{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002,
      "grad_norm": 0.3647916913032532,
      "learning_rate": 2e-08,
      "loss": 0.5569,
      "step": 10
    },
    {
      "epoch": 0.004,
      "grad_norm": 0.7772038578987122,
      "learning_rate": 4e-08,
      "loss": 0.5903,
      "step": 20
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.4401947259902954,
      "learning_rate": 6e-08,
      "loss": 0.7141,
      "step": 30
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.5190258622169495,
      "learning_rate": 8e-08,
      "loss": 0.5588,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.29183804988861084,
      "learning_rate": 1e-07,
      "loss": 0.6131,
      "step": 50
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.0,
      "learning_rate": 1.2e-07,
      "loss": 0.5862,
      "step": 60
    },
    {
      "epoch": 0.014,
      "grad_norm": 0.7411805987358093,
      "learning_rate": 1.4e-07,
      "loss": 0.6266,
      "step": 70
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.7374635338783264,
      "learning_rate": 1.6e-07,
      "loss": 0.7009,
      "step": 80
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.4058414697647095,
      "learning_rate": 1.8e-07,
      "loss": 0.452,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.22336336970329285,
      "learning_rate": 2e-07,
      "loss": 0.5137,
      "step": 100
    },
    {
      "epoch": 0.022,
      "grad_norm": 1.8679215908050537,
      "learning_rate": 2.1999999999999998e-07,
      "loss": 0.8075,
      "step": 110
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.37687167525291443,
      "learning_rate": 2.4e-07,
      "loss": 0.5353,
      "step": 120
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.5401660799980164,
      "learning_rate": 2.6e-07,
      "loss": 0.514,
      "step": 130
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.24486061930656433,
      "learning_rate": 2.8e-07,
      "loss": 0.4853,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.628027617931366,
      "learning_rate": 3e-07,
      "loss": 0.6838,
      "step": 150
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.5709905624389648,
      "learning_rate": 3.2e-07,
      "loss": 0.4199,
      "step": 160
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.3342764377593994,
      "learning_rate": 3.4000000000000003e-07,
      "loss": 0.6163,
      "step": 170
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.3317417800426483,
      "learning_rate": 3.6e-07,
      "loss": 0.6549,
      "step": 180
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.0,
      "learning_rate": 3.7999999999999996e-07,
      "loss": 0.6489,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.4099644422531128,
      "learning_rate": 4e-07,
      "loss": 0.7401,
      "step": 200
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.5073985457420349,
      "learning_rate": 4.1999999999999995e-07,
      "loss": 0.5801,
      "step": 210
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.0,
      "learning_rate": 4.3999999999999997e-07,
      "loss": 0.6649,
      "step": 220
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.4103664457798004,
      "learning_rate": 4.6e-07,
      "loss": 0.5962,
      "step": 230
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.0,
      "learning_rate": 4.8e-07,
      "loss": 0.2218,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8555938005447388,
      "learning_rate": 5e-07,
      "loss": 0.6688,
      "step": 250
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.18369270861148834,
      "learning_rate": 5.2e-07,
      "loss": 0.5419,
      "step": 260
    },
    {
      "epoch": 0.054,
      "grad_norm": 1.0677138566970825,
      "learning_rate": 5.4e-07,
      "loss": 0.7335,
      "step": 270
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.6109152436256409,
      "learning_rate": 5.6e-07,
      "loss": 0.548,
      "step": 280
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.4339801073074341,
      "learning_rate": 5.8e-07,
      "loss": 0.5819,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5122050642967224,
      "learning_rate": 6e-07,
      "loss": 0.5329,
      "step": 300
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.5236920714378357,
      "learning_rate": 6.2e-07,
      "loss": 0.6093,
      "step": 310
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.7817385196685791,
      "learning_rate": 6.4e-07,
      "loss": 0.46,
      "step": 320
    },
    {
      "epoch": 0.066,
      "grad_norm": 0.7094607353210449,
      "learning_rate": 6.6e-07,
      "loss": 0.6628,
      "step": 330
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.6325480341911316,
      "learning_rate": 6.800000000000001e-07,
      "loss": 0.7074,
      "step": 340
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7745736241340637,
      "learning_rate": 7e-07,
      "loss": 0.4944,
      "step": 350
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.6897551417350769,
      "learning_rate": 7.2e-07,
      "loss": 0.5193,
      "step": 360
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.0,
      "learning_rate": 7.4e-07,
      "loss": 0.5934,
      "step": 370
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.5348847508430481,
      "learning_rate": 7.599999999999999e-07,
      "loss": 0.6671,
      "step": 380
    },
    {
      "epoch": 0.078,
      "grad_norm": 1.388077974319458,
      "learning_rate": 7.799999999999999e-07,
      "loss": 0.6666,
      "step": 390
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3098819851875305,
      "learning_rate": 8e-07,
      "loss": 0.5733,
      "step": 400
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.5731427073478699,
      "learning_rate": 8.199999999999999e-07,
      "loss": 0.8238,
      "step": 410
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.34736743569374084,
      "learning_rate": 8.399999999999999e-07,
      "loss": 0.7012,
      "step": 420
    },
    {
      "epoch": 0.086,
      "grad_norm": 0.7676399350166321,
      "learning_rate": 8.599999999999999e-07,
      "loss": 0.5746,
      "step": 430
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.3169877231121063,
      "learning_rate": 8.799999999999999e-07,
      "loss": 0.4205,
      "step": 440
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.9648386240005493,
      "learning_rate": 9e-07,
      "loss": 0.5958,
      "step": 450
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.0,
      "learning_rate": 9.2e-07,
      "loss": 0.4632,
      "step": 460
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.3318028748035431,
      "learning_rate": 9.399999999999999e-07,
      "loss": 0.694,
      "step": 470
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.6652544140815735,
      "learning_rate": 9.6e-07,
      "loss": 0.5028,
      "step": 480
    },
    {
      "epoch": 0.098,
      "grad_norm": 1.2257837057113647,
      "learning_rate": 9.8e-07,
      "loss": 0.683,
      "step": 490
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.051938533782959,
      "learning_rate": 1e-06,
      "loss": 0.6468,
      "step": 500
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.36456993222236633,
      "learning_rate": 9.999878153526972e-07,
      "loss": 0.3102,
      "step": 510
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.6668845415115356,
      "learning_rate": 9.99951262004652e-07,
      "loss": 1.0302,
      "step": 520
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.8017757534980774,
      "learning_rate": 9.998903417374226e-07,
      "loss": 0.4671,
      "step": 530
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.7417317032814026,
      "learning_rate": 9.99805057520177e-07,
      "loss": 0.4329,
      "step": 540
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.0,
      "learning_rate": 9.996954135095478e-07,
      "loss": 0.5071,
      "step": 550
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.7268297076225281,
      "learning_rate": 9.99561415049429e-07,
      "loss": 0.6039,
      "step": 560
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.3723108470439911,
      "learning_rate": 9.99403068670717e-07,
      "loss": 0.3555,
      "step": 570
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.61618971824646,
      "learning_rate": 9.992203820909905e-07,
      "loss": 0.4968,
