1{"event": "train", "step": 20, "loss": -6.125255107879639, "grad_norm": 0.3796319365501404, "lr": 3.149999999999999e-05, "ex_per_s": 102.25168583007203, "data_frac": 0.9424836720312698, "mem_mb": 86.535168, "nll": -6.125255107879639, "elapsed_s": 12.5}2{"event": "train", "step": 40, "loss": -5.896065711975098, "grad_norm": 0.6874927878379822, "lr": 6.149999999999999e-05, "ex_per_s": 195.0280065074371, "data_frac": 0.9018478824092862, "mem_mb": 86.535168, "nll": -5.896065711975098, "elapsed_s": 13.1}3{"event": "train", "step": 60, "loss": -5.9789228439331055, "grad_norm": 0.4916608929634094, "lr": 9.149999999999999e-05, "ex_per_s": 280.17736725780213, "data_frac": 0.8646189004737875, "mem_mb": 86.535168, "nll": -5.9789228439331055, "elapsed_s": 13.7}4{"event": "train", "step": 80, "loss": -6.066742420196533, "grad_norm": 1.588425636291504, "lr": 0.0001215, "ex_per_s": 358.7808455513969, "data_frac": 0.8311763298068959, "mem_mb": 86.535168, "nll": -6.066742420196533, "elapsed_s": 14.3}5{"event": "train", "step": 100, "loss": -5.947286128997803, "grad_norm": 0.5080704092979431, "lr": 0.0001515, "ex_per_s": 431.6043399126615, "data_frac": 0.8005418298635899, "mem_mb": 86.535168, "nll": -5.947286128997803, "elapsed_s": 14.8}6{"event": "train", "step": 120, "loss": -5.91217565536499, "grad_norm": 0.4620712101459503, "lr": 0.00018149999999999997, "ex_per_s": 499.1518997157032, "data_frac": 0.7722483748559595, "mem_mb": 86.535168, "nll": -5.91217565536499, "elapsed_s": 15.4}7{"event": "train", "step": 140, "loss": -5.8317718505859375, "grad_norm": 0.5333558320999146, "lr": 0.00021149999999999996, "ex_per_s": 562.1182777807793, "data_frac": 0.7461182076598931, "mem_mb": 86.535168, "nll": -5.8317718505859375, "elapsed_s": 16.0}8{"event": "train", "step": 160, "loss": -5.7350592613220215, "grad_norm": 0.9302918314933777, "lr": 0.0002415, "ex_per_s": 620.9708941568955, "data_frac": 0.7218591342174253, "mem_mb": 86.535168, "nll": -5.7350592613220215, "elapsed_s": 16.5}9{"event": "train", "step": 180, "loss": -5.9726176261901855, "grad_norm": 1.2345480918884277, "lr": 0.0002715, "ex_per_s": 675.984533090299, "data_frac": 0.699121590801557, "mem_mb": 86.535168, "nll": -5.9726176261901855, "elapsed_s": 17.1}10{"event": "train", "step": 200, "loss": -5.972184181213379, "grad_norm": 0.4575006067752838, "lr": 0.0003, "ex_per_s": 727.4299145495207, "data_frac": 0.6777038401505592, "mem_mb": 86.535168, "nll": -5.972184181213379, "elapsed_s": 17.6}11{"event": "train", "step": 220, "loss": -6.032626628875732, "grad_norm": 0.9814556837081909, "lr": 0.00029998154617398326, "ex_per_s": 775.9147184645624, "data_frac": 0.6577372187010585, "mem_mb": 86.535168, "nll": -6.032626628875732, "elapsed_s": 18.2}12{"event": "train", "step": 240, "loss": -5.914045810699463, "grad_norm": 0.16603775322437286, "lr": 0.000299926189741025, "ex_per_s": 821.1967195576616, "data_frac": 0.6393217527547201, "mem_mb": 86.535168, "nll": -5.914045810699463, "elapsed_s": 18.7}13{"event": "train", "step": 260, "loss": -5.887772560119629, "grad_norm": 0.13920240104198456, "lr": 0.0002998339458350211, "ex_per_s": 864.2749610484323, "data_frac": 0.6216616101875178, "mem_mb": 86.535168, "nll": -5.887772560119629, "elapsed_s": 19.3}14{"event": "train", "step": 280, "loss": -5.944588661193848, "grad_norm": 0.15885789692401886, "lr": 0.0002997048396745345, "ex_per_s": 904.996892133025, "data_frac": 0.6049911610343955, "mem_mb": 86.535168, "nll": -5.944588661193848, "elapsed_s": 19.8}15{"event": "train", "step": 300, "loss": -5.967859745025635, "grad_norm": 0.14105236530303955, "lr": 0.0002995389065559004, "ex_per_s": 943.4396725417434, "data_frac": 0.5891662936518692, "mem_mb": 86.535168, "nll": -5.967859745025635, "elapsed_s": 20.4}16{"event": "train", "step": 320, "loss": -5.965723991394043, "grad_norm": 0.7944684028625488, "lr": 0.0002993361918435765, "ex_per_s": 979.8565390128973, "data_frac": 0.5741746268449176, "mem_mb": 86.535168, "nll": -5.965723991394043, "elapsed_s": 20.9}17{"event": "train", "step": 340, "loss": -5.833776950836182, "grad_norm": 0.4443264901638031, "lr": 0.00029909675095774114, "ex_per_s": 1014.4416244038487, "data_frac": 0.5599776855877528, "mem_mb": 86.535168, "nll": -5.833776950836182, "elapsed_s": 21.5}18{"event": "train", "step": 360, "loss": -5.99638032913208, "grad_norm": 0.2409820407629013, "lr": 0.00029882064935914164, "ex_per_s": 1047.256588514622, "data_frac": 0.5464635537843977, "mem_mb": 86.535168, "nll": -5.99638032913208, "elapsed_s": 22.0}19{"event": "train", "step": 380, "loss": -6.007456302642822, "grad_norm": 1.1606980562210083, "lr": 0.00029850796253119824, "ex_per_s": 1078.5548010036239, "data_frac": 0.5336468784719737, "mem_mb": 86.535168, "nll": -6.007456302642822, "elapsed_s": 22.6}20{"event": "train", "step": 400, "loss": -5.9950361251831055, "grad_norm": 1.4828273057937622, "lr": 0.0002981587759593675, "ex_per_s": 1108.265742628338, "data_frac": 0.5213997809977967, "mem_mb": 86.535168, "nll": -5.9950361251831055, "elapsed_s": 23.1}21{"event": "train", "step": 420, "loss": -5.965256690979004, "grad_norm": 0.5013834834098816, "lr": 0.0002977731851077716, "ex_per_s": 1136.6166307064866, "data_frac": 0.5097333635737786, "mem_mb": 86.535168, "nll": -5.965256690979004, "elapsed_s": 23.7}22{"event": "train", "step": 440, "loss": -5.940114974975586, "grad_norm": 0.5711237192153931, "lr": 0.0002973512953930992, "ex_per_s": 1163.6742172116128, "data_frac": 0.4985943141920172, "mem_mb": 86.535168, "nll": -5.940114974975586, "elapsed_s": 24.2}23{"event": "train", "step": 460, "loss": -6.101609706878662, "grad_norm": 0.7427289485931396, "lr": 0.0002968932221557859, "ex_per_s": 1189.58367922261, "data_frac": 0.48797642487668336, "mem_mb": 86.535168, "nll": -6.101609706878662, "elapsed_s": 24.8}24{"event": "train", "step": 480, "loss": -5.876972198486328, "grad_norm": 0.6469749212265015, "lr": 0.0002963990906284808, "ex_per_s": 1214.3622961810254, "data_frac": 0.47781100034009, "mem_mb": 86.535168, "nll": -5.876972198486328, "elapsed_s": 25.3}25{"event": "train", "step": 500, "loss": -5.923430919647217, "grad_norm": 0.7032956480979919, "lr": 0.00029586903590180956, "ex_per_s": 1237.958630876482, "data_frac": 0.4680179521613217, "mem_mb": 86.535168, "nll": -5.923430919647217, "elapsed_s": 25.9}26{"event": "val", "step": 500, "val_nll": -6.440096941311381, "best": -6.440096941311381, "bad": 0, "mem_mb": 86.535168, "elapsed_s": 88.4}27{"event": "train", "step": 520, "loss": -6.0408220291137695, "grad_norm": 0.1931159347295761, "lr": 0.00029530320288744205, "ex_per_s": 374.1415110609644, "data_frac": 0.13614354650813612, "mem_mb": 