SPB Git

spb/wp41_uqo

Public
51.7 KB · 210 lines
1{"event": "train", "step": 20, "loss": -6.078861713409424, "grad_norm": 0.4547577202320099, "lr": 3.149999999999999e-05, "ex_per_s": 91.32427216148434, "data_frac": 0.8642638219307126, "mem_mb": 107.833344, "nll": -6.078861713409424, "elapsed_s": 14.1}2{"event": "train", "step": 40, "loss": -5.839817047119141, "grad_norm": 0.6904654502868652, "lr": 6.149999999999999e-05, "ex_per_s": 163.0708082153018, "data_frac": 0.7724274148271076, "mem_mb": 107.833344, "nll": -5.839817047119141, "elapsed_s": 15.7}3{"event": "train", "step": 60, "loss": -5.904033184051514, "grad_norm": 0.44781145453453064, "lr": 9.149999999999999e-05, "ex_per_s": 220.87854549021168, "data_frac": 0.6982364195807818, "mem_mb": 107.833344, "nll": -5.904033184051514, "elapsed_s": 17.4}4{"event": "train", "step": 80, "loss": -6.056328773498535, "grad_norm": 0.550475001335144, "lr": 0.0001215, "ex_per_s": 268.5134960912202, "data_frac": 0.6372573983971267, "mem_mb": 107.833344, "nll": -6.056328773498535, "elapsed_s": 19.1}5{"event": "train", "step": 100, "loss": -5.897518157958984, "grad_norm": 1.6881122589111328, "lr": 0.0001515, "ex_per_s": 308.5477566196074, "data_frac": 0.5863537996464951, "mem_mb": 107.833344, "nll": -5.897518157958984, "elapsed_s": 20.8}6{"event": "train", "step": 120, "loss": -5.831510543823242, "grad_norm": 2.607893466949463, "lr": 0.00018149999999999997, "ex_per_s": 342.67193187805645, "data_frac": 0.5432450217912818, "mem_mb": 107.833344, "nll": -5.831510543823242, "elapsed_s": 22.5}7{"event": "train", "step": 140, "loss": -5.806246280670166, "grad_norm": 1.066223382949829, "lr": 0.00021149999999999996, "ex_per_s": 372.01741563080674, "data_frac": 0.5059531010523952, "mem_mb": 107.833344, "nll": -5.806246280670166, "elapsed_s": 24.1}8{"event": "train", "step": 160, "loss": -5.6979756355285645, "grad_norm": 1.9667901992797852, "lr": 0.0002415, "ex_per_s": 397.6222670239104, "data_frac": 0.47361052571831613, "mem_mb": 107.833344, "nll": -5.6979756355285645, "elapsed_s": 25.8}9{"event": "train", "step": 180, "loss": -5.92747163772583, "grad_norm": 1.6286942958831787, "lr": 0.0002715, "ex_per_s": 420.1319035756648, "data_frac": 0.4452101046358239, "mem_mb": 107.833344, "nll": -5.92747163772583, "elapsed_s": 27.5}10{"event": "train", "step": 200, "loss": -5.921437740325928, "grad_norm": 1.144769310951233, "lr": 0.0003, "ex_per_s": 440.10834023029514, "data_frac": 0.42013305124650285, "mem_mb": 107.833344, "nll": -5.921437740325928, "elapsed_s": 29.1}11{"event": "train", "step": 220, "loss": -6.002372741699219, "grad_norm": 1.0037473440170288, "lr": 0.00029998154617398326, "ex_per_s": 457.7807054950397, "data_frac": 0.397660512228079, "mem_mb": 107.833344, "nll": -6.002372741699219, "elapsed_s": 30.8}12{"event": "train", "step": 240, "loss": -5.901665687561035, "grad_norm": 0.5465152263641357, "lr": 0.000299926189741025, "ex_per_s": 473.74944060556385, "data_frac": 0.37758606934696004, "mem_mb": 107.833344, "nll": -5.901665687561035, "elapsed_s": 32.5}13{"event": "train", "step": 260, "loss": -5.855159759521484, "grad_norm": 1.4677324295043945, "lr": 0.0002998339458350211, "ex_per_s": 488.17566026297965, "data_frac": 0.3594772090325357, "mem_mb": 107.833344, "nll": -5.855159759521484, "elapsed_s": 34.1}14{"event": "train", "step": 280, "loss": -5.921430587768555, "grad_norm": 0.19351482391357422, "lr": 0.0002997048396745345, "ex_per_s": 501.2707326003244, "data_frac": 0.3430519139164159, "mem_mb": 107.833344, "nll": -5.921430587768555, "elapsed_s": 35.8}15{"event": "train", "step": 300, "loss": -5.933881759643555, "grad_norm": 0.7669237852096558, "lr": 0.0002995389065559004, "ex_per_s": 513.1870360586589, "data_frac": 0.3280864361998215, "mem_mb": 107.833344, "nll": -5.933881759643555, "elapsed_s": 37.5}16{"event": "train", "step": 320, "loss": -5.937410831451416, "grad_norm": 0.6171549558639526, "lr": 0.0002993361918435765, "ex_per_s": 524.0952516877094, "data_frac": 0.31439489026789563, "mem_mb": 107.833344, "nll": -5.937410831451416, "elapsed_s": 39.1}17{"event": "train", "step": 340, "loss": -5.803722381591797, "grad_norm": 0.5185009241104126, "lr": 0.00029909675095774114, "ex_per_s": 534.1184081883531, "data_frac": 0.3018264037683778, "mem_mb": 107.833344, "nll": -5.803722381591797, "elapsed_s": 40.8}18{"event": "train", "step": 360, "loss": -5.967191219329834, "grad_norm": 2.0146749019622803, "lr": 0.00029882064935914164, "ex_per_s": 543.3414323004718, "data_frac": 0.290236449373344, "mem_mb": 107.833344, "nll": -5.967191219329834, "elapsed_s": 42.4}19{"event": "train", "step": 380, "loss": -5.9731926918029785, "grad_norm": 1.5997990369796753, "lr": 0.00029850796253119824, "ex_per_s": 551.878095285005, "data_frac": 0.27953125859980793, "mem_mb": 107.833344, "nll": -5.9731926918029785, "elapsed_s": 44.1}20{"event": "train", "step": 400, "loss": -5.970386981964111, "grad_norm": 1.7778751850128174, "lr": 0.0002981587759593675, "ex_per_s": 559.7675253038483, "data_frac": 0.26959096209962086, "mem_mb": 107.833344, "nll": -5.970386981964111, "elapsed_s": 45.8}21{"event": "train", "step": 420, "loss": -5.940792083740234, "grad_norm": 1.2141106128692627, "lr": 0.0002977731851077716, "ex_per_s": 567.098119416765, "data_frac": 0.2603468268859685, "mem_mb": 107.833344, "nll": -5.940792083740234, "elapsed_s": 47.4}22{"event": "train", "step": 440, "loss": -5.909468173980713, "grad_norm": 0.4346475303173065, "lr": 0.0002973512953930992, "ex_per_s": 573.9792050604112, "data_frac": 0.2517463752496795, "mem_mb": 107.833344, "nll": -5.909468173980713, "elapsed_s": 49.1}23{"event": "train", "step": 460, "loss": -6.071146011352539, "grad_norm": 1.0715789794921875, "lr": 0.0002968932221557859, "ex_per_s": 580.4159280753846, "data_frac": 0.24373080100398167, "mem_mb": 107.833344, "nll": -6.071146011352539, "elapsed_s": 50.8}24{"event": "train", "step": 480, "loss": -5.8616862297058105, "grad_norm": 0.9082559943199158, "lr": 0.0002963990906284808, "ex_per_s": 586.559415930414, "data_frac": 0.23625848576073272, "mem_mb": 107.833344, "nll": -5.8616862297058105, "elapsed_s": 52.4}25{"event": "train", "step": 500, "loss": -5.89613676071167, "grad_norm": 0.6814314723014832, "lr": 0.00029586903590180956, "ex_per_s": 592.3208376764089, "data_frac": 0.22923808165825105, "mem_mb": 107.833344, "nll": -5.89613676071167, "elapsed_s": 54.1}26{"event": "val", "step": 500, "val_nll": -6.391830374921728, "best": -6.391830374921728, "bad": 0, "mem_mb": 107.833344, "elapsed_s": 117.8}27{"event": "train", "step": 520, "loss": -5.982941627502441, "grad_norm": 0.2663559019565582, "lr": 0.00029530320288744205, "ex_per_s": 