      "step": 580
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.5067716836929321,
      "learning_rate": 9.990133642141357e-07,
      "loss": 0.6656,
      "step": 590
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5604730248451233,
      "learning_rate": 9.98782025129912e-07,
      "loss": 0.5863,
      "step": 600
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.839195728302002,
      "learning_rate": 9.9852637611346e-07,
      "loss": 0.6651,
      "step": 610
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.0,
      "learning_rate": 9.982464296247522e-07,
      "loss": 0.468,
      "step": 620
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.35427430272102356,
      "learning_rate": 9.97942199307985e-07,
      "loss": 0.6606,
      "step": 630
    },
    {
      "epoch": 0.128,
      "grad_norm": 1.1575822830200195,
      "learning_rate": 9.976136999909155e-07,
      "loss": 0.6023,
      "step": 640
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.45604798197746277,
      "learning_rate": 9.972609476841365e-07,
      "loss": 0.5876,
      "step": 650
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.0,
      "learning_rate": 9.968839595802981e-07,
      "loss": 0.3609,
      "step": 660
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.4722711443901062,
      "learning_rate": 9.964827540532684e-07,
      "loss": 0.3629,
      "step": 670
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.34166085720062256,
      "learning_rate": 9.960573506572389e-07,
      "loss": 0.3346,
      "step": 680
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.6107274293899536,
      "learning_rate": 9.956077701257707e-07,
      "loss": 0.501,
      "step": 690
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.40759605169296265,
      "learning_rate": 9.95134034370785e-07,
      "loss": 0.4942,
      "step": 700
    },
    {
      "epoch": 0.142,
      "grad_norm": 0.8941773176193237,
      "learning_rate": 9.946361664814943e-07,
      "loss": 0.5462,
      "step": 710
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.416051983833313,
      "learning_rate": 9.941141907232763e-07,
      "loss": 0.4108,
      "step": 720
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.7638356685638428,
      "learning_rate": 9.93568132536494e-07,
      "loss": 0.417,
      "step": 730
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.25665077567100525,
      "learning_rate": 9.929980185352525e-07,
      "loss": 0.5837,
      "step": 740
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.3630926609039307,
      "learning_rate": 9.92403876506104e-07,
      "loss": 0.5312,
      "step": 750
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.37974926829338074,
      "learning_rate": 9.91785735406693e-07,
      "loss": 0.5809,
      "step": 760
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.46899351477622986,
      "learning_rate": 9.911436253643443e-07,
      "loss": 0.4575,
      "step": 770
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.788362443447113,
      "learning_rate": 9.904775776745956e-07,
      "loss": 0.4922,
      "step": 780
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.9416216611862183,
      "learning_rate": 9.89787624799672e-07,
      "loss": 0.4482,
      "step": 790
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.40510323643684387,
      "learning_rate": 9.890738003669027e-07,
      "loss": 0.3883,
      "step": 800
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.22145682573318481,
      "learning_rate": 9.883361391670839e-07,
      "loss": 0.3739,
      "step": 810
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.6583501696586609,
      "learning_rate": 9.875746771527815e-07,
      "loss": 0.4527,
      "step": 820
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.4808110296726227,
      "learning_rate": 9.8678945143658e-07,
      "loss": 0.5489,
      "step": 830
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.44500961899757385,
      "learning_rate": 9.859805002892731e-07,
      "loss": 0.4352,
      "step": 840
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.31563296914100647,
      "learning_rate": 9.851478631379982e-07,
      "loss": 0.7334,
      "step": 850
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.0,
      "learning_rate": 9.842915805643156e-07,
      "loss": 0.2776,
      "step": 860
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.546096920967102,
      "learning_rate": 9.834116943022297e-07,
      "loss": 0.4341,
      "step": 870
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.0,
      "learning_rate": 9.825082472361556e-07,
      "loss": 0.3616,
      "step": 880
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.48355451226234436,
      "learning_rate": 9.81581283398829e-07,
      "loss": 0.375,
      "step": 890
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.3133198618888855,
      "learning_rate": 9.806308479691594e-07,
      "loss": 0.3654,
      "step": 900
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.40422284603118896,
      "learning_rate": 9.796569872700287e-07,
      "loss": 0.4661,
      "step": 910
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.0,
      "learning_rate": 9.786597487660335e-07,
      "loss": 0.2929,
      "step": 920
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.6017171740531921,
      "learning_rate": 9.776391810611718e-07,
      "loss": 0.5152,
      "step": 930
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.7013118863105774,
      "learning_rate": 9.765953338964734e-07,
      "loss": 0.3121,
      "step": 940
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.35765185952186584,
      "learning_rate": 9.755282581475767e-07,
      "loss": 0.3817,
      "step": 950
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.4983702301979065,
      "learning_rate": 9.744380058222482e-07,
      "loss": 0.5493,
      "step": 960
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.5736071467399597,
      "learning_rate": 9.733246300578482e-07,
      "loss": 0.4638,
      "step": 970
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.3114262521266937,
      "learning_rate": 9.721881851187405e-07,
      "loss": 0.3779,
      "step": 980
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.7032444477081299,
      "learning_rate": 9.710287263936483e-07,
      "loss": 0.5224,
      "step": 990
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5082822442054749,
      "learning_rate": 9.698463103929541e-07,
      "loss": 0.3522,
      "step": 1000
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.29254528880119324,
      "learning_rate": 9.686409947459457e-07,
      "loss": 0.5438,
      "step": 1010
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.5429019927978516,
      "learning_rate": 9.674128381980071e-07,
      "loss": 0.3603,
      "step": 1020
    },
    {
      "epoch": 0.206,
      "grad_norm": 0.6649332046508789,
      "learning_rate": 9.661619006077561e-07,
      "loss": 0.4164,
      "step": 1030
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.4190509021282196,
      "learning_rate": 9.648882429441256e-07,
      "loss": 0.3494,
      "step": 1040
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.0,
      "learning_rate": 9.635919272833937e-07,
      "loss": 0.2614,
      "step": 1050
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.5640019774436951,
      "learning_rate": 9.622730168061567e-07,
      "loss": 0.4488,
      "step": 1060
    },
    {
      "epoch": 0.214,
      "grad_norm": 0.4980197250843048,
      "learning_rate": 9.609315757942502e-07,
      "loss": 0.4833,
      "step": 1070
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.5959740281105042,
      "learning_rate": 9.595676696276171e-07,
      "loss": 0.5463,
      "step": 1080
    },
    {
      "epoch": 0.218,
      "grad_norm": 0.5091766715049744,
      "learning_rate": 9.581813647811197e-07,
      "loss": 0.4226,
      "step": 1090
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.7578771114349365,
      "learning_rate": 9.567727288213004e-07,
      "loss": 0.4465,
      "step": 1100
    },
    {
      "epoch": 0.222,
      "grad_norm": 0.849757969379425,
      "learning_rate": 9.553418304030885e-07,
      "loss": 0.4341,
      "step": 1110
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.2365778684616089,