86.535168, "nll": -6.0408220291137695, "elapsed_s": 89.0}28{"event": "train", "step": 540, "loss": -5.880037784576416, "grad_norm": 0.5140377283096313, "lr": 0.0002947017462784745, "ex_per_s": 386.135647127649, "data_frac": 0.13542576083836222, "mem_mb": 86.535168, "nll": -5.880037784576416, "elapsed_s": 89.5}29{"event": "train", "step": 560, "loss": -5.972921848297119, "grad_norm": 0.13513579964637756, "lr": 0.00029406483050713824, "ex_per_s": 397.9885682251376, "data_frac": 0.13471854513344397, "mem_mb": 86.535168, "nll": -5.972921848297119, "elapsed_s": 90.1}30{"event": "train", "step": 580, "loss": -5.925281047821045, "grad_norm": 0.18868859112262726, "lr": 0.0002933926296998457, "ex_per_s": 409.69471478445564, "data_frac": 0.13401887277654514, "mem_mb": 86.535168, "nll": -5.925281047821045, "elapsed_s": 90.6}31{"event": "train", "step": 600, "loss": -5.972788333892822, "grad_norm": 0.7661541700363159, "lr": 0.0002926853276295856, "ex_per_s": 421.25284291764723, "data_frac": 0.13332661346291783, "mem_mb": 86.535168, "nll": -5.972788333892822, "elapsed_s": 91.2}32{"event": "train", "step": 620, "loss": -5.952682971954346, "grad_norm": 0.50300532579422, "lr": 0.00029194311766568174, "ex_per_s": 432.6778654774453, "data_frac": 0.1326455077835345, "mem_mb": 86.535168, "nll": -5.952682971954346, "elapsed_s": 91.7}33{"event": "train", "step": 640, "loss": -6.066080570220947, "grad_norm": 0.3846137523651123, "lr": 0.0002911662027209272, "ex_per_s": 443.9681133264166, "data_frac": 0.13197251642792082, "mem_mb": 86.535168, "nll": -6.066080570220947, "elapsed_s": 92.3}34{"event": "train", "step": 660, "loss": -6.009441375732422, "grad_norm": 0.992946982383728, "lr": 0.00029035479519611045, "ex_per_s": 455.1251021416592, "data_frac": 0.1313078738673729, "mem_mb": 86.535168, "nll": -6.009441375732422, "elapsed_s": 92.8}35{"event": "train", "step": 680, "loss": -6.008297920227051, "grad_norm": 0.20602625608444214, "lr": 0.0002895091169219469, "ex_per_s": 466.15045582717624, "data_frac": 0.13064784633772278, "mem_mb": 86.535168, "nll": -6.008297920227051, "elapsed_s": 93.4}36{"event": "train", "step": 700, "loss": -5.9879841804504395, "grad_norm": 0.6699579954147339, "lr": 0.00028862939909843273, "ex_per_s": 477.0509887708429, "data_frac": 0.12999730273809024, "mem_mb": 86.535168, "nll": -5.9879841804504395, "elapsed_s": 93.9}37{"event": "train", "step": 720, "loss": -6.061961650848389, "grad_norm": 0.4948371946811676, "lr": 0.0002877158822316369, "ex_per_s": 487.8238463423371, "data_frac": 0.12935554891085327, "mem_mb": 86.535168, "nll": -6.061961650848389, "elapsed_s": 94.5}38{"event": "train", "step": 740, "loss": -6.03506326675415, "grad_norm": 0.17898045480251312, "lr": 0.00028676881606794945, "ex_per_s": 498.4685825421074, "data_frac": 0.12872263619859453, "mem_mb": 86.535168, "nll": -6.03506326675415, "elapsed_s": 95.0}39{"event": "train", "step": 760, "loss": -5.955129146575928, "grad_norm": 1.276550054550171, "lr": 0.000285788459525803, "ex_per_s": 508.9934086420363, "data_frac": 0.12809478862851278, "mem_mb": 86.535168, "nll": -5.955129146575928, "elapsed_s": 95.6}40{"event": "train", "step": 780, "loss": -5.8963165283203125, "grad_norm": 0.251977801322937, "lr": 0.00028477508062488754, "ex_per_s": 519.3964664654598, "data_frac": 0.1274739916033611, "mem_mb": 86.535168, "nll": -5.8963165283203125, "elapsed_s": 96.1}41{"event": "train", "step": 800, "loss": -5.81559419631958, "grad_norm": 0.24524758756160736, "lr": 0.000283728956412876, "ex_per_s": 529.6802666722061, "data_frac": 0.12686020624739977, "mem_mb": 86.535168, "nll": -5.81559419631958, "elapsed_s": 96.7}42{"event": "train", "step": 820, "loss": -5.934140682220459, "grad_norm": 0.4769822061061859, "lr": 0.00028265037288968236, "ex_per_s": 539.8532575854763, "data_frac": 0.1262557484915123, "mem_mb": 86.535168, "nll": -5.934140682220459, "elapsed_s": 97.2}43{"event": "train", "step": 840, "loss": -5.807830333709717, "grad_norm": 0.8032633066177368, "lr": 0.0002815396249292725, "ex_per_s": 549.9091948539135, "data_frac": 0.1256532476518025, "mem_mb": 86.535168, "nll": -5.807830333709717, "elapsed_s": 97.8}44{"event": "train", "step": 860, "loss": -5.918150424957275, "grad_norm": 1.3579069375991821, "lr": 0.00028039701619904774, "ex_per_s": 559.8532620777963, "data_frac": 0.12506067778205698, "mem_mb": 86.535168, "nll": -5.918150424957275, "elapsed_s": 98.3}45{"event": "train", "step": 880, "loss": -5.87001895904541, "grad_norm": 0.5429680943489075, "lr": 0.00027922285907682625, "ex_per_s": 569.684422716792, "data_frac": 0.12447495273560828, "mem_mb": 86.535168, "nll": -5.87001895904541, "elapsed_s": 98.9}46{"event": "train", "step": 900, "loss": -5.98760461807251, "grad_norm": 0.46259579062461853, "lr": 0.00027801747456544134, "ex_per_s": 579.4120697849037, "data_frac": 0.12389706862309081, "mem_mb": 86.535168, "nll": -5.98760461807251, "elapsed_s": 99.4}47{"event": "train", "step": 920, "loss": -6.027410984039307, "grad_norm": 0.47599130868911743, "lr": 0.00027678119220498265, "ex_per_s": 589.0325091368621, "data_frac": 0.12332503304271845, "mem_mb": 86.535168, "nll": -6.027410984039307, "elapsed_s": 100.0}48{"event": "train", "step": 940, "loss": -6.081259727478027, "grad_norm": 1.0870716571807861, "lr": 0.0002755143499827035, "ex_per_s": 598.5465208966897, "data_frac": 0.1227577846361597, "mem_mb": 86.535168, "nll": -6.081259727478027, "elapsed_s": 100.5}49{"event": "train", "step": 960, "loss": -5.967348098754883, "grad_norm": 0.4104170799255371, "lr": 0.00027421729424061787, "ex_per_s": 607.9579435791717, "data_frac": 0.12219730404010333, "mem_mb": 86.535168, "nll": -5.967348098754883, "elapsed_s": 101.1}50{"event": "train", "step": 980, "loss": -5.926362991333008, "grad_norm": 0.39338523149490356, "lr": 0.00027289037958081483, "ex_per_s": 617.2605487916545, "data_frac": 0.12164244807082716, "mem_mb": 86.535168, "nll": -5.926362991333008, "elapsed_s": 101.6}51{"event": "train", "step": 1000, "loss": -6.072170734405518, "grad_norm": 0.2915230393409729, "lr": 0.0002715339687685131, "ex_per_s": 626.4716823086571, "data_frac": 0.12109556274316373, "mem_mb": 86.535168, "nll": -6.072170734405518, "elapsed_s": 102.2}52{"event": "val", "step": 1000, "val_nll": -6.446803593708008, "best": -6.446803593708008, "bad": 0, "mem_mb": 86.535168, "elapsed_s": 164.1}53{"event": "train", "step": 1020, "loss": -6.030152320861816, "grad_norm": 0.2121785283088684, "lr": 0.0002701484326328852, "ex_per_s": 396.4768972140215, "data_frac": 0.07520271404782929, "mem_mb": 86.535168, "nll": -6.030152320861816, "elapsed_s": 164.7}54{"event": "train", "step": 