{
  "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": 4.727615833282471,
      "learning_rate": 1.1976047904191617e-05,
      "loss": 1.062,
      "step": 10
    },
    {
      "epoch": 0.11976047904191617,
      "grad_norm": 2.2185091972351074,
      "learning_rate": 2.3952095808383234e-05,
      "loss": 0.6786,
      "step": 20
    },
    {
      "epoch": 0.17964071856287425,
      "grad_norm": 1.7475155591964722,
      "learning_rate": 3.592814371257485e-05,
      "loss": 0.3714,
      "step": 30
    },
    {
      "epoch": 0.23952095808383234,
      "grad_norm": 0.7705522179603577,
      "learning_rate": 4.790419161676647e-05,
      "loss": 0.2662,
      "step": 40
    },
    {
      "epoch": 0.2994011976047904,
      "grad_norm": 0.5850515365600586,
      "learning_rate": 5.988023952095808e-05,
      "loss": 0.2129,
      "step": 50
    },
    {
      "epoch": 0.3592814371257485,
      "grad_norm": 0.5091089010238647,
      "learning_rate": 7.18562874251497e-05,
      "loss": 0.1831,
      "step": 60
    },
    {
      "epoch": 0.41916167664670656,
      "grad_norm": 0.8252758383750916,
      "learning_rate": 8.383233532934131e-05,
      "loss": 0.1614,
      "step": 70
    },
    {
      "epoch": 0.47904191616766467,
      "grad_norm": 0.8221730589866638,
      "learning_rate": 9.580838323353294e-05,
      "loss": 0.1408,
      "step": 80
    },
    {
      "epoch": 0.5389221556886228,
      "grad_norm": 0.6895884871482849,
      "learning_rate": 0.00010778443113772456,
      "loss": 0.1253,
      "step": 90
    },
    {
      "epoch": 0.5988023952095808,
      "grad_norm": 1.0663549900054932,
      "learning_rate": 0.00011976047904191617,
      "loss": 0.1049,
      "step": 100
    },
    {
      "epoch": 0.6586826347305389,
      "grad_norm": 0.4257476031780243,
      "learning_rate": 0.0001317365269461078,
      "loss": 0.1017,
      "step": 110
    },
    {
      "epoch": 0.718562874251497,
      "grad_norm": 0.46928706765174866,
      "learning_rate": 0.0001437125748502994,
      "loss": 0.0887,
      "step": 120
    },
    {
      "epoch": 0.7784431137724551,
      "grad_norm": 0.40188413858413696,
      "learning_rate": 0.00015568862275449103,
      "loss": 0.0821,
      "step": 130
    },
    {
      "epoch": 0.8383233532934131,
      "grad_norm": 0.44656357169151306,
      "learning_rate": 0.00016766467065868263,
      "loss": 0.0802,
      "step": 140
    },
    {
      "epoch": 0.8982035928143712,
      "grad_norm": 0.8075553178787231,
      "learning_rate": 0.00017964071856287425,
      "loss": 0.0768,
      "step": 150
    },
    {
      "epoch": 0.9580838323353293,
      "grad_norm": 0.7961329817771912,
      "learning_rate": 0.00019161676646706587,
      "loss": 0.0788,
      "step": 160
    },
    {
      "epoch": 1.0179640718562875,
      "grad_norm": 0.5878146886825562,
      "learning_rate": 0.00019999955550902674,
      "loss": 0.0751,
      "step": 170
    },
    {
      "epoch": 1.0778443113772456,
      "grad_norm": 0.4572954773902893,
      "learning_rate": 0.0001999916535572039,
      "loss": 0.0684,
      "step": 180
    },
    {
      "epoch": 1.1377245508982037,
      "grad_norm": 0.3957609236240387,
      "learning_rate": 0.00019997387492660495,
      "loss": 0.0651,
      "step": 190
    },
    {
      "epoch": 1.1976047904191618,
      "grad_norm": 0.3396591544151306,
      "learning_rate": 0.00019994622137331462,
      "loss": 0.0655,
      "step": 200
    },
    {
      "epoch": 1.2574850299401197,
      "grad_norm": 0.4531398415565491,
      "learning_rate": 0.00019990869562881356,
      "loss": 0.0648,
      "step": 210
    },
    {
      "epoch": 1.3173652694610778,
      "grad_norm": 0.36353209614753723,
      "learning_rate": 0.00019986130139970835,
      "loss": 0.0619,
      "step": 220
    },
    {
      "epoch": 1.377245508982036,
      "grad_norm": 0.392230361700058,
      "learning_rate": 0.00019980404336736554,
      "loss": 0.0583,
      "step": 230
    },
    {
      "epoch": 1.437125748502994,
      "grad_norm": 0.34288108348846436,
      "learning_rate": 0.00019973692718744902,
      "loss": 0.0599,
      "step": 240
    },
    {
      "epoch": 1.4970059880239521,
      "grad_norm": 0.40226370096206665,
      "learning_rate": 0.0001996599594893617,
      "loss": 0.0538,
      "step": 250
    },
    {
      "epoch": 1.55688622754491,
      "grad_norm": 0.26641878485679626,
      "learning_rate": 0.00019957314787559036,
      "loss": 0.0557,
      "step": 260
    },
    {
      "epoch": 1.6167664670658684,
      "grad_norm": 0.4017570912837982,
      "learning_rate": 0.00019947650092095494,
      "loss": 0.0514,
      "step": 270
    },
    {
      "epoch": 1.6766467065868262,
      "grad_norm": 0.2945254445075989,
      "learning_rate": 0.00019937002817176146,
      "loss": 0.0506,
      "step": 280
    },
    {
      "epoch": 1.7365269461077846,
      "grad_norm": 0.2502083480358124,
      "learning_rate": 0.00019925374014485912,
      "loss": 0.0518,
      "step": 290
    },
    {
      "epoch": 1.7964071856287425,
      "grad_norm": 0.3017183840274811,
      "learning_rate": 0.00019912764832660156,
      "loss": 0.0541,
      "step": 300
    },
    {
      "epoch": 1.8562874251497006,
      "grad_norm": 0.23154252767562866,
      "learning_rate": 0.00019899176517171208,
      "loss": 0.0501,
      "step": 310
    },
    {
      "epoch": 1.9161676646706587,
      "grad_norm": 0.2425275295972824,
      "learning_rate": 0.0001988461041020537,
      "loss": 0.0437,
      "step": 320
    },
    {
      "epoch": 1.9760479041916168,
      "grad_norm": 0.2978982925415039,
      "learning_rate": 0.0001986906795053032,
      "loss": 0.045,
      "step": 330
    },
    {
      "epoch": 2.035928143712575,
      "grad_norm": 0.3514959514141083,
      "learning_rate": 0.00019852550673353004,
      "loss": 0.0492,
      "step": 340
    },
    {
      "epoch": 2.095808383233533,
      "grad_norm": 0.488469660282135,
      "learning_rate": 0.00019835060210168017,
      "loss": 0.0479,
      "step": 350
    },
    {
      "epoch": 2.155688622754491,
      "grad_norm": 0.26303964853286743,
      "learning_rate": 0.00019816598288596412,
      "loss": 0.0475,
      "step": 360
    },
    {
      "epoch": 2.215568862275449,
      "grad_norm": 0.25606536865234375,
      "learning_rate": 0.00019797166732215076,
      "loss": 0.0413,
      "step": 370
    },
    {
      "epoch": 2.2754491017964074,
      "grad_norm": 0.291185200214386,
      "learning_rate": 0.0001977676746037662,
      "loss": 0.0443,
      "step": 380
    },
    {
      "epoch": 2.3353293413173652,
      "grad_norm": 0.314779669046402,
      "learning_rate": 0.0001975540248801977,
      "loss": 0.044,
      "step": 390
    },
    {
      "epoch": 2.3952095808383236,
      "grad_norm": 0.259378582239151,
      "learning_rate": 0.00019733073925470352,
      "loss": 0.0413,
      "step": 400
    },
    {
      "epoch": 2.4550898203592815,
      "grad_norm": 0.5142641067504883,
      "learning_rate": 0.00019709783978232842,
      "loss": 0.0453,
      "step": 410
    },
    {
      "epoch": 2.5149700598802394,
      "grad_norm": 0.33504176139831543,
      "learning_rate": 0.00019685534946772536,
      "loss": 0.0438,
