SPB Git

spb/wp41_uqo

Public
51.5 KB · 210 lines
1{"event": "train", "step": 20, "loss": -5.94957971572876, "grad_norm": 0.2116481363773346, "lr": 3.149999999999999e-05, "ex_per_s": 90.88044648229604, "data_frac": 0.8696838025342071, "mem_mb": 91.579392, "nll": -5.94957971572876, "elapsed_s": 14.1}2{"event": "train", "step": 40, "loss": -5.891289234161377, "grad_norm": 0.7901540994644165, "lr": 6.149999999999999e-05, "ex_per_s": 161.36977872927685, "data_frac": 0.7747447542737731, "mem_mb": 91.579392, "nll": -5.891289234161377, "elapsed_s": 15.9}3{"event": "train", "step": 60, "loss": -6.002704620361328, "grad_norm": 0.8907306790351868, "lr": 9.149999999999999e-05, "ex_per_s": 218.10885222160107, "data_frac": 0.700857312952148, "mem_mb": 91.579392, "nll": -6.002704620361328, "elapsed_s": 17.6}4{"event": "train", "step": 80, "loss": -5.882932662963867, "grad_norm": 0.8805440068244934, "lr": 0.0001215, "ex_per_s": 265.33762922859506, "data_frac": 0.6399378515850748, "mem_mb": 91.579392, "nll": -5.882932662963867, "elapsed_s": 19.3}5{"event": "train", "step": 100, "loss": -5.9663496017456055, "grad_norm": 0.17753344774246216, "lr": 0.0001515, "ex_per_s": 305.2298698749745, "data_frac": 0.5894062520172331, "mem_mb": 91.579392, "nll": -5.9663496017456055, "elapsed_s": 21.0}6{"event": "train", "step": 120, "loss": -6.030718803405762, "grad_norm": 0.43187183141708374, "lr": 0.00018149999999999997, "ex_per_s": 339.231876507599, "data_frac": 0.5464030510828911, "mem_mb": 91.579392, "nll": -6.030718803405762, "elapsed_s": 22.7}7{"event": "train", "step": 140, "loss": -5.899853706359863, "grad_norm": 0.3915857672691345, "lr": 0.00021149999999999996, "ex_per_s": 368.546121368775, "data_frac": 0.5092896198763928, "mem_mb": 91.579392, "nll": -5.899853706359863, "elapsed_s": 24.4}8{"event": "train", "step": 160, "loss": -5.948191165924072, "grad_norm": 0.19415301084518433, "lr": 0.0002415, "ex_per_s": 394.08142411816317, "data_frac": 0.4769819599894846, "mem_mb": 91.579392, "nll": -5.948191165924072, "elapsed_s": 26.0}9{"event": "train", "step": 180, "loss": -6.087699890136719, "grad_norm": 1.0181927680969238, "lr": 0.0002715, "ex_per_s": 416.78221572829864, "data_frac": 0.4488226530306348, "mem_mb": 91.579392, "nll": -6.087699890136719, "elapsed_s": 27.7}10{"event": "train", "step": 200, "loss": -6.037732124328613, "grad_norm": 0.5448468923568726, "lr": 0.0003, "ex_per_s": 436.94219666690344, "data_frac": 0.42387287872562035, "mem_mb": 91.579392, "nll": -6.037732124328613, "elapsed_s": 29.3}11{"event": "train", "step": 220, "loss": -6.011138916015625, "grad_norm": 0.4774741232395172, "lr": 0.00029998154617398326, "ex_per_s": 454.96429050588176, "data_frac": 0.40161075437650184, "mem_mb": 91.579392, "nll": -6.011138916015625, "elapsed_s": 31.0}12{"event": "train", "step": 240, "loss": -6.04100227355957, "grad_norm": 0.6081151962280273, "lr": 0.000299926189741025, "ex_per_s": 471.1591708907435, "data_frac": 0.3816059624127743, "mem_mb": 91.579392, "nll": -6.04100227355957, "elapsed_s": 32.6}13{"event": "train", "step": 260, "loss": -5.906200885772705, "grad_norm": 0.8801228404045105, "lr": 0.0002998339458350211, "ex_per_s": 485.80679219570084, "data_frac": 0.3635514518232712, "mem_mb": 91.579392, "nll": -5.906200885772705, "elapsed_s": 34.3}14{"event": "train", "step": 280, "loss": -5.949411869049072, "grad_norm": 0.13625586032867432, "lr": 0.0002997048396745345, "ex_per_s": 499.1447164784955, "data_frac": 0.347186218540154, "mem_mb": 91.579392, "nll": -5.949411869049072, "elapsed_s": 35.9}15{"event": "train", "step": 300, "loss": -6.037998676300049, "grad_norm": 0.4929758310317993, "lr": 0.0002995389065559004, "ex_per_s": 511.2880843676444, "data_frac": 0.33223504933052933, "mem_mb": 91.579392, "nll": -6.037998676300049, "elapsed_s": 37.6}16{"event": "train", "step": 320, "loss": -6.0662946701049805, "grad_norm": 0.5182492136955261, "lr": 0.0002993361918435765, "ex_per_s": 522.3968494328429, "data_frac": 0.31854793441739876, "mem_mb": 91.579392, "nll": -6.0662946701049805, "elapsed_s": 39.2}17{"event": "train", "step": 340, "loss": -6.095026016235352, "grad_norm": 0.24814440310001373, "lr": 0.00029909675095774114, "ex_per_s": 532.6092111996279, "data_frac": 0.3059708754883804, "mem_mb": 91.579392, "nll": -6.095026016235352, "elapsed_s": 40.9}18{"event": "train", "step": 360, "loss": -5.884676456451416, "grad_norm": 1.021917700767517, "lr": 0.00029882064935914164, "ex_per_s": 542.0407065447768, "data_frac": 0.29437045234480647, "mem_mb": 91.579392, "nll": -5.884676456451416, "elapsed_s": 42.5}19{"event": "train", "step": 380, "loss": -6.035595417022705, "grad_norm": 0.12415722757577896, "lr": 0.00029850796253119824, "ex_per_s": 550.7427359662111, "data_frac": 0.2836218309142285, "mem_mb": 91.579392, "nll": -6.035595417022705, "elapsed_s": 44.2}20{"event": "train", "step": 400, "loss": -6.083630084991455, "grad_norm": 0.22890418767929077, "lr": 0.0002981587759593675, "ex_per_s": 558.7339507946855, "data_frac": 0.27362979582693875, "mem_mb": 91.579392, "nll": -6.083630084991455, "elapsed_s": 45.9}21{"event": "train", "step": 420, "loss": -6.118023872375488, "grad_norm": 0.5518267750740051, "lr": 0.0002977731851077716, "ex_per_s": 566.2474096691171, "data_frac": 0.2643566065814754, "mem_mb": 91.579392, "nll": -6.118023872375488, "elapsed_s": 47.5}22{"event": "train", "step": 440, "loss": -6.053918838500977, "grad_norm": 0.20377293229103088, "lr": 0.0002973512953930992, "ex_per_s": 573.1908089024436, "data_frac": 0.2556817308836953, "mem_mb": 91.579392, "nll": -6.053918838500977, "elapsed_s": 49.2}23{"event": "train", "step": 460, "loss": -6.029244899749756, "grad_norm": 0.3832295835018158, "lr": 0.0002968932221557859, "ex_per_s": 579.7389717273403, "data_frac": 0.24759394493178113, "mem_mb": 91.579392, "nll": -6.029244899749756, "elapsed_s": 50.8}24{"event": "train", "step": 480, "loss": -5.92629861831665, "grad_norm": 0.2188265323638916, "lr": 0.0002963990906284808, "ex_per_s": 585.8357072742411, "data_frac": 0.2400056769890098, "mem_mb": 91.579392, "nll": -5.92629861831665, "elapsed_s": 52.5}25{"event": "train", "step": 500, "loss": -5.970642566680908, "grad_norm": 0.6088058352470398, "lr": 0.00029586903590180956, "ex_per_s": 591.5450996007673, "data_frac": 0.23288233007556236, "mem_mb": 91.579392, "nll": -5.970642566680908, "elapsed_s": 54.1}26{"event": "val", "step": 500, "val_nll": -6.441875132379649, "best": -6.441875132379649, "bad": 0, "mem_mb": 91.579392, "elapsed_s": 118.9}27{"event": "train", "step": 520, "loss": -6.028384685516357, "grad_norm": 0.23169000446796417, "lr": 0.00029530320288744205, "ex_per_s": 