{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 19.92814371257485,
  "eval_steps": 500,
  "global_step": 3328,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.059880239520958084,
      "grad_norm": 6.431809425354004,
      "learning_rate": 1.1976047904191617e-05,
      "loss": 1.1211,
      "step": 10
    },
    {
      "epoch": 0.11976047904191617,
      "grad_norm": 2.930279493331909,
      "learning_rate": 2.3952095808383234e-05,
      "loss": 0.6469,
      "step": 20
    },
    {
      "epoch": 0.17964071856287425,
      "grad_norm": 2.0707571506500244,
      "learning_rate": 3.592814371257485e-05,
      "loss": 0.3656,
      "step": 30
    },
    {
      "epoch": 0.23952095808383234,
      "grad_norm": 1.0708367824554443,
      "learning_rate": 4.790419161676647e-05,
      "loss": 0.2645,
      "step": 40
    },
    {
      "epoch": 0.2994011976047904,
      "grad_norm": 0.8269807696342468,
      "learning_rate": 5.988023952095808e-05,
      "loss": 0.2082,
      "step": 50
    },
    {
      "epoch": 0.3592814371257485,
      "grad_norm": 0.9306111931800842,
      "learning_rate": 7.18562874251497e-05,
      "loss": 0.1856,
      "step": 60
    },
    {
      "epoch": 0.41916167664670656,
      "grad_norm": 2.0250155925750732,
      "learning_rate": 8.383233532934131e-05,
      "loss": 0.1663,
      "step": 70
    },
    {
      "epoch": 0.47904191616766467,
      "grad_norm": 1.3040319681167603,
      "learning_rate": 9.580838323353294e-05,
      "loss": 0.1629,
      "step": 80
    },
    {
      "epoch": 0.5389221556886228,
      "grad_norm": 0.7521260380744934,
      "learning_rate": 0.00010778443113772456,
      "loss": 0.1357,
      "step": 90
    },
    {
      "epoch": 0.5988023952095808,
      "grad_norm": 1.0798639059066772,
      "learning_rate": 0.00011976047904191617,
      "loss": 0.1219,
      "step": 100
    },
    {
      "epoch": 0.6586826347305389,
      "grad_norm": 1.1869535446166992,
      "learning_rate": 0.0001317365269461078,
      "loss": 0.1231,
      "step": 110
    },
    {
      "epoch": 0.718562874251497,
      "grad_norm": 0.6623774766921997,
      "learning_rate": 0.0001437125748502994,
      "loss": 0.1056,
      "step": 120
    },
    {
      "epoch": 0.7784431137724551,
      "grad_norm": 0.7135199308395386,
      "learning_rate": 0.00015568862275449103,
      "loss": 0.0958,
      "step": 130
    },
    {
      "epoch": 0.8383233532934131,
      "grad_norm": 0.5620641112327576,
      "learning_rate": 0.00016766467065868263,
      "loss": 0.1049,
      "step": 140
    },
    {
      "epoch": 0.8982035928143712,
      "grad_norm": 0.9924416542053223,
      "learning_rate": 0.00017964071856287425,
      "loss": 0.094,
      "step": 150
    },
    {
      "epoch": 0.9580838323353293,
      "grad_norm": 1.8146198987960815,
      "learning_rate": 0.00019161676646706587,
      "loss": 0.0982,
      "step": 160
    },
    {
      "epoch": 1.0179640718562875,
      "grad_norm": 0.9982231259346008,
      "learning_rate": 0.00019999955550902674,
      "loss": 0.0987,
      "step": 170
    },
    {
      "epoch": 1.0778443113772456,
      "grad_norm": 0.7121732831001282,
      "learning_rate": 0.0001999916535572039,
      "loss": 0.0865,
      "step": 180
    },
    {
      "epoch": 1.1377245508982037,
      "grad_norm": 0.7211266160011292,
      "learning_rate": 0.00019997387492660495,
      "loss": 0.0815,
      "step": 190
    },
    {
      "epoch": 1.1976047904191618,
      "grad_norm": 0.7090721130371094,
      "learning_rate": 0.00019994622137331462,
      "loss": 0.078,
      "step": 200
    },
    {
      "epoch": 1.2574850299401197,
      "grad_norm": 0.6617732644081116,
      "learning_rate": 0.00019990869562881356,
      "loss": 0.0751,
      "step": 210
    },
    {
      "epoch": 1.3173652694610778,
      "grad_norm": 0.9563634395599365,
      "learning_rate": 0.00019986130139970835,
      "loss": 0.077,
      "step": 220
    },
    {
      "epoch": 1.377245508982036,
      "grad_norm": 0.4591192901134491,
      "learning_rate": 0.00019980404336736554,
      "loss": 0.0742,
      "step": 230
    },
    {
      "epoch": 1.437125748502994,
      "grad_norm": 0.3945522606372833,
      "learning_rate": 0.00019973692718744902,
      "loss": 0.073,
      "step": 240
    },
    {
      "epoch": 1.4970059880239521,
      "grad_norm": 0.5244777202606201,
      "learning_rate": 0.0001996599594893617,
      "loss": 0.0689,
      "step": 250
    },
    {
      "epoch": 1.55688622754491,
      "grad_norm": 0.5341875553131104,
      "learning_rate": 0.00019957314787559036,
      "loss": 0.069,
      "step": 260
    },
    {
      "epoch": 1.6167664670658684,
      "grad_norm": 0.33649489283561707,
      "learning_rate": 0.00019947650092095494,
      "loss": 0.0651,
      "step": 270
    },
    {
      "epoch": 1.6766467065868262,
      "grad_norm": 0.48326146602630615,
      "learning_rate": 0.00019937002817176146,
      "loss": 0.0664,
      "step": 280
    },
    {
      "epoch": 1.7365269461077846,
      "grad_norm": 0.4517434239387512,
      "learning_rate": 0.00019925374014485912,
      "loss": 0.0669,
      "step": 290
    },
    {
      "epoch": 1.7964071856287425,
      "grad_norm": 0.529296338558197,
      "learning_rate": 0.00019912764832660156,
      "loss": 0.0665,
      "step": 300
    },
    {
      "epoch": 1.8562874251497006,
      "grad_norm": 0.5424647927284241,
      "learning_rate": 0.00019899176517171208,
      "loss": 0.0598,
      "step": 310
    },
    {
      "epoch": 1.9161676646706587,
      "grad_norm": 0.3814735412597656,
      "learning_rate": 0.0001988461041020537,
      "loss": 0.054,
      "step": 320
    },
    {
      "epoch": 1.9760479041916168,
      "grad_norm": 0.48093119263648987,
      "learning_rate": 0.0001986906795053032,
      "loss": 0.0553,
      "step": 330
    },
    {
      "epoch": 2.035928143712575,
      "grad_norm": 0.5698864459991455,
      "learning_rate": 0.00019852550673353004,
      "loss": 0.0586,
      "step": 340
    },
    {
      "epoch": 2.095808383233533,
      "grad_norm": 0.839526355266571,
      "learning_rate": 0.00019835060210168017,
      "loss": 0.0565,
      "step": 350
    },
    {
      "epoch": 2.155688622754491,
      "grad_norm": 0.5425819754600525,
      "learning_rate": 0.00019816598288596412,
      "loss": 0.0573,
      "step": 360
    },
    {
      "epoch": 2.215568862275449,
      "grad_norm": 0.2914687693119049,
      "learning_rate": 0.00019797166732215076,
      "loss": 0.052,
      "step": 370
    },
    {
      "epoch": 2.2754491017964074,
      "grad_norm": 0.5953413248062134,
      "learning_rate": 0.0001977676746037662,
      "loss": 0.054,
      "step": 380
    },
    {
      "epoch": 2.3353293413173652,
      "grad_norm": 0.42740774154663086,
      "learning_rate": 0.0001975540248801977,
      "loss": 0.0554,
      "step": 390
    },
    {
      "epoch": 2.3952095808383236,
      "grad_norm": 0.37728944420814514,
      "learning_rate": 0.00019733073925470352,
      "loss": 0.0513,
      "step": 400
    },
    {
      "epoch": 2.4550898203592815,
      "grad_norm": 0.475818932056427,
      "learning_rate": 0.00019709783978232842,
      "loss": 0.0542,
      "step": 410
    },
    {
      "epoch": 2.5149700598802394,
