1{"event": "train", "step": 20, "loss": -6.12005615234375, "grad_norm": 0.3674488365650177, "lr": 3.149999999999999e-05, "ex_per_s": 91.73564390958343, "data_frac": 0.8706991994935112, "mem_mb": 91.210752, "nll": -6.12005615234375, "elapsed_s": 14.0}2{"event": "train", "step": 40, "loss": -5.888593673706055, "grad_norm": 0.9757420420646667, "lr": 6.149999999999999e-05, "ex_per_s": 162.448184723972, "data_frac": 0.7794348518134484, "mem_mb": 91.210752, "nll": -5.888593673706055, "elapsed_s": 15.8}3{"event": "train", "step": 60, "loss": -5.98077917098999, "grad_norm": 0.21594838798046112, "lr": 9.149999999999999e-05, "ex_per_s": 218.9340421487789, "data_frac": 0.7011489406189537, "mem_mb": 91.210752, "nll": -5.98077917098999, "elapsed_s": 17.6}4{"event": "train", "step": 80, "loss": -6.080472946166992, "grad_norm": 0.33754172921180725, "lr": 0.0001215, "ex_per_s": 266.3199666479293, "data_frac": 0.6404581269089846, "mem_mb": 91.210752, "nll": -6.080472946166992, "elapsed_s": 19.3}5{"event": "train", "step": 100, "loss": -5.950660228729248, "grad_norm": 0.15621012449264526, "lr": 0.0001515, "ex_per_s": 306.3006554538768, "data_frac": 0.5897827444469287, "mem_mb": 91.210752, "nll": -5.950660228729248, "elapsed_s": 20.9}6{"event": "train", "step": 120, "loss": -5.911417484283447, "grad_norm": 0.8098359107971191, "lr": 0.00018149999999999997, "ex_per_s": 340.26698113120847, "data_frac": 0.5464515421428975, "mem_mb": 91.210752, "nll": -5.911417484283447, "elapsed_s": 22.6}7{"event": "train", "step": 140, "loss": -5.830864429473877, "grad_norm": 0.7668975591659546, "lr": 0.00021149999999999996, "ex_per_s": 369.60401289576106, "data_frac": 0.5092450709066326, "mem_mb": 91.210752, "nll": -5.830864429473877, "elapsed_s": 24.3}8{"event": "train", "step": 160, "loss": -5.737630844116211, "grad_norm": 0.521062970161438, "lr": 0.0002415, "ex_per_s": 395.2034192358443, "data_frac": 0.47689659104964455, "mem_mb": 91.210752, "nll": -5.737630844116211, "elapsed_s": 26.0}9{"event": "train", "step": 180, "loss": -5.976109027862549, "grad_norm": 1.000357747077942, "lr": 0.0002715, "ex_per_s": 417.68097297545603, "data_frac": 0.44844399853817135, "mem_mb": 91.210752, "nll": -5.976109027862549, "elapsed_s": 27.6}10{"event": "train", "step": 200, "loss": -5.972109794616699, "grad_norm": 0.4469493627548218, "lr": 0.0003, "ex_per_s": 437.57615319565065, "data_frac": 0.42321294881544635, "mem_mb": 91.210752, "nll": -5.972109794616699, "elapsed_s": 29.3}11{"event": "train", "step": 220, "loss": -6.031529426574707, "grad_norm": 1.137394666671753, "lr": 0.00029998154617398326, "ex_per_s": 455.4176516178602, "data_frac": 0.40082812939636336, "mem_mb": 91.210752, "nll": -6.031529426574707, "elapsed_s": 31.0}12{"event": "train", "step": 240, "loss": -5.915230751037598, "grad_norm": 0.20762112736701965, "lr": 0.000299926189741025, "ex_per_s": 471.4174199579262, "data_frac": 0.38069937481854316, "mem_mb": 91.210752, "nll": -5.915230751037598, "elapsed_s": 32.6}13{"event": "train", "step": 260, "loss": -5.888279914855957, "grad_norm": 0.1409970372915268, "lr": 0.0002998339458350211, "ex_per_s": 485.8353441770624, "data_frac": 0.362524962041992, "mem_mb": 91.210752, "nll": -5.888279914855957, "elapsed_s": 34.3}14{"event": "train", "step": 280, "loss": -5.945128917694092, "grad_norm": 0.1190439909696579, "lr": 0.0002997048396745345, "ex_per_s": 498.9662428120432, "data_frac": 0.3460818439258751, "mem_mb": 91.210752, "nll": -5.945128917694092, "elapsed_s": 36.0}15{"event": "train", "step": 300, "loss": -5.968123912811279, "grad_norm": 0.24222058057785034, "lr": 0.0002995389065559004, "ex_per_s": 510.93363510557975, "data_frac": 0.33109264301925295, "mem_mb": 91.210752, "nll": -5.968123912811279, "elapsed_s": 37.6}16{"event": "train", "step": 320, "loss": -5.967358589172363, "grad_norm": 0.3248257339000702, "lr": 0.0002993361918435765, "ex_per_s": 521.8917669525179, "data_frac": 0.3173754080697634, "mem_mb": 91.210752, "nll": -5.967358589172363, "elapsed_s": 39.3}17{"event": "train", "step": 340, "loss": -5.834793567657471, "grad_norm": 0.3113088309764862, "lr": 0.00029909675095774114, "ex_per_s": 531.9807850297708, "data_frac": 0.3047925145916486, "mem_mb": 91.210752, "nll": -5.834793567657471, "elapsed_s": 40.9}18{"event": "train", "step": 360, "loss": -5.996526718139648, "grad_norm": 0.24073287844657898, "lr": 0.00029882064935914164, "ex_per_s": 541.2132099066033, "data_frac": 0.29315400273837067, "mem_mb": 91.210752, "nll": -5.996526718139648, "elapsed_s": 42.6}19{"event": "train", "step": 380, "loss": -6.005407333374023, "grad_norm": 1.191633701324463, "lr": 0.00029850796253119824, "ex_per_s": 549.7788572843014, "data_frac": 0.28240119503006317, "mem_mb": 91.210752, "nll": -6.005407333374023, "elapsed_s": 44.3}20{"event": "train", "step": 400, "loss": -5.994297027587891, "grad_norm": 1.4911694526672363, "lr": 0.0002981587759593675, "ex_per_s": 557.776263154419, "data_frac": 0.2724625146432643, "mem_mb": 91.210752, "nll": -5.994297027587891, "elapsed_s": 45.9}21{"event": "train", "step": 420, "loss": -5.964649200439453, "grad_norm": 0.5808878540992737, "lr": 0.0002977731851077716, "ex_per_s": 565.2996964472637, "data_frac": 0.26324933563320274, "mem_mb": 91.210752, "nll": -5.964649200439453, "elapsed_s": 47.6}22{"event": "train", "step": 440, "loss": -5.938328742980957, "grad_norm": 0.7643961906433105, "lr": 0.0002973512953930992, "ex_per_s": 572.3110888768905, "data_frac": 0.2546526017497761, "mem_mb": 91.210752, "nll": -5.938328742980957, "elapsed_s": 49.2}23{"event": "train", "step": 460, "loss": -6.101223945617676, "grad_norm": 0.5868762135505676, "lr": 0.0002968932221557859, "ex_per_s": 578.850200684911, "data_frac": 0.24661362934537015, "mem_mb": 91.210752, "nll": -6.101223945617676, "elapsed_s": 50.9}24{"event": "train", "step": 480, "loss": -5.877638339996338, "grad_norm": 0.2692728042602539, "lr": 0.0002963990906284808, "ex_per_s": 584.9780385722701, "data_frac": 0.23907898522149443, "mem_mb": 91.210752, "nll": -5.877638339996338, "elapsed_s": 52.6}25{"event": "train", "step": 500, "loss": -5.925544261932373, "grad_norm": 0.5457468032836914, "lr": 0.00029586903590180956, "ex_per_s": 590.693240290869, "data_frac": 0.2322341247668225, "mem_mb": 91.210752, "nll": -5.925544261932373, "elapsed_s": 54.2}26{"event": "val", "step": 500, "val_nll": -6.439823932501979, "best": -6.439823932501979, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 119.1}27{"event": "train", "step": 520, "loss": -6.040815830230713, "grad_norm": 0.2539583742618561, "lr": 0.00029530320288744205, "ex_per_s": 274.88353710014894, "data_frac": 0.10403135218233905, "mem_mb": 