278.6447165547169, "data_frac": 0.10378706859326077, "mem_mb": 107.833344, "nll": -5.982941627502441, "elapsed_s": 119.5}28{"event": "train", "step": 540, "loss": -5.872440814971924, "grad_norm": 0.5536114573478699, "lr": 0.0002947017462784745, "ex_per_s": 285.36844872401116, "data_frac": 0.10244519122599105, "mem_mb": 107.833344, "nll": -5.872440814971924, "elapsed_s": 121.1}29{"event": "train", "step": 560, "loss": -5.9392194747924805, "grad_norm": 0.20005904138088226, "lr": 0.00029406483050713824, "ex_per_s": 291.9106645625973, "data_frac": 0.10113966830314854, "mem_mb": 107.833344, "nll": -5.9392194747924805, "elapsed_s": 122.8}30{"event": "train", "step": 580, "loss": -5.893912315368652, "grad_norm": 1.3167750835418701, "lr": 0.0002933926296998457, "ex_per_s": 298.2769001833245, "data_frac": 0.09986984294817941, "mem_mb": 107.833344, "nll": -5.893912315368652, "elapsed_s": 124.5}31{"event": "train", "step": 600, "loss": -5.746176242828369, "grad_norm": 1.9514375925064087, "lr": 0.0002926853276295856, "ex_per_s": 304.4750266788319, "data_frac": 0.0986342206357223, "mem_mb": 107.833344, "nll": -5.746176242828369, "elapsed_s": 126.2}32{"event": "train", "step": 620, "loss": -5.914329528808594, "grad_norm": 0.15852728486061096, "lr": 0.00029194311766568174, "ex_per_s": 310.5102702201401, "data_frac": 0.09743110477802518, "mem_mb": 107.833344, "nll": -5.914329528808594, "elapsed_s": 127.8}33{"event": "train", "step": 640, "loss": -6.0334153175354, "grad_norm": 1.2140780687332153, "lr": 0.0002911662027209272, "ex_per_s": 316.38984402230244, "data_frac": 0.09625797467779096, "mem_mb": 107.833344, "nll": -6.0334153175354, "elapsed_s": 129.5}34{"event": "train", "step": 660, "loss": -5.857203960418701, "grad_norm": 3.8919763565063477, "lr": 0.00029035479519611045, "ex_per_s": 322.1145429464392, "data_frac": 0.09511835094090489, "mem_mb": 107.833344, "nll": -5.857203960418701, "elapsed_s": 131.2}35{"event": "train", "step": 680, "loss": -5.978630065917969, "grad_norm": 1.0302412509918213, "lr": 0.0002895091169219469, "ex_per_s": 327.6990163976406, "data_frac": 0.09400306565785209, "mem_mb": 107.833344, "nll": -5.978630065917969, "elapsed_s": 132.8}36{"event": "train", "step": 700, "loss": -5.940977096557617, "grad_norm": 2.7490224838256836, "lr": 0.00028862939909843273, "ex_per_s": 333.1463303346346, "data_frac": 0.0929161513786894, "mem_mb": 107.833344, "nll": -5.940977096557617, "elapsed_s": 134.5}37{"event": "train", "step": 720, "loss": -6.032063007354736, "grad_norm": 0.16528140008449554, "lr": 0.0002877158822316369, "ex_per_s": 338.4596356076678, "data_frac": 0.0918541422333009, "mem_mb": 107.833344, "nll": -6.032063007354736, "elapsed_s": 136.2}38{"event": "train", "step": 740, "loss": -5.98847770690918, "grad_norm": 0.30324673652648926, "lr": 0.00028676881606794945, "ex_per_s": 343.64721185700034, "data_frac": 0.09082099243473786, "mem_mb": 107.833344, "nll": -5.98847770690918, "elapsed_s": 137.9}39{"event": "train", "step": 760, "loss": -5.928562164306641, "grad_norm": 1.5995711088180542, "lr": 0.000285788459525803, "ex_per_s": 348.70714645188133, "data_frac": 0.08991759285522627, "mem_mb": 107.833344, "nll": -5.928562164306641, "elapsed_s": 139.5}40{"event": "train", "step": 780, "loss": -5.846802234649658, "grad_norm": 0.17351554334163666, "lr": 0.00028477508062488754, "ex_per_s": 353.6506106778963, "data_frac": 0.08893190085514321, "mem_mb": 107.833344, "nll": -5.846802234649658, "elapsed_s": 141.2}41{"event": "train", "step": 800, "loss": -5.789047718048096, "grad_norm": 0.523369550704956, "lr": 0.000283728956412876, "ex_per_s": 358.4767306817447, "data_frac": 0.08797028365358767, "mem_mb": 107.833344, "nll": -5.789047718048096, "elapsed_s": 142.9}42{"event": "train", "step": 820, "loss": -5.903843402862549, "grad_norm": 0.6612290740013123, "lr": 0.00028265037288968236, "ex_per_s": 363.1898832472882, "data_frac": 0.08702992383533668, "mem_mb": 107.833344, "nll": -5.903843402862549, "elapsed_s": 144.5}43{"event": "train", "step": 840, "loss": -5.7512922286987305, "grad_norm": 1.9344267845153809, "lr": 0.0002815396249292725, "ex_per_s": 367.8052562808979, "data_frac": 0.08611586051581883, "mem_mb": 107.833344, "nll": -5.7512922286987305, "elapsed_s": 146.2}44{"event": "train", "step": 860, "loss": -5.897014141082764, "grad_norm": 0.5821326971054077, "lr": 0.00028039701619904774, "ex_per_s": 372.31194259930834, "data_frac": 0.08522158556353819, "mem_mb": 107.833344, "nll": -5.897014141082764, "elapsed_s": 147.9}45{"event": "train", "step": 880, "loss": -5.843864440917969, "grad_norm": 0.690083920955658, "lr": 0.00027922285907682625, "ex_per_s": 376.721193239836, "data_frac": 0.08434592992171747, "mem_mb": 107.833344, "nll": -5.843864440917969, "elapsed_s": 149.5}46{"event": "train", "step": 900, "loss": -5.9409356117248535, "grad_norm": 0.25723984837532043, "lr": 0.00027801747456544134, "ex_per_s": 381.0312585059186, "data_frac": 0.08349451329656962, "mem_mb": 107.833344, "nll": -5.9409356117248535, "elapsed_s": 151.2}47{"event": "train", "step": 920, "loss": -5.973232269287109, "grad_norm": 1.3262370824813843, "lr": 0.00027678119220498265, "ex_per_s": 385.25692934040444, "data_frac": 0.08265546408387092, "mem_mb": 107.833344, "nll": -5.973232269287109, "elapsed_s": 152.9}48{"event": "train", "step": 940, "loss": -6.0237555503845215, "grad_norm": 0.6899561285972595, "lr": 0.0002755143499827035, "ex_per_s": 389.399816865086, "data_frac": 0.08183620886539522, "mem_mb": 107.833344, "nll": -6.0237555503845215, "elapsed_s": 154.5}49{"event": "train", "step": 960, "loss": -5.925710678100586, "grad_norm": 1.255665898323059, "lr": 0.00027421729424061787, "ex_per_s": 393.451632717161, "data_frac": 0.0810362405241369, "mem_mb": 107.833344, "nll": -5.925710678100586, "elapsed_s": 156.2}50{"event": "train", "step": 980, "loss": -5.8946356773376465, "grad_norm": 0.621134877204895, "lr": 0.00027289037958081483, "ex_per_s": 397.41854387978736, "data_frac": 0.08025040901743086, "mem_mb": 107.833344, "nll": -5.8946356773376465, "elapsed_s": 157.9}51{"event": "train", "step": 1000, "loss": -6.036588668823242, "grad_norm": 0.26344168186187744, "lr": 0.0002715339687685131, "ex_per_s": 401.29879854840516, "data_frac": 0.07947914349331435, "mem_mb": 107.833344, "nll": -6.036588668823242, "elapsed_s": 159.5}52{"event": "val", "step": 1000, "val_nll": -6.393413394690595, "best": -6.393413394690595, "bad": 0, "mem_mb": 107.833344, "elapsed_s": 222.7}53{"event": "train", "step": 1020, "loss": -5.988218307495117, "grad_norm": 0.2535741329193115, "lr": 0.0002701484326328852, "ex_per_s": 290.8959476423647, "data_frac": 0.05653340750011817, "mem_mb": 