{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.204663212435234,
  "eval_steps": 500,
  "global_step": 3939,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.025906735751295335,
      "grad_norm": 7.820053577423096,
      "learning_rate": 1.015228426395939e-05,
      "loss": 1.2662,
      "step": 10
    },
    {
      "epoch": 0.05181347150259067,
      "grad_norm": 2.624858856201172,
      "learning_rate": 2.030456852791878e-05,
      "loss": 0.829,
      "step": 20
    },
    {
      "epoch": 0.07772020725388601,
      "grad_norm": 3.9165711402893066,
      "learning_rate": 3.0456852791878175e-05,
      "loss": 0.4265,
      "step": 30
    },
    {
      "epoch": 0.10362694300518134,
      "grad_norm": 1.9645638465881348,
      "learning_rate": 4.060913705583756e-05,
      "loss": 0.3073,
      "step": 40
    },
    {
      "epoch": 0.12953367875647667,
      "grad_norm": 0.6432750225067139,
      "learning_rate": 5.076142131979695e-05,
      "loss": 0.234,
      "step": 50
    },
    {
      "epoch": 0.15544041450777202,
      "grad_norm": 2.0933594703674316,
      "learning_rate": 6.091370558375635e-05,
      "loss": 0.2026,
      "step": 60
    },
    {
      "epoch": 0.18134715025906736,
      "grad_norm": 0.8744344115257263,
      "learning_rate": 7.106598984771574e-05,
      "loss": 0.1762,
      "step": 70
    },
    {
      "epoch": 0.20725388601036268,
      "grad_norm": 1.3427742719650269,
      "learning_rate": 8.121827411167512e-05,
      "loss": 0.1651,
      "step": 80
    },
    {
      "epoch": 0.23316062176165803,
      "grad_norm": 0.8046813607215881,
      "learning_rate": 9.137055837563452e-05,
      "loss": 0.153,
      "step": 90
    },
    {
      "epoch": 0.25906735751295334,
      "grad_norm": 0.8898593783378601,
      "learning_rate": 0.0001015228426395939,
      "loss": 0.1349,
      "step": 100
    },
    {
      "epoch": 0.2849740932642487,
      "grad_norm": 0.5029700994491577,
      "learning_rate": 0.0001116751269035533,
      "loss": 0.131,
      "step": 110
    },
    {
      "epoch": 0.31088082901554404,
      "grad_norm": 0.5996863842010498,
      "learning_rate": 0.0001218274111675127,
      "loss": 0.1167,
      "step": 120
    },
    {
      "epoch": 0.33678756476683935,
      "grad_norm": 0.935474693775177,
      "learning_rate": 0.00013197969543147207,
      "loss": 0.103,
      "step": 130
    },
    {
      "epoch": 0.3626943005181347,
      "grad_norm": 0.9033486247062683,
      "learning_rate": 0.00014213197969543148,
      "loss": 0.1048,
      "step": 140
    },
    {
      "epoch": 0.38860103626943004,
      "grad_norm": 0.8507540822029114,
      "learning_rate": 0.00015228426395939087,
      "loss": 0.1061,
      "step": 150
    },
    {
      "epoch": 0.41450777202072536,
      "grad_norm": 0.8174213767051697,
      "learning_rate": 0.00016243654822335025,
      "loss": 0.1012,
      "step": 160
    },
    {
      "epoch": 0.44041450777202074,
      "grad_norm": 0.5809940099716187,
      "learning_rate": 0.00017258883248730966,
      "loss": 0.0953,
      "step": 170
    },
    {
      "epoch": 0.46632124352331605,
      "grad_norm": 0.5646868944168091,
      "learning_rate": 0.00018274111675126904,
      "loss": 0.0842,
      "step": 180
    },
    {
      "epoch": 0.49222797927461137,
      "grad_norm": 0.7231911420822144,
      "learning_rate": 0.00019289340101522843,
      "loss": 0.0846,
      "step": 190
    },
    {
      "epoch": 0.5181347150259067,
      "grad_norm": 0.5057425498962402,
      "learning_rate": 0.00019999968282097247,
      "loss": 0.0806,
      "step": 200
    },
    {
      "epoch": 0.5440414507772021,
      "grad_norm": 0.4385872483253479,
      "learning_rate": 0.0001999940441386779,
      "loss": 0.0811,
      "step": 210
    },
    {
      "epoch": 0.5699481865284974,
      "grad_norm": 0.49294012784957886,
      "learning_rate": 0.0001999813574910152,
      "loss": 0.0778,
      "step": 220
    },
    {
      "epoch": 0.5958549222797928,
      "grad_norm": 0.8409459590911865,
      "learning_rate": 0.0001999616237721881,
      "loss": 0.0777,
      "step": 230
    },
    {
      "epoch": 0.6217616580310881,
      "grad_norm": 0.9242765307426453,
      "learning_rate": 0.00019993484437310518,
      "loss": 0.0837,
      "step": 240
    },
    {
      "epoch": 0.6476683937823834,
      "grad_norm": 0.6676081418991089,
      "learning_rate": 0.00019990102118128155,
      "loss": 0.0722,
      "step": 250
    },
    {
      "epoch": 0.6735751295336787,
      "grad_norm": 0.36002111434936523,
      "learning_rate": 0.0001998601565807061,
      "loss": 0.0755,
      "step": 260
    },
    {
      "epoch": 0.6994818652849741,
      "grad_norm": 0.25635281205177307,
      "learning_rate": 0.0001998122534516733,
      "loss": 0.0651,
      "step": 270
    },
    {
      "epoch": 0.7253886010362695,
      "grad_norm": 0.7322975993156433,
      "learning_rate": 0.00019975731517058012,
      "loss": 0.0664,
      "step": 280
    },
    {
      "epoch": 0.7512953367875648,
      "grad_norm": 0.43520084023475647,
      "learning_rate": 0.00019969534560968823,
      "loss": 0.0653,
      "step": 290
    },
    {
      "epoch": 0.7772020725388601,
      "grad_norm": 0.39980387687683105,
      "learning_rate": 0.000199626349136851,
      "loss": 0.0677,
      "step": 300
    },
    {
      "epoch": 0.8031088082901554,
      "grad_norm": 0.4908325970172882,
      "learning_rate": 0.0001995503306152055,
      "loss": 0.0665,
      "step": 310
    },
    {
      "epoch": 0.8290155440414507,
      "grad_norm": 0.566394567489624,
      "learning_rate": 0.00019946729540283,
      "loss": 0.0592,
      "step": 320
    },
    {
      "epoch": 0.8549222797927462,
      "grad_norm": 0.3903208374977112,
      "learning_rate": 0.0001993772493523661,
      "loss": 0.0625,
      "step": 330
    },
    {
      "epoch": 0.8808290155440415,
      "grad_norm": 0.31053635478019714,
      "learning_rate": 0.00019928019881060625,
      "loss": 0.0668,
      "step": 340
    },
    {
      "epoch": 0.9067357512953368,
      "grad_norm": 0.33370479941368103,
      "learning_rate": 0.00019917615061804637,
      "loss": 0.0545,
      "step": 350
    },
    {
      "epoch": 0.9326424870466321,
      "grad_norm": 0.5482617020606995,
      "learning_rate": 0.00019906511210840386,
      "loss": 0.0628,
      "step": 360
    },
    {
      "epoch": 0.9585492227979274,
      "grad_norm": 0.5890028476715088,
      "learning_rate": 0.00019894709110810057,
      "loss": 0.0586,
      "step": 370
    },
    {
      "epoch": 0.9844559585492227,
      "grad_norm": 0.6143954396247864,
      "learning_rate": 0.00019882209593571117,
      "loss": 0.0596,
      "step": 380
    },
    {
      "epoch": 1.0103626943005182,
      "grad_norm": 0.27864834666252136,
      "learning_rate": 0.00019869013540137683,
      "loss": 0.0538,
      "step": 390
    },
    {
      "epoch": 1.0362694300518134,
      "grad_norm": 0.6054168343544006,
      "learning_rate": 0.00019855121880618432,
      "loss": 0.0552,
      "step": 400
    },
    {
      "epoch": 1.0621761658031088,
      "grad_norm": 0.3929780125617981,
      "learning_rate": 0.00019840535594151023,
      "loss": 0.0564,
      "step": 410
    },
    {
      "epoch": 1.0880829015544042,
      "grad_norm": 0.45577749609947205,
      "learning_rate": 0.00019825255708833108,
      "loss": 0.0538,
      "step": 420
    },
    {
      "epoch": 1.1139896373056994,
      "grad_norm": 0.4095696806907654,
      "learning_rate": 0.00019809283301649866,
      "loss": 0.0555,
      "step": 430
    },
    {
      "epoch": 1.1398963730569949,
      "grad_norm": 0.3670872747898102,
      "learning_rate": 0.00019792619498398064,
      "loss": 0.052,
      "step": 440
    },
    {
      "epoch": 1.16580310880829,
      "grad_norm": 0.40029576420783997,
      "learning_rate": 0.00019775265473606743,
      "loss": 0.0533,
      "step": 450
    },
    {
      "epoch": 1.1917098445595855,
      "grad_norm": 0.4621596038341522,
      "learning_rate": 0.00019757222450454408,
      "loss": 0.0481,
      "step": 460
    },
    {
      "epoch": 1.2176165803108807,
      "grad_norm": 0.5111753940582275,
      "learning_rate": 0.00019738491700682832,
      "loss": 0.0493,
      "step": 470
    },
    {
      "epoch": 1.2435233160621761,
      "grad_norm": 0.28371885418891907,
      "learning_rate": 0.00019719074544507391,
      "loss": 0.0489,
      "step": 480
    },
    {
      "epoch": 1.2694300518134716,
      "grad_norm": 0.4142397344112396,
      "learning_rate": 0.0001969897235052405,
      "loss": 0.0507,