91.210752, "nll": -6.040815830230713, "elapsed_s": 121.1}28{"event": "train", "step": 540, "loss": -5.878683567047119, "grad_norm": 0.5774997472763062, "lr": 0.0002947017462784745, "ex_per_s": 281.57065565058144, "data_frac": 0.10272432285909487, "mem_mb": 91.210752, "nll": -5.878683567047119, "elapsed_s": 122.8}29{"event": "train", "step": 560, "loss": -5.972832202911377, "grad_norm": 0.0884946808218956, "lr": 0.00029406483050713824, "ex_per_s": 288.0785690765495, "data_frac": 0.10144634365241852, "mem_mb": 91.210752, "nll": -5.972832202911377, "elapsed_s": 124.5}30{"event": "train", "step": 580, "loss": -5.924684047698975, "grad_norm": 0.09403974562883377, "lr": 0.0002933926296998457, "ex_per_s": 294.4141465927031, "data_frac": 0.10020071324051306, "mem_mb": 91.210752, "nll": -5.924684047698975, "elapsed_s": 126.1}31{"event": "train", "step": 600, "loss": -5.972785949707031, "grad_norm": 0.710675060749054, "lr": 0.0002926853276295856, "ex_per_s": 300.58093590361733, "data_frac": 0.09898690207560375, "mem_mb": 91.210752, "nll": -5.972785949707031, "elapsed_s": 127.8}32{"event": "train", "step": 620, "loss": -5.952574729919434, "grad_norm": 0.38336241245269775, "lr": 0.00029194311766568174, "ex_per_s": 306.60983655304585, "data_frac": 0.09781068505491204, "mem_mb": 91.210752, "nll": -5.952574729919434, "elapsed_s": 129.5}33{"event": "train", "step": 640, "loss": -6.06535005569458, "grad_norm": 0.4723568260669708, "lr": 0.0002911662027209272, "ex_per_s": 312.4776688898185, "data_frac": 0.09666210158416436, "mem_mb": 91.210752, "nll": -6.06535005569458, "elapsed_s": 131.1}34{"event": "train", "step": 660, "loss": -6.008022785186768, "grad_norm": 1.0304464101791382, "lr": 0.00029035479519611045, "ex_per_s": 318.1978909418732, "data_frac": 0.09554248162571051, "mem_mb": 91.210752, "nll": -6.008022785186768, "elapsed_s": 132.8}35{"event": "train", "step": 680, "loss": -6.006795406341553, "grad_norm": 0.1172376275062561, "lr": 0.0002895091169219469, "ex_per_s": 323.7707680009293, "data_frac": 0.09444949136229418, "mem_mb": 91.210752, "nll": -6.006795406341553, "elapsed_s": 134.5}36{"event": "train", "step": 700, "loss": -5.987449645996094, "grad_norm": 0.657974362373352, "lr": 0.00028862939909843273, "ex_per_s": 329.2043296004172, "data_frac": 0.09338837392780842, "mem_mb": 91.210752, "nll": -5.987449645996094, "elapsed_s": 136.1}37{"event": "train", "step": 720, "loss": -6.063577175140381, "grad_norm": 0.22192463278770447, "lr": 0.0002877158822316369, "ex_per_s": 334.51402203890177, "data_frac": 0.09234990055481393, "mem_mb": 91.210752, "nll": -6.063577175140381, "elapsed_s": 137.8}38{"event": "train", "step": 740, "loss": -6.033608436584473, "grad_norm": 0.24840837717056274, "lr": 0.00028676881606794945, "ex_per_s": 339.69664207118984, "data_frac": 0.0913346681564584, "mem_mb": 91.210752, "nll": -6.033608436584473, "elapsed_s": 139.5}39{"event": "train", "step": 760, "loss": -5.959526062011719, "grad_norm": 0.7832713723182678, "lr": 0.000285788459525803, "ex_per_s": 344.75480583662835, "data_frac": 0.09034393890192459, "mem_mb": 91.210752, "nll": -5.959526062011719, "elapsed_s": 141.1}40{"event": "train", "step": 780, "loss": -5.894853591918945, "grad_norm": 0.28174686431884766, "lr": 0.00028477508062488754, "ex_per_s": 349.6976760747438, "data_frac": 0.0893764078003901, "mem_mb": 91.210752, "nll": -5.894853591918945, "elapsed_s": 142.8}41{"event": "train", "step": 800, "loss": -5.815494537353516, "grad_norm": 0.27376091480255127, "lr": 0.000283728956412876, "ex_per_s": 354.52403626801737, "data_frac": 0.0884309896225955, "mem_mb": 91.210752, "nll": -5.815494537353516, "elapsed_s": 144.5}42{"event": "train", "step": 820, "loss": -5.933772087097168, "grad_norm": 0.26308685541152954, "lr": 0.00028265037288968236, "ex_per_s": 359.2428694960809, "data_frac": 0.08750650167324749, "mem_mb": 91.210752, "nll": -5.933772087097168, "elapsed_s": 146.1}43{"event": "train", "step": 840, "loss": -5.806540012359619, "grad_norm": 0.5614535808563232, "lr": 0.0002815396249292725, "ex_per_s": 363.8520342544253, "data_frac": 0.0866006304747935, "mem_mb": 91.210752, "nll": -5.806540012359619, "elapsed_s": 147.8}44{"event": "train", "step": 860, "loss": -5.914388179779053, "grad_norm": 1.4593393802642822, "lr": 0.00028039701619904774, "ex_per_s": 368.35048589374173, "data_frac": 0.08571393372582568, "mem_mb": 91.210752, "nll": -5.914388179779053, "elapsed_s": 149.5}45{"event": "train", "step": 880, "loss": -5.868752479553223, "grad_norm": 0.6328785419464111, "lr": 0.00027922285907682625, "ex_per_s": 372.75517801517674, "data_frac": 0.08484876791873505, "mem_mb": 91.210752, "nll": -5.868752479553223, "elapsed_s": 151.1}46{"event": "train", "step": 900, "loss": -5.986020088195801, "grad_norm": 0.3657004237174988, "lr": 0.00027801747456544134, "ex_per_s": 377.06135961086187, "data_frac": 0.084000475175651, "mem_mb": 91.210752, "nll": -5.986020088195801, "elapsed_s": 152.8}47{"event": "train", "step": 920, "loss": -6.024444103240967, "grad_norm": 0.09846056252717972, "lr": 0.00027678119220498265, "ex_per_s": 381.2711278265619, "data_frac": 0.08317377757150242, "mem_mb": 91.210752, "nll": -6.024444103240967, "elapsed_s": 154.5}48{"event": "train", "step": 940, "loss": -6.080338001251221, "grad_norm": 0.927926778793335, "lr": 0.0002755143499827035, "ex_per_s": 385.38795686754247, "data_frac": 0.08236540873613485, "mem_mb": 91.210752, "nll": -6.080338001251221, "elapsed_s": 156.1}49{"event": "train", "step": 960, "loss": -5.966891765594482, "grad_norm": 0.422894686460495, "lr": 0.00027421729424061787, "ex_per_s": 389.4221608429255, "data_frac": 0.0815707217200297, "mem_mb": 91.210752, "nll": -5.966891765594482, "elapsed_s": 157.8}50{"event": "train", "step": 980, "loss": -5.92490291595459, "grad_norm": 0.44107887148857117, "lr": 0.00027289037958081483, "ex_per_s": 393.37187936948357, "data_frac": 0.08079275102137953, "mem_mb": 91.210752, "nll": -5.92490291595459, "elapsed_s": 159.5}51{"event": "train", "step": 1000, "loss": -6.0729851722717285, "grad_norm": 0.10190042108297348, "lr": 0.0002715339687685131, "ex_per_s": 397.24202111673344, "data_frac": 0.08003162143817642, "mem_mb": 91.210752, "nll": -6.0729851722717285, "elapsed_s": 161.2}52{"event": "val", "step": 1000, "val_nll": -6.446426667424411, "best": -6.446426667424411, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 224.7}53{"event": "train", "step": 1020, "loss": -6.030945777893066, "grad_norm": 0.08971472829580307, "lr": 0.0002701484326328852, "ex_per_s": 288.32543550407496, "data_frac": 0.05700617659822447, "mem_mb": 91.210752, "nll": -6.030945777893066, "elapsed_s": 