      "step": 420
    },
    {
      "epoch": 2.5748502994011977,
      "grad_norm": 0.33974334597587585,
      "learning_rate": 0.00019660329226288273,
      "loss": 0.0436,
      "step": 430
    },
    {
      "epoch": 2.6347305389221556,
      "grad_norm": 0.29948851466178894,
      "learning_rate": 0.00019634169306475917,
      "loss": 0.0404,
      "step": 440
    },
    {
      "epoch": 2.694610778443114,
      "grad_norm": 0.2677200436592102,
      "learning_rate": 0.00019607057771282372,
      "loss": 0.0423,
      "step": 450
    },
    {
      "epoch": 2.754491017964072,
      "grad_norm": 0.2578097879886627,
      "learning_rate": 0.00019578997298650395,
      "loss": 0.0385,
      "step": 460
    },
    {
      "epoch": 2.81437125748503,
      "grad_norm": 0.2699326276779175,
      "learning_rate": 0.00019549990660254073,
      "loss": 0.0402,
      "step": 470
    },
    {
      "epoch": 2.874251497005988,
      "grad_norm": 0.22369785606861115,
      "learning_rate": 0.00019520040721225033,
      "loss": 0.0402,
      "step": 480
    },
    {
      "epoch": 2.934131736526946,
      "grad_norm": 0.23244518041610718,
      "learning_rate": 0.00019489150439869465,
      "loss": 0.0393,
      "step": 490
    },
    {
      "epoch": 2.9940119760479043,
      "grad_norm": 0.2809353768825531,
      "learning_rate": 0.000194573228673759,
      "loss": 0.0402,
      "step": 500
    },
    {
      "epoch": 3.053892215568862,
      "grad_norm": 0.1485118865966797,
      "learning_rate": 0.0001942456114751382,
      "loss": 0.0364,
      "step": 510
    },
    {
      "epoch": 3.1137724550898205,
      "grad_norm": 0.35376912355422974,
      "learning_rate": 0.00019390868516323156,
      "loss": 0.0358,
      "step": 520
    },
    {
      "epoch": 3.1736526946107784,
      "grad_norm": 0.24453511834144592,
      "learning_rate": 0.00019356248301794632,
      "loss": 0.0366,
      "step": 530
    },
    {
      "epoch": 3.2335329341317367,
      "grad_norm": 0.2331007570028305,
      "learning_rate": 0.00019320703923541045,
      "loss": 0.0387,
      "step": 540
    },
    {
      "epoch": 3.2934131736526946,
      "grad_norm": 0.2162063866853714,
      "learning_rate": 0.00019284238892459485,
      "loss": 0.0373,
      "step": 550
    },
    {
      "epoch": 3.3532934131736525,
      "grad_norm": 0.22156478464603424,
      "learning_rate": 0.00019246856810384576,
      "loss": 0.042,
      "step": 560
    },
    {
      "epoch": 3.413173652694611,
      "grad_norm": 0.31770089268684387,
      "learning_rate": 0.0001920856136973266,
      "loss": 0.0421,
      "step": 570
    },
    {
      "epoch": 3.4730538922155687,
      "grad_norm": 0.37347620725631714,
      "learning_rate": 0.0001916935635313711,
      "loss": 0.0354,
      "step": 580
    },
    {
      "epoch": 3.532934131736527,
      "grad_norm": 0.2527390420436859,
      "learning_rate": 0.00019129245633074703,
      "loss": 0.0378,
      "step": 590
    },
    {
      "epoch": 3.592814371257485,
      "grad_norm": 0.2797776162624359,
      "learning_rate": 0.0001908823317148309,
      "loss": 0.039,
      "step": 600
    },
    {
      "epoch": 3.6526946107784433,
      "grad_norm": 0.3039669692516327,
      "learning_rate": 0.00019046323019369474,
      "loss": 0.0376,
      "step": 610
    },
    {
      "epoch": 3.712574850299401,
      "grad_norm": 0.3194757103919983,
      "learning_rate": 0.0001900351931641047,
      "loss": 0.0364,
      "step": 620
    },
    {
      "epoch": 3.772455089820359,
      "grad_norm": 0.29081788659095764,
      "learning_rate": 0.00018959826290543207,
      "loss": 0.0387,
      "step": 630
    },
    {
      "epoch": 3.8323353293413174,
      "grad_norm": 0.2967931628227234,
      "learning_rate": 0.00018915248257547713,
      "loss": 0.0363,
      "step": 640
    },
    {
      "epoch": 3.8922155688622757,
      "grad_norm": 0.29823440313339233,
      "learning_rate": 0.0001886978962062062,
      "loss": 0.0413,
      "step": 650
    },
    {
      "epoch": 3.9520958083832336,
      "grad_norm": 0.2680460810661316,
      "learning_rate": 0.00018823454869940242,
      "loss": 0.0381,
      "step": 660
    },
    {
      "epoch": 4.0119760479041915,
      "grad_norm": 0.18794329464435577,
      "learning_rate": 0.00018776248582223063,
      "loss": 0.035,
      "step": 670
    },
    {
      "epoch": 4.07185628742515,
      "grad_norm": 0.31890541315078735,
      "learning_rate": 0.00018728175420271658,
      "loss": 0.0309,
      "step": 680
    },
    {
      "epoch": 4.131736526946108,
      "grad_norm": 0.2308373749256134,
      "learning_rate": 0.00018679240132514136,
      "loss": 0.0338,
      "step": 690
    },
    {
      "epoch": 4.191616766467066,
      "grad_norm": 0.21351340413093567,
      "learning_rate": 0.0001862944755253511,
      "loss": 0.0319,
      "step": 700
    },
    {
      "epoch": 4.251497005988024,
      "grad_norm": 0.22991211712360382,
      "learning_rate": 0.00018578802598598256,
      "loss": 0.0339,
      "step": 710
    },
    {
      "epoch": 4.311377245508982,
      "grad_norm": 0.1862100064754486,
      "learning_rate": 0.00018527310273160522,
      "loss": 0.0323,
      "step": 720
    },
    {
      "epoch": 4.37125748502994,
      "grad_norm": 0.34272634983062744,
      "learning_rate": 0.00018474975662377996,
      "loss": 0.0324,
      "step": 730
    },
    {
      "epoch": 4.431137724550898,
      "grad_norm": 0.3091452419757843,
      "learning_rate": 0.00018421803935603532,
      "loss": 0.0338,
      "step": 740
    },
    {
      "epoch": 4.491017964071856,
      "grad_norm": 0.31789442896842957,
      "learning_rate": 0.0001836780034487614,
      "loss": 0.0343,
      "step": 750
    },
    {
      "epoch": 4.550898203592815,
      "grad_norm": 0.46529826521873474,
      "learning_rate": 0.00018312970224402224,
      "loss": 0.0359,
      "step": 760
    },
    {
      "epoch": 4.610778443113772,
      "grad_norm": 0.3199298679828644,
      "learning_rate": 0.00018257318990028682,
      "loss": 0.0375,
      "step": 770
    },
    {
      "epoch": 4.6706586826347305,
      "grad_norm": 0.30249571800231934,
      "learning_rate": 0.00018200852138707975,
      "loss": 0.039,
      "step": 780
    },
    {
      "epoch": 4.730538922155689,
      "grad_norm": 0.18588168919086456,
      "learning_rate": 0.00018143575247955135,
      "loss": 0.0338,
      "step": 790
    },
    {
      "epoch": 4.790419161676647,
      "grad_norm": 0.2602488100528717,
      "learning_rate": 0.0001808549397529689,
      "loss": 0.034,
      "step": 800
    },
    {
      "epoch": 4.850299401197605,
      "grad_norm": 0.3416588306427002,
      "learning_rate": 0.00018026614057712795,
      "loss": 0.0321,
      "step": 810
    },
    {
      "epoch": 4.910179640718563,
      "grad_norm": 0.31246596574783325,
      "learning_rate": 0.00017966941311068594,
      "loss": 0.0313,
      "step": 820
    },
    {
      "epoch": 4.970059880239521,
      "grad_norm": 0.20728591084480286,
      "learning_rate": 0.00017906481629541743,
      "loss": 0.0315,
      "step": 830
    },
    {
      "epoch": 5.029940119760479,
      "grad_norm": 0.2376144826412201,
      "learning_rate": 0.00017845240985039218,