      "grad_norm": 0.6117169260978699,
      "learning_rate": 0.00019685534946772536,
      "loss": 0.0554,
      "step": 420
    },
    {
      "epoch": 2.5748502994011977,
      "grad_norm": 0.34023669362068176,
      "learning_rate": 0.00019660329226288273,
      "loss": 0.0535,
      "step": 430
    },
    {
      "epoch": 2.6347305389221556,
      "grad_norm": 0.4431716501712799,
      "learning_rate": 0.00019634169306475917,
      "loss": 0.0496,
      "step": 440
    },
    {
      "epoch": 2.694610778443114,
      "grad_norm": 0.5750408172607422,
      "learning_rate": 0.00019607057771282372,
      "loss": 0.0505,
      "step": 450
    },
    {
      "epoch": 2.754491017964072,
      "grad_norm": 0.45190536975860596,
      "learning_rate": 0.00019578997298650395,
      "loss": 0.0438,
      "step": 460
    },
    {
      "epoch": 2.81437125748503,
      "grad_norm": 0.37500932812690735,
      "learning_rate": 0.00019549990660254073,
      "loss": 0.049,
      "step": 470
    },
    {
      "epoch": 2.874251497005988,
      "grad_norm": 0.35688552260398865,
      "learning_rate": 0.00019520040721225033,
      "loss": 0.0507,
      "step": 480
    },
    {
      "epoch": 2.934131736526946,
      "grad_norm": 0.3029610216617584,
      "learning_rate": 0.00019489150439869465,
      "loss": 0.0497,
      "step": 490
    },
    {
      "epoch": 2.9940119760479043,
      "grad_norm": 0.4592169523239136,
      "learning_rate": 0.000194573228673759,
      "loss": 0.0504,
      "step": 500
    },
    {
      "epoch": 3.053892215568862,
      "grad_norm": 0.3194161355495453,
      "learning_rate": 0.0001942456114751382,
      "loss": 0.0481,
      "step": 510
    },
    {
      "epoch": 3.1137724550898205,
      "grad_norm": 0.5735427141189575,
      "learning_rate": 0.00019390868516323156,
      "loss": 0.0463,
      "step": 520
    },
    {
      "epoch": 3.1736526946107784,
      "grad_norm": 0.251126766204834,
      "learning_rate": 0.00019356248301794632,
      "loss": 0.0466,
      "step": 530
    },
    {
      "epoch": 3.2335329341317367,
      "grad_norm": 0.2263738214969635,
      "learning_rate": 0.00019320703923541045,
      "loss": 0.0475,
      "step": 540
    },
    {
      "epoch": 3.2934131736526946,
      "grad_norm": 0.3312239348888397,
      "learning_rate": 0.00019284238892459485,
      "loss": 0.0451,
      "step": 550
    },
    {
      "epoch": 3.3532934131736525,
      "grad_norm": 0.27982234954833984,
      "learning_rate": 0.00019246856810384576,
      "loss": 0.0489,
      "step": 560
    },
    {
      "epoch": 3.413173652694611,
      "grad_norm": 0.34601065516471863,
      "learning_rate": 0.0001920856136973266,
      "loss": 0.0481,
      "step": 570
    },
    {
      "epoch": 3.4730538922155687,
      "grad_norm": 0.4402002692222595,
      "learning_rate": 0.0001916935635313711,
      "loss": 0.0437,
      "step": 580
    },
    {
      "epoch": 3.532934131736527,
      "grad_norm": 0.6538058519363403,
      "learning_rate": 0.00019129245633074703,
      "loss": 0.0478,
      "step": 590
    },
    {
      "epoch": 3.592814371257485,
      "grad_norm": 0.36621013283729553,
      "learning_rate": 0.0001908823317148309,
      "loss": 0.0436,
      "step": 600
    },
    {
      "epoch": 3.6526946107784433,
      "grad_norm": 0.36551448702812195,
      "learning_rate": 0.00019046323019369474,
      "loss": 0.044,
      "step": 610
    },
    {
      "epoch": 3.712574850299401,
      "grad_norm": 0.4750276505947113,
      "learning_rate": 0.0001900351931641047,
      "loss": 0.0453,
      "step": 620
    },
    {
      "epoch": 3.772455089820359,
      "grad_norm": 0.6237766742706299,
      "learning_rate": 0.00018959826290543207,
      "loss": 0.0466,
      "step": 630
    },
    {
      "epoch": 3.8323353293413174,
      "grad_norm": 0.45154473185539246,
      "learning_rate": 0.00018915248257547713,
      "loss": 0.0448,
      "step": 640
    },
    {
      "epoch": 3.8922155688622757,
      "grad_norm": 0.3980957865715027,
      "learning_rate": 0.0001886978962062062,
      "loss": 0.0502,
      "step": 650
    },
    {
      "epoch": 3.9520958083832336,
      "grad_norm": 0.3399762213230133,
      "learning_rate": 0.00018823454869940242,
      "loss": 0.046,
      "step": 660
    },
    {
      "epoch": 4.0119760479041915,
      "grad_norm": 0.3906604051589966,
      "learning_rate": 0.00018776248582223063,
      "loss": 0.0416,
      "step": 670
    },
    {
      "epoch": 4.07185628742515,
      "grad_norm": 0.35462263226509094,
      "learning_rate": 0.00018728175420271658,
      "loss": 0.0385,
      "step": 680
    },
    {
      "epoch": 4.131736526946108,
      "grad_norm": 0.43025287985801697,
      "learning_rate": 0.00018679240132514136,
      "loss": 0.042,
      "step": 690
    },
    {
      "epoch": 4.191616766467066,
      "grad_norm": 0.31433719396591187,
      "learning_rate": 0.0001862944755253511,
      "loss": 0.0381,
      "step": 700
    },
    {
      "epoch": 4.251497005988024,
      "grad_norm": 0.545958936214447,
      "learning_rate": 0.00018578802598598256,
      "loss": 0.0402,
      "step": 710
    },
    {
      "epoch": 4.311377245508982,
      "grad_norm": 0.283829003572464,
      "learning_rate": 0.00018527310273160522,
      "loss": 0.0409,
      "step": 720
    },
    {
      "epoch": 4.37125748502994,
      "grad_norm": 0.3808535635471344,
      "learning_rate": 0.00018474975662377996,
      "loss": 0.0411,
      "step": 730
    },
    {
      "epoch": 4.431137724550898,
      "grad_norm": 0.25326570868492126,
      "learning_rate": 0.00018421803935603532,
      "loss": 0.0431,
      "step": 740
    },
    {
      "epoch": 4.491017964071856,
      "grad_norm": 0.6720901131629944,
      "learning_rate": 0.0001836780034487614,
      "loss": 0.0436,
      "step": 750
    },
    {
      "epoch": 4.550898203592815,
      "grad_norm": 0.4183627665042877,
      "learning_rate": 0.00018312970224402224,
      "loss": 0.0408,
      "step": 760
    },
    {
      "epoch": 4.610778443113772,
      "grad_norm": 0.2710108160972595,
      "learning_rate": 0.00018257318990028682,
      "loss": 0.0427,
      "step": 770
    },
    {
      "epoch": 4.6706586826347305,
      "grad_norm": 0.38718074560165405,
      "learning_rate": 0.00018200852138707975,
      "loss": 0.0467,
      "step": 780
    },
    {
      "epoch": 4.730538922155689,
      "grad_norm": 0.2398432046175003,
      "learning_rate": 0.00018143575247955135,
      "loss": 0.0406,
      "step": 790
    },
    {
      "epoch": 4.790419161676647,
      "grad_norm": 0.3367202877998352,
      "learning_rate": 0.0001808549397529689,
      "loss": 0.0407,
      "step": 800
    },
    {
      "epoch": 4.850299401197605,
      "grad_norm": 0.30790024995803833,
      "learning_rate": 0.00018026614057712795,
      "loss": 0.0404,
      "step": 810
    },
    {
      "epoch": 4.910179640718563,
      "grad_norm": 0.266544908285141,
      "learning_rate": 0.00017966941311068594,
      "loss": 0.0373,
      "step": 820
    },
    {
      "epoch": 4.970059880239521,
      "grad_norm": 0.2760760188102722,
      "learning_rate": 0.00017906481629541743,
      "loss": 0.0376,
      "step": 830
    },
    {
      "epoch": 5.029940119760479,