      "learning_rate": 9.538887392664543e-07,
      "loss": 0.252,
      "step": 1120
    },
    {
      "epoch": 0.226,
      "grad_norm": 0.28902509808540344,
      "learning_rate": 9.524135262330098e-07,
      "loss": 0.4412,
      "step": 1130
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.29451924562454224,
      "learning_rate": 9.509162632025569e-07,
      "loss": 0.383,
      "step": 1140
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6582534313201904,
      "learning_rate": 9.493970231495834e-07,
      "loss": 0.498,
      "step": 1150
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.43566933274269104,
      "learning_rate": 9.478558801197064e-07,
      "loss": 0.4814,
      "step": 1160
    },
    {
      "epoch": 0.234,
      "grad_norm": 0.5926059484481812,
      "learning_rate": 9.462929092260628e-07,
      "loss": 0.3324,
      "step": 1170
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.2790277302265167,
      "learning_rate": 9.447081866456487e-07,
      "loss": 0.2978,
      "step": 1180
    },
    {
      "epoch": 0.238,
      "grad_norm": 0.26161783933639526,
      "learning_rate": 9.431017896156073e-07,
      "loss": 0.3284,
      "step": 1190
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.7597343921661377,
      "learning_rate": 9.414737964294634e-07,
      "loss": 0.3644,
      "step": 1200
    },
    {
      "epoch": 0.242,
      "grad_norm": 0.5080224275588989,
      "learning_rate": 9.398242864333083e-07,
      "loss": 0.3615,
      "step": 1210
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.9933916330337524,
      "learning_rate": 9.381533400219317e-07,
      "loss": 0.5431,
      "step": 1220
    },
    {
      "epoch": 0.246,
      "grad_norm": 0.6644570231437683,
      "learning_rate": 9.364610386349047e-07,
      "loss": 0.4106,
      "step": 1230
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.6827860474586487,
      "learning_rate": 9.347474647526095e-07,
      "loss": 0.4181,
      "step": 1240
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.580407440662384,
      "learning_rate": 9.330127018922193e-07,
      "loss": 0.2969,
      "step": 1250
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.7604784369468689,
      "learning_rate": 9.312568346036287e-07,
      "loss": 0.4105,
      "step": 1260
    },
    {
      "epoch": 0.254,
      "grad_norm": 0.42066580057144165,
      "learning_rate": 9.294799484653322e-07,
      "loss": 0.3617,
      "step": 1270
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.8827233910560608,
      "learning_rate": 9.276821300802533e-07,
      "loss": 0.488,
      "step": 1280
    },
    {
      "epoch": 0.258,
      "grad_norm": 0.586871325969696,
      "learning_rate": 9.258634670715237e-07,
      "loss": 0.2852,
      "step": 1290
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.2628524601459503,
      "learning_rate": 9.240240480782129e-07,
      "loss": 0.3642,
      "step": 1300
    },
    {
      "epoch": 0.262,
      "grad_norm": 0.4368833601474762,
      "learning_rate": 9.221639627510075e-07,
      "loss": 0.3637,
      "step": 1310
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.5321241617202759,
      "learning_rate": 9.202833017478421e-07,
      "loss": 0.3316,
      "step": 1320
    },
    {
      "epoch": 0.266,
      "grad_norm": 0.2961817681789398,
      "learning_rate": 9.183821567294808e-07,
      "loss": 0.3285,
      "step": 1330
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.16444119811058044,
      "learning_rate": 9.164606203550497e-07,
      "loss": 0.3077,
      "step": 1340
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.28540509939193726,
      "learning_rate": 9.145187862775208e-07,
      "loss": 0.347,
      "step": 1350
    },
    {
      "epoch": 0.272,
      "grad_norm": 2.098088264465332,
      "learning_rate": 9.125567491391475e-07,
      "loss": 0.4526,
      "step": 1360
    },
    {
      "epoch": 0.274,
      "grad_norm": 0.4260193109512329,
      "learning_rate": 9.10574604566852e-07,
      "loss": 0.3004,
      "step": 1370
    },
    {
      "epoch": 0.276,
      "grad_norm": 3.244032621383667,
      "learning_rate": 9.085724491675642e-07,
      "loss": 0.4287,
      "step": 1380
    },
    {
      "epoch": 0.278,
      "grad_norm": 0.31434857845306396,
      "learning_rate": 9.065503805235137e-07,
      "loss": 0.2818,
      "step": 1390
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3869690001010895,
      "learning_rate": 9.045084971874737e-07,
      "loss": 0.247,
      "step": 1400
    },
    {
      "epoch": 0.282,
      "grad_norm": 2.252443790435791,
      "learning_rate": 9.02446898677957e-07,
      "loss": 0.4637,
      "step": 1410
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.4407874047756195,
      "learning_rate": 9.003656854743666e-07,
      "loss": 0.321,
      "step": 1420
    },
    {
      "epoch": 0.286,
      "grad_norm": 0.2926657795906067,
      "learning_rate": 8.982649590120981e-07,
      "loss": 0.3643,
      "step": 1430
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.8922311663627625,
      "learning_rate": 8.961448216775953e-07,
      "loss": 0.4458,
      "step": 1440
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.5968508124351501,
      "learning_rate": 8.940053768033608e-07,
      "loss": 0.3387,
      "step": 1450
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.502495288848877,
      "learning_rate": 8.918467286629198e-07,
      "loss": 0.2984,
      "step": 1460
    },
    {
      "epoch": 0.294,
      "grad_norm": 0.2826671898365021,
      "learning_rate": 8.896689824657371e-07,
      "loss": 0.2773,
      "step": 1470
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.6212120056152344,
      "learning_rate": 8.874722443520898e-07,
      "loss": 0.3047,
      "step": 1480
    },
    {
      "epoch": 0.298,
      "grad_norm": 0.35743391513824463,
      "learning_rate": 8.852566213878946e-07,
      "loss": 0.249,
      "step": 1490
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.21640463173389435,
      "learning_rate": 8.83022221559489e-07,
      "loss": 0.5102,
      "step": 1500
    },
    {
      "epoch": 0.302,
      "grad_norm": 0.5322903990745544,
      "learning_rate": 8.807691537683684e-07,
      "loss": 0.4785,
      "step": 1510
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.5649182200431824,
      "learning_rate": 8.784975278258782e-07,
      "loss": 0.4196,
      "step": 1520
    },
    {
      "epoch": 0.306,
      "grad_norm": 0.8210440278053284,
      "learning_rate": 8.762074544478621e-07,
      "loss": 0.4145,
      "step": 1530
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.5836989283561707,
      "learning_rate": 8.73899045249266e-07,
      "loss": 0.3771,
      "step": 1540
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3912012279033661,
      "learning_rate": 8.71572412738697e-07,
      "loss": 0.275,
      "step": 1550
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.9098830819129944,
      "learning_rate": 8.69227670312942e-07,
      "loss": 0.3984,
      "step": 1560
    },
    {
      "epoch": 0.314,
      "grad_norm": 0.34057459235191345,
      "learning_rate": 8.668649322514381e-07,
      "loss": 0.198,
      "step": 1570
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.5625966787338257,
      "learning_rate": 8.644843137107057e-07,
      "loss": 0.3518,
      "step": 1580
    },
    {
      "epoch": 0.318,
      "grad_norm": 0.42884501814842224,
      "learning_rate": 8.620859307187338e-07,
      "loss": 0.2761,
      "step": 1590
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.45647886395454407,
      "learning_rate": 8.596699001693255e-07,
      "loss": 0.2784,
      "step": 1600
    },
    {
      "epoch": 0.322,
      "grad_norm": 0.6005297899246216,
      "learning_rate": 8.572363398164016e-07,
      "loss": 0.314,
      "step": 1610
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.0,
      "learning_rate": 8.547853682682604e-07,
      "loss": 0.3661,
      "step": 1620
    },
    {
      "epoch": 0.326,
      "grad_norm": 0.1766682118177414,
      "learning_rate": 8.523171049817973e-07,
      "loss": 0.3609,
      "step": 1630
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.3773224353790283,
      "learning_rate": 8.498316702566826e-07,
      "loss": 0.3768,
      "step": 1640