      "loss": 0.0301,
      "step": 840
    },
    {
      "epoch": 5.089820359281437,
      "grad_norm": 0.2364414483308792,
      "learning_rate": 0.0001778322542660764,
      "loss": 0.0299,
      "step": 850
    },
    {
      "epoch": 5.149700598802395,
      "grad_norm": 0.22445429861545563,
      "learning_rate": 0.0001772044107983577,
      "loss": 0.0296,
      "step": 860
    },
    {
      "epoch": 5.209580838323353,
      "grad_norm": 0.2109883427619934,
      "learning_rate": 0.0001765689414624947,
      "loss": 0.0326,
      "step": 870
    },
    {
      "epoch": 5.269461077844311,
      "grad_norm": 0.17161297798156738,
      "learning_rate": 0.00017592590902699138,
      "loss": 0.0321,
      "step": 880
    },
    {
      "epoch": 5.3293413173652695,
      "grad_norm": 0.18407276272773743,
      "learning_rate": 0.00017527537700739708,
      "loss": 0.0339,
      "step": 890
    },
    {
      "epoch": 5.389221556886228,
      "grad_norm": 0.1970146745443344,
      "learning_rate": 0.00017461740966003282,
      "loss": 0.0356,
      "step": 900
    },
    {
      "epoch": 5.449101796407185,
      "grad_norm": 0.1647009700536728,
      "learning_rate": 0.00017395207197564442,
      "loss": 0.0303,
      "step": 910
    },
    {
      "epoch": 5.508982035928144,
      "grad_norm": 0.17965902388095856,
      "learning_rate": 0.00017327942967298285,
      "loss": 0.0273,
      "step": 920
    },
    {
      "epoch": 5.568862275449102,
      "grad_norm": 0.2345510870218277,
      "learning_rate": 0.0001725995491923131,
      "loss": 0.0274,
      "step": 930
    },
    {
      "epoch": 5.62874251497006,
      "grad_norm": 0.33345136046409607,
      "learning_rate": 0.00017191249768885137,
      "loss": 0.0341,
      "step": 940
    },
    {
      "epoch": 5.688622754491018,
      "grad_norm": 0.1947910636663437,
      "learning_rate": 0.00017121834302613186,
      "loss": 0.029,
      "step": 950
    },
    {
      "epoch": 5.748502994011976,
      "grad_norm": 0.17892029881477356,
      "learning_rate": 0.00017051715376930363,
      "loss": 0.0343,
      "step": 960
    },
    {
      "epoch": 5.808383233532934,
      "grad_norm": 0.174368754029274,
      "learning_rate": 0.00016980899917835783,
      "loss": 0.0299,
      "step": 970
    },
    {
      "epoch": 5.868263473053892,
      "grad_norm": 0.24912439286708832,
      "learning_rate": 0.00016909394920128695,
      "loss": 0.0295,
      "step": 980
    },
    {
      "epoch": 5.92814371257485,
      "grad_norm": 0.2845669984817505,
      "learning_rate": 0.00016837207446717518,
      "loss": 0.0289,
      "step": 990
    },
    {
      "epoch": 5.9880239520958085,
      "grad_norm": 0.22086086869239807,
      "learning_rate": 0.00016764344627922254,
      "loss": 0.0327,
      "step": 1000
    },
    {
      "epoch": 6.047904191616767,
      "grad_norm": 0.3092862665653229,
      "learning_rate": 0.0001669081366077016,
      "loss": 0.026,
      "step": 1010
    },
    {
      "epoch": 6.107784431137724,
      "grad_norm": 0.3457212746143341,
      "learning_rate": 0.00016616621808284864,
      "loss": 0.0291,
      "step": 1020
    },
    {
      "epoch": 6.167664670658683,
      "grad_norm": 0.19416464865207672,
      "learning_rate": 0.00016541776398768966,
      "loss": 0.0297,
      "step": 1030
    },
    {
      "epoch": 6.227544910179641,
      "grad_norm": 0.24748080968856812,
      "learning_rate": 0.00016466284825080184,
      "loss": 0.028,
      "step": 1040
    },
    {
      "epoch": 6.287425149700598,
      "grad_norm": 0.20852451026439667,
      "learning_rate": 0.00016390154543901122,
      "loss": 0.0237,
      "step": 1050
    },
    {
      "epoch": 6.347305389221557,
      "grad_norm": 0.2635754346847534,
      "learning_rate": 0.00016313393075002736,
      "loss": 0.0267,
      "step": 1060
    },
    {
      "epoch": 6.407185628742515,
      "grad_norm": 0.28253108263015747,
      "learning_rate": 0.00016236008000501573,
      "loss": 0.0257,
      "step": 1070
    },
    {
      "epoch": 6.467065868263473,
      "grad_norm": 0.22715069353580475,
      "learning_rate": 0.00016158006964110837,
      "loss": 0.0308,
      "step": 1080
    },
    {
      "epoch": 6.526946107784431,
      "grad_norm": 0.2551535964012146,
      "learning_rate": 0.0001607939767038539,
      "loss": 0.0265,
      "step": 1090
    },
    {
      "epoch": 6.586826347305389,
      "grad_norm": 0.1504376381635666,
      "learning_rate": 0.00016000187883960734,
      "loss": 0.0264,
      "step": 1100
    },
    {
      "epoch": 6.6467065868263475,
      "grad_norm": 0.3545721173286438,
      "learning_rate": 0.00015920385428786057,
      "loss": 0.026,
      "step": 1110
    },
    {
      "epoch": 6.706586826347305,
      "grad_norm": 0.16215580701828003,
      "learning_rate": 0.0001583999818735142,
      "loss": 0.024,
      "step": 1120
    },
    {
      "epoch": 6.766467065868263,
      "grad_norm": 0.24395804107189178,
      "learning_rate": 0.00015759034099909158,
      "loss": 0.0275,
      "step": 1130
    },
    {
      "epoch": 6.826347305389222,
      "grad_norm": 0.18488508462905884,
      "learning_rate": 0.00015677501163689604,
      "loss": 0.0299,
      "step": 1140
    },
    {
      "epoch": 6.88622754491018,
      "grad_norm": 0.1932033747434616,
      "learning_rate": 0.00015595407432111134,
      "loss": 0.0259,
      "step": 1150
    },
    {
      "epoch": 6.946107784431137,
      "grad_norm": 0.19903336465358734,
      "learning_rate": 0.00015512761013984711,
      "loss": 0.0278,
      "step": 1160
    },
    {
      "epoch": 7.005988023952096,
      "grad_norm": 0.24045361578464508,
      "learning_rate": 0.00015429570072712925,
      "loss": 0.029,
      "step": 1170
    },
    {
      "epoch": 7.065868263473054,
      "grad_norm": 0.24332760274410248,
      "learning_rate": 0.00015345842825483666,
      "loss": 0.0271,
      "step": 1180
    },
    {
      "epoch": 7.125748502994012,
      "grad_norm": 0.2306133508682251,
      "learning_rate": 0.00015261587542458443,
      "loss": 0.0232,
      "step": 1190
    },
    {
      "epoch": 7.18562874251497,
      "grad_norm": 0.2224716693162918,
      "learning_rate": 0.0001517681254595554,
      "loss": 0.0246,
      "step": 1200
    },
    {
      "epoch": 7.245508982035928,
      "grad_norm": 0.2902887463569641,
      "learning_rate": 0.00015091526209627944,
      "loss": 0.024,
      "step": 1210
    },
    {
      "epoch": 7.3053892215568865,
      "grad_norm": 0.35143372416496277,
      "learning_rate": 0.0001500573695763625,
      "loss": 0.0248,
      "step": 1220
    },
    {
      "epoch": 7.365269461077844,
      "grad_norm": 0.3402245342731476,
      "learning_rate": 0.00014919453263816565,
      "loss": 0.0256,
      "step": 1230
    },
    {
      "epoch": 7.425149700598802,
      "grad_norm": 0.17631883919239044,
      "learning_rate": 0.0001483268365084351,
      "loss": 0.0274,
      "step": 1240
    },
    {
      "epoch": 7.485029940119761,
      "grad_norm": 0.1852467656135559,
      "learning_rate": 0.0001474543668938837,
      "loss": 0.0242,
      "step": 1250
    },
    {
      "epoch": 7.544910179640718,
      "grad_norm": 0.20840848982334137,
      "learning_rate": 0.0001465772099727256,