226.5}54{"event": "train", "step": 1040, "loss": -5.9730377197265625, "grad_norm": 0.34442105889320374, "lr": 0.00026873414996567596, "ex_per_s": 291.8245795733822, "data_frac": 0.056642910586932486, "mem_mb": 91.210752, "nll": -5.9730377197265625, "elapsed_s": 228.1}55{"event": "train", "step": 1060, "loss": -6.180736541748047, "grad_norm": 0.14116819202899933, "lr": 0.0002672915074176452, "ex_per_s": 295.27236669064104, "data_frac": 0.05628476808490039, "mem_mb": 91.210752, "nll": -6.180736541748047, "elapsed_s": 229.8}56{"event": "train", "step": 1080, "loss": -5.814763069152832, "grad_norm": 0.5020195841789246, "lr": 0.00026582089939286075, "ex_per_s": 298.6713195458147, "data_frac": 0.05593113597212545, "mem_mb": 91.210752, "nll": -5.814763069152832, "elapsed_s": 231.5}57{"event": "train", "step": 1100, "loss": -6.053089618682861, "grad_norm": 0.1253257542848587, "lr": 0.0002643227279408728, "ex_per_s": 302.02246404000914, "data_frac": 0.05558352442454096, "mem_mb": 91.210752, "nll": -6.053089618682861, "elapsed_s": 233.1}58{"event": "train", "step": 1120, "loss": -6.051904678344727, "grad_norm": 0.4769086241722107, "lr": 0.000262797402646797, "ex_per_s": 305.324267260457, "data_frac": 0.05524090977993542, "mem_mb": 91.210752, "nll": -6.051904678344727, "elapsed_s": 234.8}59{"event": "train", "step": 1140, "loss": -6.020127773284912, "grad_norm": 0.13551904261112213, "lr": 0.0002612453405193386, "ex_per_s": 308.5758886625969, "data_frac": 0.054905254159892794, "mem_mb": 91.210752, "nll": -6.020127773284912, "elapsed_s": 236.5}60{"event": "train", "step": 1160, "loss": -5.970531940460205, "grad_norm": 0.5845354199409485, "lr": 0.0002596669658767859, "ex_per_s": 311.7863671476687, "data_frac": 0.05457193067215393, "mem_mb": 91.210752, "nll": -5.970531940460205, "elapsed_s": 238.2}61{"event": "train", "step": 1180, "loss": -5.956852912902832, "grad_norm": 0.16032469272613525, "lr": 0.00025806271023100637, "ex_per_s": 314.95141553703115, "data_frac": 0.05424414540427583, "mem_mb": 91.210752, "nll": -5.956852912902832, "elapsed_s": 239.8}62{"event": "train", "step": 1200, "loss": -5.945013523101807, "grad_norm": 0.19988079369068146, "lr": 0.00025643301216947504, "ex_per_s": 318.07351618730144, "data_frac": 0.05391965143984167, "mem_mb": 91.210752, "nll": -5.945013523101807, "elapsed_s": 241.5}63{"event": "train", "step": 1220, "loss": -6.03377628326416, "grad_norm": 0.5703750848770142, "lr": 0.0002547783172353695, "ex_per_s": 321.1536141106761, "data_frac": 0.05359988049432382, "mem_mb": 91.210752, "nll": -6.03377628326416, "elapsed_s": 243.2}64{"event": "train", "step": 1240, "loss": -6.059384822845459, "grad_norm": 0.07035098224878311, "lr": 0.00025309907780576277, "ex_per_s": 324.19270176312784, "data_frac": 0.05328354895437358, "mem_mb": 91.210752, "nll": -6.059384822845459, "elapsed_s": 244.8}65{"event": "train", "step": 1260, "loss": -6.054232597351074, "grad_norm": 0.21063868701457977, "lr": 0.0002513957529679477, "ex_per_s": 327.1925053815025, "data_frac": 0.052973094046996586, "mem_mb": 91.210752, "nll": -6.054232597351074, "elapsed_s": 246.5}66{"event": "train", "step": 1280, "loss": -6.043889999389648, "grad_norm": 0.5397829413414001, "lr": 0.00024966880839392746, "ex_per_s": 330.1516871473011, "data_frac": 0.05266587434462745, "mem_mb": 91.210752, "nll": -6.043889999389648, "elapsed_s": 248.2}67{"event": "train", "step": 1300, "loss": -5.867954254150391, "grad_norm": 0.779882550239563, "lr": 0.0002479187162131052, "ex_per_s": 333.07285440704067, "data_frac": 0.05236300818151369, "mem_mb": 91.210752, "nll": -5.867954254150391, "elapsed_s": 249.8}68{"event": "train", "step": 1320, "loss": -5.829307556152344, "grad_norm": 0.1990368366241455, "lr": 0.00024614595488320833, "ex_per_s": 335.95335848456216, "data_frac": 0.0520664832721634, "mem_mb": 91.210752, "nll": -5.829307556152344, "elapsed_s": 251.5}69{"event": "train", "step": 1340, "loss": -5.952577114105225, "grad_norm": 0.33566397428512573, "lr": 0.00024435100905948387, "ex_per_s": 338.8002013122292, "data_frac": 0.051773376913181915, "mem_mb": 91.210752, "nll": -5.952577114105225, "elapsed_s": 253.2}70{"event": "train", "step": 1360, "loss": -6.038855075836182, "grad_norm": 0.6013423800468445, "lr": 0.0002425343694621974, "ex_per_s": 341.6111059190257, "data_frac": 0.051483750045818295, "mem_mb": 91.210752, "nll": -6.038855075836182, "elapsed_s": 254.8}71{"event": "train", "step": 1380, "loss": -6.0677690505981445, "grad_norm": 0.16446533799171448, "lr": 0.00024069653274247583, "ex_per_s": 344.3742261337258, "data_frac": 0.05119732977037362, "mem_mb": 91.210752, "nll": -6.0677690505981445, "elapsed_s": 256.5}72{"event": "train", "step": 1400, "loss": -5.796565055847168, "grad_norm": 0.1724846065044403, "lr": 0.0002388380013465276, "ex_per_s": 347.11416733571485, "data_frac": 0.05091506004150302, "mem_mb": 91.210752, "nll": -5.796565055847168, "elapsed_s": 258.2}73{"event": "train", "step": 1420, "loss": -5.988810062408447, "grad_norm": 0.8083173632621765, "lr": 0.00023695928337827914, "ex_per_s": 349.82034123834643, "data_frac": 0.05063710119676124, "mem_mb": 91.210752, "nll": -5.988810062408447, "elapsed_s": 259.8}74{"event": "train", "step": 1440, "loss": -5.744534015655518, "grad_norm": 0.17077358067035675, "lr": 0.00023506089246046403, "ex_per_s": 352.4939175049454, "data_frac": 0.050361335826592725, "mem_mb": 91.210752, "nll": -5.744534015655518, "elapsed_s": 261.5}75{"event": "train", "step": 1460, "loss": -5.803011417388916, "grad_norm": 0.4680931866168976, "lr": 0.00023314334759420388, "ex_per_s": 355.1309699734762, "data_frac": 0.05008994315449973, "mem_mb": 91.210752, "nll": -5.803011417388916, "elapsed_s": 263.2}76{"event": "train", "step": 1480, "loss": -6.047210693359375, "grad_norm": 0.19411218166351318, "lr": 0.0002312071730171186, "ex_per_s": 357.73491853331825, "data_frac": 0.049821075035673006, "mem_mb": 91.210752, "nll": -6.047210693359375, "elapsed_s": 264.8}77{"event": "train", "step": 1500, "loss": -5.95872163772583, "grad_norm": 0.2060663402080536, "lr": 0.0002292528980600049, "ex_per_s": 360.30151116372673, "data_frac": 0.049556115985541334, "mem_mb": 91.210752, "nll": -5.95872163772583, "elapsed_s": 266.5}78{"event": "val", "step": 1500, "val_nll": -6.447684537039633, "best": -6.447684537039633, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 330.1}79{"event": "train", "step": 1520, "loss": -5.952143669128418, "grad_norm": 0.47579720616340637, "lr": 0.00022728105700212285, "ex_per_s": 293.18779488398525, "data_frac": 0.03983145448974624, "mem_mb": 91.210752, "nll": -5.952143669128418, "elapsed_s": 331.8}80{"event": "train", "step": 