SPB Git

spb/wp41_uqo

Public
50.6 KB · 210 lines
1{"event": "train", "step": 20, "loss": -6.119615077972412, "grad_norm": 1.1472111940383911, "lr": 3.149999999999999e-05, "ex_per_s": 97.42686446450303, "data_frac": 0.9047030707802327, "mem_mb": 87.04, "nll": -6.119615077972412, "elapsed_s": 13.2}2{"event": "train", "step": 40, "loss": -5.901360988616943, "grad_norm": 0.25533005595207214, "lr": 6.149999999999999e-05, "ex_per_s": 180.1784302752409, "data_frac": 0.8374164859275169, "mem_mb": 87.04, "nll": -5.901360988616943, "elapsed_s": 14.2}3{"event": "train", "step": 60, "loss": -5.979992389678955, "grad_norm": 0.5383658409118652, "lr": 9.149999999999999e-05, "ex_per_s": 251.78144767778707, "data_frac": 0.7808707235831434, "mem_mb": 87.04, "nll": -5.979992389678955, "elapsed_s": 15.3}4{"event": "train", "step": 80, "loss": -6.070039749145508, "grad_norm": 1.3372337818145752, "lr": 0.0001215, "ex_per_s": 314.283884410966, "data_frac": 0.7316877240423939, "mem_mb": 87.04, "nll": -6.070039749145508, "elapsed_s": 16.3}5{"event": "train", "step": 100, "loss": -5.951157093048096, "grad_norm": 0.2783014476299286, "lr": 0.0001515, "ex_per_s": 369.297949467453, "data_frac": 0.6884130352991752, "mem_mb": 87.04, "nll": -5.951157093048096, "elapsed_s": 17.4}6{"event": "train", "step": 120, "loss": -5.911116600036621, "grad_norm": 0.7938029766082764, "lr": 0.00018149999999999997, "ex_per_s": 417.93823533605587, "data_frac": 0.6498163931616078, "mem_mb": 87.04, "nll": -5.911116600036621, "elapsed_s": 18.4}7{"event": "train", "step": 140, "loss": -5.833921432495117, "grad_norm": 0.41110730171203613, "lr": 0.00021149999999999996, "ex_per_s": 461.4936565222246, "data_frac": 0.6155635026430886, "mem_mb": 87.04, "nll": -5.833921432495117, "elapsed_s": 19.4}8{"event": "train", "step": 160, "loss": -5.7349371910095215, "grad_norm": 0.9349613189697266, "lr": 0.0002415, "ex_per_s": 500.56993683668367, "data_frac": 0.5847206034511524, "mem_mb": 87.04, "nll": -5.7349371910095215, "elapsed_s": 20.5}9{"event": "train", "step": 180, "loss": -5.975215435028076, "grad_norm": 1.1550108194351196, "lr": 0.0002715, "ex_per_s": 535.8565712344247, "data_frac": 0.5568163393462624, "mem_mb": 87.04, "nll": -5.975215435028076, "elapsed_s": 21.5}10{"event": "train", "step": 200, "loss": -5.96942138671875, "grad_norm": 0.9434659481048584, "lr": 0.0003, "ex_per_s": 567.8704006589811, "data_frac": 0.5314828054990434, "mem_mb": 87.04, "nll": -5.96942138671875, "elapsed_s": 22.6}11{"event": "train", "step": 220, "loss": -6.027349948883057, "grad_norm": 1.3355013132095337, "lr": 0.00029998154617398326, "ex_per_s": 596.9974082849319, "data_frac": 0.5083291205596753, "mem_mb": 87.04, "nll": -6.027349948883057, "elapsed_s": 23.6}12{"event": "train", "step": 240, "loss": -5.913966178894043, "grad_norm": 0.3661142587661743, "lr": 0.000299926189741025, "ex_per_s": 623.4338017901072, "data_frac": 0.48742110514592807, "mem_mb": 87.04, "nll": -5.913966178894043, "elapsed_s": 24.7}13{"event": "train", "step": 260, "loss": -5.888825416564941, "grad_norm": 0.10295317322015762, "lr": 0.0002998339458350211, "ex_per_s": 647.9807252125054, "data_frac": 0.46800855283971815, "mem_mb": 87.04, "nll": -5.888825416564941, "elapsed_s": 25.7}14{"event": "train", "step": 280, "loss": -5.945367336273193, "grad_norm": 0.15692193806171417, "lr": 0.0002997048396745345, "ex_per_s": 670.6467376911172, "data_frac": 0.450142062196, "mem_mb": 87.04, "nll": -5.945367336273193, "elapsed_s": 26.7}15{"event": "train", "step": 300, "loss": -5.9680352210998535, "grad_norm": 0.22284311056137085, "lr": 0.0002995389065559004, "ex_per_s": 691.6298921810514, "data_frac": 0.4336246632668083, "mem_mb": 87.04, "nll": -5.9680352210998535, "elapsed_s": 27.8}16{"event": "train", "step": 320, "loss": -5.966297626495361, "grad_norm": 0.584196150302887, "lr": 0.0002993361918435765, "ex_per_s": 711.0850873510526, "data_frac": 0.41829782880808253, "mem_mb": 87.04, "nll": -5.966297626495361, "elapsed_s": 28.8}17{"event": "train", "step": 340, "loss": -5.835437297821045, "grad_norm": 0.408099889755249, "lr": 0.00029909675095774114, "ex_per_s": 729.1702979687257, "data_frac": 0.404040873554566, "mem_mb": 87.04, "nll": -5.835437297821045, "elapsed_s": 29.9}18{"event": "train", "step": 360, "loss": -5.995971202850342, "grad_norm": 0.2877524793148041, "lr": 0.00029882064935914164, "ex_per_s": 745.9942373875297, "data_frac": 0.3907216361066014, "mem_mb": 87.04, "nll": -5.995971202850342, "elapsed_s": 30.9}19{"event": "train", "step": 380, "loss": -6.005551815032959, "grad_norm": 1.1697487831115723, "lr": 0.00029850796253119824, "ex_per_s": 761.8409424351084, "data_frac": 0.3783201888839942, "mem_mb": 87.04, "nll": -6.005551815032959, "elapsed_s": 31.9}20{"event": "train", "step": 400, "loss": -5.995796203613281, "grad_norm": 1.3492388725280762, "lr": 0.0002981587759593675, "ex_per_s": 776.6878535706238, "data_frac": 0.3667037215599246, "mem_mb": 87.04, "nll": -5.995796203613281, "elapsed_s": 33.0}21{"event": "train", "step": 420, "loss": -5.964417934417725, "grad_norm": 0.6013784408569336, "lr": 0.0002977731851077716, "ex_per_s": 790.6947858312503, "data_frac": 0.3558206684854892, "mem_mb": 87.04, "nll": -5.964417934417725, "elapsed_s": 34.0}22{"event": "train", "step": 440, "loss": -5.93910026550293, "grad_norm": 0.7144365310668945, "lr": 0.0002973512953930992, "ex_per_s": 803.839658409458, "data_frac": 0.3455691098467227, "mem_mb": 87.04, "nll": -5.93910026550293, "elapsed_s": 35.1}23{"event": "train", "step": 460, "loss": -6.102421760559082, "grad_norm": 0.5779339075088501, "lr": 0.0002968932221557859, "ex_per_s": 816.2341678202853, "data_frac": 0.3359036977245236, "mem_mb": 87.04, "nll": -6.102421760559082, "elapsed_s": 36.1}24{"event": "train", "step": 480, "loss": -5.878465175628662, "grad_norm": 0.2829596698284149, "lr": 0.0002963990906284808, "ex_per_s": 827.933995558346, "data_frac": 0.32677762685939343, "mem_mb": 87.04, "nll": -5.878465175628662, "elapsed_s": 37.1}25{"event": "train", "step": 500, "loss": -5.925892353057861, "grad_norm": 0.6030600070953369, "lr": 0.00029586903590180956, "ex_per_s": 838.992726362516, "data_frac": 0.31814667657647494, "mem_mb": 87.04, "nll": -5.925892353057861, "elapsed_s": 38.2}26{"event": "val", "step": 500, "val_nll": -6.43959913024846, "best": -6.43959913024846, "bad": 0, "mem_mb": 87.04, "elapsed_s": 101.0}27{"event": "train", "step": 520, "loss": -6.040494918823242, "grad_norm": 0.36138930916786194, "lr": 0.00029530320288744205, "ex_per_s": 325.9388655897531, "data_frac": 