107.833344, "nll": -5.988218307495117, "elapsed_s": 224.5}54{"event": "train", "step": 1040, "loss": -5.937881946563721, "grad_norm": 0.22067518532276154, "lr": 0.00026873414996567596, "ex_per_s": 294.4044151117305, "data_frac": 0.05616380422564218, "mem_mb": 107.833344, "nll": -5.937881946563721, "elapsed_s": 226.1}55{"event": "train", "step": 1060, "loss": -6.147581577301025, "grad_norm": 1.076682209968567, "lr": 0.0002672915074176452, "ex_per_s": 297.8662961239216, "data_frac": 0.05579906628909593, "mem_mb": 107.833344, "nll": -6.147581577301025, "elapsed_s": 227.8}56{"event": "train", "step": 1080, "loss": -5.800237655639648, "grad_norm": 0.3834872543811798, "lr": 0.00026582089939286075, "ex_per_s": 301.2737355383076, "data_frac": 0.05544209603363889, "mem_mb": 107.833344, "nll": -5.800237655639648, "elapsed_s": 229.5}57{"event": "train", "step": 1100, "loss": -6.034992694854736, "grad_norm": 0.32199788093566895, "lr": 0.0002643227279408728, "ex_per_s": 304.63668883774756, "data_frac": 0.055088765930304644, "mem_mb": 107.833344, "nll": -6.034992694854736, "elapsed_s": 231.1}58{"event": "train", "step": 1120, "loss": -6.004941940307617, "grad_norm": 0.24643611907958984, "lr": 0.000262797402646797, "ex_per_s": 307.9539997398212, "data_frac": 0.054746458004994024, "mem_mb": 107.833344, "nll": -6.004941940307617, "elapsed_s": 232.8}59{"event": "train", "step": 1140, "loss": -5.985411167144775, "grad_norm": 0.20577466487884521, "lr": 0.0002612453405193386, "ex_per_s": 311.2275013926455, "data_frac": 0.054405051653472196, "mem_mb": 107.833344, "nll": -5.985411167144775, "elapsed_s": 234.5}60{"event": "train", "step": 1160, "loss": -5.935714244842529, "grad_norm": 0.46368923783302307, "lr": 0.0002596669658767859, "ex_per_s": 314.4572092318745, "data_frac": 0.05406846221960052, "mem_mb": 107.833344, "nll": -5.935714244842529, "elapsed_s": 236.1}61{"event": "train", "step": 1180, "loss": -5.921591758728027, "grad_norm": 0.23439240455627441, "lr": 0.00025806271023100637, "ex_per_s": 317.64039784160775, "data_frac": 0.053735668549775906, "mem_mb": 107.833344, "nll": -5.921591758728027, "elapsed_s": 237.8}62{"event": "train", "step": 1200, "loss": -5.922958850860596, "grad_norm": 0.23676902055740356, "lr": 0.00025643301216947504, "ex_per_s": 320.7751893980181, "data_frac": 0.05340901515229355, "mem_mb": 107.833344, "nll": -5.922958850860596, "elapsed_s": 239.5}63{"event": "train", "step": 1220, "loss": -5.991818904876709, "grad_norm": 1.0197410583496094, "lr": 0.0002547783172353695, "ex_per_s": 323.8720522064911, "data_frac": 0.053087167730395055, "mem_mb": 107.833344, "nll": -5.991818904876709, "elapsed_s": 241.1}64{"event": "train", "step": 1240, "loss": -6.0357584953308105, "grad_norm": 0.42604777216911316, "lr": 0.00025309907780576277, "ex_per_s": 326.92552600288025, "data_frac": 0.05276906484558267, "mem_mb": 107.833344, "nll": -6.0357584953308105, "elapsed_s": 242.8}65{"event": "train", "step": 1260, "loss": -6.025754451751709, "grad_norm": 0.25623348355293274, "lr": 0.0002513957529679477, "ex_per_s": 329.9394471289064, "data_frac": 0.052455581202319565, "mem_mb": 107.833344, "nll": -6.025754451751709, "elapsed_s": 244.5}66{"event": "train", "step": 1280, "loss": -6.010115623474121, "grad_norm": 1.287559151649475, "lr": 0.00024966880839392746, "ex_per_s": 332.91256650980756, "data_frac": 0.052144850837442074, "mem_mb": 107.833344, "nll": -6.010115623474121, "elapsed_s": 246.1}67{"event": "train", "step": 1300, "loss": -5.8366780281066895, "grad_norm": 0.6154155135154724, "lr": 0.0002479187162131052, "ex_per_s": 335.8453355908644, "data_frac": 0.051838182356398484, "mem_mb": 107.833344, "nll": -5.8366780281066895, "elapsed_s": 247.8}68{"event": "train", "step": 1320, "loss": -5.762935161590576, "grad_norm": 0.3297886848449707, "lr": 0.00024614595488320833, "ex_per_s": 338.74072139354564, "data_frac": 0.05153752977933261, "mem_mb": 107.833344, "nll": -5.762935161590576, "elapsed_s": 249.4}69{"event": "train", "step": 1340, "loss": -5.928321838378906, "grad_norm": 0.1777764856815338, "lr": 0.00024435100905948387, "ex_per_s": 341.59788834584754, "data_frac": 0.05124076999549221, "mem_mb": 107.833344, "nll": -5.928321838378906, "elapsed_s": 251.1}70{"event": "train", "step": 1360, "loss": -5.960689067840576, "grad_norm": 1.5447614192962646, "lr": 0.0002425343694621974, "ex_per_s": 344.4136911303179, "data_frac": 0.05094886399598604, "mem_mb": 107.833344, "nll": -5.960689067840576, "elapsed_s": 252.8}71{"event": "train", "step": 1380, "loss": -5.934229850769043, "grad_norm": 1.3904041051864624, "lr": 0.00024069653274247583, "ex_per_s": 347.2007713603834, "data_frac": 0.05065940284935048, "mem_mb": 107.833344, "nll": -5.934229850769043, "elapsed_s": 254.4}72{"event": "train", "step": 1400, "loss": -5.761836051940918, "grad_norm": 0.33711862564086914, "lr": 0.0002388380013465276, "ex_per_s": 349.9541299936778, "data_frac": 0.05037479964149972, "mem_mb": 107.833344, "nll": -5.761836051940918, "elapsed_s": 256.1}73{"event": "train", "step": 1420, "loss": -5.971724033355713, "grad_norm": 1.004765510559082, "lr": 0.00023695928337827914, "ex_per_s": 352.67417305844805, "data_frac": 0.0500935365482716, "mem_mb": 107.833344, "nll": -5.971724033355713, "elapsed_s": 257.7}74{"event": "train", "step": 1440, "loss": -5.695178985595703, "grad_norm": 1.5354465246200562, "lr": 0.00023506089246046403, "ex_per_s": 355.36054453487134, "data_frac": 0.04981641638461898, "mem_mb": 107.833344, "nll": -5.695178985595703, "elapsed_s": 259.4}75{"event": "train", "step": 1460, "loss": -5.661196708679199, "grad_norm": 2.4736056327819824, "lr": 0.00023314334759420388, "ex_per_s": 358.01373267528055, "data_frac": 0.0495438397923343, "mem_mb": 107.833344, "nll": -5.661196708679199, "elapsed_s": 261.0}76{"event": "train", "step": 1480, "loss": -6.028436183929443, "grad_norm": 0.5933417081832886, "lr": 0.0002312071730171186, "ex_per_s": 360.63371998845577, "data_frac": 0.04927494394227648, "mem_mb": 107.833344, "nll": -6.028436183929443, "elapsed_s": 262.7}77{"event": "train", "step": 1500, "loss": -5.926986217498779, "grad_norm": 0.3081326186656952, "lr": 0.0002292528980600049, "ex_per_s": 363.22282650602364, "data_frac": 0.04900910831610267, "mem_mb": 107.833344, "nll": -5.926986217498779, "elapsed_s": 264.3}78{"event": "val", "step": 1500, "val_nll": -6.392438699657211, "best": -6.393413394690595, "bad": 1, "mem_mb": 107.833344, "elapsed_s": 327.7}79{"event": "train", "step": 1520, "loss": -5.925544738769531, "grad_norm": 0.482468843460083, "lr": 0.00022728105700212285, "ex_per_s": 295.3699832111299, "data_frac": 0.039365662683979764, "mem_mb": 107.833344, "nll": -5.925544738769531, "elapsed_s": 