loss": -5.737489700317383, "grad_norm": 0.5463359355926514, "lr": 0.0002415, "ex_per_s": 395.6325423554575, "data_frac": 0.47651525308631426, "mem_mb": 91.210752, "nll": -5.737489700317383, "elapsed_s": 25.9}9{"event": "train", "step": 180, "loss": -5.975886344909668, "grad_norm": 1.009723424911499, "lr": 0.0002715, "ex_per_s": 418.19801908232563, "data_frac": 0.44813640802142773, "mem_mb": 91.210752, "nll": -5.975886344909668, "elapsed_s": 27.6}10{"event": "train", "step": 200, "loss": -5.971966743469238, "grad_norm": 0.4755775034427643, "lr": 0.0003, "ex_per_s": 438.13912549873976, "data_frac": 0.4229731964293505, "mem_mb": 91.210752, "nll": -5.971966743469238, "elapsed_s": 29.3}11{"event": "train", "step": 220, "loss": -6.031113624572754, "grad_norm": 1.1688822507858276, "lr": 0.00029998154617398326, "ex_per_s": 456.04909883702993, "data_frac": 0.4006211512280868, "mem_mb": 91.210752, "nll": -6.031113624572754, "elapsed_s": 30.9}12{"event": "train", "step": 240, "loss": -5.915210723876953, "grad_norm": 0.21985410153865814, "lr": 0.000299926189741025, "ex_per_s": 472.04129892012617, "data_frac": 0.38045665121072864, "mem_mb": 91.210752, "nll": -5.915210723876953, "elapsed_s": 32.6}13{"event": "train", "step": 260, "loss": -5.888294696807861, "grad_norm": 0.13914845883846283, "lr": 0.0002998339458350211, "ex_per_s": 486.5039343415738, "data_frac": 0.36228883668973616, "mem_mb": 91.210752, "nll": -5.888294696807861, "elapsed_s": 34.2}14{"event": "train", "step": 280, "loss": -5.94511604309082, "grad_norm": 0.1131725013256073, "lr": 0.0002997048396745345, "ex_per_s": 499.67930984223597, "data_frac": 0.3458393244338804, "mem_mb": 91.210752, "nll": -5.94511604309082, "elapsed_s": 35.9}15{"event": "train", "step": 300, "loss": -5.9681077003479, "grad_norm": 0.24137750267982483, "lr": 0.0002995389065559004, "ex_per_s": 511.68244317367584, "data_frac": 0.33083811803147933, "mem_mb": 91.210752, "nll": -5.9681077003479, "elapsed_s": 37.6}16{"event": "train", "step": 320, "loss": -5.967286586761475, "grad_norm": 0.3391331434249878, "lr": 0.0002993361918435765, "ex_per_s": 522.6397039799718, "data_frac": 0.31709210886708084, "mem_mb": 91.210752, "nll": -5.967286586761475, "elapsed_s": 39.2}17{"event": "train", "step": 340, "loss": -5.834740161895752, "grad_norm": 0.31500959396362305, "lr": 0.00029909675095774114, "ex_per_s": 532.7004534202184, "data_frac": 0.30446007137077474, "mem_mb": 91.210752, "nll": -5.834740161895752, "elapsed_s": 40.9}18{"event": "train", "step": 360, "loss": -5.996451377868652, "grad_norm": 0.24469149112701416, "lr": 0.00029882064935914164, "ex_per_s": 541.9679246204156, "data_frac": 0.29282669039126946, "mem_mb": 91.210752, "nll": -5.996451377868652, "elapsed_s": 42.6}19{"event": "train", "step": 380, "loss": -6.005426406860352, "grad_norm": 1.188538908958435, "lr": 0.00029850796253119824, "ex_per_s": 550.5606578125754, "data_frac": 0.2820700473015796, "mem_mb": 91.210752, "nll": -6.005426406860352, "elapsed_s": 44.2}20{"event": "train", "step": 400, "loss": -5.994283676147461, "grad_norm": 1.4908181428909302, "lr": 0.0002981587759593675, "ex_per_s": 558.51171917354, "data_frac": 0.2720863808214432, "mem_mb": 91.210752, "nll": -5.994283676147461, "elapsed_s": 45.9}21{"event": "train", "step": 420, "loss": -5.964623928070068, "grad_norm": 0.5802336931228638, "lr": 0.0002977731851077716, "ex_per_s": 565.9117400906019, "data_frac": 0.26280431319095593, "mem_mb": 91.210752, "nll": -5.964623928070068, "elapsed_s": 47.5}22{"event": "train", "step": 440, "loss": -5.938300132751465, "grad_norm": 0.7682151794433594, "lr": 0.0002973512953930992, "ex_per_s": 572.4730000677642, "data_frac": 0.2540220457426144, "mem_mb": 91.210752, "nll": -5.938300132751465, "elapsed_s": 49.2}23{"event": "train", "step": 460, "loss": -6.101249694824219, "grad_norm": 0.5825731158256531, "lr": 0.0002968932221557859, "ex_per_s": 578.9160519867748, "data_frac": 0.24594171277960403, "mem_mb": 91.210752, "nll": -6.101249694824219, "elapsed_s": 50.9}24{"event": "train", "step": 480, "loss": -5.877638339996338, "grad_norm": 0.27511727809906006, "lr": 0.0002963990906284808, "ex_per_s": 584.9294013436464, "data_frac": 0.238361848907958, "mem_mb": 91.210752, "nll": -5.877638339996338, "elapsed_s": 52.6}25{"event": "train", "step": 500, "loss": -5.925490379333496, "grad_norm": 0.5436809659004211, "lr": 0.00029586903590180956, "ex_per_s": 590.5987998751792, "data_frac": 0.23125831338426198, "mem_mb": 91.210752, "nll": -5.925490379333496, "elapsed_s": 54.2}26{"event": "val", "step": 500, "val_nll": -6.439822963520271, "best": -6.439822963520271, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 118.1}27{"event": "train", "step": 520, "loss": -6.040792465209961, "grad_norm": 0.2529070973396301, "lr": 0.00029530320288744205, "ex_per_s": 277.74218425403296, "data_frac": 0.10468289931130995, "mem_mb": 91.210752, "nll": -6.040792465209961, "elapsed_s": 119.9}28{"event": "train", "step": 540, "loss": -5.878663063049316, "grad_norm": 0.5834041833877563, "lr": 0.0002947017462784745, "ex_per_s": 284.44010702362436, "data_frac": 0.10334230534224881, "mem_mb": 91.210752, "nll": -5.878663063049316, "elapsed_s": 121.5}29{"event": "train", "step": 560, "loss": -5.972829341888428, "grad_norm": 0.08392803370952606, "lr": 0.00029406483050713824, "ex_per_s": 290.94418900686674, "data_frac": 0.10202272520865865, "mem_mb": 91.210752, "nll": -5.972829341888428, "elapsed_s": 123.2}30{"event": "train", "step": 580, "loss": -5.924657344818115, "grad_norm": 0.09499597549438477, "lr": 0.0002933926296998457, "ex_per_s": 297.28359688766903, "data_frac": 0.10074199075636428, "mem_mb": 91.210752, "nll": -5.924657344818115, "elapsed_s": 124.9}31{"event": "train", "step": 600, "loss": -5.972743511199951, "grad_norm": 0.715925395488739, "lr": 0.0002926853276295856, "ex_per_s": 303.4532945474882, "data_frac": 0.09949558909838284, "mem_mb": 91.210752, "nll": -5.972743511199951, "elapsed_s": 126.6}32{"event": "train", "step": 620, "loss": -5.952538967132568, "grad_norm": 0.3831275403499603, "lr": 0.00029194311766568174, "ex_per_s": 309.4746238080621, "data_frac": 0.09828846271583953, "mem_mb": 91.210752, "nll": -5.952538967132568, "elapsed_s": 128.3}33{"event": "train", "step": 640, "loss": -6.065334796905518, "grad_norm": 0.47480061650276184, "lr": 0.0002911662027209272, "ex_per_s": 315.33573676181635, "data_frac": 0.09711559122335403, "mem_mb": 