SPB Git

spb/wp41_uqo

Public
51.4 KB · 210 lines
1{"event": "train", "step": 20, "loss": -5.841698169708252, "grad_norm": 2.221902847290039, "lr": 3.149999999999999e-05, "ex_per_s": 89.46594519001587, "data_frac": 0.8704356272729106, "mem_mb": 91.210752, "nll": -5.841698169708252, "elapsed_s": 14.4}2{"event": "train", "step": 40, "loss": -6.059584617614746, "grad_norm": 0.4851023852825165, "lr": 6.149999999999999e-05, "ex_per_s": 159.16100264116633, "data_frac": 0.7795587245203344, "mem_mb": 91.210752, "nll": -6.059584617614746, "elapsed_s": 16.1}3{"event": "train", "step": 60, "loss": -5.859298229217529, "grad_norm": 0.21644479036331177, "lr": 9.149999999999999e-05, "ex_per_s": 215.9713008512616, "data_frac": 0.7061641967228485, "mem_mb": 91.210752, "nll": -5.859298229217529, "elapsed_s": 17.8}4{"event": "train", "step": 80, "loss": -5.971600532531738, "grad_norm": 0.42947861552238464, "lr": 0.0001215, "ex_per_s": 263.03512137825135, "data_frac": 0.645611043997347, "mem_mb": 91.210752, "nll": -5.971600532531738, "elapsed_s": 19.5}5{"event": "train", "step": 100, "loss": -5.978066921234131, "grad_norm": 0.6544821262359619, "lr": 0.0001515, "ex_per_s": 302.75440911053107, "data_frac": 0.5951647461763836, "mem_mb": 91.210752, "nll": -5.978066921234131, "elapsed_s": 21.2}6{"event": "train", "step": 120, "loss": -5.92750883102417, "grad_norm": 0.3328477144241333, "lr": 0.00018149999999999997, "ex_per_s": 336.69015375137616, "data_frac": 0.552052323067389, "mem_mb": 91.210752, "nll": -5.92750883102417, "elapsed_s": 22.9}7{"event": "train", "step": 140, "loss": -5.746175289154053, "grad_norm": 0.46641993522644043, "lr": 0.00021149999999999996, "ex_per_s": 365.9813020293279, "data_frac": 0.5148262205535885, "mem_mb": 91.210752, "nll": -5.746175289154053, "elapsed_s": 24.5}8{"event": "train", "step": 160, "loss": -6.076200008392334, "grad_norm": 0.5024569034576416, "lr": 0.0002415, "ex_per_s": 391.49960970063023, "data_frac": 0.4823506074424761, "mem_mb": 91.210752, "nll": -6.076200008392334, "elapsed_s": 26.2}9{"event": "train", "step": 180, "loss": -6.100988864898682, "grad_norm": 0.6710880398750305, "lr": 0.0002715, "ex_per_s": 414.0297875444513, "data_frac": 0.4538243773358024, "mem_mb": 91.210752, "nll": -6.100988864898682, "elapsed_s": 27.9}10{"event": "train", "step": 200, "loss": -5.90971040725708, "grad_norm": 0.3817109167575836, "lr": 0.0003, "ex_per_s": 434.04183513614225, "data_frac": 0.428555462524484, "mem_mb": 91.210752, "nll": -5.90971040725708, "elapsed_s": 29.5}11{"event": "train", "step": 220, "loss": -6.084378719329834, "grad_norm": 0.532283365726471, "lr": 0.00029998154617398326, "ex_per_s": 451.92603221667423, "data_frac": 0.4059839384198192, "mem_mb": 91.210752, "nll": -6.084378719329834, "elapsed_s": 31.2}12{"event": "train", "step": 240, "loss": -5.911529541015625, "grad_norm": 1.2363072633743286, "lr": 0.000299926189741025, "ex_per_s": 467.9857529887083, "data_frac": 0.38570100248650696, "mem_mb": 91.210752, "nll": -5.911529541015625, "elapsed_s": 32.9}13{"event": "train", "step": 260, "loss": -5.966259956359863, "grad_norm": 1.2497937679290771, "lr": 0.0002998339458350211, "ex_per_s": 482.5287171718354, "data_frac": 0.3674118659491537, "mem_mb": 91.210752, "nll": -5.966259956359863, "elapsed_s": 34.5}14{"event": "train", "step": 280, "loss": -6.010699272155762, "grad_norm": 0.6864197850227356, "lr": 0.0002997048396745345, "ex_per_s": 495.71335143748666, "data_frac": 0.35078946163395447, "mem_mb": 91.210752, "nll": -6.010699272155762, "elapsed_s": 36.2}15{"event": "train", "step": 300, "loss": -6.1084675788879395, "grad_norm": 0.4321875274181366, "lr": 0.0002995389065559004, "ex_per_s": 507.73912258707765, "data_frac": 0.3356361523688532, "mem_mb": 91.210752, "nll": -6.1084675788879395, "elapsed_s": 37.9}16{"event": "train", "step": 320, "loss": -5.853294849395752, "grad_norm": 0.5281363725662231, "lr": 0.0002993361918435765, "ex_per_s": 518.7309240081188, "data_frac": 0.3217540255849219, "mem_mb": 91.210752, "nll": -5.853294849395752, "elapsed_s": 39.5}17{"event": "train", "step": 340, "loss": -5.927958965301514, "grad_norm": 0.7523049116134644, "lr": 0.00029909675095774114, "ex_per_s": 528.8431538071023, "data_frac": 0.30899993414090743, "mem_mb": 91.210752, "nll": -5.927958965301514, "elapsed_s": 41.2}18{"event": "train", "step": 360, "loss": -5.958866119384766, "grad_norm": 0.3970949947834015, "lr": 0.00029882064935914164, "ex_per_s": 538.0255507001726, "data_frac": 0.2971645070653516, "mem_mb": 91.210752, "nll": -5.958866119384766, "elapsed_s": 42.9}19{"event": "train", "step": 380, "loss": -5.98062801361084, "grad_norm": 0.4921177625656128, "lr": 0.00029850796253119824, "ex_per_s": 546.4376138812648, "data_frac": 0.2861804921375221, "mem_mb": 91.210752, "nll": -5.98062801361084, "elapsed_s": 44.5}20{"event": "train", "step": 400, "loss": -6.012863636016846, "grad_norm": 0.25698816776275635, "lr": 0.0002981587759593675, "ex_per_s": 554.4340177754423, "data_frac": 0.27610178534947954, "mem_mb": 91.210752, "nll": -6.012863636016846, "elapsed_s": 46.2}21{"event": "train", "step": 420, "loss": -6.123285293579102, "grad_norm": 0.38354426622390747, "lr": 0.0002977731851077716, "ex_per_s": 561.8754112850934, "data_frac": 0.26670483673412737, "mem_mb": 91.210752, "nll": -6.123285293579102, "elapsed_s": 47.9}22{"event": "train", "step": 440, "loss": -5.917429447174072, "grad_norm": 0.1223885789513588, "lr": 0.0002973512953930992, "ex_per_s": 568.8143795867533, "data_frac": 0.25794446801776977, "mem_mb": 91.210752, "nll": -5.917429447174072, "elapsed_s": 49.6}23{"event": "train", "step": 460, "loss": -5.933666706085205, "grad_norm": 0.6267595291137695, "lr": 0.0002968932221557859, "ex_per_s": 575.301054990373, "data_frac": 0.2497586483481915, "mem_mb": 91.210752, "nll": -5.933666706085205, "elapsed_s": 51.2}24{"event": "train", "step": 480, "loss": -6.024855136871338, "grad_norm": 0.6737233400344849, "lr": 0.0002963990906284808, "ex_per_s": 581.3746732963045, "data_frac": 0.2420846149204284, "mem_mb": 91.210752, "nll": -6.024855136871338, "elapsed_s": 52.9}25{"event": "train", "step": 500, "loss": -5.83712100982666, "grad_norm": 0.15794390439987183, "lr": 0.00029586903590180956, "ex_per_s": 587.0855894340897, "data_frac": 0.23488361006133363, "mem_mb": 91.210752, "nll": -5.83712100982666, "elapsed_s": 54.5}26{"event": "val", "step": 500, "val_nll": -6.444735161041836, "best": -6.444735161041836, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 118.7}27{"event": "train", "step": 520, "loss": -6.034851551055908, "grad_norm": 0.9165611863136292, "lr": 0.00029530320288744205, "ex_per_s": 