329.4}80{"event": "train", "step": 1540, "loss": -5.845150947570801, "grad_norm": 0.2051572948694229, "lr": 0.00022529218892512886, "ex_per_s": 297.7443250366525, "data_frac": 0.03920041232721277, "mem_mb": 107.833344, "nll": -5.845150947570801, "elapsed_s": 331.1}81{"event": "train", "step": 1560, "loss": -5.894013404846191, "grad_norm": 0.5693384408950806, "lr": 0.00022328683756569652, "ex_per_s": 300.09377685096547, "data_frac": 0.039038914346128925, "mem_mb": 107.833344, "nll": -5.894013404846191, "elapsed_s": 332.7}82{"event": "train", "step": 1580, "loss": -5.9503326416015625, "grad_norm": 0.5666359066963196, "lr": 0.0002212655511668642, "ex_per_s": 302.4209660226815, "data_frac": 0.038875723160121306, "mem_mb": 107.833344, "nll": -5.9503326416015625, "elapsed_s": 334.4}83{"event": "train", "step": 1600, "loss": -5.931901454925537, "grad_norm": 0.16876180469989777, "lr": 0.00021922888232815089, "ex_per_s": 304.72573495566405, "data_frac": 0.03871536028575151, "mem_mb": 107.833344, "nll": -5.931901454925537, "elapsed_s": 336.1}84{"event": "train", "step": 1620, "loss": -5.841105937957764, "grad_norm": 0.4480663537979126, "lr": 0.00021717738785448102, "ex_per_s": 307.0071114237594, "data_frac": 0.038556861842969505, "mem_mb": 107.833344, "nll": -5.841105937957764, "elapsed_s": 337.8}85{"event": "train", "step": 1640, "loss": -6.054318428039551, "grad_norm": 0.7727385759353638, "lr": 0.00021511162860395977, "ex_per_s": 309.2675403857308, "data_frac": 0.03839867481790049, "mem_mb": 107.833344, "nll": -6.054318428039551, "elapsed_s": 339.4}86{"event": "train", "step": 1660, "loss": -5.988401889801025, "grad_norm": 1.146742582321167, "lr": 0.0002130321693345397, "ex_per_s": 311.50525362755224, "data_frac": 0.03824465641644509, "mem_mb": 107.833344, "nll": -5.988401889801025, "elapsed_s": 341.1}87{"event": "train", "step": 1680, "loss": -5.838001728057861, "grad_norm": 0.49990779161453247, "lr": 0.0002109395785496218, "ex_per_s": 313.7218607461316, "data_frac": 0.038090946198368086, "mem_mb": 107.833344, "nll": -5.838001728057861, "elapsed_s": 342.8}88{"event": "train", "step": 1700, "loss": -5.824907302856445, "grad_norm": 2.6474342346191406, "lr": 0.00020883442834263225, "ex_per_s": 315.9165694209953, "data_frac": 0.03793811992185743, "mem_mb": 107.833344, "nll": -5.824907302856445, "elapsed_s": 344.4}89{"event": "train", "step": 1720, "loss": -5.8605523109436035, "grad_norm": 0.987604558467865, "lr": 0.00020671729424061788, "ex_per_s": 318.0910808149127, "data_frac": 0.03778730008490325, "mem_mb": 107.833344, "nll": -5.8605523109436035, "elapsed_s": 346.1}90{"event": "train", "step": 1740, "loss": -6.246664047241211, "grad_norm": 1.1393071413040161, "lr": 0.00020458875504690275, "ex_per_s": 320.24319361341577, "data_frac": 0.037636994819026715, "mem_mb": 107.833344, "nll": -6.246664047241211, "elapsed_s": 347.8}91{"event": "train", "step": 1760, "loss": -5.864348888397217, "grad_norm": 1.4316226243972778, "lr": 0.000202449392682849, "ex_per_s": 322.37859522697323, "data_frac": 0.037488624749339436, "mem_mb": 107.833344, "nll": -5.864348888397217, "elapsed_s": 349.4}92{"event": "train", "step": 1780, "loss": -5.8454813957214355, "grad_norm": 0.5364003777503967, "lr": 0.00020029979202876563, "ex_per_s": 324.494365332878, "data_frac": 0.037341265168161504, "mem_mb": 107.833344, "nll": -5.8454813957214355, "elapsed_s": 351.1}93{"event": "train", "step": 1800, "loss": -6.012762069702148, "grad_norm": 1.2668790817260742, "lr": 0.00019814054076400787, "ex_per_s": 326.58991689473004, "data_frac": 0.03719939912175206, "mem_mb": 107.833344, "nll": -6.012762069702148, "elapsed_s": 352.8}94{"event": "train", "step": 1820, "loss": -5.905048847198486, "grad_norm": 0.7887094616889954, "lr": 0.00019597222920631147, "ex_per_s": 328.6701915222431, "data_frac": 0.037056629823607336, "mem_mb": 107.833344, "nll": -5.905048847198486, "elapsed_s": 354.4}95{"event": "train", "step": 1840, "loss": -6.049783229827881, "grad_norm": 0.5747636556625366, "lr": 0.0001937954501504058, "ex_per_s": 330.72978740068044, "data_frac": 0.03691397393424992, "mem_mb": 107.833344, "nll": -6.049783229827881, "elapsed_s": 356.1}96{"event": "train", "step": 1860, "loss": -6.034173011779785, "grad_norm": 0.35871174931526184, "lr": 0.0001916107987059498, "ex_per_s": 332.77242676911067, "data_frac": 0.036772677295207626, "mem_mb": 107.833344, "nll": -6.034173011779785, "elapsed_s": 357.8}97{"event": "train", "step": 1880, "loss": -5.922298908233643, "grad_norm": 0.1988658457994461, "lr": 0.00018941887213483484, "ex_per_s": 334.7935092713221, "data_frac": 0.036634922667638306, "mem_mb": 107.833344, "nll": -5.922298908233643, "elapsed_s": 359.4}98{"event": "train", "step": 1900, "loss": -5.955255508422852, "grad_norm": 0.8963642716407776, "lr": 0.00018722026968789906, "ex_per_s": 336.79738308369747, "data_frac": 0.036496983994005874, "mem_mb": 107.833344, "nll": -5.955255508422852, "elapsed_s": 361.1}99{"event": "train", "step": 1920, "loss": -5.933923244476318, "grad_norm": 0.2805110216140747, "lr": 0.00018501559244109864, "ex_per_s": 338.78332814749, "data_frac": 0.0363600726638322, "mem_mb": 107.833344, "nll": -5.933923244476318, "elapsed_s": 362.8}100{"event": "train", "step": 1940, "loss": -6.028850555419922, "grad_norm": 0.7048793435096741, "lr": 0.00018280544313117932, "ex_per_s": 340.7503434873697, "data_frac": 0.03622348134047789, "mem_mb": 107.833344, "nll": -6.028850555419922, "elapsed_s": 364.4}101{"event": "train", "step": 1960, "loss": -5.896194934844971, "grad_norm": 0.17662997543811798, "lr": 0.00018059042599089415, "ex_per_s": 342.70116881628473, "data_frac": 0.03608870508219495, "mem_mb": 107.833344, "nll": -5.896194934844971, "elapsed_s": 366.1}102{"event": "train", "step": 1980, "loss": -5.882737636566162, "grad_norm": 2.024052381515503, "lr": 0.00017837114658381247, "ex_per_s": 344.6333304731371, "data_frac": 0.035956754844546086, "mem_mb": 107.833344, "nll": -5.882737636566162, "elapsed_s": 367.7}103{"event": "train", "step": 2000, "loss": -5.894165515899658, "grad_norm": 0.20714247226715088, "lr": 0.00017614821163876486, "ex_per_s": 346.54826377061846, "data_frac": 0.03582630094813454, "mem_mb": 107.833344, "nll": -5.894165515899658, "elapsed_s": 369.4}104{"event": "val", "step": 2000, "val_nll": -6.397167645657429, "best": -6.397167645657429, "bad": 0, "mem_mb": 107.833344, "elapsed_s": 432.5}105{"event": "train", "step": 2020, "loss": -5.899983882904053, "grad_norm": 1.16838800907135, "lr": 0.00017392222888396941, "ex_per_s": 297.73316106923374, "data_frac": 0.03050158776762585, "mem_mb": 107.833344, "nll": -5.899983882904053, "elapsed_s": 