{
  "best_metric": 0.3036472499370575,
  "best_model_checkpoint": "/mmfs1/gscratch/ark/knylund/matplotalt/efficientnet-b0-classifier/checkpoint-5000",
  "epoch": 3.0,
  "eval_steps": 1000,
  "global_step": 5571,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 4.032857418060303,
      "learning_rate": 0.00019964099802548914,
      "loss": 2.1603,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.213678359985352,
      "learning_rate": 0.0001992819960509783,
      "loss": 1.6635,
      "step": 20
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.376636505126953,
      "learning_rate": 0.00019892299407646744,
      "loss": 1.3565,
      "step": 30
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.861647844314575,
      "learning_rate": 0.00019856399210195657,
      "loss": 1.0057,
      "step": 40
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.796308994293213,
      "learning_rate": 0.00019820499012744572,
      "loss": 0.8873,
      "step": 50
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.174099445343018,
      "learning_rate": 0.00019784598815293484,
      "loss": 0.9,
      "step": 60
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7922903299331665,
      "learning_rate": 0.00019748698617842397,
      "loss": 0.6295,
      "step": 70
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.818875551223755,
      "learning_rate": 0.00019712798420391315,
      "loss": 0.6851,
      "step": 80
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.673445463180542,
      "learning_rate": 0.00019676898222940228,
      "loss": 0.6127,
      "step": 90
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.8028321266174316,
      "learning_rate": 0.0001964099802548914,
      "loss": 0.4692,
      "step": 100
    },
    {
      "epoch": 0.06,
      "grad_norm": 6.368370056152344,
      "learning_rate": 0.00019605097828038055,
      "loss": 0.5857,
      "step": 110
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.711109638214111,
      "learning_rate": 0.00019569197630586968,
      "loss": 0.5111,
      "step": 120
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.1240601539611816,
      "learning_rate": 0.00019533297433135883,
      "loss": 0.6485,
      "step": 130
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.9486416578292847,
      "learning_rate": 0.00019497397235684798,
      "loss": 0.7518,
      "step": 140
    },
    {
      "epoch": 0.08,
      "grad_norm": 5.057428359985352,
      "learning_rate": 0.0001946149703823371,
      "loss": 0.4953,
      "step": 150
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.432128667831421,
      "learning_rate": 0.00019425596840782626,
      "loss": 0.7169,
      "step": 160
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.6906955242156982,
      "learning_rate": 0.0001938969664333154,
      "loss": 0.6127,
      "step": 170
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.316644668579102,
      "learning_rate": 0.00019353796445880454,
      "loss": 0.5752,
      "step": 180
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.704860210418701,
      "learning_rate": 0.00019317896248429367,
      "loss": 0.6974,
      "step": 190
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.3908371925354004,
      "learning_rate": 0.00019281996050978282,
      "loss": 0.5098,
      "step": 200
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.832446098327637,
      "learning_rate": 0.00019246095853527195,
      "loss": 0.6852,
      "step": 210
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.9463236331939697,
      "learning_rate": 0.0001921019565607611,
      "loss": 0.5206,
      "step": 220
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.601738691329956,
      "learning_rate": 0.00019174295458625022,
      "loss": 0.589,
      "step": 230
    },
    {
      "epoch": 0.13,
      "grad_norm": 5.127821445465088,
      "learning_rate": 0.00019138395261173938,
      "loss": 0.5953,
      "step": 240
    },
    {
      "epoch": 0.13,
      "grad_norm": 4.185557842254639,
      "learning_rate": 0.00019102495063722853,
      "loss": 0.4675,
      "step": 250
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.7336225509643555,
      "learning_rate": 0.00019066594866271765,
      "loss": 0.6038,
      "step": 260
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.018154621124268,
      "learning_rate": 0.00019030694668820678,
      "loss": 0.5399,
      "step": 270
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.3394994735717773,
      "learning_rate": 0.00018994794471369593,
      "loss": 0.5321,
      "step": 280
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.332233905792236,
      "learning_rate": 0.00018958894273918509,
      "loss": 0.5468,
      "step": 290
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.8721323013305664,
      "learning_rate": 0.0001892299407646742,
      "loss": 0.6093,
      "step": 300
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.815431833267212,
      "learning_rate": 0.00018887093879016336,
      "loss": 0.7448,
      "step": 310
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.5958600044250488,
      "learning_rate": 0.0001885119368156525,
      "loss": 0.5036,
      "step": 320
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.088627815246582,
      "learning_rate": 0.00018815293484114161,
      "loss": 0.6353,
      "step": 330
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.206453800201416,
      "learning_rate": 0.0001877939328666308,
      "loss": 0.449,
      "step": 340
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.1590213775634766,
      "learning_rate": 0.00018743493089211992,
      "loss": 0.626,
      "step": 350
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.7529923915863037,
      "learning_rate": 0.00018707592891760905,
      "loss": 0.5606,
      "step": 360
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.4643802642822266,
      "learning_rate": 0.0001867169269430982,
      "loss": 0.5379,
      "step": 370
    },
    {
      "epoch": 0.2,
      "grad_norm": 8.444103240966797,
      "learning_rate": 0.00018635792496858732,
      "loss": 0.4777,
      "step": 380
    },
    {
      "epoch": 0.21,
      "grad_norm": 5.32062292098999,
      "learning_rate": 0.00018599892299407648,
      "loss": 0.667,
      "step": 390
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.176698207855225,
      "learning_rate": 0.00018563992101956563,
      "loss": 0.4736,
      "step": 400
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.630472660064697,
      "learning_rate": 0.00018528091904505476,
      "loss": 0.4823,
      "step": 410
    },
    {
      "epoch": 0.23,
      "grad_norm": 3.107424020767212,
      "learning_rate": 0.00018492191707054388,
      "loss": 0.6637,
      "step": 420
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.1830148696899414,
      "learning_rate": 0.00018456291509603303,
      "loss": 0.4474,
      "step": 430
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.216266393661499,
      "learning_rate": 0.00018420391312152219,
      "loss": 0.3902,
      "step": 440
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.395972728729248,
      "learning_rate": 0.0001838449111470113,
      "loss": 0.4703,
      "step": 450
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6726592779159546,
      "learning_rate": 0.00018348590917250046,
      "loss": 0.4576,
      "step": 460
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.86993408203125,
      "learning_rate": 0.0001831269071979896,
      "loss": 0.5224,
      "step": 470
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.105071544647217,
      "learning_rate": 0.00018276790522347874,
      "loss": 0.4184,
      "step": 480
    },
    {
      "epoch": 0.26,
      "grad_norm": 3.1365551948547363,
      "learning_rate": 0.0001824089032489679,
      "loss": 0.4105,
      "step": 490
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.22435188293457,
      "learning_rate": 0.00018204990127445702,
      "loss": 0.6113,
      "step": 500
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.1273105144500732,
      "learning_rate": 0.00018169089929994615,
      "loss": 0.4885,
      "step": 510
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.1334537267684937,
      "learning_rate": 0.0001813318973254353,
      "loss": 0.4339,
      "step": 520
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.3947975635528564,
      "learning_rate": 0.00018097289535092442,
      "loss": 0.4574,
      "step": 530
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.6882948875427246,
      "learning_rate": 0.00018061389337641358,
      "loss": 0.4964,
      "step": 540
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.1019885540008545,
      "learning_rate": 0.00018025489140190273,
      "loss": 0.401,
      "step": 550
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.382169008255005,
      "learning_rate": 0.00017989588942739186,
      "loss": 0.3039,
      "step": 560
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.942392110824585,
      "learning_rate": 0.000179536887452881,
      "loss": 0.3974,
      "step": 570
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.2216510772705078,
      "learning_rate": 0.00017917788547837013,
      "loss": 0.3715,
      "step": 580
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.215604782104492,
      "learning_rate": 0.00017881888350385926,
      "loss": 0.4073,
      "step": 590
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.8399810791015625,
      "learning_rate": 0.00017845988152934844,
      "loss": 0.3651,
      "step": 600
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4025568962097168,
      "learning_rate": 0.00017810087955483757,
      "loss": 0.3249,
      "step": 610
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.812524318695068,
      "learning_rate": 0.0001777418775803267,
      "loss": 0.4492,
      "step": 620
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.6712212562561035,
      "learning_rate": 0.00017738287560581584,
      "loss": 0.585,
      "step": 630
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.6772258281707764,
      "learning_rate": 0.00017702387363130497,
      "loss": 0.6295,
      "step": 640
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.9459309577941895,
      "learning_rate": 0.00017666487165679412,
      "loss": 0.4852,
      "step": 650
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.4058187007904053,
      "learning_rate": 0.00017630586968228327,
      "loss": 0.4026,
      "step": 660
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.554550290107727,
      "learning_rate": 0.0001759468677077724,
      "loss": 0.4278,
      "step": 670
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.242212772369385,
      "learning_rate": 0.00017558786573326153,
      "loss": 0.3727,
      "step": 680
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.0075364112854004,
      "learning_rate": 0.00017522886375875068,
      "loss": 0.4446,
      "step": 690
    },
    {
      "epoch": 0.38,
      "grad_norm": 3.4899425506591797,
      "learning_rate": 0.00017486986178423983,
      "loss": 0.3842,
      "step": 700
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.058801651000977,