275.59361212391076, "data_frac": 0.10443794001383615, "mem_mb": 91.579392, "nll": -6.028384685516357, "elapsed_s": 120.8}28{"event": "train", "step": 540, "loss": -5.9163713455200195, "grad_norm": 0.7237868905067444, "lr": 0.0002947017462784745, "ex_per_s": 282.288225286166, "data_frac": 0.10311956557861612, "mem_mb": 91.579392, "nll": -5.9163713455200195, "elapsed_s": 122.5}29{"event": "train", "step": 560, "loss": -6.127859115600586, "grad_norm": 0.517238974571228, "lr": 0.00029406483050713824, "ex_per_s": 288.80069385084744, "data_frac": 0.10183132176602996, "mem_mb": 91.579392, "nll": -6.127859115600586, "elapsed_s": 124.1}30{"event": "train", "step": 580, "loss": -6.000555038452148, "grad_norm": 1.1438493728637695, "lr": 0.0002933926296998457, "ex_per_s": 295.1100874040306, "data_frac": 0.10058024674223941, "mem_mb": 91.579392, "nll": -6.000555038452148, "elapsed_s": 125.8}31{"event": "train", "step": 600, "loss": -5.9979939460754395, "grad_norm": 0.5206809043884277, "lr": 0.0002926853276295856, "ex_per_s": 301.25231069178886, "data_frac": 0.09939239744873818, "mem_mb": 91.579392, "nll": -5.9979939460754395, "elapsed_s": 127.5}32{"event": "train", "step": 620, "loss": -6.043325424194336, "grad_norm": 1.0280752182006836, "lr": 0.00029194311766568174, "ex_per_s": 307.24621001200916, "data_frac": 0.09821414206471413, "mem_mb": 91.579392, "nll": -6.043325424194336, "elapsed_s": 129.2}33{"event": "train", "step": 640, "loss": -6.177517890930176, "grad_norm": 1.1026183366775513, "lr": 0.0002911662027209272, "ex_per_s": 313.10992136089675, "data_frac": 0.09705463036428016, "mem_mb": 91.579392, "nll": -6.177517890930176, "elapsed_s": 130.9}34{"event": "train", "step": 660, "loss": -6.072023868560791, "grad_norm": 0.08402319252490997, "lr": 0.00029035479519611045, "ex_per_s": 318.8219292650539, "data_frac": 0.09592327195947854, "mem_mb": 91.579392, "nll": -6.072023868560791, "elapsed_s": 132.5}35{"event": "train", "step": 680, "loss": -6.133881568908691, "grad_norm": 0.11885223537683487, "lr": 0.0002895091169219469, "ex_per_s": 324.392910592487, "data_frac": 0.09482225146642617, "mem_mb": 91.579392, "nll": -6.133881568908691, "elapsed_s": 134.2}36{"event": "train", "step": 700, "loss": -5.9270782470703125, "grad_norm": 0.21360726654529572, "lr": 0.00028862939909843273, "ex_per_s": 329.82411789239836, "data_frac": 0.09374501074828685, "mem_mb": 91.579392, "nll": -5.9270782470703125, "elapsed_s": 135.9}37{"event": "train", "step": 720, "loss": -6.013575077056885, "grad_norm": 1.1387791633605957, "lr": 0.0002877158822316369, "ex_per_s": 335.12708735805705, "data_frac": 0.09269562709131234, "mem_mb": 91.579392, "nll": -6.013575077056885, "elapsed_s": 137.5}38{"event": "train", "step": 740, "loss": -5.99160623550415, "grad_norm": 0.5805383920669556, "lr": 0.00028676881606794945, "ex_per_s": 340.29247551865706, "data_frac": 0.09166965402829383, "mem_mb": 91.579392, "nll": -5.99160623550415, "elapsed_s": 139.2}39{"event": "train", "step": 760, "loss": -6.107082843780518, "grad_norm": 0.4453311860561371, "lr": 0.000285788459525803, "ex_per_s": 345.33962221950355, "data_frac": 0.09066785642647911, "mem_mb": 91.579392, "nll": -6.107082843780518, "elapsed_s": 140.9}40{"event": "train", "step": 780, "loss": -5.834163665771484, "grad_norm": 0.13779251277446747, "lr": 0.00028477508062488754, "ex_per_s": 350.26943896076665, "data_frac": 0.08968889406384853, "mem_mb": 91.579392, "nll": -5.834163665771484, "elapsed_s": 142.6}41{"event": "train", "step": 800, "loss": -6.0819525718688965, "grad_norm": 0.27272337675094604, "lr": 0.000283728956412876, "ex_per_s": 355.0825908718909, "data_frac": 0.0887351884622766, "mem_mb": 91.579392, "nll": -6.0819525718688965, "elapsed_s": 144.2}42{"event": "train", "step": 820, "loss": -6.049243450164795, "grad_norm": 0.6756024360656738, "lr": 0.00028265037288968236, "ex_per_s": 359.7726320498096, "data_frac": 0.08780155160302723, "mem_mb": 91.579392, "nll": -6.049243450164795, "elapsed_s": 145.9}43{"event": "train", "step": 840, "loss": -6.031419277191162, "grad_norm": 0.36348414421081543, "lr": 0.0002815396249292725, "ex_per_s": 364.36642303510826, "data_frac": 0.0868892478617299, "mem_mb": 91.579392, "nll": -6.031419277191162, "elapsed_s": 147.6}44{"event": "train", "step": 860, "loss": -5.8390374183654785, "grad_norm": 0.5765975117683411, "lr": 0.00028039701619904774, "ex_per_s": 368.8627419834907, "data_frac": 0.08599669912857141, "mem_mb": 91.579392, "nll": -5.8390374183654785, "elapsed_s": 149.3}45{"event": "train", "step": 880, "loss": -5.9931440353393555, "grad_norm": 1.2543022632598877, "lr": 0.00027922285907682625, "ex_per_s": 373.25735920903304, "data_frac": 0.08512626429143746, "mem_mb": 91.579392, "nll": -5.9931440353393555, "elapsed_s": 150.9}46{"event": "train", "step": 900, "loss": -5.907224178314209, "grad_norm": 0.09142596274614334, "lr": 0.00027801747456544134, "ex_per_s": 377.5604527539621, "data_frac": 0.08427404135851606, "mem_mb": 91.579392, "nll": -5.907224178314209, "elapsed_s": 152.6}47{"event": "train", "step": 920, "loss": -6.114495277404785, "grad_norm": 0.5376896858215332, "lr": 0.00027678119220498265, "ex_per_s": 381.77151264168975, "data_frac": 0.08344187550257529, "mem_mb": 91.579392, "nll": -6.114495277404785, "elapsed_s": 154.3}48{"event": "train", "step": 940, "loss": -5.982850551605225, "grad_norm": 0.945375382900238, "lr": 0.0002755143499827035, "ex_per_s": 385.89454384465705, "data_frac": 0.08262744938737117, "mem_mb": 91.579392, "nll": -5.982850551605225, "elapsed_s": 155.9}49{"event": "train", "step": 960, "loss": -5.988513946533203, "grad_norm": 0.6454148888587952, "lr": 0.00027421729424061787, "ex_per_s": 389.9348130251058, "data_frac": 0.08183136698922193, "mem_mb": 91.579392, "nll": -5.988513946533203, "elapsed_s": 157.6}50{"event": "train", "step": 980, "loss": -5.889679908752441, "grad_norm": 0.7950974106788635, "lr": 0.00027289037958081483, "ex_per_s": 393.89569459120366, "data_frac": 0.08105236401849335, "mem_mb": 91.579392, "nll": -5.889679908752441, "elapsed_s": 159.3}51{"event": "train", "step": 1000, "loss": -5.917909145355225, "grad_norm": 0.42069244384765625, "lr": 0.0002715339687685131, "ex_per_s": 397.78058150690686, "data_frac": 0.08029150240911016, "mem_mb": 91.579392, "nll": -5.917909145355225, "elapsed_s": 160.9}52{"event": "val", "step": 1000, "val_nll": -6.446090712802322, "best": -6.446090712802322, "bad": 0, "mem_mb": 91.579392, "elapsed_s": 224.6}53{"event": "train", "step": 1020, "loss": -6.104353904724121, "grad_norm": 0.3607009947299957, "lr": 0.0002701484326328852, "ex_per_s": 288.5005079344019, "data_frac": 0.057150577567544866, "mem_mb": 