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.8536304235458374,
      "learning_rate": 8.473291852294986e-07,
      "loss": 0.3376,
      "step": 1650
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.40247610211372375,
      "learning_rate": 8.448097718678348e-07,
      "loss": 0.3568,
      "step": 1660
    },
    {
      "epoch": 0.334,
      "grad_norm": 0.711200475692749,
      "learning_rate": 8.422735529643443e-07,
      "loss": 0.245,
      "step": 1670
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.271049439907074,
      "learning_rate": 8.397206521307583e-07,
      "loss": 0.2278,
      "step": 1680
    },
    {
      "epoch": 0.338,
      "grad_norm": 0.7005532383918762,
      "learning_rate": 8.371511937918617e-07,
      "loss": 0.3255,
      "step": 1690
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.2691313922405243,
      "learning_rate": 8.34565303179429e-07,
      "loss": 0.2114,
      "step": 1700
    },
    {
      "epoch": 0.342,
      "grad_norm": 0.2832666337490082,
      "learning_rate": 8.319631063261207e-07,
      "loss": 0.3853,
      "step": 1710
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.5301315188407898,
      "learning_rate": 8.293447300593402e-07,
      "loss": 0.3129,
      "step": 1720
    },
    {
      "epoch": 0.346,
      "grad_norm": 1.369352102279663,
      "learning_rate": 8.267103019950528e-07,
      "loss": 0.2986,
      "step": 1730
    },
    {
      "epoch": 0.348,
      "grad_norm": 1.4075043201446533,
      "learning_rate": 8.240599505315654e-07,
      "loss": 0.3362,
      "step": 1740
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.5943218469619751,
      "learning_rate": 8.213938048432696e-07,
      "loss": 0.2817,
      "step": 1750
    },
    {
      "epoch": 0.352,
      "grad_norm": 1.011788249015808,
      "learning_rate": 8.187119948743449e-07,
      "loss": 0.3021,
      "step": 1760
    },
    {
      "epoch": 0.354,
      "grad_norm": 0.20643028616905212,
      "learning_rate": 8.160146513324254e-07,
      "loss": 0.38,
      "step": 1770
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.623815655708313,
      "learning_rate": 8.133019056822302e-07,
      "loss": 0.3966,
      "step": 1780
    },
    {
      "epoch": 0.358,
      "grad_norm": 0.9031881093978882,
      "learning_rate": 8.105738901391551e-07,
      "loss": 0.25,
      "step": 1790
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.29175126552581787,
      "learning_rate": 8.07830737662829e-07,
      "loss": 0.262,
      "step": 1800
    },
    {
      "epoch": 0.362,
      "grad_norm": 0.6108918190002441,
      "learning_rate": 8.050725819506339e-07,
      "loss": 0.2751,
      "step": 1810
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.32167190313339233,
      "learning_rate": 8.022995574311875e-07,
      "loss": 0.2439,
      "step": 1820
    },
    {
      "epoch": 0.366,
      "grad_norm": 1.1157898902893066,
      "learning_rate": 7.995117992577928e-07,
      "loss": 0.3475,
      "step": 1830
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.0,
      "learning_rate": 7.967094433018508e-07,
      "loss": 0.3295,
      "step": 1840
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3029947280883789,
      "learning_rate": 7.938926261462365e-07,
      "loss": 0.257,
      "step": 1850
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.2989325225353241,
      "learning_rate": 7.910614850786447e-07,
      "loss": 0.3749,
      "step": 1860
    },
    {
      "epoch": 0.374,
      "grad_norm": 0.3622772693634033,
      "learning_rate": 7.882161580848966e-07,
      "loss": 0.373,
      "step": 1870
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.49973657727241516,
      "learning_rate": 7.853567838422159e-07,
      "loss": 0.3807,
      "step": 1880
    },
    {
      "epoch": 0.378,
      "grad_norm": 0.33801835775375366,
      "learning_rate": 7.82483501712469e-07,
      "loss": 0.1554,
      "step": 1890
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8740425705909729,
      "learning_rate": 7.795964517353733e-07,
      "loss": 0.2553,
      "step": 1900
    },
    {
      "epoch": 0.382,
      "grad_norm": 0.6293899416923523,
      "learning_rate": 7.76695774621672e-07,
      "loss": 0.2488,
      "step": 1910
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.6553785800933838,
      "learning_rate": 7.737816117462751e-07,
      "loss": 0.2794,
      "step": 1920
    },
    {
      "epoch": 0.386,
      "grad_norm": 0.6106793880462646,
      "learning_rate": 7.7085410514137e-07,
      "loss": 0.2829,
      "step": 1930
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.6467621326446533,
      "learning_rate": 7.679133974894982e-07,
      "loss": 0.3499,
      "step": 1940
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3079082667827606,
      "learning_rate": 7.649596321166024e-07,
      "loss": 0.1809,
      "step": 1950
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.0,
      "learning_rate": 7.619929529850396e-07,
      "loss": 0.2201,
      "step": 1960
    },
    {
      "epoch": 0.394,
      "grad_norm": 0.7928106784820557,
      "learning_rate": 7.590135046865651e-07,
      "loss": 0.2696,
      "step": 1970
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.4864124357700348,
      "learning_rate": 7.560214324352858e-07,
      "loss": 0.295,
      "step": 1980
    },
    {
      "epoch": 0.398,
      "grad_norm": 0.41892245411872864,
      "learning_rate": 7.530168820605818e-07,
      "loss": 0.1679,
      "step": 1990
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.443067193031311,
      "learning_rate": 7.5e-07,
      "loss": 0.3207,
      "step": 2000
    },
    {
      "epoch": 0.402,
      "grad_norm": 0.29227420687675476,
      "learning_rate": 7.469709332921154e-07,
      "loss": 0.2237,
      "step": 2010
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.9847503304481506,
      "learning_rate": 7.439298295693663e-07,
      "loss": 0.2219,
      "step": 2020
    },
    {
      "epoch": 0.406,
      "grad_norm": 0.5221825838088989,
      "learning_rate": 7.408768370508576e-07,
      "loss": 0.2063,
      "step": 2030
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.781994640827179,
      "learning_rate": 7.378121045351377e-07,
      "loss": 0.2827,
      "step": 2040
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4918607771396637,
      "learning_rate": 7.347357813929454e-07,
      "loss": 0.2603,
      "step": 2050
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.2811141610145569,
      "learning_rate": 7.316480175599308e-07,
      "loss": 0.1887,
      "step": 2060
    },
    {
      "epoch": 0.414,
      "grad_norm": 0.6080925464630127,
      "learning_rate": 7.285489635293471e-07,
      "loss": 0.2102,
      "step": 2070
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.4952521026134491,
      "learning_rate": 7.254387703447153e-07,
      "loss": 0.2386,
      "step": 2080
    },
    {
      "epoch": 0.418,
      "grad_norm": 0.44920068979263306,
      "learning_rate": 7.223175895924637e-07,
      "loss": 0.203,
      "step": 2090
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4240502417087555,
      "learning_rate": 7.191855733945386e-07,
      "loss": 0.268,
      "step": 2100
    },
    {
      "epoch": 0.422,
      "grad_norm": 1.42499577999115,
      "learning_rate": 7.160428744009912e-07,
      "loss": 0.3531,
      "step": 2110
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.6819471716880798,
      "learning_rate": 7.128896457825363e-07,
      "loss": 0.3363,
      "step": 2120
    },
    {
      "epoch": 0.426,
      "grad_norm": 0.4226500391960144,
      "learning_rate": 7.097260412230885e-07,
      "loss": 0.3331,
      "step": 2130
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.32036811113357544,
      "learning_rate": 7.065522149122709e-07,
      "loss": 0.2405,
      "step": 2140
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.4470788538455963,
      "learning_rate": 7.033683215379002e-07,
      "loss": 0.1376,
      "step": 2150
    },
    {
      "epoch": 0.432,
      "grad_norm": 1.326379418373108,
      "learning_rate": 7.001745162784475e-07,
      "loss": 0.4589,
      "step": 2160
    },
    {
      "epoch": 0.434,
      "grad_norm": 0.26921626925468445,
      "learning_rate": 6.969709547954755e-07,
      "loss": 0.2441,
      "step": 2170