91.210752, "nll": -6.065334796905518, "elapsed_s": 129.9}34{"event": "train", "step": 660, "loss": -6.008012294769287, "grad_norm": 1.0299180746078491, "lr": 0.00029035479519611045, "ex_per_s": 321.05868887975726, "data_frac": 0.09596988743175099, "mem_mb": 91.210752, "nll": -6.008012294769287, "elapsed_s": 131.6}35{"event": "train", "step": 680, "loss": -6.006779193878174, "grad_norm": 0.11547449976205826, "lr": 0.0002895091169219469, "ex_per_s": 326.6388585747468, "data_frac": 0.09485498083392145, "mem_mb": 91.210752, "nll": -6.006779193878174, "elapsed_s": 133.3}36{"event": "train", "step": 700, "loss": -5.987523555755615, "grad_norm": 0.6487700343132019, "lr": 0.00028862939909843273, "ex_per_s": 332.0793682913161, "data_frac": 0.09376235391047448, "mem_mb": 91.210752, "nll": -5.987523555755615, "elapsed_s": 135.0}37{"event": "train", "step": 720, "loss": -6.063596725463867, "grad_norm": 0.21977651119232178, "lr": 0.0002877158822316369, "ex_per_s": 337.3952292246339, "data_frac": 0.09269930273425989, "mem_mb": 91.210752, "nll": -6.063596725463867, "elapsed_s": 136.6}38{"event": "train", "step": 740, "loss": -6.033612251281738, "grad_norm": 0.2448177933692932, "lr": 0.00028676881606794945, "ex_per_s": 342.5796821811089, "data_frac": 0.09166243051586517, "mem_mb": 91.210752, "nll": -6.033612251281738, "elapsed_s": 138.3}39{"event": "train", "step": 760, "loss": -5.959536075592041, "grad_norm": 0.7799402475357056, "lr": 0.000285788459525803, "ex_per_s": 347.63779494981503, "data_frac": 0.09064963024331248, "mem_mb": 91.210752, "nll": -5.959536075592041, "elapsed_s": 140.0}40{"event": "train", "step": 780, "loss": -5.894865036010742, "grad_norm": 0.28068578243255615, "lr": 0.00028477508062488754, "ex_per_s": 352.58148627809214, "data_frac": 0.08966211448328235, "mem_mb": 91.210752, "nll": -5.894865036010742, "elapsed_s": 141.6}41{"event": "train", "step": 800, "loss": -5.8154826164245605, "grad_norm": 0.27545416355133057, "lr": 0.000283728956412876, "ex_per_s": 357.4073581007506, "data_frac": 0.08869983666317756, "mem_mb": 91.210752, "nll": -5.8154826164245605, "elapsed_s": 143.3}42{"event": "train", "step": 820, "loss": -5.933736801147461, "grad_norm": 0.2651994228363037, "lr": 0.00028265037288968236, "ex_per_s": 362.13288483273607, "data_frac": 0.08775739349881032, "mem_mb": 91.210752, "nll": -5.933736801147461, "elapsed_s": 145.0}43{"event": "train", "step": 840, "loss": -5.806473731994629, "grad_norm": 0.5664976239204407, "lr": 0.0002815396249292725, "ex_per_s": 366.7448219176946, "data_frac": 0.08683921051857937, "mem_mb": 91.210752, "nll": -5.806473731994629, "elapsed_s": 146.6}44{"event": "train", "step": 860, "loss": -5.914696216583252, "grad_norm": 1.4353175163269043, "lr": 0.00028039701619904774, "ex_per_s": 371.2495503863039, "data_frac": 0.08594245219684155, "mem_mb": 91.210752, "nll": -5.914696216583252, "elapsed_s": 148.3}45{"event": "train", "step": 880, "loss": -5.868708610534668, "grad_norm": 0.6343095898628235, "lr": 0.00027922285907682625, "ex_per_s": 375.6558070908673, "data_frac": 0.08506852188228793, "mem_mb": 91.210752, "nll": -5.868708610534668, "elapsed_s": 150.0}46{"event": "train", "step": 900, "loss": -5.985992431640625, "grad_norm": 0.3600236475467682, "lr": 0.00027801747456544134, "ex_per_s": 379.9857974923583, "data_frac": 0.08421458649866387, "mem_mb": 91.210752, "nll": -5.985992431640625, "elapsed_s": 151.6}47{"event": "train", "step": 920, "loss": -6.024289608001709, "grad_norm": 0.11485781520605087, "lr": 0.00027678119220498265, "ex_per_s": 384.2220093175315, "data_frac": 0.08338030288850104, "mem_mb": 91.210752, "nll": -6.024289608001709, "elapsed_s": 153.3}48{"event": "train", "step": 940, "loss": -6.080376148223877, "grad_norm": 0.9066921472549438, "lr": 0.0002755143499827035, "ex_per_s": 388.3659011424366, "data_frac": 0.08256051678927358, "mem_mb": 91.210752, "nll": -6.080376148223877, "elapsed_s": 154.9}49{"event": "train", "step": 960, "loss": -5.96680212020874, "grad_norm": 0.4307335317134857, "lr": 0.00027421729424061787, "ex_per_s": 392.43943268755106, "data_frac": 0.08176242556157642, "mem_mb": 91.210752, "nll": -5.96680212020874, "elapsed_s": 156.6}50{"event": "train", "step": 980, "loss": -5.924592971801758, "grad_norm": 0.46913886070251465, "lr": 0.00027289037958081483, "ex_per_s": 396.42589745117164, "data_frac": 0.0809835549557866, "mem_mb": 91.210752, "nll": -5.924592971801758, "elapsed_s": 158.3}51{"event": "train", "step": 1000, "loss": -6.071223258972168, "grad_norm": 0.15925239026546478, "lr": 0.0002715339687685131, "ex_per_s": 400.33607263995805, "data_frac": 0.08022089971113795, "mem_mb": 91.210752, "nll": -6.071223258972168, "elapsed_s": 159.9}52{"event": "val", "step": 1000, "val_nll": -6.445527173841883, "best": -6.445527173841883, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 223.4}53{"event": "train", "step": 1020, "loss": -6.029529094696045, "grad_norm": 0.09336287528276443, "lr": 0.0002701484326328852, "ex_per_s": 289.9404122745717, "data_frac": 0.057011825130897184, "mem_mb": 91.210752, "nll": -6.029529094696045, "elapsed_s": 225.2}54{"event": "train", "step": 1040, "loss": -5.9705939292907715, "grad_norm": 0.21520906686782837, "lr": 0.00026873414996567596, "ex_per_s": 293.4252341612947, "data_frac": 0.056664168109131434, "mem_mb": 91.210752, "nll": -5.9705939292907715, "elapsed_s": 226.9}55{"event": "train", "step": 1060, "loss": -6.178966522216797, "grad_norm": 0.25049909949302673, "lr": 0.0002672915074176452, "ex_per_s": 296.88107565255945, "data_frac": 0.056298416594209665, "mem_mb": 91.210752, "nll": -6.178966522216797, "elapsed_s": 228.6}56{"event": "train", "step": 1080, "loss": -5.813770771026611, "grad_norm": 0.13968683779239655, "lr": 0.00026582089939286075, "ex_per_s": 300.2840972849226, "data_frac": 0.05594301589498006, "mem_mb": 91.210752, "nll": -5.813770771026611, "elapsed_s": 230.2}57{"event": "train", "step": 1100, "loss": -6.052749156951904, "grad_norm": 0.12900885939598083, "lr": 0.0002643227279408728, "ex_per_s": 303.6395205583628, "data_frac": 0.055589403480859015, "mem_mb": 91.210752, "nll": -6.052749156951904, "elapsed_s": 231.9}58{"event": "train", "step": 1120, "loss": -6.054229736328125, "grad_norm": 0.31064099073410034, "lr": 0.000262797402646797, "ex_per_s": 306.9469719386329, "data_frac": 0.05524209240053451, "mem_mb": 91.210752, "nll": -6.054229736328125, "elapsed_s": 233.6}59{"event": "train", "step": 