      "learning_rate": 0.00017451085980972896,
      "loss": 0.4403,
      "step": 710
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.838331401348114,
      "learning_rate": 0.0001741518578352181,
      "loss": 0.3213,
      "step": 720
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.9423627853393555,
      "learning_rate": 0.00017379285586070723,
      "loss": 0.4725,
      "step": 730
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.4071078300476074,
      "learning_rate": 0.00017343385388619636,
      "loss": 0.4542,
      "step": 740
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.141202926635742,
      "learning_rate": 0.00017307485191168554,
      "loss": 0.415,
      "step": 750
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.7862131595611572,
      "learning_rate": 0.00017271584993717467,
      "loss": 0.506,
      "step": 760
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.758727550506592,
      "learning_rate": 0.0001723568479626638,
      "loss": 0.3793,
      "step": 770
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.440414309501648,
      "learning_rate": 0.00017199784598815294,
      "loss": 0.4049,
      "step": 780
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.1473422050476074,
      "learning_rate": 0.00017163884401364207,
      "loss": 0.3329,
      "step": 790
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.930641174316406,
      "learning_rate": 0.00017127984203913122,
      "loss": 0.4517,
      "step": 800
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.398808240890503,
      "learning_rate": 0.00017092084006462038,
      "loss": 0.439,
      "step": 810
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.5552713871002197,
      "learning_rate": 0.0001705618380901095,
      "loss": 0.4203,
      "step": 820
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.3895692825317383,
      "learning_rate": 0.00017020283611559865,
      "loss": 0.3735,
      "step": 830
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.32607364654541,
      "learning_rate": 0.00016984383414108778,
      "loss": 0.3951,
      "step": 840
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.7037802934646606,
      "learning_rate": 0.00016948483216657693,
      "loss": 0.2747,
      "step": 850
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.201627254486084,
      "learning_rate": 0.00016912583019206606,
      "loss": 0.3614,
      "step": 860
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.009246349334717,
      "learning_rate": 0.0001687668282175552,
      "loss": 0.3911,
      "step": 870
    },
    {
      "epoch": 0.47,
      "grad_norm": 3.186776876449585,
      "learning_rate": 0.00016840782624304434,
      "loss": 0.373,
      "step": 880
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6319185495376587,
      "learning_rate": 0.0001680488242685335,
      "loss": 0.3213,
      "step": 890
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.854424953460693,
      "learning_rate": 0.00016768982229402264,
      "loss": 0.4254,
      "step": 900
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.399857997894287,
      "learning_rate": 0.00016733082031951177,
      "loss": 0.3312,
      "step": 910
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.601319789886475,
      "learning_rate": 0.00016697181834500092,
      "loss": 0.4948,
      "step": 920
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.637693405151367,
      "learning_rate": 0.00016661281637049004,
      "loss": 0.5873,
      "step": 930
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8648672103881836,
      "learning_rate": 0.00016625381439597917,
      "loss": 0.2871,
      "step": 940
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.556180238723755,
      "learning_rate": 0.00016589481242146832,
      "loss": 0.505,
      "step": 950
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.072059154510498,
      "learning_rate": 0.00016553581044695748,
      "loss": 0.405,
      "step": 960
    },
    {
      "epoch": 0.52,
      "grad_norm": 3.141235828399658,
      "learning_rate": 0.0001651768084724466,
      "loss": 0.5541,
      "step": 970
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.269875526428223,
      "learning_rate": 0.00016481780649793575,
      "loss": 0.3842,
      "step": 980
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.289591312408447,
      "learning_rate": 0.00016445880452342488,
      "loss": 0.3436,
      "step": 990
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.728127479553223,
      "learning_rate": 0.000164099802548914,
      "loss": 0.4198,
      "step": 1000
    },
    {
      "epoch": 0.54,
      "eval_accuracy": 0.8758246936852027,
      "eval_loss": 0.36355388164520264,
      "eval_runtime": 26.7203,
      "eval_samples_per_second": 158.831,
      "eval_steps_per_second": 19.873,
      "step": 1000
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5636017322540283,
      "learning_rate": 0.00016374080057440319,
      "loss": 0.4572,
      "step": 1010
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.419159173965454,
      "learning_rate": 0.0001633817985998923,
      "loss": 0.287,
      "step": 1020
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.939203977584839,
      "learning_rate": 0.00016302279662538144,
      "loss": 0.355,
      "step": 1030
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.203690528869629,
      "learning_rate": 0.0001626637946508706,
      "loss": 0.406,
      "step": 1040
    },
    {
      "epoch": 0.57,
      "grad_norm": 3.8833212852478027,
      "learning_rate": 0.00016230479267635971,
      "loss": 0.3715,
      "step": 1050
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.7199671268463135,
      "learning_rate": 0.00016194579070184887,
      "loss": 0.3713,
      "step": 1060
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.960907220840454,
      "learning_rate": 0.00016158678872733802,
      "loss": 0.3635,
      "step": 1070
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5699355602264404,
      "learning_rate": 0.00016122778675282715,
      "loss": 0.3228,
      "step": 1080
    },
    {
      "epoch": 0.59,
      "grad_norm": 3.974890947341919,
      "learning_rate": 0.00016086878477831627,
      "loss": 0.5052,
      "step": 1090
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.8382863998413086,
      "learning_rate": 0.00016050978280380542,
      "loss": 0.3811,
      "step": 1100
    },
    {
      "epoch": 0.6,
      "grad_norm": 3.3032066822052,
      "learning_rate": 0.00016015078082929458,
      "loss": 0.4694,
      "step": 1110
    },
    {
      "epoch": 0.6,
      "grad_norm": 3.2820935249328613,
      "learning_rate": 0.0001597917788547837,
      "loss": 0.3933,
      "step": 1120
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.1728293895721436,
      "learning_rate": 0.00015943277688027285,
      "loss": 0.4353,
      "step": 1130
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.178715229034424,
      "learning_rate": 0.00015907377490576198,
      "loss": 0.4013,
      "step": 1140
    },
    {
      "epoch": 0.62,
      "grad_norm": 3.2325494289398193,
      "learning_rate": 0.00015871477293125113,
      "loss": 0.4729,
      "step": 1150
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6329644918441772,
      "learning_rate": 0.00015835577095674029,
      "loss": 0.3281,
      "step": 1160
    },
    {
      "epoch": 0.63,
      "grad_norm": 2.220376968383789,
      "learning_rate": 0.0001579967689822294,
      "loss": 0.3198,
      "step": 1170
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.842592477798462,
      "learning_rate": 0.00015763776700771856,
      "loss": 0.4726,
      "step": 1180
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.8406704664230347,
      "learning_rate": 0.0001572787650332077,
      "loss": 0.4375,
      "step": 1190
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.6387284994125366,
      "learning_rate": 0.00015691976305869682,
      "loss": 0.4196,
      "step": 1200
    },
    {
      "epoch": 0.65,
      "grad_norm": 2.7015557289123535,
      "learning_rate": 0.00015656076108418597,
      "loss": 0.3169,
      "step": 1210
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.334587335586548,
      "learning_rate": 0.00015620175910967512,
      "loss": 0.4468,
      "step": 1220
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.266781806945801,
      "learning_rate": 0.00015584275713516425,
      "loss": 0.4902,
      "step": 1230
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.053922414779663,
      "learning_rate": 0.0001554837551606534,
      "loss": 0.369,
      "step": 1240
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.4680334329605103,
      "learning_rate": 0.00015512475318614252,
      "loss": 0.4692,
      "step": 1250
    },
    {
      "epoch": 0.68,
      "grad_norm": 3.4532482624053955,
      "learning_rate": 0.00015476575121163168,
      "loss": 0.5083,
      "step": 1260
    },
    {
      "epoch": 0.68,
      "grad_norm": 3.088257074356079,
      "learning_rate": 0.00015440674923712083,
      "loss": 0.366,
      "step": 1270
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.26444411277771,
      "learning_rate": 0.00015404774726260996,
      "loss": 0.2412,
      "step": 1280
    },
    {
      "epoch": 0.69,
      "grad_norm": 3.667440176010132,
      "learning_rate": 0.00015368874528809908,
      "loss": 0.3248,
      "step": 1290
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.9644229412078857,
      "learning_rate": 0.00015332974331358823,
      "loss": 0.5926,
      "step": 1300
    },
    {
      "epoch": 0.71,
      "grad_norm": 4.429443359375,
      "learning_rate": 0.00015297074133907736,
      "loss": 0.3579,
      "step": 1310
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.854767918586731,
      "learning_rate": 0.0001526117393645665,
      "loss": 0.4213,
      "step": 1320
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.6535346508026123,
      "learning_rate": 0.00015225273739005566,
      "loss": 0.364,
      "step": 1330
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8327158093452454,
      "learning_rate": 0.0001518937354155448,
      "loss": 0.252,
      "step": 1340
    },
    {
      "epoch": 0.73,
      "grad_norm": 3.9177310466766357,
      "learning_rate": 0.00015153473344103392,
      "loss": 0.3365,
      "step": 1350
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7887599468231201,
      "learning_rate": 0.00015117573146652307,
      "loss": 0.2267,
      "step": 1360
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.150704860687256,
      "learning_rate": 0.00015081672949201222,
      "loss": 0.3312,
      "step": 1370
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.1817049980163574,
      "learning_rate": 0.00015045772751750135,
      "loss": 0.251,
      "step": 1380
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.4935834407806396,
      "learning_rate": 0.0001500987255429905,
      "loss": 0.2915,
      "step": 1390
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.7289029359817505,
      "learning_rate": 0.00014973972356847963,
      "loss": 0.3386,
      "step": 1400