      "step": 490
    },
    {
      "epoch": 1.2953367875647668,
      "grad_norm": 0.266857773065567,
      "learning_rate": 0.0001967818653561286,
      "loss": 0.0515,
      "step": 500
    },
    {
      "epoch": 1.3212435233160622,
      "grad_norm": 0.32568779587745667,
      "learning_rate": 0.0001965671856483812,
      "loss": 0.0504,
      "step": 510
    },
    {
      "epoch": 1.3471502590673574,
      "grad_norm": 0.43539974093437195,
      "learning_rate": 0.00019634569951345088,
      "loss": 0.0476,
      "step": 520
    },
    {
      "epoch": 1.3730569948186528,
      "grad_norm": 0.40603071451187134,
      "learning_rate": 0.0001961174225625336,
      "loss": 0.0519,
      "step": 530
    },
    {
      "epoch": 1.3989637305699483,
      "grad_norm": 0.5506306886672974,
      "learning_rate": 0.00019588237088546807,
      "loss": 0.0506,
      "step": 540
    },
    {
      "epoch": 1.4248704663212435,
      "grad_norm": 0.4322110414505005,
      "learning_rate": 0.00019564056104960176,
      "loss": 0.0412,
      "step": 550
    },
    {
      "epoch": 1.450777202072539,
      "grad_norm": 0.30257806181907654,
      "learning_rate": 0.00019539201009862332,
      "loss": 0.049,
      "step": 560
    },
    {
      "epoch": 1.4766839378238341,
      "grad_norm": 0.28851810097694397,
      "learning_rate": 0.00019513673555136105,
      "loss": 0.0503,
      "step": 570
    },
    {
      "epoch": 1.5025906735751295,
      "grad_norm": 0.39474016427993774,
      "learning_rate": 0.00019487475540054828,
      "loss": 0.0497,
      "step": 580
    },
    {
      "epoch": 1.528497409326425,
      "grad_norm": 0.32942134141921997,
      "learning_rate": 0.00019460608811155508,
      "loss": 0.0492,
      "step": 590
    },
    {
      "epoch": 1.5544041450777202,
      "grad_norm": 0.2994638979434967,
      "learning_rate": 0.00019433075262108678,
      "loss": 0.0471,
      "step": 600
    },
    {
      "epoch": 1.5803108808290154,
      "grad_norm": 0.3796844482421875,
      "learning_rate": 0.00019404876833584923,
      "loss": 0.0457,
      "step": 610
    },
    {
      "epoch": 1.6062176165803108,
      "grad_norm": 0.30895495414733887,
      "learning_rate": 0.00019376015513118092,
      "loss": 0.0475,
      "step": 620
    },
    {
      "epoch": 1.6321243523316062,
      "grad_norm": 0.3315885663032532,
      "learning_rate": 0.00019346493334965217,
      "loss": 0.0461,
      "step": 630
    },
    {
      "epoch": 1.6580310880829017,
      "grad_norm": 0.4402955174446106,
      "learning_rate": 0.00019316312379963118,
      "loss": 0.0502,
      "step": 640
    },
    {
      "epoch": 1.6839378238341969,
      "grad_norm": 0.46544697880744934,
      "learning_rate": 0.00019285474775381759,
      "loss": 0.0487,
      "step": 650
    },
    {
      "epoch": 1.709844559585492,
      "grad_norm": 0.48502856492996216,
      "learning_rate": 0.0001925398269477427,
      "loss": 0.0429,
      "step": 660
    },
    {
      "epoch": 1.7357512953367875,
      "grad_norm": 0.23488780856132507,
      "learning_rate": 0.000192218383578238,
      "loss": 0.0448,
      "step": 670
    },
    {
      "epoch": 1.761658031088083,
      "grad_norm": 0.3346535265445709,
      "learning_rate": 0.00019189044030187017,
      "loss": 0.0424,
      "step": 680
    },
    {
      "epoch": 1.7875647668393784,
      "grad_norm": 0.31202977895736694,
      "learning_rate": 0.00019155602023334444,
      "loss": 0.037,
      "step": 690
    },
    {
      "epoch": 1.8134715025906736,
      "grad_norm": 0.2789950966835022,
      "learning_rate": 0.0001912151469438753,
      "loss": 0.0394,
      "step": 700
    },
    {
      "epoch": 1.8393782383419688,
      "grad_norm": 0.3802800476551056,
      "learning_rate": 0.00019086784445952515,
      "loss": 0.0442,
      "step": 710
    },
    {
      "epoch": 1.8652849740932642,
      "grad_norm": 0.26964104175567627,
      "learning_rate": 0.0001905141372595107,
      "loss": 0.0409,
      "step": 720
    },
    {
      "epoch": 1.8911917098445596,
      "grad_norm": 0.35952338576316833,
      "learning_rate": 0.00019015405027447777,
      "loss": 0.0448,
      "step": 730
    },
    {
      "epoch": 1.917098445595855,
      "grad_norm": 0.32625675201416016,
      "learning_rate": 0.00018978760888474397,
      "loss": 0.0456,
      "step": 740
    },
    {
      "epoch": 1.9430051813471503,
      "grad_norm": 0.37657493352890015,
      "learning_rate": 0.00018941483891850983,
      "loss": 0.0444,
      "step": 750
    },
    {
      "epoch": 1.9689119170984455,
      "grad_norm": 0.31168240308761597,
      "learning_rate": 0.00018903576665003832,
      "loss": 0.0418,
      "step": 760
    },
    {
      "epoch": 1.994818652849741,
      "grad_norm": 0.263838529586792,
      "learning_rate": 0.000188650418797803,
      "loss": 0.0458,
      "step": 770
    },
    {
      "epoch": 2.0207253886010363,
      "grad_norm": 0.32005709409713745,
      "learning_rate": 0.00018825882252260467,
      "loss": 0.0444,
      "step": 780
    },
    {
      "epoch": 2.0466321243523318,
      "grad_norm": 0.3012694716453552,
      "learning_rate": 0.0001878610054256571,
      "loss": 0.0366,
      "step": 790
    },
    {
      "epoch": 2.0725388601036268,
      "grad_norm": 0.2204688936471939,
      "learning_rate": 0.00018745699554664148,
      "loss": 0.0389,
      "step": 800
    },
    {
      "epoch": 2.098445595854922,
      "grad_norm": 0.28166186809539795,
      "learning_rate": 0.0001870468213617302,
      "loss": 0.0377,
      "step": 810
    },
    {
      "epoch": 2.1243523316062176,
      "grad_norm": 0.34278395771980286,
      "learning_rate": 0.00018663051178157965,
      "loss": 0.0385,
      "step": 820
    },
    {
      "epoch": 2.150259067357513,
      "grad_norm": 0.29428401589393616,
      "learning_rate": 0.00018620809614929243,
      "loss": 0.0427,
      "step": 830
    },
    {
      "epoch": 2.1761658031088085,
      "grad_norm": 0.3556409478187561,
      "learning_rate": 0.00018577960423834936,
      "loss": 0.0412,
      "step": 840
    },
    {
      "epoch": 2.2020725388601035,
      "grad_norm": 0.20572839677333832,
      "learning_rate": 0.00018534506625051067,
      "loss": 0.0386,
      "step": 850
    },
    {
      "epoch": 2.227979274611399,
      "grad_norm": 0.30463582277297974,
      "learning_rate": 0.0001849045128136874,
      "loss": 0.042,
      "step": 860
    },
    {
      "epoch": 2.2538860103626943,
      "grad_norm": 0.2660544514656067,
      "learning_rate": 0.00018445797497978266,
      "loss": 0.0382,
      "step": 870
    },
    {
      "epoch": 2.2797927461139897,
      "grad_norm": 0.2419629544019699,
      "learning_rate": 0.00018400548422250278,
      "loss": 0.0436,
      "step": 880
    },
    {
      "epoch": 2.305699481865285,
      "grad_norm": 0.2734587788581848,
      "learning_rate": 0.00018354707243513926,
      "loss": 0.0414,
      "step": 890
    },
    {
      "epoch": 2.33160621761658,
      "grad_norm": 0.3013145625591278,
      "learning_rate": 0.00018308277192832038,
      "loss": 0.0424,
      "step": 900
    },
    {
      "epoch": 2.3575129533678756,
      "grad_norm": 0.23408107459545135,
      "learning_rate": 0.0001826126154277343,
      "loss": 0.0381,
      "step": 910
    },
    {
      "epoch": 2.383419689119171,
      "grad_norm": 0.24882006645202637,
      "learning_rate": 0.0001821366360718219,
      "loss": 0.0361,
      "step": 920
    },
    {
      "epoch": 2.4093264248704664,
      "grad_norm": 0.3461746573448181,
      "learning_rate": 0.0001816548674094415,
      "loss": 0.0412,
      "step": 930
    },
    {
      "epoch": 2.4352331606217614,
      "grad_norm": 0.3645137548446655,
      "learning_rate": 0.00018116734339750407,
      "loss": 0.0411,
      "step": 940
    },
    {
      "epoch": 2.461139896373057,
      "grad_norm": 0.5230131149291992,
      "learning_rate": 0.0001806740983985797,
      "loss": 0.0413,
      "step": 950
    },
    {
      "epoch": 2.4870466321243523,
      "grad_norm": 0.48028966784477234,
      "learning_rate": 0.00018017516717847563,
      "loss": 0.0413,
      "step": 960
    },
    {
      "epoch": 2.5129533678756477,
      "grad_norm": 0.3212891221046448,
      "learning_rate": 0.000179670584903786,
      "loss": 0.0427,
      "step": 970
    },
    {
      "epoch": 2.538860103626943,
      "grad_norm": 0.3119347095489502,
      "learning_rate": 0.00017916038713941296,
      "loss": 0.0376,
      "step": 980
    },
    {
      "epoch": 2.5647668393782386,
      "grad_norm": 0.31051719188690186,