1040, "loss": -5.9733452796936035, "grad_norm": 0.14980149269104004, "lr": 0.00026873414996567596, "ex_per_s": 402.9041762558615, "data_frac": 0.075018527599603, "mem_mb": 86.535168, "nll": -5.9733452796936035, "elapsed_s": 165.2}55{"event": "train", "step": 1060, "loss": -6.183572292327881, "grad_norm": 0.1940319538116455, "lr": 0.0002672915074176452, "ex_per_s": 409.28462449939127, "data_frac": 0.07483481316864268, "mem_mb": 86.535168, "nll": -6.183572292327881, "elapsed_s": 165.8}56{"event": "train", "step": 1080, "loss": -5.816099166870117, "grad_norm": 0.5615946650505066, "lr": 0.00026582089939286075, "ex_per_s": 415.62516103559176, "data_frac": 0.07465253023951705, "mem_mb": 86.535168, "nll": -5.816099166870117, "elapsed_s": 166.3}57{"event": "train", "step": 1100, "loss": -6.054073333740234, "grad_norm": 0.10386236011981964, "lr": 0.0002643227279408728, "ex_per_s": 421.91939842206415, "data_frac": 0.07447216487380989, "mem_mb": 86.535168, "nll": -6.054073333740234, "elapsed_s": 166.9}58{"event": "train", "step": 1120, "loss": -6.053536415100098, "grad_norm": 0.24231714010238647, "lr": 0.000262797402646797, "ex_per_s": 428.17851658654365, "data_frac": 0.07429283671740329, "mem_mb": 86.535168, "nll": -6.053536415100098, "elapsed_s": 167.4}59{"event": "train", "step": 1140, "loss": -6.019773960113525, "grad_norm": 0.17143279314041138, "lr": 0.0002612453405193386, "ex_per_s": 434.3965760450516, "data_frac": 0.07411466058312048, "mem_mb": 86.535168, "nll": -6.019773960113525, "elapsed_s": 168.0}60{"event": "train", "step": 1160, "loss": -5.971249103546143, "grad_norm": 0.44152316451072693, "lr": 0.0002596669658767859, "ex_per_s": 440.5738120032218, "data_frac": 0.07393659860725534, "mem_mb": 86.535168, "nll": -5.971249103546143, "elapsed_s": 168.5}61{"event": "train", "step": 1180, "loss": -5.957459926605225, "grad_norm": 0.145237997174263, "lr": 0.00025806271023100637, "ex_per_s": 446.6973743851224, "data_frac": 0.07376301152339468, "mem_mb": 86.535168, "nll": -5.957459926605225, "elapsed_s": 169.1}62{"event": "train", "step": 1200, "loss": -5.947926044464111, "grad_norm": 0.20981232821941376, "lr": 0.00025643301216947504, "ex_per_s": 452.7809062008137, "data_frac": 0.07358964209879496, "mem_mb": 86.535168, "nll": -5.947926044464111, "elapsed_s": 169.6}63{"event": "train", "step": 1220, "loss": -6.033280849456787, "grad_norm": 0.6388020515441895, "lr": 0.0002547783172353695, "ex_per_s": 458.8371863072323, "data_frac": 0.07341593090217856, "mem_mb": 86.535168, "nll": -6.033280849456787, "elapsed_s": 170.2}64{"event": "train", "step": 1240, "loss": -6.05922794342041, "grad_norm": 0.06508813053369522, "lr": 0.00025309907780576277, "ex_per_s": 464.8534584582782, "data_frac": 0.07324486916650153, "mem_mb": 86.535168, "nll": -6.05922794342041, "elapsed_s": 170.7}65{"event": "train", "step": 1260, "loss": -6.053707122802734, "grad_norm": 0.20556485652923584, "lr": 0.0002513957529679477, "ex_per_s": 470.8344919179028, "data_frac": 0.07307438416892185, "mem_mb": 86.535168, "nll": -6.053707122802734, "elapsed_s": 171.3}66{"event": "train", "step": 1280, "loss": -6.047028064727783, "grad_norm": 0.7456238269805908, "lr": 0.00024966880839392746, "ex_per_s": 476.7769195231425, "data_frac": 0.07290356060741521, "mem_mb": 86.535168, "nll": -6.047028064727783, "elapsed_s": 171.8}67{"event": "train", "step": 1300, "loss": -5.865945339202881, "grad_norm": 0.9461830854415894, "lr": 0.0002479187162131052, "ex_per_s": 482.6773280982753, "data_frac": 0.07273454072584482, "mem_mb": 86.535168, "nll": -5.865945339202881, "elapsed_s": 172.4}68{"event": "train", "step": 1320, "loss": -5.828519821166992, "grad_norm": 0.2013840228319168, "lr": 0.00024614595488320833, "ex_per_s": 488.5408971320578, "data_frac": 0.07256641788801213, "mem_mb": 86.535168, "nll": -5.828519821166992, "elapsed_s": 172.9}69{"event": "train", "step": 1340, "loss": -5.9520673751831055, "grad_norm": 0.2511095404624939, "lr": 0.00024435100905948387, "ex_per_s": 494.367704668644, "data_frac": 0.07239789336048169, "mem_mb": 86.535168, "nll": -5.9520673751831055, "elapsed_s": 173.5}70{"event": "train", "step": 1360, "loss": -6.037209987640381, "grad_norm": 0.6684756278991699, "lr": 0.0002425343694621974, "ex_per_s": 500.1574775567907, "data_frac": 0.07223210168916348, "mem_mb": 86.535168, "nll": -6.037209987640381, "elapsed_s": 174.0}71{"event": "train", "step": 1380, "loss": -6.0689826011657715, "grad_norm": 0.26380905508995056, "lr": 0.00024069653274247583, "ex_per_s": 505.91370900454, "data_frac": 0.07206754117354817, "mem_mb": 86.535168, "nll": -6.0689826011657715, "elapsed_s": 174.6}72{"event": "train", "step": 1400, "loss": -5.796529769897461, "grad_norm": 0.12343746423721313, "lr": 0.0002388380013465276, "ex_per_s": 511.6307575526417, "data_frac": 0.07190261727790098, "mem_mb": 86.535168, "nll": -5.796529769897461, "elapsed_s": 175.1}73{"event": "train", "step": 1420, "loss": -5.989380836486816, "grad_norm": 0.7883347868919373, "lr": 0.00023695928337827914, "ex_per_s": 517.3144750765632, "data_frac": 0.07173978136896449, "mem_mb": 86.535168, "nll": -5.989380836486816, "elapsed_s": 175.7}74{"event": "train", "step": 1440, "loss": -5.744483470916748, "grad_norm": 0.09928842633962631, "lr": 0.00023506089246046403, "ex_per_s": 522.9634047973352, "data_frac": 0.07157791375937003, "mem_mb": 86.535168, "nll": -5.744483470916748, "elapsed_s": 176.2}75{"event": "train", "step": 1460, "loss": -5.801549911499023, "grad_norm": 0.26376378536224365, "lr": 0.00023314334759420388, "ex_per_s": 528.577012272664, "data_frac": 0.07141744532331183, "mem_mb": 86.535168, "nll": -5.801549911499023, "elapsed_s": 176.8}76{"event": "train", "step": 1480, "loss": -6.0453362464904785, "grad_norm": 0.14677578210830688, "lr": 0.0002312071730171186, "ex_per_s": 534.1561226966672, "data_frac": 0.07125758043275, "mem_mb": 86.535168, "nll": -6.0453362464904785, "elapsed_s": 177.3}77{"event": "train", "step": 1500, "loss": -5.958207130432129, "grad_norm": 0.4842979609966278, "lr": 0.0002292528980600049, "ex_per_s": 539.7029919365231, "data_frac": 0.07109913653552191, "mem_mb": 86.535168, "nll": -5.958207130432129, "elapsed_s": 177.9}78{"event": "val", "step": 1500, "val_nll": -6.447797691832402, "best": -6.447797691832402, "bad": 0, "mem_mb": 86.535168, "elapsed_s": 239.7}79{"event": "train", "step": 1520, "loss": -5.951138019561768, "grad_norm": 0.5999360680580139, "lr": 0.00022728105700212285, "ex_per_s": 404.9449219650629, "data_frac": 0.05269006788798363, "mem_mb": 86.535168, "nll": -5.951138019561768, "elapsed_s": 240.2}80{"event": "train", "step": 