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.6607092618942261,
      "learning_rate": 6.937577932260514e-07,
      "loss": 0.2402,
      "step": 2180
    },
    {
      "epoch": 0.438,
      "grad_norm": 0.40199458599090576,
      "learning_rate": 6.905351881751371e-07,
      "loss": 0.2452,
      "step": 2190
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.0,
      "learning_rate": 6.87303296707956e-07,
      "loss": 0.1521,
      "step": 2200
    },
    {
      "epoch": 0.442,
      "grad_norm": 0.5950799584388733,
      "learning_rate": 6.840622763423391e-07,
      "loss": 0.2665,
      "step": 2210
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.33319640159606934,
      "learning_rate": 6.80812285041046e-07,
      "loss": 0.2881,
      "step": 2220
    },
    {
      "epoch": 0.446,
      "grad_norm": 0.2309122532606125,
      "learning_rate": 6.775534812040686e-07,
      "loss": 0.2833,
      "step": 2230
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.5239594578742981,
      "learning_rate": 6.742860236609076e-07,
      "loss": 0.2009,
      "step": 2240
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.32858043909072876,
      "learning_rate": 6.710100716628344e-07,
      "loss": 0.2794,
      "step": 2250
    },
    {
      "epoch": 0.452,
      "grad_norm": 1.1937505006790161,
      "learning_rate": 6.677257848751276e-07,
      "loss": 0.2411,
      "step": 2260
    },
    {
      "epoch": 0.454,
      "grad_norm": 0.4281042218208313,
      "learning_rate": 6.644333233692916e-07,
      "loss": 0.1991,
      "step": 2270
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.3388921916484833,
      "learning_rate": 6.611328476152556e-07,
      "loss": 0.3334,
      "step": 2280
    },
    {
      "epoch": 0.458,
      "grad_norm": 0.3999691903591156,
      "learning_rate": 6.578245184735512e-07,
      "loss": 0.3575,
      "step": 2290
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5984915494918823,
      "learning_rate": 6.545084971874736e-07,
      "loss": 0.3567,
      "step": 2300
    },
    {
      "epoch": 0.462,
      "grad_norm": 0.5047705769538879,
      "learning_rate": 6.511849453752223e-07,
      "loss": 0.2235,
      "step": 2310
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.19550199806690216,
      "learning_rate": 6.478540250220233e-07,
      "loss": 0.3879,
      "step": 2320
    },
    {
      "epoch": 0.466,
      "grad_norm": 0.7670475244522095,
      "learning_rate": 6.445158984722358e-07,
      "loss": 0.2557,
      "step": 2330
    },
    {
      "epoch": 0.468,
      "grad_norm": 0.20435401797294617,
      "learning_rate": 6.411707284214383e-07,
      "loss": 0.2397,
      "step": 2340
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2008686661720276,
      "learning_rate": 6.378186779084995e-07,
      "loss": 0.3854,
      "step": 2350
    },
    {
      "epoch": 0.472,
      "grad_norm": 1.6063778400421143,
      "learning_rate": 6.344599103076328e-07,
      "loss": 0.2838,
      "step": 2360
    },
    {
      "epoch": 0.474,
      "grad_norm": 0.9098937511444092,
      "learning_rate": 6.310945893204324e-07,
      "loss": 0.3221,
      "step": 2370
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.4095039963722229,
      "learning_rate": 6.277228789678953e-07,
      "loss": 0.2223,
      "step": 2380
    },
    {
      "epoch": 0.478,
      "grad_norm": 0.31623950600624084,
      "learning_rate": 6.243449435824276e-07,
      "loss": 0.2595,
      "step": 2390
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6135280728340149,
      "learning_rate": 6.209609477998338e-07,
      "loss": 0.19,
      "step": 2400
    },
    {
      "epoch": 0.482,
      "grad_norm": 0.8254637122154236,
      "learning_rate": 6.17571056551295e-07,
      "loss": 0.2618,
      "step": 2410
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.33638447523117065,
      "learning_rate": 6.141754350553279e-07,
      "loss": 0.3483,
      "step": 2420
    },
    {
      "epoch": 0.486,
      "grad_norm": 0.2547425925731659,
      "learning_rate": 6.107742488097338e-07,
      "loss": 0.184,
      "step": 2430
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.23442941904067993,
      "learning_rate": 6.073676635835316e-07,
      "loss": 0.3503,
      "step": 2440
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.33613914251327515,
      "learning_rate": 6.039558454088795e-07,
      "loss": 0.3457,
      "step": 2450
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.5778683423995972,
      "learning_rate": 6.005389605729824e-07,
      "loss": 0.3099,
      "step": 2460
    },
    {
      "epoch": 0.494,
      "grad_norm": 0.0,
      "learning_rate": 5.97117175609986e-07,
      "loss": 0.2064,
      "step": 2470
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.3998168110847473,
      "learning_rate": 5.936906572928624e-07,
      "loss": 0.2502,
      "step": 2480
    },
    {
      "epoch": 0.498,
      "grad_norm": 0.35442838072776794,
      "learning_rate": 5.9025957262528e-07,
      "loss": 0.2189,
      "step": 2490
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.31255003809928894,
      "learning_rate": 5.868240888334652e-07,
      "loss": 0.2579,
      "step": 2500
    },
    {
      "epoch": 0.502,
      "grad_norm": 0.3876568675041199,
      "learning_rate": 5.833843733580512e-07,
      "loss": 0.291,
      "step": 2510
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.7023966908454895,
      "learning_rate": 5.799405938459174e-07,
      "loss": 0.1659,
      "step": 2520
    },
    {
      "epoch": 0.506,
      "grad_norm": 0.42696765065193176,
      "learning_rate": 5.764929181420191e-07,
      "loss": 0.2178,
      "step": 2530
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.24111975729465485,
      "learning_rate": 5.730415142812058e-07,
      "loss": 0.2218,
      "step": 2540
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.253818154335022,
      "learning_rate": 5.695865504800327e-07,
      "loss": 0.2646,
      "step": 2550
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.4682632088661194,
      "learning_rate": 5.661281951285612e-07,
      "loss": 0.3102,
      "step": 2560
    },
    {
      "epoch": 0.514,
      "grad_norm": 0.7523025870323181,
      "learning_rate": 5.626666167821521e-07,
      "loss": 0.3413,
      "step": 2570
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.3234543204307556,
      "learning_rate": 5.592019841532506e-07,
      "loss": 0.1998,
      "step": 2580
    },
    {
      "epoch": 0.518,
      "grad_norm": 0.9460989832878113,
      "learning_rate": 5.557344661031627e-07,
      "loss": 0.2994,
      "step": 2590
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.8611673712730408,
      "learning_rate": 5.522642316338268e-07,
      "loss": 0.2624,
      "step": 2600
    },
    {
      "epoch": 0.522,
      "grad_norm": 0.3594551086425781,
      "learning_rate": 5.487914498795747e-07,
      "loss": 0.2005,
      "step": 2610
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.5597927570343018,
      "learning_rate": 5.453162900988901e-07,
      "loss": 0.3244,
      "step": 2620
    },
    {
      "epoch": 0.526,
      "grad_norm": 0.3440828025341034,
      "learning_rate": 5.418389216661578e-07,
      "loss": 0.2086,
      "step": 2630
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.4287198781967163,
      "learning_rate": 5.383595140634093e-07,
      "loss": 0.2158,
      "step": 2640
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.7096256017684937,
      "learning_rate": 5.348782368720625e-07,
      "loss": 0.2943,
      "step": 2650
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.32011571526527405,
      "learning_rate": 5.313952597646567e-07,
      "loss": 0.3096,
      "step": 2660
    },
    {
      "epoch": 0.534,
      "grad_norm": 0.47610387206077576,
      "learning_rate": 5.27910752496582e-07,
      "loss": 0.3096,
      "step": 2670
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.0,
      "learning_rate": 5.244248848978067e-07,
      "loss": 0.2443,
      "step": 2680
    },
    {
      "epoch": 0.538,
      "grad_norm": 0.4076542854309082,
      "learning_rate": 5.209378268645997e-07,
      "loss": 0.2085,
      "step": 2690
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8643516898155212,
      "learning_rate": 5.174497483512505e-07,
      "loss": 0.3603,