      "learning_rate": 0.00017864460984606,
      "loss": 0.0463,
      "step": 990
    },
    {
      "epoch": 2.5906735751295336,
      "grad_norm": 0.3582245111465454,
      "learning_rate": 0.0001781232893776974,
      "loss": 0.0351,
      "step": 1000
    },
    {
      "epoch": 2.616580310880829,
      "grad_norm": 0.28858813643455505,
      "learning_rate": 0.0001775964624789996,
      "loss": 0.0393,
      "step": 1010
    },
    {
      "epoch": 2.6424870466321244,
      "grad_norm": 0.21984124183654785,
      "learning_rate": 0.00017706416628275563,
      "loss": 0.0432,
      "step": 1020
    },
    {
      "epoch": 2.66839378238342,
      "grad_norm": 0.32943961024284363,
      "learning_rate": 0.0001765264383072515,
      "loss": 0.0354,
      "step": 1030
    },
    {
      "epoch": 2.694300518134715,
      "grad_norm": 0.16112935543060303,
      "learning_rate": 0.0001759833164536261,
      "loss": 0.035,
      "step": 1040
    },
    {
      "epoch": 2.7202072538860103,
      "grad_norm": 0.27633556723594666,
      "learning_rate": 0.0001754348390031996,
      "loss": 0.0344,
      "step": 1050
    },
    {
      "epoch": 2.7461139896373057,
      "grad_norm": 0.29028812050819397,
      "learning_rate": 0.0001748810446147751,
      "loss": 0.0332,
      "step": 1060
    },
    {
      "epoch": 2.772020725388601,
      "grad_norm": 0.268162339925766,
      "learning_rate": 0.00017432197232191424,
      "loss": 0.0381,
      "step": 1070
    },
    {
      "epoch": 2.7979274611398965,
      "grad_norm": 0.207249715924263,
      "learning_rate": 0.00017375766153018542,
      "loss": 0.0403,
      "step": 1080
    },
    {
      "epoch": 2.823834196891192,
      "grad_norm": 0.26709848642349243,
      "learning_rate": 0.00017318815201438682,
      "loss": 0.035,
      "step": 1090
    },
    {
      "epoch": 2.849740932642487,
      "grad_norm": 0.3294147253036499,
      "learning_rate": 0.00017261348391574258,
      "loss": 0.0353,
      "step": 1100
    },
    {
      "epoch": 2.8756476683937824,
      "grad_norm": 0.34221452474594116,
      "learning_rate": 0.00017203369773907375,
      "loss": 0.0323,
      "step": 1110
    },
    {
      "epoch": 2.901554404145078,
      "grad_norm": 0.23212362825870514,
      "learning_rate": 0.00017144883434994313,
      "loss": 0.0348,
      "step": 1120
    },
    {
      "epoch": 2.927461139896373,
      "grad_norm": 0.3442625105381012,
      "learning_rate": 0.00017085893497177502,
      "loss": 0.0384,
      "step": 1130
    },
    {
      "epoch": 2.9533678756476682,
      "grad_norm": 0.3061060607433319,
      "learning_rate": 0.0001702640411829497,
      "loss": 0.0364,
      "step": 1140
    },
    {
      "epoch": 2.9792746113989637,
      "grad_norm": 0.2969331443309784,
      "learning_rate": 0.00016966419491387267,
      "loss": 0.0345,
      "step": 1150
    },
    {
      "epoch": 3.005181347150259,
      "grad_norm": 0.2101631909608841,
      "learning_rate": 0.0001690594384440194,
      "loss": 0.0336,
      "step": 1160
    },
    {
      "epoch": 3.0310880829015545,
      "grad_norm": 0.2539927661418915,
      "learning_rate": 0.00016844981439895505,
      "loss": 0.0381,
      "step": 1170
    },
    {
      "epoch": 3.05699481865285,
      "grad_norm": 0.26166775822639465,
      "learning_rate": 0.00016783536574733047,
      "loss": 0.037,
      "step": 1180
    },
    {
      "epoch": 3.082901554404145,
      "grad_norm": 0.24203582108020782,
      "learning_rate": 0.00016721613579785317,
      "loss": 0.0323,
      "step": 1190
    },
    {
      "epoch": 3.1088082901554404,
      "grad_norm": 0.25355836749076843,
      "learning_rate": 0.000166592168196235,
      "loss": 0.0375,
      "step": 1200
    },
    {
      "epoch": 3.134715025906736,
      "grad_norm": 0.25193431973457336,
      "learning_rate": 0.0001659635069221159,
      "loss": 0.0308,
      "step": 1210
    },
    {
      "epoch": 3.160621761658031,
      "grad_norm": 0.3023054599761963,
      "learning_rate": 0.00016533019628596368,
      "loss": 0.0331,
      "step": 1220
    },
    {
      "epoch": 3.186528497409326,
      "grad_norm": 0.23343996703624725,
      "learning_rate": 0.00016469228092595137,
      "loss": 0.0322,
      "step": 1230
    },
    {
      "epoch": 3.2124352331606216,
      "grad_norm": 0.21372058987617493,
      "learning_rate": 0.00016404980580481042,
      "loss": 0.0328,
      "step": 1240
    },
    {
      "epoch": 3.238341968911917,
      "grad_norm": 0.23146772384643555,
      "learning_rate": 0.00016340281620666204,
      "loss": 0.032,
      "step": 1250
    },
    {
      "epoch": 3.2642487046632125,
      "grad_norm": 0.2798649072647095,
      "learning_rate": 0.00016275135773382494,
      "loss": 0.0338,
      "step": 1260
    },
    {
      "epoch": 3.290155440414508,
      "grad_norm": 0.1371963620185852,
      "learning_rate": 0.00016209547630360155,
      "loss": 0.0307,
      "step": 1270
    },
    {
      "epoch": 3.3160621761658033,
      "grad_norm": 0.3256252408027649,
      "learning_rate": 0.00016143521814504118,
      "loss": 0.0371,
      "step": 1280
    },
    {
      "epoch": 3.3419689119170983,
      "grad_norm": 0.23841211199760437,
      "learning_rate": 0.00016077062979568196,
      "loss": 0.0327,
      "step": 1290
    },
    {
      "epoch": 3.3678756476683938,
      "grad_norm": 0.19018740952014923,
      "learning_rate": 0.00016010175809827047,
      "loss": 0.0344,
      "step": 1300
    },
    {
      "epoch": 3.393782383419689,
      "grad_norm": 0.20630723237991333,
      "learning_rate": 0.00015942865019746014,
      "loss": 0.0326,
      "step": 1310
    },
    {
      "epoch": 3.4196891191709846,
      "grad_norm": 0.29622000455856323,
      "learning_rate": 0.0001587513535364884,
      "loss": 0.0287,
      "step": 1320
    },
    {
      "epoch": 3.4455958549222796,
      "grad_norm": 0.22663627564907074,
      "learning_rate": 0.00015806991585383257,
      "loss": 0.0347,
      "step": 1330
    },
    {
      "epoch": 3.471502590673575,
      "grad_norm": 0.1253904402256012,
      "learning_rate": 0.00015738438517984513,
      "loss": 0.0328,
      "step": 1340
    },
    {
      "epoch": 3.4974093264248705,
      "grad_norm": 0.3609561622142792,
      "learning_rate": 0.00015669480983336842,
      "loss": 0.0315,
      "step": 1350
    },
    {
      "epoch": 3.523316062176166,
      "grad_norm": 0.2557467222213745,
      "learning_rate": 0.0001560012384183288,
      "loss": 0.0314,
      "step": 1360
    },
    {
      "epoch": 3.5492227979274613,
      "grad_norm": 0.16117317974567413,
      "learning_rate": 0.00015530371982031097,
      "loss": 0.0317,
      "step": 1370
    },
    {
      "epoch": 3.5751295336787567,
      "grad_norm": 0.21727634966373444,
      "learning_rate": 0.0001546023032031123,
      "loss": 0.0295,
      "step": 1380
    },
    {
      "epoch": 3.6010362694300517,
      "grad_norm": 0.22786779701709747,
      "learning_rate": 0.00015389703800527756,
      "loss": 0.0307,
      "step": 1390
    },
    {
      "epoch": 3.626943005181347,
      "grad_norm": 0.28124475479125977,
      "learning_rate": 0.0001531879739366142,
      "loss": 0.0314,
      "step": 1400
    },
    {
      "epoch": 3.6528497409326426,
      "grad_norm": 0.20782004296779633,
      "learning_rate": 0.00015247516097468893,
      "loss": 0.0292,
      "step": 1410
    },
    {
      "epoch": 3.6787564766839376,
      "grad_norm": 0.23655350506305695,
      "learning_rate": 0.00015175864936130466,
      "loss": 0.0314,
      "step": 1420
    },
    {
      "epoch": 3.704663212435233,
      "grad_norm": 0.2071056067943573,
      "learning_rate": 0.00015103848959895978,
      "loss": 0.031,
      "step": 1430
    },
    {
      "epoch": 3.7305699481865284,
      "grad_norm": 0.2913356423377991,
      "learning_rate": 0.00015031473244728793,
      "loss": 0.0302,
      "step": 1440
    },
    {
      "epoch": 3.756476683937824,
      "grad_norm": 0.2031751126050949,
      "learning_rate": 0.00014958742891948092,
      "loss": 0.0317,
      "step": 1450
    },
    {
      "epoch": 3.7823834196891193,
      "grad_norm": 0.3000684380531311,
      "learning_rate": 0.0001488566302786926,
      "loss": 0.0343,
      "step": 1460
    },
    {
      "epoch": 3.8082901554404147,
      "grad_norm": 0.27631327509880066,
      "learning_rate": 0.00014812238803442598,
      "loss": 0.0317,
      "step": 1470
    },
    {
      "epoch": 3.8341968911917097,
      "grad_norm": 0.3157835602760315,
      "learning_rate": 0.00014738475393890235,
      "loss": 0.033,