      "grad_norm": 0.23295895755290985,
      "learning_rate": 0.00017845240985039218,
      "loss": 0.0375,
      "step": 840
    },
    {
      "epoch": 5.089820359281437,
      "grad_norm": 0.3514467477798462,
      "learning_rate": 0.0001778322542660764,
      "loss": 0.037,
      "step": 850
    },
    {
      "epoch": 5.149700598802395,
      "grad_norm": 0.3003256618976593,
      "learning_rate": 0.0001772044107983577,
      "loss": 0.0379,
      "step": 860
    },
    {
      "epoch": 5.209580838323353,
      "grad_norm": 0.24236004054546356,
      "learning_rate": 0.0001765689414624947,
      "loss": 0.0399,
      "step": 870
    },
    {
      "epoch": 5.269461077844311,
      "grad_norm": 0.22592489421367645,
      "learning_rate": 0.00017592590902699138,
      "loss": 0.0388,
      "step": 880
    },
    {
      "epoch": 5.3293413173652695,
      "grad_norm": 0.21647202968597412,
      "learning_rate": 0.00017527537700739708,
      "loss": 0.0409,
      "step": 890
    },
    {
      "epoch": 5.389221556886228,
      "grad_norm": 0.37862035632133484,
      "learning_rate": 0.00017461740966003282,
      "loss": 0.0429,
      "step": 900
    },
    {
      "epoch": 5.449101796407185,
      "grad_norm": 0.29174819588661194,
      "learning_rate": 0.00017395207197564442,
      "loss": 0.0402,
      "step": 910
    },
    {
      "epoch": 5.508982035928144,
      "grad_norm": 0.2290957272052765,
      "learning_rate": 0.00017327942967298285,
      "loss": 0.0372,
      "step": 920
    },
    {
      "epoch": 5.568862275449102,
      "grad_norm": 0.23877893388271332,
      "learning_rate": 0.0001725995491923131,
      "loss": 0.0348,
      "step": 930
    },
    {
      "epoch": 5.62874251497006,
      "grad_norm": 0.3068605959415436,
      "learning_rate": 0.00017191249768885137,
      "loss": 0.0434,
      "step": 940
    },
    {
      "epoch": 5.688622754491018,
      "grad_norm": 0.2592070698738098,
      "learning_rate": 0.00017121834302613186,
      "loss": 0.0349,
      "step": 950
    },
    {
      "epoch": 5.748502994011976,
      "grad_norm": 0.2851567268371582,
      "learning_rate": 0.00017051715376930363,
      "loss": 0.0394,
      "step": 960
    },
    {
      "epoch": 5.808383233532934,
      "grad_norm": 0.3089054226875305,
      "learning_rate": 0.00016980899917835783,
      "loss": 0.0356,
      "step": 970
    },
    {
      "epoch": 5.868263473053892,
      "grad_norm": 0.26863303780555725,
      "learning_rate": 0.00016909394920128695,
      "loss": 0.0366,
      "step": 980
    },
    {
      "epoch": 5.92814371257485,
      "grad_norm": 0.3612504005432129,
      "learning_rate": 0.00016837207446717518,
      "loss": 0.0375,
      "step": 990
    },
    {
      "epoch": 5.9880239520958085,
      "grad_norm": 0.24680382013320923,
      "learning_rate": 0.00016764344627922254,
      "loss": 0.0419,
      "step": 1000
    },
    {
      "epoch": 6.047904191616767,
      "grad_norm": 0.2882663905620575,
      "learning_rate": 0.0001669081366077016,
      "loss": 0.0337,
      "step": 1010
    },
    {
      "epoch": 6.107784431137724,
      "grad_norm": 0.3996296226978302,
      "learning_rate": 0.00016616621808284864,
      "loss": 0.0371,
      "step": 1020
    },
    {
      "epoch": 6.167664670658683,
      "grad_norm": 0.19218039512634277,
      "learning_rate": 0.00016541776398768966,
      "loss": 0.0374,
      "step": 1030
    },
    {
      "epoch": 6.227544910179641,
      "grad_norm": 0.21666273474693298,
      "learning_rate": 0.00016466284825080184,
      "loss": 0.0319,
      "step": 1040
    },
    {
      "epoch": 6.287425149700598,
      "grad_norm": 0.267782986164093,
      "learning_rate": 0.00016390154543901122,
      "loss": 0.0291,
      "step": 1050
    },
    {
      "epoch": 6.347305389221557,
      "grad_norm": 0.30645301938056946,
      "learning_rate": 0.00016313393075002736,
      "loss": 0.0331,
      "step": 1060
    },
    {
      "epoch": 6.407185628742515,
      "grad_norm": 0.29111963510513306,
      "learning_rate": 0.00016236008000501573,
      "loss": 0.0328,
      "step": 1070
    },
    {
      "epoch": 6.467065868263473,
      "grad_norm": 1.8472002744674683,
      "learning_rate": 0.00016158006964110837,
      "loss": 0.0392,
      "step": 1080
    },
    {
      "epoch": 6.526946107784431,
      "grad_norm": 0.2733491063117981,
      "learning_rate": 0.0001607939767038539,
      "loss": 0.0355,
      "step": 1090
    },
    {
      "epoch": 6.586826347305389,
      "grad_norm": 0.3126026391983032,
      "learning_rate": 0.00016000187883960734,
      "loss": 0.0347,
      "step": 1100
    },
    {
      "epoch": 6.6467065868263475,
      "grad_norm": 0.1978524774312973,
      "learning_rate": 0.00015920385428786057,
      "loss": 0.0335,
      "step": 1110
    },
    {
      "epoch": 6.706586826347305,
      "grad_norm": 0.181430846452713,
      "learning_rate": 0.0001583999818735142,
      "loss": 0.0318,
      "step": 1120
    },
    {
      "epoch": 6.766467065868263,
      "grad_norm": 0.29239514470100403,
      "learning_rate": 0.00015759034099909158,
      "loss": 0.034,
      "step": 1130
    },
    {
      "epoch": 6.826347305389222,
      "grad_norm": 0.22587193548679352,
      "learning_rate": 0.00015677501163689604,
      "loss": 0.0361,
      "step": 1140
    },
    {
      "epoch": 6.88622754491018,
      "grad_norm": 0.20451228320598602,
      "learning_rate": 0.00015595407432111134,
      "loss": 0.032,
      "step": 1150
    },
    {
      "epoch": 6.946107784431137,
      "grad_norm": 0.26598894596099854,
      "learning_rate": 0.00015512761013984711,
      "loss": 0.0341,
      "step": 1160
    },
    {
      "epoch": 7.005988023952096,
      "grad_norm": 0.34581083059310913,
      "learning_rate": 0.00015429570072712925,
      "loss": 0.0351,
      "step": 1170
    },
    {
      "epoch": 7.065868263473054,
      "grad_norm": 0.31990233063697815,
      "learning_rate": 0.00015345842825483666,
      "loss": 0.0369,
      "step": 1180
    },
    {
      "epoch": 7.125748502994012,
      "grad_norm": 0.17922641336917877,
      "learning_rate": 0.00015261587542458443,
      "loss": 0.0288,
      "step": 1190
    },
    {
      "epoch": 7.18562874251497,
      "grad_norm": 0.3264837861061096,
      "learning_rate": 0.0001517681254595554,
      "loss": 0.0299,
      "step": 1200
    },
    {
      "epoch": 7.245508982035928,
      "grad_norm": 1.1214922666549683,
      "learning_rate": 0.00015091526209627944,
      "loss": 0.0292,
      "step": 1210
    },
    {
      "epoch": 7.3053892215568865,
      "grad_norm": 0.3472778797149658,
      "learning_rate": 0.0001500573695763625,
      "loss": 0.031,
      "step": 1220
    },
    {
      "epoch": 7.365269461077844,
      "grad_norm": 0.3422839641571045,
      "learning_rate": 0.00014919453263816565,
      "loss": 0.0312,
      "step": 1230
    },
    {
      "epoch": 7.425149700598802,
      "grad_norm": 0.1827375441789627,
      "learning_rate": 0.0001483268365084351,
      "loss": 0.0329,
      "step": 1240
    },
    {
      "epoch": 7.485029940119761,
      "grad_norm": 0.265622079372406,
      "learning_rate": 0.0001474543668938837,
      "loss": 0.0288,
      "step": 1250
    },
    {
      "epoch": 7.544910179640718,