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.7266626358032227,
      "learning_rate": 0.00014938072159396875,
      "loss": 0.3036,
      "step": 1410
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.1115903854370117,
      "learning_rate": 0.00014902171961945793,
      "loss": 0.2455,
      "step": 1420
    },
    {
      "epoch": 0.77,
      "grad_norm": 5.557359218597412,
      "learning_rate": 0.00014866271764494706,
      "loss": 0.315,
      "step": 1430
    },
    {
      "epoch": 0.78,
      "grad_norm": 2.8962290287017822,
      "learning_rate": 0.00014830371567043618,
      "loss": 0.3313,
      "step": 1440
    },
    {
      "epoch": 0.78,
      "grad_norm": 5.260874271392822,
      "learning_rate": 0.00014794471369592533,
      "loss": 0.4113,
      "step": 1450
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.290341854095459,
      "learning_rate": 0.00014758571172141446,
      "loss": 0.4856,
      "step": 1460
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.5758588314056396,
      "learning_rate": 0.0001472267097469036,
      "loss": 0.3505,
      "step": 1470
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.3954672813415527,
      "learning_rate": 0.00014686770777239277,
      "loss": 0.3395,
      "step": 1480
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.661501884460449,
      "learning_rate": 0.0001465087057978819,
      "loss": 0.2511,
      "step": 1490
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6223273277282715,
      "learning_rate": 0.00014614970382337104,
      "loss": 0.369,
      "step": 1500
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.306932210922241,
      "learning_rate": 0.00014579070184886017,
      "loss": 0.3082,
      "step": 1510
    },
    {
      "epoch": 0.82,
      "grad_norm": 3.3667421340942383,
      "learning_rate": 0.00014543169987434932,
      "loss": 0.4531,
      "step": 1520
    },
    {
      "epoch": 0.82,
      "grad_norm": 2.718759536743164,
      "learning_rate": 0.00014507269789983845,
      "loss": 0.2769,
      "step": 1530
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.9687058925628662,
      "learning_rate": 0.0001447136959253276,
      "loss": 0.4445,
      "step": 1540
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.725067615509033,
      "learning_rate": 0.00014435469395081673,
      "loss": 0.2666,
      "step": 1550
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.1680444478988647,
      "learning_rate": 0.00014399569197630588,
      "loss": 0.2961,
      "step": 1560
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.628370761871338,
      "learning_rate": 0.00014363669000179503,
      "loss": 0.3832,
      "step": 1570
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.524216651916504,
      "learning_rate": 0.00014327768802728416,
      "loss": 0.3385,
      "step": 1580
    },
    {
      "epoch": 0.86,
      "grad_norm": 3.9513893127441406,
      "learning_rate": 0.0001429186860527733,
      "loss": 0.4203,
      "step": 1590
    },
    {
      "epoch": 0.86,
      "grad_norm": 3.5642659664154053,
      "learning_rate": 0.00014255968407826244,
      "loss": 0.3039,
      "step": 1600
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.5025720596313477,
      "learning_rate": 0.00014220068210375156,
      "loss": 0.2882,
      "step": 1610
    },
    {
      "epoch": 0.87,
      "grad_norm": 3.6991374492645264,
      "learning_rate": 0.00014184168012924074,
      "loss": 0.4565,
      "step": 1620
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.279611825942993,
      "learning_rate": 0.00014148267815472987,
      "loss": 0.4616,
      "step": 1630
    },
    {
      "epoch": 0.88,
      "grad_norm": 4.0941243171691895,
      "learning_rate": 0.000141123676180219,
      "loss": 0.3271,
      "step": 1640
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.057126998901367,
      "learning_rate": 0.00014076467420570814,
      "loss": 0.3351,
      "step": 1650
    },
    {
      "epoch": 0.89,
      "grad_norm": 4.880453109741211,
      "learning_rate": 0.00014040567223119727,
      "loss": 0.318,
      "step": 1660
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.4076123237609863,
      "learning_rate": 0.0001400466702566864,
      "loss": 0.2955,
      "step": 1670
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.2481727600097656,
      "learning_rate": 0.00013968766828217558,
      "loss": 0.2446,
      "step": 1680
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.711183071136475,
      "learning_rate": 0.0001393286663076647,
      "loss": 0.4932,
      "step": 1690
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.7555177211761475,
      "learning_rate": 0.00013896966433315383,
      "loss": 0.5549,
      "step": 1700
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.4186011552810669,
      "learning_rate": 0.00013861066235864298,
      "loss": 0.2987,
      "step": 1710
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.8629070520401,
      "learning_rate": 0.0001382516603841321,
      "loss": 0.4236,
      "step": 1720
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.2015318870544434,
      "learning_rate": 0.00013789265840962126,
      "loss": 0.4508,
      "step": 1730
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.301048517227173,
      "learning_rate": 0.0001375336564351104,
      "loss": 0.3198,
      "step": 1740
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.4173504114151,
      "learning_rate": 0.00013717465446059954,
      "loss": 0.4391,
      "step": 1750
    },
    {
      "epoch": 0.95,
      "grad_norm": 5.957512378692627,
      "learning_rate": 0.00013681565248608866,
      "loss": 0.3438,
      "step": 1760
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.6446846723556519,
      "learning_rate": 0.00013645665051157781,
      "loss": 0.4001,
      "step": 1770
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.2408043146133423,
      "learning_rate": 0.00013609764853706697,
      "loss": 0.303,
      "step": 1780
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.955267071723938,
      "learning_rate": 0.0001357386465625561,
      "loss": 0.2302,
      "step": 1790
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.4929084777832031,
      "learning_rate": 0.00013537964458804525,
      "loss": 0.3113,
      "step": 1800
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.9054404497146606,
      "learning_rate": 0.00013502064261353437,
      "loss": 0.4578,
      "step": 1810
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.36942720413208,
      "learning_rate": 0.00013466164063902352,
      "loss": 0.1819,
      "step": 1820
    },
    {
      "epoch": 0.99,
      "grad_norm": 3.5076136589050293,
      "learning_rate": 0.00013430263866451268,
      "loss": 0.35,
      "step": 1830
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.2791285514831543,
      "learning_rate": 0.0001339436366900018,
      "loss": 0.2727,
      "step": 1840
    },
    {
      "epoch": 1.0,
      "grad_norm": 3.5197854042053223,
      "learning_rate": 0.00013358463471549095,
      "loss": 0.3642,
      "step": 1850
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.8509562015533447,
      "learning_rate": 0.00013322563274098008,
      "loss": 0.405,
      "step": 1860
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.810483694076538,
      "learning_rate": 0.0001328666307664692,
      "loss": 0.2358,
      "step": 1870
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.4791336059570312,
      "learning_rate": 0.00013250762879195836,
      "loss": 0.2444,
      "step": 1880
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.5583447217941284,
      "learning_rate": 0.0001321486268174475,
      "loss": 0.2832,
      "step": 1890
    },
    {
      "epoch": 1.02,
      "grad_norm": 4.119895935058594,
      "learning_rate": 0.00013178962484293664,
      "loss": 0.3502,
      "step": 1900
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.3282105922698975,
      "learning_rate": 0.0001314306228684258,
      "loss": 0.3043,
      "step": 1910
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.9217724800109863,
      "learning_rate": 0.00013107162089391491,
      "loss": 0.2994,
      "step": 1920
    },
    {
      "epoch": 1.04,
      "grad_norm": 3.4529683589935303,
      "learning_rate": 0.00013071261891940407,
      "loss": 0.3582,
      "step": 1930
    },
    {
      "epoch": 1.04,
      "grad_norm": 3.3998777866363525,
      "learning_rate": 0.00013035361694489322,
      "loss": 0.3677,
      "step": 1940
    },
    {
      "epoch": 1.05,
      "grad_norm": 4.907555103302002,
      "learning_rate": 0.00012999461497038235,
      "loss": 0.2766,
      "step": 1950
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.1131783723831177,
      "learning_rate": 0.00012963561299587147,
      "loss": 0.2385,
      "step": 1960
    },
    {
      "epoch": 1.06,
      "grad_norm": 2.5900931358337402,
      "learning_rate": 0.00012927661102136062,
      "loss": 0.2512,
      "step": 1970
    },
    {
      "epoch": 1.07,
      "grad_norm": 3.0974531173706055,
      "learning_rate": 0.00012891760904684978,
      "loss": 0.3595,
      "step": 1980
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.1107558012008667,
      "learning_rate": 0.0001285586070723389,
      "loss": 0.2754,
      "step": 1990
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.5847731828689575,
      "learning_rate": 0.00012819960509782806,
      "loss": 0.2037,
      "step": 2000
    },
    {
      "epoch": 1.08,
      "eval_accuracy": 0.8774740810556079,
      "eval_loss": 0.34774747490882874,
      "eval_runtime": 25.626,
      "eval_samples_per_second": 165.613,
      "eval_steps_per_second": 20.721,
      "step": 2000
    },
    {
      "epoch": 1.08,
      "grad_norm": 4.522055149078369,
      "learning_rate": 0.00012784060312331718,
      "loss": 0.2866,
      "step": 2010
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.3559322357177734,
      "learning_rate": 0.0001274816011488063,
      "loss": 0.152,
      "step": 2020
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.509596586227417,
      "learning_rate": 0.00012712259917429546,
      "loss": 0.253,
      "step": 2030
    },
    {
      "epoch": 1.1,
      "grad_norm": 3.271685838699341,
      "learning_rate": 0.0001267635971997846,
      "loss": 0.4021,
      "step": 2040
    },
    {
      "epoch": 1.1,
      "grad_norm": 3.338609457015991,
      "learning_rate": 0.00012640459522527374,
      "loss": 0.2471,
      "step": 2050
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.7912172079086304,
      "learning_rate": 0.0001260455932507629,
      "loss": 0.2271,
      "step": 2060
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.5808847546577454,
      "learning_rate": 0.00012568659127625202,
      "loss": 0.3226,
      "step": 2070
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.444607973098755,
      "learning_rate": 0.00012532758930174117,
      "loss": 0.344,
      "step": 2080
    },
    {
      "epoch": 1.13,
      "grad_norm": 4.3131561279296875,
      "learning_rate": 0.00012496858732723032,
      "loss": 0.2386,
      "step": 2090
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.8824583292007446,