434.3}106{"event": "train", "step": 2040, "loss": -5.853193759918213, "grad_norm": 0.38793277740478516, "lr": 0.000171693806880885, "ex_per_s": 299.5283818726306, "data_frac": 0.030409803996350914, "mem_mb": 107.833344, "nll": -5.853193759918213, "elapsed_s": 435.9}107{"event": "train", "step": 2060, "loss": -6.108746528625488, "grad_norm": 0.25437477231025696, "lr": 0.0001694635548578366, "ex_per_s": 301.30988948286074, "data_frac": 0.030319008712160717, "mem_mb": 107.833344, "nll": -6.108746528625488, "elapsed_s": 437.6}108{"event": "train", "step": 2080, "loss": -5.944520950317383, "grad_norm": 0.30934756994247437, "lr": 0.0001672320825434584, "ex_per_s": 303.0779908984772, "data_frac": 0.030229334839490703, "mem_mb": 107.833344, "nll": -5.944520950317383, "elapsed_s": 439.3}109{"event": "train", "step": 2100, "loss": -5.879607677459717, "grad_norm": 0.7044767141342163, "lr": 0.000165, "ex_per_s": 304.83242406258705, "data_frac": 0.03014015614215138, "mem_mb": 107.833344, "nll": -5.879607677459717, "elapsed_s": 440.9}110{"event": "train", "step": 2120, "loss": -6.126163005828857, "grad_norm": 0.15331979095935822, "lr": 0.0001627679174565416, "ex_per_s": 306.5746305455251, "data_frac": 0.030051648239540804, "mem_mb": 107.833344, "nll": -6.126163005828857, "elapsed_s": 442.6}111{"event": "train", "step": 2140, "loss": -5.933512210845947, "grad_norm": 0.3582135736942291, "lr": 0.00016053644514216336, "ex_per_s": 308.3033448945127, "data_frac": 0.02996298623401857, "mem_mb": 107.833344, "nll": -5.933512210845947, "elapsed_s": 444.3}112{"event": "train", "step": 2160, "loss": -5.8968892097473145, "grad_norm": 0.4182126522064209, "lr": 0.00015830619311911498, "ex_per_s": 310.0178250926579, "data_frac": 0.029876145239535558, "mem_mb": 107.833344, "nll": -5.8968892097473145, "elapsed_s": 446.0}113{"event": "train", "step": 2180, "loss": -6.037850379943848, "grad_norm": 0.33288368582725525, "lr": 0.00015607777111603058, "ex_per_s": 311.7201946979869, "data_frac": 0.029789642740643635, "mem_mb": 107.833344, "nll": -6.037850379943848, "elapsed_s": 447.6}114{"event": "train", "step": 2200, "loss": -5.942330360412598, "grad_norm": 0.24846844375133514, "lr": 0.00015385178836123514, "ex_per_s": 313.4102384987008, "data_frac": 0.029704401726088948, "mem_mb": 107.833344, "nll": -5.942330360412598, "elapsed_s": 449.3}115{"event": "train", "step": 2220, "loss": -5.794544219970703, "grad_norm": 0.6400553584098816, "lr": 0.00015162885341618752, "ex_per_s": 315.08419342166286, "data_frac": 0.02962192514783435, "mem_mb": 107.833344, "nll": -5.794544219970703, "elapsed_s": 451.0}116{"event": "train", "step": 2240, "loss": -5.932826042175293, "grad_norm": 0.8686032295227051, "lr": 0.00014940957400910585, "ex_per_s": 316.7513891986504, "data_frac": 0.029540020154697214, "mem_mb": 107.833344, "nll": -5.932826042175293, "elapsed_s": 452.6}117{"event": "train", "step": 2260, "loss": -5.662790775299072, "grad_norm": 0.8145262598991394, "lr": 0.00014719455686882068, "ex_per_s": 318.4088743359563, "data_frac": 0.02945698291202384, "mem_mb": 107.833344, "nll": -5.662790775299072, "elapsed_s": 454.3}118{"event": "train", "step": 2280, "loss": -5.913989543914795, "grad_norm": 0.6972472667694092, "lr": 0.0001449844075589013, "ex_per_s": 320.05554661313255, "data_frac": 0.029374505613270693, "mem_mb": 107.833344, "nll": -5.913989543914795, "elapsed_s": 456.0}119{"event": "train", "step": 2300, "loss": -5.959708213806152, "grad_norm": 0.1949561983346939, "lr": 0.0001427797303121009, "ex_per_s": 321.6899955589802, "data_frac": 0.029291778481969455, "mem_mb": 107.833344, "nll": -5.959708213806152, "elapsed_s": 457.6}120{"event": "train", "step": 2320, "loss": -6.052186489105225, "grad_norm": 0.20010913908481598, "lr": 0.00014058112786516518, "ex_per_s": 323.3123941146767, "data_frac": 0.029210371586102717, "mem_mb": 107.833344, "nll": -6.052186489105225, "elapsed_s": 459.3}121{"event": "train", "step": 2340, "loss": -5.826751708984375, "grad_norm": 0.362045556306839, "lr": 0.00013838920129405016, "ex_per_s": 324.92273919787846, "data_frac": 0.029129060587251624, "mem_mb": 107.833344, "nll": -5.826751708984375, "elapsed_s": 461.0}122{"event": "train", "step": 2360, "loss": -5.899625778198242, "grad_norm": 0.3320169448852539, "lr": 0.0001362045498495942, "ex_per_s": 326.52186758202754, "data_frac": 0.029048627050761007, "mem_mb": 107.833344, "nll": -5.899625778198242, "elapsed_s": 462.6}123{"event": "train", "step": 2380, "loss": -5.910579204559326, "grad_norm": 0.5055805444717407, "lr": 0.00013402777079368853, "ex_per_s": 328.10942656731436, "data_frac": 0.028968403253220493, "mem_mb": 107.833344, "nll": -5.910579204559326, "elapsed_s": 464.3}124{"event": "train", "step": 2400, "loss": -5.930047512054443, "grad_norm": 0.5086067914962769, "lr": 0.00013185945923599213, "ex_per_s": 329.68382866691667, "data_frac": 0.02888890861690548, "mem_mb": 107.833344, "nll": -5.930047512054443, "elapsed_s": 465.9}125{"event": "train", "step": 2420, "loss": -5.980703830718994, "grad_norm": 0.5380949974060059, "lr": 0.00012970020797123432, "ex_per_s": 331.24400729681184, "data_frac": 0.028809249182909124, "mem_mb": 107.833344, "nll": -5.980703830718994, "elapsed_s": 467.6}126{"event": "train", "step": 2440, "loss": -6.063994884490967, "grad_norm": 0.9947973489761353, "lr": 0.000127550607317151, "ex_per_s": 332.7938000422632, "data_frac": 0.028729853856768364, "mem_mb": 107.833344, "nll": -6.063994884490967, "elapsed_s": 469.3}127{"event": "train", "step": 2460, "loss": -6.107862949371338, "grad_norm": 0.332720547914505, "lr": 0.00012541124495309727, "ex_per_s": 334.33339586886126, "data_frac": 0.02865205219612736, "mem_mb": 107.833344, "nll": -6.107862949371338, "elapsed_s": 471.0}128{"event": "train", "step": 2480, "loss": -5.922675132751465, "grad_norm": 0.18116086721420288, "lr": 0.00012328270575938212, "ex_per_s": 335.860596233069, "data_frac": 0.02857423465570758, "mem_mb": 107.833344, "nll": -5.922675132751465, "elapsed_s": 472.6}129{"event": "train", "step": 2500, "loss": -5.719053745269775, "grad_norm": 0.7611356973648071, "lr": 0.00012116557165736775, "ex_per_s": 337.3769605904166, "data_frac": 0.028498854533043134, "mem_mb": 107.833344, "nll": -5.719053745269775, "elapsed_s": 474.3}130{"event": "val", "step": 2500, "val_nll": -6.3994315438002305, "best": -6.3994315438002305, "bad": 0, "mem_mb": 107.833344, "elapsed_s": 537.5}131{"event": "train", "step": 2520, "loss": -5.92976188659668, "grad_norm": 0.6460935473442078, "lr": 0.00011906042145037817, "ex_per_s": 299.09814226694834, "data_frac": 0.02508546869245908, "mem_mb": 107.833344, "nll": -5.92976188659668, "elapsed_s": 539.3}132{"event": "train", "step": 