276.38583650043176, "data_frac": 0.10642099098729689, "mem_mb": 91.210752, "nll": -6.034851551055908, "elapsed_s": 120.5}28{"event": "train", "step": 540, "loss": -5.9142279624938965, "grad_norm": 0.4891327917575836, "lr": 0.0002947017462784745, "ex_per_s": 283.0844437417091, "data_frac": 0.10505744949412839, "mem_mb": 91.210752, "nll": -5.9142279624938965, "elapsed_s": 122.1}29{"event": "train", "step": 560, "loss": -5.97721004486084, "grad_norm": 0.6215461492538452, "lr": 0.00029406483050713824, "ex_per_s": 289.6046638039733, "data_frac": 0.1037259160052384, "mem_mb": 91.210752, "nll": -5.97721004486084, "elapsed_s": 123.8}30{"event": "train", "step": 580, "loss": -5.929139137268066, "grad_norm": 0.3790610730648041, "lr": 0.0002933926296998457, "ex_per_s": 295.9453690311816, "data_frac": 0.10243766274564996, "mem_mb": 91.210752, "nll": -5.929139137268066, "elapsed_s": 125.5}31{"event": "train", "step": 600, "loss": -5.946424961090088, "grad_norm": 0.830741822719574, "lr": 0.0002926853276295856, "ex_per_s": 302.1273064510888, "data_frac": 0.10118110958645514, "mem_mb": 91.210752, "nll": -5.946424961090088, "elapsed_s": 127.1}32{"event": "train", "step": 620, "loss": -5.946561813354492, "grad_norm": 0.54784095287323, "lr": 0.00029194311766568174, "ex_per_s": 308.1509544938387, "data_frac": 0.09995562777620445, "mem_mb": 91.210752, "nll": -5.946561813354492, "elapsed_s": 128.8}33{"event": "train", "step": 640, "loss": -5.796878337860107, "grad_norm": 0.4023013710975647, "lr": 0.0002911662027209272, "ex_per_s": 314.0201050654932, "data_frac": 0.09875911908625405, "mem_mb": 91.210752, "nll": -5.796878337860107, "elapsed_s": 130.5}34{"event": "train", "step": 660, "loss": -5.877902507781982, "grad_norm": 0.6141157746315002, "lr": 0.00029035479519611045, "ex_per_s": 319.7399012649202, "data_frac": 0.09759341093618978, "mem_mb": 91.210752, "nll": -5.877902507781982, "elapsed_s": 132.2}35{"event": "train", "step": 680, "loss": -5.962996959686279, "grad_norm": 0.739201009273529, "lr": 0.0002895091169219469, "ex_per_s": 325.31887731385376, "data_frac": 0.09645784263602047, "mem_mb": 91.210752, "nll": -5.962996959686279, "elapsed_s": 133.8}36{"event": "train", "step": 700, "loss": -5.923465728759766, "grad_norm": 0.1536826640367508, "lr": 0.00028862939909843273, "ex_per_s": 330.7616174622503, "data_frac": 0.09534941325282888, "mem_mb": 91.210752, "nll": -5.923465728759766, "elapsed_s": 135.5}37{"event": "train", "step": 720, "loss": -6.019242286682129, "grad_norm": 0.9659088850021362, "lr": 0.0002877158822316369, "ex_per_s": 336.0711488670879, "data_frac": 0.09426581032610955, "mem_mb": 91.210752, "nll": -6.019242286682129, "elapsed_s": 137.2}38{"event": "train", "step": 740, "loss": -5.993554592132568, "grad_norm": 0.5114173293113708, "lr": 0.00028676881606794945, "ex_per_s": 341.2539851119868, "data_frac": 0.09321133539995019, "mem_mb": 91.210752, "nll": -5.993554592132568, "elapsed_s": 138.8}39{"event": "train", "step": 760, "loss": -5.856978893280029, "grad_norm": 0.5998989343643188, "lr": 0.000285788459525803, "ex_per_s": 346.31191617023313, "data_frac": 0.09218078682150432, "mem_mb": 91.210752, "nll": -5.856978893280029, "elapsed_s": 140.5}40{"event": "train", "step": 780, "loss": -5.9066290855407715, "grad_norm": 0.20180554687976837, "lr": 0.00028477508062488754, "ex_per_s": 351.2533598408766, "data_frac": 0.09117489798512333, "mem_mb": 91.210752, "nll": -5.9066290855407715, "elapsed_s": 142.2}41{"event": "train", "step": 800, "loss": -5.870351791381836, "grad_norm": 0.8280360102653503, "lr": 0.000283728956412876, "ex_per_s": 356.0779588676668, "data_frac": 0.09019241828394238, "mem_mb": 91.210752, "nll": -5.870351791381836, "elapsed_s": 143.8}42{"event": "train", "step": 820, "loss": -6.032659530639648, "grad_norm": 0.16700775921344757, "lr": 0.00028265037288968236, "ex_per_s": 360.7919446498986, "data_frac": 0.08933912948522148, "mem_mb": 91.210752, "nll": -6.032659530639648, "elapsed_s": 145.5}43{"event": "train", "step": 840, "loss": -5.9502716064453125, "grad_norm": 0.3526519536972046, "lr": 0.0002815396249292725, "ex_per_s": 365.4043902901845, "data_frac": 0.08841217662662379, "mem_mb": 91.210752, "nll": -5.9502716064453125, "elapsed_s": 147.2}44{"event": "train", "step": 860, "loss": -5.842617511749268, "grad_norm": 0.8445549607276917, "lr": 0.00028039701619904774, "ex_per_s": 369.92519682802293, "data_frac": 0.0875016308199915, "mem_mb": 91.210752, "nll": -5.842617511749268, "elapsed_s": 148.8}45{"event": "train", "step": 880, "loss": -5.958684921264648, "grad_norm": 0.763663113117218, "lr": 0.00027922285907682625, "ex_per_s": 374.3451086205249, "data_frac": 0.08661324345318468, "mem_mb": 91.210752, "nll": -5.958684921264648, "elapsed_s": 150.5}46{"event": "train", "step": 900, "loss": -6.028291702270508, "grad_norm": 0.11522895097732544, "lr": 0.00027801747456544134, "ex_per_s": 378.67099797199216, "data_frac": 0.08573976674602703, "mem_mb": 91.210752, "nll": -6.028291702270508, "elapsed_s": 152.2}47{"event": "train", "step": 920, "loss": -5.976678848266602, "grad_norm": 0.9938505291938782, "lr": 0.00027678119220498265, "ex_per_s": 382.8998116085248, "data_frac": 0.08488421929317384, "mem_mb": 91.210752, "nll": -5.976678848266602, "elapsed_s": 153.8}48{"event": "train", "step": 940, "loss": -6.060321807861328, "grad_norm": 0.24588420987129211, "lr": 0.0002755143499827035, "ex_per_s": 387.0382887323654, "data_frac": 0.08404782515078739, "mem_mb": 91.210752, "nll": -6.060321807861328, "elapsed_s": 155.5}49{"event": "train", "step": 960, "loss": -5.965737819671631, "grad_norm": 0.3825717270374298, "lr": 0.00027421729424061787, "ex_per_s": 391.0834179922066, "data_frac": 0.0832262657812068, "mem_mb": 91.210752, "nll": -5.965737819671631, "elapsed_s": 157.1}50{"event": "train", "step": 980, "loss": -5.8697710037231445, "grad_norm": 0.5800141096115112, "lr": 0.00027289037958081483, "ex_per_s": 395.0503388460904, "data_frac": 0.08242525331630404, "mem_mb": 91.210752, "nll": -5.8697710037231445, "elapsed_s": 158.8}51{"event": "train", "step": 1000, "loss": -5.9044976234436035, "grad_norm": 0.5019832849502563, "lr": 0.0002715339687685131, "ex_per_s": 398.93919552947744, "data_frac": 0.08164247824515894, "mem_mb": 91.210752, "nll": -5.9044976234436035, "elapsed_s": 160.5}52{"event": "val", "step": 1000, "val_nll": -6.444183498657211, "best": -6.444735161041836, "bad": 1, "mem_mb": 91.210752, "elapsed_s": 223.9}53{"event": "train", "step": 1020, "loss": -6.124957084655762, "grad_norm": 0.09187275916337967, "lr": 0.0002701484326328852, "ex_per_s": 289.3027208586239, "data_frac": 0.0580981354105164, "mem_mb": 