91.579392, "nll": -6.104353904724121, "elapsed_s": 226.3}54{"event": "train", "step": 1040, "loss": -5.962338447570801, "grad_norm": 0.3525364100933075, "lr": 0.00026873414996567596, "ex_per_s": 291.9965546748563, "data_frac": 0.05678452830351608, "mem_mb": 91.579392, "nll": -5.962338447570801, "elapsed_s": 228.0}55{"event": "train", "step": 1060, "loss": -5.966161727905273, "grad_norm": 0.5405707955360413, "lr": 0.0002672915074176452, "ex_per_s": 295.4417182445846, "data_frac": 0.05642507330613331, "mem_mb": 91.579392, "nll": -5.966161727905273, "elapsed_s": 229.7}56{"event": "train", "step": 1080, "loss": -5.96868896484375, "grad_norm": 0.28380367159843445, "lr": 0.00026582089939286075, "ex_per_s": 298.8359460322803, "data_frac": 0.056069098882394475, "mem_mb": 91.579392, "nll": -5.96868896484375, "elapsed_s": 231.3}57{"event": "train", "step": 1100, "loss": -5.954532623291016, "grad_norm": 0.4074774384498596, "lr": 0.0002643227279408728, "ex_per_s": 302.1831356261486, "data_frac": 0.055718283370807045, "mem_mb": 91.579392, "nll": -5.954532623291016, "elapsed_s": 233.0}58{"event": "train", "step": 1120, "loss": -5.898386001586914, "grad_norm": 0.6750671863555908, "lr": 0.000262797402646797, "ex_per_s": 305.485631991928, "data_frac": 0.055372296142365635, "mem_mb": 91.579392, "nll": -5.898386001586914, "elapsed_s": 234.7}59{"event": "train", "step": 1140, "loss": -5.885149955749512, "grad_norm": 0.10479969531297684, "lr": 0.0002612453405193386, "ex_per_s": 308.73817282983754, "data_frac": 0.05503076648912269, "mem_mb": 91.579392, "nll": -5.885149955749512, "elapsed_s": 236.4}60{"event": "train", "step": 1160, "loss": -5.998284816741943, "grad_norm": 0.431646466255188, "lr": 0.0002596669658767859, "ex_per_s": 311.9476407444815, "data_frac": 0.05469677058413658, "mem_mb": 91.579392, "nll": -5.998284816741943, "elapsed_s": 238.0}61{"event": "train", "step": 1180, "loss": -6.108011722564697, "grad_norm": 0.712592601776123, "lr": 0.00025806271023100637, "ex_per_s": 315.1133623273111, "data_frac": 0.05436826258613227, "mem_mb": 91.579392, "nll": -6.108011722564697, "elapsed_s": 239.7}62{"event": "train", "step": 1200, "loss": -5.80920934677124, "grad_norm": 1.655475378036499, "lr": 0.00025643301216947504, "ex_per_s": 318.2356488090116, "data_frac": 0.05404214168566545, "mem_mb": 91.579392, "nll": -5.80920934677124, "elapsed_s": 241.4}63{"event": "train", "step": 1220, "loss": -6.011447429656982, "grad_norm": 0.22742709517478943, "lr": 0.0002547783172353695, "ex_per_s": 321.31604487550817, "data_frac": 0.05371977467163255, "mem_mb": 91.579392, "nll": -6.011447429656982, "elapsed_s": 243.0}64{"event": "train", "step": 1240, "loss": -5.960742473602295, "grad_norm": 0.1427316814661026, "lr": 0.00025309907780576277, "ex_per_s": 324.3538730199514, "data_frac": 0.05340316545811868, "mem_mb": 91.579392, "nll": -5.960742473602295, "elapsed_s": 244.7}65{"event": "train", "step": 1260, "loss": -6.0267839431762695, "grad_norm": 0.5216888189315796, "lr": 0.0002513957529679477, "ex_per_s": 327.3463115819698, "data_frac": 0.05309225111636002, "mem_mb": 91.579392, "nll": -6.0267839431762695, "elapsed_s": 246.4}66{"event": "train", "step": 1280, "loss": -6.0334577560424805, "grad_norm": 0.6216946840286255, "lr": 0.00024966880839392746, "ex_per_s": 330.3073695555652, "data_frac": 0.052784584715311535, "mem_mb": 91.579392, "nll": -6.0334577560424805, "elapsed_s": 248.1}67{"event": "train", "step": 1300, "loss": -5.958864212036133, "grad_norm": 0.15597069263458252, "lr": 0.0002479187162131052, "ex_per_s": 333.2272203579011, "data_frac": 0.052482525563997635, "mem_mb": 91.579392, "nll": -5.958864212036133, "elapsed_s": 249.7}68{"event": "train", "step": 1320, "loss": -5.930840969085693, "grad_norm": 0.6150991916656494, "lr": 0.00024614595488320833, "ex_per_s": 336.110718760124, "data_frac": 0.052184385764102725, "mem_mb": 91.579392, "nll": -5.930840969085693, "elapsed_s": 251.4}69{"event": "train", "step": 1340, "loss": -5.924959182739258, "grad_norm": 1.0264945030212402, "lr": 0.00024435100905948387, "ex_per_s": 338.9555158237355, "data_frac": 0.05189030084046634, "mem_mb": 91.579392, "nll": -5.924959182739258, "elapsed_s": 253.1}70{"event": "train", "step": 1360, "loss": -5.9779276847839355, "grad_norm": 0.3283599019050598, "lr": 0.0002425343694621974, "ex_per_s": 341.76399659058904, "data_frac": 0.051599028486210216, "mem_mb": 91.579392, "nll": -5.9779276847839355, "elapsed_s": 254.7}71{"event": "train", "step": 1380, "loss": -6.056778430938721, "grad_norm": 0.20625431835651398, "lr": 0.00024069653274247583, "ex_per_s": 344.5365405164008, "data_frac": 0.05130991791657011, "mem_mb": 91.579392, "nll": -6.056778430938721, "elapsed_s": 256.4}72{"event": "train", "step": 1400, "loss": -5.888830184936523, "grad_norm": 0.11533966660499573, "lr": 0.0002388380013465276, "ex_per_s": 347.26917569012267, "data_frac": 0.051026465548784115, "mem_mb": 91.579392, "nll": -5.888830184936523, "elapsed_s": 258.1}73{"event": "train", "step": 1420, "loss": -5.8813395500183105, "grad_norm": 0.595119059085846, "lr": 0.00023695928337827914, "ex_per_s": 349.96831975127554, "data_frac": 0.050748164044720705, "mem_mb": 91.579392, "nll": -5.8813395500183105, "elapsed_s": 259.7}74{"event": "train", "step": 1440, "loss": -6.007033824920654, "grad_norm": 0.0798485279083252, "lr": 0.00023506089246046403, "ex_per_s": 352.6356616893761, "data_frac": 0.050471193935893235, "mem_mb": 91.579392, "nll": -6.007033824920654, "elapsed_s": 261.4}75{"event": "train", "step": 1460, "loss": -5.985134124755859, "grad_norm": 0.10572867840528488, "lr": 0.00023314334759420388, "ex_per_s": 355.2700662738849, "data_frac": 0.050198178513809484, "mem_mb": 91.579392, "nll": -5.985134124755859, "elapsed_s": 263.1}76{"event": "train", "step": 1480, "loss": -5.984589099884033, "grad_norm": 0.42784300446510315, "lr": 0.0002312071730171186, "ex_per_s": 357.86760618322035, "data_frac": 0.0499284489938446, "mem_mb": 91.579392, "nll": -5.984589099884033, "elapsed_s": 264.7}77{"event": "train", "step": 1500, "loss": -5.9649810791015625, "grad_norm": 0.12090575695037842, "lr": 0.0002292528980600049, "ex_per_s": 360.423999099571, "data_frac": 0.04965957781626294, "mem_mb": 91.579392, "nll": -5.9649810791015625, "elapsed_s": 266.4}78{"event": "val", "step": 1500, "val_nll": -6.447718861080702, "best": -6.447718861080702, "bad": 0, "mem_mb": 91.579392, "elapsed_s": 329.9}79{"event": "train", "step": 1520, "loss": -5.841037273406982, "grad_norm": 1.4327553510665894, "lr": 0.00022728105700212285, "ex_per_s": 293.3715671075258, "data_frac": 0.039927125638661694, "mem_mb": 91.579392, "nll": -5.841037273406982, "elapsed_s": 