0.11893625512150265, "mem_mb": 87.04, "nll": -6.040494918823242, "elapsed_s": 102.1}28{"event": "train", "step": 540, "loss": -5.879533767700195, "grad_norm": 0.6168885827064514, "lr": 0.0002947017462784745, "ex_per_s": 335.0347399160215, "data_frac": 0.11782036739296306, "mem_mb": 87.04, "nll": -5.879533767700195, "elapsed_s": 103.2}29{"event": "train", "step": 560, "loss": -5.972998142242432, "grad_norm": 0.1831420361995697, "lr": 0.00029406483050713824, "ex_per_s": 343.95049859221984, "data_frac": 0.11672713785384461, "mem_mb": 87.04, "nll": -5.972998142242432, "elapsed_s": 104.2}30{"event": "train", "step": 580, "loss": -5.9249162673950195, "grad_norm": 0.13599076867103577, "lr": 0.0002933926296998457, "ex_per_s": 352.6882607214253, "data_frac": 0.11565480359547176, "mem_mb": 87.04, "nll": -5.9249162673950195, "elapsed_s": 105.3}31{"event": "train", "step": 600, "loss": -5.972914695739746, "grad_norm": 0.7049604654312134, "lr": 0.0002926853276295856, "ex_per_s": 361.2536696951482, "data_frac": 0.1146028436800017, "mem_mb": 87.04, "nll": -5.972914695739746, "elapsed_s": 106.3}32{"event": "train", "step": 620, "loss": -5.951656818389893, "grad_norm": 0.463882714509964, "lr": 0.00029194311766568174, "ex_per_s": 369.64898576874356, "data_frac": 0.1135703238522797, "mem_mb": 87.04, "nll": -5.951656818389893, "elapsed_s": 107.4}33{"event": "train", "step": 640, "loss": -6.065169334411621, "grad_norm": 0.39711618423461914, "lr": 0.0002911662027209272, "ex_per_s": 377.8867212227948, "data_frac": 0.11256027666287331, "mem_mb": 87.04, "nll": -6.065169334411621, "elapsed_s": 108.4}34{"event": "train", "step": 660, "loss": -6.006849765777588, "grad_norm": 1.0623191595077515, "lr": 0.00029035479519611045, "ex_per_s": 385.9642452962785, "data_frac": 0.11156887926880686, "mem_mb": 87.04, "nll": -6.006849765777588, "elapsed_s": 109.5}35{"event": "train", "step": 680, "loss": -6.007784843444824, "grad_norm": 0.103029765188694, "lr": 0.0002895091169219469, "ex_per_s": 393.88621550838945, "data_frac": 0.11059494220289279, "mem_mb": 87.04, "nll": -6.007784843444824, "elapsed_s": 110.5}36{"event": "train", "step": 700, "loss": -5.98783016204834, "grad_norm": 0.5830191373825073, "lr": 0.00028862939909843273, "ex_per_s": 401.6646050656893, "data_frac": 0.10964197576389716, "mem_mb": 87.04, "nll": -5.98783016204834, "elapsed_s": 111.6}37{"event": "train", "step": 720, "loss": -6.063215255737305, "grad_norm": 0.2581166625022888, "lr": 0.0002877158822316369, "ex_per_s": 409.2952035401951, "data_frac": 0.10870490500036163, "mem_mb": 87.04, "nll": -6.063215255737305, "elapsed_s": 112.6}38{"event": "train", "step": 740, "loss": -6.033407211303711, "grad_norm": 0.2590143084526062, "lr": 0.00028676881606794945, "ex_per_s": 416.78685369700315, "data_frac": 0.10778581678982253, "mem_mb": 87.04, "nll": -6.033407211303711, "elapsed_s": 113.7}39{"event": "train", "step": 760, "loss": -5.958136081695557, "grad_norm": 0.8965355753898621, "lr": 0.000285788459525803, "ex_per_s": 424.14463154640947, "data_frac": 0.10688536289302013, "mem_mb": 87.04, "nll": -5.958136081695557, "elapsed_s": 114.7}40{"event": "train", "step": 780, "loss": -5.895252704620361, "grad_norm": 0.24063441157341003, "lr": 0.00028477508062488754, "ex_per_s": 431.36470334772093, "data_frac": 0.10600093011184171, "mem_mb": 87.04, "nll": -5.895252704620361, "elapsed_s": 115.8}41{"event": "train", "step": 800, "loss": -5.81594181060791, "grad_norm": 0.2322651594877243, "lr": 0.000283728956412876, "ex_per_s": 438.45828644031343, "data_frac": 0.10513588221812091, "mem_mb": 87.04, "nll": -5.81594181060791, "elapsed_s": 116.8}42{"event": "train", "step": 820, "loss": -5.93446159362793, "grad_norm": 0.5022519826889038, "lr": 0.00028265037288968236, "ex_per_s": 445.43121655920453, "data_frac": 0.10428488995353284, "mem_mb": 87.04, "nll": -5.93446159362793, "elapsed_s": 117.8}43{"event": "train", "step": 840, "loss": -5.8066020011901855, "grad_norm": 0.7184305191040039, "lr": 0.0002815396249292725, "ex_per_s": 452.27499003748875, "data_frac": 0.10344938568117738, "mem_mb": 87.04, "nll": -5.8066020011901855, "elapsed_s": 118.9}44{"event": "train", "step": 860, "loss": -5.915968894958496, "grad_norm": 1.4325928688049316, "lr": 0.00028039701619904774, "ex_per_s": 458.99869798809385, "data_frac": 0.10263338245594181, "mem_mb": 87.04, "nll": -5.915968894958496, "elapsed_s": 119.9}45{"event": "train", "step": 880, "loss": -5.8717498779296875, "grad_norm": 0.2285817265510559, "lr": 0.00027922285907682625, "ex_per_s": 465.60294073843824, "data_frac": 0.10182618225632092, "mem_mb": 87.04, "nll": -5.8717498779296875, "elapsed_s": 121.0}46{"event": "train", "step": 900, "loss": -5.9877543449401855, "grad_norm": 0.25965768098831177, "lr": 0.00027801747456544134, "ex_per_s": 472.1021666424029, "data_frac": 0.10103600388090356, "mem_mb": 87.04, "nll": -5.9877543449401855, "elapsed_s": 122.0}47{"event": "train", "step": 920, "loss": -6.02508544921875, "grad_norm": 0.6515376567840576, "lr": 0.00027678119220498265, "ex_per_s": 478.4918483181546, "data_frac": 0.10025743866374162, "mem_mb": 87.04, "nll": -6.02508544921875, "elapsed_s": 123.1}48{"event": "train", "step": 940, "loss": -6.086118221282959, "grad_norm": 0.3991756737232208, "lr": 0.0002755143499827035, "ex_per_s": 484.7775731772618, "data_frac": 0.0994905380101489, "mem_mb": 87.04, "nll": -6.086118221282959, "elapsed_s": 124.1}49{"event": "train", "step": 960, "loss": -5.968112468719482, "grad_norm": 0.13628406822681427, "lr": 0.00027421729424061787, "ex_per_s": 490.9518434396184, "data_frac": 0.09873537215037982, "mem_mb": 87.04, "nll": -5.968112468719482, "elapsed_s": 125.2}50{"event": "train", "step": 980, "loss": -5.926452159881592, "grad_norm": 0.162456676363945, "lr": 0.00027289037958081483, "ex_per_s": 497.0301094255349, "data_frac": 0.09799539762980759, "mem_mb": 87.04, "nll": -5.926452159881592, "elapsed_s": 126.2}51{"event": "train", "step": 1000, "loss": -6.073229789733887, "grad_norm": 0.26093387603759766, "lr": 0.0002715339687685131, "ex_per_s": 503.00835908583974, "data_frac": 0.09726720211742296, "mem_mb": 87.04, "nll": -6.073229789733887, "elapsed_s": 127.3}52{"event": "val", "step": 1000, "val_nll": -6.4474136287521455, "best": -6.4474136287521455, "bad": 0, "mem_mb": 87.04, "elapsed_s": 189.4}53{"event": "train", "step": 1020, "loss": -6.031726360321045, "grad_norm": 0.12809501588344574, "lr": 0.0002701484326328852, "ex_per_s": 342.7682516360012, "data_frac": 0.06503411110648084, "mem_mb": 87.04, "nll": -6.031726360321045, "elapsed_s": 