2540, "loss": -5.992037296295166, "grad_norm": 0.14566488564014435, "lr": 0.00011696783066546026, "ex_per_s": 300.53845928218664, "data_frac": 0.025028687234637505, "mem_mb": 107.833344, "nll": -5.992037296295166, "elapsed_s": 540.9}133{"event": "train", "step": 2560, "loss": -5.924813270568848, "grad_norm": 0.3907863199710846, "lr": 0.00011488837139604017, "ex_per_s": 301.9719926059591, "data_frac": 0.0249715935767254, "mem_mb": 107.833344, "nll": -5.924813270568848, "elapsed_s": 542.6}134{"event": "train", "step": 2580, "loss": -5.906313419342041, "grad_norm": 0.29225048422813416, "lr": 0.00011282261214551898, "ex_per_s": 303.39695169006586, "data_frac": 0.02491523204617773, "mem_mb": 107.833344, "nll": -5.906313419342041, "elapsed_s": 544.3}135{"event": "train", "step": 2600, "loss": -5.832542419433594, "grad_norm": 0.6862005591392517, "lr": 0.00011077111767184915, "ex_per_s": 304.8132133249624, "data_frac": 0.02485931221680292, "mem_mb": 107.833344, "nll": -5.832542419433594, "elapsed_s": 546.0}136{"event": "train", "step": 2620, "loss": -5.834075927734375, "grad_norm": 1.4808331727981567, "lr": 0.00010873444883313583, "ex_per_s": 306.22048879227555, "data_frac": 0.024803201054674096, "mem_mb": 107.833344, "nll": -5.834075927734375, "elapsed_s": 547.6}137{"event": "train", "step": 2640, "loss": -5.960031986236572, "grad_norm": 0.27051523327827454, "lr": 0.00010671316243430345, "ex_per_s": 307.619480042353, "data_frac": 0.024747922218444567, "mem_mb": 107.833344, "nll": -5.960031986236572, "elapsed_s": 549.3}138{"event": "train", "step": 2660, "loss": -5.770601749420166, "grad_norm": 0.20710638165473938, "lr": 0.00010470781107487112, "ex_per_s": 309.0089759040845, "data_frac": 0.024692742127260112, "mem_mb": 107.833344, "nll": -5.770601749420166, "elapsed_s": 551.0}139{"event": "train", "step": 2680, "loss": -5.942142486572266, "grad_norm": 0.4472395181655884, "lr": 0.00010271894299787715, "ex_per_s": 310.3895968653273, "data_frac": 0.02463789935907728, "mem_mb": 107.833344, "nll": -5.942142486572266, "elapsed_s": 552.6}140{"event": "train", "step": 2700, "loss": -6.101226806640625, "grad_norm": 0.9745652675628662, "lr": 0.00010074710193999506, "ex_per_s": 311.76098786841374, "data_frac": 0.02458560047789811, "mem_mb": 107.833344, "nll": -6.101226806640625, "elapsed_s": 554.3}141{"event": "train", "step": 2720, "loss": -5.8509345054626465, "grad_norm": 0.39205971360206604, "lr": 9.879282698288144e-05, "ex_per_s": 313.12676394764543, "data_frac": 0.024531923107321045, "mem_mb": 107.833344, "nll": -5.8509345054626465, "elapsed_s": 556.0}142{"event": "train", "step": 2740, "loss": -6.093481063842773, "grad_norm": 0.29937657713890076, "lr": 9.685665240579613e-05, "ex_per_s": 314.48367523497706, "data_frac": 0.02447817037255436, "mem_mb": 107.833344, "nll": -6.093481063842773, "elapsed_s": 557.7}143{"event": "train", "step": 2760, "loss": -5.931821823120117, "grad_norm": 0.4646196961402893, "lr": 9.4939107539536e-05, "ex_per_s": 315.83302674633643, "data_frac": 0.024424872295411373, "mem_mb": 107.833344, "nll": -5.931821823120117, "elapsed_s": 559.3}144{"event": "train", "step": 2780, "loss": -5.974265098571777, "grad_norm": 0.7091771364212036, "lr": 9.304071662172087e-05, "ex_per_s": 317.1752345203907, "data_frac": 0.024371938143233515, "mem_mb": 107.833344, "nll": -5.974265098571777, "elapsed_s": 561.0}145{"event": "train", "step": 2800, "loss": -5.954347133636475, "grad_norm": 1.3524482250213623, "lr": 9.116199865347241e-05, "ex_per_s": 318.5098076182935, "data_frac": 0.024318816750308836, "mem_mb": 107.833344, "nll": -5.954347133636475, "elapsed_s": 562.7}146{"event": "train", "step": 2820, "loss": -5.977738380432129, "grad_norm": 0.4127412736415863, "lr": 8.930346725752419e-05, "ex_per_s": 319.8376597590541, "data_frac": 0.024266571463779238, "mem_mb": 107.833344, "nll": -5.977738380432129, "elapsed_s": 564.3}147{"event": "train", "step": 2840, "loss": -5.852445125579834, "grad_norm": 0.5236789584159851, "lr": 8.746563053780262e-05, "ex_per_s": 321.1573856646905, "data_frac": 0.024214295514090284, "mem_mb": 107.833344, "nll": -5.852445125579834, "elapsed_s": 566.0}148{"event": "train", "step": 2860, "loss": -6.022107124328613, "grad_norm": 0.4373396039009094, "lr": 8.564899094051614e-05, "ex_per_s": 322.47048983678246, "data_frac": 0.024162658134855047, "mem_mb": 107.833344, "nll": -6.022107124328613, "elapsed_s": 567.7}149{"event": "train", "step": 2880, "loss": -5.745233058929443, "grad_norm": 0.16559730470180511, "lr": 8.385404511679161e-05, "ex_per_s": 323.7760756656253, "data_frac": 0.024111627549959654, "mem_mb": 107.833344, "nll": -5.745233058929443, "elapsed_s": 569.3}150{"event": "train", "step": 2900, "loss": -5.725635528564453, "grad_norm": 1.9291961193084717, "lr": 8.208128378689484e-05, "ex_per_s": 325.0734852954457, "data_frac": 0.02406168627814455, "mem_mb": 107.833344, "nll": -5.725635528564453, "elapsed_s": 571.0}151{"event": "train", "step": 2920, "loss": -5.946223735809326, "grad_norm": 0.9593944549560547, "lr": 8.03311916060725e-05, "ex_per_s": 326.36468808110317, "data_frac": 0.024010394128116196, "mem_mb": 107.833344, "nll": -5.946223735809326, "elapsed_s": 572.7}152{"event": "train", "step": 2940, "loss": -5.792087078094482, "grad_norm": 2.5735175609588623, "lr": 7.860424703205225e-05, "ex_per_s": 327.64662299191656, "data_frac": 0.023961603122226584, "mem_mb": 107.833344, "nll": -5.792087078094482, "elapsed_s": 574.3}153{"event": "train", "step": 2960, "loss": -5.913642883300781, "grad_norm": 0.21530680358409882, "lr": 7.690092219423722e-05, "ex_per_s": 328.92315290997567, "data_frac": 0.02391173143921466, "mem_mb": 107.833344, "nll": -5.913642883300781, "elapsed_s": 576.0}154{"event": "train", "step": 2980, "loss": -5.607391357421875, "grad_norm": 6.090331554412842, "lr": 7.522168276463048e-05, "ex_per_s": 330.19158313772886, "data_frac": 0.023861953572442657, "mem_mb": 107.833344, "nll": -5.607391357421875, "elapsed_s": 577.6}155{"event": "train", "step": 3000, "loss": -5.958337783813477, "grad_norm": 0.15975210070610046, "lr": 7.356698783052496e-05, "ex_per_s": 331.4537127303271, "data_frac": 0.023812689012320823, "mem_mb": 107.833344, "nll": -5.958337783813477, "elapsed_s": 579.3}156{"event": "val", "step": 3000, "val_nll": -6.3984386182675905, "best": -6.3994315438002305, "bad": 1, "mem_mb": 107.833344, "elapsed_s": 642.8}157{"event": "train", "step": 3020, "loss": -5.9363694190979, "grad_norm": 0.8629269599914551, "lr": 7.193728976899362e-05, "ex_per_s": 299.89764583013607, "data_frac": 0.021419979458085103, "mem_mb": 107.833344, "nll": -5.9363694190979, "elapsed_s": 644.5}158{"event": "train", "step": 