1140, "loss": -6.019798278808594, "grad_norm": 0.1488422453403473, "lr": 0.0002612453405193386, "ex_per_s": 310.2064663696853, "data_frac": 0.05489823702367992, "mem_mb": 91.210752, "nll": -6.019798278808594, "elapsed_s": 235.2}60{"event": "train", "step": 1160, "loss": -5.966041564941406, "grad_norm": 0.778268575668335, "lr": 0.0002596669658767859, "ex_per_s": 313.4248281613623, "data_frac": 0.05455992085260633, "mem_mb": 91.210752, "nll": -5.966041564941406, "elapsed_s": 236.9}61{"event": "train", "step": 1180, "loss": -5.957314491271973, "grad_norm": 0.17309662699699402, "lr": 0.00025806271023100637, "ex_per_s": 316.59825789689035, "data_frac": 0.05422547295188577, "mem_mb": 91.210752, "nll": -5.957314491271973, "elapsed_s": 238.6}62{"event": "train", "step": 1200, "loss": -5.944145679473877, "grad_norm": 0.1845875084400177, "lr": 0.00025643301216947504, "ex_per_s": 319.7270066705095, "data_frac": 0.05389512316110087, "mem_mb": 91.210752, "nll": -5.944145679473877, "elapsed_s": 240.2}63{"event": "train", "step": 1220, "loss": -6.03354549407959, "grad_norm": 0.34642255306243896, "lr": 0.0002547783172353695, "ex_per_s": 322.8117208004143, "data_frac": 0.05356981749727167, "mem_mb": 91.210752, "nll": -6.03354549407959, "elapsed_s": 241.9}64{"event": "train", "step": 1240, "loss": -6.057979583740234, "grad_norm": 0.19741255044937134, "lr": 0.00025309907780576277, "ex_per_s": 325.85721662653543, "data_frac": 0.05324967486693659, "mem_mb": 91.210752, "nll": -6.057979583740234, "elapsed_s": 243.6}65{"event": "train", "step": 1260, "loss": -6.0540313720703125, "grad_norm": 0.2969401180744171, "lr": 0.0002513957529679477, "ex_per_s": 328.8571021764372, "data_frac": 0.05293396269159261, "mem_mb": 91.210752, "nll": -6.0540313720703125, "elapsed_s": 245.3}66{"event": "train", "step": 1280, "loss": -6.048571586608887, "grad_norm": 0.4108867645263672, "lr": 0.00024966880839392746, "ex_per_s": 331.8211884047714, "data_frac": 0.05262382156926423, "mem_mb": 91.210752, "nll": -6.048571586608887, "elapsed_s": 246.9}67{"event": "train", "step": 1300, "loss": -5.8655924797058105, "grad_norm": 0.3472103178501129, "lr": 0.0002479187162131052, "ex_per_s": 334.73981598222457, "data_frac": 0.05231568618124602, "mem_mb": 91.210752, "nll": -5.8655924797058105, "elapsed_s": 248.6}68{"event": "train", "step": 1320, "loss": -5.825611591339111, "grad_norm": 0.291996568441391, "lr": 0.00024614595488320833, "ex_per_s": 337.6186470491824, "data_frac": 0.052015000320977764, "mem_mb": 91.210752, "nll": -5.825611591339111, "elapsed_s": 250.3}69{"event": "train", "step": 1340, "loss": -5.95272159576416, "grad_norm": 0.6569538116455078, "lr": 0.00024435100905948387, "ex_per_s": 340.4620313922484, "data_frac": 0.051717332265476125, "mem_mb": 91.210752, "nll": -5.95272159576416, "elapsed_s": 251.9}70{"event": "train", "step": 1360, "loss": -6.037216663360596, "grad_norm": 0.288521409034729, "lr": 0.0002425343694621974, "ex_per_s": 343.26826420672285, "data_frac": 0.05142258870621125, "mem_mb": 91.210752, "nll": -6.037216663360596, "elapsed_s": 253.6}71{"event": "train", "step": 1380, "loss": -6.063647270202637, "grad_norm": 0.21889418363571167, "lr": 0.00024069653274247583, "ex_per_s": 346.0374366748012, "data_frac": 0.051129335469156025, "mem_mb": 91.210752, "nll": -6.063647270202637, "elapsed_s": 255.3}72{"event": "train", "step": 1400, "loss": -5.79464054107666, "grad_norm": 0.1381206512451172, "lr": 0.0002388380013465276, "ex_per_s": 348.7727125060349, "data_frac": 0.050843084576010826, "mem_mb": 91.210752, "nll": -5.79464054107666, "elapsed_s": 256.9}73{"event": "train", "step": 1420, "loss": -5.990894794464111, "grad_norm": 0.5114552974700928, "lr": 0.00023695928337827914, "ex_per_s": 351.473892138157, "data_frac": 0.050560545382147375, "mem_mb": 91.210752, "nll": -5.990894794464111, "elapsed_s": 258.6}74{"event": "train", "step": 1440, "loss": -5.742600917816162, "grad_norm": 0.485858291387558, "lr": 0.00023506089246046403, "ex_per_s": 354.13783744903316, "data_frac": 0.05028087851841541, "mem_mb": 91.210752, "nll": -5.742600917816162, "elapsed_s": 260.3}75{"event": "train", "step": 1460, "loss": -5.79490852355957, "grad_norm": 0.11229640990495682, "lr": 0.00023314334759420388, "ex_per_s": 356.76458853490925, "data_frac": 0.050005649743990244, "mem_mb": 91.210752, "nll": -5.79490852355957, "elapsed_s": 262.0}76{"event": "train", "step": 1480, "loss": -6.046689510345459, "grad_norm": 0.17627036571502686, "lr": 0.0002312071730171186, "ex_per_s": 359.3646403774535, "data_frac": 0.04973466840219766, "mem_mb": 91.210752, "nll": -6.046689510345459, "elapsed_s": 263.6}77{"event": "train", "step": 1500, "loss": -5.955654144287109, "grad_norm": 0.44177567958831787, "lr": 0.0002292528980600049, "ex_per_s": 361.9312944115242, "data_frac": 0.04946399257575438, "mem_mb": 91.210752, "nll": -5.955654144287109, "elapsed_s": 265.3}78{"event": "val", "step": 1500, "val_nll": -6.445288776055581, "best": -6.445527173841883, "bad": 1, "mem_mb": 91.210752, "elapsed_s": 328.7}79{"event": "train", "step": 1520, "loss": -5.949131965637207, "grad_norm": 0.2375475913286209, "lr": 0.00022728105700212285, "ex_per_s": 294.42719130980925, "data_frac": 0.03974476671012212, "mem_mb": 91.210752, "nll": -5.949131965637207, "elapsed_s": 330.4}80{"event": "train", "step": 1540, "loss": -5.87663459777832, "grad_norm": 0.20667387545108795, "lr": 0.00022529218892512886, "ex_per_s": 296.7964659631359, "data_frac": 0.039582378666588, "mem_mb": 91.210752, "nll": -5.87663459777832, "elapsed_s": 332.1}81{"event": "train", "step": 1560, "loss": -5.980387210845947, "grad_norm": 0.15004071593284607, "lr": 0.00022328683756569652, "ex_per_s": 299.14214438428655, "data_frac": 0.03941921587171449, "mem_mb": 91.210752, "nll": -5.980387210845947, "elapsed_s": 333.8}82{"event": "train", "step": 1580, "loss": -5.978658676147461, "grad_norm": 0.34112992882728577, "lr": 0.0002212655511668642, "ex_per_s": 301.46428980458035, "data_frac": 0.03925862394129312, "mem_mb": 91.210752, "nll": -5.978658676147461, "elapsed_s": 335.5}83{"event": "train", "step": 1600, "loss": -5.954066753387451, "grad_norm": 0.12589457631111145, "lr": 0.00021922888232815089, "ex_per_s": 303.7649817437447, "data_frac": 0.039098715293141115, "mem_mb": 91.210752, "nll": -5.954066753387451, "elapsed_s": 337.1}84{"event": "train", "step": 1620, "loss": -5.879581928253174, "grad_norm": 0.07455980777740479, "lr": 