      "step": 2700
    },
    {
      "epoch": 0.542,
      "grad_norm": 0.3362959027290344,
      "learning_rate": 5.139608193617844e-07,
      "loss": 0.2014,
      "step": 2710
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.35498881340026855,
      "learning_rate": 5.104712099416785e-07,
      "loss": 0.2257,
      "step": 2720
    },
    {
      "epoch": 0.546,
      "grad_norm": 0.48586198687553406,
      "learning_rate": 5.069810901695727e-07,
      "loss": 0.2808,
      "step": 2730
    },
    {
      "epoch": 0.548,
      "grad_norm": 0.35722529888153076,
      "learning_rate": 5.034906301489807e-07,
      "loss": 0.13,
      "step": 2740
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.3690369725227356,
      "learning_rate": 5e-07,
      "loss": 0.3317,
      "step": 2750
    },
    {
      "epoch": 0.552,
      "grad_norm": 1.0494160652160645,
      "learning_rate": 4.965093698510192e-07,
      "loss": 0.2685,
      "step": 2760
    },
    {
      "epoch": 0.554,
      "grad_norm": 0.4894683063030243,
      "learning_rate": 4.930189098304274e-07,
      "loss": 0.2293,
      "step": 2770
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.6006208062171936,
      "learning_rate": 4.895287900583216e-07,
      "loss": 0.247,
      "step": 2780
    },
    {
      "epoch": 0.558,
      "grad_norm": 0.27367955446243286,
      "learning_rate": 4.860391806382156e-07,
      "loss": 0.2167,
      "step": 2790
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.0968672037124634,
      "learning_rate": 4.825502516487496e-07,
      "loss": 0.2455,
      "step": 2800
    },
    {
      "epoch": 0.562,
      "grad_norm": 0.0,
      "learning_rate": 4.790621731354002e-07,
      "loss": 0.2411,
      "step": 2810
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.3560165762901306,
      "learning_rate": 4.7557511510219335e-07,
      "loss": 0.1861,
      "step": 2820
    },
    {
      "epoch": 0.566,
      "grad_norm": 0.3030807673931122,
      "learning_rate": 4.7208924750341805e-07,
      "loss": 0.1857,
      "step": 2830
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.5121121406555176,
      "learning_rate": 4.686047402353433e-07,
      "loss": 0.2738,
      "step": 2840
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2215707153081894,
      "learning_rate": 4.6512176312793735e-07,
      "loss": 0.2099,
      "step": 2850
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.0,
      "learning_rate": 4.6164048593659065e-07,
      "loss": 0.1693,
      "step": 2860
    },
    {
      "epoch": 0.574,
      "grad_norm": 0.5592711567878723,
      "learning_rate": 4.5816107833384233e-07,
      "loss": 0.3155,
      "step": 2870
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.49754905700683594,
      "learning_rate": 4.5468370990110997e-07,
      "loss": 0.208,
      "step": 2880
    },
    {
      "epoch": 0.578,
      "grad_norm": 0.6235384941101074,
      "learning_rate": 4.512085501204253e-07,
      "loss": 0.2906,
      "step": 2890
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.35063299536705017,
      "learning_rate": 4.477357683661733e-07,
      "loss": 0.291,
      "step": 2900
    },
    {
      "epoch": 0.582,
      "grad_norm": 0.8492065668106079,
      "learning_rate": 4.442655338968373e-07,
      "loss": 0.2493,
      "step": 2910
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.31068241596221924,
      "learning_rate": 4.407980158467495e-07,
      "loss": 0.2199,
      "step": 2920
    },
    {
      "epoch": 0.586,
      "grad_norm": 0.16246047616004944,
      "learning_rate": 4.3733338321784777e-07,
      "loss": 0.2611,
      "step": 2930
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.5197228789329529,
      "learning_rate": 4.338718048714387e-07,
      "loss": 0.2822,
      "step": 2940
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9255732297897339,
      "learning_rate": 4.304134495199674e-07,
      "loss": 0.228,
      "step": 2950
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.6209078431129456,
      "learning_rate": 4.2695848571879424e-07,
      "loss": 0.2105,
      "step": 2960
    },
    {
      "epoch": 0.594,
      "grad_norm": 0.699409544467926,
      "learning_rate": 4.23507081857981e-07,
      "loss": 0.2484,
      "step": 2970
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.596609890460968,
      "learning_rate": 4.200594061540826e-07,
      "loss": 0.2553,
      "step": 2980
    },
    {
      "epoch": 0.598,
      "grad_norm": 0.5129525661468506,
      "learning_rate": 4.166156266419489e-07,
      "loss": 0.2247,
      "step": 2990
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.39060530066490173,
      "learning_rate": 4.131759111665348e-07,
      "loss": 0.3284,
      "step": 3000
    },
    {
      "epoch": 0.602,
      "grad_norm": 0.41641587018966675,
      "learning_rate": 4.0974042737472005e-07,
      "loss": 0.3142,
      "step": 3010
    },
    {
      "epoch": 0.604,
      "grad_norm": 0.9066689014434814,
      "learning_rate": 4.0630934270713755e-07,
      "loss": 0.2861,
      "step": 3020
    },
    {
      "epoch": 0.606,
      "grad_norm": 0.5587870478630066,
      "learning_rate": 4.028828243900141e-07,
      "loss": 0.2254,
      "step": 3030
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.47786420583724976,
      "learning_rate": 3.9946103942701775e-07,
      "loss": 0.2112,
      "step": 3040
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.6182255744934082,
      "learning_rate": 3.960441545911204e-07,
      "loss": 0.3442,
      "step": 3050
    },
    {
      "epoch": 0.612,
      "grad_norm": 0.8767617344856262,
      "learning_rate": 3.9263233641646836e-07,
      "loss": 0.294,
      "step": 3060
    },
    {
      "epoch": 0.614,
      "grad_norm": 0.0,
      "learning_rate": 3.8922575119026635e-07,
      "loss": 0.1191,
      "step": 3070
    },
    {
      "epoch": 0.616,
      "grad_norm": 0.3942841589450836,
      "learning_rate": 3.8582456494467206e-07,
      "loss": 0.2486,
      "step": 3080
    },
    {
      "epoch": 0.618,
      "grad_norm": 0.457390159368515,
      "learning_rate": 3.8242894344870495e-07,
      "loss": 0.291,
      "step": 3090
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.34195929765701294,
      "learning_rate": 3.790390522001662e-07,
      "loss": 0.2265,
      "step": 3100
    },
    {
      "epoch": 0.622,
      "grad_norm": 0.2878647744655609,
      "learning_rate": 3.7565505641757266e-07,
      "loss": 0.274,
      "step": 3110
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.0,
      "learning_rate": 3.722771210321048e-07,
      "loss": 0.1274,
      "step": 3120
    },
    {
      "epoch": 0.626,
      "grad_norm": 0.38419225811958313,
      "learning_rate": 3.689054106795677e-07,
      "loss": 0.2817,
      "step": 3130
    },
    {
      "epoch": 0.628,
      "grad_norm": 0.4910019636154175,
      "learning_rate": 3.6554008969236715e-07,
      "loss": 0.2044,
      "step": 3140
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2987191379070282,
      "learning_rate": 3.621813220915004e-07,
      "loss": 0.2311,
      "step": 3150
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.3759290874004364,
      "learning_rate": 3.5882927157856167e-07,
      "loss": 0.2061,
      "step": 3160
    },
    {
      "epoch": 0.634,
      "grad_norm": 0.37429606914520264,
      "learning_rate": 3.554841015277641e-07,
      "loss": 0.2124,
      "step": 3170
    },
    {
      "epoch": 0.636,
      "grad_norm": 0.2639157772064209,
      "learning_rate": 3.521459749779768e-07,
      "loss": 0.2445,
      "step": 3180
    },
    {
      "epoch": 0.638,
      "grad_norm": 0.2917717695236206,
      "learning_rate": 3.488150546247778e-07,
      "loss": 0.325,
      "step": 3190
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2606163024902344,
      "learning_rate": 3.454915028125263e-07,
      "loss": 0.1581,
      "step": 3200
    },
    {
      "epoch": 0.642,
      "grad_norm": 0.0,
      "learning_rate": 3.421754815264488e-07,
      "loss": 0.2244,
      "step": 3210
    },
    {
      "epoch": 0.644,
      "grad_norm": 0.6836721897125244,
      "learning_rate": 3.388671523847445e-07,
      "loss": 0.178,
      "step": 3220
    },
    {
      "epoch": 0.646,
      "grad_norm": 0.36950552463531494,
      "learning_rate": 3.3556667663070835e-07,
      "loss": 0.2215,