1540, "loss": -5.879750728607178, "grad_norm": 0.08371023088693619, "lr": 0.00022529218892512886, "ex_per_s": 295.5549549280959, "data_frac": 0.039668491191615085, "mem_mb": 91.210752, "nll": -5.879750728607178, "elapsed_s": 333.5}81{"event": "train", "step": 1560, "loss": -5.982535362243652, "grad_norm": 0.43554168939590454, "lr": 0.00022328683756569652, "ex_per_s": 297.8978846727703, "data_frac": 0.039506432901486, "mem_mb": 91.210752, "nll": -5.982535362243652, "elapsed_s": 335.2}82{"event": "train", "step": 1580, "loss": -5.982180595397949, "grad_norm": 0.07394397258758545, "lr": 0.0002212655511668642, "ex_per_s": 300.2133208327886, "data_frac": 0.03934850365873748, "mem_mb": 91.210752, "nll": -5.982180595397949, "elapsed_s": 336.9}83{"event": "train", "step": 1600, "loss": -5.957630157470703, "grad_norm": 0.26167020201683044, "lr": 0.00021922888232815089, "ex_per_s": 302.5115885905393, "data_frac": 0.03919027113012891, "mem_mb": 91.210752, "nll": -5.957630157470703, "elapsed_s": 338.5}84{"event": "train", "step": 1620, "loss": -5.880764007568359, "grad_norm": 0.1619175523519516, "lr": 0.00021717738785448102, "ex_per_s": 304.7879623512781, "data_frac": 0.03903390015365849, "mem_mb": 91.210752, "nll": -5.880764007568359, "elapsed_s": 340.2}85{"event": "train", "step": 1640, "loss": -6.077530860900879, "grad_norm": 0.13726632297039032, "lr": 0.00021511162860395977, "ex_per_s": 307.0410533426903, "data_frac": 0.03887851823053184, "mem_mb": 91.210752, "nll": -6.077530860900879, "elapsed_s": 341.9}86{"event": "train", "step": 1660, "loss": -6.041773796081543, "grad_norm": 0.7374817728996277, "lr": 0.0002130321693345397, "ex_per_s": 309.272614872387, "data_frac": 0.038725945445058904, "mem_mb": 91.210752, "nll": -6.041773796081543, "elapsed_s": 343.6}87{"event": "train", "step": 1680, "loss": -5.86683988571167, "grad_norm": 0.1506103277206421, "lr": 0.0002109395785496218, "ex_per_s": 311.48407257278376, "data_frac": 0.03857521547972262, "mem_mb": 91.210752, "nll": -5.86683988571167, "elapsed_s": 345.2}88{"event": "train", "step": 1700, "loss": -5.926962852478027, "grad_norm": 0.6337291598320007, "lr": 0.00020883442834263225, "ex_per_s": 313.6729200343791, "data_frac": 0.03842520303712326, "mem_mb": 91.210752, "nll": -5.926962852478027, "elapsed_s": 346.9}89{"event": "train", "step": 1720, "loss": -5.889098167419434, "grad_norm": 0.7394022345542908, "lr": 0.00020671729424061788, "ex_per_s": 315.8429784044018, "data_frac": 0.03827616379972629, "mem_mb": 91.210752, "nll": -5.889098167419434, "elapsed_s": 348.6}90{"event": "train", "step": 1740, "loss": -6.273293495178223, "grad_norm": 0.3273378908634186, "lr": 0.00020458875504690275, "ex_per_s": 317.9934468122258, "data_frac": 0.03812868369063761, "mem_mb": 91.210752, "nll": -6.273293495178223, "elapsed_s": 350.2}91{"event": "train", "step": 1760, "loss": -5.932523727416992, "grad_norm": 0.27622073888778687, "lr": 0.000202449392682849, "ex_per_s": 320.1233847381502, "data_frac": 0.03798343734653488, "mem_mb": 91.210752, "nll": -5.932523727416992, "elapsed_s": 351.9}92{"event": "train", "step": 1780, "loss": -5.872031211853027, "grad_norm": 0.5262339115142822, "lr": 0.00020029979202876563, "ex_per_s": 322.23509131764314, "data_frac": 0.037838826184243286, "mem_mb": 91.210752, "nll": -5.872031211853027, "elapsed_s": 353.6}93{"event": "train", "step": 1800, "loss": -6.048254013061523, "grad_norm": 1.1212674379348755, "lr": 0.00019814054076400787, "ex_per_s": 324.32851255167395, "data_frac": 0.037695898547804155, "mem_mb": 91.210752, "nll": -6.048254013061523, "elapsed_s": 355.2}94{"event": "train", "step": 1820, "loss": -5.933318614959717, "grad_norm": 0.9279379844665527, "lr": 0.00019597222920631147, "ex_per_s": 326.39658014531284, "data_frac": 0.037556105731613054, "mem_mb": 91.210752, "nll": -5.933318614959717, "elapsed_s": 356.9}95{"event": "train", "step": 1840, "loss": -6.0722198486328125, "grad_norm": 0.42181745171546936, "lr": 0.0001937954501504058, "ex_per_s": 328.4503583457103, "data_frac": 0.03741538869311534, "mem_mb": 91.210752, "nll": -6.0722198486328125, "elapsed_s": 358.6}96{"event": "train", "step": 1860, "loss": -6.052122592926025, "grad_norm": 0.14684821665287018, "lr": 0.0001916107987059498, "ex_per_s": 330.4846081998558, "data_frac": 0.03727651580869571, "mem_mb": 91.210752, "nll": -6.052122592926025, "elapsed_s": 360.2}97{"event": "train", "step": 1880, "loss": -5.948825359344482, "grad_norm": 0.4162394106388092, "lr": 0.00018941887213483484, "ex_per_s": 332.50204325473834, "data_frac": 0.03713859452388076, "mem_mb": 91.210752, "nll": -5.948825359344482, "elapsed_s": 361.9}98{"event": "train", "step": 1900, "loss": -5.982935428619385, "grad_norm": 0.2586296498775482, "lr": 0.00018722026968789906, "ex_per_s": 334.50036724070895, "data_frac": 0.03700219893202046, "mem_mb": 91.210752, "nll": -5.982935428619385, "elapsed_s": 363.6}99{"event": "train", "step": 1920, "loss": -5.963915824890137, "grad_norm": 0.4236266016960144, "lr": 0.00018501559244109864, "ex_per_s": 336.48066902988273, "data_frac": 0.036866743439074146, "mem_mb": 91.210752, "nll": -5.963915824890137, "elapsed_s": 365.2}100{"event": "train", "step": 1940, "loss": -6.049343585968018, "grad_norm": 0.33647871017456055, "lr": 0.00018280544313117932, "ex_per_s": 338.44328540042955, "data_frac": 0.03673359324177524, "mem_mb": 91.210752, "nll": -6.049343585968018, "elapsed_s": 366.9}101{"event": "train", "step": 1960, "loss": -5.939020156860352, "grad_norm": 0.07539704442024231, "lr": 0.00018059042599089415, "ex_per_s": 340.3911416508639, "data_frac": 0.03660073871083354, "mem_mb": 91.210752, "nll": -5.939020156860352, "elapsed_s": 368.6}102{"event": "train", "step": 1980, "loss": -5.9647955894470215, "grad_norm": 0.6281135082244873, "lr": 0.00017837114658381247, "ex_per_s": 342.32255842820456, "data_frac": 0.03646958611767274, "mem_mb": 91.210752, "nll": -5.9647955894470215, "elapsed_s": 370.2}103{"event": "train", "step": 2000, "loss": -5.915889263153076, "grad_norm": 0.41833072900772095, "lr": 0.00017614821163876486, "ex_per_s": 344.23485022317544, "data_frac": 0.03634017412851188, "mem_mb": 91.210752, "nll": -5.915889263153076, "elapsed_s": 371.9}104{"event": "val", "step": 2000, "val_nll": -6.448475636641461, "best": -6.448475636641461, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 435.4}105{"event": "train", "step": 2020, "loss": -5.928366661071777, "grad_norm": 0.5644353628158569, "lr": 0.00017392222888396941, "ex_per_s": 295.78700592799544, "data_frac": 0.030946657476934432, "mem_mb": 91.210752, "nll": -5.928366661071777, "elapsed_s": 437.1}106{"event": "train", "step": 