190.5}54{"event": "train", "step": 1040, "loss": -5.971897602081299, "grad_norm": 0.3107963502407074, "lr": 0.00026873414996567596, "ex_per_s": 347.5746852955716, "data_frac": 0.0647291905018748, "mem_mb": 87.04, "nll": -5.971897602081299, "elapsed_s": 191.5}55{"event": "train", "step": 1060, "loss": -6.182488918304443, "grad_norm": 0.14623722434043884, "lr": 0.0002672915074176452, "ex_per_s": 352.33010786329226, "data_frac": 0.06442782407271207, "mem_mb": 87.04, "nll": -6.182488918304443, "elapsed_s": 192.6}56{"event": "train", "step": 1080, "loss": -5.815884590148926, "grad_norm": 0.4739453196525574, "lr": 0.00026582089939286075, "ex_per_s": 357.0335178358918, "data_frac": 0.06412869946437712, "mem_mb": 87.04, "nll": -5.815884590148926, "elapsed_s": 193.6}57{"event": "train", "step": 1100, "loss": -6.053974151611328, "grad_norm": 0.1078750267624855, "lr": 0.0002643227279408728, "ex_per_s": 361.68741207441826, "data_frac": 0.06383261777072467, "mem_mb": 87.04, "nll": -6.053974151611328, "elapsed_s": 194.7}58{"event": "train", "step": 1120, "loss": -6.0527520179748535, "grad_norm": 0.35383832454681396, "lr": 0.000262797402646797, "ex_per_s": 366.2918888971138, "data_frac": 0.06354140156802583, "mem_mb": 87.04, "nll": -6.0527520179748535, "elapsed_s": 195.7}59{"event": "train", "step": 1140, "loss": -6.019047260284424, "grad_norm": 0.2714374363422394, "lr": 0.0002612453405193386, "ex_per_s": 370.85001348094715, "data_frac": 0.06325265023769767, "mem_mb": 87.04, "nll": -6.019047260284424, "elapsed_s": 196.8}60{"event": "train", "step": 1160, "loss": -5.969729900360107, "grad_norm": 0.5510554313659668, "lr": 0.0002596669658767859, "ex_per_s": 375.35774372384327, "data_frac": 0.06296720732251034, "mem_mb": 87.04, "nll": -5.969729900360107, "elapsed_s": 197.8}61{"event": "train", "step": 1180, "loss": -5.9577813148498535, "grad_norm": 0.14138782024383545, "lr": 0.00025806271023100637, "ex_per_s": 379.8187659430967, "data_frac": 0.06268386168143504, "mem_mb": 87.04, "nll": -5.9577813148498535, "elapsed_s": 198.9}62{"event": "train", "step": 1200, "loss": -5.944911003112793, "grad_norm": 0.24028746783733368, "lr": 0.00025643301216947504, "ex_per_s": 384.2257863675813, "data_frac": 0.06240641011602148, "mem_mb": 87.04, "nll": -5.944911003112793, "elapsed_s": 199.9}63{"event": "train", "step": 1220, "loss": -6.033281326293945, "grad_norm": 0.586449384689331, "lr": 0.0002547783172353695, "ex_per_s": 388.59376706650755, "data_frac": 0.06213009313087525, "mem_mb": 87.04, "nll": -6.033281326293945, "elapsed_s": 201.0}64{"event": "train", "step": 1240, "loss": -6.058996677398682, "grad_norm": 0.0708232894539833, "lr": 0.00025309907780576277, "ex_per_s": 392.9157936557887, "data_frac": 0.0618559893042658, "mem_mb": 87.04, "nll": -6.058996677398682, "elapsed_s": 202.0}65{"event": "train", "step": 1260, "loss": -6.054467678070068, "grad_norm": 0.19770516455173492, "lr": 0.0002513957529679477, "ex_per_s": 397.1938430501902, "data_frac": 0.061584561295837476, "mem_mb": 87.04, "nll": -6.054467678070068, "elapsed_s": 203.1}66{"event": "train", "step": 1280, "loss": -6.04311990737915, "grad_norm": 0.5940806865692139, "lr": 0.00024966880839392746, "ex_per_s": 401.4279228000933, "data_frac": 0.06131605038844838, "mem_mb": 87.04, "nll": -6.04311990737915, "elapsed_s": 204.1}67{"event": "train", "step": 1300, "loss": -5.868980884552002, "grad_norm": 0.5589854121208191, "lr": 0.0002479187162131052, "ex_per_s": 405.6163036095865, "data_frac": 0.061053766690630305, "mem_mb": 87.04, "nll": -5.868980884552002, "elapsed_s": 205.1}68{"event": "train", "step": 1320, "loss": -5.830454349517822, "grad_norm": 0.1983889639377594, "lr": 0.00024614595488320833, "ex_per_s": 409.75733266409713, "data_frac": 0.06079465633999823, "mem_mb": 87.04, "nll": -5.830454349517822, "elapsed_s": 206.2}69{"event": "train", "step": 1340, "loss": -5.951794624328613, "grad_norm": 0.49315333366394043, "lr": 0.00024435100905948387, "ex_per_s": 413.8655572505596, "data_frac": 0.06053524688202376, "mem_mb": 87.04, "nll": -5.951794624328613, "elapsed_s": 207.2}70{"event": "train", "step": 1360, "loss": -6.040543079376221, "grad_norm": 0.5307888388633728, "lr": 0.0002425343694621974, "ex_per_s": 417.93496636502584, "data_frac": 0.060278384345070816, "mem_mb": 87.04, "nll": -6.040543079376221, "elapsed_s": 208.3}71{"event": "train", "step": 1380, "loss": -6.069918155670166, "grad_norm": 0.27632570266723633, "lr": 0.00024069653274247583, "ex_per_s": 421.962167270441, "data_frac": 0.06002342209930659, "mem_mb": 87.04, "nll": -6.069918155670166, "elapsed_s": 209.3}72{"event": "train", "step": 1400, "loss": -5.796263694763184, "grad_norm": 0.10512220114469528, "lr": 0.0002388380013465276, "ex_per_s": 425.947575703343, "data_frac": 0.059771875039308184, "mem_mb": 87.04, "nll": -5.796263694763184, "elapsed_s": 210.4}73{"event": "train", "step": 1420, "loss": -5.988466739654541, "grad_norm": 0.7622658014297485, "lr": 0.00023695928337827914, "ex_per_s": 429.88828541302684, "data_frac": 0.05952351505036243, "mem_mb": 87.04, "nll": -5.988466739654541, "elapsed_s": 211.4}74{"event": "train", "step": 1440, "loss": -5.743991374969482, "grad_norm": 0.30678337812423706, "lr": 0.00023506089246046403, "ex_per_s": 433.79006360893203, "data_frac": 0.05928107590863881, "mem_mb": 87.04, "nll": -5.743991374969482, "elapsed_s": 212.5}75{"event": "train", "step": 1460, "loss": -5.801860809326172, "grad_norm": 0.3518563508987427, "lr": 0.00023314334759420388, "ex_per_s": 437.6475328876936, "data_frac": 0.059059791307766014, "mem_mb": 87.04, "nll": -5.801860809326172, "elapsed_s": 213.5}76{"event": "train", "step": 1480, "loss": -6.046921253204346, "grad_norm": 0.1288769543170929, "lr": 0.0002312071730171186, "ex_per_s": 441.46446568440064, "data_frac": 0.05883712075531497, "mem_mb": 87.04, "nll": -6.046921253204346, "elapsed_s": 214.6}77{"event": "train", "step": 1500, "loss": -5.95939302444458, "grad_norm": 0.39597898721694946, "lr": 0.0002292528980600049, "ex_per_s": 445.2556339950976, "data_frac": 0.05861062727809929, "mem_mb": 87.04, "nll": -5.95939302444458, "elapsed_s": 215.6}78{"event": "val", "step": 1500, "val_nll": -6.446964149618753, "best": -6.4474136287521455, "bad": 1, "mem_mb": 87.04, "elapsed_s": 278.0}79{"event": "train", "step": 1520, "loss": -5.950689792633057, "grad_norm": 0.6414943933486938, "lr": 0.00022728105700212285, "ex_per_s": 348.5482540702528, "data_frac": 0.0453132069362438, "mem_mb": 87.04, "nll": -5.950689792633057, "elapsed_s": 279.1}80{"event": "train", "step": 