      "grad_norm": 0.56901615858078,
      "learning_rate": 0.0001465772099727256,
      "loss": 0.0322,
      "step": 1260
    },
    {
      "epoch": 7.604790419161676,
      "grad_norm": 0.2764129638671875,
      "learning_rate": 0.00014569545238616368,
      "loss": 0.029,
      "step": 1270
    },
    {
      "epoch": 7.664670658682635,
      "grad_norm": 0.3648919463157654,
      "learning_rate": 0.00014480918122983174,
      "loss": 0.0316,
      "step": 1280
    },
    {
      "epoch": 7.724550898203593,
      "grad_norm": 0.25707149505615234,
      "learning_rate": 0.00014391848404519158,
      "loss": 0.033,
      "step": 1290
    },
    {
      "epoch": 7.7844311377245505,
      "grad_norm": 0.4237244427204132,
      "learning_rate": 0.0001430234488108861,
      "loss": 0.0331,
      "step": 1300
    },
    {
      "epoch": 7.844311377245509,
      "grad_norm": 0.6553322672843933,
      "learning_rate": 0.00014212416393404915,
      "loss": 0.0293,
      "step": 1310
    },
    {
      "epoch": 7.904191616766467,
      "grad_norm": 0.4958610236644745,
      "learning_rate": 0.00014122071824157318,
      "loss": 0.0333,
      "step": 1320
    },
    {
      "epoch": 7.9640718562874255,
      "grad_norm": 0.24589641392230988,
      "learning_rate": 0.0001403132009713354,
      "loss": 0.0339,
      "step": 1330
    },
    {
      "epoch": 8.023952095808383,
      "grad_norm": 0.4254026710987091,
      "learning_rate": 0.0001394017017633831,
      "loss": 0.0361,
      "step": 1340
    },
    {
      "epoch": 8.08383233532934,
      "grad_norm": 0.1916535496711731,
      "learning_rate": 0.00013848631065107973,
      "loss": 0.0318,
      "step": 1350
    },
    {
      "epoch": 8.1437125748503,
      "grad_norm": 0.3515986204147339,
      "learning_rate": 0.00013756711805221168,
      "loss": 0.0327,
      "step": 1360
    },
    {
      "epoch": 8.203592814371257,
      "grad_norm": 0.2998508810997009,
      "learning_rate": 0.00013664421476005723,
      "loss": 0.0337,
      "step": 1370
    },
    {
      "epoch": 8.263473053892216,
      "grad_norm": 0.3755238354206085,
      "learning_rate": 0.0001357176919344186,
      "loss": 0.0282,
      "step": 1380
    },
    {
      "epoch": 8.323353293413174,
      "grad_norm": 0.28876182436943054,
      "learning_rate": 0.0001347876410926175,
      "loss": 0.0279,
      "step": 1390
    },
    {
      "epoch": 8.383233532934131,
      "grad_norm": 0.18108633160591125,
      "learning_rate": 0.00013385415410045555,
      "loss": 0.0244,
      "step": 1400
    },
    {
      "epoch": 8.44311377245509,
      "grad_norm": 0.24347344040870667,
      "learning_rate": 0.00013291732316314024,
      "loss": 0.0268,
      "step": 1410
    },
    {
      "epoch": 8.502994011976048,
      "grad_norm": 0.18795382976531982,
      "learning_rate": 0.00013197724081617736,
      "loss": 0.0266,
      "step": 1420
    },
    {
      "epoch": 8.562874251497005,
      "grad_norm": 0.2501884698867798,
      "learning_rate": 0.00013103399991623078,
      "loss": 0.0323,
      "step": 1430
    },
    {
      "epoch": 8.622754491017965,
      "grad_norm": 0.24649246037006378,
      "learning_rate": 0.00013008769363195052,
      "loss": 0.0282,
      "step": 1440
    },
    {
      "epoch": 8.682634730538922,
      "grad_norm": 0.2907567322254181,
      "learning_rate": 0.00012913841543477005,
      "loss": 0.0261,
      "step": 1450
    },
    {
      "epoch": 8.74251497005988,
      "grad_norm": 0.4828208386898041,
      "learning_rate": 0.00012818625908967367,
      "loss": 0.0301,
      "step": 1460
    },
    {
      "epoch": 8.802395209580839,
      "grad_norm": 0.2910805642604828,
      "learning_rate": 0.0001272313186459348,
      "loss": 0.0265,
      "step": 1470
    },
    {
      "epoch": 8.862275449101796,
      "grad_norm": 0.18343299627304077,
      "learning_rate": 0.00012627368842782626,
      "loss": 0.0274,
      "step": 1480
    },
    {
      "epoch": 8.922155688622755,
      "grad_norm": 0.2735447883605957,
      "learning_rate": 0.00012531346302530348,
      "loss": 0.0314,
      "step": 1490
    },
    {
      "epoch": 8.982035928143713,
      "grad_norm": 0.5026355385780334,
      "learning_rate": 0.0001243507372846613,
      "loss": 0.0234,
      "step": 1500
    },
    {
      "epoch": 9.04191616766467,
      "grad_norm": 0.3467860519886017,
      "learning_rate": 0.00012338560629916544,
      "loss": 0.0271,
      "step": 1510
    },
    {
      "epoch": 9.10179640718563,
      "grad_norm": 0.2949310839176178,
      "learning_rate": 0.00012241816539965996,
      "loss": 0.0289,
      "step": 1520
    },
    {
      "epoch": 9.161676646706587,
      "grad_norm": 0.3735937476158142,
      "learning_rate": 0.00012144851014515055,
      "loss": 0.0268,
      "step": 1530
    },
    {
      "epoch": 9.221556886227544,
      "grad_norm": 0.26456427574157715,
      "learning_rate": 0.000120476736313366,
      "loss": 0.0295,
      "step": 1540
    },
    {
      "epoch": 9.281437125748504,
      "grad_norm": 0.2651910185813904,
      "learning_rate": 0.0001195029398912976,
      "loss": 0.0271,
      "step": 1550
    },
    {
      "epoch": 9.341317365269461,
      "grad_norm": 0.310340017080307,
      "learning_rate": 0.00011852721706571803,
      "loss": 0.0301,
      "step": 1560
    },
    {
      "epoch": 9.401197604790418,
      "grad_norm": 0.23329517245292664,
      "learning_rate": 0.00011754966421368066,
      "loss": 0.0281,
      "step": 1570
    },
    {
      "epoch": 9.461077844311378,
      "grad_norm": 0.21749205887317657,
      "learning_rate": 0.00011657037789299961,
      "loss": 0.0263,
      "step": 1580
    },
    {
      "epoch": 9.520958083832335,
      "grad_norm": 0.4612361490726471,
      "learning_rate": 0.00011558945483271253,
      "loss": 0.0298,
      "step": 1590
    },
    {
      "epoch": 9.580838323353294,
      "grad_norm": 0.19917744398117065,
      "learning_rate": 0.00011460699192352608,
      "loss": 0.0276,
      "step": 1600
    },
    {
      "epoch": 9.640718562874252,
      "grad_norm": 0.22414059937000275,
      "learning_rate": 0.00011362308620824549,
      "loss": 0.0261,
      "step": 1610
    },
    {
      "epoch": 9.70059880239521,
      "grad_norm": 0.3830195963382721,
      "learning_rate": 0.00011263783487218924,
      "loss": 0.025,
      "step": 1620
    },
    {
      "epoch": 9.760479041916168,
      "grad_norm": 0.2302185744047165,
      "learning_rate": 0.00011165133523358955,
      "loss": 0.0197,
      "step": 1630
    },
    {
      "epoch": 9.820359281437126,
      "grad_norm": 0.22359217703342438,
      "learning_rate": 0.00011066368473397977,
      "loss": 0.0263,
      "step": 1640
    },
    {
      "epoch": 9.880239520958083,
      "grad_norm": 0.29255324602127075,
      "learning_rate": 0.0001096749809285695,
      "loss": 0.0252,
      "step": 1650
    },
    {
      "epoch": 9.940119760479043,
      "grad_norm": 0.1755807250738144,
      "learning_rate": 0.0001086853214766088,
      "loss": 0.024,
      "step": 1660
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.3928946554660797,
      "learning_rate": 0.00010769480413174162,
      "loss": 0.0237,
      "step": 1670
    },
    {
      "epoch": 10.059880239520957,