331.6}80{"event": "train", "step": 1540, "loss": -5.9852142333984375, "grad_norm": 0.18406890332698822, "lr": 0.00022529218892512886, "ex_per_s": 295.7402399794592, "data_frac": 0.03976337736181234, "mem_mb": 91.579392, "nll": -5.9852142333984375, "elapsed_s": 333.3}81{"event": "train", "step": 1560, "loss": -6.165699005126953, "grad_norm": 0.584041953086853, "lr": 0.00022328683756569652, "ex_per_s": 298.0845277555608, "data_frac": 0.039600628090388466, "mem_mb": 91.579392, "nll": -6.165699005126953, "elapsed_s": 335.0}82{"event": "train", "step": 1580, "loss": -5.9922661781311035, "grad_norm": 0.16625213623046875, "lr": 0.0002212655511668642, "ex_per_s": 300.4049038857167, "data_frac": 0.039439859652338385, "mem_mb": 91.579392, "nll": -5.9922661781311035, "elapsed_s": 336.7}83{"event": "train", "step": 1600, "loss": -6.089993476867676, "grad_norm": 0.26647377014160156, "lr": 0.00021922888232815089, "ex_per_s": 302.70269599772456, "data_frac": 0.03928085802759383, "mem_mb": 91.579392, "nll": -6.089993476867676, "elapsed_s": 338.3}84{"event": "train", "step": 1620, "loss": -5.900341510772705, "grad_norm": 0.36864301562309265, "lr": 0.00021717738785448102, "ex_per_s": 304.97778201596634, "data_frac": 0.039123764920477574, "mem_mb": 91.579392, "nll": -5.900341510772705, "elapsed_s": 340.0}85{"event": "train", "step": 1640, "loss": -5.871169090270996, "grad_norm": 0.4906729459762573, "lr": 0.00021511162860395977, "ex_per_s": 307.23098578513276, "data_frac": 0.03896810118906225, "mem_mb": 91.579392, "nll": -5.871169090270996, "elapsed_s": 341.7}86{"event": "train", "step": 1660, "loss": -6.159986972808838, "grad_norm": 0.7234247922897339, "lr": 0.0002130321693345397, "ex_per_s": 309.461407665998, "data_frac": 0.03881387319763168, "mem_mb": 91.579392, "nll": -6.159986972808838, "elapsed_s": 343.4}87{"event": "train", "step": 1680, "loss": -6.123255729675293, "grad_norm": 0.15443100035190582, "lr": 0.0002109395785496218, "ex_per_s": 311.670577972437, "data_frac": 0.03866158396704662, "mem_mb": 91.579392, "nll": -6.123255729675293, "elapsed_s": 345.0}88{"event": "train", "step": 1700, "loss": -5.88673210144043, "grad_norm": 0.11225607246160507, "lr": 0.00020883442834263225, "ex_per_s": 313.85772594254263, "data_frac": 0.03851134590386987, "mem_mb": 91.579392, "nll": -5.88673210144043, "elapsed_s": 346.7}89{"event": "train", "step": 1720, "loss": -5.983457088470459, "grad_norm": 0.11715499311685562, "lr": 0.00020671729424061788, "ex_per_s": 316.0255942457668, "data_frac": 0.038363682554233614, "mem_mb": 91.579392, "nll": -5.983457088470459, "elapsed_s": 348.4}90{"event": "train", "step": 1740, "loss": -5.83555793762207, "grad_norm": 0.34796759486198425, "lr": 0.00020458875504690275, "ex_per_s": 318.1709196735659, "data_frac": 0.03821577362970864, "mem_mb": 91.579392, "nll": -5.83555793762207, "elapsed_s": 350.0}91{"event": "train", "step": 1760, "loss": -5.88642692565918, "grad_norm": 0.19127467274665833, "lr": 0.000202449392682849, "ex_per_s": 320.3001522395606, "data_frac": 0.03806886159678799, "mem_mb": 91.579392, "nll": -5.88642692565918, "elapsed_s": 351.7}92{"event": "train", "step": 1780, "loss": -6.07855224609375, "grad_norm": 0.3666149377822876, "lr": 0.00020029979202876563, "ex_per_s": 322.4097458892502, "data_frac": 0.03792452765284563, "mem_mb": 91.579392, "nll": -6.07855224609375, "elapsed_s": 353.4}93{"event": "train", "step": 1800, "loss": -5.988064289093018, "grad_norm": 0.23937836289405823, "lr": 0.00019814054076400787, "ex_per_s": 324.4966220308453, "data_frac": 0.03778108444699318, "mem_mb": 91.579392, "nll": -5.988064289093018, "elapsed_s": 355.1}94{"event": "train", "step": 1820, "loss": -5.886514186859131, "grad_norm": 0.7281148433685303, "lr": 0.00019597222920631147, "ex_per_s": 326.5669145501589, "data_frac": 0.03763859885461962, "mem_mb": 91.579392, "nll": -5.886514186859131, "elapsed_s": 356.7}95{"event": "train", "step": 1840, "loss": -5.862462520599365, "grad_norm": 0.18867474794387817, "lr": 0.0001937954501504058, "ex_per_s": 328.6207491884657, "data_frac": 0.037496971200398355, "mem_mb": 91.579392, "nll": -5.862462520599365, "elapsed_s": 358.4}96{"event": "train", "step": 1860, "loss": -5.895153522491455, "grad_norm": 0.2523593306541443, "lr": 0.0001916107987059498, "ex_per_s": 330.6551802845985, "data_frac": 0.03735834786797847, "mem_mb": 91.579392, "nll": -5.895153522491455, "elapsed_s": 360.1}97{"event": "train", "step": 1880, "loss": -5.9668965339660645, "grad_norm": 0.3193817436695099, "lr": 0.00018941887213483484, "ex_per_s": 332.6735880351311, "data_frac": 0.037219649853557905, "mem_mb": 91.579392, "nll": -5.9668965339660645, "elapsed_s": 361.7}98{"event": "train", "step": 1900, "loss": -5.935783863067627, "grad_norm": 0.2168026715517044, "lr": 0.00018722026968789906, "ex_per_s": 334.6732935615021, "data_frac": 0.037082911721745285, "mem_mb": 91.579392, "nll": -5.935783863067627, "elapsed_s": 363.4}99{"event": "train", "step": 1920, "loss": -5.876211643218994, "grad_norm": 0.09452507644891739, "lr": 0.00018501559244109864, "ex_per_s": 336.6549066194252, "data_frac": 0.036948114883067026, "mem_mb": 91.579392, "nll": -5.876211643218994, "elapsed_s": 365.0}100{"event": "train", "step": 1940, "loss": -6.085047721862793, "grad_norm": 0.481357604265213, "lr": 0.00018280544313117932, "ex_per_s": 338.6171259823768, "data_frac": 0.03681472418164728, "mem_mb": 91.579392, "nll": -6.085047721862793, "elapsed_s": 366.7}101{"event": "train", "step": 1960, "loss": -6.073001384735107, "grad_norm": 0.22772958874702454, "lr": 0.00018059042599089415, "ex_per_s": 340.56326027841516, "data_frac": 0.03668369954973749, "mem_mb": 91.579392, "nll": -6.073001384735107, "elapsed_s": 368.4}102{"event": "train", "step": 1980, "loss": -5.975799560546875, "grad_norm": 0.4277324676513672, "lr": 0.00017837114658381247, "ex_per_s": 342.49100670994903, "data_frac": 0.03655295093481028, "mem_mb": 91.579392, "nll": -5.975799560546875, "elapsed_s": 370.0}103{"event": "train", "step": 2000, "loss": -6.096059322357178, "grad_norm": 0.38019195199012756, "lr": 0.00017614821163876486, "ex_per_s": 344.40328648871116, "data_frac": 0.03642256631192814, "mem_mb": 91.579392, "nll": -6.096059322357178, "elapsed_s": 371.7}104{"event": "val", "step": 2000, "val_nll": -6.449384604102312, "best": -6.449384604102312, "bad": 0, "mem_mb": 91.579392, "elapsed_s": 435.2}105{"event": "train", "step": 2020, "loss": -5.974090099334717, "grad_norm": 0.3281496465206146, "lr": 0.00017392222888396941, "ex_per_s": 295.91810960922385, "data_frac": 0.031013583151158937, "mem_mb": 91.579392, "nll": -5.974090099334717, "elapsed_s": 