91.210752, "nll": -6.124957084655762, "elapsed_s": 225.7}54{"event": "train", "step": 1040, "loss": -5.955892562866211, "grad_norm": 0.4091689884662628, "lr": 0.00026873414996567596, "ex_per_s": 292.79858859381255, "data_frac": 0.05772123513278478, "mem_mb": 91.210752, "nll": -5.955892562866211, "elapsed_s": 227.4}55{"event": "train", "step": 1060, "loss": -6.031679153442383, "grad_norm": 0.8243312835693359, "lr": 0.0002672915074176452, "ex_per_s": 296.2470191142492, "data_frac": 0.057347228610658676, "mem_mb": 91.210752, "nll": -6.031679153442383, "elapsed_s": 229.0}56{"event": "train", "step": 1080, "loss": -5.761965751647949, "grad_norm": 0.36728495359420776, "lr": 0.00026582089939286075, "ex_per_s": 299.6451921510246, "data_frac": 0.05697877553054745, "mem_mb": 91.210752, "nll": -5.761965751647949, "elapsed_s": 230.7}57{"event": "train", "step": 1100, "loss": -5.856236934661865, "grad_norm": 0.4474188983440399, "lr": 0.0002643227279408728, "ex_per_s": 302.99420515979597, "data_frac": 0.05661421126741551, "mem_mb": 91.210752, "nll": -5.856236934661865, "elapsed_s": 232.4}58{"event": "train", "step": 1120, "loss": -6.09713077545166, "grad_norm": 0.85405033826828, "lr": 0.000262797402646797, "ex_per_s": 306.2942797661541, "data_frac": 0.056255286450043344, "mem_mb": 91.210752, "nll": -6.09713077545166, "elapsed_s": 234.1}59{"event": "train", "step": 1140, "loss": -6.2340087890625, "grad_norm": 0.08634272962808609, "lr": 0.0002612453405193386, "ex_per_s": 309.54595162006007, "data_frac": 0.05590352705984866, "mem_mb": 91.210752, "nll": -6.2340087890625, "elapsed_s": 235.7}60{"event": "train", "step": 1160, "loss": -6.126562118530273, "grad_norm": 1.0171434879302979, "lr": 0.0002596669658767859, "ex_per_s": 312.7503338947537, "data_frac": 0.055555326834931446, "mem_mb": 91.210752, "nll": -6.126562118530273, "elapsed_s": 237.4}61{"event": "train", "step": 1180, "loss": -6.069647789001465, "grad_norm": 0.20892098546028137, "lr": 0.00025806271023100637, "ex_per_s": 315.917024113929, "data_frac": 0.055211355896834874, "mem_mb": 91.210752, "nll": -6.069647789001465, "elapsed_s": 239.1}62{"event": "train", "step": 1200, "loss": -5.950080394744873, "grad_norm": 0.21578292548656464, "lr": 0.00025643301216947504, "ex_per_s": 319.0380525329636, "data_frac": 0.054872190679442365, "mem_mb": 91.210752, "nll": -5.950080394744873, "elapsed_s": 240.8}63{"event": "train", "step": 1220, "loss": -5.9877142906188965, "grad_norm": 0.58149653673172, "lr": 0.0002547783172353695, "ex_per_s": 322.1222216981353, "data_frac": 0.05453976095359237, "mem_mb": 91.210752, "nll": -5.9877142906188965, "elapsed_s": 242.4}64{"event": "train", "step": 1240, "loss": -5.9434123039245605, "grad_norm": 0.15840184688568115, "lr": 0.00025309907780576277, "ex_per_s": 325.16315139500017, "data_frac": 0.054210706849805446, "mem_mb": 91.210752, "nll": -5.9434123039245605, "elapsed_s": 244.1}65{"event": "train", "step": 1260, "loss": -6.081459045410156, "grad_norm": 0.12046205997467041, "lr": 0.0002513957529679477, "ex_per_s": 328.16081113636875, "data_frac": 0.05388704360814865, "mem_mb": 91.210752, "nll": -6.081459045410156, "elapsed_s": 245.8}66{"event": "train", "step": 1280, "loss": -6.033250331878662, "grad_norm": 0.14856822788715363, "lr": 0.00024966880839392746, "ex_per_s": 331.12069624380797, "data_frac": 0.053572185410730305, "mem_mb": 91.210752, "nll": -6.033250331878662, "elapsed_s": 247.4}67{"event": "train", "step": 1300, "loss": -5.913089752197266, "grad_norm": 0.2796117067337036, "lr": 0.0002479187162131052, "ex_per_s": 334.04238002166005, "data_frac": 0.05325870064458046, "mem_mb": 91.210752, "nll": -5.913089752197266, "elapsed_s": 249.1}68{"event": "train", "step": 1320, "loss": -5.8588361740112305, "grad_norm": 0.6998174786567688, "lr": 0.00024614595488320833, "ex_per_s": 336.9254120412962, "data_frac": 0.05294830762371241, "mem_mb": 91.210752, "nll": -5.8588361740112305, "elapsed_s": 250.8}69{"event": "train", "step": 1340, "loss": -5.821824550628662, "grad_norm": 0.3291776776313782, "lr": 0.00024435100905948387, "ex_per_s": 339.7699671336833, "data_frac": 0.05263923767932849, "mem_mb": 91.210752, "nll": -5.821824550628662, "elapsed_s": 252.4}70{"event": "train", "step": 1360, "loss": -5.916557788848877, "grad_norm": 0.8000309467315674, "lr": 0.0002425343694621974, "ex_per_s": 342.58099600259663, "data_frac": 0.05233660260674035, "mem_mb": 91.210752, "nll": -5.916557788848877, "elapsed_s": 254.1}71{"event": "train", "step": 1380, "loss": -6.058549404144287, "grad_norm": 0.2735864520072937, "lr": 0.00024069653274247583, "ex_per_s": 345.35575693065846, "data_frac": 0.05203720677053999, "mem_mb": 91.210752, "nll": -6.058549404144287, "elapsed_s": 255.8}72{"event": "train", "step": 1400, "loss": -6.279260158538818, "grad_norm": 0.9549837708473206, "lr": 0.0002388380013465276, "ex_per_s": 348.0960871243598, "data_frac": 0.05174270517199415, "mem_mb": 91.210752, "nll": -6.279260158538818, "elapsed_s": 257.4}73{"event": "train", "step": 1420, "loss": -6.054855823516846, "grad_norm": 0.3866163492202759, "lr": 0.00023695928337827914, "ex_per_s": 350.80248650263223, "data_frac": 0.05145239282789407, "mem_mb": 91.210752, "nll": -6.054855823516846, "elapsed_s": 259.1}74{"event": "train", "step": 1440, "loss": -5.838022708892822, "grad_norm": 0.4068060517311096, "lr": 0.00023506089246046403, "ex_per_s": 353.47293875068647, "data_frac": 0.05116538121694274, "mem_mb": 91.210752, "nll": -5.838022708892822, "elapsed_s": 260.8}75{"event": "train", "step": 1460, "loss": -5.933524131774902, "grad_norm": 0.30775031447410583, "lr": 0.00023314334759420388, "ex_per_s": 356.10997050205987, "data_frac": 0.05088353621878928, "mem_mb": 91.210752, "nll": -5.933524131774902, "elapsed_s": 262.4}76{"event": "train", "step": 1480, "loss": -5.978257656097412, "grad_norm": 0.6662095189094543, "lr": 0.0002312071730171186, "ex_per_s": 358.71514906978797, "data_frac": 0.050605289736866614, "mem_mb": 91.210752, "nll": -5.978257656097412, "elapsed_s": 264.1}77{"event": "train", "step": 1500, "loss": -6.007508754730225, "grad_norm": 0.0881936326622963, "lr": 0.0002292528980600049, "ex_per_s": 361.28480985825024, "data_frac": 0.050329172722192006, "mem_mb": 91.210752, "nll": -6.007508754730225, "elapsed_s": 265.8}78{"event": "val", "step": 1500, "val_nll": -6.447400228185234, "best": -6.447400228185234, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 329.2}79{"event": "train", "step": 1520, "loss": -5.936246871948242, "grad_norm": 0.4227154850959778, "lr": 0.00022728105700212285, "ex_per_s": 294.0185073134008, "data_frac": 0.04045431752054295, "mem_mb": 91.210752, "nll": -5.936246871948242, "elapsed_s": 