1540, "loss": -5.879042625427246, "grad_norm": 0.12087566405534744, "lr": 0.00022529218892512886, "ex_per_s": 351.8111114486328, "data_frac": 0.04518099383438117, "mem_mb": 87.04, "nll": -5.879042625427246, "elapsed_s": 280.2}81{"event": "train", "step": 1560, "loss": -5.98291540145874, "grad_norm": 0.3042891025543213, "lr": 0.00022328683756569652, "ex_per_s": 355.0521716686144, "data_frac": 0.04504754304434296, "mem_mb": 87.04, "nll": -5.98291540145874, "elapsed_s": 281.2}82{"event": "train", "step": 1580, "loss": -5.98119592666626, "grad_norm": 0.09364987164735794, "lr": 0.0002212655511668642, "ex_per_s": 358.26677726999566, "data_frac": 0.044914440349132265, "mem_mb": 87.04, "nll": -5.98119592666626, "elapsed_s": 282.3}83{"event": "train", "step": 1600, "loss": -5.9569783210754395, "grad_norm": 0.21926671266555786, "lr": 0.00021922888232815089, "ex_per_s": 361.46001864035316, "data_frac": 0.04478181626175426, "mem_mb": 87.04, "nll": -5.9569783210754395, "elapsed_s": 283.3}84{"event": "train", "step": 1620, "loss": -5.881229400634766, "grad_norm": 0.10480878502130508, "lr": 0.00021717738785448102, "ex_per_s": 364.6285735135267, "data_frac": 0.044650268902244133, "mem_mb": 87.04, "nll": -5.881229400634766, "elapsed_s": 284.4}85{"event": "train", "step": 1640, "loss": -6.0777177810668945, "grad_norm": 0.10108496993780136, "lr": 0.00021511162860395977, "ex_per_s": 367.7745225629641, "data_frac": 0.04451990327164222, "mem_mb": 87.04, "nll": -6.0777177810668945, "elapsed_s": 285.4}86{"event": "train", "step": 1660, "loss": -6.0403218269348145, "grad_norm": 0.7576424479484558, "lr": 0.0002130321693345397, "ex_per_s": 370.89767810294717, "data_frac": 0.044390961072116944, "mem_mb": 87.04, "nll": -6.0403218269348145, "elapsed_s": 286.5}87{"event": "train", "step": 1680, "loss": -5.866042137145996, "grad_norm": 0.15602430701255798, "lr": 0.0002109395785496218, "ex_per_s": 373.99863137228203, "data_frac": 0.044262408911917975, "mem_mb": 87.04, "nll": -5.866042137145996, "elapsed_s": 287.5}88{"event": "train", "step": 1700, "loss": -5.928844928741455, "grad_norm": 0.4501587152481079, "lr": 0.00020883442834263225, "ex_per_s": 377.0755867703141, "data_frac": 0.044134947145250035, "mem_mb": 87.04, "nll": -5.928844928741455, "elapsed_s": 288.6}89{"event": "train", "step": 1720, "loss": -5.887063503265381, "grad_norm": 0.8516375422477722, "lr": 0.00020671729424061788, "ex_per_s": 380.13176798496056, "data_frac": 0.044008957103471066, "mem_mb": 87.04, "nll": -5.887063503265381, "elapsed_s": 289.6}90{"event": "train", "step": 1740, "loss": -6.273763179779053, "grad_norm": 0.3694813549518585, "lr": 0.00020458875504690275, "ex_per_s": 383.16474966911477, "data_frac": 0.04388292257778543, "mem_mb": 87.04, "nll": -6.273763179779053, "elapsed_s": 290.7}91{"event": "train", "step": 1760, "loss": -5.930388450622559, "grad_norm": 0.2435314953327179, "lr": 0.000202449392682849, "ex_per_s": 386.1774326678764, "data_frac": 0.043758057627534666, "mem_mb": 87.04, "nll": -5.930388450622559, "elapsed_s": 291.7}92{"event": "train", "step": 1780, "loss": -5.871505260467529, "grad_norm": 0.4761614203453064, "lr": 0.00020029979202876563, "ex_per_s": 389.1685963167462, "data_frac": 0.043634389902130935, "mem_mb": 87.04, "nll": -5.871505260467529, "elapsed_s": 292.8}93{"event": "train", "step": 1800, "loss": -6.047906875610352, "grad_norm": 1.1375110149383545, "lr": 0.00019814054076400787, "ex_per_s": 392.137362487686, "data_frac": 0.04351152947689346, "mem_mb": 87.04, "nll": -6.047906875610352, "elapsed_s": 293.8}94{"event": "train", "step": 1820, "loss": -5.934260845184326, "grad_norm": 0.9085363149642944, "lr": 0.00019597222920631147, "ex_per_s": 395.0851834948086, "data_frac": 0.04338980685100685, "mem_mb": 87.04, "nll": -5.934260845184326, "elapsed_s": 294.8}95{"event": "train", "step": 1840, "loss": -6.072961807250977, "grad_norm": 0.4065357744693756, "lr": 0.0001937954501504058, "ex_per_s": 398.01376659700634, "data_frac": 0.04326918062990926, "mem_mb": 87.04, "nll": -6.072961807250977, "elapsed_s": 295.9}96{"event": "train", "step": 1860, "loss": -6.0517778396606445, "grad_norm": 0.16695263981819153, "lr": 0.0001916107987059498, "ex_per_s": 400.9230754791837, "data_frac": 0.04315002719538534, "mem_mb": 87.04, "nll": -6.0517778396606445, "elapsed_s": 296.9}97{"event": "train", "step": 1880, "loss": -5.947168827056885, "grad_norm": 0.4264964759349823, "lr": 0.00018941887213483484, "ex_per_s": 403.81033559472735, "data_frac": 0.04303215258373481, "mem_mb": 87.04, "nll": -5.947168827056885, "elapsed_s": 298.0}98{"event": "train", "step": 1900, "loss": -5.982675552368164, "grad_norm": 0.25677943229675293, "lr": 0.00018722026968789906, "ex_per_s": 406.68047578989234, "data_frac": 0.042913571506912454, "mem_mb": 87.04, "nll": -5.982675552368164, "elapsed_s": 299.0}99{"event": "train", "step": 1920, "loss": -5.9642791748046875, "grad_norm": 0.40626761317253113, "lr": 0.00018501559244109864, "ex_per_s": 409.5280647190599, "data_frac": 0.042797248984051184, "mem_mb": 87.04, "nll": -5.9642791748046875, "elapsed_s": 300.1}100{"event": "train", "step": 1940, "loss": -6.049821853637695, "grad_norm": 0.2240692675113678, "lr": 0.00018280544313117932, "ex_per_s": 412.3571849968655, "data_frac": 0.042682796978803, "mem_mb": 87.04, "nll": -6.049821853637695, "elapsed_s": 301.1}101{"event": "train", "step": 1960, "loss": -5.938291549682617, "grad_norm": 0.08028226345777512, "lr": 0.00018059042599089415, "ex_per_s": 415.1649216568842, "data_frac": 0.042569093230903805, "mem_mb": 87.04, "nll": -5.938291549682617, "elapsed_s": 302.2}102{"event": "train", "step": 1980, "loss": -5.965019226074219, "grad_norm": 0.5312119126319885, "lr": 0.00017837114658381247, "ex_per_s": 417.9570137923497, "data_frac": 0.04245409356514114, "mem_mb": 87.04, "nll": -5.965019226074219, "elapsed_s": 303.2}103{"event": "train", "step": 2000, "loss": -5.914764881134033, "grad_norm": 0.43267321586608887, "lr": 0.00017614821163876486, "ex_per_s": 420.73201271118063, "data_frac": 0.04233995919259847, "mem_mb": 87.04, "nll": -5.914764881134033, "elapsed_s": 304.3}104{"event": "val", "step": 2000, "val_nll": -6.449428591413162, "best": -6.449428591413162, "bad": 0, "mem_mb": 87.04, "elapsed_s": 366.4}105{"event": "train", "step": 2020, "loss": -5.926745891571045, "grad_norm": 0.44131433963775635, "lr": 0.00017392222888396941, "ex_per_s": 351.78931147986464, "data_frac": 0.035077099339005886, "mem_mb": 87.04, "nll": -5.926745891571045, "elapsed_s": 367.5}106{"event": "train", "step": 