      "loss": 0.027,
      "step": 1260
    },
    {
      "epoch": 7.604790419161676,
      "grad_norm": 0.19811859726905823,
      "learning_rate": 0.00014569545238616368,
      "loss": 0.0229,
      "step": 1270
    },
    {
      "epoch": 7.664670658682635,
      "grad_norm": 0.15369650721549988,
      "learning_rate": 0.00014480918122983174,
      "loss": 0.0266,
      "step": 1280
    },
    {
      "epoch": 7.724550898203593,
      "grad_norm": 0.4044421315193176,
      "learning_rate": 0.00014391848404519158,
      "loss": 0.0251,
      "step": 1290
    },
    {
      "epoch": 7.7844311377245505,
      "grad_norm": 0.17981642484664917,
      "learning_rate": 0.0001430234488108861,
      "loss": 0.0248,
      "step": 1300
    },
    {
      "epoch": 7.844311377245509,
      "grad_norm": 0.2915557324886322,
      "learning_rate": 0.00014212416393404915,
      "loss": 0.0231,
      "step": 1310
    },
    {
      "epoch": 7.904191616766467,
      "grad_norm": 0.20533527433872223,
      "learning_rate": 0.00014122071824157318,
      "loss": 0.0257,
      "step": 1320
    },
    {
      "epoch": 7.9640718562874255,
      "grad_norm": 0.15892498195171356,
      "learning_rate": 0.0001403132009713354,
      "loss": 0.0253,
      "step": 1330
    },
    {
      "epoch": 8.023952095808383,
      "grad_norm": 0.34935975074768066,
      "learning_rate": 0.0001394017017633831,
      "loss": 0.0261,
      "step": 1340
    },
    {
      "epoch": 8.08383233532934,
      "grad_norm": 0.1838160902261734,
      "learning_rate": 0.00013848631065107973,
      "loss": 0.0231,
      "step": 1350
    },
    {
      "epoch": 8.1437125748503,
      "grad_norm": 0.24909768998622894,
      "learning_rate": 0.00013756711805221168,
      "loss": 0.0256,
      "step": 1360
    },
    {
      "epoch": 8.203592814371257,
      "grad_norm": 0.2749064266681671,
      "learning_rate": 0.00013664421476005723,
      "loss": 0.0258,
      "step": 1370
    },
    {
      "epoch": 8.263473053892216,
      "grad_norm": 0.24632593989372253,
      "learning_rate": 0.0001357176919344186,
      "loss": 0.023,
      "step": 1380
    },
    {
      "epoch": 8.323353293413174,
      "grad_norm": 0.22254560887813568,
      "learning_rate": 0.0001347876410926175,
      "loss": 0.0249,
      "step": 1390
    },
    {
      "epoch": 8.383233532934131,
      "grad_norm": 0.1739770621061325,
      "learning_rate": 0.00013385415410045555,
      "loss": 0.0215,
      "step": 1400
    },
    {
      "epoch": 8.44311377245509,
      "grad_norm": 0.15595784783363342,
      "learning_rate": 0.00013291732316314024,
      "loss": 0.023,
      "step": 1410
    },
    {
      "epoch": 8.502994011976048,
      "grad_norm": 0.12900565564632416,
      "learning_rate": 0.00013197724081617736,
      "loss": 0.0227,
      "step": 1420
    },
    {
      "epoch": 8.562874251497005,
      "grad_norm": 0.18408159911632538,
      "learning_rate": 0.00013103399991623078,
      "loss": 0.0249,
      "step": 1430
    },
    {
      "epoch": 8.622754491017965,
      "grad_norm": 0.3930126130580902,
      "learning_rate": 0.00013008769363195052,
      "loss": 0.0235,
      "step": 1440
    },
    {
      "epoch": 8.682634730538922,
      "grad_norm": 0.22931550443172455,
      "learning_rate": 0.00012913841543477005,
      "loss": 0.02,
      "step": 1450
    },
    {
      "epoch": 8.74251497005988,
      "grad_norm": 0.1572323441505432,
      "learning_rate": 0.00012818625908967367,
      "loss": 0.0244,
      "step": 1460
    },
    {
      "epoch": 8.802395209580839,
      "grad_norm": 0.09360220283269882,
      "learning_rate": 0.0001272313186459348,
      "loss": 0.021,
      "step": 1470
    },
    {
      "epoch": 8.862275449101796,
      "grad_norm": 0.17760924994945526,
      "learning_rate": 0.00012627368842782626,
      "loss": 0.0221,
      "step": 1480
    },
    {
      "epoch": 8.922155688622755,
      "grad_norm": 0.18415001034736633,
      "learning_rate": 0.00012531346302530348,
      "loss": 0.0255,
      "step": 1490
    },
    {
      "epoch": 8.982035928143713,
      "grad_norm": 0.479499489068985,
      "learning_rate": 0.0001243507372846613,
      "loss": 0.0207,
      "step": 1500
    },
    {
      "epoch": 9.04191616766467,
      "grad_norm": 0.22728319466114044,
      "learning_rate": 0.00012338560629916544,
      "loss": 0.0253,
      "step": 1510
    },
    {
      "epoch": 9.10179640718563,
      "grad_norm": 0.1806812733411789,
      "learning_rate": 0.00012241816539965996,
      "loss": 0.0255,
      "step": 1520
    },
    {
      "epoch": 9.161676646706587,
      "grad_norm": 0.26954516768455505,
      "learning_rate": 0.00012144851014515055,
      "loss": 0.0219,
      "step": 1530
    },
    {
      "epoch": 9.221556886227544,
      "grad_norm": 0.30244171619415283,
      "learning_rate": 0.000120476736313366,
      "loss": 0.0253,
      "step": 1540
    },
    {
      "epoch": 9.281437125748504,
      "grad_norm": 0.1826360821723938,
      "learning_rate": 0.0001195029398912976,
      "loss": 0.022,
      "step": 1550
    },
    {
      "epoch": 9.341317365269461,
      "grad_norm": 0.22383596003055573,
      "learning_rate": 0.00011852721706571803,
      "loss": 0.0253,
      "step": 1560
    },
    {
      "epoch": 9.401197604790418,
      "grad_norm": 0.1451132595539093,
      "learning_rate": 0.00011754966421368066,
      "loss": 0.0227,
      "step": 1570
    },
    {
      "epoch": 9.461077844311378,
      "grad_norm": 0.1684962958097458,
      "learning_rate": 0.00011657037789299961,
      "loss": 0.022,
      "step": 1580
    },
    {
      "epoch": 9.520958083832335,
      "grad_norm": 0.18602758646011353,
      "learning_rate": 0.00011558945483271253,
      "loss": 0.0238,
      "step": 1590
    },
    {
      "epoch": 9.580838323353294,
      "grad_norm": 0.1612151712179184,
      "learning_rate": 0.00011460699192352608,
      "loss": 0.0229,
      "step": 1600
    },
    {
      "epoch": 9.640718562874252,
      "grad_norm": 0.1520804762840271,
      "learning_rate": 0.00011362308620824549,
      "loss": 0.0215,
      "step": 1610
    },
    {
      "epoch": 9.70059880239521,
      "grad_norm": 0.23648953437805176,
      "learning_rate": 0.00011263783487218924,
      "loss": 0.0205,
      "step": 1620
    },
    {
      "epoch": 9.760479041916168,
      "grad_norm": 0.17972023785114288,
      "learning_rate": 0.00011165133523358955,
      "loss": 0.0171,
      "step": 1630
    },
    {
      "epoch": 9.820359281437126,
      "grad_norm": 0.20409125089645386,
      "learning_rate": 0.00011066368473397977,
      "loss": 0.0229,
      "step": 1640
    },
    {
      "epoch": 9.880239520958083,
      "grad_norm": 0.1429627388715744,
      "learning_rate": 0.0001096749809285695,
      "loss": 0.0212,
      "step": 1650
    },
    {
      "epoch": 9.940119760479043,
      "grad_norm": 0.20382557809352875,
      "learning_rate": 0.0001086853214766088,
      "loss": 0.019,
      "step": 1660
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.25044798851013184,
      "learning_rate": 0.00010769480413174162,
      "loss": 0.0192,
      "step": 1670
    },
    {
      "epoch": 10.059880239520957,
      "grad_norm": 0.18214142322540283,