330.9}80{"event": "train", "step": 1540, "loss": -5.780516147613525, "grad_norm": 0.17959490418434143, "lr": 0.00022529218892512886, "ex_per_s": 296.388604957868, "data_frac": 0.04028455366779963, "mem_mb": 91.210752, "nll": -5.780516147613525, "elapsed_s": 332.6}81{"event": "train", "step": 1560, "loss": -5.897628307342529, "grad_norm": 0.5487073659896851, "lr": 0.00022328683756569652, "ex_per_s": 298.7344207453439, "data_frac": 0.04011613229105049, "mem_mb": 91.210752, "nll": -5.897628307342529, "elapsed_s": 334.3}82{"event": "train", "step": 1580, "loss": -5.984220504760742, "grad_norm": 0.0938587635755539, "lr": 0.0002212655511668642, "ex_per_s": 301.056664049833, "data_frac": 0.039949221789806075, "mem_mb": 91.210752, "nll": -5.984220504760742, "elapsed_s": 335.9}83{"event": "train", "step": 1600, "loss": -5.975241661071777, "grad_norm": 0.5109139680862427, "lr": 0.00021922888232815089, "ex_per_s": 303.35626912303707, "data_frac": 0.03978321534018351, "mem_mb": 91.210752, "nll": -5.975241661071777, "elapsed_s": 337.6}84{"event": "train", "step": 1620, "loss": -5.743502140045166, "grad_norm": 0.3065767288208008, "lr": 0.00021717738785448102, "ex_per_s": 305.63222330289653, "data_frac": 0.03961923665974815, "mem_mb": 91.210752, "nll": -5.743502140045166, "elapsed_s": 339.3}85{"event": "train", "step": 1640, "loss": -5.914535045623779, "grad_norm": 0.3715212047100067, "lr": 0.00021511162860395977, "ex_per_s": 307.8852549860389, "data_frac": 0.03945796416017068, "mem_mb": 91.210752, "nll": -5.914535045623779, "elapsed_s": 340.9}86{"event": "train", "step": 1660, "loss": -5.873894214630127, "grad_norm": 0.5153968334197998, "lr": 0.0002130321693345397, "ex_per_s": 310.119934858751, "data_frac": 0.03929746554075998, "mem_mb": 91.210752, "nll": -5.873894214630127, "elapsed_s": 342.6}87{"event": "train", "step": 1680, "loss": -5.834245204925537, "grad_norm": 0.912449061870575, "lr": 0.0002109395785496218, "ex_per_s": 312.3339527102966, "data_frac": 0.039138630676676625, "mem_mb": 91.210752, "nll": -5.834245204925537, "elapsed_s": 344.3}88{"event": "train", "step": 1700, "loss": -6.136268138885498, "grad_norm": 0.19005858898162842, "lr": 0.00020883442834263225, "ex_per_s": 314.52624411732324, "data_frac": 0.03898042684458081, "mem_mb": 91.210752, "nll": -6.136268138885498, "elapsed_s": 346.0}89{"event": "train", "step": 1720, "loss": -5.883195400238037, "grad_norm": 0.11153122782707214, "lr": 0.00020671729424061788, "ex_per_s": 316.6943689646688, "data_frac": 0.03882721768086129, "mem_mb": 91.210752, "nll": -5.883195400238037, "elapsed_s": 347.6}90{"event": "train", "step": 1740, "loss": -5.946223258972168, "grad_norm": 0.4607234001159668, "lr": 0.00020458875504690275, "ex_per_s": 318.84760212786813, "data_frac": 0.038673796068177474, "mem_mb": 91.210752, "nll": -5.946223258972168, "elapsed_s": 349.3}91{"event": "train", "step": 1760, "loss": -6.031312942504883, "grad_norm": 0.7836379408836365, "lr": 0.000202449392682849, "ex_per_s": 320.97987678434083, "data_frac": 0.038520437124751235, "mem_mb": 91.210752, "nll": -6.031312942504883, "elapsed_s": 351.0}92{"event": "train", "step": 1780, "loss": -6.126735210418701, "grad_norm": 0.1428418755531311, "lr": 0.00020029979202876563, "ex_per_s": 323.0930276080454, "data_frac": 0.038369230581021534, "mem_mb": 91.210752, "nll": -6.126735210418701, "elapsed_s": 352.6}93{"event": "train", "step": 1800, "loss": -5.958287239074707, "grad_norm": 0.19570878148078918, "lr": 0.00019814054076400787, "ex_per_s": 325.18595629833317, "data_frac": 0.03822003994141412, "mem_mb": 91.210752, "nll": -5.958287239074707, "elapsed_s": 354.3}94{"event": "train", "step": 1820, "loss": -6.012476921081543, "grad_norm": 0.22258028388023376, "lr": 0.00019597222920631147, "ex_per_s": 327.2595967604427, "data_frac": 0.03807166911481434, "mem_mb": 91.210752, "nll": -6.012476921081543, "elapsed_s": 356.0}95{"event": "train", "step": 1840, "loss": -5.966746807098389, "grad_norm": 0.43454253673553467, "lr": 0.0001937954501504058, "ex_per_s": 329.305151359124, "data_frac": 0.037926613588376956, "mem_mb": 91.210752, "nll": -5.966746807098389, "elapsed_s": 357.6}96{"event": "train", "step": 1860, "loss": -5.915466785430908, "grad_norm": 0.8064450025558472, "lr": 0.0001916107987059498, "ex_per_s": 331.33337474785276, "data_frac": 0.0377811545042387, "mem_mb": 91.210752, "nll": -5.915466785430908, "elapsed_s": 359.3}97{"event": "train", "step": 1880, "loss": -5.904940128326416, "grad_norm": 0.3579712212085724, "lr": 0.00018941887213483484, "ex_per_s": 333.3457756156384, "data_frac": 0.03763630459227163, "mem_mb": 91.210752, "nll": -5.904940128326416, "elapsed_s": 361.0}98{"event": "train", "step": 1900, "loss": -6.203967094421387, "grad_norm": 0.5230263471603394, "lr": 0.00018722026968789906, "ex_per_s": 335.33942143619737, "data_frac": 0.03749383268281809, "mem_mb": 91.210752, "nll": -6.203967094421387, "elapsed_s": 362.7}99{"event": "train", "step": 1920, "loss": -6.179316520690918, "grad_norm": 0.47064581513404846, "lr": 0.00018501559244109864, "ex_per_s": 337.3151127821602, "data_frac": 0.037351593931545746, "mem_mb": 91.210752, "nll": -6.179316520690918, "elapsed_s": 364.3}100{"event": "train", "step": 1940, "loss": -5.9429192543029785, "grad_norm": 0.14980104565620422, "lr": 0.00018280544313117932, "ex_per_s": 339.2707092347774, "data_frac": 0.037211760865222, "mem_mb": 91.210752, "nll": -5.9429192543029785, "elapsed_s": 366.0}101{"event": "train", "step": 1960, "loss": -5.89815092086792, "grad_norm": 1.0434833765029907, "lr": 0.00018059042599089415, "ex_per_s": 341.20908210299507, "data_frac": 0.037074875838719, "mem_mb": 91.210752, "nll": -5.89815092086792, "elapsed_s": 367.7}102{"event": "train", "step": 1980, "loss": -5.978249549865723, "grad_norm": 0.5677815675735474, "lr": 0.00017837114658381247, "ex_per_s": 343.1351501595609, "data_frac": 0.036938273511627304, "mem_mb": 91.210752, "nll": -5.978249549865723, "elapsed_s": 369.3}103{"event": "train", "step": 2000, "loss": -5.901158809661865, "grad_norm": 0.13491590321063995, "lr": 0.00017614821163876486, "ex_per_s": 345.04362126572306, "data_frac": 0.0368025149411549, "mem_mb": 91.210752, "nll": -5.901158809661865, "elapsed_s": 371.0}104{"event": "val", "step": 2000, "val_nll": -6.447393487998027, "best": -6.447400228185234, "bad": 1, "mem_mb": 91.210752, "elapsed_s": 434.4}105{"event": "train", "step": 2020, "loss": -5.957119941711426, "grad_norm": 0.17238225042819977, "lr": 0.00017392222888396941, "ex_per_s": 296.4699137584362, "data_frac": 0.03133389449277718, "mem_mb": 91.210752, "nll": -5.957119941711426, "elapsed_s": 