1540, "loss": -5.879615783691406, "grad_norm": 0.12561018764972687, "lr": 0.00022529218892512886, "ex_per_s": 409.33572865200915, "data_frac": 0.05261377968762112, "mem_mb": 86.535168, "nll": -5.879615783691406, "elapsed_s": 240.8}81{"event": "train", "step": 1560, "loss": -5.981464385986328, "grad_norm": 0.4761824905872345, "lr": 0.00022328683756569652, "ex_per_s": 413.7046252237354, "data_frac": 0.05253797555077122, "mem_mb": 86.535168, "nll": -5.981464385986328, "elapsed_s": 241.3}82{"event": "train", "step": 1580, "loss": -5.981413841247559, "grad_norm": 0.09112240374088287, "lr": 0.0002212655511668642, "ex_per_s": 418.0548461755958, "data_frac": 0.05246349137134388, "mem_mb": 86.535168, "nll": -5.981413841247559, "elapsed_s": 241.9}83{"event": "train", "step": 1600, "loss": -5.95588493347168, "grad_norm": 0.28630590438842773, "lr": 0.00021922888232815089, "ex_per_s": 422.3851470352824, "data_frac": 0.05238875985719339, "mem_mb": 86.535168, "nll": -5.95588493347168, "elapsed_s": 242.4}84{"event": "train", "step": 1620, "loss": -5.882150173187256, "grad_norm": 0.11628701537847519, "lr": 0.00021717738785448102, "ex_per_s": 426.6946668121606, "data_frac": 0.05231366613040553, "mem_mb": 86.535168, "nll": -5.882150173187256, "elapsed_s": 243.0}85{"event": "train", "step": 1640, "loss": -6.077833652496338, "grad_norm": 0.534850537776947, "lr": 0.00021511162860395977, "ex_per_s": 430.985420940174, "data_frac": 0.052238848994627135, "mem_mb": 86.535168, "nll": -6.077833652496338, "elapsed_s": 243.5}86{"event": "train", "step": 1660, "loss": -6.038712501525879, "grad_norm": 0.7111098766326904, "lr": 0.0002130321693345397, "ex_per_s": 435.2543381722473, "data_frac": 0.052165029840024635, "mem_mb": 86.535168, "nll": -6.038712501525879, "elapsed_s": 244.1}87{"event": "train", "step": 1680, "loss": -5.866774082183838, "grad_norm": 0.1457652449607849, "lr": 0.0002109395785496218, "ex_per_s": 439.5065074692959, "data_frac": 0.05209215807713872, "mem_mb": 86.535168, "nll": -5.866774082183838, "elapsed_s": 244.6}88{"event": "train", "step": 1700, "loss": -5.9275617599487305, "grad_norm": 0.5554318428039551, "lr": 0.00020883442834263225, "ex_per_s": 443.74045965029177, "data_frac": 0.05201828950551986, "mem_mb": 86.535168, "nll": -5.9275617599487305, "elapsed_s": 245.2}89{"event": "train", "step": 1720, "loss": -5.887142658233643, "grad_norm": 0.9149301052093506, "lr": 0.00020671729424061788, "ex_per_s": 447.9549530439897, "data_frac": 0.05194535133533239, "mem_mb": 86.535168, "nll": -5.887142658233643, "elapsed_s": 245.8}90{"event": "train", "step": 1740, "loss": -6.273637771606445, "grad_norm": 0.45222505927085876, "lr": 0.00020458875504690275, "ex_per_s": 452.14995236327746, "data_frac": 0.051872918405756814, "mem_mb": 86.535168, "nll": -6.273637771606445, "elapsed_s": 246.3}91{"event": "train", "step": 1760, "loss": -5.9288811683654785, "grad_norm": 0.34395721554756165, "lr": 0.000202449392682849, "ex_per_s": 456.3267954101742, "data_frac": 0.05180073555668092, "mem_mb": 86.535168, "nll": -5.9288811683654785, "elapsed_s": 246.9}92{"event": "train", "step": 1780, "loss": -5.870924949645996, "grad_norm": 0.4353761076927185, "lr": 0.00020029979202876563, "ex_per_s": 460.48530976022255, "data_frac": 0.051729366504874905, "mem_mb": 86.535168, "nll": -5.870924949645996, "elapsed_s": 247.4}93{"event": "train", "step": 1800, "loss": -6.046781063079834, "grad_norm": 1.2656985521316528, "lr": 0.00019814054076400787, "ex_per_s": 464.62651557130835, "data_frac": 0.05165897880040057, "mem_mb": 86.535168, "nll": -6.046781063079834, "elapsed_s": 248.0}94{"event": "train", "step": 1820, "loss": -5.93690824508667, "grad_norm": 0.7726545333862305, "lr": 0.00019597222920631147, "ex_per_s": 468.7489809415864, "data_frac": 0.05158800533573372, "mem_mb": 86.535168, "nll": -5.93690824508667, "elapsed_s": 248.5}95{"event": "train", "step": 1840, "loss": -6.074361324310303, "grad_norm": 0.3398592174053192, "lr": 0.0001937954501504058, "ex_per_s": 472.8535942238926, "data_frac": 0.051517240177543826, "mem_mb": 86.535168, "nll": -6.074361324310303, "elapsed_s": 249.1}96{"event": "train", "step": 1860, "loss": -6.050648212432861, "grad_norm": 0.11062514781951904, "lr": 0.0001916107987059498, "ex_per_s": 476.9278459360416, "data_frac": 0.05144794646142646, "mem_mb": 86.535168, "nll": -6.050648212432861, "elapsed_s": 249.6}97{"event": "train", "step": 1880, "loss": -5.948655128479004, "grad_norm": 0.07030808925628662, "lr": 0.00018941887213483484, "ex_per_s": 480.98852877525866, "data_frac": 0.05138049393324369, "mem_mb": 86.535168, "nll": -5.948655128479004, "elapsed_s": 250.2}98{"event": "train", "step": 1900, "loss": -5.983978748321533, "grad_norm": 0.11149947345256805, "lr": 0.00018722026968789906, "ex_per_s": 485.038301742791, "data_frac": 0.05131160367258751, "mem_mb": 86.535168, "nll": -5.983978748321533, "elapsed_s": 250.7}99{"event": "train", "step": 1920, "loss": -5.965317249298096, "grad_norm": 0.36936843395233154, "lr": 0.00018501559244109864, "ex_per_s": 489.06891574947184, "data_frac": 0.05124252794697613, "mem_mb": 86.535168, "nll": -5.965317249298096, "elapsed_s": 251.3}100{"event": "train", "step": 1940, "loss": -6.051220417022705, "grad_norm": 0.2596738636493683, "lr": 0.00018280544313117932, "ex_per_s": 493.0835303615984, "data_frac": 0.05117300948045501, "mem_mb": 86.535168, "nll": -6.051220417022705, "elapsed_s": 251.8}101{"event": "train", "step": 1960, "loss": -5.939828395843506, "grad_norm": 0.13565145432949066, "lr": 0.00018059042599089415, "ex_per_s": 497.07956875823515, "data_frac": 0.05110423320402504, "mem_mb": 86.535168, "nll": -5.939828395843506, "elapsed_s": 252.4}102{"event": "train", "step": 1980, "loss": -5.964278697967529, "grad_norm": 0.9947322607040405, "lr": 0.00017837114658381247, "ex_per_s": 501.05949829590793, "data_frac": 0.05103512213045098, "mem_mb": 86.535168, "nll": -5.964278697967529, "elapsed_s": 252.9}103{"event": "train", "step": 2000, "loss": -5.914780616760254, "grad_norm": 0.3267081677913666, "lr": 0.00017614821163876486, "ex_per_s": 505.01080403060865, "data_frac": 0.05096958236399877, "mem_mb": 86.535168, "nll": -5.914780616760254, "elapsed_s": 253.5}104{"event": "val", "step": 2000, "val_nll": -6.449032247923463, "best": -6.449032247923463, "bad": 0, "mem_mb": 86.535168, "elapsed_s": 315.2}105{"event": "train", "step": 2020, "loss": -5.92490816116333, "grad_norm": 0.5178285241127014, "lr": 0.00017392222888396941, "ex_per_s": 409.3810518148253, "data_frac": 0.04094404755876861, "mem_mb": 86.535168, "nll": -5.92490816116333, "elapsed_s": 315.8}106{"event": "train", "step": 