2040, "loss": -5.873067855834961, "grad_norm": 0.4074225127696991, "lr": 0.000171693806880885, "ex_per_s": 354.2610505208139, "data_frac": 0.03500345137912596, "mem_mb": 87.04, "nll": -5.873067855834961, "elapsed_s": 368.6}107{"event": "train", "step": 2060, "loss": -6.143547534942627, "grad_norm": 0.39751240611076355, "lr": 0.0001694635548578366, "ex_per_s": 356.7179247764465, "data_frac": 0.03492994905377225, "mem_mb": 87.04, "nll": -6.143547534942627, "elapsed_s": 369.6}108{"event": "train", "step": 2080, "loss": -5.96115779876709, "grad_norm": 0.3052837550640106, "lr": 0.0001672320825434584, "ex_per_s": 359.1605013270115, "data_frac": 0.03485681876379437, "mem_mb": 87.04, "nll": -5.96115779876709, "elapsed_s": 370.7}109{"event": "train", "step": 2100, "loss": -5.8991007804870605, "grad_norm": 1.077431082725525, "lr": 0.000165, "ex_per_s": 361.59163637429225, "data_frac": 0.03478435478056541, "mem_mb": 87.04, "nll": -5.8991007804870605, "elapsed_s": 371.7}110{"event": "train", "step": 2120, "loss": -6.151405334472656, "grad_norm": 0.5631482601165771, "lr": 0.0001627679174565416, "ex_per_s": 364.00882962761574, "data_frac": 0.0347121475859664, "mem_mb": 87.04, "nll": -6.151405334472656, "elapsed_s": 372.8}111{"event": "train", "step": 2140, "loss": -5.977453708648682, "grad_norm": 0.10758768022060394, "lr": 0.00016053644514216336, "ex_per_s": 366.41329260547064, "data_frac": 0.03464007142704537, "mem_mb": 87.04, "nll": -5.977453708648682, "elapsed_s": 373.8}112{"event": "train", "step": 2160, "loss": -5.926970481872559, "grad_norm": 0.48392271995544434, "lr": 0.00015830619311911498, "ex_per_s": 368.80343578618454, "data_frac": 0.03456909640494594, "mem_mb": 87.04, "nll": -5.926970481872559, "elapsed_s": 374.9}113{"event": "train", "step": 2180, "loss": -6.084055423736572, "grad_norm": 0.20775550603866577, "lr": 0.00015607777111603058, "ex_per_s": 371.1806774753066, "data_frac": 0.034498344697351, "mem_mb": 87.04, "nll": -6.084055423736572, "elapsed_s": 375.9}114{"event": "train", "step": 2200, "loss": -5.964260578155518, "grad_norm": 0.5382903218269348, "lr": 0.00015385178836123514, "ex_per_s": 373.54676264432123, "data_frac": 0.034428219257502723, "mem_mb": 87.04, "nll": -5.964260578155518, "elapsed_s": 377.0}115{"event": "train", "step": 2220, "loss": -5.826135635375977, "grad_norm": 0.7455757260322571, "lr": 0.00015162885341618752, "ex_per_s": 375.8971108429543, "data_frac": 0.034357430848258395, "mem_mb": 87.04, "nll": -5.826135635375977, "elapsed_s": 378.0}116{"event": "train", "step": 2240, "loss": -5.963090896606445, "grad_norm": 0.7711625695228577, "lr": 0.00014940957400910585, "ex_per_s": 378.2349639642339, "data_frac": 0.034287973885477954, "mem_mb": 87.04, "nll": -5.963090896606445, "elapsed_s": 379.0}117{"event": "train", "step": 2260, "loss": -5.738098621368408, "grad_norm": 0.8115771412849426, "lr": 0.00014719455686882068, "ex_per_s": 380.5603039165787, "data_frac": 0.034219267644778444, "mem_mb": 87.04, "nll": -5.738098621368408, "elapsed_s": 380.1}118{"event": "train", "step": 2280, "loss": -6.010741710662842, "grad_norm": 0.41634348034858704, "lr": 0.0001449844075589013, "ex_per_s": 382.87150430446934, "data_frac": 0.03415283175759979, "mem_mb": 87.04, "nll": -6.010741710662842, "elapsed_s": 381.1}119{"event": "train", "step": 2300, "loss": -6.001887798309326, "grad_norm": 0.1703479140996933, "lr": 0.0001427797303121009, "ex_per_s": 385.17127332298304, "data_frac": 0.03408450528859575, "mem_mb": 87.04, "nll": -6.001887798309326, "elapsed_s": 382.2}120{"event": "train", "step": 2320, "loss": -6.070018291473389, "grad_norm": 0.39208412170410156, "lr": 0.00014058112786516518, "ex_per_s": 387.45892416984105, "data_frac": 0.0340163721731201, "mem_mb": 87.04, "nll": -6.070018291473389, "elapsed_s": 383.2}121{"event": "train", "step": 2340, "loss": -5.853186130523682, "grad_norm": 0.38892120122909546, "lr": 0.00013838920129405016, "ex_per_s": 389.73427209864445, "data_frac": 0.033949552394937264, "mem_mb": 87.04, "nll": -5.853186130523682, "elapsed_s": 384.3}122{"event": "train", "step": 2360, "loss": -5.929654121398926, "grad_norm": 0.08632367849349976, "lr": 0.0001362045498495942, "ex_per_s": 391.9972800928464, "data_frac": 0.03388221499038046, "mem_mb": 87.04, "nll": -5.929654121398926, "elapsed_s": 385.3}123{"event": "train", "step": 2380, "loss": -5.932918071746826, "grad_norm": 0.10027097910642624, "lr": 0.00013402777079368853, "ex_per_s": 394.2489707395062, "data_frac": 0.03381455738151721, "mem_mb": 87.04, "nll": -5.932918071746826, "elapsed_s": 386.4}124{"event": "train", "step": 2400, "loss": -5.956467628479004, "grad_norm": 0.8867030739784241, "lr": 0.00013185945923599213, "ex_per_s": 396.48971031730997, "data_frac": 0.03374878414896794, "mem_mb": 87.04, "nll": -5.956467628479004, "elapsed_s": 387.4}125{"event": "train", "step": 2420, "loss": -5.996216773986816, "grad_norm": 0.26220500469207764, "lr": 0.00012970020797123432, "ex_per_s": 398.71989788273675, "data_frac": 0.03368279567318557, "mem_mb": 87.04, "nll": -5.996216773986816, "elapsed_s": 388.5}126{"event": "train", "step": 2440, "loss": -6.099379062652588, "grad_norm": 0.4166848063468933, "lr": 0.000127550607317151, "ex_per_s": 400.93868007495604, "data_frac": 0.033617167769708314, "mem_mb": 87.04, "nll": -6.099379062652588, "elapsed_s": 389.5}127{"event": "train", "step": 2460, "loss": -6.134914875030518, "grad_norm": 0.11265060305595398, "lr": 0.00012541124495309727, "ex_per_s": 403.14379229622875, "data_frac": 0.03355124991035432, "mem_mb": 87.04, "nll": -6.134914875030518, "elapsed_s": 390.6}128{"event": "train", "step": 2480, "loss": -5.945562362670898, "grad_norm": 0.11165693402290344, "lr": 0.00012328270575938212, "ex_per_s": 405.33494841901273, "data_frac": 0.033486293103078706, "mem_mb": 87.04, "nll": -5.945562362670898, "elapsed_s": 391.6}129{"event": "train", "step": 2500, "loss": -5.7568135261535645, "grad_norm": 0.6138439178466797, "lr": 0.00012116557165736775, "ex_per_s": 407.5134339660475, "data_frac": 0.03342351997336849, "mem_mb": 87.04, "nll": -5.7568135261535645, "elapsed_s": 392.7}130{"event": "val", "step": 2500, "val_nll": -6.450397919521789, "best": -6.450397919521789, "bad": 0, "mem_mb": 87.04, "elapsed_s": 454.7}131{"event": "train", "step": 2520, "loss": -5.96063756942749, "grad_norm": 0.8228510022163391, "lr": 0.00011906042145037817, "ex_per_s": 353.86943851732144, "data_frac": 0.02881484191596148, "mem_mb": 87.04, "nll": -5.96063756942749, "elapsed_s": 455.8}132{"event": "train", "step": 