436.1}106{"event": "train", "step": 2040, "loss": -5.938762664794922, "grad_norm": 0.38799163699150085, "lr": 0.000171693806880885, "ex_per_s": 298.2592542843733, "data_frac": 0.031238825983834838, "mem_mb": 91.210752, "nll": -5.938762664794922, "elapsed_s": 437.8}107{"event": "train", "step": 2060, "loss": -5.921187877655029, "grad_norm": 0.24151603877544403, "lr": 0.0001694635548578366, "ex_per_s": 300.0356948568406, "data_frac": 0.031144050921832894, "mem_mb": 91.210752, "nll": -5.921187877655029, "elapsed_s": 439.5}108{"event": "train", "step": 2080, "loss": -5.874792575836182, "grad_norm": 0.42840006947517395, "lr": 0.0001672320825434584, "ex_per_s": 301.798269984877, "data_frac": 0.031050193757559214, "mem_mb": 91.210752, "nll": -5.874792575836182, "elapsed_s": 441.1}109{"event": "train", "step": 2100, "loss": -5.890578269958496, "grad_norm": 0.6374815702438354, "lr": 0.000165, "ex_per_s": 303.5466320519301, "data_frac": 0.030957630303614744, "mem_mb": 91.210752, "nll": -5.890578269958496, "elapsed_s": 442.8}110{"event": "train", "step": 2120, "loss": -6.140714645385742, "grad_norm": 0.5037065148353577, "lr": 0.0001627679174565416, "ex_per_s": 305.28307408109515, "data_frac": 0.030864813405369323, "mem_mb": 91.210752, "nll": -6.140714645385742, "elapsed_s": 444.5}111{"event": "train", "step": 2140, "loss": -5.9174957275390625, "grad_norm": 0.4180409610271454, "lr": 0.00016053644514216336, "ex_per_s": 307.00645598077836, "data_frac": 0.030773841863412206, "mem_mb": 91.210752, "nll": -5.9174957275390625, "elapsed_s": 446.2}112{"event": "train", "step": 2160, "loss": -5.943973064422607, "grad_norm": 0.1712675839662552, "lr": 0.00015830619311911498, "ex_per_s": 308.7021480102245, "data_frac": 0.03069107302224122, "mem_mb": 91.210752, "nll": -5.943973064422607, "elapsed_s": 447.9}113{"event": "train", "step": 2180, "loss": -6.0250630378723145, "grad_norm": 0.099614717066288, "lr": 0.00015607777111603058, "ex_per_s": 310.40065753015637, "data_frac": 0.030606852969685374, "mem_mb": 91.210752, "nll": -6.0250630378723145, "elapsed_s": 449.5}114{"event": "train", "step": 2200, "loss": -5.851351261138916, "grad_norm": 0.2950785458087921, "lr": 0.00015385178836123514, "ex_per_s": 312.08597334314794, "data_frac": 0.030523522909509156, "mem_mb": 91.210752, "nll": -5.851351261138916, "elapsed_s": 451.2}115{"event": "train", "step": 2220, "loss": -5.836081027984619, "grad_norm": 0.36219754815101624, "lr": 0.00015162885341618752, "ex_per_s": 313.75258112583344, "data_frac": 0.03044797350300214, "mem_mb": 91.210752, "nll": -5.836081027984619, "elapsed_s": 452.9}116{"event": "train", "step": 2240, "loss": -5.946961879730225, "grad_norm": 0.2174311876296997, "lr": 0.00014940957400910585, "ex_per_s": 315.4207109805873, "data_frac": 0.030361103387712425, "mem_mb": 91.210752, "nll": -5.946961879730225, "elapsed_s": 454.5}117{"event": "train", "step": 2260, "loss": -5.894201755523682, "grad_norm": 0.47290608286857605, "lr": 0.00014719455686882068, "ex_per_s": 317.07350628577245, "data_frac": 0.030275986654863887, "mem_mb": 91.210752, "nll": -5.894201755523682, "elapsed_s": 456.2}118{"event": "train", "step": 2280, "loss": -6.071601390838623, "grad_norm": 0.16182540357112885, "lr": 0.0001449844075589013, "ex_per_s": 318.7151461272925, "data_frac": 0.0301902553798464, "mem_mb": 91.210752, "nll": -6.071601390838623, "elapsed_s": 457.9}119{"event": "train", "step": 2300, "loss": -6.027144432067871, "grad_norm": 0.08688570559024811, "lr": 0.0001427797303121009, "ex_per_s": 320.34568988462667, "data_frac": 0.03010518897920798, "mem_mb": 91.210752, "nll": -6.027144432067871, "elapsed_s": 459.5}120{"event": "train", "step": 2320, "loss": -5.988244533538818, "grad_norm": 0.10214187949895859, "lr": 0.00014058112786516518, "ex_per_s": 321.9621867719951, "data_frac": 0.0300212412186145, "mem_mb": 91.210752, "nll": -5.988244533538818, "elapsed_s": 461.2}121{"event": "train", "step": 2340, "loss": -6.036227226257324, "grad_norm": 0.5591873526573181, "lr": 0.00013838920129405016, "ex_per_s": 323.56659004574135, "data_frac": 0.02993803020470805, "mem_mb": 91.210752, "nll": -6.036227226257324, "elapsed_s": 462.9}122{"event": "train", "step": 2360, "loss": -5.982945442199707, "grad_norm": 0.3772628605365753, "lr": 0.0001362045498495942, "ex_per_s": 325.1601776522902, "data_frac": 0.02985431569041764, "mem_mb": 91.210752, "nll": -5.982945442199707, "elapsed_s": 464.6}123{"event": "train", "step": 2380, "loss": -6.030460357666016, "grad_norm": 1.1027617454528809, "lr": 0.00013402777079368853, "ex_per_s": 326.74334555216547, "data_frac": 0.029771296213088765, "mem_mb": 91.210752, "nll": -6.030460357666016, "elapsed_s": 466.2}124{"event": "train", "step": 2400, "loss": -6.053019046783447, "grad_norm": 0.26526615023612976, "lr": 0.00013185945923599213, "ex_per_s": 328.3141768192354, "data_frac": 0.029689326722552915, "mem_mb": 91.210752, "nll": -6.053019046783447, "elapsed_s": 467.9}125{"event": "train", "step": 2420, "loss": -6.053780555725098, "grad_norm": 0.9656879901885986, "lr": 0.00012970020797123432, "ex_per_s": 329.874346863751, "data_frac": 0.02960790351943836, "mem_mb": 91.210752, "nll": -6.053780555725098, "elapsed_s": 469.6}126{"event": "train", "step": 2440, "loss": -6.011946678161621, "grad_norm": 0.19278423488140106, "lr": 0.000127550607317151, "ex_per_s": 331.4233358554344, "data_frac": 0.02952708282131383, "mem_mb": 91.210752, "nll": -6.011946678161621, "elapsed_s": 471.2}127{"event": "train", "step": 2460, "loss": -5.939724445343018, "grad_norm": 0.21075624227523804, "lr": 0.00012541124495309727, "ex_per_s": 332.96176500761305, "data_frac": 0.02944628986991054, "mem_mb": 91.210752, "nll": -5.939724445343018, "elapsed_s": 472.9}128{"event": "train", "step": 2480, "loss": -5.924521446228027, "grad_norm": 0.1529516726732254, "lr": 0.00012328270575938212, "ex_per_s": 334.4895491864278, "data_frac": 0.029366261594967094, "mem_mb": 91.210752, "nll": -5.924521446228027, "elapsed_s": 474.6}129{"event": "train", "step": 2500, "loss": -5.857922554016113, "grad_norm": 0.14147904515266418, "lr": 0.00012116557165736775, "ex_per_s": 336.00630683148336, "data_frac": 0.029287671144124547, "mem_mb": 91.210752, "nll": -5.857922554016113, "elapsed_s": 476.2}130{"event": "val", "step": 2500, "val_nll": -6.449225811476839, "best": -6.449225811476839, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 539.7}131{"event": "train", "step": 2520, "loss": -5.896664142608643, "grad_norm": 0.09527081996202469, "lr": 0.00011906042145037817, "ex_per_s": 297.91281598845643, "data_frac": 0.025783290845219114, "mem_mb": 91.210752, "nll": -5.896664142608643, "elapsed_s": 