      "step": 1480
    },
    {
      "epoch": 3.860103626943005,
      "grad_norm": 0.29543039202690125,
      "learning_rate": 0.00014664377998341391,
      "loss": 0.031,
      "step": 1490
    },
    {
      "epoch": 3.8860103626943006,
      "grad_norm": 0.21288472414016724,
      "learning_rate": 0.000145899518394659,
      "loss": 0.032,
      "step": 1500
    },
    {
      "epoch": 3.911917098445596,
      "grad_norm": 0.257420152425766,
      "learning_rate": 0.00014515202163106088,
      "loss": 0.0327,
      "step": 1510
    },
    {
      "epoch": 3.937823834196891,
      "grad_norm": 0.23928193747997284,
      "learning_rate": 0.00014440134237907065,
      "loss": 0.0299,
      "step": 1520
    },
    {
      "epoch": 3.9637305699481864,
      "grad_norm": 0.22903333604335785,
      "learning_rate": 0.0001436475335494532,
      "loss": 0.0325,
      "step": 1530
    },
    {
      "epoch": 3.989637305699482,
      "grad_norm": 0.2012854814529419,
      "learning_rate": 0.0001428906482735583,
      "loss": 0.0309,
      "step": 1540
    },
    {
      "epoch": 4.015544041450777,
      "grad_norm": 0.2412451058626175,
      "learning_rate": 0.00014213073989957533,
      "loss": 0.0313,
      "step": 1550
    },
    {
      "epoch": 4.041450777202073,
      "grad_norm": 0.22772116959095,
      "learning_rate": 0.0001413678619887734,
      "loss": 0.0334,
      "step": 1560
    },
    {
      "epoch": 4.067357512953368,
      "grad_norm": 0.22949784994125366,
      "learning_rate": 0.000140602068311726,
      "loss": 0.0308,
      "step": 1570
    },
    {
      "epoch": 4.0932642487046635,
      "grad_norm": 0.21911077201366425,
      "learning_rate": 0.00013983341284452093,
      "loss": 0.0263,
      "step": 1580
    },
    {
      "epoch": 4.119170984455959,
      "grad_norm": 0.19424454867839813,
      "learning_rate": 0.00013906194976495616,
      "loss": 0.0317,
      "step": 1590
    },
    {
      "epoch": 4.1450777202072535,
      "grad_norm": 0.23268303275108337,
      "learning_rate": 0.00013828773344872082,
      "loss": 0.0306,
      "step": 1600
    },
    {
      "epoch": 4.170984455958549,
      "grad_norm": 0.253647118806839,
      "learning_rate": 0.000137510818465563,
      "loss": 0.029,
      "step": 1610
    },
    {
      "epoch": 4.196891191709844,
      "grad_norm": 0.15061113238334656,
      "learning_rate": 0.00013673125957544286,
      "loss": 0.0248,
      "step": 1620
    },
    {
      "epoch": 4.22279792746114,
      "grad_norm": 0.2819027602672577,
      "learning_rate": 0.00013594911172467374,
      "loss": 0.0308,
      "step": 1630
    },
    {
      "epoch": 4.248704663212435,
      "grad_norm": 0.24925965070724487,
      "learning_rate": 0.00013516443004204853,
      "loss": 0.027,
      "step": 1640
    },
    {
      "epoch": 4.274611398963731,
      "grad_norm": 0.2695538401603699,
      "learning_rate": 0.00013437726983495462,
      "loss": 0.0286,
      "step": 1650
    },
    {
      "epoch": 4.300518134715026,
      "grad_norm": 0.21748502552509308,
      "learning_rate": 0.00013358768658547517,
      "loss": 0.0263,
      "step": 1660
    },
    {
      "epoch": 4.3264248704663215,
      "grad_norm": 0.19919273257255554,
      "learning_rate": 0.00013279573594647873,
      "loss": 0.0286,
      "step": 1670
    },
    {
      "epoch": 4.352331606217617,
      "grad_norm": 0.29898253083229065,
      "learning_rate": 0.0001320014737376967,
      "loss": 0.0299,
      "step": 1680
    },
    {
      "epoch": 4.3782383419689115,
      "grad_norm": 0.26166197657585144,
      "learning_rate": 0.00013120495594178867,
      "loss": 0.0306,
      "step": 1690
    },
    {
      "epoch": 4.404145077720207,
      "grad_norm": 0.23769961297512054,
      "learning_rate": 0.0001304062387003968,
      "loss": 0.0333,
      "step": 1700
    },
    {
      "epoch": 4.430051813471502,
      "grad_norm": 0.2202085703611374,
      "learning_rate": 0.00012960537831018863,
      "loss": 0.032,
      "step": 1710
    },
    {
      "epoch": 4.455958549222798,
      "grad_norm": 0.2457539141178131,
      "learning_rate": 0.00012880243121888898,
      "loss": 0.0262,
      "step": 1720
    },
    {
      "epoch": 4.481865284974093,
      "grad_norm": 0.2208835929632187,
      "learning_rate": 0.0001279974540213015,
      "loss": 0.0311,
      "step": 1730
    },
    {
      "epoch": 4.507772020725389,
      "grad_norm": 0.19063198566436768,
      "learning_rate": 0.0001271905034553196,
      "loss": 0.0269,
      "step": 1740
    },
    {
      "epoch": 4.533678756476684,
      "grad_norm": 0.2229800969362259,
      "learning_rate": 0.00012638163639792715,
      "loss": 0.0292,
      "step": 1750
    },
    {
      "epoch": 4.5595854922279795,
      "grad_norm": 0.16816496849060059,
      "learning_rate": 0.0001255709098611898,
      "loss": 0.0266,
      "step": 1760
    },
    {
      "epoch": 4.585492227979275,
      "grad_norm": 0.22993479669094086,
      "learning_rate": 0.0001247583809882365,
      "loss": 0.0266,
      "step": 1770
    },
    {
      "epoch": 4.61139896373057,
      "grad_norm": 0.2233438789844513,
      "learning_rate": 0.00012394410704923167,
      "loss": 0.0303,
      "step": 1780
    },
    {
      "epoch": 4.637305699481866,
      "grad_norm": 0.2839880883693695,
      "learning_rate": 0.0001231281454373389,
      "loss": 0.0298,
      "step": 1790
    },
    {
      "epoch": 4.66321243523316,
      "grad_norm": 0.22099633514881134,
      "learning_rate": 0.0001223105536646752,
      "loss": 0.0302,
      "step": 1800
    },
    {
      "epoch": 4.689119170984456,
      "grad_norm": 0.18007872998714447,
      "learning_rate": 0.00012149138935825792,
      "loss": 0.0242,
      "step": 1810
    },
    {
      "epoch": 4.715025906735751,
      "grad_norm": 0.2904217541217804,
      "learning_rate": 0.00012067071025594234,
      "loss": 0.0284,
      "step": 1820
    },
    {
      "epoch": 4.740932642487047,
      "grad_norm": 0.2809178829193115,
      "learning_rate": 0.00011984857420235263,
      "loss": 0.0277,
      "step": 1830
    },
    {
      "epoch": 4.766839378238342,
      "grad_norm": 0.23288775980472565,
      "learning_rate": 0.00011902503914480428,
      "loss": 0.0284,
      "step": 1840
    },
    {
      "epoch": 4.7927461139896375,
      "grad_norm": 0.1514245569705963,
      "learning_rate": 0.00011820016312922015,
      "loss": 0.031,
      "step": 1850
    },
    {
      "epoch": 4.818652849740933,
      "grad_norm": 0.12524273991584778,
      "learning_rate": 0.00011737400429603888,
      "loss": 0.031,
      "step": 1860
    },
    {
      "epoch": 4.844559585492228,
      "grad_norm": 0.22305960953235626,
      "learning_rate": 0.00011654662087611707,
      "loss": 0.0294,
      "step": 1870
    },
    {
      "epoch": 4.870466321243523,
      "grad_norm": 0.2521384358406067,
      "learning_rate": 0.00011571807118662492,
      "loss": 0.0276,
      "step": 1880
    },
    {
      "epoch": 4.896373056994818,
      "grad_norm": 0.180227592587471,
      "learning_rate": 0.00011488841362693576,
      "loss": 0.0289,
      "step": 1890
    },
    {
      "epoch": 4.922279792746114,
      "grad_norm": 0.22024022042751312,
      "learning_rate": 0.00011405770667450996,
      "loss": 0.0266,
      "step": 1900
    },
    {
      "epoch": 4.948186528497409,
      "grad_norm": 0.19573189318180084,
      "learning_rate": 0.00011322600888077303,
      "loss": 0.0282,
      "step": 1910
    },
    {
      "epoch": 4.974093264248705,
      "grad_norm": 0.1752632111310959,
      "learning_rate": 0.00011239337886698887,
      "loss": 0.0269,
      "step": 1920
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.2073243260383606,
      "learning_rate": 0.00011155987532012779,
      "loss": 0.0268,
      "step": 1930
    },
    {
      "epoch": 5.025906735751295,
      "grad_norm": 0.21750442683696747,
      "learning_rate": 0.00011072555698873013,
      "loss": 0.0249,
      "step": 1940
    },
    {
      "epoch": 5.051813471502591,
      "grad_norm": 0.1471615731716156,
      "learning_rate": 0.00010989048267876529,
      "loss": 0.0231,
      "step": 1950
    },
    {
      "epoch": 5.077720207253886,
      "grad_norm": 0.21576045453548431,
      "learning_rate": 0.00010905471124948707,
      "loss": 0.0245,
      "step": 1960
    },
    {
      "epoch": 5.103626943005182,
      "grad_norm": 0.1472284197807312,
      "learning_rate": 0.00010821830160928486,
      "loss": 0.0247,
      "step": 1970
    },
    {
      "epoch": 5.129533678756476,
      "grad_norm": 0.354701429605484,