      "learning_rate": 0.00012460958535271945,
      "loss": 0.1935,
      "step": 2100
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.8956676125526428,
      "learning_rate": 0.00012425058337820857,
      "loss": 0.3488,
      "step": 2110
    },
    {
      "epoch": 1.14,
      "grad_norm": 3.4252536296844482,
      "learning_rate": 0.00012389158140369772,
      "loss": 0.2186,
      "step": 2120
    },
    {
      "epoch": 1.15,
      "grad_norm": 4.158681392669678,
      "learning_rate": 0.00012353257942918685,
      "loss": 0.21,
      "step": 2130
    },
    {
      "epoch": 1.15,
      "grad_norm": 4.56991720199585,
      "learning_rate": 0.000123173577454676,
      "loss": 0.3828,
      "step": 2140
    },
    {
      "epoch": 1.16,
      "grad_norm": 2.266087532043457,
      "learning_rate": 0.00012281457548016516,
      "loss": 0.373,
      "step": 2150
    },
    {
      "epoch": 1.16,
      "grad_norm": 2.2584266662597656,
      "learning_rate": 0.00012245557350565428,
      "loss": 0.3405,
      "step": 2160
    },
    {
      "epoch": 1.17,
      "grad_norm": 3.8184754848480225,
      "learning_rate": 0.00012209657153114343,
      "loss": 0.3754,
      "step": 2170
    },
    {
      "epoch": 1.17,
      "grad_norm": 3.0375304222106934,
      "learning_rate": 0.00012173756955663256,
      "loss": 0.2974,
      "step": 2180
    },
    {
      "epoch": 1.18,
      "grad_norm": 2.4245946407318115,
      "learning_rate": 0.00012137856758212171,
      "loss": 0.1965,
      "step": 2190
    },
    {
      "epoch": 1.18,
      "grad_norm": 3.8202474117279053,
      "learning_rate": 0.00012101956560761085,
      "loss": 0.3214,
      "step": 2200
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.085423469543457,
      "learning_rate": 0.00012066056363309999,
      "loss": 0.4175,
      "step": 2210
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.4140416383743286,
      "learning_rate": 0.00012030156165858913,
      "loss": 0.33,
      "step": 2220
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.262833118438721,
      "learning_rate": 0.00011994255968407826,
      "loss": 0.2509,
      "step": 2230
    },
    {
      "epoch": 1.21,
      "grad_norm": 2.558087110519409,
      "learning_rate": 0.00011958355770956742,
      "loss": 0.313,
      "step": 2240
    },
    {
      "epoch": 1.21,
      "grad_norm": 4.801666736602783,
      "learning_rate": 0.00011922455573505655,
      "loss": 0.243,
      "step": 2250
    },
    {
      "epoch": 1.22,
      "grad_norm": 4.679200172424316,
      "learning_rate": 0.00011886555376054569,
      "loss": 0.4214,
      "step": 2260
    },
    {
      "epoch": 1.22,
      "grad_norm": 4.6267924308776855,
      "learning_rate": 0.00011850655178603483,
      "loss": 0.3274,
      "step": 2270
    },
    {
      "epoch": 1.23,
      "grad_norm": 7.414104461669922,
      "learning_rate": 0.00011814754981152396,
      "loss": 0.3157,
      "step": 2280
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.789473056793213,
      "learning_rate": 0.00011778854783701312,
      "loss": 0.4056,
      "step": 2290
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.3517308235168457,
      "learning_rate": 0.00011742954586250226,
      "loss": 0.2379,
      "step": 2300
    },
    {
      "epoch": 1.24,
      "grad_norm": 2.3322317600250244,
      "learning_rate": 0.0001170705438879914,
      "loss": 0.3424,
      "step": 2310
    },
    {
      "epoch": 1.25,
      "grad_norm": 2.0581860542297363,
      "learning_rate": 0.00011671154191348052,
      "loss": 0.289,
      "step": 2320
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.9208920001983643,
      "learning_rate": 0.00011635253993896966,
      "loss": 0.3423,
      "step": 2330
    },
    {
      "epoch": 1.26,
      "grad_norm": 4.830653190612793,
      "learning_rate": 0.00011599353796445883,
      "loss": 0.2381,
      "step": 2340
    },
    {
      "epoch": 1.27,
      "grad_norm": 3.273254871368408,
      "learning_rate": 0.00011563453598994795,
      "loss": 0.4048,
      "step": 2350
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.556912660598755,
      "learning_rate": 0.00011527553401543709,
      "loss": 0.1976,
      "step": 2360
    },
    {
      "epoch": 1.28,
      "grad_norm": 3.5109705924987793,
      "learning_rate": 0.00011491653204092623,
      "loss": 0.2245,
      "step": 2370
    },
    {
      "epoch": 1.28,
      "grad_norm": 3.6279704570770264,
      "learning_rate": 0.00011455753006641537,
      "loss": 0.3154,
      "step": 2380
    },
    {
      "epoch": 1.29,
      "grad_norm": 2.629523992538452,
      "learning_rate": 0.00011419852809190452,
      "loss": 0.2974,
      "step": 2390
    },
    {
      "epoch": 1.29,
      "grad_norm": 6.597686290740967,
      "learning_rate": 0.00011383952611739366,
      "loss": 0.3215,
      "step": 2400
    },
    {
      "epoch": 1.3,
      "grad_norm": 3.6411619186401367,
      "learning_rate": 0.00011348052414288279,
      "loss": 0.3519,
      "step": 2410
    },
    {
      "epoch": 1.3,
      "grad_norm": 4.907269477844238,
      "learning_rate": 0.00011312152216837193,
      "loss": 0.2902,
      "step": 2420
    },
    {
      "epoch": 1.31,
      "grad_norm": 4.394081115722656,
      "learning_rate": 0.00011276252019386107,
      "loss": 0.2661,
      "step": 2430
    },
    {
      "epoch": 1.31,
      "grad_norm": 4.089559078216553,
      "learning_rate": 0.0001124035182193502,
      "loss": 0.2106,
      "step": 2440
    },
    {
      "epoch": 1.32,
      "grad_norm": 8.708992004394531,
      "learning_rate": 0.00011204451624483936,
      "loss": 0.2477,
      "step": 2450
    },
    {
      "epoch": 1.32,
      "grad_norm": 2.78205943107605,
      "learning_rate": 0.0001116855142703285,
      "loss": 0.2677,
      "step": 2460
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.323694109916687,
      "learning_rate": 0.00011132651229581764,
      "loss": 0.3387,
      "step": 2470
    },
    {
      "epoch": 1.34,
      "grad_norm": 2.025190591812134,
      "learning_rate": 0.00011096751032130676,
      "loss": 0.407,
      "step": 2480
    },
    {
      "epoch": 1.34,
      "grad_norm": 2.0524258613586426,
      "learning_rate": 0.0001106085083467959,
      "loss": 0.3032,
      "step": 2490
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.9103044271469116,
      "learning_rate": 0.00011024950637228507,
      "loss": 0.1874,
      "step": 2500
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.8446240425109863,
      "learning_rate": 0.00010989050439777419,
      "loss": 0.3678,
      "step": 2510
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.9754679799079895,
      "learning_rate": 0.00010953150242326333,
      "loss": 0.2411,
      "step": 2520
    },
    {
      "epoch": 1.36,
      "grad_norm": 3.324087619781494,
      "learning_rate": 0.00010917250044875247,
      "loss": 0.2345,
      "step": 2530
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.892694115638733,
      "learning_rate": 0.00010881349847424161,
      "loss": 0.1868,
      "step": 2540
    },
    {
      "epoch": 1.37,
      "grad_norm": 5.160618305206299,
      "learning_rate": 0.00010845449649973076,
      "loss": 0.3081,
      "step": 2550
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.6300907135009766,
      "learning_rate": 0.0001080954945252199,
      "loss": 0.4015,
      "step": 2560
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.7582225799560547,
      "learning_rate": 0.00010773649255070904,
      "loss": 0.346,
      "step": 2570
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.9022983312606812,
      "learning_rate": 0.00010737749057619817,
      "loss": 0.2479,
      "step": 2580
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.9787120819091797,
      "learning_rate": 0.0001070184886016873,
      "loss": 0.1605,
      "step": 2590
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.39937639236450195,
      "learning_rate": 0.00010665948662717646,
      "loss": 0.2324,
      "step": 2600
    },
    {
      "epoch": 1.41,
      "grad_norm": 4.440683841705322,
      "learning_rate": 0.0001063004846526656,
      "loss": 0.2951,
      "step": 2610
    },
    {
      "epoch": 1.41,
      "grad_norm": 7.80417013168335,
      "learning_rate": 0.00010594148267815474,
      "loss": 0.3216,
      "step": 2620
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.8643627762794495,
      "learning_rate": 0.00010558248070364388,
      "loss": 0.3748,
      "step": 2630
    },
    {
      "epoch": 1.42,
      "grad_norm": 2.565984010696411,
      "learning_rate": 0.000105223478729133,
      "loss": 0.2482,
      "step": 2640
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.808835506439209,
      "learning_rate": 0.00010486447675462217,
      "loss": 0.1928,
      "step": 2650
    },
    {
      "epoch": 1.43,
      "grad_norm": 4.27629280090332,
      "learning_rate": 0.0001045054747801113,
      "loss": 0.2212,
      "step": 2660
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.2094106674194336,
      "learning_rate": 0.00010414647280560043,
      "loss": 0.3034,
      "step": 2670
    },
    {
      "epoch": 1.44,
      "grad_norm": 3.1522459983825684,
      "learning_rate": 0.00010378747083108957,
      "loss": 0.3882,
      "step": 2680
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.6770269870758057,
      "learning_rate": 0.00010342846885657871,
      "loss": 0.2733,
      "step": 2690
    },
    {
      "epoch": 1.45,
      "grad_norm": 2.9456825256347656,
      "learning_rate": 0.00010306946688206786,
      "loss": 0.2525,
      "step": 2700
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.2579903602600098,
      "learning_rate": 0.000102710464907557,
      "loss": 0.2838,
      "step": 2710
    },
    {
      "epoch": 1.46,
      "grad_norm": 4.77798318862915,
      "learning_rate": 0.00010235146293304614,
      "loss": 0.2237,
      "step": 2720
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.9022595882415771,
      "learning_rate": 0.00010199246095853528,
      "loss": 0.2609,
      "step": 2730
    },
    {
      "epoch": 1.48,
      "grad_norm": 3.534769058227539,
      "learning_rate": 0.0001016334589840244,
      "loss": 0.236,
      "step": 2740
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.7139040231704712,
      "learning_rate": 0.00010127445700951357,
      "loss": 0.2902,
      "step": 2750
    },
    {
      "epoch": 1.49,
      "grad_norm": 3.206073760986328,
      "learning_rate": 0.0001009154550350027,
      "loss": 0.2345,
      "step": 2760
    },
    {
      "epoch": 1.49,
      "grad_norm": 2.6459364891052246,
      "learning_rate": 0.00010055645306049184,
      "loss": 0.2473,
      "step": 2770
    },
    {
      "epoch": 1.5,
      "grad_norm": 3.378709554672241,
      "learning_rate": 0.00010019745108598098,
      "loss": 0.2749,
      "step": 2780
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.6295195817947388,
      "learning_rate": 9.983844911147013e-05,
      "loss": 0.2706,
      "step": 2790
    },
    {
      "epoch": 1.51,
      "grad_norm": 2.157210111618042,
      "learning_rate": 9.947944713695925e-05,
      "loss": 0.2334,
      "step": 2800
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.8217396140098572,