541.4}132{"event": "train", "step": 2540, "loss": -5.879454135894775, "grad_norm": 0.08107165992259979, "lr": 0.00011696783066546026, "ex_per_s": 299.35127698512457, "data_frac": 0.025726136153535605, "mem_mb": 91.210752, "nll": -5.879454135894775, "elapsed_s": 543.1}133{"event": "train", "step": 2560, "loss": -5.97968053817749, "grad_norm": 0.4299287497997284, "lr": 0.00011488837139604017, "ex_per_s": 300.7826856906762, "data_frac": 0.02566761373418008, "mem_mb": 91.210752, "nll": -5.97968053817749, "elapsed_s": 544.8}134{"event": "train", "step": 2580, "loss": -6.025605201721191, "grad_norm": 0.12945927679538727, "lr": 0.00011282261214551898, "ex_per_s": 302.2040226985587, "data_frac": 0.025609879410086238, "mem_mb": 91.210752, "nll": -6.025605201721191, "elapsed_s": 546.4}135{"event": "train", "step": 2600, "loss": -5.880365371704102, "grad_norm": 0.3245372772216797, "lr": 0.00011077111767184915, "ex_per_s": 303.6144158325089, "data_frac": 0.02555368581811273, "mem_mb": 91.210752, "nll": -5.880365371704102, "elapsed_s": 548.1}136{"event": "train", "step": 2620, "loss": -6.031259059906006, "grad_norm": 0.09070190042257309, "lr": 0.00010873444883313583, "ex_per_s": 305.01826810329, "data_frac": 0.025497882294643552, "mem_mb": 91.210752, "nll": -6.031259059906006, "elapsed_s": 549.8}137{"event": "train", "step": 2640, "loss": -6.0807085037231445, "grad_norm": 0.41239723563194275, "lr": 0.00010671316243430345, "ex_per_s": 306.4144718424856, "data_frac": 0.02544048012948255, "mem_mb": 91.210752, "nll": -6.0807085037231445, "elapsed_s": 551.5}138{"event": "train", "step": 2660, "loss": -6.020235061645508, "grad_norm": 0.6605398058891296, "lr": 0.00010470781107487112, "ex_per_s": 307.8025197683191, "data_frac": 0.025383628613799534, "mem_mb": 91.210752, "nll": -6.020235061645508, "elapsed_s": 553.1}139{"event": "train", "step": 2680, "loss": -5.85435152053833, "grad_norm": 0.0970563068985939, "lr": 0.00010271894299787715, "ex_per_s": 309.1810261185055, "data_frac": 0.025328365267069424, "mem_mb": 91.210752, "nll": -5.85435152053833, "elapsed_s": 554.8}140{"event": "train", "step": 2700, "loss": -5.906038761138916, "grad_norm": 0.14178670942783356, "lr": 0.00010074710193999506, "ex_per_s": 310.55267579788864, "data_frac": 0.025272563370696364, "mem_mb": 91.210752, "nll": -5.906038761138916, "elapsed_s": 556.5}141{"event": "train", "step": 2720, "loss": -5.7374677658081055, "grad_norm": 0.25024330615997314, "lr": 9.879282698288144e-05, "ex_per_s": 311.9161717284227, "data_frac": 0.025217071684537302, "mem_mb": 91.210752, "nll": -5.7374677658081055, "elapsed_s": 558.1}142{"event": "train", "step": 2740, "loss": -6.002330780029297, "grad_norm": 0.08502272516489029, "lr": 9.685665240579613e-05, "ex_per_s": 313.27145005775424, "data_frac": 0.025161165898587003, "mem_mb": 91.210752, "nll": -6.002330780029297, "elapsed_s": 559.8}143{"event": "train", "step": 2760, "loss": -5.968435764312744, "grad_norm": 0.5589243173599243, "lr": 9.4939107539536e-05, "ex_per_s": 314.6193194960946, "data_frac": 0.02510607248368531, "mem_mb": 91.210752, "nll": -5.968435764312744, "elapsed_s": 561.5}144{"event": "train", "step": 2780, "loss": -6.026730537414551, "grad_norm": 0.14232084155082703, "lr": 9.304071662172087e-05, "ex_per_s": 315.9593925089095, "data_frac": 0.025050795445657366, "mem_mb": 91.210752, "nll": -6.026730537414551, "elapsed_s": 563.2}145{"event": "train", "step": 2800, "loss": -6.006741523742676, "grad_norm": 0.8277803659439087, "lr": 9.116199865347241e-05, "ex_per_s": 317.2921814577588, "data_frac": 0.024995654282957722, "mem_mb": 91.210752, "nll": -6.006741523742676, "elapsed_s": 564.8}146{"event": "train", "step": 2820, "loss": -5.920007705688477, "grad_norm": 0.13874268531799316, "lr": 8.930346725752419e-05, "ex_per_s": 318.61690263238654, "data_frac": 0.02494181536859788, "mem_mb": 91.210752, "nll": -5.920007705688477, "elapsed_s": 566.5}147{"event": "train", "step": 2840, "loss": -5.909499168395996, "grad_norm": 0.22277526557445526, "lr": 8.746563053780262e-05, "ex_per_s": 319.93060663476325, "data_frac": 0.02488905857887405, "mem_mb": 91.210752, "nll": -5.909499168395996, "elapsed_s": 568.2}148{"event": "train", "step": 2860, "loss": -6.098891735076904, "grad_norm": 0.4815010130405426, "lr": 8.564899094051614e-05, "ex_per_s": 321.2418671590058, "data_frac": 0.02483691303069254, "mem_mb": 91.210752, "nll": -6.098891735076904, "elapsed_s": 569.8}149{"event": "train", "step": 2880, "loss": -6.070987224578857, "grad_norm": 0.5919878482818604, "lr": 8.385404511679161e-05, "ex_per_s": 322.5468336374265, "data_frac": 0.024784102064778647, "mem_mb": 91.210752, "nll": -6.070987224578857, "elapsed_s": 571.5}150{"event": "train", "step": 2900, "loss": -5.9061279296875, "grad_norm": 0.6070135831832886, "lr": 8.208128378689484e-05, "ex_per_s": 323.8440840817052, "data_frac": 0.024731442400884488, "mem_mb": 91.210752, "nll": -5.9061279296875, "elapsed_s": 573.2}151{"event": "train", "step": 2920, "loss": -6.076800346374512, "grad_norm": 0.09847317636013031, "lr": 8.03311916060725e-05, "ex_per_s": 325.1332991428924, "data_frac": 0.024679355316303678, "mem_mb": 91.210752, "nll": -6.076800346374512, "elapsed_s": 574.8}152{"event": "train", "step": 2940, "loss": -6.0899248123168945, "grad_norm": 0.14553995430469513, "lr": 7.860424703205225e-05, "ex_per_s": 326.4154554868473, "data_frac": 0.024627879369899693, "mem_mb": 91.210752, "nll": -6.0899248123168945, "elapsed_s": 576.5}153{"event": "train", "step": 2960, "loss": -5.910630226135254, "grad_norm": 0.30287966132164, "lr": 7.690092219423722e-05, "ex_per_s": 327.6904226288028, "data_frac": 0.024576638081818477, "mem_mb": 91.210752, "nll": -5.910630226135254, "elapsed_s": 578.2}154{"event": "train", "step": 2980, "loss": -5.932413101196289, "grad_norm": 0.42161619663238525, "lr": 7.522168276463048e-05, "ex_per_s": 328.9584509365773, "data_frac": 0.02452574357152057, "mem_mb": 91.210752, "nll": -5.932413101196289, "elapsed_s": 579.8}155{"event": "train", "step": 3000, "loss": -6.120960712432861, "grad_norm": 0.08288842439651489, "lr": 7.356698783052496e-05, "ex_per_s": 330.21867403023856, "data_frac": 0.024475105699094505, "mem_mb": 91.210752, "nll": -6.120960712432861, "elapsed_s": 581.5}156{"event": "val", "step": 3000, "val_nll": -6.450616433821653, "best": -6.450616433821653, "bad": 0, "mem_mb": 91.210752, "elapsed_s": 644.7}157{"event": "train", "step": 3020, "loss": -6.036746501922607, "grad_norm": 0.17894531786441803, "lr": 7.193728976899362e-05, "ex_per_s": 298.9829125142033, "data_frac": 0.022032171263485743, "mem_mb": 91.210752, "nll": -6.036746501922607, "elapsed_s": 