2040, "loss": -5.873110771179199, "grad_norm": 0.17219310998916626, "lr": 0.000171693806880885, "ex_per_s": 412.71354583799365, "data_frac": 0.04090693088300497, "mem_mb": 86.535168, "nll": -5.873110771179199, "elapsed_s": 316.4}107{"event": "train", "step": 2060, "loss": -6.1432671546936035, "grad_norm": 0.20360790193080902, "lr": 0.0001694635548578366, "ex_per_s": 416.0316664734773, "data_frac": 0.040870233664994884, "mem_mb": 86.535168, "nll": -6.1432671546936035, "elapsed_s": 316.9}108{"event": "train", "step": 2080, "loss": -5.9612836837768555, "grad_norm": 0.6130344271659851, "lr": 0.0001672320825434584, "ex_per_s": 419.3383276424029, "data_frac": 0.04083337588333576, "mem_mb": 86.535168, "nll": -5.9612836837768555, "elapsed_s": 317.5}109{"event": "train", "step": 2100, "loss": -5.902692794799805, "grad_norm": 0.7701839804649353, "lr": 0.000165, "ex_per_s": 422.63688820984555, "data_frac": 0.04079623419362833, "mem_mb": 86.535168, "nll": -5.902692794799805, "elapsed_s": 318.0}110{"event": "train", "step": 2120, "loss": -6.151300430297852, "grad_norm": 0.5778326988220215, "lr": 0.0001627679174565416, "ex_per_s": 425.9250229720973, "data_frac": 0.040759714885903235, "mem_mb": 86.535168, "nll": -6.151300430297852, "elapsed_s": 318.6}111{"event": "train", "step": 2140, "loss": -5.978779315948486, "grad_norm": 0.27007821202278137, "lr": 0.00016053644514216336, "ex_per_s": 429.2004304811345, "data_frac": 0.04072342596504845, "mem_mb": 86.535168, "nll": -5.978779315948486, "elapsed_s": 319.1}112{"event": "train", "step": 2160, "loss": -5.924803256988525, "grad_norm": 0.9890788197517395, "lr": 0.00015830619311911498, "ex_per_s": 432.4631074554742, "data_frac": 0.040687163180593296, "mem_mb": 86.535168, "nll": -5.924803256988525, "elapsed_s": 319.7}113{"event": "train", "step": 2180, "loss": -6.0841522216796875, "grad_norm": 0.10617180168628693, "lr": 0.00015607777111603058, "ex_per_s": 435.7159364488431, "data_frac": 0.04064992381517876, "mem_mb": 86.535168, "nll": -6.0841522216796875, "elapsed_s": 320.2}114{"event": "train", "step": 2200, "loss": -5.964648723602295, "grad_norm": 0.39339104294776917, "lr": 0.00015385178836123514, "ex_per_s": 438.9594888005124, "data_frac": 0.04061321911376946, "mem_mb": 86.535168, "nll": -5.964648723602295, "elapsed_s": 320.8}115{"event": "train", "step": 2220, "loss": -5.823565483093262, "grad_norm": 1.0467565059661865, "lr": 0.00015162885341618752, "ex_per_s": 442.1923577435025, "data_frac": 0.04057772623749856, "mem_mb": 86.535168, "nll": -5.823565483093262, "elapsed_s": 321.3}116{"event": "train", "step": 2240, "loss": -5.962638854980469, "grad_norm": 0.7181340456008911, "lr": 0.00014940957400910585, "ex_per_s": 445.41180662888144, "data_frac": 0.040541552820508704, "mem_mb": 86.535168, "nll": -5.962638854980469, "elapsed_s": 321.9}117{"event": "train", "step": 2260, "loss": -5.739428520202637, "grad_norm": 0.5915848016738892, "lr": 0.00014719455686882068, "ex_per_s": 448.62156704991753, "data_frac": 0.04050600183847366, "mem_mb": 86.535168, "nll": -5.739428520202637, "elapsed_s": 322.4}118{"event": "train", "step": 2280, "loss": -6.011369228363037, "grad_norm": 0.25450658798217773, "lr": 0.0001449844075589013, "ex_per_s": 451.81899545744517, "data_frac": 0.040470766972315636, "mem_mb": 86.535168, "nll": -6.011369228363037, "elapsed_s": 323.0}119{"event": "train", "step": 2300, "loss": -6.000100612640381, "grad_norm": 0.10042679309844971, "lr": 0.0001427797303121009, "ex_per_s": 455.00651123075477, "data_frac": 0.04043447005036335, "mem_mb": 86.535168, "nll": -6.000100612640381, "elapsed_s": 323.5}120{"event": "train", "step": 2320, "loss": -6.070340156555176, "grad_norm": 0.39716431498527527, "lr": 0.00014058112786516518, "ex_per_s": 458.18205887064704, "data_frac": 0.04039976860642143, "mem_mb": 86.535168, "nll": -6.070340156555176, "elapsed_s": 324.1}121{"event": "train", "step": 2340, "loss": -5.851423740386963, "grad_norm": 0.2872275114059448, "lr": 0.00013838920129405016, "ex_per_s": 461.3460109218306, "data_frac": 0.040364660637674604, "mem_mb": 86.535168, "nll": -5.851423740386963, "elapsed_s": 324.6}122{"event": "train", "step": 2360, "loss": -5.928483963012695, "grad_norm": 0.13307888805866241, "lr": 0.0001362045498495942, "ex_per_s": 464.5014829402268, "data_frac": 0.040330363265695744, "mem_mb": 86.535168, "nll": -5.928483963012695, "elapsed_s": 325.2}123{"event": "train", "step": 2380, "loss": -5.932003974914551, "grad_norm": 0.07020774483680725, "lr": 0.00013402777079368853, "ex_per_s": 467.64714952751444, "data_frac": 0.04029384622058506, "mem_mb": 86.535168, "nll": -5.932003974914551, "elapsed_s": 325.7}124{"event": "train", "step": 2400, "loss": -5.959148406982422, "grad_norm": 0.7208147048950195, "lr": 0.00013185945923599213, "ex_per_s": 470.78101894252, "data_frac": 0.04025863769813724, "mem_mb": 86.535168, "nll": -5.959148406982422, "elapsed_s": 326.3}125{"event": "train", "step": 2420, "loss": -5.996665000915527, "grad_norm": 0.12690407037734985, "lr": 0.00012970020797123432, "ex_per_s": 473.9045375458015, "data_frac": 0.040223943557954986, "mem_mb": 86.535168, "nll": -5.996665000915527, "elapsed_s": 326.8}126{"event": "train", "step": 2440, "loss": -6.099179744720459, "grad_norm": 0.43491557240486145, "lr": 0.000127550607317151, "ex_per_s": 477.01865628407865, "data_frac": 0.04018911770439641, "mem_mb": 86.535168, "nll": -6.099179744720459, "elapsed_s": 327.4}127{"event": "train", "step": 2460, "loss": -6.13151216506958, "grad_norm": 0.13079553842544556, "lr": 0.00012541124495309727, "ex_per_s": 480.1214225158201, "data_frac": 0.04015431140132893, "mem_mb": 86.535168, "nll": -6.13151216506958, "elapsed_s": 327.9}128{"event": "train", "step": 2480, "loss": -5.945183753967285, "grad_norm": 0.13411901891231537, "lr": 0.00012328270575938212, "ex_per_s": 483.21456696351754, "data_frac": 0.04011979509656233, "mem_mb": 86.535168, "nll": -5.945183753967285, "elapsed_s": 328.5}129{"event": "train", "step": 2500, "loss": -5.758757591247559, "grad_norm": 0.46311846375465393, "lr": 0.00012116557165736775, "ex_per_s": 486.2954333664823, "data_frac": 0.040085753154997204, "mem_mb": 86.535168, "nll": -5.758757591247559, "elapsed_s": 329.0}130{"event": "val", "step": 2500, "val_nll": -6.44968405509289, "best": -6.44968405509289, "bad": 0, "mem_mb": 86.535168, "elapsed_s": 390.9}131{"event": "train", "step": 2520, "loss": -5.963291645050049, "grad_norm": 0.4600125551223755, "lr": 0.00011906042145037817, "ex_per_s": 412.01977573060424, "data_frac": 0.033722044973586204, "mem_mb": 86.535168, "nll": -5.963291645050049, "elapsed_s": 391.4}132{"event": "train", "step": 