      "grad_norm": 0.2258201539516449,
      "learning_rate": 0.0001067035267323504,
      "loss": 0.0263,
      "step": 1680
    },
    {
      "epoch": 10.119760479041917,
      "grad_norm": 0.24732491374015808,
      "learning_rate": 0.00010571158719189192,
      "loss": 0.0273,
      "step": 1690
    },
    {
      "epoch": 10.179640718562874,
      "grad_norm": 0.2847275733947754,
      "learning_rate": 0.00010471908348922596,
      "loss": 0.0263,
      "step": 1700
    },
    {
      "epoch": 10.239520958083832,
      "grad_norm": 0.1886613517999649,
      "learning_rate": 0.00010372611365893758,
      "loss": 0.0285,
      "step": 1710
    },
    {
      "epoch": 10.29940119760479,
      "grad_norm": 0.3526057302951813,
      "learning_rate": 0.00010273277578165341,
      "loss": 0.0273,
      "step": 1720
    },
    {
      "epoch": 10.359281437125748,
      "grad_norm": 0.3312857449054718,
      "learning_rate": 0.00010173916797435405,
      "loss": 0.0259,
      "step": 1730
    },
    {
      "epoch": 10.419161676646706,
      "grad_norm": 0.3997904360294342,
      "learning_rate": 0.00010074538838068243,
      "loss": 0.0234,
      "step": 1740
    },
    {
      "epoch": 10.479041916167665,
      "grad_norm": 0.29786816239356995,
      "learning_rate": 9.975153516124972e-05,
      "loss": 0.027,
      "step": 1750
    },
    {
      "epoch": 10.538922155688622,
      "grad_norm": 0.37553417682647705,
      "learning_rate": 9.875770648393936e-05,
      "loss": 0.0282,
      "step": 1760
    },
    {
      "epoch": 10.598802395209582,
      "grad_norm": 0.20235197246074677,
      "learning_rate": 9.776400051421078e-05,
      "loss": 0.0211,
      "step": 1770
    },
    {
      "epoch": 10.658682634730539,
      "grad_norm": 0.20965729653835297,
      "learning_rate": 9.677051540540286e-05,
      "loss": 0.0271,
      "step": 1780
    },
    {
      "epoch": 10.718562874251496,
      "grad_norm": 0.29233694076538086,
      "learning_rate": 9.577734928903899e-05,
      "loss": 0.025,
      "step": 1790
    },
    {
      "epoch": 10.778443113772456,
      "grad_norm": 0.258211225271225,
      "learning_rate": 9.47846002651341e-05,
      "loss": 0.0245,
      "step": 1800
    },
    {
      "epoch": 10.838323353293413,
      "grad_norm": 0.19290903210639954,
      "learning_rate": 9.379236639250473e-05,
      "loss": 0.0221,
      "step": 1810
    },
    {
      "epoch": 10.89820359281437,
      "grad_norm": 0.2504674196243286,
      "learning_rate": 9.28007456790833e-05,
      "loss": 0.0232,
      "step": 1820
    },
    {
      "epoch": 10.95808383233533,
      "grad_norm": 0.2591340243816376,
      "learning_rate": 9.180983607223757e-05,
      "loss": 0.0247,
      "step": 1830
    },
    {
      "epoch": 11.017964071856287,
      "grad_norm": 0.3230155408382416,
      "learning_rate": 9.081973544909557e-05,
      "loss": 0.0212,
      "step": 1840
    },
    {
      "epoch": 11.077844311377245,
      "grad_norm": 0.21772272884845734,
      "learning_rate": 8.983054160687801e-05,
      "loss": 0.0249,
      "step": 1850
    },
    {
      "epoch": 11.137724550898204,
      "grad_norm": 0.31186336278915405,
      "learning_rate": 8.88423522532383e-05,
      "loss": 0.0254,
      "step": 1860
    },
    {
      "epoch": 11.197604790419161,
      "grad_norm": 0.20252284407615662,
      "learning_rate": 8.785526499661137e-05,
      "loss": 0.0233,
      "step": 1870
    },
    {
      "epoch": 11.25748502994012,
      "grad_norm": 0.2314423769712448,
      "learning_rate": 8.686937733657264e-05,
      "loss": 0.025,
      "step": 1880
    },
    {
      "epoch": 11.317365269461078,
      "grad_norm": 0.16507509350776672,
      "learning_rate": 8.588478665420717e-05,
      "loss": 0.0253,
      "step": 1890
    },
    {
      "epoch": 11.377245508982035,
      "grad_norm": 0.3258945941925049,
      "learning_rate": 8.490159020249107e-05,
      "loss": 0.0219,
      "step": 1900
    },
    {
      "epoch": 11.437125748502995,
      "grad_norm": 0.20880267024040222,
      "learning_rate": 8.391988509668534e-05,
      "loss": 0.0248,
      "step": 1910
    },
    {
      "epoch": 11.497005988023952,
      "grad_norm": 0.4254096448421478,
      "learning_rate": 8.293976830474312e-05,
      "loss": 0.0272,
      "step": 1920
    },
    {
      "epoch": 11.55688622754491,
      "grad_norm": 0.15054860711097717,
      "learning_rate": 8.196133663773191e-05,
      "loss": 0.0202,
      "step": 1930
    },
    {
      "epoch": 11.616766467065869,
      "grad_norm": 0.18735750019550323,
      "learning_rate": 8.09846867402709e-05,
      "loss": 0.0249,
      "step": 1940
    },
    {
      "epoch": 11.676646706586826,
      "grad_norm": 0.1882217973470688,
      "learning_rate": 8.000991508098492e-05,
      "loss": 0.0204,
      "step": 1950
    },
    {
      "epoch": 11.736526946107784,
      "grad_norm": 0.23175270855426788,
      "learning_rate": 7.903711794297572e-05,
      "loss": 0.0231,
      "step": 1960
    },
    {
      "epoch": 11.796407185628743,
      "grad_norm": 0.18652957677841187,
      "learning_rate": 7.806639141431178e-05,
      "loss": 0.0223,
      "step": 1970
    },
    {
      "epoch": 11.8562874251497,
      "grad_norm": 0.4685545563697815,
      "learning_rate": 7.709783137853692e-05,
      "loss": 0.0246,
      "step": 1980
    },
    {
      "epoch": 11.91616766467066,
      "grad_norm": 0.2099016010761261,
      "learning_rate": 7.613153350519969e-05,
      "loss": 0.0224,
      "step": 1990
    },
    {
      "epoch": 11.976047904191617,
      "grad_norm": 0.23307521641254425,
      "learning_rate": 7.516759324040338e-05,
      "loss": 0.0209,
      "step": 2000
    },
    {
      "epoch": 12.035928143712574,
      "grad_norm": 0.8399657607078552,
      "learning_rate": 7.42061057973785e-05,
      "loss": 0.0215,
      "step": 2010
    },
    {
      "epoch": 12.095808383233534,
      "grad_norm": 0.389076292514801,
      "learning_rate": 7.324716614707793e-05,
      "loss": 0.0227,
      "step": 2020
    },
    {
      "epoch": 12.155688622754491,
      "grad_norm": 0.25040480494499207,
      "learning_rate": 7.229086900879629e-05,
      "loss": 0.0231,
      "step": 2030
    },
    {
      "epoch": 12.215568862275449,
      "grad_norm": 0.2006189376115799,
      "learning_rate": 7.133730884081395e-05,
      "loss": 0.022,
      "step": 2040
    },
    {
      "epoch": 12.275449101796408,
      "grad_norm": 0.24044956266880035,
      "learning_rate": 7.038657983106702e-05,
      "loss": 0.0196,
      "step": 2050
    },
    {
      "epoch": 12.335329341317365,
      "grad_norm": 0.15739458799362183,
      "learning_rate": 6.943877588784382e-05,
      "loss": 0.0232,
      "step": 2060
    },
    {
      "epoch": 12.395209580838323,
      "grad_norm": 0.18294060230255127,
      "learning_rate": 6.849399063050912e-05,
      "loss": 0.0193,
      "step": 2070
    },
    {
      "epoch": 12.455089820359282,
      "grad_norm": 0.17852729558944702,
      "learning_rate": 6.755231738025693e-05,
      "loss": 0.0225,
      "step": 2080
    },
    {
      "epoch": 12.51497005988024,
      "grad_norm": 0.1694345921278,
      "learning_rate": 6.661384915089268e-05,
      "loss": 0.0181,
      "step": 2090
    },
    {
      "epoch": 12.574850299401197,