646.5}158{"event": "train", "step": 3040, "loss": -5.882773399353027, "grad_norm": 0.3920639753341675, "lr": 7.033303412321404e-05, "ex_per_s": 300.1847114140302, "data_frac": 0.021992489562275005, "mem_mb": 91.210752, "nll": -5.882773399353027, "elapsed_s": 648.2}159{"event": "train", "step": 3060, "loss": -5.975804328918457, "grad_norm": 0.7562133073806763, "lr": 6.875465948066139e-05, "ex_per_s": 301.37928658490284, "data_frac": 0.021955251240699275, "mem_mb": 91.210752, "nll": -5.975804328918457, "elapsed_s": 649.9}160{"event": "train", "step": 3080, "loss": -5.9666852951049805, "grad_norm": 0.7162729501724243, "lr": 6.720259735320299e-05, "ex_per_s": 302.5698232609742, "data_frac": 0.021915573779540645, "mem_mb": 91.210752, "nll": -5.9666852951049805, "elapsed_s": 651.5}161{"event": "train", "step": 3100, "loss": -5.8737030029296875, "grad_norm": 0.27623119950294495, "lr": 6.567727205912722e-05, "ex_per_s": 303.75331871797704, "data_frac": 0.02187660428806628, "mem_mb": 91.210752, "nll": -5.8737030029296875, "elapsed_s": 653.2}162{"event": "train", "step": 3120, "loss": -5.808856010437012, "grad_norm": 0.6026402115821838, "lr": 6.417910060713926e-05, "ex_per_s": 304.9323220327071, "data_frac": 0.02183867518050722, "mem_mb": 91.210752, "nll": -5.808856010437012, "elapsed_s": 654.9}163{"event": "train", "step": 3140, "loss": -5.8936543464660645, "grad_norm": 0.3619730472564697, "lr": 6.270849258235485e-05, "ex_per_s": 306.1047600203862, "data_frac": 0.02180012994115382, "mem_mb": 91.210752, "nll": -5.8936543464660645, "elapsed_s": 656.6}164{"event": "train", "step": 3160, "loss": -6.002748489379883, "grad_norm": 0.6164283156394958, "lr": 6.126585003432406e-05, "ex_per_s": 307.2711450576598, "data_frac": 0.021761639505113606, "mem_mb": 91.210752, "nll": -6.002748489379883, "elapsed_s": 658.2}165{"event": "train", "step": 3180, "loss": -5.957095623016357, "grad_norm": 0.09600992500782013, "lr": 5.985156736711483e-05, "ex_per_s": 308.43216352029725, "data_frac": 0.02172390590281414, "mem_mb": 91.210752, "nll": -5.957095623016357, "elapsed_s": 659.9}166{"event": "train", "step": 3200, "loss": -5.943420886993408, "grad_norm": 0.16108688712120056, "lr": 5.846603123148688e-05, "ex_per_s": 309.5885215122738, "data_frac": 0.021685873728641586, "mem_mb": 91.210752, "nll": -5.943420886993408, "elapsed_s": 661.6}167{"event": "train", "step": 3220, "loss": -6.071151256561279, "grad_norm": 0.4444780647754669, "lr": 5.710962041918516e-05, "ex_per_s": 310.7381932875755, "data_frac": 0.02164853054392323, "mem_mb": 91.210752, "nll": -6.071151256561279, "elapsed_s": 663.2}168{"event": "train", "step": 3240, "loss": -6.03569221496582, "grad_norm": 0.2179240882396698, "lr": 5.578270575938211e-05, "ex_per_s": 311.8825049863731, "data_frac": 0.02161186908212885, "mem_mb": 91.210752, "nll": -6.03569221496582, "elapsed_s": 664.9}169{"event": "train", "step": 3260, "loss": -5.935572624206543, "grad_norm": 0.4176536798477173, "lr": 5.448565001729653e-05, "ex_per_s": 313.0231077208364, "data_frac": 0.021574987137859762, "mem_mb": 91.210752, "nll": -5.935572624206543, "elapsed_s": 666.6}170{"event": "train", "step": 3280, "loss": -6.101743698120117, "grad_norm": 0.15495151281356812, "lr": 5.3218807795017296e-05, "ex_per_s": 314.1580857730005, "data_frac": 0.021537794778585456, "mem_mb": 91.210752, "nll": -6.101743698120117, "elapsed_s": 668.2}171{"event": "train", "step": 3300, "loss": -6.029427528381348, "grad_norm": 0.14668011665344238, "lr": 5.198252543455864e-05, "ex_per_s": 315.2848298100551, "data_frac": 0.021502138976552242, "mem_mb": 91.210752, "nll": -6.029427528381348, "elapsed_s": 669.9}172{"event": "train", "step": 3320, "loss": -6.0077290534973145, "grad_norm": 0.38335496187210083, "lr": 5.07771409231737e-05, "ex_per_s": 316.4084566613482, "data_frac": 0.021465297879677804, "mem_mb": 91.210752, "nll": -6.0077290534973145, "elapsed_s": 671.6}173{"event": "train", "step": 3340, "loss": -5.840697288513184, "grad_norm": 0.2743399143218994, "lr": 4.960298380095221e-05, "ex_per_s": 317.5269357848861, "data_frac": 0.02142860844037099, "mem_mb": 91.210752, "nll": -5.840697288513184, "elapsed_s": 673.2}174{"event": "train", "step": 3360, "loss": -5.977883815765381, "grad_norm": 0.6379706263542175, "lr": 4.846037507072751e-05, "ex_per_s": 318.6398657020545, "data_frac": 0.021391952678264527, "mem_mb": 91.210752, "nll": -5.977883815765381, "elapsed_s": 674.9}175{"event": "train", "step": 3380, "loss": -6.073073387145996, "grad_norm": 0.1310640573501587, "lr": 4.7349627110317556e-05, "ex_per_s": 319.74792511978126, "data_frac": 0.021355622826176392, "mem_mb": 91.210752, "nll": -6.073073387145996, "elapsed_s": 676.6}176{"event": "train", "step": 3400, "loss": -6.146839141845703, "grad_norm": 0.599957287311554, "lr": 4.6271043587123986e-05, "ex_per_s": 320.85006020215087, "data_frac": 0.02131962676633379, "mem_mb": 91.210752, "nll": -6.146839141845703, "elapsed_s": 678.2}177{"event": "train", "step": 3420, "loss": -6.219030857086182, "grad_norm": 1.1495457887649536, "lr": 4.52249193751124e-05, "ex_per_s": 321.94749248711133, "data_frac": 0.021283595745913104, "mem_mb": 91.210752, "nll": -6.219030857086182, "elapsed_s": 679.9}178{"event": "train", "step": 3440, "loss": -6.017535209655762, "grad_norm": 0.5250397324562073, "lr": 4.4211540474196926e-05, "ex_per_s": 323.0391300511273, "data_frac": 0.021247745593084507, "mem_mb": 91.210752, "nll": -6.017535209655762, "elapsed_s": 681.6}179{"event": "train", "step": 3460, "loss": -6.149810314178467, "grad_norm": 0.24079850316047668, "lr": 4.3231183932050546e-05, "ex_per_s": 324.1261316839248, "data_frac": 0.021212392417742233, "mem_mb": 91.210752, "nll": -6.149810314178467, "elapsed_s": 683.2}180{"event": "train", "step": 3480, "loss": -5.897778511047363, "grad_norm": 0.11566495150327682, "lr": 4.228411776836306e-05, "ex_per_s": 325.20866393412524, "data_frac": 0.021177241180399205, "mem_mb": 91.210752, "nll": -5.897778511047363, "elapsed_s": 684.9}181{"event": "train", "step": 3500, "loss": -5.95402193069458, "grad_norm": 0.5003106594085693, "lr": 4.137060090156723e-05, "ex_per_s": 326.2855646077063, "data_frac": 0.021143304440440523, "mem_mb": 91.210752, "nll": -5.95402193069458, "elapsed_s": 686.6}182{"event": "val", "step": 3500, "val_nll": -6.450419456993132, "best": -6.450616433821653, "bad": 1, "mem_mb": 91.210752, "elapsed_s": 749.7}183{"event": "train", "step": 3520, "loss": -5.933611869812012, "grad_norm": 0.20921331644058228, "lr": 4.049088307805306e-05, "ex_per_s": 299.81376337852123, "data_frac": 0.019332599384455796, "mem_mb": 91.210752, "nll": -5.933611869812012, "elapsed_s": 