2540, "loss": -6.030059814453125, "grad_norm": 0.33369743824005127, "lr": 0.00011696783066546026, "ex_per_s": 414.70251227264816, "data_frac": 0.03370307768556627, "mem_mb": 86.535168, "nll": -6.030059814453125, "elapsed_s": 392.0}133{"event": "train", "step": 2560, "loss": -5.94873571395874, "grad_norm": 0.18910224735736847, "lr": 0.00011488837139604017, "ex_per_s": 417.37030612453765, "data_frac": 0.033685408557762996, "mem_mb": 86.535168, "nll": -5.94873571395874, "elapsed_s": 392.6}134{"event": "train", "step": 2580, "loss": -5.960561275482178, "grad_norm": 0.10775568336248398, "lr": 0.00011282261214551898, "ex_per_s": 420.03144612444447, "data_frac": 0.03366756085096034, "mem_mb": 86.535168, "nll": -5.960561275482178, "elapsed_s": 393.1}135{"event": "train", "step": 2600, "loss": -5.8475775718688965, "grad_norm": 0.0966213196516037, "lr": 0.00011077111767184915, "ex_per_s": 422.69608389793234, "data_frac": 0.03364928825392546, "mem_mb": 86.535168, "nll": -5.8475775718688965, "elapsed_s": 393.7}136{"event": "train", "step": 2620, "loss": -5.88977575302124, "grad_norm": 0.8857305645942688, "lr": 0.00010873444883313583, "ex_per_s": 425.3535612231035, "data_frac": 0.03362944457504846, "mem_mb": 86.535168, "nll": -5.88977575302124, "elapsed_s": 394.2}137{"event": "train", "step": 2640, "loss": -5.987175941467285, "grad_norm": 0.09312697499990463, "lr": 0.00010671316243430345, "ex_per_s": 428.0035259436602, "data_frac": 0.033609422808944474, "mem_mb": 86.535168, "nll": -5.987175941467285, "elapsed_s": 394.8}138{"event": "train", "step": 2660, "loss": -5.820704460144043, "grad_norm": 0.05977461859583855, "lr": 0.00010470781107487112, "ex_per_s": 430.6455912729237, "data_frac": 0.03358972984047459, "mem_mb": 86.535168, "nll": -5.820704460144043, "elapsed_s": 395.3}139{"event": "train", "step": 2680, "loss": -5.958130836486816, "grad_norm": 0.5838608145713806, "lr": 0.00010271894299787715, "ex_per_s": 433.2809756348707, "data_frac": 0.03357017510691194, "mem_mb": 86.535168, "nll": -5.958130836486816, "elapsed_s": 395.9}140{"event": "train", "step": 2700, "loss": -6.131473064422607, "grad_norm": 0.597090482711792, "lr": 0.00010074710193999506, "ex_per_s": 435.90911542733926, "data_frac": 0.03355031935276131, "mem_mb": 86.535168, "nll": -6.131473064422607, "elapsed_s": 396.4}141{"event": "train", "step": 2720, "loss": -5.884420394897461, "grad_norm": 0.16195715963840485, "lr": 9.879282698288144e-05, "ex_per_s": 438.52907925283574, "data_frac": 0.0335303917179311, "mem_mb": 86.535168, "nll": -5.884420394897461, "elapsed_s": 397.0}142{"event": "train", "step": 2740, "loss": -6.1258864402771, "grad_norm": 0.09413597732782364, "lr": 9.685665240579613e-05, "ex_per_s": 441.14306732078643, "data_frac": 0.033511073005241775, "mem_mb": 86.535168, "nll": -6.1258864402771, "elapsed_s": 397.5}143{"event": "train", "step": 2760, "loss": -5.960720539093018, "grad_norm": 0.3622843027114868, "lr": 9.4939107539536e-05, "ex_per_s": 443.75004111631375, "data_frac": 0.033491216025069916, "mem_mb": 86.535168, "nll": -5.960720539093018, "elapsed_s": 398.1}144{"event": "train", "step": 2780, "loss": -6.011152744293213, "grad_norm": 1.0129261016845703, "lr": 9.304071662172087e-05, "ex_per_s": 446.3502042915668, "data_frac": 0.03347142974477652, "mem_mb": 86.535168, "nll": -6.011152744293213, "elapsed_s": 398.6}145{"event": "train", "step": 2800, "loss": -5.98100471496582, "grad_norm": 0.5051150918006897, "lr": 9.116199865347241e-05, "ex_per_s": 448.9430296606717, "data_frac": 0.033451858918817745, "mem_mb": 86.535168, "nll": -5.98100471496582, "elapsed_s": 399.2}146{"event": "train", "step": 2820, "loss": -6.009163856506348, "grad_norm": 0.08719000220298767, "lr": 8.930346725752419e-05, "ex_per_s": 451.5272964231853, "data_frac": 0.03343326922311434, "mem_mb": 86.535168, "nll": -6.009163856506348, "elapsed_s": 399.7}147{"event": "train", "step": 2840, "loss": -5.8768157958984375, "grad_norm": 0.2376178503036499, "lr": 8.746563053780262e-05, "ex_per_s": 454.1062113822387, "data_frac": 0.033414264036979224, "mem_mb": 86.535168, "nll": -5.8768157958984375, "elapsed_s": 400.3}148{"event": "train", "step": 2860, "loss": -6.050480365753174, "grad_norm": 0.13200536370277405, "lr": 8.564899094051614e-05, "ex_per_s": 456.6781130375221, "data_frac": 0.03339521149719946, "mem_mb": 86.535168, "nll": -6.050480365753174, "elapsed_s": 400.8}149{"event": "train", "step": 2880, "loss": -5.77029275894165, "grad_norm": 0.44065576791763306, "lr": 8.385404511679161e-05, "ex_per_s": 459.2432914928382, "data_frac": 0.03337562520434173, "mem_mb": 86.535168, "nll": -5.77029275894165, "elapsed_s": 401.4}150{"event": "train", "step": 2900, "loss": -5.792809963226318, "grad_norm": 0.7220891118049622, "lr": 8.208128378689484e-05, "ex_per_s": 461.8002790121132, "data_frac": 0.03335678884572673, "mem_mb": 86.535168, "nll": -5.792809963226318, "elapsed_s": 401.9}151{"event": "train", "step": 2920, "loss": -5.968239784240723, "grad_norm": 0.6517534852027893, "lr": 8.03311916060725e-05, "ex_per_s": 464.35107276754235, "data_frac": 0.033337495342398944, "mem_mb": 86.535168, "nll": -5.968239784240723, "elapsed_s": 402.5}152{"event": "train", "step": 2940, "loss": -5.901467323303223, "grad_norm": 0.5381296873092651, "lr": 7.860424703205225e-05, "ex_per_s": 466.8951373269635, "data_frac": 0.03331878816199549, "mem_mb": 86.535168, "nll": -5.901467323303223, "elapsed_s": 403.0}153{"event": "train", "step": 2960, "loss": -5.936922073364258, "grad_norm": 0.2195059061050415, "lr": 7.690092219423722e-05, "ex_per_s": 469.4311824355728, "data_frac": 0.033299980951196295, "mem_mb": 86.535168, "nll": -5.936922073364258, "elapsed_s": 403.6}154{"event": "train", "step": 2980, "loss": -5.888759136199951, "grad_norm": 0.6943848133087158, "lr": 7.522168276463048e-05, "ex_per_s": 471.9605520252125, "data_frac": 0.03328113552193277, "mem_mb": 86.535168, "nll": -5.888759136199951, "elapsed_s": 404.1}155{"event": "train", "step": 3000, "loss": -5.99542236328125, "grad_norm": 0.25452950596809387, "lr": 7.356698783052496e-05, "ex_per_s": 474.47631273895934, "data_frac": 0.03326352429682012, "mem_mb": 86.535168, "nll": -5.99542236328125, "elapsed_s": 404.7}156{"event": "val", "step": 3000, "val_nll": -6.449487894925626, "best": -6.44968405509289, "bad": 1, "mem_mb": 86.535168, "elapsed_s": 466.4}157{"event": "train", "step": 3020, "loss": -5.958747863769531, "grad_norm": 1.2017390727996826, "lr": 7.193728976899362e-05, "ex_per_s": 413.9038678841138, "data_frac": 0.0288486886488475, "mem_mb": 86.535168, "nll": -5.958747863769531, "elapsed_s": 467.0}158{"event": "train", "step": 