2540, "loss": -6.031038761138916, "grad_norm": 0.2783392667770386, "lr": 0.00011696783066546026, "ex_per_s": 355.86292694174443, "data_frac": 0.0287699149075272, "mem_mb": 87.04, "nll": -6.031038761138916, "elapsed_s": 456.8}133{"event": "train", "step": 2560, "loss": -5.948510646820068, "grad_norm": 0.1430380642414093, "lr": 0.00011488837139604017, "ex_per_s": 357.8469259230383, "data_frac": 0.028726203142729573, "mem_mb": 87.04, "nll": -5.948510646820068, "elapsed_s": 457.9}134{"event": "train", "step": 2580, "loss": -5.96066951751709, "grad_norm": 0.15230979025363922, "lr": 0.00011282261214551898, "ex_per_s": 359.81879376267744, "data_frac": 0.02868158222934515, "mem_mb": 87.04, "nll": -5.96066951751709, "elapsed_s": 458.9}135{"event": "train", "step": 2600, "loss": -5.849329471588135, "grad_norm": 0.15821196138858795, "lr": 0.00011077111767184915, "ex_per_s": 361.78319010236737, "data_frac": 0.0286373930908375, "mem_mb": 87.04, "nll": -5.849329471588135, "elapsed_s": 460.0}136{"event": "train", "step": 2620, "loss": -5.8889079093933105, "grad_norm": 0.9614686369895935, "lr": 0.00010873444883313583, "ex_per_s": 363.73699710933874, "data_frac": 0.028594448156410917, "mem_mb": 87.04, "nll": -5.8889079093933105, "elapsed_s": 461.0}137{"event": "train", "step": 2640, "loss": -5.987636566162109, "grad_norm": 0.38030269742012024, "lr": 0.00010671316243430345, "ex_per_s": 365.6816284053561, "data_frac": 0.028552707071625893, "mem_mb": 87.04, "nll": -5.987636566162109, "elapsed_s": 462.1}138{"event": "train", "step": 2660, "loss": -5.819762229919434, "grad_norm": 0.21608659625053406, "lr": 0.00010470781107487112, "ex_per_s": 367.6189712435255, "data_frac": 0.028509010734300594, "mem_mb": 87.04, "nll": -5.819762229919434, "elapsed_s": 463.1}139{"event": "train", "step": 2680, "loss": -5.9606194496154785, "grad_norm": 0.28313860297203064, "lr": 0.00010271894299787715, "ex_per_s": 369.5476888570068, "data_frac": 0.028465779765219707, "mem_mb": 87.04, "nll": -5.9606194496154785, "elapsed_s": 464.2}140{"event": "train", "step": 2700, "loss": -6.134004592895508, "grad_norm": 0.640190839767456, "lr": 0.00010074710193999506, "ex_per_s": 371.467735027518, "data_frac": 0.028422443173540566, "mem_mb": 87.04, "nll": -6.134004592895508, "elapsed_s": 465.2}141{"event": "train", "step": 2720, "loss": -5.883052349090576, "grad_norm": 0.545653760433197, "lr": 9.879282698288144e-05, "ex_per_s": 373.3782395792918, "data_frac": 0.02837932125940466, "mem_mb": 87.04, "nll": -5.883052349090576, "elapsed_s": 466.3}142{"event": "train", "step": 2740, "loss": -6.127445697784424, "grad_norm": 0.11376520246267319, "lr": 9.685665240579613e-05, "ex_per_s": 375.28042944334646, "data_frac": 0.028335068011231333, "mem_mb": 87.04, "nll": -6.127445697784424, "elapsed_s": 467.3}143{"event": "train", "step": 2760, "loss": -5.95934534072876, "grad_norm": 0.27952632308006287, "lr": 9.4939107539536e-05, "ex_per_s": 377.1746595946528, "data_frac": 0.028292221871174382, "mem_mb": 87.04, "nll": -5.95934534072876, "elapsed_s": 468.4}144{"event": "train", "step": 2780, "loss": -6.01336669921875, "grad_norm": 0.8153471350669861, "lr": 9.304071662172087e-05, "ex_per_s": 379.05953342567966, "data_frac": 0.02824963596077116, "mem_mb": 87.04, "nll": -6.01336669921875, "elapsed_s": 469.4}145{"event": "train", "step": 2800, "loss": -5.9800872802734375, "grad_norm": 0.8266234993934631, "lr": 9.116199865347241e-05, "ex_per_s": 380.93713702284043, "data_frac": 0.028207593991312733, "mem_mb": 87.04, "nll": -5.9800872802734375, "elapsed_s": 470.4}146{"event": "train", "step": 2820, "loss": -6.00831413269043, "grad_norm": 0.2931225299835205, "lr": 8.930346725752419e-05, "ex_per_s": 382.8063551673675, "data_frac": 0.028165252648067567, "mem_mb": 87.04, "nll": -6.00831413269043, "elapsed_s": 471.5}147{"event": "train", "step": 2840, "loss": -5.881725788116455, "grad_norm": 0.131366565823555, "lr": 8.746563053780262e-05, "ex_per_s": 384.6666041070263, "data_frac": 0.028123031447088862, "mem_mb": 87.04, "nll": -5.881725788116455, "elapsed_s": 472.5}148{"event": "train", "step": 2860, "loss": -6.048248767852783, "grad_norm": 0.22457239031791687, "lr": 8.564899094051614e-05, "ex_per_s": 386.51993308211576, "data_frac": 0.02808108768219185, "mem_mb": 87.04, "nll": -6.048248767852783, "elapsed_s": 473.6}149{"event": "train", "step": 2880, "loss": -5.772135257720947, "grad_norm": 0.412826806306839, "lr": 8.385404511679161e-05, "ex_per_s": 388.36464980484465, "data_frac": 0.02803910437315891, "mem_mb": 87.04, "nll": -5.772135257720947, "elapsed_s": 474.6}150{"event": "train", "step": 2900, "loss": -5.796346187591553, "grad_norm": 0.5453027486801147, "lr": 8.208128378689484e-05, "ex_per_s": 390.20070799833724, "data_frac": 0.0279981250865312, "mem_mb": 87.04, "nll": -5.796346187591553, "elapsed_s": 475.7}151{"event": "train", "step": 2920, "loss": -5.9706315994262695, "grad_norm": 0.7435406446456909, "lr": 8.03311916060725e-05, "ex_per_s": 392.0300087929824, "data_frac": 0.02795681700787153, "mem_mb": 87.04, "nll": -5.9706315994262695, "elapsed_s": 476.7}152{"event": "train", "step": 2940, "loss": -5.903410911560059, "grad_norm": 0.540775716304779, "lr": 7.860424703205225e-05, "ex_per_s": 393.85127395722037, "data_frac": 0.027915507046086474, "mem_mb": 87.04, "nll": -5.903410911560059, "elapsed_s": 477.8}153{"event": "train", "step": 2960, "loss": -5.937546730041504, "grad_norm": 0.42116114497184753, "lr": 7.690092219423722e-05, "ex_per_s": 395.66450037464534, "data_frac": 0.02787501345609403, "mem_mb": 87.04, "nll": -5.937546730041504, "elapsed_s": 478.8}154{"event": "train", "step": 2980, "loss": -5.893056392669678, "grad_norm": 0.6120806336402893, "lr": 7.522168276463048e-05, "ex_per_s": 397.47120682629105, "data_frac": 0.02783428639823081, "mem_mb": 87.04, "nll": -5.893056392669678, "elapsed_s": 479.9}155{"event": "train", "step": 3000, "loss": -5.995741844177246, "grad_norm": 0.15290744602680206, "lr": 7.356698783052496e-05, "ex_per_s": 399.2702238897191, "data_frac": 0.02779413175451365, "mem_mb": 87.04, "nll": -5.995741844177246, "elapsed_s": 480.9}156{"event": "val", "step": 3000, "val_nll": -6.450188497274488, "best": -6.450397919521789, "bad": 1, "mem_mb": 87.04, "elapsed_s": 542.9}157{"event": "train", "step": 3020, "loss": -5.961538314819336, "grad_norm": 0.9350239634513855, "lr": 7.193728976899362e-05, "ex_per_s": 355.2871724464645, "data_frac": 0.024587368846635316, "mem_mb": 87.04, "nll": -5.961538314819336, "elapsed_s": 544.0}158{"event": "train", "step": 3040, "loss": -5.861483573913574, "grad_norm": 