      "learning_rate": 9.91204451624484e-05,
      "loss": 0.3544,
      "step": 2810
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.1436272859573364,
      "learning_rate": 9.876144318793755e-05,
      "loss": 0.2514,
      "step": 2820
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.6002836227416992,
      "learning_rate": 9.840244121342667e-05,
      "loss": 0.2121,
      "step": 2830
    },
    {
      "epoch": 1.53,
      "grad_norm": 5.171236038208008,
      "learning_rate": 9.804343923891582e-05,
      "loss": 0.333,
      "step": 2840
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.5939213037490845,
      "learning_rate": 9.768443726440496e-05,
      "loss": 0.2132,
      "step": 2850
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.8077958226203918,
      "learning_rate": 9.732543528989409e-05,
      "loss": 0.2008,
      "step": 2860
    },
    {
      "epoch": 1.55,
      "grad_norm": 2.7641491889953613,
      "learning_rate": 9.696643331538324e-05,
      "loss": 0.2443,
      "step": 2870
    },
    {
      "epoch": 1.55,
      "grad_norm": 3.886826515197754,
      "learning_rate": 9.660743134087238e-05,
      "loss": 0.2784,
      "step": 2880
    },
    {
      "epoch": 1.56,
      "grad_norm": 3.848173141479492,
      "learning_rate": 9.624842936636152e-05,
      "loss": 0.4413,
      "step": 2890
    },
    {
      "epoch": 1.56,
      "grad_norm": 4.851912975311279,
      "learning_rate": 9.588942739185066e-05,
      "loss": 0.3506,
      "step": 2900
    },
    {
      "epoch": 1.57,
      "grad_norm": 4.759628772735596,
      "learning_rate": 9.55304254173398e-05,
      "loss": 0.2806,
      "step": 2910
    },
    {
      "epoch": 1.57,
      "grad_norm": 3.028917074203491,
      "learning_rate": 9.517142344282894e-05,
      "loss": 0.1563,
      "step": 2920
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.7735349535942078,
      "learning_rate": 9.481242146831808e-05,
      "loss": 0.2664,
      "step": 2930
    },
    {
      "epoch": 1.58,
      "grad_norm": 5.392376899719238,
      "learning_rate": 9.445341949380722e-05,
      "loss": 0.4742,
      "step": 2940
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.3702976703643799,
      "learning_rate": 9.409441751929637e-05,
      "loss": 0.2374,
      "step": 2950
    },
    {
      "epoch": 1.59,
      "grad_norm": 3.8237831592559814,
      "learning_rate": 9.37354155447855e-05,
      "loss": 0.3373,
      "step": 2960
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.340137004852295,
      "learning_rate": 9.337641357027465e-05,
      "loss": 0.1712,
      "step": 2970
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.852940320968628,
      "learning_rate": 9.301741159576379e-05,
      "loss": 0.3893,
      "step": 2980
    },
    {
      "epoch": 1.61,
      "grad_norm": 4.046962738037109,
      "learning_rate": 9.265840962125291e-05,
      "loss": 0.2705,
      "step": 2990
    },
    {
      "epoch": 1.62,
      "grad_norm": 3.1118102073669434,
      "learning_rate": 9.229940764674206e-05,
      "loss": 0.2858,
      "step": 3000
    },
    {
      "epoch": 1.62,
      "eval_accuracy": 0.8982092365692743,
      "eval_loss": 0.31681233644485474,
      "eval_runtime": 28.8932,
      "eval_samples_per_second": 146.886,
      "eval_steps_per_second": 18.378,
      "step": 3000
    },
    {
      "epoch": 1.62,
      "grad_norm": 3.118884563446045,
      "learning_rate": 9.19404056722312e-05,
      "loss": 0.2141,
      "step": 3010
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.300128698348999,
      "learning_rate": 9.158140369772034e-05,
      "loss": 0.2447,
      "step": 3020
    },
    {
      "epoch": 1.63,
      "grad_norm": 5.444098472595215,
      "learning_rate": 9.122240172320948e-05,
      "loss": 0.3292,
      "step": 3030
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.3221828937530518,
      "learning_rate": 9.086339974869862e-05,
      "loss": 0.2242,
      "step": 3040
    },
    {
      "epoch": 1.64,
      "grad_norm": 6.229465007781982,
      "learning_rate": 9.050439777418776e-05,
      "loss": 0.2202,
      "step": 3050
    },
    {
      "epoch": 1.65,
      "grad_norm": 3.1482324600219727,
      "learning_rate": 9.01453957996769e-05,
      "loss": 0.2703,
      "step": 3060
    },
    {
      "epoch": 1.65,
      "grad_norm": 2.2517826557159424,
      "learning_rate": 8.978639382516604e-05,
      "loss": 0.3036,
      "step": 3070
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.955870270729065,
      "learning_rate": 8.942739185065519e-05,
      "loss": 0.2429,
      "step": 3080
    },
    {
      "epoch": 1.66,
      "grad_norm": 5.1806135177612305,
      "learning_rate": 8.906838987614432e-05,
      "loss": 0.2818,
      "step": 3090
    },
    {
      "epoch": 1.67,
      "grad_norm": 5.079287052154541,
      "learning_rate": 8.870938790163347e-05,
      "loss": 0.4519,
      "step": 3100
    },
    {
      "epoch": 1.67,
      "grad_norm": 5.5971221923828125,
      "learning_rate": 8.835038592712261e-05,
      "loss": 0.3289,
      "step": 3110
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.7511892318725586,
      "learning_rate": 8.799138395261173e-05,
      "loss": 0.229,
      "step": 3120
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.807925820350647,
      "learning_rate": 8.763238197810089e-05,
      "loss": 0.1914,
      "step": 3130
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.941755771636963,
      "learning_rate": 8.727338000359003e-05,
      "loss": 0.3117,
      "step": 3140
    },
    {
      "epoch": 1.7,
      "grad_norm": 2.525628089904785,
      "learning_rate": 8.691437802907917e-05,
      "loss": 0.184,
      "step": 3150
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.7301961183547974,
      "learning_rate": 8.65553760545683e-05,
      "loss": 0.3211,
      "step": 3160
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.8699519634246826,
      "learning_rate": 8.619637408005744e-05,
      "loss": 0.4774,
      "step": 3170
    },
    {
      "epoch": 1.71,
      "grad_norm": 4.752264499664307,
      "learning_rate": 8.583737210554658e-05,
      "loss": 0.3038,
      "step": 3180
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.078616976737976,
      "learning_rate": 8.547837013103572e-05,
      "loss": 0.1999,
      "step": 3190
    },
    {
      "epoch": 1.72,
      "grad_norm": 2.705883502960205,
      "learning_rate": 8.511936815652487e-05,
      "loss": 0.1434,
      "step": 3200
    },
    {
      "epoch": 1.73,
      "grad_norm": 3.39790678024292,
      "learning_rate": 8.4760366182014e-05,
      "loss": 0.3326,
      "step": 3210
    },
    {
      "epoch": 1.73,
      "grad_norm": 4.738800525665283,
      "learning_rate": 8.440136420750314e-05,
      "loss": 0.3452,
      "step": 3220
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.4218274354934692,
      "learning_rate": 8.404236223299229e-05,
      "loss": 0.2557,
      "step": 3230
    },
    {
      "epoch": 1.74,
      "grad_norm": 4.442851543426514,
      "learning_rate": 8.368336025848143e-05,
      "loss": 0.2807,
      "step": 3240
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.8416496515274048,
      "learning_rate": 8.332435828397056e-05,
      "loss": 0.2253,
      "step": 3250
    },
    {
      "epoch": 1.76,
      "grad_norm": 2.569054365158081,
      "learning_rate": 8.296535630945971e-05,
      "loss": 0.291,
      "step": 3260
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.4429701864719391,
      "learning_rate": 8.260635433494885e-05,
      "loss": 0.2412,
      "step": 3270
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.4425578117370605,
      "learning_rate": 8.224735236043799e-05,
      "loss": 0.308,
      "step": 3280
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.7549358606338501,
      "learning_rate": 8.188835038592713e-05,
      "loss": 0.2046,
      "step": 3290
    },
    {
      "epoch": 1.78,
      "grad_norm": 5.6480536460876465,
      "learning_rate": 8.152934841141627e-05,
      "loss": 0.2321,
      "step": 3300
    },
    {
      "epoch": 1.78,
      "grad_norm": 5.329677104949951,
      "learning_rate": 8.11703464369054e-05,
      "loss": 0.2555,
      "step": 3310
    },
    {
      "epoch": 1.79,
      "grad_norm": 2.327784776687622,
      "learning_rate": 8.081134446239454e-05,
      "loss": 0.2767,
      "step": 3320
    },
    {
      "epoch": 1.79,
      "grad_norm": 2.5506343841552734,
      "learning_rate": 8.04523424878837e-05,
      "loss": 0.244,
      "step": 3330
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.2340612411499023,
      "learning_rate": 8.009334051337282e-05,
      "loss": 0.2596,
      "step": 3340
    },
    {
      "epoch": 1.8,
      "grad_norm": 4.444881439208984,
      "learning_rate": 7.973433853886196e-05,
      "loss": 0.3693,
      "step": 3350
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.2533739805221558,
      "learning_rate": 7.937533656435111e-05,
      "loss": 0.2085,
      "step": 3360
    },
    {
      "epoch": 1.81,
      "grad_norm": 3.2997021675109863,
      "learning_rate": 7.901633458984024e-05,
      "loss": 0.2568,
      "step": 3370
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.5801811218261719,
      "learning_rate": 7.865733261532939e-05,
      "loss": 0.2121,
      "step": 3380
    },
    {
      "epoch": 1.83,
      "grad_norm": 2.1749870777130127,
      "learning_rate": 7.829833064081853e-05,
      "loss": 0.306,
      "step": 3390
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.134340763092041,
      "learning_rate": 7.793932866630767e-05,
      "loss": 0.1628,
      "step": 3400
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.5507072806358337,
      "learning_rate": 7.758032669179681e-05,
      "loss": 0.1448,
      "step": 3410
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.2097892761230469,
      "learning_rate": 7.722132471728595e-05,
      "loss": 0.2467,
      "step": 3420
    },
    {
      "epoch": 1.85,
      "grad_norm": 3.975356101989746,
      "learning_rate": 7.686232274277509e-05,
      "loss": 0.3024,
      "step": 3430
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.4087692499160767,
      "learning_rate": 7.650332076826423e-05,
      "loss": 0.1599,
      "step": 3440
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.547336220741272,
      "learning_rate": 7.614431879375337e-05,
      "loss": 0.1841,
      "step": 3450
    },
    {
      "epoch": 1.86,
      "grad_norm": 2.1175031661987305,
      "learning_rate": 7.578531681924252e-05,
      "loss": 0.2962,
      "step": 3460
    },
    {
      "epoch": 1.87,
      "grad_norm": 2.945688486099243,
      "learning_rate": 7.542631484473164e-05,
      "loss": 0.3062,
      "step": 3470
    },
    {
      "epoch": 1.87,
      "grad_norm": 4.48527193069458,
      "learning_rate": 7.506731287022078e-05,
      "loss": 0.2825,
      "step": 3480
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.1994941234588623,
      "learning_rate": 7.470831089570994e-05,
      "loss": 0.247,
      "step": 3490
    },
    {
      "epoch": 1.88,
      "grad_norm": 5.999602317810059,
      "learning_rate": 7.434930892119906e-05,
      "loss": 0.2539,