3040, "loss": -5.860010147094727, "grad_norm": 0.8937140703201294, "lr": 7.033303412321404e-05, "ex_per_s": 416.15439393867257, "data_frac": 0.02883803366127355, "mem_mb": 86.535168, "nll": -5.860010147094727, "elapsed_s": 467.5}159{"event": "train", "step": 3060, "loss": -6.071678638458252, "grad_norm": 0.4028576910495758, "lr": 6.875465948066139e-05, "ex_per_s": 418.3990781053251, "data_frac": 0.028827039435632706, "mem_mb": 86.535168, "nll": -6.071678638458252, "elapsed_s": 468.1}160{"event": "train", "step": 3080, "loss": -5.969563961029053, "grad_norm": 0.08275040984153748, "lr": 6.720259735320299e-05, "ex_per_s": 420.6392226356724, "data_frac": 0.028815985384833712, "mem_mb": 86.535168, "nll": -5.969563961029053, "elapsed_s": 468.6}161{"event": "train", "step": 3100, "loss": -6.016390800476074, "grad_norm": 0.2627483010292053, "lr": 6.567727205912722e-05, "ex_per_s": 422.8747173436162, "data_frac": 0.028804781464413818, "mem_mb": 86.535168, "nll": -6.016390800476074, "elapsed_s": 469.2}162{"event": "train", "step": 3120, "loss": -6.091325759887695, "grad_norm": 0.42851361632347107, "lr": 6.417910060713926e-05, "ex_per_s": 425.10415819961554, "data_frac": 0.028793777966501666, "mem_mb": 86.535168, "nll": -6.091325759887695, "elapsed_s": 469.7}163{"event": "train", "step": 3140, "loss": -6.011155128479004, "grad_norm": 0.14957058429718018, "lr": 6.270849258235485e-05, "ex_per_s": 427.3285741592263, "data_frac": 0.02878293298103422, "mem_mb": 86.535168, "nll": -6.011155128479004, "elapsed_s": 470.3}164{"event": "train", "step": 3160, "loss": -6.019983291625977, "grad_norm": 0.15608350932598114, "lr": 6.126585003432406e-05, "ex_per_s": 429.54187129707475, "data_frac": 0.02877248048531877, "mem_mb": 86.535168, "nll": -6.019983291625977, "elapsed_s": 470.8}165{"event": "train", "step": 3180, "loss": -6.09419059753418, "grad_norm": 0.26855650544166565, "lr": 5.985156736711483e-05, "ex_per_s": 431.75194171480155, "data_frac": 0.02876313133799992, "mem_mb": 86.535168, "nll": -6.09419059753418, "elapsed_s": 471.4}166{"event": "train", "step": 3200, "loss": -6.002453804016113, "grad_norm": 0.18004706501960754, "lr": 5.846603123148688e-05, "ex_per_s": 433.9609729206291, "data_frac": 0.02875244212260365, "mem_mb": 86.535168, "nll": -6.002453804016113, "elapsed_s": 471.9}167{"event": "train", "step": 3220, "loss": -5.916248798370361, "grad_norm": 0.09150106459856033, "lr": 5.710962041918516e-05, "ex_per_s": 436.16474271590903, "data_frac": 0.028741234631933222, "mem_mb": 86.535168, "nll": -5.916248798370361, "elapsed_s": 472.5}168{"event": "train", "step": 3240, "loss": -5.801613807678223, "grad_norm": 0.4568430483341217, "lr": 5.578270575938211e-05, "ex_per_s": 438.3639930355754, "data_frac": 0.028730335425191482, "mem_mb": 86.535168, "nll": -5.801613807678223, "elapsed_s": 473.0}169{"event": "train", "step": 3260, "loss": -5.960525989532471, "grad_norm": 0.31972217559814453, "lr": 5.448565001729653e-05, "ex_per_s": 440.55769965270576, "data_frac": 0.02871945499280501, "mem_mb": 86.535168, "nll": -5.960525989532471, "elapsed_s": 473.6}170{"event": "train", "step": 3280, "loss": -6.0181708335876465, "grad_norm": 0.3915044367313385, "lr": 5.3218807795017296e-05, "ex_per_s": 442.74729883622865, "data_frac": 0.028708655938903038, "mem_mb": 86.535168, "nll": -6.0181708335876465, "elapsed_s": 474.1}171{"event": "train", "step": 3300, "loss": -5.803713321685791, "grad_norm": 0.1261027753353119, "lr": 5.198252543455864e-05, "ex_per_s": 444.93059199429916, "data_frac": 0.028698118710830646, "mem_mb": 86.535168, "nll": -5.803713321685791, "elapsed_s": 474.7}172{"event": "train", "step": 3320, "loss": -5.990504741668701, "grad_norm": 0.1781640499830246, "lr": 5.07771409231737e-05, "ex_per_s": 447.10882701907315, "data_frac": 0.028687327819398072, "mem_mb": 86.535168, "nll": -5.990504741668701, "elapsed_s": 475.2}173{"event": "train", "step": 3340, "loss": -5.931906223297119, "grad_norm": 0.09085994958877563, "lr": 4.960298380095221e-05, "ex_per_s": 449.28138788970244, "data_frac": 0.028677457894652425, "mem_mb": 86.535168, "nll": -5.931906223297119, "elapsed_s": 475.8}174{"event": "train", "step": 3360, "loss": -5.911627292633057, "grad_norm": 0.10811258852481842, "lr": 4.846037507072751e-05, "ex_per_s": 451.44731480375964, "data_frac": 0.028668037237682187, "mem_mb": 86.535168, "nll": -5.911627292633057, "elapsed_s": 476.3}175{"event": "train", "step": 3380, "loss": -6.120527267456055, "grad_norm": 0.17647716403007507, "lr": 4.7349627110317556e-05, "ex_per_s": 453.61091517943953, "data_frac": 0.028657455721873098, "mem_mb": 86.535168, "nll": -6.120527267456055, "elapsed_s": 476.9}176{"event": "train", "step": 3400, "loss": -5.880710124969482, "grad_norm": 0.5955110788345337, "lr": 4.6271043587123986e-05, "ex_per_s": 455.7692023144996, "data_frac": 0.028646963227486482, "mem_mb": 86.535168, "nll": -5.880710124969482, "elapsed_s": 477.4}177{"event": "train", "step": 3420, "loss": -5.837514877319336, "grad_norm": 0.13344787061214447, "lr": 4.52249193751124e-05, "ex_per_s": 457.9230699425996, "data_frac": 0.028636482822608934, "mem_mb": 86.535168, "nll": -5.837514877319336, "elapsed_s": 478.0}178{"event": "train", "step": 3440, "loss": -5.917680263519287, "grad_norm": 0.08200714737176895, "lr": 4.4211540474196926e-05, "ex_per_s": 460.07118083395426, "data_frac": 0.028626196346947172, "mem_mb": 86.535168, "nll": -5.917680263519287, "elapsed_s": 478.5}179{"event": "train", "step": 3460, "loss": -6.059091567993164, "grad_norm": 0.30697768926620483, "lr": 4.3231183932050546e-05, "ex_per_s": 462.21505567604027, "data_frac": 0.028615670785212176, "mem_mb": 86.535168, "nll": -6.059091567993164, "elapsed_s": 479.1}180{"event": "train", "step": 3480, "loss": -5.918484687805176, "grad_norm": 0.10291261225938797, "lr": 4.228411776836306e-05, "ex_per_s": 464.3552540432731, "data_frac": 0.028605114859907528, "mem_mb": 86.535168, "nll": -5.918484687805176, "elapsed_s": 479.6}181{"event": "train", "step": 3500, "loss": -6.038266658782959, "grad_norm": 0.23202019929885864, "lr": 4.137060090156723e-05, "ex_per_s": 466.48933376778535, "data_frac": 0.028594559294946392, "mem_mb": 86.535168, "nll": -6.038266658782959, "elapsed_s": 480.2}182{"event": "val", "step": 3500, "val_nll": -6.4504765663426005, "best": -6.4504765663426005, "bad": 0, "mem_mb": 86.535168, "elapsed_s": 542.0}183{"event": "train", "step": 3520, "loss": -5.935113906860352, "grad_norm": 0.3620780110359192, "lr": 4.049088307805306e-05, "ex_per_s": 415.1888739217289, "data_frac": 0.02532593861129629, "mem_mb": 86.535168, "nll": -5.935113906860352, "elapsed_s": 542.6}184{"event": "train", "step": 