3040, "loss": -5.791289329528809, "grad_norm": 2.0522408485412598, "lr": 7.033303412321404e-05, "ex_per_s": 301.1022418595421, "data_frac": 0.021381794784960523, "mem_mb": 107.833344, "nll": -5.791289329528809, "elapsed_s": 646.2}159{"event": "train", "step": 3060, "loss": -6.018851280212402, "grad_norm": 0.5756768584251404, "lr": 6.875465948066139e-05, "ex_per_s": 302.3010397918525, "data_frac": 0.0213435931219388, "mem_mb": 107.833344, "nll": -6.018851280212402, "elapsed_s": 647.9}160{"event": "train", "step": 3080, "loss": -5.9473490715026855, "grad_norm": 0.13481874763965607, "lr": 6.720259735320299e-05, "ex_per_s": 303.4948877519141, "data_frac": 0.021305213172781765, "mem_mb": 107.833344, "nll": -5.9473490715026855, "elapsed_s": 649.5}161{"event": "train", "step": 3100, "loss": -5.981713771820068, "grad_norm": 1.0422130823135376, "lr": 6.567727205912722e-05, "ex_per_s": 304.68290199747725, "data_frac": 0.0212673798981643, "mem_mb": 107.833344, "nll": -5.981713771820068, "elapsed_s": 651.2}162{"event": "train", "step": 3120, "loss": -6.0689826011657715, "grad_norm": 0.383851557970047, "lr": 6.417910060713926e-05, "ex_per_s": 305.86494607218384, "data_frac": 0.02122972157851023, "mem_mb": 107.833344, "nll": -6.0689826011657715, "elapsed_s": 652.9}163{"event": "train", "step": 3140, "loss": -5.986001968383789, "grad_norm": 0.29263022541999817, "lr": 6.270849258235485e-05, "ex_per_s": 307.0395215504437, "data_frac": 0.02119384808510633, "mem_mb": 107.833344, "nll": -5.986001968383789, "elapsed_s": 654.6}164{"event": "train", "step": 3160, "loss": -5.992488861083984, "grad_norm": 0.11407465487718582, "lr": 6.126585003432406e-05, "ex_per_s": 308.2093241622843, "data_frac": 0.021156915355050276, "mem_mb": 107.833344, "nll": -5.992488861083984, "elapsed_s": 656.2}165{"event": "train", "step": 3180, "loss": -6.032766342163086, "grad_norm": 0.7635337710380554, "lr": 5.985156736711483e-05, "ex_per_s": 309.37425799831107, "data_frac": 0.021119810438927772, "mem_mb": 107.833344, "nll": -6.032766342163086, "elapsed_s": 657.9}166{"event": "train", "step": 3200, "loss": -5.962067127227783, "grad_norm": 0.4976271986961365, "lr": 5.846603123148688e-05, "ex_per_s": 310.53349129874175, "data_frac": 0.021083130968099908, "mem_mb": 107.833344, "nll": -5.962067127227783, "elapsed_s": 659.6}167{"event": "train", "step": 3220, "loss": -5.85957670211792, "grad_norm": 0.9409359097480774, "lr": 5.710962041918516e-05, "ex_per_s": 311.6851524996358, "data_frac": 0.021047097974512614, "mem_mb": 107.833344, "nll": -5.85957670211792, "elapsed_s": 661.2}168{"event": "train", "step": 3240, "loss": -5.771299362182617, "grad_norm": 0.17690595984458923, "lr": 5.578270575938211e-05, "ex_per_s": 312.8320748402252, "data_frac": 0.02101109604283043, "mem_mb": 107.833344, "nll": -5.771299362182617, "elapsed_s": 662.9}169{"event": "train", "step": 3260, "loss": -5.925745964050293, "grad_norm": 0.2579410970211029, "lr": 5.448565001729653e-05, "ex_per_s": 313.9743827193447, "data_frac": 0.020974827980508648, "mem_mb": 107.833344, "nll": -5.925745964050293, "elapsed_s": 664.6}170{"event": "train", "step": 3280, "loss": -5.963176250457764, "grad_norm": 0.23690474033355713, "lr": 5.3218807795017296e-05, "ex_per_s": 315.1103482020645, "data_frac": 0.02093895299329857, "mem_mb": 107.833344, "nll": -5.963176250457764, "elapsed_s": 666.2}171{"event": "train", "step": 3300, "loss": -5.720840930938721, "grad_norm": 2.147510290145874, "lr": 5.198252543455864e-05, "ex_per_s": 316.24118178863864, "data_frac": 0.02090329114114642, "mem_mb": 107.833344, "nll": -5.720840930938721, "elapsed_s": 667.9}172{"event": "train", "step": 3320, "loss": -5.93883752822876, "grad_norm": 0.31516414880752563, "lr": 5.07771409231737e-05, "ex_per_s": 317.36609899737743, "data_frac": 0.020867512682461106, "mem_mb": 107.833344, "nll": -5.93883752822876, "elapsed_s": 669.6}173{"event": "train", "step": 3340, "loss": -5.889064311981201, "grad_norm": 0.42035260796546936, "lr": 4.960298380095221e-05, "ex_per_s": 318.4862420502944, "data_frac": 0.02083198793834394, "mem_mb": 107.833344, "nll": -5.889064311981201, "elapsed_s": 671.2}174{"event": "train", "step": 3360, "loss": -5.887307167053223, "grad_norm": 0.30520325899124146, "lr": 4.846037507072751e-05, "ex_per_s": 319.60125286629443, "data_frac": 0.02079664571864915, "mem_mb": 107.833344, "nll": -5.887307167053223, "elapsed_s": 672.9}175{"event": "train", "step": 3380, "loss": -6.076573371887207, "grad_norm": 0.36024823784828186, "lr": 4.7349627110317556e-05, "ex_per_s": 320.70933075515757, "data_frac": 0.020763167941090108, "mem_mb": 107.833344, "nll": -6.076573371887207, "elapsed_s": 674.5}176{"event": "train", "step": 3400, "loss": -5.850979328155518, "grad_norm": 1.1655066013336182, "lr": 4.6271043587123986e-05, "ex_per_s": 321.8142767988545, "data_frac": 0.020728498586886995, "mem_mb": 107.833344, "nll": -5.850979328155518, "elapsed_s": 676.2}177{"event": "train", "step": 3420, "loss": -5.810490131378174, "grad_norm": 0.8732507228851318, "lr": 4.52249193751124e-05, "ex_per_s": 322.91551689849376, "data_frac": 0.020693905903551763, "mem_mb": 107.833344, "nll": -5.810490131378174, "elapsed_s": 677.9}178{"event": "train", "step": 3440, "loss": -5.886451244354248, "grad_norm": 0.3542679250240326, "lr": 4.4211540474196926e-05, "ex_per_s": 324.01261086719285, "data_frac": 0.02065967787754825, "mem_mb": 107.833344, "nll": -5.886451244354248, "elapsed_s": 679.5}179{"event": "train", "step": 3460, "loss": -6.031919479370117, "grad_norm": 0.4259660840034485, "lr": 4.3231183932050546e-05, "ex_per_s": 325.10494021526495, "data_frac": 0.020625772665825187, "mem_mb": 107.833344, "nll": -6.031919479370117, "elapsed_s": 681.2}180{"event": "train", "step": 3480, "loss": -5.88660192489624, "grad_norm": 0.731090247631073, "lr": 4.228411776836306e-05, "ex_per_s": 326.1928131797367, "data_frac": 0.020592119455794942, "mem_mb": 107.833344, "nll": -5.88660192489624, "elapsed_s": 682.8}181{"event": "train", "step": 3500, "loss": -6.007784366607666, "grad_norm": 0.29522883892059326, "lr": 4.137060090156723e-05, "ex_per_s": 327.2754004214328, "data_frac": 0.020558670223450246, "mem_mb": 107.833344, "nll": -6.007784366607666, "elapsed_s": 684.5}182{"event": "val", "step": 3500, "val_nll": -6.3999923274301365, "best": -6.3999923274301365, "bad": 0, "mem_mb": 107.833344, "elapsed_s": 747.8}183{"event": "train", "step": 3520, "loss": -5.914472579956055, "grad_norm": 0.524591863155365, "lr": 4.049088307805306e-05, "ex_per_s": 300.5704915648809, "data_frac": 0.018790244214890905, "mem_mb": 107.833344, "nll": -5.914472579956055, "elapsed_s": 749.6}184{"event": "train", "step": 