      "step": 3500
    },
    {
      "epoch": 1.89,
      "grad_norm": 2.8415539264678955,
      "learning_rate": 7.399030694668822e-05,
      "loss": 0.2171,
      "step": 3510
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.3422552347183228,
      "learning_rate": 7.363130497217735e-05,
      "loss": 0.3478,
      "step": 3520
    },
    {
      "epoch": 1.9,
      "grad_norm": 2.4383304119110107,
      "learning_rate": 7.327230299766648e-05,
      "loss": 0.1589,
      "step": 3530
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.2923927307128906,
      "learning_rate": 7.291330102315563e-05,
      "loss": 0.2454,
      "step": 3540
    },
    {
      "epoch": 1.91,
      "grad_norm": 7.643916130065918,
      "learning_rate": 7.255429904864477e-05,
      "loss": 0.2569,
      "step": 3550
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.6370854377746582,
      "learning_rate": 7.219529707413391e-05,
      "loss": 0.2397,
      "step": 3560
    },
    {
      "epoch": 1.92,
      "grad_norm": 2.2898166179656982,
      "learning_rate": 7.183629509962305e-05,
      "loss": 0.311,
      "step": 3570
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.21601971983909607,
      "learning_rate": 7.147729312511219e-05,
      "loss": 0.132,
      "step": 3580
    },
    {
      "epoch": 1.93,
      "grad_norm": 5.631169319152832,
      "learning_rate": 7.111829115060133e-05,
      "loss": 0.3239,
      "step": 3590
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.8647130727767944,
      "learning_rate": 7.075928917609047e-05,
      "loss": 0.2938,
      "step": 3600
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.4058655500411987,
      "learning_rate": 7.04002872015796e-05,
      "loss": 0.2243,
      "step": 3610
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.8157402276992798,
      "learning_rate": 7.004128522706876e-05,
      "loss": 0.3979,
      "step": 3620
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.5826122760772705,
      "learning_rate": 6.968228325255788e-05,
      "loss": 0.1669,
      "step": 3630
    },
    {
      "epoch": 1.96,
      "grad_norm": 3.8877973556518555,
      "learning_rate": 6.932328127804704e-05,
      "loss": 0.1809,
      "step": 3640
    },
    {
      "epoch": 1.97,
      "grad_norm": 2.680852174758911,
      "learning_rate": 6.896427930353618e-05,
      "loss": 0.2838,
      "step": 3650
    },
    {
      "epoch": 1.97,
      "grad_norm": 2.5472583770751953,
      "learning_rate": 6.86052773290253e-05,
      "loss": 0.3837,
      "step": 3660
    },
    {
      "epoch": 1.98,
      "grad_norm": 2.85748291015625,
      "learning_rate": 6.824627535451445e-05,
      "loss": 0.227,
      "step": 3670
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.5653326511383057,
      "learning_rate": 6.78872733800036e-05,
      "loss": 0.3072,
      "step": 3680
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.8465702533721924,
      "learning_rate": 6.752827140549273e-05,
      "loss": 0.2379,
      "step": 3690
    },
    {
      "epoch": 1.99,
      "grad_norm": 2.931452989578247,
      "learning_rate": 6.716926943098187e-05,
      "loss": 0.1828,
      "step": 3700
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.0221519470214844,
      "learning_rate": 6.681026745647101e-05,
      "loss": 0.3451,
      "step": 3710
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.943457841873169,
      "learning_rate": 6.645126548196015e-05,
      "loss": 0.2619,
      "step": 3720
    },
    {
      "epoch": 2.01,
      "grad_norm": 2.9193050861358643,
      "learning_rate": 6.609226350744929e-05,
      "loss": 0.2775,
      "step": 3730
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.2334821224212646,
      "learning_rate": 6.573326153293844e-05,
      "loss": 0.24,
      "step": 3740
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.4643306732177734,
      "learning_rate": 6.537425955842758e-05,
      "loss": 0.1615,
      "step": 3750
    },
    {
      "epoch": 2.02,
      "grad_norm": 4.313879013061523,
      "learning_rate": 6.501525758391671e-05,
      "loss": 0.322,
      "step": 3760
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.9670339822769165,
      "learning_rate": 6.465625560940586e-05,
      "loss": 0.159,
      "step": 3770
    },
    {
      "epoch": 2.04,
      "grad_norm": 2.7353298664093018,
      "learning_rate": 6.4297253634895e-05,
      "loss": 0.2219,
      "step": 3780
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.9934465885162354,
      "learning_rate": 6.393825166038412e-05,
      "loss": 0.1663,
      "step": 3790
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.8405640125274658,
      "learning_rate": 6.357924968587328e-05,
      "loss": 0.242,
      "step": 3800
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.6964789628982544,
      "learning_rate": 6.322024771136242e-05,
      "loss": 0.1724,
      "step": 3810
    },
    {
      "epoch": 2.06,
      "grad_norm": 3.5647034645080566,
      "learning_rate": 6.286124573685156e-05,
      "loss": 0.2968,
      "step": 3820
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.890254259109497,
      "learning_rate": 6.25022437623407e-05,
      "loss": 0.2064,
      "step": 3830
    },
    {
      "epoch": 2.07,
      "grad_norm": 3.5268359184265137,
      "learning_rate": 6.214324178782983e-05,
      "loss": 0.2538,
      "step": 3840
    },
    {
      "epoch": 2.07,
      "grad_norm": 6.591871738433838,
      "learning_rate": 6.178423981331897e-05,
      "loss": 0.1802,
      "step": 3850
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.2522422671318054,
      "learning_rate": 6.142523783880811e-05,
      "loss": 0.1511,
      "step": 3860
    },
    {
      "epoch": 2.08,
      "grad_norm": 3.9219987392425537,
      "learning_rate": 6.106623586429726e-05,
      "loss": 0.2375,
      "step": 3870
    },
    {
      "epoch": 2.09,
      "grad_norm": 3.851166009902954,
      "learning_rate": 6.07072338897864e-05,
      "loss": 0.2537,
      "step": 3880
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.0214225053787231,
      "learning_rate": 6.034823191527553e-05,
      "loss": 0.2132,
      "step": 3890
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.24583332240581512,
      "learning_rate": 5.998922994076468e-05,
      "loss": 0.1845,
      "step": 3900
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.3417609930038452,
      "learning_rate": 5.9630227966253815e-05,
      "loss": 0.2113,
      "step": 3910
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.993512511253357,
      "learning_rate": 5.927122599174296e-05,
      "loss": 0.2232,
      "step": 3920
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.5137494802474976,
      "learning_rate": 5.89122240172321e-05,
      "loss": 0.1665,
      "step": 3930
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.766516923904419,
      "learning_rate": 5.855322204272123e-05,
      "loss": 0.1327,
      "step": 3940
    },
    {
      "epoch": 2.13,
      "grad_norm": 2.160499095916748,
      "learning_rate": 5.819422006821038e-05,
      "loss": 0.1939,
      "step": 3950
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.33353400230407715,
      "learning_rate": 5.783521809369952e-05,
      "loss": 0.1866,
      "step": 3960
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.14930427074432373,
      "learning_rate": 5.747621611918865e-05,
      "loss": 0.3156,
      "step": 3970
    },
    {
      "epoch": 2.14,
      "grad_norm": 2.471959352493286,
      "learning_rate": 5.71172141446778e-05,
      "loss": 0.3412,
      "step": 3980
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.0565524101257324,
      "learning_rate": 5.6758212170166935e-05,
      "loss": 0.1895,
      "step": 3990
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.38892602920532227,
      "learning_rate": 5.639921019565608e-05,
      "loss": 0.2103,
      "step": 4000
    },
    {
      "epoch": 2.15,
      "eval_accuracy": 0.9033930254476908,
      "eval_loss": 0.3191111385822296,
      "eval_runtime": 29.6926,
      "eval_samples_per_second": 142.931,
      "eval_steps_per_second": 17.883,
      "step": 4000
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.1914218664169312,
      "learning_rate": 5.604020822114522e-05,
      "loss": 0.1546,
      "step": 4010
    },
    {
      "epoch": 2.16,
      "grad_norm": 7.335626602172852,
      "learning_rate": 5.568120624663435e-05,
      "loss": 0.2709,
      "step": 4020
    },
    {
      "epoch": 2.17,
      "grad_norm": 3.775116443634033,
      "learning_rate": 5.53222042721235e-05,
      "loss": 0.2181,
      "step": 4030
    },
    {
      "epoch": 2.18,
      "grad_norm": 3.1225345134735107,
      "learning_rate": 5.496320229761264e-05,
      "loss": 0.1933,
      "step": 4040
    },
    {
      "epoch": 2.18,
      "grad_norm": 6.653754711151123,
      "learning_rate": 5.460420032310178e-05,
      "loss": 0.2501,
      "step": 4050
    },
    {
      "epoch": 2.19,
      "grad_norm": 5.019111156463623,
      "learning_rate": 5.424519834859092e-05,
      "loss": 0.2231,
      "step": 4060
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.7963464260101318,
      "learning_rate": 5.3886196374080055e-05,
      "loss": 0.2554,
      "step": 4070
    },
    {
      "epoch": 2.2,
      "grad_norm": 2.2074685096740723,
      "learning_rate": 5.35271943995692e-05,
      "loss": 0.1698,
      "step": 4080
    },
    {
      "epoch": 2.2,
      "grad_norm": 2.9906063079833984,
      "learning_rate": 5.316819242505834e-05,
      "loss": 0.1265,
      "step": 4090
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.4613714218139648,
      "learning_rate": 5.2809190450547486e-05,
      "loss": 0.22,
      "step": 4100
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.3449119329452515,
      "learning_rate": 5.2450188476036625e-05,
      "loss": 0.2703,
      "step": 4110
    },
    {
      "epoch": 2.22,
      "grad_norm": 2.9305408000946045,
      "learning_rate": 5.209118650152576e-05,
      "loss": 0.1684,
      "step": 4120
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.8008801937103271,
      "learning_rate": 5.17321845270149e-05,
      "loss": 0.1492,
      "step": 4130
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.2380738258361816,
      "learning_rate": 5.137318255250404e-05,
      "loss": 0.2236,
      "step": 4140
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.0907912254333496,
      "learning_rate": 5.101418057799319e-05,
      "loss": 0.1822,
      "step": 4150
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.3513331413269043,
      "learning_rate": 5.065517860348232e-05,
      "loss": 0.2087,
      "step": 4160
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.3731173276901245,
      "learning_rate": 5.029617662897146e-05,
      "loss": 0.1842,
      "step": 4170
    },
    {
      "epoch": 2.25,
      "grad_norm": 4.231403827667236,
      "learning_rate": 4.99371746544606e-05,
      "loss": 0.2008,
      "step": 4180
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.6611512899398804,
      "learning_rate": 4.9578172679949745e-05,
      "loss": 0.1885,
      "step": 4190
    },
    {
      "epoch": 2.26,
      "grad_norm": 2.624540090560913,