      "grad_norm": 0.2561972141265869,
      "learning_rate": 6.567867863964561e-05,
      "loss": 0.0197,
      "step": 2100
    },
    {
      "epoch": 12.634730538922156,
      "grad_norm": 0.16482268273830414,
      "learning_rate": 6.474689821801295e-05,
      "loss": 0.0179,
      "step": 2110
    },
    {
      "epoch": 12.694610778443113,
      "grad_norm": 1.6696979999542236,
      "learning_rate": 6.38185999226356e-05,
      "loss": 0.0197,
      "step": 2120
    },
    {
      "epoch": 12.754491017964071,
      "grad_norm": 0.22506001591682434,
      "learning_rate": 6.289387544620736e-05,
      "loss": 0.0218,
      "step": 2130
    },
    {
      "epoch": 12.81437125748503,
      "grad_norm": 0.23911543190479279,
      "learning_rate": 6.197281612841795e-05,
      "loss": 0.0217,
      "step": 2140
    },
    {
      "epoch": 12.874251497005988,
      "grad_norm": 0.20184627175331116,
      "learning_rate": 6.10555129469309e-05,
      "loss": 0.0168,
      "step": 2150
    },
    {
      "epoch": 12.934131736526947,
      "grad_norm": 0.17834153771400452,
      "learning_rate": 6.014205650839737e-05,
      "loss": 0.019,
      "step": 2160
    },
    {
      "epoch": 12.994011976047904,
      "grad_norm": 0.2051975578069687,
      "learning_rate": 5.923253703950617e-05,
      "loss": 0.0192,
      "step": 2170
    },
    {
      "epoch": 13.053892215568862,
      "grad_norm": 0.21173317730426788,
      "learning_rate": 5.832704437807195e-05,
      "loss": 0.0165,
      "step": 2180
    },
    {
      "epoch": 13.113772455089821,
      "grad_norm": 0.4522579610347748,
      "learning_rate": 5.742566796416129e-05,
      "loss": 0.0185,
      "step": 2190
    },
    {
      "epoch": 13.173652694610778,
      "grad_norm": 0.22313633561134338,
      "learning_rate": 5.652849683125823e-05,
      "loss": 0.0172,
      "step": 2200
    },
    {
      "epoch": 13.233532934131736,
      "grad_norm": 0.2275288850069046,
      "learning_rate": 5.563561959747019e-05,
      "loss": 0.0242,
      "step": 2210
    },
    {
      "epoch": 13.293413173652695,
      "grad_norm": 0.15697140991687775,
      "learning_rate": 5.4747124456774445e-05,
      "loss": 0.0207,
      "step": 2220
    },
    {
      "epoch": 13.353293413173652,
      "grad_norm": 0.1576000303030014,
      "learning_rate": 5.38630991703069e-05,
      "loss": 0.0217,
      "step": 2230
    },
    {
      "epoch": 13.41317365269461,
      "grad_norm": 0.42864474654197693,
      "learning_rate": 5.2983631057693484e-05,
      "loss": 0.0201,
      "step": 2240
    },
    {
      "epoch": 13.47305389221557,
      "grad_norm": 0.24114832282066345,
      "learning_rate": 5.210880698842514e-05,
      "loss": 0.0187,
      "step": 2250
    },
    {
      "epoch": 13.532934131736527,
      "grad_norm": 0.2340080291032791,
      "learning_rate": 5.123871337327724e-05,
      "loss": 0.0202,
      "step": 2260
    },
    {
      "epoch": 13.592814371257486,
      "grad_norm": 0.22986619174480438,
      "learning_rate": 5.037343615577448e-05,
      "loss": 0.0211,
      "step": 2270
    },
    {
      "epoch": 13.652694610778443,
      "grad_norm": 0.09167616814374924,
      "learning_rate": 4.951306080370165e-05,
      "loss": 0.0177,
      "step": 2280
    },
    {
      "epoch": 13.7125748502994,
      "grad_norm": 0.1504812389612198,
      "learning_rate": 4.865767230066166e-05,
      "loss": 0.0153,
      "step": 2290
    },
    {
      "epoch": 13.77245508982036,
      "grad_norm": 0.16760465502738953,
      "learning_rate": 4.78073551376812e-05,
      "loss": 0.0177,
      "step": 2300
    },
    {
      "epoch": 13.832335329341317,
      "grad_norm": 0.1385156363248825,
      "learning_rate": 4.6962193304865335e-05,
      "loss": 0.0183,
      "step": 2310
    },
    {
      "epoch": 13.892215568862275,
      "grad_norm": 0.5168209075927734,
      "learning_rate": 4.6122270283100944e-05,
      "loss": 0.0184,
      "step": 2320
    },
    {
      "epoch": 13.952095808383234,
      "grad_norm": 0.25558820366859436,
      "learning_rate": 4.528766903581149e-05,
      "loss": 0.0171,
      "step": 2330
    },
    {
      "epoch": 14.011976047904191,
      "grad_norm": 0.2101954072713852,
      "learning_rate": 4.44584720007618e-05,
      "loss": 0.022,
      "step": 2340
    },
    {
      "epoch": 14.071856287425149,
      "grad_norm": 0.3249596655368805,
      "learning_rate": 4.363476108191552e-05,
      "loss": 0.0176,
      "step": 2350
    },
    {
      "epoch": 14.131736526946108,
      "grad_norm": 0.48448604345321655,
      "learning_rate": 4.2816617641344934e-05,
      "loss": 0.0189,
      "step": 2360
    },
    {
      "epoch": 14.191616766467066,
      "grad_norm": 0.3032410144805908,
      "learning_rate": 4.200412249119463e-05,
      "loss": 0.0176,
      "step": 2370
    },
    {
      "epoch": 14.251497005988025,
      "grad_norm": 0.2361571490764618,
      "learning_rate": 4.119735588569899e-05,
      "loss": 0.0165,
      "step": 2380
    },
    {
      "epoch": 14.311377245508982,
      "grad_norm": 0.2228900045156479,
      "learning_rate": 4.039639751325525e-05,
      "loss": 0.0178,
      "step": 2390
    },
    {
      "epoch": 14.37125748502994,
      "grad_norm": 0.25487351417541504,
      "learning_rate": 3.960132648855226e-05,
      "loss": 0.0191,
      "step": 2400
    },
    {
      "epoch": 14.431137724550899,
      "grad_norm": 0.13682517409324646,
      "learning_rate": 3.881222134475589e-05,
      "loss": 0.0161,
      "step": 2410
    },
    {
      "epoch": 14.491017964071856,
      "grad_norm": 0.29112255573272705,
      "learning_rate": 3.802916002575192e-05,
      "loss": 0.0152,
      "step": 2420
    },
    {
      "epoch": 14.550898203592814,
      "grad_norm": 0.1447286605834961,
      "learning_rate": 3.725221987844721e-05,
      "loss": 0.0208,
      "step": 2430
    },
    {
      "epoch": 14.610778443113773,
      "grad_norm": 0.2142002284526825,
      "learning_rate": 3.6481477645129624e-05,
      "loss": 0.0183,
      "step": 2440
    },
    {
      "epoch": 14.67065868263473,
      "grad_norm": 0.47556108236312866,
      "learning_rate": 3.571700945588789e-05,
      "loss": 0.0179,
      "step": 2450
    },
    {
      "epoch": 14.730538922155688,
      "grad_norm": 0.11879216134548187,
      "learning_rate": 3.495889082109185e-05,
      "loss": 0.0189,
      "step": 2460
    },
    {
      "epoch": 14.790419161676647,
      "grad_norm": 0.18499957025051117,
      "learning_rate": 3.4207196623933857e-05,
      "loss": 0.0171,
      "step": 2470
    },
    {
      "epoch": 14.850299401197605,
      "grad_norm": 0.3223067820072174,
      "learning_rate": 3.346200111303226e-05,
      "loss": 0.0165,
      "step": 2480
    },
    {
      "epoch": 14.910179640718562,
      "grad_norm": 0.2720371186733246,
      "learning_rate": 3.272337789509751e-05,
      "loss": 0.0175,
      "step": 2490
    },
    {
      "epoch": 14.970059880239521,
      "grad_norm": 0.20944693684577942,
      "learning_rate": 3.199139992766155e-05,
      "loss": 0.0195,
      "step": 2500
    },
    {
      "epoch": 15.029940119760479,
      "grad_norm": 0.18262909352779388,
      "learning_rate": 3.1266139511871465e-05,
      "loss": 0.015,
      "step": 2510
    },
    {