2040, "loss": -5.873862266540527, "grad_norm": 0.35582831501960754, "lr": 0.000171693806880885, "ex_per_s": 297.57629972159754, "data_frac": 0.03085671586726657, "mem_mb": 91.210752, "nll": -5.873862266540527, "elapsed_s": 438.8}107{"event": "train", "step": 2060, "loss": -6.145205497741699, "grad_norm": 0.2957032322883606, "lr": 0.0001694635548578366, "ex_per_s": 299.35300555312705, "data_frac": 0.030767323208846283, "mem_mb": 91.210752, "nll": -6.145205497741699, "elapsed_s": 440.5}108{"event": "train", "step": 2080, "loss": -5.962329864501953, "grad_norm": 0.46098586916923523, "lr": 0.0001672320825434584, "ex_per_s": 301.1154313970829, "data_frac": 0.0306779789087719, "mem_mb": 91.210752, "nll": -5.962329864501953, "elapsed_s": 442.1}109{"event": "train", "step": 2100, "loss": -5.899989128112793, "grad_norm": 0.9632615447044373, "lr": 0.000165, "ex_per_s": 302.863322620945, "data_frac": 0.030589839646998872, "mem_mb": 91.210752, "nll": -5.899989128112793, "elapsed_s": 443.8}110{"event": "train", "step": 2120, "loss": -6.151390552520752, "grad_norm": 0.507055401802063, "lr": 0.0001627679174565416, "ex_per_s": 304.5991358140052, "data_frac": 0.030502190113299105, "mem_mb": 91.210752, "nll": -6.151390552520752, "elapsed_s": 445.5}111{"event": "train", "step": 2140, "loss": -5.978710651397705, "grad_norm": 0.10240347683429718, "lr": 0.00016053644514216336, "ex_per_s": 306.32278006587865, "data_frac": 0.030415401128543934, "mem_mb": 91.210752, "nll": -5.978710651397705, "elapsed_s": 447.2}112{"event": "train", "step": 2160, "loss": -5.9275617599487305, "grad_norm": 0.31116387248039246, "lr": 0.00015830619311911498, "ex_per_s": 308.0342463610345, "data_frac": 0.03032916144464977, "mem_mb": 91.210752, "nll": -5.9275617599487305, "elapsed_s": 448.8}113{"event": "train", "step": 2180, "loss": -6.082663536071777, "grad_norm": 0.32297834753990173, "lr": 0.00015607777111603058, "ex_per_s": 309.7340578801805, "data_frac": 0.03024319893538794, "mem_mb": 91.210752, "nll": -6.082663536071777, "elapsed_s": 450.5}114{"event": "train", "step": 2200, "loss": -5.963308334350586, "grad_norm": 0.5499728322029114, "lr": 0.00015385178836123514, "ex_per_s": 311.4224710736082, "data_frac": 0.030158948360305575, "mem_mb": 91.210752, "nll": -5.963308334350586, "elapsed_s": 452.2}115{"event": "train", "step": 2220, "loss": -5.8256425857543945, "grad_norm": 0.7449060678482056, "lr": 0.00015162885341618752, "ex_per_s": 313.0971038020291, "data_frac": 0.030074201139257958, "mem_mb": 91.210752, "nll": -5.8256425857543945, "elapsed_s": 453.8}116{"event": "train", "step": 2240, "loss": -5.961644172668457, "grad_norm": 0.713131308555603, "lr": 0.00014940957400910585, "ex_per_s": 314.76235786855216, "data_frac": 0.02999088528062552, "mem_mb": 91.210752, "nll": -5.961644172668457, "elapsed_s": 455.5}117{"event": "train", "step": 2260, "loss": -5.736505508422852, "grad_norm": 0.8857096433639526, "lr": 0.00014719455686882068, "ex_per_s": 316.4120809601385, "data_frac": 0.02990895107993875, "mem_mb": 91.210752, "nll": -5.736505508422852, "elapsed_s": 457.2}118{"event": "train", "step": 2280, "loss": -6.010248184204102, "grad_norm": 0.5078855752944946, "lr": 0.0001449844075589013, "ex_per_s": 318.0521072915076, "data_frac": 0.029827278588273047, "mem_mb": 91.210752, "nll": -6.010248184204102, "elapsed_s": 458.8}119{"event": "train", "step": 2300, "loss": -6.002241134643555, "grad_norm": 0.10627080500125885, "lr": 0.0001427797303121009, "ex_per_s": 319.68039314703555, "data_frac": 0.029746062729018533, "mem_mb": 91.210752, "nll": -6.002241134643555, "elapsed_s": 460.5}120{"event": "train", "step": 2320, "loss": -6.069090843200684, "grad_norm": 0.4285055696964264, "lr": 0.00014058112786516518, "ex_per_s": 321.29479724192225, "data_frac": 0.02966451668135126, "mem_mb": 91.210752, "nll": -6.069090843200684, "elapsed_s": 462.2}121{"event": "train", "step": 2340, "loss": -5.851993083953857, "grad_norm": 0.366129070520401, "lr": 0.00013838920129405016, "ex_per_s": 322.895680171639, "data_frac": 0.029584034726191588, "mem_mb": 91.210752, "nll": -5.851993083953857, "elapsed_s": 463.8}122{"event": "train", "step": 2360, "loss": -5.930131435394287, "grad_norm": 0.08191855996847153, "lr": 0.0001362045498495942, "ex_per_s": 324.48632746994275, "data_frac": 0.029504225391317345, "mem_mb": 91.210752, "nll": -5.930131435394287, "elapsed_s": 465.5}123{"event": "train", "step": 2380, "loss": -5.932521343231201, "grad_norm": 0.13215112686157227, "lr": 0.00013402777079368853, "ex_per_s": 326.0633402914763, "data_frac": 0.029425651273127303, "mem_mb": 91.210752, "nll": -5.932521343231201, "elapsed_s": 467.2}124{"event": "train", "step": 2400, "loss": -5.956867218017578, "grad_norm": 0.875450849533081, "lr": 0.00013185945923599213, "ex_per_s": 327.63167679397577, "data_frac": 0.029347347919390684, "mem_mb": 91.210752, "nll": -5.956867218017578, "elapsed_s": 468.9}125{"event": "train", "step": 2420, "loss": -5.996809482574463, "grad_norm": 0.3518834412097931, "lr": 0.00012970020797123432, "ex_per_s": 329.18899872101287, "data_frac": 0.029269245236294143, "mem_mb": 91.210752, "nll": -5.996809482574463, "elapsed_s": 470.5}126{"event": "train", "step": 2440, "loss": -6.0994157791137695, "grad_norm": 0.40163108706474304, "lr": 0.000127550607317151, "ex_per_s": 330.7330814792081, "data_frac": 0.0291919042354175, "mem_mb": 91.210752, "nll": -6.0994157791137695, "elapsed_s": 472.2}127{"event": "train", "step": 2460, "loss": -6.132595062255859, "grad_norm": 0.11878899484872818, "lr": 0.00012541124495309727, "ex_per_s": 332.27015320970304, "data_frac": 0.029115458577042234, "mem_mb": 91.210752, "nll": -6.132595062255859, "elapsed_s": 473.9}128{"event": "train", "step": 2480, "loss": -5.945192337036133, "grad_norm": 0.1363336443901062, "lr": 0.00012328270575938212, "ex_per_s": 333.79469573338264, "data_frac": 0.02903970883223855, "mem_mb": 91.210752, "nll": -5.945192337036133, "elapsed_s": 475.5}129{"event": "train", "step": 2500, "loss": -5.756168365478516, "grad_norm": 0.7004461288452148, "lr": 0.00012116557165736775, "ex_per_s": 335.3067652160474, "data_frac": 0.028964348128281556, "mem_mb": 91.210752, "nll": -5.756168365478516, "elapsed_s": 477.2}130{"event": "val", "step": 2500, "val_nll": -6.449929799895296, "best": -6.449929799895296, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 540.5}131{"event": "train", "step": 2520, "loss": -5.9605793952941895, "grad_norm": 0.7435715198516846, "lr": 0.00011906042145037817, "ex_per_s": 297.4466682027227, "data_frac": 0.02551274479807002, "mem_mb": 91.210752, "nll": -5.9605793952941895, "elapsed_s": 542.3}132{"event": "train", "step": 