0.00021717738785448102, "ex_per_s": 306.0433647819607, "data_frac": 0.03894106511349863, "mem_mb": 91.210752, "nll": -5.879581928253174, "elapsed_s": 338.8}85{"event": "train", "step": 1640, "loss": -6.076663017272949, "grad_norm": 0.35954803228378296, "lr": 0.00021511162860395977, "ex_per_s": 308.2988119651278, "data_frac": 0.038785852801327333, "mem_mb": 91.210752, "nll": -6.076663017272949, "elapsed_s": 340.5}86{"event": "train", "step": 1660, "loss": -6.041078567504883, "grad_norm": 0.37353748083114624, "lr": 0.0002130321693345397, "ex_per_s": 310.5321835219835, "data_frac": 0.03863274138808066, "mem_mb": 91.210752, "nll": -6.041078567504883, "elapsed_s": 342.2}87{"event": "train", "step": 1680, "loss": -5.863365650177002, "grad_norm": 0.22345809638500214, "lr": 0.0002109395785496218, "ex_per_s": 312.7452032899627, "data_frac": 0.03847888859280399, "mem_mb": 91.210752, "nll": -5.863365650177002, "elapsed_s": 343.8}88{"event": "train", "step": 1700, "loss": -5.928807258605957, "grad_norm": 0.5705450773239136, "lr": 0.00020883442834263225, "ex_per_s": 314.93769251300193, "data_frac": 0.03832554069480668, "mem_mb": 91.210752, "nll": -5.928807258605957, "elapsed_s": 345.5}89{"event": "train", "step": 1720, "loss": -5.884502410888672, "grad_norm": 0.681938111782074, "lr": 0.00020671729424061788, "ex_per_s": 317.10951552412496, "data_frac": 0.0381745806181202, "mem_mb": 91.210752, "nll": -5.884502410888672, "elapsed_s": 347.2}90{"event": "train", "step": 1740, "loss": -6.271241188049316, "grad_norm": 0.6645751595497131, "lr": 0.00020458875504690275, "ex_per_s": 319.26131845678157, "data_frac": 0.03802428052143502, "mem_mb": 91.210752, "nll": -6.271241188049316, "elapsed_s": 348.8}91{"event": "train", "step": 1760, "loss": -5.924862384796143, "grad_norm": 0.6479564309120178, "lr": 0.000202449392682849, "ex_per_s": 321.3932993029716, "data_frac": 0.03787861628982093, "mem_mb": 91.210752, "nll": -5.924862384796143, "elapsed_s": 350.5}92{"event": "train", "step": 1780, "loss": -5.8675665855407715, "grad_norm": 0.7065231204032898, "lr": 0.00020029979202876563, "ex_per_s": 323.5013681887699, "data_frac": 0.03773509149660419, "mem_mb": 91.210752, "nll": -5.8675665855407715, "elapsed_s": 352.2}93{"event": "train", "step": 1800, "loss": -6.0464019775390625, "grad_norm": 1.0556100606918335, "lr": 0.00019814054076400787, "ex_per_s": 325.597125011037, "data_frac": 0.03759161321646137, "mem_mb": 91.210752, "nll": -6.0464019775390625, "elapsed_s": 353.9}94{"event": "train", "step": 1820, "loss": -5.934177398681641, "grad_norm": 0.6749878525733948, "lr": 0.00019597222920631147, "ex_per_s": 327.67405922858603, "data_frac": 0.037447924767895775, "mem_mb": 91.210752, "nll": -5.934177398681641, "elapsed_s": 355.5}95{"event": "train", "step": 1840, "loss": -6.071023464202881, "grad_norm": 0.560168981552124, "lr": 0.0001937954501504058, "ex_per_s": 329.73147817125977, "data_frac": 0.03730592292507447, "mem_mb": 91.210752, "nll": -6.071023464202881, "elapsed_s": 357.2}96{"event": "train", "step": 1860, "loss": -6.049586296081543, "grad_norm": 0.10110259056091309, "lr": 0.0001916107987059498, "ex_per_s": 331.7680195689794, "data_frac": 0.03716582702794178, "mem_mb": 91.210752, "nll": -6.049586296081543, "elapsed_s": 358.8}97{"event": "train", "step": 1880, "loss": -5.9473090171813965, "grad_norm": 0.0767834261059761, "lr": 0.00018941887213483484, "ex_per_s": 333.789163366619, "data_frac": 0.03702622729964076, "mem_mb": 91.210752, "nll": -5.9473090171813965, "elapsed_s": 360.5}98{"event": "train", "step": 1900, "loss": -5.98015832901001, "grad_norm": 0.1728878915309906, "lr": 0.00018722026968789906, "ex_per_s": 335.78952345116124, "data_frac": 0.0368875929914166, "mem_mb": 91.210752, "nll": -5.98015832901001, "elapsed_s": 362.2}99{"event": "train", "step": 1920, "loss": -5.961526393890381, "grad_norm": 0.1958780437707901, "lr": 0.00018501559244109864, "ex_per_s": 337.77196171187643, "data_frac": 0.036750668152303136, "mem_mb": 91.210752, "nll": -5.961526393890381, "elapsed_s": 363.8}100{"event": "train", "step": 1940, "loss": -6.046903610229492, "grad_norm": 0.1673535704612732, "lr": 0.00018280544313117932, "ex_per_s": 339.7345135516827, "data_frac": 0.036616049951114456, "mem_mb": 91.210752, "nll": -6.046903610229492, "elapsed_s": 365.5}101{"event": "train", "step": 1960, "loss": -5.9344377517700195, "grad_norm": 0.3267311155796051, "lr": 0.00018059042599089415, "ex_per_s": 341.67969248769515, "data_frac": 0.03648327531483391, "mem_mb": 91.210752, "nll": -5.9344377517700195, "elapsed_s": 367.2}102{"event": "train", "step": 1980, "loss": -5.960544109344482, "grad_norm": 0.903177797794342, "lr": 0.00017837114658381247, "ex_per_s": 343.60308901840256, "data_frac": 0.03634993690384121, "mem_mb": 91.210752, "nll": -5.960544109344482, "elapsed_s": 368.8}103{"event": "train", "step": 2000, "loss": -5.913472652435303, "grad_norm": 0.14637182652950287, "lr": 0.00017614821163876486, "ex_per_s": 345.5079884594028, "data_frac": 0.03621856842111737, "mem_mb": 91.210752, "nll": -5.913472652435303, "elapsed_s": 370.5}104{"event": "val", "step": 2000, "val_nll": -6.444991876640591, "best": -6.445527173841883, "bad": 2, "mem_mb": 91.210752, "elapsed_s": 433.8}105{"event": "train", "step": 2020, "loss": -5.923578262329102, "grad_norm": 1.0220565795898438, "lr": 0.00017392222888396941, "ex_per_s": 296.85479189456123, "data_frac": 0.03083723097677188, "mem_mb": 91.210752, "nll": -5.923578262329102, "elapsed_s": 435.5}106{"event": "train", "step": 2040, "loss": -5.873042583465576, "grad_norm": 0.23823031783103943, "lr": 0.000171693806880885, "ex_per_s": 298.646873220202, "data_frac": 0.030745915599574156, "mem_mb": 91.210752, "nll": -5.873042583465576, "elapsed_s": 437.2}107{"event": "train", "step": 2060, "loss": -6.140944957733154, "grad_norm": 0.26851576566696167, "lr": 0.0001694635548578366, "ex_per_s": 300.4250449635986, "data_frac": 0.030654297458056933, "mem_mb": 91.210752, "nll": -6.140944957733154, "elapsed_s": 438.9}108{"event": "train", "step": 2080, "loss": -5.960919380187988, "grad_norm": 0.5917397141456604, "lr": 0.0001672320825434584, "ex_per_s": 302.18887243523, "data_frac": 0.03056366615761909, "mem_mb": 91.210752, "nll": -5.960919380187988, "elapsed_s": 440.6}109{"event": "train", "step": 2100, "loss": -5.902200222015381, "grad_norm": 0.4123203158378601, "lr": 0.000165, "ex_per_s": 303.939491477057, "data_frac": 