0.7513971924781799, "lr": 7.033303412321404e-05, "ex_per_s": 356.95109674823965, "data_frac": 0.024557823194060774, "mem_mb": 87.04, "nll": -5.861483573913574, "elapsed_s": 545.1}159{"event": "train", "step": 3060, "loss": -6.073173522949219, "grad_norm": 0.2579517662525177, "lr": 6.875465948066139e-05, "ex_per_s": 358.61006550455, "data_frac": 0.024528217434616768, "mem_mb": 87.04, "nll": -6.073173522949219, "elapsed_s": 546.1}160{"event": "train", "step": 3080, "loss": -5.9698357582092285, "grad_norm": 0.30502694845199585, "lr": 6.720259735320299e-05, "ex_per_s": 360.2623308315738, "data_frac": 0.024498868193687882, "mem_mb": 87.04, "nll": -5.9698357582092285, "elapsed_s": 547.2}161{"event": "train", "step": 3100, "loss": -6.0161004066467285, "grad_norm": 0.3001844882965088, "lr": 6.567727205912722e-05, "ex_per_s": 361.90880356031494, "data_frac": 0.02447079149602727, "mem_mb": 87.04, "nll": -6.0161004066467285, "elapsed_s": 548.2}162{"event": "train", "step": 3120, "loss": -6.091786861419678, "grad_norm": 0.6344705820083618, "lr": 6.417910060713926e-05, "ex_per_s": 363.54911370708845, "data_frac": 0.02444228592180963, "mem_mb": 87.04, "nll": -6.091786861419678, "elapsed_s": 549.3}163{"event": "train", "step": 3140, "loss": -6.0084052085876465, "grad_norm": 0.3295203447341919, "lr": 6.270849258235485e-05, "ex_per_s": 365.18280267638175, "data_frac": 0.024414283461590826, "mem_mb": 87.04, "nll": -6.0084052085876465, "elapsed_s": 550.3}164{"event": "train", "step": 3160, "loss": -6.022407054901123, "grad_norm": 0.12153351306915283, "lr": 6.126585003432406e-05, "ex_per_s": 366.8100360931024, "data_frac": 0.024385427588309062, "mem_mb": 87.04, "nll": -6.022407054901123, "elapsed_s": 551.4}165{"event": "train", "step": 3180, "loss": -6.096247673034668, "grad_norm": 0.13388048112392426, "lr": 5.985156736711483e-05, "ex_per_s": 368.4309263259209, "data_frac": 0.024356734941381783, "mem_mb": 87.04, "nll": -6.096247673034668, "elapsed_s": 552.4}166{"event": "train", "step": 3200, "loss": -6.001851558685303, "grad_norm": 0.4614557921886444, "lr": 5.846603123148688e-05, "ex_per_s": 370.04608876418433, "data_frac": 0.024328037377331302, "mem_mb": 87.04, "nll": -6.001851558685303, "elapsed_s": 553.5}167{"event": "train", "step": 3220, "loss": -5.916229724884033, "grad_norm": 0.15701963007450104, "lr": 5.710962041918516e-05, "ex_per_s": 371.65479536079835, "data_frac": 0.024299388374161686, "mem_mb": 87.04, "nll": -5.916229724884033, "elapsed_s": 554.5}168{"event": "train", "step": 3240, "loss": -5.803135395050049, "grad_norm": 0.44982677698135376, "lr": 5.578270575938211e-05, "ex_per_s": 373.2588663140697, "data_frac": 0.024270940063817054, "mem_mb": 87.04, "nll": -5.803135395050049, "elapsed_s": 555.6}169{"event": "train", "step": 3260, "loss": -5.960736274719238, "grad_norm": 0.4541875720024109, "lr": 5.448565001729653e-05, "ex_per_s": 374.85804180505556, "data_frac": 0.024242307190987588, "mem_mb": 87.04, "nll": -5.960736274719238, "elapsed_s": 556.6}170{"event": "train", "step": 3280, "loss": -6.016822338104248, "grad_norm": 0.3475340008735657, "lr": 5.3218807795017296e-05, "ex_per_s": 376.44924874260727, "data_frac": 0.024213802380419418, "mem_mb": 87.04, "nll": -6.016822338104248, "elapsed_s": 557.7}171{"event": "train", "step": 3300, "loss": -5.808777332305908, "grad_norm": 0.1888248473405838, "lr": 5.198252543455864e-05, "ex_per_s": 378.03574783821654, "data_frac": 0.024185249050636147, "mem_mb": 87.04, "nll": -5.808777332305908, "elapsed_s": 558.7}172{"event": "train", "step": 3320, "loss": -5.990631580352783, "grad_norm": 0.3589250445365906, "lr": 5.07771409231737e-05, "ex_per_s": 379.61656145604184, "data_frac": 0.02415713688925267, "mem_mb": 87.04, "nll": -5.990631580352783, "elapsed_s": 559.7}173{"event": "train", "step": 3340, "loss": -5.931220531463623, "grad_norm": 0.29501235485076904, "lr": 4.960298380095221e-05, "ex_per_s": 381.1916868163463, "data_frac": 0.024129205471265923, "mem_mb": 87.04, "nll": -5.931220531463623, "elapsed_s": 560.8}174{"event": "train", "step": 3360, "loss": -5.908580780029297, "grad_norm": 0.4806421399116516, "lr": 4.846037507072751e-05, "ex_per_s": 382.76141874901975, "data_frac": 0.024101838505553568, "mem_mb": 87.04, "nll": -5.908580780029297, "elapsed_s": 561.8}175{"event": "train", "step": 3380, "loss": -6.123459339141846, "grad_norm": 0.47969064116477966, "lr": 4.7349627110317556e-05, "ex_per_s": 384.3251830113956, "data_frac": 0.024075601114927844, "mem_mb": 87.04, "nll": -6.123459339141846, "elapsed_s": 562.9}176{"event": "train", "step": 3400, "loss": -5.883491039276123, "grad_norm": 0.5333940982818604, "lr": 4.6271043587123986e-05, "ex_per_s": 385.88387410798276, "data_frac": 0.02404838285775352, "mem_mb": 87.04, "nll": -5.883491039276123, "elapsed_s": 563.9}177{"event": "train", "step": 3420, "loss": -5.83681058883667, "grad_norm": 0.29943934082984924, "lr": 4.52249193751124e-05, "ex_per_s": 387.4372657806797, "data_frac": 0.02402093741117327, "mem_mb": 87.04, "nll": -5.83681058883667, "elapsed_s": 565.0}178{"event": "train", "step": 3440, "loss": -5.91804838180542, "grad_norm": 0.17160546779632568, "lr": 4.4211540474196926e-05, "ex_per_s": 388.98542396436136, "data_frac": 0.023993398440054062, "mem_mb": 87.04, "nll": -5.91804838180542, "elapsed_s": 566.0}179{"event": "train", "step": 3460, "loss": -6.059709548950195, "grad_norm": 0.15575508773326874, "lr": 4.3231183932050546e-05, "ex_per_s": 390.5275427967594, "data_frac": 0.023966026761557027, "mem_mb": 87.04, "nll": -6.059709548950195, "elapsed_s": 567.1}180{"event": "train", "step": 3480, "loss": -5.918010234832764, "grad_norm": 0.09876300394535065, "lr": 4.228411776836306e-05, "ex_per_s": 392.06458390877464, "data_frac": 0.02393916134216457, "mem_mb": 87.04, "nll": -5.918010234832764, "elapsed_s": 568.1}181{"event": "train", "step": 3500, "loss": -6.038009166717529, "grad_norm": 0.08236965537071228, "lr": 4.137060090156723e-05, "ex_per_s": 393.59552113173584, "data_frac": 0.023911950589060163, "mem_mb": 87.04, "nll": -6.038009166717529, "elapsed_s": 569.1}182{"event": "val", "step": 3500, "val_nll": -6.451227885057247, "best": -6.451227885057247, "bad": 0, "mem_mb": 87.04, "elapsed_s": 631.2}183{"event": "train", "step": 3520, "loss": -5.938910961151123, "grad_norm": 0.45747262239456177, "lr": 4.049088307805306e-05, "ex_per_s": 356.29088506320295, "data_frac": 0.021537317162999604, "mem_mb": 87.04, "nll": -5.938910961151123, "elapsed_s": 632.3}184{"event": "train", "step": 3540, "loss": -6.137465000152588, "grad_norm": 