      "step": 3230
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.7161170840263367,
      "learning_rate": 3.3227421512487255e-07,
      "loss": 0.1874,
      "step": 3240
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.33053773641586304,
      "learning_rate": 3.2898992833716563e-07,
      "loss": 0.1942,
      "step": 3250
    },
    {
      "epoch": 0.652,
      "grad_norm": 0.39361345767974854,
      "learning_rate": 3.257139763390925e-07,
      "loss": 0.2147,
      "step": 3260
    },
    {
      "epoch": 0.654,
      "grad_norm": 0.34917065501213074,
      "learning_rate": 3.2244651879593156e-07,
      "loss": 0.2706,
      "step": 3270
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.29661568999290466,
      "learning_rate": 3.191877149589539e-07,
      "loss": 0.2537,
      "step": 3280
    },
    {
      "epoch": 0.658,
      "grad_norm": 1.5755507946014404,
      "learning_rate": 3.15937723657661e-07,
      "loss": 0.2887,
      "step": 3290
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.30394697189331055,
      "learning_rate": 3.1269670329204393e-07,
      "loss": 0.1861,
      "step": 3300
    },
    {
      "epoch": 0.662,
      "grad_norm": 0.666857123374939,
      "learning_rate": 3.0946481182486297e-07,
      "loss": 0.2361,
      "step": 3310
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.4149036109447479,
      "learning_rate": 3.0624220677394854e-07,
      "loss": 0.3015,
      "step": 3320
    },
    {
      "epoch": 0.666,
      "grad_norm": 0.5183743238449097,
      "learning_rate": 3.0302904520452443e-07,
      "loss": 0.2615,
      "step": 3330
    },
    {
      "epoch": 0.668,
      "grad_norm": 0.2939796447753906,
      "learning_rate": 2.9982548372155256e-07,
      "loss": 0.2192,
      "step": 3340
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.0,
      "learning_rate": 2.9663167846209996e-07,
      "loss": 0.1743,
      "step": 3350
    },
    {
      "epoch": 0.672,
      "grad_norm": 1.0357635021209717,
      "learning_rate": 2.9344778508772914e-07,
      "loss": 0.1922,
      "step": 3360
    },
    {
      "epoch": 0.674,
      "grad_norm": 0.22886469960212708,
      "learning_rate": 2.902739587769114e-07,
      "loss": 0.2038,
      "step": 3370
    },
    {
      "epoch": 0.676,
      "grad_norm": 0.5248194336891174,
      "learning_rate": 2.8711035421746363e-07,
      "loss": 0.1597,
      "step": 3380
    },
    {
      "epoch": 0.678,
      "grad_norm": 0.26883548498153687,
      "learning_rate": 2.8395712559900874e-07,
      "loss": 0.2412,
      "step": 3390
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.0,
      "learning_rate": 2.808144266054612e-07,
      "loss": 0.3742,
      "step": 3400
    },
    {
      "epoch": 0.682,
      "grad_norm": 0.5440124273300171,
      "learning_rate": 2.776824104075364e-07,
      "loss": 0.1784,
      "step": 3410
    },
    {
      "epoch": 0.684,
      "grad_norm": 0.0,
      "learning_rate": 2.745612296552847e-07,
      "loss": 0.1994,
      "step": 3420
    },
    {
      "epoch": 0.686,
      "grad_norm": 0.5974320769309998,
      "learning_rate": 2.71451036470653e-07,
      "loss": 0.251,
      "step": 3430
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.0,
      "learning_rate": 2.683519824400692e-07,
      "loss": 0.1304,
      "step": 3440
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.6028246283531189,
      "learning_rate": 2.6526421860705473e-07,
      "loss": 0.1921,
      "step": 3450
    },
    {
      "epoch": 0.692,
      "grad_norm": 0.27002623677253723,
      "learning_rate": 2.621878954648623e-07,
      "loss": 0.2107,
      "step": 3460
    },
    {
      "epoch": 0.694,
      "grad_norm": 0.47267985343933105,
      "learning_rate": 2.591231629491423e-07,
      "loss": 0.2283,
      "step": 3470
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.0,
      "learning_rate": 2.5607017043063353e-07,
      "loss": 0.1611,
      "step": 3480
    },
    {
      "epoch": 0.698,
      "grad_norm": 0.5085638165473938,
      "learning_rate": 2.530290667078846e-07,
      "loss": 0.1577,
      "step": 3490
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.6995971202850342,
      "learning_rate": 2.500000000000001e-07,
      "loss": 0.2268,
      "step": 3500
    },
    {
      "epoch": 0.702,
      "grad_norm": 1.0333393812179565,
      "learning_rate": 2.469831179394182e-07,
      "loss": 0.2646,
      "step": 3510
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.5018756985664368,
      "learning_rate": 2.439785675647143e-07,
      "loss": 0.2841,
      "step": 3520
    },
    {
      "epoch": 0.706,
      "grad_norm": 0.4787593483924866,
      "learning_rate": 2.4098649531343494e-07,
      "loss": 0.1832,
      "step": 3530
    },
    {
      "epoch": 0.708,
      "grad_norm": 0.20107752084732056,
      "learning_rate": 2.380070470149605e-07,
      "loss": 0.2042,
      "step": 3540
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.6337828636169434,
      "learning_rate": 2.350403678833976e-07,
      "loss": 0.2196,
      "step": 3550
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.5903733968734741,
      "learning_rate": 2.3208660251050156e-07,
      "loss": 0.1625,
      "step": 3560
    },
    {
      "epoch": 0.714,
      "grad_norm": 0.463150292634964,
      "learning_rate": 2.2914589485863012e-07,
      "loss": 0.1901,
      "step": 3570
    },
    {
      "epoch": 0.716,
      "grad_norm": 0.2560785710811615,
      "learning_rate": 2.262183882537249e-07,
      "loss": 0.2732,
      "step": 3580
    },
    {
      "epoch": 0.718,
      "grad_norm": 0.8427993059158325,
      "learning_rate": 2.23304225378328e-07,
      "loss": 0.2463,
      "step": 3590
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.26840728521347046,
      "learning_rate": 2.2040354826462664e-07,
      "loss": 0.2391,
      "step": 3600
    },
    {
      "epoch": 0.722,
      "grad_norm": 1.2490087747573853,
      "learning_rate": 2.1751649828753106e-07,
      "loss": 0.2908,
      "step": 3610
    },
    {
      "epoch": 0.724,
      "grad_norm": 0.19666559994220734,
      "learning_rate": 2.146432161577842e-07,
      "loss": 0.1933,
      "step": 3620
    },
    {
      "epoch": 0.726,
      "grad_norm": 0.41633477807044983,
      "learning_rate": 2.117838419151034e-07,
      "loss": 0.1907,
      "step": 3630
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.29759231209754944,
      "learning_rate": 2.0893851492135532e-07,
      "loss": 0.2814,
      "step": 3640
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.5257274508476257,
      "learning_rate": 2.0610737385376348e-07,
      "loss": 0.2377,
      "step": 3650
    },
    {
      "epoch": 0.732,
      "grad_norm": 1.2364554405212402,
      "learning_rate": 2.0329055669814933e-07,
      "loss": 0.3568,
      "step": 3660
    },
    {
      "epoch": 0.734,
      "grad_norm": 1.1709978580474854,
      "learning_rate": 2.0048820074220711e-07,
      "loss": 0.2385,
      "step": 3670
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.9572620391845703,
      "learning_rate": 1.9770044256881258e-07,
      "loss": 0.1968,
      "step": 3680
    },
    {
      "epoch": 0.738,
      "grad_norm": 0.2426954209804535,
      "learning_rate": 1.9492741804936618e-07,
      "loss": 0.175,
      "step": 3690
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5391064286231995,
      "learning_rate": 1.9216926233717084e-07,
      "loss": 0.2462,
      "step": 3700
    },
    {
      "epoch": 0.742,
      "grad_norm": 0.5289126038551331,
      "learning_rate": 1.8942610986084484e-07,
      "loss": 0.2174,
      "step": 3710
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.0,
      "learning_rate": 1.8669809431776988e-07,
      "loss": 0.1438,
      "step": 3720
    },
    {
      "epoch": 0.746,
      "grad_norm": 0.551702618598938,
      "learning_rate": 1.8398534866757455e-07,
      "loss": 0.2802,
      "step": 3730
    },
    {
      "epoch": 0.748,
      "grad_norm": 0.3379834294319153,
      "learning_rate": 1.812880051256551e-07,
      "loss": 0.1889,
      "step": 3740
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.0366861820220947,
      "learning_rate": 1.7860619515673032e-07,
      "loss": 0.2649,
      "step": 3750
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.3923557698726654,
      "learning_rate": 1.7594004946843454e-07,