436.9}106{"event": "train", "step": 2040, "loss": -5.930271148681641, "grad_norm": 0.8239365220069885, "lr": 0.000171693806880885, "ex_per_s": 297.70791931531807, "data_frac": 0.030922746381147267, "mem_mb": 91.579392, "nll": -5.930271148681641, "elapsed_s": 438.6}107{"event": "train", "step": 2060, "loss": -5.808361530303955, "grad_norm": 0.8019617795944214, "lr": 0.0001694635548578366, "ex_per_s": 299.48124253848124, "data_frac": 0.03083318822106444, "mem_mb": 91.579392, "nll": -5.808361530303955, "elapsed_s": 440.3}108{"event": "train", "step": 2080, "loss": -5.969672203063965, "grad_norm": 0.12784934043884277, "lr": 0.0001672320825434584, "ex_per_s": 301.2444845018163, "data_frac": 0.030743793240677823, "mem_mb": 91.579392, "nll": -5.969672203063965, "elapsed_s": 441.9}109{"event": "train", "step": 2100, "loss": -6.069770812988281, "grad_norm": 0.23664960265159607, "lr": 0.000165, "ex_per_s": 302.9929495139307, "data_frac": 0.030655452894779254, "mem_mb": 91.579392, "nll": -6.069770812988281, "elapsed_s": 443.6}110{"event": "train", "step": 2120, "loss": -5.894711017608643, "grad_norm": 0.07654909789562225, "lr": 0.0001627679174565416, "ex_per_s": 304.7286345003758, "data_frac": 0.030567284104070936, "mem_mb": 91.579392, "nll": -5.894711017608643, "elapsed_s": 445.3}111{"event": "train", "step": 2140, "loss": -5.945315837860107, "grad_norm": 0.20338764786720276, "lr": 0.00016053644514216336, "ex_per_s": 306.45051737995396, "data_frac": 0.03048114187853767, "mem_mb": 91.579392, "nll": -5.945315837860107, "elapsed_s": 447.0}112{"event": "train", "step": 2160, "loss": -5.786678314208984, "grad_norm": 0.45357638597488403, "lr": 0.00015830619311911498, "ex_per_s": 308.16239473005146, "data_frac": 0.030394791870938912, "mem_mb": 91.579392, "nll": -5.786678314208984, "elapsed_s": 448.6}113{"event": "train", "step": 2180, "loss": -6.002985000610352, "grad_norm": 0.5127912759780884, "lr": 0.00015607777111603058, "ex_per_s": 309.86176238318325, "data_frac": 0.03030848874886051, "mem_mb": 91.579392, "nll": -6.002985000610352, "elapsed_s": 450.3}114{"event": "train", "step": 2200, "loss": -5.9732818603515625, "grad_norm": 0.4110260307788849, "lr": 0.00015385178836123514, "ex_per_s": 311.5498429375368, "data_frac": 0.030223095278919695, "mem_mb": 91.579392, "nll": -5.9732818603515625, "elapsed_s": 452.0}115{"event": "train", "step": 2220, "loss": -6.015933990478516, "grad_norm": 0.234901562333107, "lr": 0.00015162885341618752, "ex_per_s": 313.2263498093277, "data_frac": 0.03013892359849348, "mem_mb": 91.579392, "nll": -6.015933990478516, "elapsed_s": 453.6}116{"event": "train", "step": 2240, "loss": -5.959944248199463, "grad_norm": 0.6605212688446045, "lr": 0.00014940957400910585, "ex_per_s": 314.8908566233601, "data_frac": 0.030055238817670398, "mem_mb": 91.579392, "nll": -5.959944248199463, "elapsed_s": 455.3}117{"event": "train", "step": 2260, "loss": -6.009171009063721, "grad_norm": 0.28705543279647827, "lr": 0.00014719455686882068, "ex_per_s": 316.54391968821693, "data_frac": 0.029972050154257576, "mem_mb": 91.579392, "nll": -6.009171009063721, "elapsed_s": 457.0}118{"event": "train", "step": 2280, "loss": -6.0487847328186035, "grad_norm": 0.20994557440280914, "lr": 0.0001449844075589013, "ex_per_s": 318.1851312971936, "data_frac": 0.029889412378339652, "mem_mb": 91.579392, "nll": -6.0487847328186035, "elapsed_s": 458.6}119{"event": "train", "step": 2300, "loss": -5.9552202224731445, "grad_norm": 0.6435462832450867, "lr": 0.0001427797303121009, "ex_per_s": 319.81553127871683, "data_frac": 0.02980759728860534, "mem_mb": 91.579392, "nll": -5.9552202224731445, "elapsed_s": 460.3}120{"event": "train", "step": 2320, "loss": -6.015156269073486, "grad_norm": 0.5915181040763855, "lr": 0.00014058112786516518, "ex_per_s": 321.43391128950674, "data_frac": 0.029726450466272324, "mem_mb": 91.579392, "nll": -6.015156269073486, "elapsed_s": 462.0}121{"event": "train", "step": 2340, "loss": -5.951956748962402, "grad_norm": 0.8240927457809448, "lr": 0.00013838920129405016, "ex_per_s": 323.0405411140692, "data_frac": 0.0296458465094408, "mem_mb": 91.579392, "nll": -5.951956748962402, "elapsed_s": 463.6}122{"event": "train", "step": 2360, "loss": -5.918355464935303, "grad_norm": 0.6677366495132446, "lr": 0.0001362045498495942, "ex_per_s": 324.6353485443261, "data_frac": 0.02956544407829887, "mem_mb": 91.579392, "nll": -5.918355464935303, "elapsed_s": 465.3}123{"event": "train", "step": 2380, "loss": -6.0002264976501465, "grad_norm": 0.22691325843334198, "lr": 0.00013402777079368853, "ex_per_s": 326.21855927278517, "data_frac": 0.02948594099938377, "mem_mb": 91.579392, "nll": -6.0002264976501465, "elapsed_s": 467.0}124{"event": "train", "step": 2400, "loss": -5.9064154624938965, "grad_norm": 0.26540473103523254, "lr": 0.00013185945923599213, "ex_per_s": 327.79177444755123, "data_frac": 0.029408384659030323, "mem_mb": 91.579392, "nll": -5.9064154624938965, "elapsed_s": 468.6}125{"event": "train", "step": 2420, "loss": -5.9514241218566895, "grad_norm": 0.20363137125968933, "lr": 0.00012970020797123432, "ex_per_s": 329.3535552977805, "data_frac": 0.029330705450656466, "mem_mb": 91.579392, "nll": -5.9514241218566895, "elapsed_s": 470.3}126{"event": "train", "step": 2440, "loss": -5.933928966522217, "grad_norm": 0.18642286956310272, "lr": 0.000127550607317151, "ex_per_s": 330.9076330909229, "data_frac": 0.029253564434401073, "mem_mb": 91.579392, "nll": -5.933928966522217, "elapsed_s": 472.0}127{"event": "train", "step": 2460, "loss": -6.097858428955078, "grad_norm": 0.23755963146686554, "lr": 0.00012541124495309727, "ex_per_s": 332.45133150202184, "data_frac": 0.029177125432707816, "mem_mb": 91.579392, "nll": -6.097858428955078, "elapsed_s": 473.6}128{"event": "train", "step": 2480, "loss": -5.786258697509766, "grad_norm": 0.736402153968811, "lr": 0.00012328270575938212, "ex_per_s": 333.98355776780403, "data_frac": 0.02910091622432516, "mem_mb": 91.579392, "nll": -5.786258697509766, "elapsed_s": 475.3}129{"event": "train", "step": 2500, "loss": -5.94600772857666, "grad_norm": 0.08854123204946518, "lr": 0.00012116557165736775, "ex_per_s": 335.5062501058005, "data_frac": 0.02902523700635689, "mem_mb": 91.579392, "nll": -5.94600772857666, "elapsed_s": 476.9}130{"event": "val", "step": 2500, "val_nll": -6.444557715633881, "best": -6.449384604102312, "bad": 1, "mem_mb": 91.579392, "elapsed_s": 540.1}131{"event": "train", "step": 2520, "loss": -5.996769905090332, "grad_norm": 0.08772371709346771, "lr": 0.00011906042145037817, "ex_per_s": 297.65542387079057, "data_frac": 0.02557036790740021, "mem_mb": 91.579392, "nll": -5.996769905090332, "elapsed_s": 