      "learning_rate": 0.00010738131271153156,
      "loss": 0.0255,
      "step": 1980
    },
    {
      "epoch": 5.155440414507772,
      "grad_norm": 0.16570764780044556,
      "learning_rate": 0.00010654380355042852,
      "loss": 0.0236,
      "step": 1990
    },
    {
      "epoch": 5.181347150259067,
      "grad_norm": 0.20196418464183807,
      "learning_rate": 0.00010570583315684708,
      "loss": 0.0278,
      "step": 2000
    },
    {
      "epoch": 5.2072538860103625,
      "grad_norm": 0.25251758098602295,
      "learning_rate": 0.00010486746059416816,
      "loss": 0.0258,
      "step": 2010
    },
    {
      "epoch": 5.233160621761658,
      "grad_norm": 0.2057216763496399,
      "learning_rate": 0.00010402874495411902,
      "loss": 0.0287,
      "step": 2020
    },
    {
      "epoch": 5.259067357512953,
      "grad_norm": 0.27119141817092896,
      "learning_rate": 0.00010318974535260844,
      "loss": 0.0252,
      "step": 2030
    },
    {
      "epoch": 5.284974093264249,
      "grad_norm": 0.3158484995365143,
      "learning_rate": 0.00010235052092555972,
      "loss": 0.0254,
      "step": 2040
    },
    {
      "epoch": 5.310880829015544,
      "grad_norm": 0.2271166890859604,
      "learning_rate": 0.00010151113082474295,
      "loss": 0.0276,
      "step": 2050
    },
    {
      "epoch": 5.33678756476684,
      "grad_norm": 0.28330063819885254,
      "learning_rate": 0.00010067163421360535,
      "loss": 0.025,
      "step": 2060
    },
    {
      "epoch": 5.362694300518135,
      "grad_norm": 0.1780189722776413,
      "learning_rate": 9.983209026310149e-05,
      "loss": 0.0261,
      "step": 2070
    },
    {
      "epoch": 5.38860103626943,
      "grad_norm": 0.22980116307735443,
      "learning_rate": 9.899255814752263e-05,
      "loss": 0.0224,
      "step": 2080
    },
    {
      "epoch": 5.414507772020725,
      "grad_norm": 0.21145443618297577,
      "learning_rate": 9.815309704032583e-05,
      "loss": 0.0276,
      "step": 2090
    },
    {
      "epoch": 5.4404145077720205,
      "grad_norm": 0.18889708817005157,
      "learning_rate": 9.731376610996313e-05,
      "loss": 0.0284,
      "step": 2100
    },
    {
      "epoch": 5.466321243523316,
      "grad_norm": 0.1767049878835678,
      "learning_rate": 9.647462451571127e-05,
      "loss": 0.0253,
      "step": 2110
    },
    {
      "epoch": 5.492227979274611,
      "grad_norm": 0.16297735273838043,
      "learning_rate": 9.563573140350192e-05,
      "loss": 0.0232,
      "step": 2120
    },
    {
      "epoch": 5.518134715025907,
      "grad_norm": 0.13897372782230377,
      "learning_rate": 9.47971459017527e-05,
      "loss": 0.0252,
      "step": 2130
    },
    {
      "epoch": 5.544041450777202,
      "grad_norm": 0.2665773630142212,
      "learning_rate": 9.395892711719962e-05,
      "loss": 0.0234,
      "step": 2140
    },
    {
      "epoch": 5.569948186528498,
      "grad_norm": 0.2184874713420868,
      "learning_rate": 9.312113413073113e-05,
      "loss": 0.0226,
      "step": 2150
    },
    {
      "epoch": 5.595854922279793,
      "grad_norm": 0.24830220639705658,
      "learning_rate": 9.228382599322374e-05,
      "loss": 0.0232,
      "step": 2160
    },
    {
      "epoch": 5.6217616580310885,
      "grad_norm": 0.1857186257839203,
      "learning_rate": 9.144706172138002e-05,
      "loss": 0.0204,
      "step": 2170
    },
    {
      "epoch": 5.647668393782383,
      "grad_norm": 0.18585319817066193,
      "learning_rate": 9.061090029356867e-05,
      "loss": 0.0208,
      "step": 2180
    },
    {
      "epoch": 5.6735751295336785,
      "grad_norm": 0.19226165115833282,
      "learning_rate": 8.977540064566771e-05,
      "loss": 0.0225,
      "step": 2190
    },
    {
      "epoch": 5.699481865284974,
      "grad_norm": 0.2323295623064041,
      "learning_rate": 8.89406216669103e-05,
      "loss": 0.0276,
      "step": 2200
    },
    {
      "epoch": 5.725388601036269,
      "grad_norm": 0.17143680155277252,
      "learning_rate": 8.810662219573421e-05,
      "loss": 0.0284,
      "step": 2210
    },
    {
      "epoch": 5.751295336787565,
      "grad_norm": 0.21579648554325104,
      "learning_rate": 8.727346101563425e-05,
      "loss": 0.0262,
      "step": 2220
    },
    {
      "epoch": 5.77720207253886,
      "grad_norm": 0.21625088155269623,
      "learning_rate": 8.64411968510195e-05,
      "loss": 0.0236,
      "step": 2230
    },
    {
      "epoch": 5.803108808290156,
      "grad_norm": 0.25434359908103943,
      "learning_rate": 8.56098883630738e-05,
      "loss": 0.0234,
      "step": 2240
    },
    {
      "epoch": 5.829015544041451,
      "grad_norm": 0.26135268807411194,
      "learning_rate": 8.477959414562126e-05,
      "loss": 0.024,
      "step": 2250
    },
    {
      "epoch": 5.8549222797927465,
      "grad_norm": 0.2323196679353714,
      "learning_rate": 8.395037272099629e-05,
      "loss": 0.0274,
      "step": 2260
    },
    {
      "epoch": 5.880829015544041,
      "grad_norm": 0.17564794421195984,
      "learning_rate": 8.312228253591867e-05,
      "loss": 0.024,
      "step": 2270
    },
    {
      "epoch": 5.9067357512953365,
      "grad_norm": 0.1802438646554947,
      "learning_rate": 8.22953819573742e-05,
      "loss": 0.0224,
      "step": 2280
    },
    {
      "epoch": 5.932642487046632,
      "grad_norm": 0.2152315080165863,
      "learning_rate": 8.146972926850054e-05,
      "loss": 0.0234,
      "step": 2290
    },
    {
      "epoch": 5.958549222797927,
      "grad_norm": 0.2047027349472046,
      "learning_rate": 8.064538266447926e-05,
      "loss": 0.0235,
      "step": 2300
    },
    {
      "epoch": 5.984455958549223,
      "grad_norm": 0.20742332935333252,
      "learning_rate": 7.982240024843404e-05,
      "loss": 0.027,
      "step": 2310
    },
    {
      "epoch": 6.010362694300518,
      "grad_norm": 0.15848736464977264,
      "learning_rate": 7.900084002733541e-05,
      "loss": 0.0216,
      "step": 2320
    },
    {
      "epoch": 6.036269430051814,
      "grad_norm": 0.20834903419017792,
      "learning_rate": 7.818075990791208e-05,
      "loss": 0.0276,
      "step": 2330
    },
    {
      "epoch": 6.062176165803109,
      "grad_norm": 0.14767374098300934,
      "learning_rate": 7.736221769256946e-05,
      "loss": 0.0223,
      "step": 2340
    },
    {
      "epoch": 6.0880829015544045,
      "grad_norm": 0.1815556287765503,
      "learning_rate": 7.654527107531559e-05,
      "loss": 0.0233,
      "step": 2350
    },
    {
      "epoch": 6.1139896373057,
      "grad_norm": 0.33385881781578064,
      "learning_rate": 7.572997763769459e-05,
      "loss": 0.0227,
      "step": 2360
    },
    {
      "epoch": 6.139896373056994,
      "grad_norm": 0.1571478247642517,
      "learning_rate": 7.491639484472819e-05,
      "loss": 0.0214,
      "step": 2370
    },
    {
      "epoch": 6.16580310880829,
      "grad_norm": 0.1373177021741867,
      "learning_rate": 7.410458004086519e-05,
      "loss": 0.0209,
      "step": 2380
    },
    {
      "epoch": 6.191709844559585,
      "grad_norm": 0.26975080370903015,
      "learning_rate": 7.329459044593976e-05,
      "loss": 0.0243,
      "step": 2390
    },
    {
      "epoch": 6.217616580310881,
      "grad_norm": 0.18673855066299438,
      "learning_rate": 7.248648315113829e-05,
      "loss": 0.0201,
      "step": 2400
    },
    {
      "epoch": 6.243523316062176,
      "grad_norm": 0.1592017412185669,
      "learning_rate": 7.16803151149755e-05,
      "loss": 0.0206,
      "step": 2410
    },
    {
      "epoch": 6.269430051813472,
      "grad_norm": 0.22650600969791412,
      "learning_rate": 7.08761431592795e-05,
      "loss": 0.0272,
      "step": 2420
    },
    {
      "epoch": 6.295336787564767,
      "grad_norm": 0.21501009166240692,
      "learning_rate": 7.007402396518717e-05,
      "loss": 0.0183,
      "step": 2430
    },
    {
      "epoch": 6.321243523316062,
      "grad_norm": 0.2062009572982788,
      "learning_rate": 6.927401406914872e-05,
      "loss": 0.0229,
      "step": 2440
    },
    {
      "epoch": 6.347150259067358,
      "grad_norm": 0.1995324045419693,
      "learning_rate": 6.847616985894296e-05,
      "loss": 0.0188,
      "step": 2450
    },
    {
      "epoch": 6.373056994818652,
      "grad_norm": 0.18677528202533722,
      "learning_rate": 6.768054756970279e-05,
      "loss": 0.0196,
      "step": 2460
    },
    {
      "epoch": 6.398963730569948,
      "grad_norm": 0.20209306478500366,
      "learning_rate": 6.688720327995153e-05,
      "loss": 0.0199,
      "step": 2470
    },
    {