      "epoch": 15.089820359281438,
      "grad_norm": 0.09555590897798538,
      "learning_rate": 3.0547668285348105e-05,
      "loss": 0.0163,
      "step": 2520
    },
    {
      "epoch": 15.149700598802395,
      "grad_norm": 0.2287481278181076,
      "learning_rate": 2.9836057215109746e-05,
      "loss": 0.0184,
      "step": 2530
    },
    {
      "epoch": 15.209580838323353,
      "grad_norm": 0.18416175246238708,
      "learning_rate": 2.9131376590562587e-05,
      "loss": 0.0161,
      "step": 2540
    },
    {
      "epoch": 15.269461077844312,
      "grad_norm": 0.18038851022720337,
      "learning_rate": 2.843369601655792e-05,
      "loss": 0.0149,
      "step": 2550
    },
    {
      "epoch": 15.32934131736527,
      "grad_norm": 0.2322443276643753,
      "learning_rate": 2.7743084406516693e-05,
      "loss": 0.0178,
      "step": 2560
    },
    {
      "epoch": 15.389221556886227,
      "grad_norm": 0.1812848597764969,
      "learning_rate": 2.705960997562277e-05,
      "loss": 0.0191,
      "step": 2570
    },
    {
      "epoch": 15.449101796407186,
      "grad_norm": 0.1792038083076477,
      "learning_rate": 2.6383340234084875e-05,
      "loss": 0.0132,
      "step": 2580
    },
    {
      "epoch": 15.508982035928144,
      "grad_norm": 0.14055480062961578,
      "learning_rate": 2.5714341980468437e-05,
      "loss": 0.0146,
      "step": 2590
    },
    {
      "epoch": 15.568862275449101,
      "grad_norm": 0.2022690773010254,
      "learning_rate": 2.5052681295097225e-05,
      "loss": 0.0156,
      "step": 2600
    },
    {
      "epoch": 15.62874251497006,
      "grad_norm": 0.11712924391031265,
      "learning_rate": 2.439842353352667e-05,
      "loss": 0.0174,
      "step": 2610
    },
    {
      "epoch": 15.688622754491018,
      "grad_norm": 0.3119458258152008,
      "learning_rate": 2.3751633320088064e-05,
      "loss": 0.0188,
      "step": 2620
    },
    {
      "epoch": 15.748502994011975,
      "grad_norm": 0.31999191641807556,
      "learning_rate": 2.311237454150539e-05,
      "loss": 0.0149,
      "step": 2630
    },
    {
      "epoch": 15.808383233532934,
      "grad_norm": 0.21171946823596954,
      "learning_rate": 2.248071034058492e-05,
      "loss": 0.0166,
      "step": 2640
    },
    {
      "epoch": 15.868263473053892,
      "grad_norm": 0.15118949115276337,
      "learning_rate": 2.185670310997835e-05,
      "loss": 0.0195,
      "step": 2650
    },
    {
      "epoch": 15.928143712574851,
      "grad_norm": 0.16680358350276947,
      "learning_rate": 2.124041448601979e-05,
      "loss": 0.0135,
      "step": 2660
    },
    {
      "epoch": 15.988023952095809,
      "grad_norm": 0.16885513067245483,
      "learning_rate": 2.0631905342637782e-05,
      "loss": 0.0154,
      "step": 2670
    },
    {
      "epoch": 16.047904191616766,
      "grad_norm": 0.2569000720977783,
      "learning_rate": 2.0031235785342407e-05,
      "loss": 0.0168,
      "step": 2680
    },
    {
      "epoch": 16.107784431137723,
      "grad_norm": 0.3208532929420471,
      "learning_rate": 1.9438465145288374e-05,
      "loss": 0.0154,
      "step": 2690
    },
    {
      "epoch": 16.16766467065868,
      "grad_norm": 0.1462489366531372,
      "learning_rate": 1.885365197341461e-05,
      "loss": 0.0157,
      "step": 2700
    },
    {
      "epoch": 16.227544910179642,
      "grad_norm": 0.1316727101802826,
      "learning_rate": 1.827685403466087e-05,
      "loss": 0.0138,
      "step": 2710
    },
    {
      "epoch": 16.2874251497006,
      "grad_norm": 0.17914234101772308,
      "learning_rate": 1.7708128302261994e-05,
      "loss": 0.0147,
      "step": 2720
    },
    {
      "epoch": 16.347305389221557,
      "grad_norm": 0.18030117452144623,
      "learning_rate": 1.7147530952120337e-05,
      "loss": 0.0149,
      "step": 2730
    },
    {
      "epoch": 16.407185628742514,
      "grad_norm": 0.18496276438236237,
      "learning_rate": 1.659511735725716e-05,
      "loss": 0.016,
      "step": 2740
    },
    {
      "epoch": 16.46706586826347,
      "grad_norm": 0.20014838874340057,
      "learning_rate": 1.605094208234288e-05,
      "loss": 0.0119,
      "step": 2750
    },
    {
      "epoch": 16.526946107784433,
      "grad_norm": 0.2455081343650818,
      "learning_rate": 1.5515058878307665e-05,
      "loss": 0.0129,
      "step": 2760
    },
    {
      "epoch": 16.58682634730539,
      "grad_norm": 0.40518850088119507,
      "learning_rate": 1.4987520677032174e-05,
      "loss": 0.0145,
      "step": 2770
    },
    {
      "epoch": 16.646706586826348,
      "grad_norm": 0.2330215722322464,
      "learning_rate": 1.4468379586119074e-05,
      "loss": 0.0178,
      "step": 2780
    },
    {
      "epoch": 16.706586826347305,
      "grad_norm": 0.29919859766960144,
      "learning_rate": 1.3957686883746191e-05,
      "loss": 0.0129,
      "step": 2790
    },
    {
      "epoch": 16.766467065868262,
      "grad_norm": 0.1759548783302307,
      "learning_rate": 1.3455493013601561e-05,
      "loss": 0.0137,
      "step": 2800
    },
    {
      "epoch": 16.82634730538922,
      "grad_norm": 0.21709783375263214,
      "learning_rate": 1.2961847579900777e-05,
      "loss": 0.0154,
      "step": 2810
    },
    {
      "epoch": 16.88622754491018,
      "grad_norm": 0.12408145517110825,
      "learning_rate": 1.2476799342487255e-05,
      "loss": 0.0147,
      "step": 2820
    },
    {
      "epoch": 16.94610778443114,
      "grad_norm": 0.17193816602230072,
      "learning_rate": 1.2000396212016218e-05,
      "loss": 0.0161,
      "step": 2830
    },
    {
      "epoch": 17.005988023952096,
      "grad_norm": 0.2605287730693817,
      "learning_rate": 1.1532685245222075e-05,
      "loss": 0.0142,
      "step": 2840
    },
    {
      "epoch": 17.065868263473053,
      "grad_norm": 0.11747428774833679,
      "learning_rate": 1.1073712640270506e-05,
      "loss": 0.0131,
      "step": 2850
    },
    {
      "epoch": 17.12574850299401,
      "grad_norm": 0.3511081337928772,
      "learning_rate": 1.0623523732195284e-05,
      "loss": 0.0147,
      "step": 2860
    },
    {
      "epoch": 17.18562874251497,
      "grad_norm": 0.3589492440223694,
      "learning_rate": 1.0182162988420163e-05,
      "loss": 0.0166,
      "step": 2870
    },
    {
      "epoch": 17.24550898203593,
      "grad_norm": 0.14912162721157074,
      "learning_rate": 9.749674004366727e-06,
      "loss": 0.0148,
      "step": 2880
    },
    {
      "epoch": 17.305389221556887,
      "grad_norm": 0.4481550455093384,
      "learning_rate": 9.326099499148177e-06,
      "loss": 0.0126,
      "step": 2890
    },
    {
      "epoch": 17.365269461077844,
      "grad_norm": 0.3854505121707916,
      "learning_rate": 8.911481311349834e-06,
      "loss": 0.0149,
      "step": 2900
    },
    {
      "epoch": 17.4251497005988,
      "grad_norm": 0.11779692023992538,
      "learning_rate": 8.505860394896414e-06,
      "loss": 0.0133,
      "step": 2910
    },
    {
      "epoch": 17.48502994011976,
      "grad_norm": 0.16993847489356995,
      "learning_rate": 8.109276815006938e-06,
      "loss": 0.0116,
      "step": 2920
    },
    {
      "epoch": 17.54491017964072,
      "grad_norm": 0.3170551657676697,
      "learning_rate": 7.721769744237184e-06,