0.6292771697044373, "lr": 3.964520480388956e-05, "ex_per_s": 357.72291621818835, "data_frac": 0.021517089846851644, "mem_mb": 87.04, "nll": -6.137465000152588, "elapsed_s": 633.4}185{"event": "train", "step": 3560, "loss": -6.140748977661133, "grad_norm": 0.07359914481639862, "lr": 3.88337972790728e-05, "ex_per_s": 359.1490271280012, "data_frac": 0.021497440398476907, "mem_mb": 87.04, "nll": -6.140748977661133, "elapsed_s": 634.4}186{"event": "train", "step": 3580, "loss": -5.9030232429504395, "grad_norm": 0.2778854966163635, "lr": 3.805688233431824e-05, "ex_per_s": 360.5718070381375, "data_frac": 0.021477334587867474, "mem_mb": 87.04, "nll": -5.9030232429504395, "elapsed_s": 635.5}187{"event": "train", "step": 3600, "loss": -6.099392890930176, "grad_norm": 0.39385804533958435, "lr": 3.7314672370414325e-05, "ex_per_s": 361.9899145214003, "data_frac": 0.02145726722847325, "mem_mb": 87.04, "nll": -6.099392890930176, "elapsed_s": 636.5}188{"event": "train", "step": 3620, "loss": -5.912858009338379, "grad_norm": 0.9359757304191589, "lr": 3.660737030015427e-05, "ex_per_s": 363.40366187569094, "data_frac": 0.021437335594135888, "mem_mb": 87.04, "nll": -5.912858009338379, "elapsed_s": 637.6}189{"event": "train", "step": 3640, "loss": -6.052988052368164, "grad_norm": 0.4626435339450836, "lr": 3.5935169492861716e-05, "ex_per_s": 364.81303189423073, "data_frac": 0.021417311734825594, "mem_mb": 87.04, "nll": -6.052988052368164, "elapsed_s": 638.6}190{"event": "train", "step": 3660, "loss": -5.834073543548584, "grad_norm": 0.5145513415336609, "lr": 3.52982537215255e-05, "ex_per_s": 366.21825512254753, "data_frac": 0.021397205495225353, "mem_mb": 87.04, "nll": -5.834073543548584, "elapsed_s": 639.6}191{"event": "train", "step": 3680, "loss": -5.991765022277832, "grad_norm": 0.10713344812393188, "lr": 3.469679711255794e-05, "ex_per_s": 367.6184103143205, "data_frac": 0.021377452255869456, "mem_mb": 87.04, "nll": -5.991765022277832, "elapsed_s": 640.7}192{"event": "train", "step": 3700, "loss": -6.06709623336792, "grad_norm": 0.7376312017440796, "lr": 3.41309640981904e-05, "ex_per_s": 369.01434701180705, "data_frac": 0.021357612826709515, "mem_mb": 87.04, "nll": -6.06709623336792, "elapsed_s": 641.7}193{"event": "train", "step": 3720, "loss": -5.948267459869385, "grad_norm": 0.523550808429718, "lr": 3.3600909371519225e-05, "ex_per_s": 370.40475463162505, "data_frac": 0.02133898793304807, "mem_mb": 87.04, "nll": -5.948267459869385, "elapsed_s": 642.8}194{"event": "train", "step": 3740, "loss": -6.190938949584961, "grad_norm": 0.20708702504634857, "lr": 3.31067778442141e-05, "ex_per_s": 371.7910105828941, "data_frac": 0.021319705286487373, "mem_mb": 87.04, "nll": -6.190938949584961, "elapsed_s": 643.8}195{"event": "train", "step": 3760, "loss": -6.148644924163818, "grad_norm": 0.17288844287395477, "lr": 3.2648704606900735e-05, "ex_per_s": 373.174643941945, "data_frac": 0.021300575629964024, "mem_mb": 87.04, "nll": -6.148644924163818, "elapsed_s": 644.9}196{"event": "train", "step": 3780, "loss": -5.9644646644592285, "grad_norm": 0.40576744079589844, "lr": 3.222681489222838e-05, "ex_per_s": 374.55273541642134, "data_frac": 0.02128124005711577, "mem_mb": 87.04, "nll": -5.9644646644592285, "elapsed_s": 645.9}197{"event": "train", "step": 3800, "loss": -6.099854469299316, "grad_norm": 0.10551419109106064, "lr": 3.1841224040632484e-05, "ex_per_s": 375.9265578513152, "data_frac": 0.021261681686226247, "mem_mb": 87.04, "nll": -6.099854469299316, "elapsed_s": 647.0}198{"event": "train", "step": 3820, "loss": -5.972358226776123, "grad_norm": 0.16502749919891357, "lr": 3.149203746880175e-05, "ex_per_s": 377.29599895092474, "data_frac": 0.021242370821755743, "mem_mb": 87.04, "nll": -5.972358226776123, "elapsed_s": 648.0}199{"event": "train", "step": 3840, "loss": -5.874516010284424, "grad_norm": 0.6420828700065613, "lr": 3.117935064085835e-05, "ex_per_s": 378.66110556353794, "data_frac": 0.021222987343767073, "mem_mb": 87.04, "nll": -5.874516010284424, "elapsed_s": 649.0}200{"event": "train", "step": 3860, "loss": -6.015261173248291, "grad_norm": 0.40917009115219116, "lr": 3.090324904225885e-05, "ex_per_s": 380.0216238197984, "data_frac": 0.021204067455909435, "mem_mb": 87.04, "nll": -6.015261173248291, "elapsed_s": 650.1}201{"event": "train", "step": 3880, "loss": -5.891748905181885, "grad_norm": 0.07437483221292496, "lr": 3.0663808156423456e-05, "ex_per_s": 381.37735187156534, "data_frac": 0.021184848185608444, "mem_mb": 87.04, "nll": -5.891748905181885, "elapsed_s": 651.1}202{"event": "train", "step": 3900, "loss": -6.184466361999512, "grad_norm": 0.6514259576797485, "lr": 3.046109344409957e-05, "ex_per_s": 382.73024752120136, "data_frac": 0.021165522731638176, "mem_mb": 87.04, "nll": -6.184466361999512, "elapsed_s": 652.2}203{"event": "train", "step": 3920, "loss": -6.035919189453125, "grad_norm": 0.0865134671330452, "lr": 3.0295160325465435e-05, "ex_per_s": 384.0759471273742, "data_frac": 0.021147659525628637, "mem_mb": 87.04, "nll": -6.035919189453125, "elapsed_s": 653.2}204{"event": "train", "step": 3940, "loss": -5.923712730407715, "grad_norm": 0.1081044152379036, "lr": 3.016605416497886e-05, "ex_per_s": 385.4205282026967, "data_frac": 0.021128793482109304, "mem_mb": 87.04, "nll": -5.923712730407715, "elapsed_s": 654.3}205{"event": "train", "step": 3960, "loss": -6.0707502365112305, "grad_norm": 0.08182890713214874, "lr": 3.0073810258974985e-05, "ex_per_s": 386.7607008602709, "data_frac": 0.021109809655629492, "mem_mb": 87.04, "nll": -6.0707502365112305, "elapsed_s": 655.3}206{"event": "train", "step": 3980, "loss": -5.802964210510254, "grad_norm": 0.12014619261026382, "lr": 3.0018453826016683e-05, "ex_per_s": 388.0969210780507, "data_frac": 0.02109099497554663, "mem_mb": 87.04, "nll": -5.802964210510254, "elapsed_s": 656.4}207{"event": "train", "step": 4000, "loss": -6.055367946624756, "grad_norm": 0.7738989591598511, "lr": 2.9999999999999997e-05, "ex_per_s": 389.42977858039075, "data_frac": 0.021072319473601567, "mem_mb": 87.04, "nll": -6.055367946624756, "elapsed_s": 657.4}208{"event": "val", "step": 4000, "val_nll": -6.451153782232082, "best": -6.451227885057247, "bad": 1, "mem_mb": 87.04, "elapsed_s": 719.5}209{"event": "done", "steps": 4000, "best_val": -6.451227885057247, "wall_s": 719.5306529998779, "ex_per_s": 355.7874830386739, "peak_mem_mb": 87.04, "n_params": 3330819, "device": "mps", "data_time_frac": 0.019251911178004226, "elapsed_s": 719.5}210
Files · bab86036e93a82a6f322d24c81650aa6a2bb946e
Menu

Theme