      "learning_rate": 0.0001067035267323504,
      "loss": 0.0219,
      "step": 1680
    },
    {
      "epoch": 10.119760479041917,
      "grad_norm": 0.17343035340309143,
      "learning_rate": 0.00010571158719189192,
      "loss": 0.0224,
      "step": 1690
    },
    {
      "epoch": 10.179640718562874,
      "grad_norm": 0.1327686607837677,
      "learning_rate": 0.00010471908348922596,
      "loss": 0.0221,
      "step": 1700
    },
    {
      "epoch": 10.239520958083832,
      "grad_norm": 0.22235077619552612,
      "learning_rate": 0.00010372611365893758,
      "loss": 0.0234,
      "step": 1710
    },
    {
      "epoch": 10.29940119760479,
      "grad_norm": 0.2747216522693634,
      "learning_rate": 0.00010273277578165341,
      "loss": 0.0222,
      "step": 1720
    },
    {
      "epoch": 10.359281437125748,
      "grad_norm": 0.22264912724494934,
      "learning_rate": 0.00010173916797435405,
      "loss": 0.0211,
      "step": 1730
    },
    {
      "epoch": 10.419161676646706,
      "grad_norm": 0.6474366784095764,
      "learning_rate": 0.00010074538838068243,
      "loss": 0.0212,
      "step": 1740
    },
    {
      "epoch": 10.479041916167665,
      "grad_norm": 0.24274206161499023,
      "learning_rate": 9.975153516124972e-05,
      "loss": 0.0214,
      "step": 1750
    },
    {
      "epoch": 10.538922155688622,
      "grad_norm": 0.19830718636512756,
      "learning_rate": 9.875770648393936e-05,
      "loss": 0.0233,
      "step": 1760
    },
    {
      "epoch": 10.598802395209582,
      "grad_norm": 0.12125121057033539,
      "learning_rate": 9.776400051421078e-05,
      "loss": 0.0194,
      "step": 1770
    },
    {
      "epoch": 10.658682634730539,
      "grad_norm": 0.16296449303627014,
      "learning_rate": 9.677051540540286e-05,
      "loss": 0.0225,
      "step": 1780
    },
    {
      "epoch": 10.718562874251496,
      "grad_norm": 0.24490657448768616,
      "learning_rate": 9.577734928903899e-05,
      "loss": 0.0213,
      "step": 1790
    },
    {
      "epoch": 10.778443113772456,
      "grad_norm": 0.16158710420131683,
      "learning_rate": 9.47846002651341e-05,
      "loss": 0.0199,
      "step": 1800
    },
    {
      "epoch": 10.838323353293413,
      "grad_norm": 0.13847924768924713,
      "learning_rate": 9.379236639250473e-05,
      "loss": 0.0185,
      "step": 1810
    },
    {
      "epoch": 10.89820359281437,
      "grad_norm": 0.18506354093551636,
      "learning_rate": 9.28007456790833e-05,
      "loss": 0.0207,
      "step": 1820
    },
    {
      "epoch": 10.95808383233533,
      "grad_norm": 0.1724933683872223,
      "learning_rate": 9.180983607223757e-05,
      "loss": 0.022,
      "step": 1830
    },
    {
      "epoch": 11.017964071856287,
      "grad_norm": 0.2975943088531494,
      "learning_rate": 9.081973544909557e-05,
      "loss": 0.0179,
      "step": 1840
    },
    {
      "epoch": 11.077844311377245,
      "grad_norm": 0.12449675798416138,
      "learning_rate": 8.983054160687801e-05,
      "loss": 0.0211,
      "step": 1850
    },
    {
      "epoch": 11.137724550898204,
      "grad_norm": 0.15286274254322052,
      "learning_rate": 8.88423522532383e-05,
      "loss": 0.0212,
      "step": 1860
    },
    {
      "epoch": 11.197604790419161,
      "grad_norm": 0.17604878544807434,
      "learning_rate": 8.785526499661137e-05,
      "loss": 0.02,
      "step": 1870
    },
    {
      "epoch": 11.25748502994012,
      "grad_norm": 0.33563774824142456,
      "learning_rate": 8.686937733657264e-05,
      "loss": 0.0222,
      "step": 1880
    },
    {
      "epoch": 11.317365269461078,
      "grad_norm": 0.1683143824338913,
      "learning_rate": 8.588478665420717e-05,
      "loss": 0.0215,
      "step": 1890
    },
    {
      "epoch": 11.377245508982035,
      "grad_norm": 0.36611148715019226,
      "learning_rate": 8.490159020249107e-05,
      "loss": 0.0203,
      "step": 1900
    },
    {
      "epoch": 11.437125748502995,
      "grad_norm": 0.13477054238319397,
      "learning_rate": 8.391988509668534e-05,
      "loss": 0.0204,
      "step": 1910
    },
    {
      "epoch": 11.497005988023952,
      "grad_norm": 0.25305667519569397,
      "learning_rate": 8.293976830474312e-05,
      "loss": 0.0222,
      "step": 1920
    },
    {
      "epoch": 11.55688622754491,
      "grad_norm": 0.13116030395030975,
      "learning_rate": 8.196133663773191e-05,
      "loss": 0.0173,
      "step": 1930
    },
    {
      "epoch": 11.616766467065869,
      "grad_norm": 0.15626917779445648,
      "learning_rate": 8.09846867402709e-05,
      "loss": 0.0208,
      "step": 1940
    },
    {
      "epoch": 11.676646706586826,
      "grad_norm": 0.13971462845802307,
      "learning_rate": 8.000991508098492e-05,
      "loss": 0.0177,
      "step": 1950
    },
    {
      "epoch": 11.736526946107784,
      "grad_norm": 0.26388251781463623,
      "learning_rate": 7.903711794297572e-05,
      "loss": 0.019,
      "step": 1960
    },
    {
      "epoch": 11.796407185628743,
      "grad_norm": 0.17708484828472137,
      "learning_rate": 7.806639141431178e-05,
      "loss": 0.0189,
      "step": 1970
    },
    {
      "epoch": 11.8562874251497,
      "grad_norm": 0.2096002697944641,
      "learning_rate": 7.709783137853692e-05,
      "loss": 0.0204,
      "step": 1980
    },
    {
      "epoch": 11.91616766467066,
      "grad_norm": 0.14451850950717926,
      "learning_rate": 7.613153350519969e-05,
      "loss": 0.0173,
      "step": 1990
    },
    {
      "epoch": 11.976047904191617,
      "grad_norm": 0.12203036993741989,
      "learning_rate": 7.516759324040338e-05,
      "loss": 0.0168,
      "step": 2000
    },
    {
      "epoch": 12.035928143712574,
      "grad_norm": 0.14412252604961395,
      "learning_rate": 7.42061057973785e-05,
      "loss": 0.0185,
      "step": 2010
    },
    {
      "epoch": 12.095808383233534,
      "grad_norm": 0.27144360542297363,
      "learning_rate": 7.324716614707793e-05,
      "loss": 0.0195,
      "step": 2020
    },
    {
      "epoch": 12.155688622754491,
      "grad_norm": 0.15820185840129852,
      "learning_rate": 7.229086900879629e-05,
      "loss": 0.0191,
      "step": 2030
    },
    {
      "epoch": 12.215568862275449,
      "grad_norm": 0.16200768947601318,
      "learning_rate": 7.133730884081395e-05,
      "loss": 0.0193,
      "step": 2040
    },
    {
      "epoch": 12.275449101796408,
      "grad_norm": 0.18649928271770477,
      "learning_rate": 7.038657983106702e-05,
      "loss": 0.0159,
      "step": 2050
    },
    {
      "epoch": 12.335329341317365,
      "grad_norm": 0.17984548211097717,
      "learning_rate": 6.943877588784382e-05,
      "loss": 0.0199,
      "step": 2060
    },
    {
      "epoch": 12.395209580838323,
      "grad_norm": 0.16168512403964996,
      "learning_rate": 6.849399063050912e-05,
      "loss": 0.0148,
      "step": 2070
    },
    {
      "epoch": 12.455089820359282,
      "grad_norm": 0.14747169613838196,
      "learning_rate": 6.755231738025693e-05,
      "loss": 0.0189,
      "step": 2080
    },
    {
      "epoch": 12.51497005988024,
      "grad_norm": 0.1350451409816742,
      "learning_rate": 6.661384915089268e-05,
      "loss": 0.0142,
      "step": 2090
    },
    {