541.9}132{"event": "train", "step": 2540, "loss": -6.120438575744629, "grad_norm": 0.26465991139411926, "lr": 0.00011696783066546026, "ex_per_s": 299.0893413807932, "data_frac": 0.02551500069127003, "mem_mb": 91.579392, "nll": -6.120438575744629, "elapsed_s": 543.6}133{"event": "train", "step": 2560, "loss": -6.178769588470459, "grad_norm": 1.0875829458236694, "lr": 0.00011488837139604017, "ex_per_s": 300.51098134129114, "data_frac": 0.025459014974823734, "mem_mb": 91.579392, "nll": -6.178769588470459, "elapsed_s": 545.2}134{"event": "train", "step": 2580, "loss": -5.760161876678467, "grad_norm": 0.8877089619636536, "lr": 0.00011282261214551898, "ex_per_s": 301.92333448643814, "data_frac": 0.02540343966280627, "mem_mb": 91.579392, "nll": -5.760161876678467, "elapsed_s": 546.9}135{"event": "train", "step": 2600, "loss": -6.126797199249268, "grad_norm": 0.11941265314817429, "lr": 0.00011077111767184915, "ex_per_s": 303.3258010783874, "data_frac": 0.025349191983870967, "mem_mb": 91.579392, "nll": -6.126797199249268, "elapsed_s": 548.6}136{"event": "train", "step": 2620, "loss": -6.117669582366943, "grad_norm": 0.07498589903116226, "lr": 0.00010873444883313583, "ex_per_s": 304.72253902323195, "data_frac": 0.02529510190233481, "mem_mb": 91.579392, "nll": -6.117669582366943, "elapsed_s": 550.3}137{"event": "train", "step": 2640, "loss": -6.0054168701171875, "grad_norm": 0.30057474970817566, "lr": 0.00010671316243430345, "ex_per_s": 306.11055814517346, "data_frac": 0.025239105428781237, "mem_mb": 91.579392, "nll": -6.0054168701171875, "elapsed_s": 552.0}138{"event": "train", "step": 2660, "loss": -6.189077377319336, "grad_norm": 0.185459703207016, "lr": 0.00010470781107487112, "ex_per_s": 307.4926743790462, "data_frac": 0.025182837488633043, "mem_mb": 91.579392, "nll": -6.189077377319336, "elapsed_s": 553.7}139{"event": "train", "step": 2680, "loss": -5.910770893096924, "grad_norm": 0.1474732607603073, "lr": 0.00010271894299787715, "ex_per_s": 308.8658377722443, "data_frac": 0.02512768737505441, "mem_mb": 91.579392, "nll": -5.910770893096924, "elapsed_s": 555.4}140{"event": "train", "step": 2700, "loss": -5.986726760864258, "grad_norm": 0.8472393155097961, "lr": 0.00010074710193999506, "ex_per_s": 310.23098500502044, "data_frac": 0.025072552489746665, "mem_mb": 91.579392, "nll": -5.986726760864258, "elapsed_s": 557.0}141{"event": "train", "step": 2720, "loss": -5.847282886505127, "grad_norm": 0.115810327231884, "lr": 9.879282698288144e-05, "ex_per_s": 311.5879809010045, "data_frac": 0.025018097525345614, "mem_mb": 91.579392, "nll": -5.847282886505127, "elapsed_s": 558.7}142{"event": "train", "step": 2740, "loss": -5.959972381591797, "grad_norm": 0.6234204173088074, "lr": 9.685665240579613e-05, "ex_per_s": 312.9342413090549, "data_frac": 0.024965092596728928, "mem_mb": 91.579392, "nll": -5.959972381591797, "elapsed_s": 560.4}143{"event": "train", "step": 2760, "loss": -5.9507832527160645, "grad_norm": 0.48219799995422363, "lr": 9.4939107539536e-05, "ex_per_s": 314.2748348441486, "data_frac": 0.02491101704595059, "mem_mb": 91.579392, "nll": -5.9507832527160645, "elapsed_s": 562.1}144{"event": "train", "step": 2780, "loss": -5.978977203369141, "grad_norm": 0.32053154706954956, "lr": 9.304071662172087e-05, "ex_per_s": 315.6035501169833, "data_frac": 0.024857450422091715, "mem_mb": 91.579392, "nll": -5.978977203369141, "elapsed_s": 563.8}145{"event": "train", "step": 2800, "loss": -6.174136638641357, "grad_norm": 0.7942374348640442, "lr": 9.116199865347241e-05, "ex_per_s": 316.92842935627016, "data_frac": 0.02480429612707411, "mem_mb": 91.579392, "nll": -6.174136638641357, "elapsed_s": 565.5}146{"event": "train", "step": 2820, "loss": -6.031755447387695, "grad_norm": 0.5387247204780579, "lr": 8.930346725752419e-05, "ex_per_s": 318.24915528428363, "data_frac": 0.024751000841476314, "mem_mb": 91.579392, "nll": -6.031755447387695, "elapsed_s": 567.1}147{"event": "train", "step": 2840, "loss": -5.876088619232178, "grad_norm": 0.15348032116889954, "lr": 8.746563053780262e-05, "ex_per_s": 319.56436858124357, "data_frac": 0.024699016886071824, "mem_mb": 91.579392, "nll": -5.876088619232178, "elapsed_s": 568.8}148{"event": "train", "step": 2860, "loss": -6.2606940269470215, "grad_norm": 0.44924652576446533, "lr": 8.564899094051614e-05, "ex_per_s": 320.87737692908235, "data_frac": 0.024646248473249754, "mem_mb": 91.579392, "nll": -6.2606940269470215, "elapsed_s": 570.5}149{"event": "train", "step": 2880, "loss": -6.003882884979248, "grad_norm": 0.2632979452610016, "lr": 8.385404511679161e-05, "ex_per_s": 322.18282948679683, "data_frac": 0.024593456995932437, "mem_mb": 91.579392, "nll": -6.003882884979248, "elapsed_s": 572.1}150{"event": "train", "step": 2900, "loss": -6.011064529418945, "grad_norm": 0.23966528475284576, "lr": 8.208128378689484e-05, "ex_per_s": 323.4801394915802, "data_frac": 0.02454091842485137, "mem_mb": 91.579392, "nll": -6.011064529418945, "elapsed_s": 573.8}151{"event": "train", "step": 2920, "loss": -5.795802593231201, "grad_norm": 0.7084763646125793, "lr": 8.03311916060725e-05, "ex_per_s": 324.7705495951174, "data_frac": 0.024489030507368018, "mem_mb": 91.579392, "nll": -5.795802593231201, "elapsed_s": 575.5}152{"event": "train", "step": 2940, "loss": -6.062953948974609, "grad_norm": 0.33902981877326965, "lr": 7.860424703205225e-05, "ex_per_s": 326.0539373594654, "data_frac": 0.0244370613145366, "mem_mb": 91.579392, "nll": -6.062953948974609, "elapsed_s": 577.1}153{"event": "train", "step": 2960, "loss": -6.048628330230713, "grad_norm": 0.15913508832454681, "lr": 7.690092219423722e-05, "ex_per_s": 327.3285257794235, "data_frac": 0.02438588505196172, "mem_mb": 91.579392, "nll": -6.048628330230713, "elapsed_s": 578.8}154{"event": "train", "step": 2980, "loss": -6.041815757751465, "grad_norm": 0.14662770926952362, "lr": 7.522168276463048e-05, "ex_per_s": 328.59667803946246, "data_frac": 0.02433454562094783, "mem_mb": 91.579392, "nll": -6.041815757751465, "elapsed_s": 580.4}155{"event": "train", "step": 3000, "loss": -6.141651630401611, "grad_norm": 0.19702623784542084, "lr": 7.356698783052496e-05, "ex_per_s": 329.85641425735975, "data_frac": 0.024283690142758392, "mem_mb": 91.579392, "nll": -6.141651630401611, "elapsed_s": 582.1}156{"event": "val", "step": 3000, "val_nll": -6.45053233793159, "best": -6.45053233793159, "bad": 0, "mem_mb": 91.579392, "elapsed_s": 646.9}157{"event": "train", "step": 3020, "loss": -5.919569969177246, "grad_norm": 0.33395522832870483, "lr": 7.193728976899362e-05, "ex_per_s": 297.9222189327516, "data_frac": 0.021807617334618362, "mem_mb": 91.579392, "nll": -5.919569969177246, "elapsed_s": 