      "epoch": 6.424870466321243,
      "grad_norm": 0.12447565048933029,
      "learning_rate": 6.609619290765045e-05,
      "loss": 0.0222,
      "step": 2480
    },
    {
      "epoch": 6.450777202072539,
      "grad_norm": 0.23713508248329163,
      "learning_rate": 6.530757220625723e-05,
      "loss": 0.0263,
      "step": 2490
    },
    {
      "epoch": 6.476683937823834,
      "grad_norm": 0.17403320968151093,
      "learning_rate": 6.452139676079631e-05,
      "loss": 0.022,
      "step": 2500
    },
    {
      "epoch": 6.5025906735751295,
      "grad_norm": 0.19337254762649536,
      "learning_rate": 6.373772198394114e-05,
      "loss": 0.0199,
      "step": 2510
    },
    {
      "epoch": 6.528497409326425,
      "grad_norm": 0.1531633585691452,
      "learning_rate": 6.295660311210843e-05,
      "loss": 0.022,
      "step": 2520
    },
    {
      "epoch": 6.55440414507772,
      "grad_norm": 0.23906753957271576,
      "learning_rate": 6.217809520156481e-05,
      "loss": 0.0203,
      "step": 2530
    },
    {
      "epoch": 6.580310880829016,
      "grad_norm": 0.1223067194223404,
      "learning_rate": 6.140225312454629e-05,
      "loss": 0.0208,
      "step": 2540
    },
    {
      "epoch": 6.606217616580311,
      "grad_norm": 0.18273130059242249,
      "learning_rate": 6.0629131565390715e-05,
      "loss": 0.0238,
      "step": 2550
    },
    {
      "epoch": 6.632124352331607,
      "grad_norm": 0.2462676465511322,
      "learning_rate": 5.9858785016683306e-05,
      "loss": 0.025,
      "step": 2560
    },
    {
      "epoch": 6.658031088082901,
      "grad_norm": 0.2025512307882309,
      "learning_rate": 5.909126777541596e-05,
      "loss": 0.0199,
      "step": 2570
    },
    {
      "epoch": 6.683937823834197,
      "grad_norm": 0.18922774493694305,
      "learning_rate": 5.832663393915991e-05,
      "loss": 0.0193,
      "step": 2580
    },
    {
      "epoch": 6.709844559585492,
      "grad_norm": 0.2345733493566513,
      "learning_rate": 5.756493740225306e-05,
      "loss": 0.0186,
      "step": 2590
    },
    {
      "epoch": 6.7357512953367875,
      "grad_norm": 0.17045198380947113,
      "learning_rate": 5.6806231852001104e-05,
      "loss": 0.0248,
      "step": 2600
    },
    {
      "epoch": 6.761658031088083,
      "grad_norm": 0.23411734402179718,
      "learning_rate": 5.605057076489342e-05,
      "loss": 0.0213,
      "step": 2610
    },
    {
      "epoch": 6.787564766839378,
      "grad_norm": 0.19447815418243408,
      "learning_rate": 5.5298007402833905e-05,
      "loss": 0.0217,
      "step": 2620
    },
    {
      "epoch": 6.813471502590674,
      "grad_norm": 0.17666564881801605,
      "learning_rate": 5.454859480938692e-05,
      "loss": 0.0194,
      "step": 2630
    },
    {
      "epoch": 6.839378238341969,
      "grad_norm": 0.193482905626297,
      "learning_rate": 5.3802385806038535e-05,
      "loss": 0.0229,
      "step": 2640
    },
    {
      "epoch": 6.865284974093264,
      "grad_norm": 0.16288697719573975,
      "learning_rate": 5.305943298847336e-05,
      "loss": 0.0208,
      "step": 2650
    },
    {
      "epoch": 6.891191709844559,
      "grad_norm": 0.2407398670911789,
      "learning_rate": 5.23197887228675e-05,
      "loss": 0.0202,
      "step": 2660
    },
    {
      "epoch": 6.917098445595855,
      "grad_norm": 0.20776060223579407,
      "learning_rate": 5.1583505142197694e-05,
      "loss": 0.0207,
      "step": 2670
    },
    {
      "epoch": 6.94300518134715,
      "grad_norm": 0.21267752349376678,
      "learning_rate": 5.08506341425665e-05,
      "loss": 0.0199,
      "step": 2680
    },
    {
      "epoch": 6.9689119170984455,
      "grad_norm": 0.21812330186367035,
      "learning_rate": 5.0121227379544775e-05,
      "loss": 0.0215,
      "step": 2690
    },
    {
      "epoch": 6.994818652849741,
      "grad_norm": 0.19859305024147034,
      "learning_rate": 4.9395336264530486e-05,
      "loss": 0.0202,
      "step": 2700
    },
    {
      "epoch": 7.020725388601036,
      "grad_norm": 0.12451327592134476,
      "learning_rate": 4.8673011961125334e-05,
      "loss": 0.0167,
      "step": 2710
    },
    {
      "epoch": 7.046632124352332,
      "grad_norm": 0.23396964371204376,
      "learning_rate": 4.7954305381528254e-05,
      "loss": 0.0178,
      "step": 2720
    },
    {
      "epoch": 7.072538860103627,
      "grad_norm": 0.19071508944034576,
      "learning_rate": 4.723926718294722e-05,
      "loss": 0.022,
      "step": 2730
    },
    {
      "epoch": 7.098445595854923,
      "grad_norm": 0.17534908652305603,
      "learning_rate": 4.6527947764028456e-05,
      "loss": 0.0158,
      "step": 2740
    },
    {
      "epoch": 7.124352331606218,
      "grad_norm": 0.17512699961662292,
      "learning_rate": 4.582039726130437e-05,
      "loss": 0.0172,
      "step": 2750
    },
    {
      "epoch": 7.150259067357513,
      "grad_norm": 0.14806455373764038,
      "learning_rate": 4.511666554565955e-05,
      "loss": 0.0178,
      "step": 2760
    },
    {
      "epoch": 7.176165803108808,
      "grad_norm": 0.17297431826591492,
      "learning_rate": 4.441680221881587e-05,
      "loss": 0.0206,
      "step": 2770
    },
    {
      "epoch": 7.2020725388601035,
      "grad_norm": 0.16674122214317322,
      "learning_rate": 4.372085660983616e-05,
      "loss": 0.0163,
      "step": 2780
    },
    {
      "epoch": 7.227979274611399,
      "grad_norm": 0.219289630651474,
      "learning_rate": 4.302887777164742e-05,
      "loss": 0.0185,
      "step": 2790
    },
    {
      "epoch": 7.253886010362694,
      "grad_norm": 0.15512730181217194,
      "learning_rate": 4.2340914477583425e-05,
      "loss": 0.0177,
      "step": 2800
    },
    {
      "epoch": 7.27979274611399,
      "grad_norm": 0.23912139236927032,
      "learning_rate": 4.1657015217946936e-05,
      "loss": 0.0188,
      "step": 2810
    },
    {
      "epoch": 7.305699481865285,
      "grad_norm": 0.20765544474124908,
      "learning_rate": 4.097722819659184e-05,
      "loss": 0.0185,
      "step": 2820
    },
    {
      "epoch": 7.331606217616581,
      "grad_norm": 0.17259199917316437,
      "learning_rate": 4.030160132752565e-05,
      "loss": 0.017,
      "step": 2830
    },
    {
      "epoch": 7.357512953367876,
      "grad_norm": 0.1474766880273819,
      "learning_rate": 3.9630182231532366e-05,
      "loss": 0.0188,
      "step": 2840
    },
    {
      "epoch": 7.383419689119171,
      "grad_norm": 0.16211964190006256,
      "learning_rate": 3.896301823281595e-05,
      "loss": 0.0192,
      "step": 2850
    },
    {
      "epoch": 7.409326424870466,
      "grad_norm": 0.1418692171573639,
      "learning_rate": 3.8300156355664665e-05,
      "loss": 0.0188,
      "step": 2860
    },
    {
      "epoch": 7.435233160621761,
      "grad_norm": 0.16895586252212524,
      "learning_rate": 3.7641643321136624e-05,
      "loss": 0.0205,
      "step": 2870
    },
    {
      "epoch": 7.461139896373057,
      "grad_norm": 0.22076182067394257,
      "learning_rate": 3.698752554376689e-05,
      "loss": 0.0178,
      "step": 2880
    },
    {
      "epoch": 7.487046632124352,
      "grad_norm": 0.15769238770008087,
      "learning_rate": 3.633784912829586e-05,
      "loss": 0.0189,
      "step": 2890
    },
    {
      "epoch": 7.512953367875648,
      "grad_norm": 0.27454307675361633,
      "learning_rate": 3.5692659866419484e-05,
      "loss": 0.0191,
      "step": 2900
    },
    {
      "epoch": 7.538860103626943,
      "grad_norm": 0.16451720893383026,
      "learning_rate": 3.5052003233562e-05,
      "loss": 0.0172,
      "step": 2910
    },
    {
      "epoch": 7.564766839378239,
      "grad_norm": 0.14524053037166595,
      "learning_rate": 3.4415924385670495e-05,
      "loss": 0.0214,
      "step": 2920
    },
    {
      "epoch": 7.590673575129534,
      "grad_norm": 0.25439292192459106,
      "learning_rate": 3.378446815603205e-05,
      "loss": 0.0198,
      "step": 2930
    },
    {
      "epoch": 7.616580310880829,
      "grad_norm": 0.2044086903333664,
      "learning_rate": 3.31576790521138e-05,
      "loss": 0.0181,
      "step": 2940
    },
    {
      "epoch": 7.642487046632124,
      "grad_norm": 0.13481000065803528,
      "learning_rate": 3.253560125242597e-05,
      "loss": 0.0197,
      "step": 2950
    },
    {
      "epoch": 7.668393782383419,
      "grad_norm": 0.13905170559883118,
      "learning_rate": 3.191827860340794e-05,
      "loss": 0.0161,
      "step": 2960
    },
    {
      "epoch": 7.694300518134715,
      "grad_norm": 0.17516188323497772,
      "learning_rate": 3.1305754616337624e-05,