0.030472682521543348, "mem_mb": 91.210752, "nll": -5.902200222015381, "elapsed_s": 442.2}110{"event": "train", "step": 2120, "loss": -6.148244380950928, "grad_norm": 0.23805509507656097, "lr": 0.0001627679174565416, "ex_per_s": 305.6775939563998, "data_frac": 0.030384015637542464, "mem_mb": 91.210752, "nll": -6.148244380950928, "elapsed_s": 443.9}111{"event": "train", "step": 2140, "loss": -5.9738616943359375, "grad_norm": 0.4584055542945862, "lr": 0.00016053644514216336, "ex_per_s": 307.4035335998125, "data_frac": 0.030295656718994514, "mem_mb": 91.210752, "nll": -5.9738616943359375, "elapsed_s": 445.6}112{"event": "train", "step": 2160, "loss": -5.922845363616943, "grad_norm": 0.6371517181396484, "lr": 0.00015830619311911498, "ex_per_s": 309.1152707802658, "data_frac": 0.030207913098032106, "mem_mb": 91.210752, "nll": -5.922845363616943, "elapsed_s": 447.3}113{"event": "train", "step": 2180, "loss": -6.0809807777404785, "grad_norm": 0.24410203099250793, "lr": 0.00015607777111603058, "ex_per_s": 310.8154783960085, "data_frac": 0.030120064947049035, "mem_mb": 91.210752, "nll": -6.0809807777404785, "elapsed_s": 448.9}114{"event": "train", "step": 2200, "loss": -5.963162899017334, "grad_norm": 0.14549662172794342, "lr": 0.00015385178836123514, "ex_per_s": 312.503298894476, "data_frac": 0.030034115430992585, "mem_mb": 91.210752, "nll": -5.963162899017334, "elapsed_s": 450.6}115{"event": "train", "step": 2220, "loss": -5.8238372802734375, "grad_norm": 0.5776700377464294, "lr": 0.00015162885341618752, "ex_per_s": 314.1760861874876, "data_frac": 0.02995110616976474, "mem_mb": 91.210752, "nll": -5.8238372802734375, "elapsed_s": 452.3}116{"event": "train", "step": 2240, "loss": -5.960683345794678, "grad_norm": 0.37187492847442627, "lr": 0.00014940957400910585, "ex_per_s": 315.83742364120417, "data_frac": 0.029866188702379393, "mem_mb": 91.210752, "nll": -5.960683345794678, "elapsed_s": 453.9}117{"event": "train", "step": 2260, "loss": -5.739213466644287, "grad_norm": 0.10437458008527756, "lr": 0.00014719455686882068, "ex_per_s": 317.4863197590792, "data_frac": 0.029782194228650243, "mem_mb": 91.210752, "nll": -5.739213466644287, "elapsed_s": 455.6}118{"event": "train", "step": 2280, "loss": -6.007909297943115, "grad_norm": 0.24961963295936584, "lr": 0.0001449844075589013, "ex_per_s": 319.12552401490495, "data_frac": 0.02969751985306193, "mem_mb": 91.210752, "nll": -6.007909297943115, "elapsed_s": 457.3}119{"event": "train", "step": 2300, "loss": -5.997223854064941, "grad_norm": 0.2744145095348358, "lr": 0.0001427797303121009, "ex_per_s": 320.75316836109323, "data_frac": 0.029613647969644604, "mem_mb": 91.210752, "nll": -5.997223854064941, "elapsed_s": 459.0}120{"event": "train", "step": 2320, "loss": -6.068382740020752, "grad_norm": 0.3490552008152008, "lr": 0.00014058112786516518, "ex_per_s": 322.368565605462, "data_frac": 0.029529830950148463, "mem_mb": 91.210752, "nll": -6.068382740020752, "elapsed_s": 460.6}121{"event": "train", "step": 2340, "loss": -5.847777843475342, "grad_norm": 0.11277725547552109, "lr": 0.00013838920129405016, "ex_per_s": 323.97598439969977, "data_frac": 0.02944747814441103, "mem_mb": 91.210752, "nll": -5.847777843475342, "elapsed_s": 462.3}122{"event": "train", "step": 2360, "loss": -5.9251627922058105, "grad_norm": 0.2377963662147522, "lr": 0.0001362045498495942, "ex_per_s": 325.5712659555257, "data_frac": 0.02936573861025147, "mem_mb": 91.210752, "nll": -5.9251627922058105, "elapsed_s": 464.0}123{"event": "train", "step": 2380, "loss": -5.933005332946777, "grad_norm": 0.12726716697216034, "lr": 0.00013402777079368853, "ex_per_s": 327.15494887328026, "data_frac": 0.029283872343244623, "mem_mb": 91.210752, "nll": -5.933005332946777, "elapsed_s": 465.6}124{"event": "train", "step": 2400, "loss": -5.953297138214111, "grad_norm": 0.5663267374038696, "lr": 0.00013185945923599213, "ex_per_s": 328.72810809373993, "data_frac": 0.029203265651783366, "mem_mb": 91.210752, "nll": -5.953297138214111, "elapsed_s": 467.3}125{"event": "train", "step": 2420, "loss": -5.99528169631958, "grad_norm": 0.3233030438423157, "lr": 0.00012970020797123432, "ex_per_s": 330.29131522927315, "data_frac": 0.029124252198485485, "mem_mb": 91.210752, "nll": -5.99528169631958, "elapsed_s": 469.0}126{"event": "train", "step": 2440, "loss": -6.099708557128906, "grad_norm": 0.42002081871032715, "lr": 0.000127550607317151, "ex_per_s": 331.84312857562884, "data_frac": 0.029045105139371023, "mem_mb": 91.210752, "nll": -6.099708557128906, "elapsed_s": 470.6}127{"event": "train", "step": 2460, "loss": -6.130716800689697, "grad_norm": 0.13377095758914948, "lr": 0.00012541124495309727, "ex_per_s": 333.3817447865207, "data_frac": 0.028967738384886144, "mem_mb": 91.210752, "nll": -6.130716800689697, "elapsed_s": 472.3}128{"event": "train", "step": 2480, "loss": -5.937493801116943, "grad_norm": 0.2645162045955658, "lr": 0.00012328270575938212, "ex_per_s": 334.91239719433594, "data_frac": 0.028890347692154487, "mem_mb": 91.210752, "nll": -5.937493801116943, "elapsed_s": 474.0}129{"event": "train", "step": 2500, "loss": -5.754216194152832, "grad_norm": 0.39647525548934937, "lr": 0.00012116557165736775, "ex_per_s": 336.432736244437, "data_frac": 0.028813371325592637, "mem_mb": 91.210752, "nll": -5.754216194152832, "elapsed_s": 475.6}130{"event": "val", "step": 2500, "val_nll": -6.446134912098296, "best": -6.446134912098296, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 538.9}131{"event": "train", "step": 2520, "loss": -5.961039066314697, "grad_norm": 0.14766743779182434, "lr": 0.00011906042145037817, "ex_per_s": 298.3341087569061, "data_frac": 0.025368255838997718, "mem_mb": 91.210752, "nll": -5.961039066314697, "elapsed_s": 540.6}132{"event": "train", "step": 2540, "loss": -6.026669025421143, "grad_norm": 0.163873553276062, "lr": 0.00011696783066546026, "ex_per_s": 299.7698179549182, "data_frac": 0.025309625480593265, "mem_mb": 91.210752, "nll": -6.026669025421143, "elapsed_s": 542.3}133{"event": "train", "step": 2560, "loss": -5.948884010314941, "grad_norm": 0.09378398209810257, "lr": 0.00011488837139604017, "ex_per_s": 301.19674449493186, "data_frac": 0.025251703340168297, "mem_mb": 91.210752, "nll": -5.948884010314941, "elapsed_s": 544.0}134{"event": "train", "step": 2580, "loss": -5.959177494049072, "grad_norm": 0.14287294447422028, "lr": 0.00011282261214551898, "ex_per_s": 302.6144088381169, "data_frac": 