648.8}158{"event": "train", "step": 3040, "loss": -5.892619609832764, "grad_norm": 0.5345608592033386, "lr": 7.033303412321404e-05, "ex_per_s": 299.1193431502245, "data_frac": 0.021768772148071963, "mem_mb": 91.579392, "nll": -5.892619609832764, "elapsed_s": 650.5}159{"event": "train", "step": 3060, "loss": -5.83868408203125, "grad_norm": 0.22688840329647064, "lr": 6.875465948066139e-05, "ex_per_s": 300.309862987775, "data_frac": 0.021730185081399643, "mem_mb": 91.579392, "nll": -5.83868408203125, "elapsed_s": 652.2}160{"event": "train", "step": 3080, "loss": -5.808419704437256, "grad_norm": 0.6835206747055054, "lr": 6.720259735320299e-05, "ex_per_s": 301.50073567815446, "data_frac": 0.02169320414940405, "mem_mb": 91.579392, "nll": -5.808419704437256, "elapsed_s": 653.8}161{"event": "train", "step": 3100, "loss": -5.965798854827881, "grad_norm": 0.0712767019867897, "lr": 6.567727205912722e-05, "ex_per_s": 302.681375453175, "data_frac": 0.021658091904100372, "mem_mb": 91.579392, "nll": -5.965798854827881, "elapsed_s": 655.5}162{"event": "train", "step": 3120, "loss": -5.990973949432373, "grad_norm": 0.37940165400505066, "lr": 6.417910060713926e-05, "ex_per_s": 303.86065605417986, "data_frac": 0.02162153746132315, "mem_mb": 91.579392, "nll": -5.990973949432373, "elapsed_s": 657.2}163{"event": "train", "step": 3140, "loss": -5.90919828414917, "grad_norm": 0.4606643319129944, "lr": 6.270849258235485e-05, "ex_per_s": 305.0352958898855, "data_frac": 0.021585676857011283, "mem_mb": 91.579392, "nll": -5.90919828414917, "elapsed_s": 658.9}164{"event": "train", "step": 3160, "loss": -6.038841247558594, "grad_norm": 0.2711588442325592, "lr": 6.126585003432406e-05, "ex_per_s": 306.2047503119815, "data_frac": 0.021549329728874193, "mem_mb": 91.579392, "nll": -6.038841247558594, "elapsed_s": 660.5}165{"event": "train", "step": 3180, "loss": -5.968332290649414, "grad_norm": 0.1674196720123291, "lr": 5.985156736711483e-05, "ex_per_s": 307.36720132400876, "data_frac": 0.021513260455435036, "mem_mb": 91.579392, "nll": -5.968332290649414, "elapsed_s": 662.2}166{"event": "train", "step": 3200, "loss": -5.949464797973633, "grad_norm": 0.510462760925293, "lr": 5.846603123148688e-05, "ex_per_s": 308.5237328958438, "data_frac": 0.021478043927054272, "mem_mb": 91.579392, "nll": -5.949464797973633, "elapsed_s": 663.8}167{"event": "train", "step": 3220, "loss": -5.848260879516602, "grad_norm": 0.4123486578464508, "lr": 5.710962041918516e-05, "ex_per_s": 309.6733526694042, "data_frac": 0.02144345478435524, "mem_mb": 91.579392, "nll": -5.848260879516602, "elapsed_s": 665.5}168{"event": "train", "step": 3240, "loss": -6.002839088439941, "grad_norm": 0.07316333800554276, "lr": 5.578270575938211e-05, "ex_per_s": 310.81770549064163, "data_frac": 0.021407442532866878, "mem_mb": 91.579392, "nll": -6.002839088439941, "elapsed_s": 667.2}169{"event": "train", "step": 3260, "loss": -6.10504150390625, "grad_norm": 0.7397664785385132, "lr": 5.448565001729653e-05, "ex_per_s": 311.951634057562, "data_frac": 0.021373651326192383, "mem_mb": 91.579392, "nll": -6.10504150390625, "elapsed_s": 668.9}170{"event": "train", "step": 3280, "loss": -5.857390403747559, "grad_norm": 0.11160846054553986, "lr": 5.3218807795017296e-05, "ex_per_s": 313.08416414437505, "data_frac": 0.02133821255763879, "mem_mb": 91.579392, "nll": -5.857390403747559, "elapsed_s": 670.5}171{"event": "train", "step": 3300, "loss": -6.00472354888916, "grad_norm": 0.1651795506477356, "lr": 5.198252543455864e-05, "ex_per_s": 314.21096665687037, "data_frac": 0.021302881767617094, "mem_mb": 91.579392, "nll": -6.00472354888916, "elapsed_s": 672.2}172{"event": "train", "step": 3320, "loss": -6.079982280731201, "grad_norm": 0.4569019675254822, "lr": 5.07771409231737e-05, "ex_per_s": 315.3314318026371, "data_frac": 0.0212670856492507, "mem_mb": 91.579392, "nll": -6.079982280731201, "elapsed_s": 673.9}173{"event": "train", "step": 3340, "loss": -5.925394058227539, "grad_norm": 0.17250800132751465, "lr": 4.960298380095221e-05, "ex_per_s": 316.44204701292523, "data_frac": 0.02123172470472135, "mem_mb": 91.579392, "nll": -5.925394058227539, "elapsed_s": 675.6}174{"event": "train", "step": 3360, "loss": -5.846305847167969, "grad_norm": 0.9420239925384521, "lr": 4.846037507072751e-05, "ex_per_s": 317.54972075912997, "data_frac": 0.02119711277391708, "mem_mb": 91.579392, "nll": -5.846305847167969, "elapsed_s": 677.2}175{"event": "train", "step": 3380, "loss": -5.905497074127197, "grad_norm": 0.22143355011940002, "lr": 4.7349627110317556e-05, "ex_per_s": 318.65380082544897, "data_frac": 0.021161135350522116, "mem_mb": 91.579392, "nll": -5.905497074127197, "elapsed_s": 678.9}176{"event": "train", "step": 3400, "loss": -6.154113292694092, "grad_norm": 0.46952563524246216, "lr": 4.6271043587123986e-05, "ex_per_s": 319.75202067894224, "data_frac": 0.0211265783204292, "mem_mb": 91.579392, "nll": -6.154113292694092, "elapsed_s": 680.6}177{"event": "train", "step": 3420, "loss": -5.970956802368164, "grad_norm": 0.2527228593826294, "lr": 4.52249193751124e-05, "ex_per_s": 320.84318959712033, "data_frac": 0.0210916824826824, "mem_mb": 91.579392, "nll": -5.970956802368164, "elapsed_s": 682.2}178{"event": "train", "step": 3440, "loss": -5.996883869171143, "grad_norm": 0.47609278559684753, "lr": 4.4211540474196926e-05, "ex_per_s": 321.93182681236857, "data_frac": 0.021057300250075465, "mem_mb": 91.579392, "nll": -5.996883869171143, "elapsed_s": 683.9}179{"event": "train", "step": 3460, "loss": -5.946842193603516, "grad_norm": 0.1551828682422638, "lr": 4.3231183932050546e-05, "ex_per_s": 323.01247367561865, "data_frac": 0.02102362021973407, "mem_mb": 91.579392, "nll": -5.946842193603516, "elapsed_s": 685.6}180{"event": "train", "step": 3480, "loss": -6.170212268829346, "grad_norm": 0.12233702093362808, "lr": 4.228411776836306e-05, "ex_per_s": 324.0884767284519, "data_frac": 0.020990244637151486, "mem_mb": 91.579392, "nll": -6.170212268829346, "elapsed_s": 687.3}181{"event": "train", "step": 3500, "loss": -5.956061840057373, "grad_norm": 0.43589547276496887, "lr": 4.137060090156723e-05, "ex_per_s": 325.15942688202745, "data_frac": 0.020957142144559404, "mem_mb": 91.579392, "nll": -5.956061840057373, "elapsed_s": 688.9}182{"event": "val", "step": 3500, "val_nll": -6.451273006874564, "best": -6.451273006874564, "bad": 0, "mem_mb": 91.579392, "elapsed_s": 752.9}183{"event": "train", "step": 3520, "loss": -5.969521999359131, "grad_norm": 0.15035198628902435, "lr": 4.049088307805306e-05, "ex_per_s": 298.55203106183626, "data_frac": 0.019149361553105364, "mem_mb": 91.579392, "nll": -5.969521999359131, "elapsed_s": 754.6}184{"event": "train", "step": 