      "loss": 0.0182,
      "step": 2970
    },
    {
      "epoch": 7.72020725388601,
      "grad_norm": 0.1584802269935608,
      "learning_rate": 3.06980724642648e-05,
      "loss": 0.0171,
      "step": 2980
    },
    {
      "epoch": 7.746113989637306,
      "grad_norm": 0.10378487408161163,
      "learning_rate": 3.009527497896817e-05,
      "loss": 0.0158,
      "step": 2990
    },
    {
      "epoch": 7.772020725388601,
      "grad_norm": 0.22542284429073334,
      "learning_rate": 2.949740464793621e-05,
      "loss": 0.0172,
      "step": 3000
    },
    {
      "epoch": 7.7979274611398965,
      "grad_norm": 0.1301916539669037,
      "learning_rate": 2.8904503611372612e-05,
      "loss": 0.0165,
      "step": 3010
    },
    {
      "epoch": 7.823834196891192,
      "grad_norm": 0.17367781698703766,
      "learning_rate": 2.8316613659226022e-05,
      "loss": 0.0153,
      "step": 3020
    },
    {
      "epoch": 7.849740932642487,
      "grad_norm": 0.16101326048374176,
      "learning_rate": 2.773377622824459e-05,
      "loss": 0.0147,
      "step": 3030
    },
    {
      "epoch": 7.875647668393782,
      "grad_norm": 0.15740449726581573,
      "learning_rate": 2.7156032399055375e-05,
      "loss": 0.0183,
      "step": 3040
    },
    {
      "epoch": 7.901554404145077,
      "grad_norm": 0.1986459195613861,
      "learning_rate": 2.6583422893268638e-05,
      "loss": 0.015,
      "step": 3050
    },
    {
      "epoch": 7.927461139896373,
      "grad_norm": 0.1751032918691635,
      "learning_rate": 2.601598807060779e-05,
      "loss": 0.0171,
      "step": 3060
    },
    {
      "epoch": 7.953367875647668,
      "grad_norm": 0.15115486085414886,
      "learning_rate": 2.545376792606473e-05,
      "loss": 0.0177,
      "step": 3070
    },
    {
      "epoch": 7.979274611398964,
      "grad_norm": 0.15944965183734894,
      "learning_rate": 2.489680208708063e-05,
      "loss": 0.0157,
      "step": 3080
    },
    {
      "epoch": 8.005181347150259,
      "grad_norm": 0.15558196604251862,
      "learning_rate": 2.4345129810753087e-05,
      "loss": 0.0182,
      "step": 3090
    },
    {
      "epoch": 8.031088082901555,
      "grad_norm": 0.21775135397911072,
      "learning_rate": 2.3798789981068914e-05,
      "loss": 0.0187,
      "step": 3100
    },
    {
      "epoch": 8.05699481865285,
      "grad_norm": 0.19147688150405884,
      "learning_rate": 2.3257821106163634e-05,
      "loss": 0.015,
      "step": 3110
    },
    {
      "epoch": 8.082901554404145,
      "grad_norm": 0.2634642720222473,
      "learning_rate": 2.2722261315607106e-05,
      "loss": 0.0157,
      "step": 3120
    },
    {
      "epoch": 8.10880829015544,
      "grad_norm": 0.14185205101966858,
      "learning_rate": 2.2192148357716168e-05,
      "loss": 0.0157,
      "step": 3130
    },
    {
      "epoch": 8.134715025906736,
      "grad_norm": 0.19837462902069092,
      "learning_rate": 2.1667519596893836e-05,
      "loss": 0.0171,
      "step": 3140
    },
    {
      "epoch": 8.160621761658032,
      "grad_norm": 0.21308404207229614,
      "learning_rate": 2.114841201099588e-05,
      "loss": 0.0171,
      "step": 3150
    },
    {
      "epoch": 8.186528497409327,
      "grad_norm": 0.16102993488311768,
      "learning_rate": 2.06348621887243e-05,
      "loss": 0.014,
      "step": 3160
    },
    {
      "epoch": 8.212435233160623,
      "grad_norm": 0.1218574196100235,
      "learning_rate": 2.012690632704861e-05,
      "loss": 0.0154,
      "step": 3170
    },
    {
      "epoch": 8.238341968911918,
      "grad_norm": 0.17196378111839294,
      "learning_rate": 1.9624580228654366e-05,
      "loss": 0.014,
      "step": 3180
    },
    {
      "epoch": 8.264248704663212,
      "grad_norm": 0.16071684658527374,
      "learning_rate": 1.9127919299419705e-05,
      "loss": 0.0162,
      "step": 3190
    },
    {
      "epoch": 8.290155440414507,
      "grad_norm": 0.151749387383461,
      "learning_rate": 1.8636958545919904e-05,
      "loss": 0.016,
      "step": 3200
    },
    {
      "epoch": 8.316062176165802,
      "grad_norm": 0.16748680174350739,
      "learning_rate": 1.8151732572959857e-05,
      "loss": 0.017,
      "step": 3210
    },
    {
      "epoch": 8.341968911917098,
      "grad_norm": 0.09705256670713425,
      "learning_rate": 1.7672275581135023e-05,
      "loss": 0.014,
      "step": 3220
    },
    {
      "epoch": 8.367875647668393,
      "grad_norm": 0.17474466562271118,
      "learning_rate": 1.719862136442083e-05,
      "loss": 0.0132,
      "step": 3230
    },
    {
      "epoch": 8.393782383419689,
      "grad_norm": 0.12089501321315765,
      "learning_rate": 1.6730803307790842e-05,
      "loss": 0.0162,
      "step": 3240
    },
    {
      "epoch": 8.419689119170984,
      "grad_norm": 0.1267276108264923,
      "learning_rate": 1.626885438486355e-05,
      "loss": 0.0151,
      "step": 3250
    },
    {
      "epoch": 8.44559585492228,
      "grad_norm": 0.17889542877674103,
      "learning_rate": 1.581280715557827e-05,
      "loss": 0.0181,
      "step": 3260
    },
    {
      "epoch": 8.471502590673575,
      "grad_norm": 0.17118024826049805,
      "learning_rate": 1.5362693763900193e-05,
      "loss": 0.0178,
      "step": 3270
    },
    {
      "epoch": 8.49740932642487,
      "grad_norm": 0.13716407120227814,
      "learning_rate": 1.4918545935554872e-05,
      "loss": 0.015,
      "step": 3280
    },
    {
      "epoch": 8.523316062176166,
      "grad_norm": 0.17024938762187958,
      "learning_rate": 1.4480394975791955e-05,
      "loss": 0.0206,
      "step": 3290
    },
    {
      "epoch": 8.549222797927461,
      "grad_norm": 0.137301504611969,
      "learning_rate": 1.4048271767178579e-05,
      "loss": 0.0161,
      "step": 3300
    },
    {
      "epoch": 8.575129533678757,
      "grad_norm": 0.12056181579828262,
      "learning_rate": 1.3622206767422907e-05,
      "loss": 0.0156,
      "step": 3310
    },
    {
      "epoch": 8.601036269430052,
      "grad_norm": 0.228641539812088,
      "learning_rate": 1.3202230007227179e-05,
      "loss": 0.0182,
      "step": 3320
    },
    {
      "epoch": 8.626943005181348,
      "grad_norm": 0.14743590354919434,
      "learning_rate": 1.278837108817098e-05,
      "loss": 0.0139,
      "step": 3330
    },
    {
      "epoch": 8.652849740932643,
      "grad_norm": 0.18240438401699066,
      "learning_rate": 1.2380659180624998e-05,
      "loss": 0.0154,
      "step": 3340
    },
    {
      "epoch": 8.678756476683938,
      "grad_norm": 0.18438772857189178,
      "learning_rate": 1.1979123021694782e-05,
      "loss": 0.0137,
      "step": 3350
    },
    {
      "epoch": 8.704663212435234,
      "grad_norm": 0.1203690692782402,
      "learning_rate": 1.1583790913195402e-05,
      "loss": 0.0146,
      "step": 3360
    },
    {
      "epoch": 8.73056994818653,
      "grad_norm": 0.11229782551527023,
      "learning_rate": 1.1194690719656486e-05,
      "loss": 0.0157,
      "step": 3370
    },
    {
      "epoch": 8.756476683937823,
      "grad_norm": 0.15702977776527405,
      "learning_rate": 1.0811849866358393e-05,
      "loss": 0.0134,
      "step": 3380
    },
    {
      "epoch": 8.782383419689118,
      "grad_norm": 0.11596658080816269,
      "learning_rate": 1.0435295337398943e-05,
      "loss": 0.0123,
      "step": 3390
    },
    {
      "epoch": 8.808290155440414,
      "grad_norm": 0.15277987718582153,
      "learning_rate": 1.0065053673791748e-05,
      "loss": 0.0171,
      "step": 3400
    },
    {
      "epoch": 8.83419689119171,
      "grad_norm": 0.17561012506484985,
      "learning_rate": 9.701150971595218e-06,
      "loss": 0.0147,
      "step": 3410
    },
    {
      "epoch": 8.860103626943005,
      "grad_norm": 0.12972599267959595,
      "learning_rate": 9.34361288007346e-06,
      "loss": 0.0132,
      "step": 3420
    },
    {
      "epoch": 8.8860103626943,
      "grad_norm": 0.14359018206596375,
      "learning_rate": 8.992464599888229e-06,
      "loss": 0.014,
      "step": 3430
    },
    {
      "epoch": 8.911917098445596,
      "grad_norm": 0.18934310972690582,
      "learning_rate": 8.647730881322813e-06,
      "loss": 0.0186,
      "step": 3440
    },
    {
      "epoch": 8.937823834196891,
      "grad_norm": 0.22511492669582367,
      "learning_rate": 8.309436022537465e-06,
      "loss": 0.0167,
      "step": 3450
    },
    {
      "epoch": 8.963730569948186,
      "grad_norm": 0.17313934862613678,
      "learning_rate": 7.97760386785682e-06,
      "loss": 0.0172,
      "step": 3460
    },
    {
      "epoch": 8.989637305699482,