2540, "loss": -6.030704975128174, "grad_norm": 0.47608399391174316, "lr": 0.00011696783066546026, "ex_per_s": 298.88400999627896, "data_frac": 0.02545682636590209, "mem_mb": 91.210752, "nll": -6.030704975128174, "elapsed_s": 543.9}133{"event": "train", "step": 2560, "loss": -5.948773384094238, "grad_norm": 0.2422042340040207, "lr": 0.00011488837139604017, "ex_per_s": 300.31186499789146, "data_frac": 0.025401112846515438, "mem_mb": 91.210752, "nll": -5.948773384094238, "elapsed_s": 545.6}134{"event": "train", "step": 2580, "loss": -5.96376371383667, "grad_norm": 0.20102456212043762, "lr": 0.00011282261214551898, "ex_per_s": 301.73247003054325, "data_frac": 0.025345439901330598, "mem_mb": 91.210752, "nll": -5.96376371383667, "elapsed_s": 547.3}135{"event": "train", "step": 2600, "loss": -5.851315498352051, "grad_norm": 0.32607194781303406, "lr": 0.00011077111767184915, "ex_per_s": 303.14330055727316, "data_frac": 0.025290627386425377, "mem_mb": 91.210752, "nll": -5.851315498352051, "elapsed_s": 549.0}136{"event": "train", "step": 2620, "loss": -5.89043664932251, "grad_norm": 0.9279305934906006, "lr": 0.00010873444883313583, "ex_per_s": 304.54634503109514, "data_frac": 0.025235784423342397, "mem_mb": 91.210752, "nll": -5.89043664932251, "elapsed_s": 550.6}137{"event": "train", "step": 2640, "loss": -5.987027168273926, "grad_norm": 0.40841832756996155, "lr": 0.00010671316243430345, "ex_per_s": 305.94112065879403, "data_frac": 0.025181784794355858, "mem_mb": 91.210752, "nll": -5.987027168273926, "elapsed_s": 552.3}138{"event": "train", "step": 2660, "loss": -5.822195053100586, "grad_norm": 0.11964211612939835, "lr": 0.00010470781107487112, "ex_per_s": 307.3281020627417, "data_frac": 0.025127797061734455, "mem_mb": 91.210752, "nll": -5.822195053100586, "elapsed_s": 554.0}139{"event": "train", "step": 2680, "loss": -5.959531784057617, "grad_norm": 0.2904872000217438, "lr": 0.00010271894299787715, "ex_per_s": 308.70577749547357, "data_frac": 0.02507439747345986, "mem_mb": 91.210752, "nll": -5.959531784057617, "elapsed_s": 555.7}140{"event": "train", "step": 2700, "loss": -6.133109092712402, "grad_norm": 0.7302188873291016, "lr": 0.00010074710193999506, "ex_per_s": 310.0713673212071, "data_frac": 0.025022110127194273, "mem_mb": 91.210752, "nll": -6.133109092712402, "elapsed_s": 557.3}141{"event": "train", "step": 2720, "loss": -5.8831353187561035, "grad_norm": 0.3669418692588806, "lr": 9.879282698288144e-05, "ex_per_s": 311.43058844635664, "data_frac": 0.02497118054028568, "mem_mb": 91.210752, "nll": -5.8831353187561035, "elapsed_s": 559.0}142{"event": "train", "step": 2740, "loss": -6.125751972198486, "grad_norm": 0.1315528154373169, "lr": 9.685665240579613e-05, "ex_per_s": 312.7840407604471, "data_frac": 0.02491890252311481, "mem_mb": 91.210752, "nll": -6.125751972198486, "elapsed_s": 560.7}143{"event": "train", "step": 2760, "loss": -5.960939407348633, "grad_norm": 0.2491128146648407, "lr": 9.4939107539536e-05, "ex_per_s": 314.1304710324608, "data_frac": 0.02486670761379252, "mem_mb": 91.210752, "nll": -5.960939407348633, "elapsed_s": 562.4}144{"event": "train", "step": 2780, "loss": -6.014832496643066, "grad_norm": 0.7507601976394653, "lr": 9.304071662172087e-05, "ex_per_s": 315.46982440274286, "data_frac": 0.024815848704248706, "mem_mb": 91.210752, "nll": -6.014832496643066, "elapsed_s": 564.0}145{"event": "train", "step": 2800, "loss": -5.980314254760742, "grad_norm": 0.9408416152000427, "lr": 9.116199865347241e-05, "ex_per_s": 316.8008758073632, "data_frac": 0.0247642406409882, "mem_mb": 91.210752, "nll": -5.980314254760742, "elapsed_s": 565.7}146{"event": "train", "step": 2820, "loss": -6.008308410644531, "grad_norm": 0.1251251995563507, "lr": 8.930346725752419e-05, "ex_per_s": 318.12378095485604, "data_frac": 0.02471325652285459, "mem_mb": 91.210752, "nll": -6.008308410644531, "elapsed_s": 567.4}147{"event": "train", "step": 2840, "loss": -5.880282878875732, "grad_norm": 0.10276840627193451, "lr": 8.746563053780262e-05, "ex_per_s": 319.4403331489338, "data_frac": 0.02466324879895198, "mem_mb": 91.210752, "nll": -5.880282878875732, "elapsed_s": 569.0}148{"event": "train", "step": 2860, "loss": -6.048560619354248, "grad_norm": 0.28456151485443115, "lr": 8.564899094051614e-05, "ex_per_s": 320.7501762546828, "data_frac": 0.024613764182076104, "mem_mb": 91.210752, "nll": -6.048560619354248, "elapsed_s": 570.7}149{"event": "train", "step": 2880, "loss": -5.772538661956787, "grad_norm": 0.3942430019378662, "lr": 8.385404511679161e-05, "ex_per_s": 322.05287401285864, "data_frac": 0.024563852918922977, "mem_mb": 91.210752, "nll": -5.772538661956787, "elapsed_s": 572.4}150{"event": "train", "step": 2900, "loss": -5.795225620269775, "grad_norm": 0.5791579484939575, "lr": 8.208128378689484e-05, "ex_per_s": 323.3482813700948, "data_frac": 0.02451429901613835, "mem_mb": 91.210752, "nll": -5.795225620269775, "elapsed_s": 574.0}151{"event": "train", "step": 2920, "loss": -5.970181941986084, "grad_norm": 0.7553589940071106, "lr": 8.03311916060725e-05, "ex_per_s": 324.63595032520954, "data_frac": 0.024465581075202785, "mem_mb": 91.210752, "nll": -5.970181941986084, "elapsed_s": 575.7}152{"event": "train", "step": 2940, "loss": -5.9047746658325195, "grad_norm": 0.3774472773075104, "lr": 7.860424703205225e-05, "ex_per_s": 325.91429288097623, "data_frac": 0.02441633217671485, "mem_mb": 91.210752, "nll": -5.9047746658325195, "elapsed_s": 577.4}153{"event": "train", "step": 2960, "loss": -5.93799352645874, "grad_norm": 0.4201558530330658, "lr": 7.690092219423722e-05, "ex_per_s": 327.1864971992843, "data_frac": 0.024368091659500487, "mem_mb": 91.210752, "nll": -5.93799352645874, "elapsed_s": 579.0}154{"event": "train", "step": 2980, "loss": -5.894800186157227, "grad_norm": 0.4688147008419037, "lr": 7.522168276463048e-05, "ex_per_s": 328.4515222141104, "data_frac": 0.024319540544468915, "mem_mb": 91.210752, "nll": -5.894800186157227, "elapsed_s": 580.7}155{"event": "train", "step": 3000, "loss": -5.9960198402404785, "grad_norm": 0.11653976887464523, "lr": 7.356698783052496e-05, "ex_per_s": 329.706886095606, "data_frac": 0.024271806288191984, "mem_mb": 91.210752, "nll": -5.9960198402404785, "elapsed_s": 582.4}156{"event": "val", "step": 3000, "val_nll": -6.4483297421796015, "best": -6.449929799895296, "bad": 1, "mem_mb": 91.210752, "elapsed_s": 645.6}157{"event": "train", "step": 3020, "loss": -5.961395263671875, "grad_norm": 0.932748556137085, "lr": 7.193728976899362e-05, "ex_per_s": 298.61560878370955, "data_frac": 0.021856931797051973, "mem_mb": 91.210752, "nll": -5.961395263671875, "elapsed_s": 647.3}158{"event": "train", "step": 