3540, "loss": -5.9866461753845215, "grad_norm": 0.4337175488471985, "lr": 3.964520480388956e-05, "ex_per_s": 299.58284753028323, "data_frac": 0.01912280208848074, "mem_mb": 91.579392, "nll": -5.9866461753845215, "elapsed_s": 756.3}185{"event": "train", "step": 3560, "loss": -5.854382514953613, "grad_norm": 0.13264714181423187, "lr": 3.88337972790728e-05, "ex_per_s": 300.6091280315512, "data_frac": 0.019096135738933366, "mem_mb": 91.579392, "nll": -5.854382514953613, "elapsed_s": 758.0}186{"event": "train", "step": 3580, "loss": -5.998902320861816, "grad_norm": 0.2292366474866867, "lr": 3.805688233431824e-05, "ex_per_s": 301.6307528712049, "data_frac": 0.019069907081513862, "mem_mb": 91.579392, "nll": -5.998902320861816, "elapsed_s": 759.6}187{"event": "train", "step": 3600, "loss": -5.876835346221924, "grad_norm": 0.21841268241405487, "lr": 3.7314672370414325e-05, "ex_per_s": 302.6471204590323, "data_frac": 0.019043960802731196, "mem_mb": 91.579392, "nll": -5.876835346221924, "elapsed_s": 761.3}188{"event": "train", "step": 3620, "loss": -5.734883785247803, "grad_norm": 0.8878716230392456, "lr": 3.660737030015427e-05, "ex_per_s": 303.65986039015024, "data_frac": 0.019018161230911736, "mem_mb": 91.579392, "nll": -5.734883785247803, "elapsed_s": 763.0}189{"event": "train", "step": 3640, "loss": -5.982384204864502, "grad_norm": 0.1558372676372528, "lr": 3.5935169492861716e-05, "ex_per_s": 304.66771449970486, "data_frac": 0.018992375917789784, "mem_mb": 91.579392, "nll": -5.982384204864502, "elapsed_s": 764.7}190{"event": "train", "step": 3660, "loss": -5.978577136993408, "grad_norm": 0.12505929172039032, "lr": 3.52982537215255e-05, "ex_per_s": 305.6713886299141, "data_frac": 0.01896686537160715, "mem_mb": 91.579392, "nll": -5.978577136993408, "elapsed_s": 766.4}191{"event": "train", "step": 3680, "loss": -5.946007251739502, "grad_norm": 0.12129829078912735, "lr": 3.469679711255794e-05, "ex_per_s": 306.6709525240401, "data_frac": 0.01894077555457562, "mem_mb": 91.579392, "nll": -5.946007251739502, "elapsed_s": 768.0}192{"event": "train", "step": 3700, "loss": -5.899984836578369, "grad_norm": 0.267341285943985, "lr": 3.41309640981904e-05, "ex_per_s": 307.6651032321577, "data_frac": 0.018915737257679983, "mem_mb": 91.579392, "nll": -5.899984836578369, "elapsed_s": 769.7}193{"event": "train", "step": 3720, "loss": -5.780359268188477, "grad_norm": 0.4593442678451538, "lr": 3.3600909371519225e-05, "ex_per_s": 308.6547236172151, "data_frac": 0.01889073000154853, "mem_mb": 91.579392, "nll": -5.780359268188477, "elapsed_s": 771.4}194{"event": "train", "step": 3740, "loss": -5.894896030426025, "grad_norm": 0.6824584007263184, "lr": 3.31067778442141e-05, "ex_per_s": 309.64191858031256, "data_frac": 0.018865758681347874, "mem_mb": 91.579392, "nll": -5.894896030426025, "elapsed_s": 773.1}195{"event": "train", "step": 3760, "loss": -6.0849995613098145, "grad_norm": 0.14638835191726685, "lr": 3.2648704606900735e-05, "ex_per_s": 310.6248393239179, "data_frac": 0.018840418835556576, "mem_mb": 91.579392, "nll": -6.0849995613098145, "elapsed_s": 774.7}196{"event": "train", "step": 3780, "loss": -5.984799861907959, "grad_norm": 0.311418354511261, "lr": 3.222681489222838e-05, "ex_per_s": 311.6036585149072, "data_frac": 0.01881532627714434, "mem_mb": 91.579392, "nll": -5.984799861907959, "elapsed_s": 776.4}197{"event": "train", "step": 3800, "loss": -5.9963059425354, "grad_norm": 0.4358578622341156, "lr": 3.1841224040632484e-05, "ex_per_s": 312.57839167613275, "data_frac": 0.018790299682368047, "mem_mb": 91.579392, "nll": -5.9963059425354, "elapsed_s": 778.1}198{"event": "train", "step": 3820, "loss": -5.8175153732299805, "grad_norm": 0.7197644710540771, "lr": 3.149203746880175e-05, "ex_per_s": 313.54852902081154, "data_frac": 0.01876543456006289, "mem_mb": 91.579392, "nll": -5.8175153732299805, "elapsed_s": 779.8}199{"event": "train", "step": 3840, "loss": -5.88870096206665, "grad_norm": 0.6957560777664185, "lr": 3.117935064085835e-05, "ex_per_s": 314.51553951280573, "data_frac": 0.018741186075787473, "mem_mb": 91.579392, "nll": -5.88870096206665, "elapsed_s": 781.4}200{"event": "train", "step": 3860, "loss": -6.032577037811279, "grad_norm": 0.37275728583335876, "lr": 3.090324904225885e-05, "ex_per_s": 315.4782857335704, "data_frac": 0.018716997612940902, "mem_mb": 91.579392, "nll": -6.032577037811279, "elapsed_s": 783.1}201{"event": "train", "step": 3880, "loss": -6.0316362380981445, "grad_norm": 0.4317234456539154, "lr": 3.0663808156423456e-05, "ex_per_s": 316.4365030049801, "data_frac": 0.01869277204375026, "mem_mb": 91.579392, "nll": -6.0316362380981445, "elapsed_s": 784.8}202{"event": "train", "step": 3900, "loss": -5.8741984367370605, "grad_norm": 0.559131920337677, "lr": 3.046109344409957e-05, "ex_per_s": 317.3912220350565, "data_frac": 0.018669134573573744, "mem_mb": 91.579392, "nll": -5.8741984367370605, "elapsed_s": 786.5}203{"event": "train", "step": 3920, "loss": -6.007130146026611, "grad_norm": 0.6891421675682068, "lr": 3.0295160325465435e-05, "ex_per_s": 318.34203934510003, "data_frac": 0.018645169975970422, "mem_mb": 91.579392, "nll": -6.007130146026611, "elapsed_s": 788.1}204{"event": "train", "step": 3940, "loss": -5.882992744445801, "grad_norm": 0.2205687165260315, "lr": 3.016605416497886e-05, "ex_per_s": 319.286342406439, "data_frac": 0.018621565956381304, "mem_mb": 91.579392, "nll": -5.882992744445801, "elapsed_s": 789.8}205{"event": "train", "step": 3960, "loss": -5.790736198425293, "grad_norm": 0.11714425683021545, "lr": 3.0073810258974985e-05, "ex_per_s": 320.22851711108547, "data_frac": 0.018598417422707424, "mem_mb": 91.579392, "nll": -5.790736198425293, "elapsed_s": 791.5}206{"event": "train", "step": 3980, "loss": -5.899009704589844, "grad_norm": 0.5891684293746948, "lr": 3.0018453826016683e-05, "ex_per_s": 321.16746105051294, "data_frac": 0.018575085626176872, "mem_mb": 91.579392, "nll": -5.899009704589844, "elapsed_s": 793.2}207{"event": "train", "step": 4000, "loss": -6.103532314300537, "grad_norm": 0.3700306713581085, "lr": 2.9999999999999997e-05, "ex_per_s": 322.1030815122647, "data_frac": 0.01855151157464586, "mem_mb": 91.579392, "nll": -6.103532314300537, "elapsed_s": 794.8}208{"event": "val", "step": 4000, "val_nll": -6.4511954462762136, "best": -6.451273006874564, "bad": 1, "mem_mb": 91.579392, "elapsed_s": 857.5}209{"event": "done", "steps": 4000, "best_val": -6.451273006874564, "wall_s": 857.5486867427826, "ex_per_s": 298.52532451814704, "peak_mem_mb": 91.579392, "n_params": 3455619, "device": "mps", "data_time_frac": 0.017193551788210516, "elapsed_s": 857.5}210
Files · main
Menu

Theme