3540, "loss": -6.113830089569092, "grad_norm": 0.7230539321899414, "lr": 3.964520480388956e-05, "ex_per_s": 301.6046433091078, "data_frac": 0.01876374154925019, "mem_mb": 107.833344, "nll": -6.113830089569092, "elapsed_s": 751.2}185{"event": "train", "step": 3560, "loss": -6.1137189865112305, "grad_norm": 0.1559121459722519, "lr": 3.88337972790728e-05, "ex_per_s": 302.63504268493324, "data_frac": 0.018736824561506348, "mem_mb": 107.833344, "nll": -6.1137189865112305, "elapsed_s": 752.9}186{"event": "train", "step": 3580, "loss": -5.879316806793213, "grad_norm": 0.24926596879959106, "lr": 3.805688233431824e-05, "ex_per_s": 303.66032303410935, "data_frac": 0.018710447515673488, "mem_mb": 107.833344, "nll": -5.879316806793213, "elapsed_s": 754.6}187{"event": "train", "step": 3600, "loss": -6.073864459991455, "grad_norm": 0.5073188543319702, "lr": 3.7314672370414325e-05, "ex_per_s": 304.6816884481191, "data_frac": 0.018684505979180147, "mem_mb": 107.833344, "nll": -6.073864459991455, "elapsed_s": 756.2}188{"event": "train", "step": 3620, "loss": -5.872777462005615, "grad_norm": 1.7644224166870117, "lr": 3.660737030015427e-05, "ex_per_s": 305.69880890445467, "data_frac": 0.01865774947603951, "mem_mb": 107.833344, "nll": -5.872777462005615, "elapsed_s": 757.9}189{"event": "train", "step": 3640, "loss": -6.00703239440918, "grad_norm": 0.6158971190452576, "lr": 3.5935169492861716e-05, "ex_per_s": 306.71129294648415, "data_frac": 0.01863106264732958, "mem_mb": 107.833344, "nll": -6.00703239440918, "elapsed_s": 759.6}190{"event": "train", "step": 3660, "loss": -5.784574031829834, "grad_norm": 0.5204005241394043, "lr": 3.52982537215255e-05, "ex_per_s": 307.7194118280861, "data_frac": 0.01860436638899639, "mem_mb": 107.833344, "nll": -5.784574031829834, "elapsed_s": 761.3}191{"event": "train", "step": 3680, "loss": -5.956212997436523, "grad_norm": 0.5591573715209961, "lr": 3.469679711255794e-05, "ex_per_s": 308.7237858027081, "data_frac": 0.018577988188488978, "mem_mb": 107.833344, "nll": -5.956212997436523, "elapsed_s": 762.9}192{"event": "train", "step": 3700, "loss": -6.038631439208984, "grad_norm": 1.0399320125579834, "lr": 3.41309640981904e-05, "ex_per_s": 309.7233095371414, "data_frac": 0.018551831214589944, "mem_mb": 107.833344, "nll": -6.038631439208984, "elapsed_s": 764.6}193{"event": "train", "step": 3720, "loss": -5.90824031829834, "grad_norm": 0.5664897561073303, "lr": 3.3600909371519225e-05, "ex_per_s": 310.7182061620061, "data_frac": 0.018525801888743337, "mem_mb": 107.833344, "nll": -5.90824031829834, "elapsed_s": 766.3}194{"event": "train", "step": 3740, "loss": -6.167071342468262, "grad_norm": 0.8735705614089966, "lr": 3.31067778442141e-05, "ex_per_s": 311.708638911317, "data_frac": 0.018500041620356535, "mem_mb": 107.833344, "nll": -6.167071342468262, "elapsed_s": 767.9}195{"event": "train", "step": 3760, "loss": -6.109151363372803, "grad_norm": 0.31054356694221497, "lr": 3.2648704606900735e-05, "ex_per_s": 312.69640824800445, "data_frac": 0.018474775174442272, "mem_mb": 107.833344, "nll": -6.109151363372803, "elapsed_s": 769.6}196{"event": "train", "step": 3780, "loss": -5.938636302947998, "grad_norm": 0.17457985877990723, "lr": 3.222681489222838e-05, "ex_per_s": 313.6796651291418, "data_frac": 0.01844933799767787, "mem_mb": 107.833344, "nll": -5.938636302947998, "elapsed_s": 771.3}197{"event": "train", "step": 3800, "loss": -6.072531223297119, "grad_norm": 0.17048710584640503, "lr": 3.1841224040632484e-05, "ex_per_s": 314.65979764415306, "data_frac": 0.018423892386973128, "mem_mb": 107.833344, "nll": -6.072531223297119, "elapsed_s": 772.9}198{"event": "train", "step": 3820, "loss": -5.890435218811035, "grad_norm": 0.880724310874939, "lr": 3.149203746880175e-05, "ex_per_s": 315.6340229613473, "data_frac": 0.018398769086724717, "mem_mb": 107.833344, "nll": -5.890435218811035, "elapsed_s": 774.6}199{"event": "train", "step": 3840, "loss": -5.857598781585693, "grad_norm": 0.48008856177330017, "lr": 3.117935064085835e-05, "ex_per_s": 316.6060038913785, "data_frac": 0.01837536573520292, "mem_mb": 107.833344, "nll": -5.857598781585693, "elapsed_s": 776.3}200{"event": "train", "step": 3860, "loss": -5.984775543212891, "grad_norm": 0.18917927145957947, "lr": 3.090324904225885e-05, "ex_per_s": 317.57434659654996, "data_frac": 0.018350320319417985, "mem_mb": 107.833344, "nll": -5.984775543212891, "elapsed_s": 777.9}201{"event": "train", "step": 3880, "loss": -5.863846302032471, "grad_norm": 0.5789470672607422, "lr": 3.0663808156423456e-05, "ex_per_s": 318.53852471587106, "data_frac": 0.018325197408363154, "mem_mb": 107.833344, "nll": -5.863846302032471, "elapsed_s": 779.6}202{"event": "train", "step": 3900, "loss": -6.149547100067139, "grad_norm": 1.6384341716766357, "lr": 3.046109344409957e-05, "ex_per_s": 319.49785131375285, "data_frac": 0.018300646980031823, "mem_mb": 107.833344, "nll": -6.149547100067139, "elapsed_s": 781.3}203{"event": "train", "step": 3920, "loss": -6.004547119140625, "grad_norm": 0.5034112930297852, "lr": 3.0295160325465435e-05, "ex_per_s": 320.4536663607659, "data_frac": 0.018276558174211813, "mem_mb": 107.833344, "nll": -6.004547119140625, "elapsed_s": 782.9}204{"event": "train", "step": 3940, "loss": -5.894385814666748, "grad_norm": 0.26846441626548767, "lr": 3.016605416497886e-05, "ex_per_s": 321.40573406834056, "data_frac": 0.01825198307034082, "mem_mb": 107.833344, "nll": -5.894385814666748, "elapsed_s": 784.6}205{"event": "train", "step": 3960, "loss": -6.031185150146484, "grad_norm": 0.33165785670280457, "lr": 3.0073810258974985e-05, "ex_per_s": 322.35343978673035, "data_frac": 0.018227795141352858, "mem_mb": 107.833344, "nll": -6.031185150146484, "elapsed_s": 786.3}206{"event": "train", "step": 3980, "loss": -5.774660110473633, "grad_norm": 0.12990102171897888, "lr": 3.0018453826016683e-05, "ex_per_s": 323.29777705445696, "data_frac": 0.018203399283394078, "mem_mb": 107.833344, "nll": -5.774660110473633, "elapsed_s": 787.9}207{"event": "train", "step": 4000, "loss": -6.022433757781982, "grad_norm": 1.168764591217041, "lr": 2.9999999999999997e-05, "ex_per_s": 324.23817604051044, "data_frac": 0.01817925693533897, "mem_mb": 107.833344, "nll": -6.022433757781982, "elapsed_s": 789.6}208{"event": "val", "step": 4000, "val_nll": -6.399150441926187, "best": -6.3999923274301365, "bad": 1, "mem_mb": 107.833344, "elapsed_s": 852.7}209{"event": "done", "steps": 4000, "best_val": -6.3999923274301365, "wall_s": 852.7440068721771, "ex_per_s": 300.20732826841595, "peak_mem_mb": 107.833344, "n_params": 3455619, "device": "mps", "data_time_frac": 0.016831904932074735, "elapsed_s": 852.7}210
Files · main
Menu

Theme