3040, "loss": -5.862003326416016, "grad_norm": 0.7020903825759888, "lr": 7.033303412321404e-05, "ex_per_s": 299.81685591347707, "data_frac": 0.021820076402726905, "mem_mb": 91.210752, "nll": -5.862003326416016, "elapsed_s": 649.0}159{"event": "train", "step": 3060, "loss": -6.072027206420898, "grad_norm": 0.2764829993247986, "lr": 6.875465948066139e-05, "ex_per_s": 301.0125036101025, "data_frac": 0.021782845230837865, "mem_mb": 91.210752, "nll": -6.072027206420898, "elapsed_s": 650.6}160{"event": "train", "step": 3080, "loss": -5.9694108963012695, "grad_norm": 0.22716213762760162, "lr": 6.720259735320299e-05, "ex_per_s": 302.2022530660211, "data_frac": 0.02174618081292531, "mem_mb": 91.210752, "nll": -5.9694108963012695, "elapsed_s": 652.3}161{"event": "train", "step": 3100, "loss": -6.014222145080566, "grad_norm": 0.3991992175579071, "lr": 6.567727205912722e-05, "ex_per_s": 303.38561995484685, "data_frac": 0.021709107643892032, "mem_mb": 91.210752, "nll": -6.014222145080566, "elapsed_s": 654.0}162{"event": "train", "step": 3120, "loss": -6.092358589172363, "grad_norm": 0.6135021448135376, "lr": 6.417910060713926e-05, "ex_per_s": 304.56260682327724, "data_frac": 0.021672370293790145, "mem_mb": 91.210752, "nll": -6.092358589172363, "elapsed_s": 655.7}163{"event": "train", "step": 3140, "loss": -6.009486675262451, "grad_norm": 0.42082226276397705, "lr": 6.270849258235485e-05, "ex_per_s": 305.73383237413486, "data_frac": 0.021635798717391786, "mem_mb": 91.210752, "nll": -6.009486675262451, "elapsed_s": 657.3}164{"event": "train", "step": 3160, "loss": -6.0227742195129395, "grad_norm": 0.1489672064781189, "lr": 6.126585003432406e-05, "ex_per_s": 306.8973307630403, "data_frac": 0.02160042613030383, "mem_mb": 91.210752, "nll": -6.0227742195129395, "elapsed_s": 659.0}165{"event": "train", "step": 3180, "loss": -6.096859931945801, "grad_norm": 0.11174474656581879, "lr": 5.985156736711483e-05, "ex_per_s": 308.05774336573904, "data_frac": 0.021564627203622806, "mem_mb": 91.210752, "nll": -6.096859931945801, "elapsed_s": 660.7}166{"event": "train", "step": 3200, "loss": -5.997987747192383, "grad_norm": 0.4225763976573944, "lr": 5.846603123148688e-05, "ex_per_s": 309.211842288764, "data_frac": 0.02152849453621592, "mem_mb": 91.210752, "nll": -5.997987747192383, "elapsed_s": 662.4}167{"event": "train", "step": 3220, "loss": -5.9146623611450195, "grad_norm": 0.17164485156536102, "lr": 5.710962041918516e-05, "ex_per_s": 310.36006037145114, "data_frac": 0.021492807908737543, "mem_mb": 91.210752, "nll": -5.9146623611450195, "elapsed_s": 664.0}168{"event": "train", "step": 3240, "loss": -5.802014350891113, "grad_norm": 0.3957598805427551, "lr": 5.578270575938211e-05, "ex_per_s": 311.50330136003896, "data_frac": 0.021457032258675283, "mem_mb": 91.210752, "nll": -5.802014350891113, "elapsed_s": 665.7}169{"event": "train", "step": 3260, "loss": -5.9607625007629395, "grad_norm": 0.3966607451438904, "lr": 5.448565001729653e-05, "ex_per_s": 312.64075070113546, "data_frac": 0.021421837466585523, "mem_mb": 91.210752, "nll": -5.9607625007629395, "elapsed_s": 667.4}170{"event": "train", "step": 3280, "loss": -6.015378475189209, "grad_norm": 0.3786124289035797, "lr": 5.3218807795017296e-05, "ex_per_s": 313.77101349449464, "data_frac": 0.021386705323336058, "mem_mb": 91.210752, "nll": -6.015378475189209, "elapsed_s": 669.1}171{"event": "train", "step": 3300, "loss": -5.804663181304932, "grad_norm": 0.2617359757423401, "lr": 5.198252543455864e-05, "ex_per_s": 314.8965695855669, "data_frac": 0.021351193940580066, "mem_mb": 91.210752, "nll": -5.804663181304932, "elapsed_s": 670.7}172{"event": "train", "step": 3320, "loss": -5.988413333892822, "grad_norm": 0.27031099796295166, "lr": 5.07771409231737e-05, "ex_per_s": 316.01735458457415, "data_frac": 0.021315907635381628, "mem_mb": 91.210752, "nll": -5.988413333892822, "elapsed_s": 672.4}173{"event": "train", "step": 3340, "loss": -5.931777477264404, "grad_norm": 0.3131471872329712, "lr": 4.960298380095221e-05, "ex_per_s": 317.13375697360584, "data_frac": 0.02128155269930546, "mem_mb": 91.210752, "nll": -5.931777477264404, "elapsed_s": 674.1}174{"event": "train", "step": 3360, "loss": -5.907993316650391, "grad_norm": 0.38611099123954773, "lr": 4.846037507072751e-05, "ex_per_s": 318.24448289834487, "data_frac": 0.02124729361253785, "mem_mb": 91.210752, "nll": -5.907993316650391, "elapsed_s": 675.8}175{"event": "train", "step": 3380, "loss": -6.123985767364502, "grad_norm": 0.5115537643432617, "lr": 4.7349627110317556e-05, "ex_per_s": 319.34793686792966, "data_frac": 0.021213456096145753, "mem_mb": 91.210752, "nll": -6.123985767364502, "elapsed_s": 677.4}176{"event": "train", "step": 3400, "loss": -5.884200096130371, "grad_norm": 0.37830492854118347, "lr": 4.6271043587123986e-05, "ex_per_s": 320.44665875163287, "data_frac": 0.021180139953845694, "mem_mb": 91.210752, "nll": -5.884200096130371, "elapsed_s": 679.1}177{"event": "train", "step": 3420, "loss": -5.8377838134765625, "grad_norm": 0.26583442091941833, "lr": 4.52249193751124e-05, "ex_per_s": 321.5445755639754, "data_frac": 0.02114689683475345, "mem_mb": 91.210752, "nll": -5.8377838134765625, "elapsed_s": 680.8}178{"event": "train", "step": 3440, "loss": -5.917542934417725, "grad_norm": 0.2095748484134674, "lr": 4.4211540474196926e-05, "ex_per_s": 322.6370825916712, "data_frac": 0.021113597138570413, "mem_mb": 91.210752, "nll": -5.917542934417725, "elapsed_s": 682.4}179{"event": "train", "step": 3460, "loss": -6.0578131675720215, "grad_norm": 0.19636762142181396, "lr": 4.3231183932050546e-05, "ex_per_s": 323.7236087957688, "data_frac": 0.02108059160535041, "mem_mb": 91.210752, "nll": -6.0578131675720215, "elapsed_s": 684.1}180{"event": "train", "step": 3480, "loss": -5.918508052825928, "grad_norm": 0.05856293812394142, "lr": 4.228411776836306e-05, "ex_per_s": 324.80566024543134, "data_frac": 0.021047687353381223, "mem_mb": 91.210752, "nll": -5.918508052825928, "elapsed_s": 685.7}181{"event": "train", "step": 3500, "loss": -6.037768363952637, "grad_norm": 0.16325785219669342, "lr": 4.137060090156723e-05, "ex_per_s": 325.8821376161718, "data_frac": 0.02101491500738016, "mem_mb": 91.210752, "nll": -6.037768363952637, "elapsed_s": 687.4}182{"event": "val", "step": 3500, "val_nll": -6.449995007517438, "best": -6.449995007517438, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 750.5}183{"event": "train", "step": 3520, "loss": -5.939347743988037, "grad_norm": 0.41785120964050293, "lr": 4.049088307805306e-05, "ex_per_s": 299.49340930701044, "data_frac": 0.019220094846414876, "mem_mb": 91.210752, "nll": -5.939347743988037, "elapsed_s": 