0.025194611688237428, "mem_mb": 91.210752, "nll": -5.959177494049072, "elapsed_s": 545.7}135{"event": "train", "step": 2600, "loss": -5.84912633895874, "grad_norm": 0.14404144883155823, "lr": 0.00011077111767184915, "ex_per_s": 304.0247312412782, "data_frac": 0.02513756755835988, "mem_mb": 91.210752, "nll": -5.84912633895874, "elapsed_s": 547.4}136{"event": "train", "step": 2620, "loss": -5.885745048522949, "grad_norm": 0.9270579814910889, "lr": 0.00010873444883313583, "ex_per_s": 305.4296182918938, "data_frac": 0.025081351094881927, "mem_mb": 91.210752, "nll": -5.885745048522949, "elapsed_s": 549.0}137{"event": "train", "step": 2640, "loss": -5.983213901519775, "grad_norm": 0.26831021904945374, "lr": 0.00010671316243430345, "ex_per_s": 306.8270829479303, "data_frac": 0.02502528842894116, "mem_mb": 91.210752, "nll": -5.983213901519775, "elapsed_s": 550.7}138{"event": "train", "step": 2660, "loss": -5.817509174346924, "grad_norm": 0.1402476280927658, "lr": 0.00010470781107487112, "ex_per_s": 308.2152648333683, "data_frac": 0.02497088243965844, "mem_mb": 91.210752, "nll": -5.817509174346924, "elapsed_s": 552.4}139{"event": "train", "step": 2680, "loss": -5.956153392791748, "grad_norm": 0.6179718971252441, "lr": 0.00010271894299787715, "ex_per_s": 309.5955892263451, "data_frac": 0.024916255544069028, "mem_mb": 91.210752, "nll": -5.956153392791748, "elapsed_s": 554.1}140{"event": "train", "step": 2700, "loss": -6.129299640655518, "grad_norm": 0.4935493469238281, "lr": 0.00010074710193999506, "ex_per_s": 310.9688518654993, "data_frac": 0.024860720290618706, "mem_mb": 91.210752, "nll": -6.129299640655518, "elapsed_s": 555.7}141{"event": "train", "step": 2720, "loss": -5.876919269561768, "grad_norm": 0.16604933142662048, "lr": 9.879282698288144e-05, "ex_per_s": 312.3345601339946, "data_frac": 0.02480563880613405, "mem_mb": 91.210752, "nll": -5.876919269561768, "elapsed_s": 557.4}142{"event": "train", "step": 2740, "loss": -6.1224894523620605, "grad_norm": 0.3284367620944977, "lr": 9.685665240579613e-05, "ex_per_s": 313.6922483639167, "data_frac": 0.024751013798969105, "mem_mb": 91.210752, "nll": -6.1224894523620605, "elapsed_s": 559.1}143{"event": "train", "step": 2760, "loss": -5.9573516845703125, "grad_norm": 0.19600693881511688, "lr": 9.4939107539536e-05, "ex_per_s": 315.041862045543, "data_frac": 0.024697468095214527, "mem_mb": 91.210752, "nll": -5.9573516845703125, "elapsed_s": 560.7}144{"event": "train", "step": 2780, "loss": -6.013204574584961, "grad_norm": 0.4243514835834503, "lr": 9.304071662172087e-05, "ex_per_s": 316.38355451041076, "data_frac": 0.0246435183709613, "mem_mb": 91.210752, "nll": -6.013204574584961, "elapsed_s": 562.4}145{"event": "train", "step": 2800, "loss": -5.9796319007873535, "grad_norm": 0.47775107622146606, "lr": 9.116199865347241e-05, "ex_per_s": 317.716771677377, "data_frac": 0.02459049483644335, "mem_mb": 91.210752, "nll": -5.9796319007873535, "elapsed_s": 564.1}146{"event": "train", "step": 2820, "loss": -6.003994464874268, "grad_norm": 0.4563171863555908, "lr": 8.930346725752419e-05, "ex_per_s": 319.0427174730289, "data_frac": 0.024537780654737405, "mem_mb": 91.210752, "nll": -6.003994464874268, "elapsed_s": 565.7}147{"event": "train", "step": 2840, "loss": -5.872187614440918, "grad_norm": 0.3455256223678589, "lr": 8.746563053780262e-05, "ex_per_s": 320.36179918025493, "data_frac": 0.024485009336103817, "mem_mb": 91.210752, "nll": -5.872187614440918, "elapsed_s": 567.4}148{"event": "train", "step": 2860, "loss": -6.049644947052002, "grad_norm": 0.08263269066810608, "lr": 8.564899094051614e-05, "ex_per_s": 321.67292758327767, "data_frac": 0.02443297179320895, "mem_mb": 91.210752, "nll": -6.049644947052002, "elapsed_s": 569.1}149{"event": "train", "step": 2880, "loss": -5.770216464996338, "grad_norm": 0.2691022753715515, "lr": 8.385404511679161e-05, "ex_per_s": 322.9759476229907, "data_frac": 0.02438085411208989, "mem_mb": 91.210752, "nll": -5.770216464996338, "elapsed_s": 570.7}150{"event": "train", "step": 2900, "loss": -5.790160655975342, "grad_norm": 0.7417678833007812, "lr": 8.208128378689484e-05, "ex_per_s": 324.2696712315928, "data_frac": 0.024330043172364205, "mem_mb": 91.210752, "nll": -5.790160655975342, "elapsed_s": 572.4}151{"event": "train", "step": 2920, "loss": -5.968519687652588, "grad_norm": 0.5523889064788818, "lr": 8.03311916060725e-05, "ex_per_s": 325.5575217821378, "data_frac": 0.024280887555300317, "mem_mb": 91.210752, "nll": -5.968519687652588, "elapsed_s": 574.1}152{"event": "train", "step": 2940, "loss": -5.893084526062012, "grad_norm": 0.352428674697876, "lr": 7.860424703205225e-05, "ex_per_s": 326.83904824549523, "data_frac": 0.02423132404447185, "mem_mb": 91.210752, "nll": -5.893084526062012, "elapsed_s": 575.7}153{"event": "train", "step": 2960, "loss": -5.930917739868164, "grad_norm": 0.11328455060720444, "lr": 7.690092219423722e-05, "ex_per_s": 328.11548379675355, "data_frac": 0.024181326308380227, "mem_mb": 91.210752, "nll": -5.930917739868164, "elapsed_s": 577.4}154{"event": "train", "step": 2980, "loss": -5.882646560668945, "grad_norm": 0.8227868676185608, "lr": 7.522168276463048e-05, "ex_per_s": 329.3870262813244, "data_frac": 0.024131790416915336, "mem_mb": 91.210752, "nll": -5.882646560668945, "elapsed_s": 579.1}155{"event": "train", "step": 3000, "loss": -5.990944862365723, "grad_norm": 0.2723524570465088, "lr": 7.356698783052496e-05, "ex_per_s": 330.6513998270682, "data_frac": 0.024083460187761457, "mem_mb": 91.210752, "nll": -5.990944862365723, "elapsed_s": 580.7}156{"event": "val", "step": 3000, "val_nll": -6.44639632448096, "best": -6.44639632448096, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 643.8}157{"event": "train", "step": 3020, "loss": -5.960368633270264, "grad_norm": 0.9696645140647888, "lr": 7.193728976899362e-05, "ex_per_s": 299.4214764398597, "data_frac": 0.021681758299895777, "mem_mb": 91.210752, "nll": -5.960368633270264, "elapsed_s": 645.6}158{"event": "train", "step": 3040, "loss": -5.856933116912842, "grad_norm": 0.6339148879051208, "lr": 7.033303412321404e-05, "ex_per_s": 300.6239574243182, "data_frac": 0.021642920657917553, "mem_mb": 91.210752, "nll": -5.856933116912842, "elapsed_s": 647.2}159{"event": "train", "step": 3060, "loss": -6.072070598602295, "grad_norm": 0.14116710424423218, "lr": 6.875465948066139e-05, "ex_per_s": 301.81966659041996, "data_frac": 0.02160374120707855, "mem_mb": 