      "learning_rate": 4.9219170705438884e-05,
      "loss": 0.2974,
      "step": 4200
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.173483967781067,
      "learning_rate": 4.886016873092802e-05,
      "loss": 0.1817,
      "step": 4210
    },
    {
      "epoch": 2.27,
      "grad_norm": 2.013545513153076,
      "learning_rate": 4.850116675641716e-05,
      "loss": 0.165,
      "step": 4220
    },
    {
      "epoch": 2.28,
      "grad_norm": 3.9248149394989014,
      "learning_rate": 4.81421647819063e-05,
      "loss": 0.1693,
      "step": 4230
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.9728902578353882,
      "learning_rate": 4.778316280739544e-05,
      "loss": 0.1985,
      "step": 4240
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.5156265497207642,
      "learning_rate": 4.7424160832884586e-05,
      "loss": 0.2339,
      "step": 4250
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.8958511352539062,
      "learning_rate": 4.7065158858373725e-05,
      "loss": 0.1843,
      "step": 4260
    },
    {
      "epoch": 2.3,
      "grad_norm": 2.751262664794922,
      "learning_rate": 4.6706156883862865e-05,
      "loss": 0.1703,
      "step": 4270
    },
    {
      "epoch": 2.3,
      "grad_norm": 2.066708564758301,
      "learning_rate": 4.6347154909352004e-05,
      "loss": 0.1751,
      "step": 4280
    },
    {
      "epoch": 2.31,
      "grad_norm": 5.061429500579834,
      "learning_rate": 4.598815293484114e-05,
      "loss": 0.2089,
      "step": 4290
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.9940416812896729,
      "learning_rate": 4.562915096033029e-05,
      "loss": 0.1938,
      "step": 4300
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.8539941906929016,
      "learning_rate": 4.527014898581942e-05,
      "loss": 0.1815,
      "step": 4310
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.626430869102478,
      "learning_rate": 4.491114701130856e-05,
      "loss": 0.211,
      "step": 4320
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.9689849019050598,
      "learning_rate": 4.4552145036797706e-05,
      "loss": 0.2467,
      "step": 4330
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.6195528507232666,
      "learning_rate": 4.4193143062286845e-05,
      "loss": 0.1992,
      "step": 4340
    },
    {
      "epoch": 2.34,
      "grad_norm": 2.0465312004089355,
      "learning_rate": 4.3834141087775984e-05,
      "loss": 0.1256,
      "step": 4350
    },
    {
      "epoch": 2.35,
      "grad_norm": 4.3072404861450195,
      "learning_rate": 4.3475139113265124e-05,
      "loss": 0.2078,
      "step": 4360
    },
    {
      "epoch": 2.35,
      "grad_norm": 2.987069606781006,
      "learning_rate": 4.311613713875426e-05,
      "loss": 0.1688,
      "step": 4370
    },
    {
      "epoch": 2.36,
      "grad_norm": 5.688981533050537,
      "learning_rate": 4.275713516424341e-05,
      "loss": 0.1867,
      "step": 4380
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.6407305002212524,
      "learning_rate": 4.239813318973255e-05,
      "loss": 0.1093,
      "step": 4390
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.10652494430542,
      "learning_rate": 4.203913121522168e-05,
      "loss": 0.1621,
      "step": 4400
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.8150840401649475,
      "learning_rate": 4.1680129240710826e-05,
      "loss": 0.19,
      "step": 4410
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.18293878436088562,
      "learning_rate": 4.1321127266199965e-05,
      "loss": 0.2173,
      "step": 4420
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.6565092206001282,
      "learning_rate": 4.0962125291689104e-05,
      "loss": 0.2949,
      "step": 4430
    },
    {
      "epoch": 2.39,
      "grad_norm": 4.2705793380737305,
      "learning_rate": 4.060312331717825e-05,
      "loss": 0.1654,
      "step": 4440
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.5598949790000916,
      "learning_rate": 4.024412134266738e-05,
      "loss": 0.1619,
      "step": 4450
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.7344173192977905,
      "learning_rate": 3.988511936815653e-05,
      "loss": 0.4215,
      "step": 4460
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.7158490419387817,
      "learning_rate": 3.952611739364567e-05,
      "loss": 0.2035,
      "step": 4470
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.7473658323287964,
      "learning_rate": 3.916711541913481e-05,
      "loss": 0.126,
      "step": 4480
    },
    {
      "epoch": 2.42,
      "grad_norm": 5.31575345993042,
      "learning_rate": 3.8808113444623946e-05,
      "loss": 0.247,
      "step": 4490
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.087374448776245,
      "learning_rate": 3.8449111470113085e-05,
      "loss": 0.2586,
      "step": 4500
    },
    {
      "epoch": 2.43,
      "grad_norm": 6.151815891265869,
      "learning_rate": 3.8090109495602224e-05,
      "loss": 0.2277,
      "step": 4510
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.4579615294933319,
      "learning_rate": 3.773110752109137e-05,
      "loss": 0.1649,
      "step": 4520
    },
    {
      "epoch": 2.44,
      "grad_norm": 2.4893856048583984,
      "learning_rate": 3.737210554658051e-05,
      "loss": 0.2459,
      "step": 4530
    },
    {
      "epoch": 2.44,
      "grad_norm": 3.96018385887146,
      "learning_rate": 3.701310357206965e-05,
      "loss": 0.2287,
      "step": 4540
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.051582098007202,
      "learning_rate": 3.665410159755879e-05,
      "loss": 0.2518,
      "step": 4550
    },
    {
      "epoch": 2.46,
      "grad_norm": 3.5769152641296387,
      "learning_rate": 3.629509962304793e-05,
      "loss": 0.2217,
      "step": 4560
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.9112540483474731,
      "learning_rate": 3.593609764853707e-05,
      "loss": 0.2797,
      "step": 4570
    },
    {
      "epoch": 2.47,
      "grad_norm": 4.061404228210449,
      "learning_rate": 3.5577095674026205e-05,
      "loss": 0.2029,
      "step": 4580
    },
    {
      "epoch": 2.47,
      "grad_norm": 5.101590633392334,
      "learning_rate": 3.521809369951535e-05,
      "loss": 0.1701,
      "step": 4590
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.21550802886486053,
      "learning_rate": 3.485909172500449e-05,
      "loss": 0.246,
      "step": 4600
    },
    {
      "epoch": 2.48,
      "grad_norm": 2.4973855018615723,
      "learning_rate": 3.450008975049363e-05,
      "loss": 0.2274,
      "step": 4610
    },
    {
      "epoch": 2.49,
      "grad_norm": 4.115890979766846,
      "learning_rate": 3.4141087775982775e-05,
      "loss": 0.1776,
      "step": 4620
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.45104721188545227,
      "learning_rate": 3.378208580147191e-05,
      "loss": 0.1482,
      "step": 4630
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.4731504917144775,
      "learning_rate": 3.342308382696105e-05,
      "loss": 0.1552,
      "step": 4640
    },
    {
      "epoch": 2.5,
      "grad_norm": 3.0015158653259277,
      "learning_rate": 3.306408185245019e-05,
      "loss": 0.2306,
      "step": 4650
    },
    {
      "epoch": 2.51,
      "grad_norm": 3.9862313270568848,
      "learning_rate": 3.270507987793933e-05,
      "loss": 0.2291,
      "step": 4660
    },
    {
      "epoch": 2.51,
      "grad_norm": 3.0735368728637695,
      "learning_rate": 3.234607790342847e-05,
      "loss": 0.1431,
      "step": 4670
    },
    {
      "epoch": 2.52,
      "grad_norm": 2.155254364013672,
      "learning_rate": 3.198707592891761e-05,
      "loss": 0.1864,
      "step": 4680
    },
    {
      "epoch": 2.53,
      "grad_norm": 3.6670236587524414,
      "learning_rate": 3.162807395440675e-05,
      "loss": 0.3039,
      "step": 4690
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.040534496307373,
      "learning_rate": 3.1269071979895895e-05,
      "loss": 0.1168,
      "step": 4700
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.5933005809783936,
      "learning_rate": 3.0910070005385034e-05,
      "loss": 0.1937,
      "step": 4710
    },
    {
      "epoch": 2.54,
      "grad_norm": 2.87385630607605,
      "learning_rate": 3.055106803087417e-05,
      "loss": 0.1457,
      "step": 4720
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.0385680198669434,
      "learning_rate": 3.019206605636331e-05,
      "loss": 0.1738,
      "step": 4730
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.6659444570541382,
      "learning_rate": 2.9833064081852452e-05,
      "loss": 0.1625,
      "step": 4740
    },
    {
      "epoch": 2.56,
      "grad_norm": 6.179418563842773,
      "learning_rate": 2.9474062107341594e-05,
      "loss": 0.183,
      "step": 4750
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.6888794898986816,
      "learning_rate": 2.911506013283073e-05,
      "loss": 0.1682,
      "step": 4760
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.9652776718139648,
      "learning_rate": 2.875605815831987e-05,
      "loss": 0.1581,
      "step": 4770
    },
    {
      "epoch": 2.57,
      "grad_norm": 2.623950719833374,
      "learning_rate": 2.8397056183809012e-05,
      "loss": 0.1231,
      "step": 4780
    },
    {
      "epoch": 2.58,
      "grad_norm": 2.0535216331481934,
      "learning_rate": 2.8038054209298154e-05,
      "loss": 0.2414,
      "step": 4790
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.4909818470478058,
      "learning_rate": 2.7679052234787293e-05,
      "loss": 0.1228,
      "step": 4800
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.1262626200914383,
      "learning_rate": 2.732005026027643e-05,
      "loss": 0.1882,
      "step": 4810
    },
    {
      "epoch": 2.6,
      "grad_norm": 5.220785140991211,
      "learning_rate": 2.696104828576557e-05,
      "loss": 0.2131,
      "step": 4820
    },
    {
      "epoch": 2.6,
      "grad_norm": 2.826752185821533,
      "learning_rate": 2.6602046311254714e-05,
      "loss": 0.2074,
      "step": 4830
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.2399777173995972,
      "learning_rate": 2.6243044336743853e-05,
      "loss": 0.1189,
      "step": 4840
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.9711045026779175,
      "learning_rate": 2.5884042362232992e-05,
      "loss": 0.1084,
      "step": 4850
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.2137917280197144,
      "learning_rate": 2.552504038772213e-05,
      "loss": 0.2382,
      "step": 4860
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.8354400396347046,
      "learning_rate": 2.5166038413211274e-05,
      "loss": 0.1869,
      "step": 4870
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.3178220987319946,
      "learning_rate": 2.4807036438700417e-05,
      "loss": 0.1731,
      "step": 4880
    },
    {
      "epoch": 2.63,
      "grad_norm": 5.171186923980713,
      "learning_rate": 2.4448034464189552e-05,
      "loss": 0.1721,
      "step": 4890
    },
    {
      "epoch": 2.64,
      "grad_norm": 2.362175703048706,
      "learning_rate": 2.4089032489678695e-05,
      "loss": 0.1693,
      "step": 4900
    },
    {
      "epoch": 2.64,
      "grad_norm": 5.108491897583008,