3540, "loss": -6.136935710906982, "grad_norm": 0.5370523929595947, "lr": 3.964520480388956e-05, "ex_per_s": 417.124727236546, "data_frac": 0.02532020184562539, "mem_mb": 86.535168, "nll": -6.136935710906982, "elapsed_s": 543.2}185{"event": "train", "step": 3560, "loss": -6.141398906707764, "grad_norm": 0.09299107640981674, "lr": 3.88337972790728e-05, "ex_per_s": 419.0555814148192, "data_frac": 0.025314305820399185, "mem_mb": 86.535168, "nll": -6.141398906707764, "elapsed_s": 543.7}186{"event": "train", "step": 3580, "loss": -5.9022016525268555, "grad_norm": 0.28306370973587036, "lr": 3.805688233431824e-05, "ex_per_s": 420.98351626697337, "data_frac": 0.025308339282031966, "mem_mb": 86.535168, "nll": -5.9022016525268555, "elapsed_s": 544.3}187{"event": "train", "step": 3600, "loss": -6.097652912139893, "grad_norm": 0.2914966940879822, "lr": 3.7314672370414325e-05, "ex_per_s": 422.9078467745829, "data_frac": 0.025302577349120184, "mem_mb": 86.535168, "nll": -6.097652912139893, "elapsed_s": 544.8}188{"event": "train", "step": 3620, "loss": -5.910172462463379, "grad_norm": 1.0272372961044312, "lr": 3.660737030015427e-05, "ex_per_s": 424.8277141685818, "data_frac": 0.025296803422717725, "mem_mb": 86.535168, "nll": -5.910172462463379, "elapsed_s": 545.4}189{"event": "train", "step": 3640, "loss": -6.053393363952637, "grad_norm": 0.5445851683616638, "lr": 3.5935169492861716e-05, "ex_per_s": 426.74298171190566, "data_frac": 0.025291164167360988, "mem_mb": 86.535168, "nll": -6.053393363952637, "elapsed_s": 545.9}190{"event": "train", "step": 3660, "loss": -5.832545757293701, "grad_norm": 0.6098777055740356, "lr": 3.52982537215255e-05, "ex_per_s": 428.65554401820594, "data_frac": 0.025285057799190165, "mem_mb": 86.535168, "nll": -5.832545757293701, "elapsed_s": 546.5}191{"event": "train", "step": 3680, "loss": -5.9932780265808105, "grad_norm": 0.25747063755989075, "lr": 3.469679711255794e-05, "ex_per_s": 430.5634066347911, "data_frac": 0.02527920012468594, "mem_mb": 86.535168, "nll": -5.9932780265808105, "elapsed_s": 547.0}192{"event": "train", "step": 3700, "loss": -6.06676721572876, "grad_norm": 0.7818856239318848, "lr": 3.41309640981904e-05, "ex_per_s": 432.4686892168065, "data_frac": 0.025273485968926958, "mem_mb": 86.535168, "nll": -6.06676721572876, "elapsed_s": 547.6}193{"event": "train", "step": 3720, "loss": -5.947232246398926, "grad_norm": 0.4552668631076813, "lr": 3.3600909371519225e-05, "ex_per_s": 434.3699958963345, "data_frac": 0.02526759516054826, "mem_mb": 86.535168, "nll": -5.947232246398926, "elapsed_s": 548.1}194{"event": "train", "step": 3740, "loss": -6.191053867340088, "grad_norm": 0.21633362770080566, "lr": 3.31067778442141e-05, "ex_per_s": 436.26753473587297, "data_frac": 0.025261920810321834, "mem_mb": 86.535168, "nll": -6.191053867340088, "elapsed_s": 548.7}195{"event": "train", "step": 3760, "loss": -6.147576332092285, "grad_norm": 0.11195147782564163, "lr": 3.2648704606900735e-05, "ex_per_s": 438.1611532922269, "data_frac": 0.02525619536208033, "mem_mb": 86.535168, "nll": -6.147576332092285, "elapsed_s": 549.2}196{"event": "train", "step": 3780, "loss": -5.96305513381958, "grad_norm": 0.11359459906816483, "lr": 3.222681489222838e-05, "ex_per_s": 440.0512820198345, "data_frac": 0.025250286094302358, "mem_mb": 86.535168, "nll": -5.96305513381958, "elapsed_s": 549.8}197{"event": "train", "step": 3800, "loss": -6.100017070770264, "grad_norm": 0.07617157697677612, "lr": 3.1841224040632484e-05, "ex_per_s": 441.9372901836306, "data_frac": 0.02524467789712414, "mem_mb": 86.535168, "nll": -6.100017070770264, "elapsed_s": 550.3}198{"event": "train", "step": 3820, "loss": -5.970897197723389, "grad_norm": 0.16164164245128632, "lr": 3.149203746880175e-05, "ex_per_s": 443.8189081601646, "data_frac": 0.02523892301231528, "mem_mb": 86.535168, "nll": -5.970897197723389, "elapsed_s": 550.9}199{"event": "train", "step": 3840, "loss": -5.871518611907959, "grad_norm": 0.8552179336547852, "lr": 3.117935064085835e-05, "ex_per_s": 445.69318483870006, "data_frac": 0.025234419742913593, "mem_mb": 86.535168, "nll": -5.871518611907959, "elapsed_s": 551.4}200{"event": "train", "step": 3860, "loss": -6.014821529388428, "grad_norm": 0.30839428305625916, "lr": 3.090324904225885e-05, "ex_per_s": 447.56485585861304, "data_frac": 0.025229837511112702, "mem_mb": 86.535168, "nll": -6.014821529388428, "elapsed_s": 552.0}201{"event": "train", "step": 3880, "loss": -5.8924407958984375, "grad_norm": 0.15562200546264648, "lr": 3.0663808156423456e-05, "ex_per_s": 449.4360605049075, "data_frac": 0.0252241837452861, "mem_mb": 86.535168, "nll": -5.8924407958984375, "elapsed_s": 552.5}202{"event": "train", "step": 3900, "loss": -6.179962158203125, "grad_norm": 0.8057243824005127, "lr": 3.046109344409957e-05, "ex_per_s": 451.3032182337553, "data_frac": 0.02521859070487621, "mem_mb": 86.535168, "nll": -6.179962158203125, "elapsed_s": 553.1}203{"event": "train", "step": 3920, "loss": -6.038008213043213, "grad_norm": 0.20454788208007812, "lr": 3.0295160325465435e-05, "ex_per_s": 453.16593699946174, "data_frac": 0.025212640618678464, "mem_mb": 86.535168, "nll": -6.038008213043213, "elapsed_s": 553.6}204{"event": "train", "step": 3940, "loss": -5.923555374145508, "grad_norm": 0.29861897230148315, "lr": 3.016605416497886e-05, "ex_per_s": 455.02591661368336, "data_frac": 0.02520741459066002, "mem_mb": 86.535168, "nll": -5.923555374145508, "elapsed_s": 554.2}205{"event": "train", "step": 3960, "loss": -6.068994045257568, "grad_norm": 0.0681753009557724, "lr": 3.0073810258974985e-05, "ex_per_s": 456.88148121947415, "data_frac": 0.02520187261111933, "mem_mb": 86.535168, "nll": -6.068994045257568, "elapsed_s": 554.7}206{"event": "train", "step": 3980, "loss": -5.8057122230529785, "grad_norm": 0.13688510656356812, "lr": 3.0018453826016683e-05, "ex_per_s": 458.7293205584508, "data_frac": 0.025197377185605132, "mem_mb": 86.535168, "nll": -5.8057122230529785, "elapsed_s": 555.3}207{"event": "train", "step": 4000, "loss": -6.0567708015441895, "grad_norm": 0.7389938831329346, "lr": 2.9999999999999997e-05, "ex_per_s": 460.5782727159962, "data_frac": 0.025191368062004806, "mem_mb": 86.535168, "nll": -6.0567708015441895, "elapsed_s": 555.8}208{"event": "val", "step": 4000, "val_nll": -6.450252340094127, "best": -6.4504765663426005, "bad": 1, "mem_mb": 86.535168, "elapsed_s": 617.6}209{"event": "done", "steps": 4000, "best_val": -6.4504765663426005, "wall_s": 617.6293749809265, "ex_per_s": 414.48805767683206, "peak_mem_mb": 86.535168, "n_params": 3200259, "device": "mps", "data_time_frac": 0.022670459804084265, "elapsed_s": 617.6}210