91.210752, "nll": -6.072070598602295, "elapsed_s": 648.9}160{"event": "train", "step": 3080, "loss": -5.963530540466309, "grad_norm": 0.09078484773635864, "lr": 6.720259735320299e-05, "ex_per_s": 303.01018593947896, "data_frac": 0.0215646902052055, "mem_mb": 91.210752, "nll": -5.963530540466309, "elapsed_s": 650.6}161{"event": "train", "step": 3100, "loss": -6.006124019622803, "grad_norm": 0.47012436389923096, "lr": 6.567727205912722e-05, "ex_per_s": 304.19378717322763, "data_frac": 0.021526341475019205, "mem_mb": 91.210752, "nll": -6.006124019622803, "elapsed_s": 652.3}162{"event": "train", "step": 3120, "loss": -6.08865213394165, "grad_norm": 0.19578295946121216, "lr": 6.417910060713926e-05, "ex_per_s": 305.3710235433516, "data_frac": 0.021489697459421316, "mem_mb": 91.210752, "nll": -6.08865213394165, "elapsed_s": 653.9}163{"event": "train", "step": 3140, "loss": -6.004762172698975, "grad_norm": 0.08299801498651505, "lr": 6.270849258235485e-05, "ex_per_s": 306.5433812049495, "data_frac": 0.02145154705041235, "mem_mb": 91.210752, "nll": -6.004762172698975, "elapsed_s": 655.6}164{"event": "train", "step": 3160, "loss": -6.021054267883301, "grad_norm": 0.10535323619842529, "lr": 6.126585003432406e-05, "ex_per_s": 307.7104526777616, "data_frac": 0.021413532352161282, "mem_mb": 91.210752, "nll": -6.021054267883301, "elapsed_s": 657.3}165{"event": "train", "step": 3180, "loss": -6.090549945831299, "grad_norm": 0.21968282759189606, "lr": 5.985156736711483e-05, "ex_per_s": 308.87143231088623, "data_frac": 0.021375394345079245, "mem_mb": 91.210752, "nll": -6.090549945831299, "elapsed_s": 659.0}166{"event": "train", "step": 3200, "loss": -5.9986796379089355, "grad_norm": 0.4217994809150696, "lr": 5.846603123148688e-05, "ex_per_s": 310.0266922996041, "data_frac": 0.021338202886161655, "mem_mb": 91.210752, "nll": -5.9986796379089355, "elapsed_s": 660.6}167{"event": "train", "step": 3220, "loss": -5.9117889404296875, "grad_norm": 0.17284750938415527, "lr": 5.710962041918516e-05, "ex_per_s": 311.1753794907543, "data_frac": 0.02130107347237286, "mem_mb": 91.210752, "nll": -5.9117889404296875, "elapsed_s": 662.3}168{"event": "train", "step": 3240, "loss": -5.797316074371338, "grad_norm": 0.3840087056159973, "lr": 5.578270575938211e-05, "ex_per_s": 312.31936506081894, "data_frac": 0.021263896328165374, "mem_mb": 91.210752, "nll": -5.797316074371338, "elapsed_s": 664.0}169{"event": "train", "step": 3260, "loss": -5.958695411682129, "grad_norm": 0.14037933945655823, "lr": 5.448565001729653e-05, "ex_per_s": 313.45780303334027, "data_frac": 0.021227035028953976, "mem_mb": 91.210752, "nll": -5.958695411682129, "elapsed_s": 665.7}170{"event": "train", "step": 3280, "loss": -6.01618766784668, "grad_norm": 0.522440493106842, "lr": 5.3218807795017296e-05, "ex_per_s": 314.59001051032874, "data_frac": 0.02119074187679782, "mem_mb": 91.210752, "nll": -6.01618766784668, "elapsed_s": 667.3}171{"event": "train", "step": 3300, "loss": -5.800528526306152, "grad_norm": 0.5905721783638, "lr": 5.198252543455864e-05, "ex_per_s": 315.7174710256915, "data_frac": 0.021153973220063134, "mem_mb": 91.210752, "nll": -5.800528526306152, "elapsed_s": 669.0}172{"event": "train", "step": 3320, "loss": -5.98220682144165, "grad_norm": 0.18632712960243225, "lr": 5.07771409231737e-05, "ex_per_s": 316.83971622154036, "data_frac": 0.021118214345129224, "mem_mb": 91.210752, "nll": -5.98220682144165, "elapsed_s": 670.7}173{"event": "train", "step": 3340, "loss": -5.927611827850342, "grad_norm": 0.17184993624687195, "lr": 4.960298380095221e-05, "ex_per_s": 317.9567665957992, "data_frac": 0.021082597513834553, "mem_mb": 91.210752, "nll": -5.927611827850342, "elapsed_s": 672.3}174{"event": "train", "step": 3360, "loss": -5.910363674163818, "grad_norm": 0.06776314228773117, "lr": 4.846037507072751e-05, "ex_per_s": 319.067165944449, "data_frac": 0.021049216309816925, "mem_mb": 91.210752, "nll": -5.910363674163818, "elapsed_s": 674.0}175{"event": "train", "step": 3380, "loss": -6.121209144592285, "grad_norm": 0.17552392184734344, "lr": 4.7349627110317556e-05, "ex_per_s": 320.17468564972904, "data_frac": 0.021013799270514597, "mem_mb": 91.210752, "nll": -6.121209144592285, "elapsed_s": 675.7}176{"event": "train", "step": 3400, "loss": -5.875268459320068, "grad_norm": 0.6892944574356079, "lr": 4.6271043587123986e-05, "ex_per_s": 321.2769024722506, "data_frac": 0.020978606615958458, "mem_mb": 91.210752, "nll": -5.875268459320068, "elapsed_s": 677.3}177{"event": "train", "step": 3420, "loss": -5.832906246185303, "grad_norm": 0.341045081615448, "lr": 4.52249193751124e-05, "ex_per_s": 322.3732640135692, "data_frac": 0.020943874127914215, "mem_mb": 91.210752, "nll": -5.832906246185303, "elapsed_s": 679.0}178{"event": "train", "step": 3440, "loss": -5.914818286895752, "grad_norm": 0.0969712883234024, "lr": 4.4211540474196926e-05, "ex_per_s": 323.46473663014416, "data_frac": 0.02090933628104648, "mem_mb": 91.210752, "nll": -5.914818286895752, "elapsed_s": 680.7}179{"event": "train", "step": 3460, "loss": -6.059756278991699, "grad_norm": 0.1431690752506256, "lr": 4.3231183932050546e-05, "ex_per_s": 324.55065990501237, "data_frac": 0.020874956534773302, "mem_mb": 91.210752, "nll": -6.059756278991699, "elapsed_s": 682.3}180{"event": "train", "step": 3480, "loss": -5.917640686035156, "grad_norm": 0.2658033072948456, "lr": 4.228411776836306e-05, "ex_per_s": 325.63158641205547, "data_frac": 0.020841049413636736, "mem_mb": 91.210752, "nll": -5.917640686035156, "elapsed_s": 684.0}181{"event": "train", "step": 3500, "loss": -6.032753944396973, "grad_norm": 0.18727131187915802, "lr": 4.137060090156723e-05, "ex_per_s": 326.706876926426, "data_frac": 0.020807470448436952, "mem_mb": 91.210752, "nll": -6.032753944396973, "elapsed_s": 685.7}182{"event": "val", "step": 3500, "val_nll": -6.446326105452635, "best": -6.44639632448096, "bad": 1, "mem_mb": 91.210752, "elapsed_s": 748.7}183{"event": "train", "step": 3520, "loss": -5.9309868812561035, "grad_norm": 0.20190085470676422, "lr": 4.049088307805306e-05, "ex_per_s": 300.2255400204723, "data_frac": 0.019027611599843768, "mem_mb": 91.210752, "nll": -5.9309868812561035, "elapsed_s": 750.4}184{"event": "train", "step": 3540, "loss": -6.134633541107178, "grad_norm": 0.34960252046585083, "lr": 3.964520480388956e-05, "ex_per_s": 301.25916051391033, "data_frac": 0.01900008595601971, "mem_mb": 91.210752, "nll": -6.134633541107178, "elapsed_s": 