751.4}184{"event": "train", "step": 3540, "loss": -5.956335067749023, "grad_norm": 0.1916598528623581, "lr": 3.964520480388956e-05, "ex_per_s": 300.8460782218439, "data_frac": 0.01930377086052194, "mem_mb": 91.210752, "nll": -5.956335067749023, "elapsed_s": 753.1}185{"event": "train", "step": 3560, "loss": -6.035600185394287, "grad_norm": 0.10926391929388046, "lr": 3.88337972790728e-05, "ex_per_s": 301.87365982549096, "data_frac": 0.01927548394590919, "mem_mb": 91.210752, "nll": -6.035600185394287, "elapsed_s": 754.8}186{"event": "train", "step": 3580, "loss": -5.947732448577881, "grad_norm": 0.3800975978374481, "lr": 3.805688233431824e-05, "ex_per_s": 302.8971293102282, "data_frac": 0.019246837477030026, "mem_mb": 91.210752, "nll": -5.947732448577881, "elapsed_s": 756.5}187{"event": "train", "step": 3600, "loss": -5.96647310256958, "grad_norm": 0.0933176577091217, "lr": 3.7314672370414325e-05, "ex_per_s": 303.91552449549505, "data_frac": 0.019219042041613246, "mem_mb": 91.210752, "nll": -5.96647310256958, "elapsed_s": 758.1}188{"event": "train", "step": 3620, "loss": -5.783346652984619, "grad_norm": 0.3731544017791748, "lr": 3.660737030015427e-05, "ex_per_s": 304.9295113079547, "data_frac": 0.01919079646013145, "mem_mb": 91.210752, "nll": -5.783346652984619, "elapsed_s": 759.8}189{"event": "train", "step": 3640, "loss": -6.0161333084106445, "grad_norm": 0.25016120076179504, "lr": 3.5935169492861716e-05, "ex_per_s": 305.9389684820276, "data_frac": 0.019163871719975748, "mem_mb": 91.210752, "nll": -6.0161333084106445, "elapsed_s": 761.5}190{"event": "train", "step": 3660, "loss": -5.810636520385742, "grad_norm": 0.25203442573547363, "lr": 3.52982537215255e-05, "ex_per_s": 306.94407704817576, "data_frac": 0.019136125397235682, "mem_mb": 91.210752, "nll": -5.810636520385742, "elapsed_s": 763.2}191{"event": "train", "step": 3680, "loss": -5.824997425079346, "grad_norm": 0.1175675094127655, "lr": 3.469679711255794e-05, "ex_per_s": 307.9449518082291, "data_frac": 0.019109184824005834, "mem_mb": 91.210752, "nll": -5.824997425079346, "elapsed_s": 764.9}192{"event": "train", "step": 3700, "loss": -6.093919277191162, "grad_norm": 0.19982127845287323, "lr": 3.41309640981904e-05, "ex_per_s": 308.9417398474746, "data_frac": 0.019082249416200472, "mem_mb": 91.210752, "nll": -6.093919277191162, "elapsed_s": 766.5}193{"event": "train", "step": 3720, "loss": -5.856237888336182, "grad_norm": 0.225404292345047, "lr": 3.3600909371519225e-05, "ex_per_s": 309.934438396725, "data_frac": 0.019055126196121243, "mem_mb": 91.210752, "nll": -5.856237888336182, "elapsed_s": 768.2}194{"event": "train", "step": 3740, "loss": -5.9651265144348145, "grad_norm": 0.1748790442943573, "lr": 3.31067778442141e-05, "ex_per_s": 310.9214539107139, "data_frac": 0.019028978417132844, "mem_mb": 91.210752, "nll": -5.9651265144348145, "elapsed_s": 769.9}195{"event": "train", "step": 3760, "loss": -5.978234767913818, "grad_norm": 0.09083538502454758, "lr": 3.2648704606900735e-05, "ex_per_s": 311.90619435492755, "data_frac": 0.019002405779001922, "mem_mb": 91.210752, "nll": -5.978234767913818, "elapsed_s": 771.6}196{"event": "train", "step": 3780, "loss": -6.139213562011719, "grad_norm": 0.3430973291397095, "lr": 3.222681489222838e-05, "ex_per_s": 312.8875219043131, "data_frac": 0.01897600305622701, "mem_mb": 91.210752, "nll": -6.139213562011719, "elapsed_s": 773.2}197{"event": "train", "step": 3800, "loss": -5.773370265960693, "grad_norm": 0.13468892872333527, "lr": 3.1841224040632484e-05, "ex_per_s": 313.8653878105569, "data_frac": 0.018949602944972607, "mem_mb": 91.210752, "nll": -5.773370265960693, "elapsed_s": 774.9}198{"event": "train", "step": 3820, "loss": -5.992419242858887, "grad_norm": 0.18388356268405914, "lr": 3.149203746880175e-05, "ex_per_s": 314.8380617878795, "data_frac": 0.018923283067785703, "mem_mb": 91.210752, "nll": -5.992419242858887, "elapsed_s": 776.6}199{"event": "train", "step": 3840, "loss": -6.083745956420898, "grad_norm": 0.2144782692193985, "lr": 3.117935064085835e-05, "ex_per_s": 315.8056965534544, "data_frac": 0.018897418177320255, "mem_mb": 91.210752, "nll": -6.083745956420898, "elapsed_s": 778.2}200{"event": "train", "step": 3860, "loss": -5.988083839416504, "grad_norm": 0.3164651393890381, "lr": 3.090324904225885e-05, "ex_per_s": 316.7693152843111, "data_frac": 0.018872382380972202, "mem_mb": 91.210752, "nll": -5.988083839416504, "elapsed_s": 779.9}201{"event": "train", "step": 3880, "loss": -6.050339698791504, "grad_norm": 0.11320865154266357, "lr": 3.0663808156423456e-05, "ex_per_s": 317.7290256744994, "data_frac": 0.018846853889193015, "mem_mb": 91.210752, "nll": -6.050339698791504, "elapsed_s": 781.6}202{"event": "train", "step": 3900, "loss": -5.989501953125, "grad_norm": 0.4598483741283417, "lr": 3.046109344409957e-05, "ex_per_s": 318.6854364188144, "data_frac": 0.018820790099535925, "mem_mb": 91.210752, "nll": -5.989501953125, "elapsed_s": 783.3}203{"event": "train", "step": 3920, "loss": -5.839423179626465, "grad_norm": 0.16585980355739594, "lr": 3.0295160325465435e-05, "ex_per_s": 319.6379557400833, "data_frac": 0.01879545180061939, "mem_mb": 91.210752, "nll": -5.839423179626465, "elapsed_s": 784.9}204{"event": "train", "step": 3940, "loss": -6.041521072387695, "grad_norm": 0.3073951005935669, "lr": 3.016605416497886e-05, "ex_per_s": 320.585414506747, "data_frac": 0.018770430596501343, "mem_mb": 91.210752, "nll": -6.041521072387695, "elapsed_s": 786.6}205{"event": "train", "step": 3960, "loss": -5.85411262512207, "grad_norm": 0.21544599533081055, "lr": 3.0073810258974985e-05, "ex_per_s": 321.5290338334556, "data_frac": 0.0187458045247689, "mem_mb": 91.210752, "nll": -5.85411262512207, "elapsed_s": 788.3}206{"event": "train", "step": 3980, "loss": -5.914238929748535, "grad_norm": 0.14839886128902435, "lr": 3.0018453826016683e-05, "ex_per_s": 322.4681305591525, "data_frac": 0.018721934919383506, "mem_mb": 91.210752, "nll": -5.914238929748535, "elapsed_s": 790.0}207{"event": "train", "step": 4000, "loss": -5.942416667938232, "grad_norm": 0.6880264282226562, "lr": 2.9999999999999997e-05, "ex_per_s": 323.4041070909109, "data_frac": 0.018697128483112887, "mem_mb": 91.210752, "nll": -5.942416667938232, "elapsed_s": 791.6}208{"event": "val", "step": 4000, "val_nll": -6.449540800666577, "best": -6.450616433821653, "bad": 2, "mem_mb": 91.210752, "elapsed_s": 854.6}209{"event": "done", "steps": 4000, "best_val": -6.450616433821653, "wall_s": 854.6393349170685, "ex_per_s": 299.54156044647937, "peak_mem_mb": 91.210752, "n_params": 3455619, "device": "mps", "data_time_frac": 0.017317550763588777, "elapsed_s": 854.6}210
Files · main
Menu

Theme