752.2}184{"event": "train", "step": 3540, "loss": -6.136312961578369, "grad_norm": 0.6520488262176514, "lr": 3.964520480388956e-05, "ex_per_s": 300.5265712776158, "data_frac": 0.019193276974576068, "mem_mb": 91.210752, "nll": -6.136312961578369, "elapsed_s": 753.9}185{"event": "train", "step": 3560, "loss": -6.139519691467285, "grad_norm": 0.10938341170549393, "lr": 3.88337972790728e-05, "ex_per_s": 301.5546456140545, "data_frac": 0.019167189103191522, "mem_mb": 91.210752, "nll": -6.139519691467285, "elapsed_s": 755.6}186{"event": "train", "step": 3580, "loss": -5.903449058532715, "grad_norm": 0.19490131735801697, "lr": 3.805688233431824e-05, "ex_per_s": 302.57877321075296, "data_frac": 0.019141142176642034, "mem_mb": 91.210752, "nll": -5.903449058532715, "elapsed_s": 757.3}187{"event": "train", "step": 3600, "loss": -6.098569869995117, "grad_norm": 0.4200204312801361, "lr": 3.7314672370414325e-05, "ex_per_s": 303.59546571297756, "data_frac": 0.01911594110390295, "mem_mb": 91.210752, "nll": -6.098569869995117, "elapsed_s": 758.9}188{"event": "train", "step": 3620, "loss": -5.910645484924316, "grad_norm": 0.9013314247131348, "lr": 3.660737030015427e-05, "ex_per_s": 304.60998808830163, "data_frac": 0.019090080644118573, "mem_mb": 91.210752, "nll": -5.910645484924316, "elapsed_s": 760.6}189{"event": "train", "step": 3640, "loss": -6.054640769958496, "grad_norm": 0.4268227815628052, "lr": 3.5935169492861716e-05, "ex_per_s": 305.61982612999975, "data_frac": 0.019064058163735232, "mem_mb": 91.210752, "nll": -6.054640769958496, "elapsed_s": 762.3}190{"event": "train", "step": 3660, "loss": -5.832155227661133, "grad_norm": 0.5942702889442444, "lr": 3.52982537215255e-05, "ex_per_s": 306.6258692921387, "data_frac": 0.01903815294630503, "mem_mb": 91.210752, "nll": -5.832155227661133, "elapsed_s": 764.0}191{"event": "train", "step": 3680, "loss": -5.991639137268066, "grad_norm": 0.17173466086387634, "lr": 3.469679711255794e-05, "ex_per_s": 307.627549649259, "data_frac": 0.019012652156174425, "mem_mb": 91.210752, "nll": -5.991639137268066, "elapsed_s": 765.6}192{"event": "train", "step": 3700, "loss": -6.065609455108643, "grad_norm": 0.8103832602500916, "lr": 3.41309640981904e-05, "ex_per_s": 308.62568583319603, "data_frac": 0.018987176950723668, "mem_mb": 91.210752, "nll": -6.065609455108643, "elapsed_s": 767.3}193{"event": "train", "step": 3720, "loss": -5.948507308959961, "grad_norm": 0.5764367580413818, "lr": 3.3600909371519225e-05, "ex_per_s": 309.6180445824395, "data_frac": 0.018961973688096618, "mem_mb": 91.210752, "nll": -5.948507308959961, "elapsed_s": 769.0}194{"event": "train", "step": 3740, "loss": -6.1916937828063965, "grad_norm": 0.11277931183576584, "lr": 3.31067778442141e-05, "ex_per_s": 310.6076978608412, "data_frac": 0.018937059404808845, "mem_mb": 91.210752, "nll": -6.1916937828063965, "elapsed_s": 770.7}195{"event": "train", "step": 3760, "loss": -6.148220539093018, "grad_norm": 0.31492987275123596, "lr": 3.2648704606900735e-05, "ex_per_s": 311.5931913058757, "data_frac": 0.018912062179896338, "mem_mb": 91.210752, "nll": -6.148220539093018, "elapsed_s": 772.3}196{"event": "train", "step": 3780, "loss": -5.9641218185424805, "grad_norm": 0.2795770764350891, "lr": 3.222681489222838e-05, "ex_per_s": 312.5751967081952, "data_frac": 0.01888694717506007, "mem_mb": 91.210752, "nll": -5.9641218185424805, "elapsed_s": 774.0}197{"event": "train", "step": 3800, "loss": -6.099551677703857, "grad_norm": 0.13714127242565155, "lr": 3.1841224040632484e-05, "ex_per_s": 313.55304568894945, "data_frac": 0.01886222750092078, "mem_mb": 91.210752, "nll": -6.099551677703857, "elapsed_s": 775.7}198{"event": "train", "step": 3820, "loss": -5.971103191375732, "grad_norm": 0.13902659714221954, "lr": 3.149203746880175e-05, "ex_per_s": 314.5284025495882, "data_frac": 0.018837715521178567, "mem_mb": 91.210752, "nll": -5.971103191375732, "elapsed_s": 777.3}199{"event": "train", "step": 3840, "loss": -5.8744001388549805, "grad_norm": 0.6238865256309509, "lr": 3.117935064085835e-05, "ex_per_s": 315.49809598938486, "data_frac": 0.018813339097112403, "mem_mb": 91.210752, "nll": -5.8744001388549805, "elapsed_s": 779.0}200{"event": "train", "step": 3860, "loss": -6.014190196990967, "grad_norm": 0.3621987998485565, "lr": 3.090324904225885e-05, "ex_per_s": 316.4654705603062, "data_frac": 0.0187890554907152, "mem_mb": 91.210752, "nll": -6.014190196990967, "elapsed_s": 780.7}201{"event": "train", "step": 3880, "loss": -5.891732692718506, "grad_norm": 0.07259105890989304, "lr": 3.0663808156423456e-05, "ex_per_s": 317.4286382643748, "data_frac": 0.018764819707860283, "mem_mb": 91.210752, "nll": -5.891732692718506, "elapsed_s": 782.3}202{"event": "train", "step": 3900, "loss": -6.181200981140137, "grad_norm": 0.7728027105331421, "lr": 3.046109344409957e-05, "ex_per_s": 318.3880446082268, "data_frac": 0.01874076686267644, "mem_mb": 91.210752, "nll": -6.181200981140137, "elapsed_s": 784.0}203{"event": "train", "step": 3920, "loss": -6.036480903625488, "grad_norm": 0.20788206160068512, "lr": 3.0295160325465435e-05, "ex_per_s": 319.34279205454806, "data_frac": 0.018717063386036584, "mem_mb": 91.210752, "nll": -6.036480903625488, "elapsed_s": 785.7}204{"event": "train", "step": 3940, "loss": -5.919674873352051, "grad_norm": 0.2805401384830475, "lr": 3.016605416497886e-05, "ex_per_s": 320.29421136307, "data_frac": 0.018693725911554233, "mem_mb": 91.210752, "nll": -5.919674873352051, "elapsed_s": 787.3}205{"event": "train", "step": 3960, "loss": -6.0679826736450195, "grad_norm": 0.13022466003894806, "lr": 3.0073810258974985e-05, "ex_per_s": 321.2417453611444, "data_frac": 0.01867032634464306, "mem_mb": 91.210752, "nll": -6.0679826736450195, "elapsed_s": 789.0}206{"event": "train", "step": 3980, "loss": -5.805078983306885, "grad_norm": 0.1275697648525238, "lr": 3.0018453826016683e-05, "ex_per_s": 322.1849293907777, "data_frac": 0.01864662346766519, "mem_mb": 91.210752, "nll": -5.805078983306885, "elapsed_s": 790.6}207{"event": "train", "step": 4000, "loss": -6.055510520935059, "grad_norm": 0.814524233341217, "lr": 2.9999999999999997e-05, "ex_per_s": 323.1245946320312, "data_frac": 0.018623907430839328, "mem_mb": 91.210752, "nll": -6.055510520935059, "elapsed_s": 792.3}208{"event": "val", "step": 4000, "val_nll": -6.44936505710733, "best": -6.449995007517438, "bad": 1, "mem_mb": 91.210752, "elapsed_s": 855.4}209{"event": "done", "steps": 4000, "best_val": -6.449995007517438, "wall_s": 855.4139831066132, "ex_per_s": 299.2703007616066, "peak_mem_mb": 91.210752, "n_params": 3455619, "device": "mps", "data_time_frac": 0.017249019328072837, "elapsed_s": 855.4}210