752.1}185{"event": "train", "step": 3560, "loss": -6.137310028076172, "grad_norm": 0.25902259349823, "lr": 3.88337972790728e-05, "ex_per_s": 302.28704003446796, "data_frac": 0.018973225748510063, "mem_mb": 91.210752, "nll": -6.137310028076172, "elapsed_s": 753.8}186{"event": "train", "step": 3580, "loss": -5.901824474334717, "grad_norm": 0.162025585770607, "lr": 3.805688233431824e-05, "ex_per_s": 303.3109729857988, "data_frac": 0.018946326139801784, "mem_mb": 91.210752, "nll": -5.901824474334717, "elapsed_s": 755.4}187{"event": "train", "step": 3600, "loss": -6.091848373413086, "grad_norm": 0.10443494468927383, "lr": 3.7314672370414325e-05, "ex_per_s": 304.33076390648375, "data_frac": 0.01891978812672103, "mem_mb": 91.210752, "nll": -6.091848373413086, "elapsed_s": 757.1}188{"event": "train", "step": 3620, "loss": -5.906505107879639, "grad_norm": 0.9878615736961365, "lr": 3.660737030015427e-05, "ex_per_s": 305.3466492779454, "data_frac": 0.018892329486863098, "mem_mb": 91.210752, "nll": -5.906505107879639, "elapsed_s": 758.8}189{"event": "train", "step": 3640, "loss": -6.05250358581543, "grad_norm": 0.3317784368991852, "lr": 3.5935169492861716e-05, "ex_per_s": 306.3580990904531, "data_frac": 0.01886537416326558, "mem_mb": 91.210752, "nll": -6.05250358581543, "elapsed_s": 760.5}190{"event": "train", "step": 3660, "loss": -5.827733516693115, "grad_norm": 0.534112274646759, "lr": 3.52982537215255e-05, "ex_per_s": 307.3646323026561, "data_frac": 0.01883868011811799, "mem_mb": 91.210752, "nll": -5.827733516693115, "elapsed_s": 762.1}191{"event": "train", "step": 3680, "loss": -5.9880499839782715, "grad_norm": 0.41866829991340637, "lr": 3.469679711255794e-05, "ex_per_s": 308.3666746227446, "data_frac": 0.018811769331896034, "mem_mb": 91.210752, "nll": -5.9880499839782715, "elapsed_s": 763.8}192{"event": "train", "step": 3700, "loss": -6.0650739669799805, "grad_norm": 0.6269816160202026, "lr": 3.41309640981904e-05, "ex_per_s": 309.3656592244569, "data_frac": 0.018784953415110266, "mem_mb": 91.210752, "nll": -6.0650739669799805, "elapsed_s": 765.5}193{"event": "train", "step": 3720, "loss": -5.940706253051758, "grad_norm": 0.5012120604515076, "lr": 3.3600909371519225e-05, "ex_per_s": 310.35980159311026, "data_frac": 0.01875889173013996, "mem_mb": 91.210752, "nll": -5.940706253051758, "elapsed_s": 767.2}194{"event": "train", "step": 3740, "loss": -6.185410022735596, "grad_norm": 0.20691636204719543, "lr": 3.31067778442141e-05, "ex_per_s": 311.3494300872181, "data_frac": 0.018732336958240066, "mem_mb": 91.210752, "nll": -6.185410022735596, "elapsed_s": 768.8}195{"event": "train", "step": 3760, "loss": -6.143742084503174, "grad_norm": 0.16626064479351044, "lr": 3.2648704606900735e-05, "ex_per_s": 312.3351225578004, "data_frac": 0.018706125245785313, "mem_mb": 91.210752, "nll": -6.143742084503174, "elapsed_s": 770.5}196{"event": "train", "step": 3780, "loss": -5.960902214050293, "grad_norm": 0.34700775146484375, "lr": 3.222681489222838e-05, "ex_per_s": 313.3160194704389, "data_frac": 0.018679823968329526, "mem_mb": 91.210752, "nll": -5.960902214050293, "elapsed_s": 772.2}197{"event": "train", "step": 3800, "loss": -6.0961432456970215, "grad_norm": 0.10071080923080444, "lr": 3.1841224040632484e-05, "ex_per_s": 314.2917554854369, "data_frac": 0.018653951956585955, "mem_mb": 91.210752, "nll": -6.0961432456970215, "elapsed_s": 773.8}198{"event": "train", "step": 3820, "loss": -5.9659504890441895, "grad_norm": 0.11943749338388443, "lr": 3.149203746880175e-05, "ex_per_s": 315.26504233337334, "data_frac": 0.018628918445710508, "mem_mb": 91.210752, "nll": -5.9659504890441895, "elapsed_s": 775.5}199{"event": "train", "step": 3840, "loss": -5.866632461547852, "grad_norm": 0.9479031562805176, "lr": 3.117935064085835e-05, "ex_per_s": 316.23421104696496, "data_frac": 0.01860322962052761, "mem_mb": 91.210752, "nll": -5.866632461547852, "elapsed_s": 777.2}200{"event": "train", "step": 3860, "loss": -6.01378870010376, "grad_norm": 0.11130700260400772, "lr": 3.090324904225885e-05, "ex_per_s": 317.1995931772877, "data_frac": 0.018577280071027377, "mem_mb": 91.210752, "nll": -6.01378870010376, "elapsed_s": 778.9}201{"event": "train", "step": 3880, "loss": -5.886845588684082, "grad_norm": 0.2641935348510742, "lr": 3.0663808156423456e-05, "ex_per_s": 318.1612216725386, "data_frac": 0.018551785275313808, "mem_mb": 91.210752, "nll": -5.886845588684082, "elapsed_s": 780.5}202{"event": "train", "step": 3900, "loss": -6.1688690185546875, "grad_norm": 0.9575124382972717, "lr": 3.046109344409957e-05, "ex_per_s": 319.11897388824593, "data_frac": 0.018526293047731308, "mem_mb": 91.210752, "nll": -6.1688690185546875, "elapsed_s": 782.2}203{"event": "train", "step": 3920, "loss": -6.036338806152344, "grad_norm": 0.383069783449173, "lr": 3.0295160325465435e-05, "ex_per_s": 320.07243238558567, "data_frac": 0.018500626160465123, "mem_mb": 91.210752, "nll": -6.036338806152344, "elapsed_s": 783.9}204{"event": "train", "step": 3940, "loss": -5.91218900680542, "grad_norm": 0.31595027446746826, "lr": 3.016605416497886e-05, "ex_per_s": 321.02224076471845, "data_frac": 0.018475944711328986, "mem_mb": 91.210752, "nll": -5.91218900680542, "elapsed_s": 785.5}205{"event": "train", "step": 3960, "loss": -6.067218780517578, "grad_norm": 0.09801644831895828, "lr": 3.0073810258974985e-05, "ex_per_s": 321.9659673448446, "data_frac": 0.018451857929322718, "mem_mb": 91.210752, "nll": -6.067218780517578, "elapsed_s": 787.2}206{"event": "train", "step": 3980, "loss": -5.800100326538086, "grad_norm": 0.11506549268960953, "lr": 3.0018453826016683e-05, "ex_per_s": 322.9064846025251, "data_frac": 0.01842721623930975, "mem_mb": 91.210752, "nll": -5.800100326538086, "elapsed_s": 788.9}207{"event": "train", "step": 4000, "loss": -6.05084753036499, "grad_norm": 0.7945476174354553, "lr": 2.9999999999999997e-05, "ex_per_s": 323.84259774628526, "data_frac": 0.01840253080277092, "mem_mb": 91.210752, "nll": -6.05084753036499, "elapsed_s": 790.6}208{"event": "val", "step": 4000, "val_nll": -6.4462719280720995, "best": -6.44639632448096, "bad": 2, "mem_mb": 91.210752, "elapsed_s": 853.5}209{"event": "done", "steps": 4000, "best_val": -6.44639632448096, "wall_s": 853.4852628707886, "ex_per_s": 299.9465967800273, "peak_mem_mb": 91.210752, "n_params": 3455619, "device": "mps", "data_time_frac": 0.017044627621086567, "elapsed_s": 853.5}210
Files · main
Menu

Theme