      "learning_rate": 2.3730030515167834e-05,
      "loss": 0.2686,
      "step": 4910
    },
    {
      "epoch": 2.65,
      "grad_norm": 3.3931541442871094,
      "learning_rate": 2.3371028540656977e-05,
      "loss": 0.1216,
      "step": 4920
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.5424547791481018,
      "learning_rate": 2.3012026566146112e-05,
      "loss": 0.1603,
      "step": 4930
    },
    {
      "epoch": 2.66,
      "grad_norm": 3.6639039516448975,
      "learning_rate": 2.2653024591635255e-05,
      "loss": 0.2891,
      "step": 4940
    },
    {
      "epoch": 2.67,
      "grad_norm": 4.780062675476074,
      "learning_rate": 2.2294022617124397e-05,
      "loss": 0.1574,
      "step": 4950
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.9114960432052612,
      "learning_rate": 2.1935020642613537e-05,
      "loss": 0.1174,
      "step": 4960
    },
    {
      "epoch": 2.68,
      "grad_norm": 2.23799467086792,
      "learning_rate": 2.1576018668102676e-05,
      "loss": 0.2473,
      "step": 4970
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.7084078788757324,
      "learning_rate": 2.1217016693591815e-05,
      "loss": 0.2229,
      "step": 4980
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.8487536907196045,
      "learning_rate": 2.0858014719080957e-05,
      "loss": 0.1544,
      "step": 4990
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.9069527387619019,
      "learning_rate": 2.0499012744570097e-05,
      "loss": 0.2037,
      "step": 5000
    },
    {
      "epoch": 2.69,
      "eval_accuracy": 0.9019792648444863,
      "eval_loss": 0.3036472499370575,
      "eval_runtime": 29.9482,
      "eval_samples_per_second": 141.711,
      "eval_steps_per_second": 17.731,
      "step": 5000
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.8534448742866516,
      "learning_rate": 2.0140010770059236e-05,
      "loss": 0.1891,
      "step": 5010
    },
    {
      "epoch": 2.7,
      "grad_norm": 2.7062954902648926,
      "learning_rate": 1.9781008795548375e-05,
      "loss": 0.1891,
      "step": 5020
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.05844093859195709,
      "learning_rate": 1.9422006821037517e-05,
      "loss": 0.22,
      "step": 5030
    },
    {
      "epoch": 2.71,
      "grad_norm": 2.087597608566284,
      "learning_rate": 1.9063004846526657e-05,
      "loss": 0.2031,
      "step": 5040
    },
    {
      "epoch": 2.72,
      "grad_norm": 2.3757686614990234,
      "learning_rate": 1.8704002872015796e-05,
      "loss": 0.1134,
      "step": 5050
    },
    {
      "epoch": 2.72,
      "grad_norm": 3.9158239364624023,
      "learning_rate": 1.8345000897504938e-05,
      "loss": 0.1942,
      "step": 5060
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.6796614527702332,
      "learning_rate": 1.7985998922994077e-05,
      "loss": 0.1914,
      "step": 5070
    },
    {
      "epoch": 2.74,
      "grad_norm": 2.11728835105896,
      "learning_rate": 1.7626996948483217e-05,
      "loss": 0.1804,
      "step": 5080
    },
    {
      "epoch": 2.74,
      "grad_norm": 4.9103522300720215,
      "learning_rate": 1.7267994973972356e-05,
      "loss": 0.1873,
      "step": 5090
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.884007453918457,
      "learning_rate": 1.6908992999461498e-05,
      "loss": 0.1804,
      "step": 5100
    },
    {
      "epoch": 2.75,
      "grad_norm": 3.7006118297576904,
      "learning_rate": 1.6549991024950637e-05,
      "loss": 0.1204,
      "step": 5110
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.42698580026626587,
      "learning_rate": 1.619098905043978e-05,
      "loss": 0.151,
      "step": 5120
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.0128672122955322,
      "learning_rate": 1.583198707592892e-05,
      "loss": 0.2026,
      "step": 5130
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.4201221466064453,
      "learning_rate": 1.5472985101418058e-05,
      "loss": 0.1172,
      "step": 5140
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.530119776725769,
      "learning_rate": 1.5113983126907199e-05,
      "loss": 0.1508,
      "step": 5150
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.0262278318405151,
      "learning_rate": 1.4754981152396338e-05,
      "loss": 0.3655,
      "step": 5160
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.333284854888916,
      "learning_rate": 1.4395979177885479e-05,
      "loss": 0.1887,
      "step": 5170
    },
    {
      "epoch": 2.79,
      "grad_norm": 3.272552967071533,
      "learning_rate": 1.4036977203374618e-05,
      "loss": 0.1912,
      "step": 5180
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.6113843321800232,
      "learning_rate": 1.3677975228863759e-05,
      "loss": 0.1772,
      "step": 5190
    },
    {
      "epoch": 2.8,
      "grad_norm": 4.440057754516602,
      "learning_rate": 1.3318973254352898e-05,
      "loss": 0.2022,
      "step": 5200
    },
    {
      "epoch": 2.81,
      "grad_norm": 5.032436847686768,
      "learning_rate": 1.2959971279842039e-05,
      "loss": 0.1968,
      "step": 5210
    },
    {
      "epoch": 2.81,
      "grad_norm": 2.491443395614624,
      "learning_rate": 1.2600969305331181e-05,
      "loss": 0.2772,
      "step": 5220
    },
    {
      "epoch": 2.82,
      "grad_norm": 2.4109225273132324,
      "learning_rate": 1.2241967330820319e-05,
      "loss": 0.2009,
      "step": 5230
    },
    {
      "epoch": 2.82,
      "grad_norm": 3.7935705184936523,
      "learning_rate": 1.188296535630946e-05,
      "loss": 0.2096,
      "step": 5240
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.8197938799858093,
      "learning_rate": 1.15239633817986e-05,
      "loss": 0.148,
      "step": 5250
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.6385130882263184,
      "learning_rate": 1.1164961407287741e-05,
      "loss": 0.121,
      "step": 5260
    },
    {
      "epoch": 2.84,
      "grad_norm": 4.050340175628662,
      "learning_rate": 1.080595943277688e-05,
      "loss": 0.1742,
      "step": 5270
    },
    {
      "epoch": 2.84,
      "grad_norm": 7.119118690490723,
      "learning_rate": 1.0446957458266021e-05,
      "loss": 0.1934,
      "step": 5280
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.18755416572093964,
      "learning_rate": 1.008795548375516e-05,
      "loss": 0.1772,
      "step": 5290
    },
    {
      "epoch": 2.85,
      "grad_norm": 7.360252857208252,
      "learning_rate": 9.728953509244301e-06,
      "loss": 0.2576,
      "step": 5300
    },
    {
      "epoch": 2.86,
      "grad_norm": 2.4663641452789307,
      "learning_rate": 9.36995153473344e-06,
      "loss": 0.1611,
      "step": 5310
    },
    {
      "epoch": 2.86,
      "grad_norm": 2.426260232925415,
      "learning_rate": 9.010949560222581e-06,
      "loss": 0.168,
      "step": 5320
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.6847411394119263,
      "learning_rate": 8.65194758571172e-06,
      "loss": 0.1494,
      "step": 5330
    },
    {
      "epoch": 2.88,
      "grad_norm": 2.8023247718811035,
      "learning_rate": 8.292945611200861e-06,
      "loss": 0.2282,
      "step": 5340
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.8137952089309692,
      "learning_rate": 7.933943636690002e-06,
      "loss": 0.2085,
      "step": 5350
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.6088232398033142,
      "learning_rate": 7.574941662179143e-06,
      "loss": 0.1314,
      "step": 5360
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.53076171875,
      "learning_rate": 7.215939687668283e-06,
      "loss": 0.2243,
      "step": 5370
    },
    {
      "epoch": 2.9,
      "grad_norm": 2.4444739818573,
      "learning_rate": 6.856937713157423e-06,
      "loss": 0.1428,
      "step": 5380
    },
    {
      "epoch": 2.9,
      "grad_norm": 3.4215807914733887,
      "learning_rate": 6.497935738646563e-06,
      "loss": 0.2424,
      "step": 5390
    },
    {
      "epoch": 2.91,
      "grad_norm": 5.3438873291015625,
      "learning_rate": 6.138933764135703e-06,
      "loss": 0.2117,
      "step": 5400
    },
    {
      "epoch": 2.91,
      "grad_norm": 3.0934898853302,
      "learning_rate": 5.779931789624844e-06,
      "loss": 0.2268,
      "step": 5410
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.5680592060089111,
      "learning_rate": 5.420929815113984e-06,
      "loss": 0.1086,
      "step": 5420
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.9947880506515503,
      "learning_rate": 5.061927840603124e-06,
      "loss": 0.1477,
      "step": 5430
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.950864315032959,
      "learning_rate": 4.702925866092264e-06,
      "loss": 0.1439,
      "step": 5440
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.8457893133163452,
      "learning_rate": 4.343923891581404e-06,
      "loss": 0.2122,
      "step": 5450
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.26061761379241943,
      "learning_rate": 3.984921917070545e-06,
      "loss": 0.1721,
      "step": 5460
    },
    {
      "epoch": 2.95,
      "grad_norm": 2.678389072418213,
      "learning_rate": 3.6259199425596842e-06,
      "loss": 0.1874,
      "step": 5470
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.591628074645996,
      "learning_rate": 3.2669179680488242e-06,
      "loss": 0.1947,
      "step": 5480
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.602062702178955,
      "learning_rate": 2.9079159935379646e-06,
      "loss": 0.2288,
      "step": 5490
    },
    {
      "epoch": 2.96,
      "grad_norm": 4.162856578826904,
      "learning_rate": 2.5489140190271046e-06,
      "loss": 0.1616,
      "step": 5500
    },
    {
      "epoch": 2.97,
      "grad_norm": 3.2213492393493652,
      "learning_rate": 2.189912044516245e-06,
      "loss": 0.1245,
      "step": 5510
    },
    {
      "epoch": 2.97,
      "grad_norm": 3.308234691619873,
      "learning_rate": 1.830910070005385e-06,
      "loss": 0.234,
      "step": 5520
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.7126364707946777,
      "learning_rate": 1.4719080954945252e-06,
      "loss": 0.2397,
      "step": 5530
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.09120042622089386,
      "learning_rate": 1.1129061209836654e-06,
      "loss": 0.1205,
      "step": 5540
    },
    {
      "epoch": 2.99,
      "grad_norm": 2.043335199356079,
      "learning_rate": 7.539041464728056e-07,
      "loss": 0.1522,
      "step": 5550
    },
    {
      "epoch": 2.99,
      "grad_norm": 4.629100799560547,
      "learning_rate": 3.9490217196194584e-07,
      "loss": 0.1202,
      "step": 5560
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.9897916316986084,
      "learning_rate": 3.5900197451085984e-08,
      "loss": 0.1182,
      "step": 5570
    },
    {
      "epoch": 3.0,
      "step": 5571,
      "total_flos": 3.2380980042571776e+17,
      "train_loss": 0.3105421944733176,
      "train_runtime": 865.4804,
      "train_samples_per_second": 102.973,
      "train_steps_per_second": 6.437
    }
  ],
  "logging_steps": 10,
  "max_steps": 5571,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "total_flos": 3.2380980042571776e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}