      "epoch": 12.574850299401197,
      "grad_norm": 0.12176797538995743,
      "learning_rate": 6.567867863964561e-05,
      "loss": 0.018,
      "step": 2100
    },
    {
      "epoch": 12.634730538922156,
      "grad_norm": 0.12629830837249756,
      "learning_rate": 6.474689821801295e-05,
      "loss": 0.0151,
      "step": 2110
    },
    {
      "epoch": 12.694610778443113,
      "grad_norm": 0.2103053331375122,
      "learning_rate": 6.38185999226356e-05,
      "loss": 0.0148,
      "step": 2120
    },
    {
      "epoch": 12.754491017964071,
      "grad_norm": 0.18460889160633087,
      "learning_rate": 6.289387544620736e-05,
      "loss": 0.0172,
      "step": 2130
    },
    {
      "epoch": 12.81437125748503,
      "grad_norm": 0.22776320576667786,
      "learning_rate": 6.197281612841795e-05,
      "loss": 0.0175,
      "step": 2140
    },
    {
      "epoch": 12.874251497005988,
      "grad_norm": 0.1890844702720642,
      "learning_rate": 6.10555129469309e-05,
      "loss": 0.0154,
      "step": 2150
    },
    {
      "epoch": 12.934131736526947,
      "grad_norm": 0.17597252130508423,
      "learning_rate": 6.014205650839737e-05,
      "loss": 0.0151,
      "step": 2160
    },
    {
      "epoch": 12.994011976047904,
      "grad_norm": 0.15856590867042542,
      "learning_rate": 5.923253703950617e-05,
      "loss": 0.0154,
      "step": 2170
    },
    {
      "epoch": 13.053892215568862,
      "grad_norm": 0.22898879647254944,
      "learning_rate": 5.832704437807195e-05,
      "loss": 0.0137,
      "step": 2180
    },
    {
      "epoch": 13.113772455089821,
      "grad_norm": 0.34772729873657227,
      "learning_rate": 5.742566796416129e-05,
      "loss": 0.0144,
      "step": 2190
    },
    {
      "epoch": 13.173652694610778,
      "grad_norm": 0.18779855966567993,
      "learning_rate": 5.652849683125823e-05,
      "loss": 0.0146,
      "step": 2200
    },
    {
      "epoch": 13.233532934131736,
      "grad_norm": 0.17576445639133453,
      "learning_rate": 5.563561959747019e-05,
      "loss": 0.0196,
      "step": 2210
    },
    {
      "epoch": 13.293413173652695,
      "grad_norm": 0.15298061072826385,
      "learning_rate": 5.4747124456774445e-05,
      "loss": 0.0178,
      "step": 2220
    },
    {
      "epoch": 13.353293413173652,
      "grad_norm": 0.14338500797748566,
      "learning_rate": 5.38630991703069e-05,
      "loss": 0.0179,
      "step": 2230
    },
    {
      "epoch": 13.41317365269461,
      "grad_norm": 0.18957915902137756,
      "learning_rate": 5.2983631057693484e-05,
      "loss": 0.0165,
      "step": 2240
    },
    {
      "epoch": 13.47305389221557,
      "grad_norm": 0.19262787699699402,
      "learning_rate": 5.210880698842514e-05,
      "loss": 0.0139,
      "step": 2250
    },
    {
      "epoch": 13.532934131736527,
      "grad_norm": 0.3551245331764221,
      "learning_rate": 5.123871337327724e-05,
      "loss": 0.0172,
      "step": 2260
    },
    {
      "epoch": 13.592814371257486,
      "grad_norm": 0.21090568602085114,
      "learning_rate": 5.037343615577448e-05,
      "loss": 0.0159,
      "step": 2270
    },
    {
      "epoch": 13.652694610778443,
      "grad_norm": 0.10332461446523666,
      "learning_rate": 4.951306080370165e-05,
      "loss": 0.0147,
      "step": 2280
    },
    {
      "epoch": 13.7125748502994,
      "grad_norm": 0.12004910409450531,
      "learning_rate": 4.865767230066166e-05,
      "loss": 0.0132,
      "step": 2290
    },
    {
      "epoch": 13.77245508982036,
      "grad_norm": 0.17986303567886353,
      "learning_rate": 4.78073551376812e-05,
      "loss": 0.0142,
      "step": 2300
    },
    {
      "epoch": 13.832335329341317,
      "grad_norm": 0.17033332586288452,
      "learning_rate": 4.6962193304865335e-05,
      "loss": 0.0145,
      "step": 2310
    },
    {
      "epoch": 13.892215568862275,
      "grad_norm": 0.5156183838844299,
      "learning_rate": 4.6122270283100944e-05,
      "loss": 0.0162,
      "step": 2320
    },
    {
      "epoch": 13.952095808383234,
      "grad_norm": 0.47306326031684875,
      "learning_rate": 4.528766903581149e-05,
      "loss": 0.0151,
      "step": 2330
    },
    {
      "epoch": 14.011976047904191,
      "grad_norm": 0.10242114216089249,
      "learning_rate": 4.44584720007618e-05,
      "loss": 0.0197,
      "step": 2340
    },
    {
      "epoch": 14.071856287425149,
      "grad_norm": 0.13059602677822113,
      "learning_rate": 4.363476108191552e-05,
      "loss": 0.0144,
      "step": 2350
    },
    {
      "epoch": 14.131736526946108,
      "grad_norm": 0.1200447678565979,
      "learning_rate": 4.2816617641344934e-05,
      "loss": 0.0158,
      "step": 2360
    },
    {
      "epoch": 14.191616766467066,
      "grad_norm": 0.15896056592464447,
      "learning_rate": 4.200412249119463e-05,
      "loss": 0.0135,
      "step": 2370
    },
    {
      "epoch": 14.251497005988025,
      "grad_norm": 0.18628185987472534,
      "learning_rate": 4.119735588569899e-05,
      "loss": 0.0131,
      "step": 2380
    },
    {
      "epoch": 14.311377245508982,
      "grad_norm": 0.1670549213886261,
      "learning_rate": 4.039639751325525e-05,
      "loss": 0.0145,
      "step": 2390
    },
    {
      "epoch": 14.37125748502994,
      "grad_norm": 0.2762540280818939,
      "learning_rate": 3.960132648855226e-05,
      "loss": 0.0167,
      "step": 2400
    },
    {
      "epoch": 14.431137724550899,
      "grad_norm": 0.09425695985555649,
      "learning_rate": 3.881222134475589e-05,
      "loss": 0.0133,
      "step": 2410
    },
    {
      "epoch": 14.491017964071856,
      "grad_norm": 0.176064133644104,
      "learning_rate": 3.802916002575192e-05,
      "loss": 0.0124,
      "step": 2420
    },
    {
      "epoch": 14.550898203592814,
      "grad_norm": 0.19551633298397064,
      "learning_rate": 3.725221987844721e-05,
      "loss": 0.0177,
      "step": 2430
    },
    {
      "epoch": 14.610778443113773,
      "grad_norm": 0.20848046243190765,
      "learning_rate": 3.6481477645129624e-05,
      "loss": 0.0154,
      "step": 2440
    },
    {
      "epoch": 14.67065868263473,
      "grad_norm": 0.2565270960330963,
      "learning_rate": 3.571700945588789e-05,
      "loss": 0.015,
      "step": 2450
    },
    {
      "epoch": 14.730538922155688,
      "grad_norm": 0.13702988624572754,
      "learning_rate": 3.495889082109185e-05,
      "loss": 0.0157,
      "step": 2460
    },
    {
      "epoch": 14.790419161676647,
      "grad_norm": 0.5474858283996582,
      "learning_rate": 3.4207196623933857e-05,
      "loss": 0.0143,
      "step": 2470
    },
    {
      "epoch": 14.850299401197605,
      "grad_norm": 0.1655379682779312,
      "learning_rate": 3.346200111303226e-05,
      "loss": 0.0152,
      "step": 2480
    },
    {
      "epoch": 14.910179640718562,
      "grad_norm": 0.13053831458091736,
      "learning_rate": 3.272337789509751e-05,
      "loss": 0.0139,
      "step": 2490
    },
    {
      "epoch": 14.970059880239521,
      "grad_norm": 0.1283467561006546,
      "learning_rate": 3.199139992766155e-05,
      "loss": 0.0161,
      "step": 2500
    },
    {
      "epoch": 15.029940119760479,
      "grad_norm": 0.13677707314491272,
      "learning_rate": 3.1266139511871465e-05,