      "loss": 0.0159,
      "step": 2930
    },
    {
      "epoch": 17.604790419161677,
      "grad_norm": 0.5289223194122314,
      "learning_rate": 7.343377458610446e-06,
      "loss": 0.0171,
      "step": 2940
    },
    {
      "epoch": 17.664670658682635,
      "grad_norm": 0.2612517476081848,
      "learning_rate": 6.9741373338368165e-06,
      "loss": 0.0161,
      "step": 2950
    },
    {
      "epoch": 17.724550898203592,
      "grad_norm": 0.1158524826169014,
      "learning_rate": 6.614085841621442e-06,
      "loss": 0.0135,
      "step": 2960
    },
    {
      "epoch": 17.78443113772455,
      "grad_norm": 0.3021811246871948,
      "learning_rate": 6.263258546061978e-06,
      "loss": 0.0161,
      "step": 2970
    },
    {
      "epoch": 17.84431137724551,
      "grad_norm": 0.4671183228492737,
      "learning_rate": 5.921690100135713e-06,
      "loss": 0.0111,
      "step": 2980
    },
    {
      "epoch": 17.904191616766468,
      "grad_norm": 0.12684480845928192,
      "learning_rate": 5.589414242276869e-06,
      "loss": 0.0118,
      "step": 2990
    },
    {
      "epoch": 17.964071856287426,
      "grad_norm": 0.18783032894134521,
      "learning_rate": 5.266463793043896e-06,
      "loss": 0.0108,
      "step": 3000
    },
    {
      "epoch": 18.023952095808383,
      "grad_norm": 0.3168010413646698,
      "learning_rate": 4.952870651877739e-06,
      "loss": 0.0149,
      "step": 3010
    },
    {
      "epoch": 18.08383233532934,
      "grad_norm": 1.2831743955612183,
      "learning_rate": 4.648665793950968e-06,
      "loss": 0.0119,
      "step": 3020
    },
    {
      "epoch": 18.143712574850298,
      "grad_norm": 0.11012054234743118,
      "learning_rate": 4.353879267108119e-06,
      "loss": 0.0141,
      "step": 3030
    },
    {
      "epoch": 18.20359281437126,
      "grad_norm": 0.41468945145606995,
      "learning_rate": 4.068540188897796e-06,
      "loss": 0.0118,
      "step": 3040
    },
    {
      "epoch": 18.263473053892216,
      "grad_norm": 0.16896358132362366,
      "learning_rate": 3.792676743696588e-06,
      "loss": 0.0139,
      "step": 3050
    },
    {
      "epoch": 18.323353293413174,
      "grad_norm": 0.21275334060192108,
      "learning_rate": 3.5263161799251154e-06,
      "loss": 0.0118,
      "step": 3060
    },
    {
      "epoch": 18.38323353293413,
      "grad_norm": 0.15216277539730072,
      "learning_rate": 3.2694848073565596e-06,
      "loss": 0.0117,
      "step": 3070
    },
    {
      "epoch": 18.44311377245509,
      "grad_norm": 0.15968835353851318,
      "learning_rate": 3.0222079945179895e-06,
      "loss": 0.0144,
      "step": 3080
    },
    {
      "epoch": 18.50299401197605,
      "grad_norm": 0.2130011022090912,
      "learning_rate": 2.7845101661844618e-06,
      "loss": 0.0114,
      "step": 3090
    },
    {
      "epoch": 18.562874251497007,
      "grad_norm": 0.14197607338428497,
      "learning_rate": 2.5564148009665778e-06,
      "loss": 0.0142,
      "step": 3100
    },
    {
      "epoch": 18.622754491017965,
      "grad_norm": 0.7464102506637573,
      "learning_rate": 2.3379444289913342e-06,
      "loss": 0.0135,
      "step": 3110
    },
    {
      "epoch": 18.682634730538922,
      "grad_norm": 0.21782080829143524,
      "learning_rate": 2.1291206296767063e-06,
      "loss": 0.0146,
      "step": 3120
    },
    {
      "epoch": 18.74251497005988,
      "grad_norm": 0.21821217238903046,
      "learning_rate": 1.9299640296001288e-06,
      "loss": 0.0112,
      "step": 3130
    },
    {
      "epoch": 18.802395209580837,
      "grad_norm": 0.2201797068119049,
      "learning_rate": 1.7404943004611596e-06,
      "loss": 0.0155,
      "step": 3140
    },
    {
      "epoch": 18.862275449101798,
      "grad_norm": 0.4531762897968292,
      "learning_rate": 1.5607301571383459e-06,
      "loss": 0.0129,
      "step": 3150
    },
    {
      "epoch": 18.922155688622755,
      "grad_norm": 0.2320559322834015,
      "learning_rate": 1.390689355840691e-06,
      "loss": 0.0152,
      "step": 3160
    },
    {
      "epoch": 18.982035928143713,
      "grad_norm": 0.12046198546886444,
      "learning_rate": 1.2303886923537677e-06,
      "loss": 0.0129,
      "step": 3170
    },
    {
      "epoch": 19.04191616766467,
      "grad_norm": 0.23176096379756927,
      "learning_rate": 1.0798440003807474e-06,
      "loss": 0.013,
      "step": 3180
    },
    {
      "epoch": 19.101796407185628,
      "grad_norm": 0.15658603608608246,
      "learning_rate": 9.390701499783827e-07,
      "loss": 0.0102,
      "step": 3190
    },
    {
      "epoch": 19.161676646706585,
      "grad_norm": 0.12516404688358307,
      "learning_rate": 8.080810460882493e-07,
      "loss": 0.0117,
      "step": 3200
    },
    {
      "epoch": 19.221556886227546,
      "grad_norm": 0.1989312320947647,
      "learning_rate": 6.868896271632785e-07,
      "loss": 0.0121,
      "step": 3210
    },
    {
      "epoch": 19.281437125748504,
      "grad_norm": 0.15833425521850586,
      "learning_rate": 5.755078638897459e-07,
      "loss": 0.0145,
      "step": 3220
    },
    {
      "epoch": 19.34131736526946,
      "grad_norm": 0.13037511706352234,
      "learning_rate": 4.739467580048618e-07,
      "loss": 0.0106,
      "step": 3230
    },
    {
      "epoch": 19.40119760479042,
      "grad_norm": 0.3514297306537628,
      "learning_rate": 3.8221634121013983e-07,
      "loss": 0.0104,
      "step": 3240
    },
    {
      "epoch": 19.461077844311376,
      "grad_norm": 0.1234828531742096,
      "learning_rate": 3.003256741803906e-07,
      "loss": 0.0137,
      "step": 3250
    },
    {
      "epoch": 19.520958083832337,
      "grad_norm": 0.1495758444070816,
      "learning_rate": 2.2828284566890345e-07,
      "loss": 0.013,
      "step": 3260
    },
    {
      "epoch": 19.580838323353294,
      "grad_norm": 0.5322373509407043,
      "learning_rate": 1.6609497170834155e-07,
      "loss": 0.0136,
      "step": 3270
    },
    {
      "epoch": 19.64071856287425,
      "grad_norm": 0.15904094278812408,
      "learning_rate": 1.1376819490795942e-07,
      "loss": 0.0151,
      "step": 3280
    },
    {
      "epoch": 19.70059880239521,
      "grad_norm": 0.43996843695640564,
      "learning_rate": 7.130768384679965e-08,
      "loss": 0.0125,
      "step": 3290
    },
    {
      "epoch": 19.760479041916167,
      "grad_norm": 0.13913209736347198,
      "learning_rate": 3.8717632563201046e-08,
      "loss": 0.0188,
      "step": 3300
    },
    {
      "epoch": 19.820359281437124,
      "grad_norm": 0.6390590667724609,
      "learning_rate": 1.600126014051906e-08,
      "loss": 0.0132,
      "step": 3310
    },
    {
      "epoch": 19.880239520958085,
      "grad_norm": 0.11968772113323212,
      "learning_rate": 3.1608103891689335e-09,
      "loss": 0.0118,
      "step": 3320
    },
    {
      "epoch": 19.92814371257485,
      "step": 3328,
      "total_flos": 4.678773863650698e+17,
      "train_loss": 0.03960529512663085,
      "train_runtime": 3740.453,
      "train_samples_per_second": 56.943,
      "train_steps_per_second": 0.89
    }
  ],
  "logging_steps": 10,
  "max_steps": 3328,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.678773863650698e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}