      "grad_norm": 0.15488480031490326,
      "learning_rate": 7.652257806089213e-06,
      "loss": 0.0141,
      "step": 3470
    },
    {
      "epoch": 9.015544041450777,
      "grad_norm": 0.11114071309566498,
      "learning_rate": 7.333420768878174e-06,
      "loss": 0.0146,
      "step": 3480
    },
    {
      "epoch": 9.041450777202073,
      "grad_norm": 0.09289427846670151,
      "learning_rate": 7.0211152290861285e-06,
      "loss": 0.0145,
      "step": 3490
    },
    {
      "epoch": 9.067357512953368,
      "grad_norm": 0.09938207268714905,
      "learning_rate": 6.715363199210423e-06,
      "loss": 0.0156,
      "step": 3500
    },
    {
      "epoch": 9.093264248704664,
      "grad_norm": 0.09943415969610214,
      "learning_rate": 6.416186229831722e-06,
      "loss": 0.0134,
      "step": 3510
    },
    {
      "epoch": 9.119170984455959,
      "grad_norm": 0.15533126890659332,
      "learning_rate": 6.123605408095146e-06,
      "loss": 0.0121,
      "step": 3520
    },
    {
      "epoch": 9.145077720207254,
      "grad_norm": 0.14299090206623077,
      "learning_rate": 5.837641356223911e-06,
      "loss": 0.0169,
      "step": 3530
    },
    {
      "epoch": 9.17098445595855,
      "grad_norm": 0.096370629966259,
      "learning_rate": 5.558314230065842e-06,
      "loss": 0.0142,
      "step": 3540
    },
    {
      "epoch": 9.196891191709845,
      "grad_norm": 0.16918356716632843,
      "learning_rate": 5.285643717672628e-06,
      "loss": 0.0126,
      "step": 3550
    },
    {
      "epoch": 9.22279792746114,
      "grad_norm": 0.23584723472595215,
      "learning_rate": 5.0196490379121885e-06,
      "loss": 0.0145,
      "step": 3560
    },
    {
      "epoch": 9.248704663212436,
      "grad_norm": 0.15384496748447418,
      "learning_rate": 4.760348939114068e-06,
      "loss": 0.0152,
      "step": 3570
    },
    {
      "epoch": 9.27461139896373,
      "grad_norm": 0.24043075740337372,
      "learning_rate": 4.507761697747981e-06,
      "loss": 0.0162,
      "step": 3580
    },
    {
      "epoch": 9.300518134715025,
      "grad_norm": 0.14302724599838257,
      "learning_rate": 4.261905117135512e-06,
      "loss": 0.0151,
      "step": 3590
    },
    {
      "epoch": 9.32642487046632,
      "grad_norm": 0.1379092037677765,
      "learning_rate": 4.022796526195416e-06,
      "loss": 0.0156,
      "step": 3600
    },
    {
      "epoch": 9.352331606217616,
      "grad_norm": 0.12064504623413086,
      "learning_rate": 3.7904527782221754e-06,
      "loss": 0.0117,
      "step": 3610
    },
    {
      "epoch": 9.378238341968911,
      "grad_norm": 0.15608225762844086,
      "learning_rate": 3.564890249698005e-06,
      "loss": 0.0134,
      "step": 3620
    },
    {
      "epoch": 9.404145077720207,
      "grad_norm": 0.2138279527425766,
      "learning_rate": 3.346124839138687e-06,
      "loss": 0.0141,
      "step": 3630
    },
    {
      "epoch": 9.430051813471502,
      "grad_norm": 0.10819461196660995,
      "learning_rate": 3.1341719659729764e-06,
      "loss": 0.015,
      "step": 3640
    },
    {
      "epoch": 9.455958549222798,
      "grad_norm": 0.1564425379037857,
      "learning_rate": 2.9290465694557402e-06,
      "loss": 0.0139,
      "step": 3650
    },
    {
      "epoch": 9.481865284974093,
      "grad_norm": 0.12781445682048798,
      "learning_rate": 2.7307631076149644e-06,
      "loss": 0.0144,
      "step": 3660
    },
    {
      "epoch": 9.507772020725389,
      "grad_norm": 0.09093884378671646,
      "learning_rate": 2.539335556232736e-06,
      "loss": 0.0122,
      "step": 3670
    },
    {
      "epoch": 9.533678756476684,
      "grad_norm": 0.09807614982128143,
      "learning_rate": 2.354777407860209e-06,
      "loss": 0.0124,
      "step": 3680
    },
    {
      "epoch": 9.55958549222798,
      "grad_norm": 0.15765872597694397,
      "learning_rate": 2.1771016708665217e-06,
      "loss": 0.0152,
      "step": 3690
    },
    {
      "epoch": 9.585492227979275,
      "grad_norm": 0.3498193621635437,
      "learning_rate": 2.0063208685219494e-06,
      "loss": 0.0154,
      "step": 3700
    },
    {
      "epoch": 9.61139896373057,
      "grad_norm": 0.16188929975032806,
      "learning_rate": 1.8424470381152137e-06,
      "loss": 0.0135,
      "step": 3710
    },
    {
      "epoch": 9.637305699481866,
      "grad_norm": 0.12133660912513733,
      "learning_rate": 1.68549173010506e-06,
      "loss": 0.0144,
      "step": 3720
    },
    {
      "epoch": 9.663212435233161,
      "grad_norm": 0.15781523287296295,
      "learning_rate": 1.53546600730613e-06,
      "loss": 0.0123,
      "step": 3730
    },
    {
      "epoch": 9.689119170984457,
      "grad_norm": 0.11044237017631531,
      "learning_rate": 1.3923804441092092e-06,
      "loss": 0.0158,
      "step": 3740
    },
    {
      "epoch": 9.715025906735752,
      "grad_norm": 0.2011040598154068,
      "learning_rate": 1.2562451257358887e-06,
      "loss": 0.014,
      "step": 3750
    },
    {
      "epoch": 9.740932642487046,
      "grad_norm": 0.14315728843212128,
      "learning_rate": 1.1270696475277565e-06,
      "loss": 0.0119,
      "step": 3760
    },
    {
      "epoch": 9.766839378238341,
      "grad_norm": 0.13683845102787018,
      "learning_rate": 1.0048631142700383e-06,
      "loss": 0.0147,
      "step": 3770
    },
    {
      "epoch": 9.792746113989637,
      "grad_norm": 0.07541488856077194,
      "learning_rate": 8.896341395499108e-07,
      "loss": 0.0108,
      "step": 3780
    },
    {
      "epoch": 9.818652849740932,
      "grad_norm": 0.15584123134613037,
      "learning_rate": 7.813908451493102e-07,
      "loss": 0.0149,
      "step": 3790
    },
    {
      "epoch": 9.844559585492227,
      "grad_norm": 0.16258586943149567,
      "learning_rate": 6.801408604725446e-07,
      "loss": 0.0154,
      "step": 3800
    },
    {
      "epoch": 9.870466321243523,
      "grad_norm": 0.1323482096195221,
      "learning_rate": 5.858913220085027e-07,
      "loss": 0.0124,
      "step": 3810
    },
    {
      "epoch": 9.896373056994818,
      "grad_norm": 0.11180088669061661,
      "learning_rate": 4.986488728276673e-07,
      "loss": 0.0131,
      "step": 3820
    },
    {
      "epoch": 9.922279792746114,
      "grad_norm": 0.1439903825521469,
      "learning_rate": 4.1841966211388914e-07,
      "loss": 0.0153,
      "step": 3830
    },
    {
      "epoch": 9.94818652849741,
      "grad_norm": 0.10643361508846283,
      "learning_rate": 3.452093447309346e-07,
      "loss": 0.0131,
      "step": 3840
    },
    {
      "epoch": 9.974093264248705,
      "grad_norm": 0.11634659022092819,
      "learning_rate": 2.790230808239591e-07,
      "loss": 0.0129,
      "step": 3850
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.1371113657951355,
      "learning_rate": 2.198655354557766e-07,
      "loss": 0.0133,
      "step": 3860
    },
    {
      "epoch": 10.025906735751295,
      "grad_norm": 0.10803582519292831,
      "learning_rate": 1.6774087827801098e-07,
      "loss": 0.0127,
      "step": 3870
    },
    {
      "epoch": 10.05181347150259,
      "grad_norm": 0.1608572006225586,
      "learning_rate": 1.2265278323728703e-07,
      "loss": 0.0122,
      "step": 3880
    },
    {
      "epoch": 10.077720207253886,
      "grad_norm": 0.14935462176799774,
      "learning_rate": 8.460442831619286e-08,
      "loss": 0.0121,
      "step": 3890
    },
    {
      "epoch": 10.103626943005182,
      "grad_norm": 0.17790275812149048,
      "learning_rate": 5.359849530932603e-08,
      "loss": 0.0154,
      "step": 3900
    },
    {
      "epoch": 10.129533678756477,
      "grad_norm": 0.2045956552028656,
      "learning_rate": 2.9637169634277872e-08,
      "loss": 0.0141,
      "step": 3910
    },
    {
      "epoch": 10.155440414507773,
      "grad_norm": 0.22062177956104279,
      "learning_rate": 1.2722140177567899e-08,
      "loss": 0.0161,
      "step": 3920
    },
    {
      "epoch": 10.181347150259068,
      "grad_norm": 0.18334221839904785,
      "learning_rate": 2.8545991756168122e-09,
      "loss": 0.0128,
      "step": 3930
    },
    {
      "epoch": 10.204663212435234,
      "step": 3939,
      "total_flos": 5.206682005553e+17,
      "train_loss": 0.0395857700919645,
      "train_runtime": 4409.4556,
      "train_samples_per_second": 57.172,
      "train_steps_per_second": 0.893
    }
  ],
  "logging_steps": 10,
  "max_steps": 3939,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 11,
  "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": 5.206682005553e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}