      "loss": 0.0126,
      "step": 2510
    },
    {
      "epoch": 15.089820359281438,
      "grad_norm": 0.10139447450637817,
      "learning_rate": 3.0547668285348105e-05,
      "loss": 0.0143,
      "step": 2520
    },
    {
      "epoch": 15.149700598802395,
      "grad_norm": 0.14132164418697357,
      "learning_rate": 2.9836057215109746e-05,
      "loss": 0.0149,
      "step": 2530
    },
    {
      "epoch": 15.209580838323353,
      "grad_norm": 0.13676463067531586,
      "learning_rate": 2.9131376590562587e-05,
      "loss": 0.0131,
      "step": 2540
    },
    {
      "epoch": 15.269461077844312,
      "grad_norm": 0.20659254491329193,
      "learning_rate": 2.843369601655792e-05,
      "loss": 0.0125,
      "step": 2550
    },
    {
      "epoch": 15.32934131736527,
      "grad_norm": 0.13563022017478943,
      "learning_rate": 2.7743084406516693e-05,
      "loss": 0.0147,
      "step": 2560
    },
    {
      "epoch": 15.389221556886227,
      "grad_norm": 0.13761834800243378,
      "learning_rate": 2.705960997562277e-05,
      "loss": 0.0159,
      "step": 2570
    },
    {
      "epoch": 15.449101796407186,
      "grad_norm": 0.11985321342945099,
      "learning_rate": 2.6383340234084875e-05,
      "loss": 0.0111,
      "step": 2580
    },
    {
      "epoch": 15.508982035928144,
      "grad_norm": 0.09939990192651749,
      "learning_rate": 2.5714341980468437e-05,
      "loss": 0.0122,
      "step": 2590
    },
    {
      "epoch": 15.568862275449101,
      "grad_norm": 0.19175921380519867,
      "learning_rate": 2.5052681295097225e-05,
      "loss": 0.0134,
      "step": 2600
    },
    {
      "epoch": 15.62874251497006,
      "grad_norm": 0.11813875287771225,
      "learning_rate": 2.439842353352667e-05,
      "loss": 0.015,
      "step": 2610
    },
    {
      "epoch": 15.688622754491018,
      "grad_norm": 0.16614750027656555,
      "learning_rate": 2.3751633320088064e-05,
      "loss": 0.0147,
      "step": 2620
    },
    {
      "epoch": 15.748502994011975,
      "grad_norm": 0.2053995579481125,
      "learning_rate": 2.311237454150539e-05,
      "loss": 0.0122,
      "step": 2630
    },
    {
      "epoch": 15.808383233532934,
      "grad_norm": 0.21074947714805603,
      "learning_rate": 2.248071034058492e-05,
      "loss": 0.0133,
      "step": 2640
    },
    {
      "epoch": 15.868263473053892,
      "grad_norm": 0.18404161930084229,
      "learning_rate": 2.185670310997835e-05,
      "loss": 0.0155,
      "step": 2650
    },
    {
      "epoch": 15.928143712574851,
      "grad_norm": 0.12167193740606308,
      "learning_rate": 2.124041448601979e-05,
      "loss": 0.0113,
      "step": 2660
    },
    {
      "epoch": 15.988023952095809,
      "grad_norm": 0.19261722266674042,
      "learning_rate": 2.0631905342637782e-05,
      "loss": 0.0124,
      "step": 2670
    },
    {
      "epoch": 16.047904191616766,
      "grad_norm": 0.152290940284729,
      "learning_rate": 2.0031235785342407e-05,
      "loss": 0.014,
      "step": 2680
    },
    {
      "epoch": 16.107784431137723,
      "grad_norm": 0.13786451518535614,
      "learning_rate": 1.9438465145288374e-05,
      "loss": 0.0125,
      "step": 2690
    },
    {
      "epoch": 16.16766467065868,
      "grad_norm": 0.11830327659845352,
      "learning_rate": 1.885365197341461e-05,
      "loss": 0.013,
      "step": 2700
    },
    {
      "epoch": 16.227544910179642,
      "grad_norm": 0.1224045529961586,
      "learning_rate": 1.827685403466087e-05,
      "loss": 0.0105,
      "step": 2710
    },
    {
      "epoch": 16.2874251497006,
      "grad_norm": 0.12195927649736404,
      "learning_rate": 1.7708128302261994e-05,
      "loss": 0.0117,
      "step": 2720
    },
    {
      "epoch": 16.347305389221557,
      "grad_norm": 0.1412719041109085,
      "learning_rate": 1.7147530952120337e-05,
      "loss": 0.0131,
      "step": 2730
    },
    {
      "epoch": 16.407185628742514,
      "grad_norm": 0.1535654515028,
      "learning_rate": 1.659511735725716e-05,
      "loss": 0.013,
      "step": 2740
    },
    {
      "epoch": 16.46706586826347,
      "grad_norm": 0.3097449839115143,
      "learning_rate": 1.605094208234288e-05,
      "loss": 0.0105,
      "step": 2750
    },
    {
      "epoch": 16.526946107784433,
      "grad_norm": 0.1134740337729454,
      "learning_rate": 1.5515058878307665e-05,
      "loss": 0.0107,
      "step": 2760
    },
    {
      "epoch": 16.58682634730539,
      "grad_norm": 0.23593778908252716,
      "learning_rate": 1.4987520677032174e-05,
      "loss": 0.0117,
      "step": 2770
    },
    {
      "epoch": 16.646706586826348,
      "grad_norm": 0.1333216279745102,
      "learning_rate": 1.4468379586119074e-05,
      "loss": 0.0151,
      "step": 2780
    },
    {
      "epoch": 16.706586826347305,
      "grad_norm": 0.2893410623073578,
      "learning_rate": 1.3957686883746191e-05,
      "loss": 0.0101,
      "step": 2790
    },
    {
      "epoch": 16.766467065868262,
      "grad_norm": 0.12125154584646225,
      "learning_rate": 1.3455493013601561e-05,
      "loss": 0.0115,
      "step": 2800
    },
    {
      "epoch": 16.82634730538922,
      "grad_norm": 0.3616254925727844,
      "learning_rate": 1.2961847579900777e-05,
      "loss": 0.0133,
      "step": 2810
    },
    {
      "epoch": 16.88622754491018,
      "grad_norm": 0.1681334674358368,
      "learning_rate": 1.2476799342487255e-05,
      "loss": 0.0126,
      "step": 2820
    },
    {
      "epoch": 16.94610778443114,
      "grad_norm": 0.14123430848121643,
      "learning_rate": 1.2000396212016218e-05,
      "loss": 0.0146,
      "step": 2830
    },
    {
      "epoch": 17.005988023952096,
      "grad_norm": 0.15791960060596466,
      "learning_rate": 1.1532685245222075e-05,
      "loss": 0.0113,
      "step": 2840
    },
    {
      "epoch": 17.065868263473053,
      "grad_norm": 0.20897255837917328,
      "learning_rate": 1.1073712640270506e-05,
      "loss": 0.0111,
      "step": 2850
    },
    {
      "epoch": 17.12574850299401,
      "grad_norm": 0.07108011096715927,
      "learning_rate": 1.0623523732195284e-05,
      "loss": 0.0123,
      "step": 2860
    },
    {
      "epoch": 17.18562874251497,
      "grad_norm": 0.2679906189441681,
      "learning_rate": 1.0182162988420163e-05,
      "loss": 0.0145,
      "step": 2870
    },
    {
      "epoch": 17.24550898203593,
      "grad_norm": 0.11328988522291183,
      "learning_rate": 9.749674004366727e-06,
      "loss": 0.0122,
      "step": 2880
    },
    {
      "epoch": 17.305389221556887,
      "grad_norm": 0.3787294924259186,
      "learning_rate": 9.326099499148177e-06,
      "loss": 0.0103,
      "step": 2890
    },
    {
      "epoch": 17.365269461077844,
      "grad_norm": 0.2220940738916397,
      "learning_rate": 8.911481311349834e-06,
      "loss": 0.0123,
      "step": 2900
    },
    {
      "epoch": 17.4251497005988,
      "grad_norm": 0.12199822813272476,
      "learning_rate": 8.505860394896414e-06,
      "loss": 0.011,
      "step": 2910
    },
    {
      "epoch": 17.48502994011976,
      "grad_norm": 0.10359202325344086,
      "learning_rate": 8.109276815006938e-06,
      "loss": 0.0102,
      "step": 2920
    },
    {
      "epoch": 17.54491017964072,
      "grad_norm": 0.49252983927726746,
      "learning_rate": 7.721769744237184e-06,
      "loss": 0.014,
      "step": 2930
    },
    {
      "epoch": 17.604790419161677,
      "grad_norm": 0.3275662064552307,
      "learning_rate": 7.343377458610446e-06,
      "loss": 0.0148,
      "step": 2940
    },
    {
      "epoch": 17.664670658682635,
      "grad_norm": 0.7134996056556702,
      "learning_rate": 6.9741373338368165e-06,
      "loss": 0.0129,
      "step": 2950
    },
    {
      "epoch": 17.724550898203592,
      "grad_norm": 0.10740635544061661,
      "learning_rate": 6.614085841621442e-06,
      "loss": 0.0111,
      "step": 2960
    },
    {
      "epoch": 17.78443113772455,
      "grad_norm": 0.13902905583381653,
      "learning_rate": 6.263258546061978e-06,
      "loss": 0.0124,
      "step": 2970
    },
    {
      "epoch": 17.84431137724551,
      "grad_norm": 0.2780143916606903,
      "learning_rate": 5.921690100135713e-06,
      "loss": 0.009,
      "step": 2980
    },
    {
      "epoch": 17.904191616766468,
      "grad_norm": 0.11749708652496338,
      "learning_rate": 5.589414242276869e-06,
      "loss": 0.0098,
      "step": 2990
    },
    {
      "epoch": 17.964071856287426,
      "grad_norm": 0.45455411076545715,
      "learning_rate": 5.266463793043896e-06,
      "loss": 0.0087,
      "step": 3000
    },
    {
      "epoch": 18.023952095808383,
      "grad_norm": 0.22297538816928864,
      "learning_rate": 4.952870651877739e-06,
      "loss": 0.0114,
      "step": 3010
    },
    {
      "epoch": 18.08383233532934,
      "grad_norm": 0.5311406254768372,
      "learning_rate": 4.648665793950968e-06,
      "loss": 0.0093,
      "step": 3020
    },
    {
      "epoch": 18.143712574850298,
      "grad_norm": 0.20427249372005463,
      "learning_rate": 4.353879267108119e-06,
      "loss": 0.0123,
      "step": 3030
    },
    {
      "epoch": 18.20359281437126,
      "grad_norm": 0.2326945662498474,
      "learning_rate": 4.068540188897796e-06,
      "loss": 0.0099,
      "step": 3040
    },
    {
      "epoch": 18.263473053892216,
      "grad_norm": 0.14544923603534698,
      "learning_rate": 3.792676743696588e-06,
      "loss": 0.0123,
      "step": 3050
    },
    {
      "epoch": 18.323353293413174,
      "grad_norm": 0.1755136400461197,
      "learning_rate": 3.5263161799251154e-06,
      "loss": 0.0095,
      "step": 3060
    },
    {
      "epoch": 18.38323353293413,
      "grad_norm": 0.10239781439304352,
      "learning_rate": 3.2694848073565596e-06,
      "loss": 0.0095,
      "step": 3070
    },
    {
      "epoch": 18.44311377245509,
      "grad_norm": 0.13217875361442566,
      "learning_rate": 3.0222079945179895e-06,
      "loss": 0.0111,
      "step": 3080
    },
    {
      "epoch": 18.50299401197605,
      "grad_norm": 0.11932236701250076,
      "learning_rate": 2.7845101661844618e-06,
      "loss": 0.0099,
      "step": 3090
    },
    {
      "epoch": 18.562874251497007,
      "grad_norm": 0.12827470898628235,
      "learning_rate": 2.5564148009665778e-06,
      "loss": 0.0125,
      "step": 3100
    },
    {
      "epoch": 18.622754491017965,
      "grad_norm": 0.44992345571517944,
      "learning_rate": 2.3379444289913342e-06,
      "loss": 0.0114,
      "step": 3110
    },
    {
      "epoch": 18.682634730538922,
      "grad_norm": 0.22988931834697723,
      "learning_rate": 2.1291206296767063e-06,
      "loss": 0.0121,
      "step": 3120
    },
    {
      "epoch": 18.74251497005988,
      "grad_norm": 0.2384101152420044,
      "learning_rate": 1.9299640296001288e-06,
      "loss": 0.0094,
      "step": 3130
    },
    {
      "epoch": 18.802395209580837,
      "grad_norm": 0.10543084144592285,
      "learning_rate": 1.7404943004611596e-06,
      "loss": 0.013,
      "step": 3140
    },
    {
      "epoch": 18.862275449101798,
      "grad_norm": 0.12687674164772034,
      "learning_rate": 1.5607301571383459e-06,
      "loss": 0.0112,
      "step": 3150
    },
    {
      "epoch": 18.922155688622755,
      "grad_norm": 0.26404881477355957,
      "learning_rate": 1.390689355840691e-06,
      "loss": 0.0133,
      "step": 3160
    },
    {
      "epoch": 18.982035928143713,
      "grad_norm": 0.11557133495807648,
      "learning_rate": 1.2303886923537677e-06,
      "loss": 0.0107,
      "step": 3170
    },
    {
      "epoch": 19.04191616766467,
      "grad_norm": 0.10522342473268509,
      "learning_rate": 1.0798440003807474e-06,
      "loss": 0.0107,
      "step": 3180
    },
    {
      "epoch": 19.101796407185628,
      "grad_norm": 0.1134905070066452,
      "learning_rate": 9.390701499783827e-07,
      "loss": 0.0086,
      "step": 3190
    },
    {
      "epoch": 19.161676646706585,
      "grad_norm": 0.1318759173154831,
      "learning_rate": 8.080810460882493e-07,
      "loss": 0.0107,
      "step": 3200
    },
    {
      "epoch": 19.221556886227546,
      "grad_norm": 0.11792134493589401,
      "learning_rate": 6.868896271632785e-07,
      "loss": 0.0099,
      "step": 3210
    },
    {
      "epoch": 19.281437125748504,
      "grad_norm": 0.18841320276260376,
      "learning_rate": 5.755078638897459e-07,
      "loss": 0.0118,
      "step": 3220
    },
    {
      "epoch": 19.34131736526946,
      "grad_norm": 0.1288711428642273,
      "learning_rate": 4.739467580048618e-07,
      "loss": 0.0092,
      "step": 3230
    },
    {
      "epoch": 19.40119760479042,
      "grad_norm": 0.3439689874649048,
      "learning_rate": 3.8221634121013983e-07,
      "loss": 0.0083,
      "step": 3240
    },
    {
      "epoch": 19.461077844311376,
      "grad_norm": 0.07992304116487503,
      "learning_rate": 3.003256741803906e-07,
      "loss": 0.012,
      "step": 3250
    },
    {
      "epoch": 19.520958083832337,
      "grad_norm": 0.15533998608589172,
      "learning_rate": 2.2828284566890345e-07,
      "loss": 0.0108,
      "step": 3260
    },
    {
      "epoch": 19.580838323353294,
      "grad_norm": 0.3759068548679352,
      "learning_rate": 1.6609497170834155e-07,
      "loss": 0.0112,
      "step": 3270
    },
    {
      "epoch": 19.64071856287425,
      "grad_norm": 0.12365289032459259,
      "learning_rate": 1.1376819490795942e-07,
      "loss": 0.0135,
      "step": 3280
    },
    {
      "epoch": 19.70059880239521,
      "grad_norm": 0.12051455676555634,
      "learning_rate": 7.130768384679965e-08,
      "loss": 0.01,
      "step": 3290
    },
    {
      "epoch": 19.760479041916167,
      "grad_norm": 0.10926169902086258,
      "learning_rate": 3.8717632563201046e-08,
      "loss": 0.0152,
      "step": 3300
    },
    {
      "epoch": 19.820359281437124,
      "grad_norm": 0.15941037237644196,
      "learning_rate": 1.600126014051906e-08,
      "loss": 0.0109,
      "step": 3310
    },
    {
      "epoch": 19.880239520958085,
      "grad_norm": 0.07696834206581116,
      "learning_rate": 3.1608103891689335e-09,
      "loss": 0.0093,
      "step": 3320
    },
    {
      "epoch": 19.92814371257485,
      "step": 3328,
      "total_flos": 4.678773863650698e+17,
      "train_loss": 0.034036972624912426,
      "train_runtime": 3742.5882,
      "train_samples_per_second": 56.91,
      "train_steps_per_second": 0.889
    }
  ],
  "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
}