      "loss": 0.2012,
      "step": 3760
    },
    {
      "epoch": 0.754,
      "grad_norm": 1.2161628007888794,
      "learning_rate": 1.7328969800494726e-07,
      "loss": 0.3344,
      "step": 3770
    },
    {
      "epoch": 0.756,
      "grad_norm": 0.23506614565849304,
      "learning_rate": 1.7065526994065972e-07,
      "loss": 0.2097,
      "step": 3780
    },
    {
      "epoch": 0.758,
      "grad_norm": 0.2504694163799286,
      "learning_rate": 1.6803689367387918e-07,
      "loss": 0.1493,
      "step": 3790
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.1120474338531494,
      "learning_rate": 1.6543469682057104e-07,
      "loss": 0.2224,
      "step": 3800
    },
    {
      "epoch": 0.762,
      "grad_norm": 0.2815347909927368,
      "learning_rate": 1.6284880620813846e-07,
      "loss": 0.209,
      "step": 3810
    },
    {
      "epoch": 0.764,
      "grad_norm": 0.31576356291770935,
      "learning_rate": 1.6027934786924185e-07,
      "loss": 0.1237,
      "step": 3820
    },
    {
      "epoch": 0.766,
      "grad_norm": 0.2061535269021988,
      "learning_rate": 1.5772644703565564e-07,
      "loss": 0.1967,
      "step": 3830
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.0,
      "learning_rate": 1.551902281321651e-07,
      "loss": 0.2432,
      "step": 3840
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.45145711302757263,
      "learning_rate": 1.5267081477050131e-07,
      "loss": 0.2098,
      "step": 3850
    },
    {
      "epoch": 0.772,
      "grad_norm": 0.5424890518188477,
      "learning_rate": 1.5016832974331723e-07,
      "loss": 0.3124,
      "step": 3860
    },
    {
      "epoch": 0.774,
      "grad_norm": 0.0,
      "learning_rate": 1.4768289501820263e-07,
      "loss": 0.1453,
      "step": 3870
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.47473931312561035,
      "learning_rate": 1.4521463173173965e-07,
      "loss": 0.1799,
      "step": 3880
    },
    {
      "epoch": 0.778,
      "grad_norm": 0.27455076575279236,
      "learning_rate": 1.4276366018359842e-07,
      "loss": 0.273,
      "step": 3890
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.0,
      "learning_rate": 1.4033009983067452e-07,
      "loss": 0.2012,
      "step": 3900
    },
    {
      "epoch": 0.782,
      "grad_norm": 0.25595951080322266,
      "learning_rate": 1.3791406928126635e-07,
      "loss": 0.1351,
      "step": 3910
    },
    {
      "epoch": 0.784,
      "grad_norm": 1.085986614227295,
      "learning_rate": 1.3551568628929432e-07,
      "loss": 0.3306,
      "step": 3920
    },
    {
      "epoch": 0.786,
      "grad_norm": 0.5502387285232544,
      "learning_rate": 1.3313506774856175e-07,
      "loss": 0.1963,
      "step": 3930
    },
    {
      "epoch": 0.788,
      "grad_norm": 0.9978371262550354,
      "learning_rate": 1.3077232968705805e-07,
      "loss": 0.2349,
      "step": 3940
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.419802188873291,
      "learning_rate": 1.284275872613028e-07,
      "loss": 0.1495,
      "step": 3950
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.5003088116645813,
      "learning_rate": 1.2610095475073413e-07,
      "loss": 0.1413,
      "step": 3960
    },
    {
      "epoch": 0.794,
      "grad_norm": 0.5178236365318298,
      "learning_rate": 1.2379254555213786e-07,
      "loss": 0.2002,
      "step": 3970
    },
    {
      "epoch": 0.796,
      "grad_norm": 1.4785641431808472,
      "learning_rate": 1.2150247217412185e-07,
      "loss": 0.1613,
      "step": 3980
    },
    {
      "epoch": 0.798,
      "grad_norm": 0.4960174858570099,
      "learning_rate": 1.192308462316317e-07,
      "loss": 0.1687,
      "step": 3990
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.4522833824157715,
      "learning_rate": 1.1697777844051104e-07,
      "loss": 0.1916,
      "step": 4000
    },
    {
      "epoch": 0.802,
      "grad_norm": 0.4012996256351471,
      "learning_rate": 1.1474337861210543e-07,
      "loss": 0.1677,
      "step": 4010
    },
    {
      "epoch": 0.804,
      "grad_norm": 1.1145271062850952,
      "learning_rate": 1.1252775564791023e-07,
      "loss": 0.2675,
      "step": 4020
    },
    {
      "epoch": 0.806,
      "grad_norm": 1.063022494316101,
      "learning_rate": 1.1033101753426282e-07,
      "loss": 0.169,
      "step": 4030
    },
    {
      "epoch": 0.808,
      "grad_norm": 0.5367292761802673,
      "learning_rate": 1.0815327133708013e-07,
      "loss": 0.2466,
      "step": 4040
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.46209707856178284,
      "learning_rate": 1.0599462319663904e-07,
      "loss": 0.2542,
      "step": 4050
    },
    {
      "epoch": 0.812,
      "grad_norm": 0.48917677998542786,
      "learning_rate": 1.038551783224047e-07,
      "loss": 0.2497,
      "step": 4060
    },
    {
      "epoch": 0.814,
      "grad_norm": 0.49316373467445374,
      "learning_rate": 1.0173504098790186e-07,
      "loss": 0.1824,
      "step": 4070
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.7999529838562012,
      "learning_rate": 9.963431452563331e-08,
      "loss": 0.2714,
      "step": 4080
    },
    {
      "epoch": 0.818,
      "grad_norm": 1.2927374839782715,
      "learning_rate": 9.755310132204297e-08,
      "loss": 0.2306,
      "step": 4090
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.6657562851905823,
      "learning_rate": 9.549150281252632e-08,
      "loss": 0.2175,
      "step": 4100
    },
    {
      "epoch": 0.822,
      "grad_norm": 0.2347097098827362,
      "learning_rate": 9.344961947648622e-08,
      "loss": 0.2158,
      "step": 4110
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.6625514626502991,
      "learning_rate": 9.142755083243575e-08,
      "loss": 0.299,
      "step": 4120
    },
    {
      "epoch": 0.826,
      "grad_norm": 0.2735314667224884,
      "learning_rate": 8.942539543314798e-08,
      "loss": 0.1775,
      "step": 4130
    },
    {
      "epoch": 0.828,
      "grad_norm": 0.48911941051483154,
      "learning_rate": 8.744325086085247e-08,
      "loss": 0.2273,
      "step": 4140
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.2327689230442047,
      "learning_rate": 8.548121372247919e-08,
      "loss": 0.2096,
      "step": 4150
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.7817517518997192,
      "learning_rate": 8.353937964495028e-08,
      "loss": 0.1858,
      "step": 4160
    },
    {
      "epoch": 0.834,
      "grad_norm": 0.698259711265564,
      "learning_rate": 8.161784327051919e-08,
      "loss": 0.1888,
      "step": 4170
    },
    {
      "epoch": 0.836,
      "grad_norm": 0.5820159316062927,
      "learning_rate": 7.971669825215787e-08,
      "loss": 0.186,
      "step": 4180
    },
    {
      "epoch": 0.838,
      "grad_norm": 0.5032526850700378,
      "learning_rate": 7.783603724899257e-08,
      "loss": 0.1618,
      "step": 4190
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.1663309335708618,
      "learning_rate": 7.597595192178702e-08,
      "loss": 0.2173,
      "step": 4200
    },
    {
      "epoch": 0.842,
      "grad_norm": 0.275285542011261,
      "learning_rate": 7.413653292847616e-08,
      "loss": 0.3055,
      "step": 4210
    },
    {
      "epoch": 0.844,
      "grad_norm": 0.6619118452072144,
      "learning_rate": 7.23178699197467e-08,
      "loss": 0.2306,
      "step": 4220
    },
    {
      "epoch": 0.846,
      "grad_norm": 0.667110800743103,
      "learning_rate": 7.052005153466778e-08,
      "loss": 0.2826,
      "step": 4230
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.374187171459198,
      "learning_rate": 6.874316539637126e-08,
      "loss": 0.2028,
      "step": 4240
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.23237831890583038,
      "learning_rate": 6.698729810778064e-08,
      "loss": 0.2217,
      "step": 4250
    },
    {
      "epoch": 0.852,
      "grad_norm": 0.6198431253433228,
      "learning_rate": 6.52525352473905e-08,
      "loss": 0.1642,
      "step": 4260
    },
    {
      "epoch": 0.854,
      "grad_norm": 0.560222864151001,
      "learning_rate": 6.353896136509524e-08,
      "loss": 0.2124,
      "step": 4270
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.38097235560417175,
      "learning_rate": 6.184665997806831e-08,
      "loss": 0.2093,
      "step": 4280
    },
    {
      "epoch": 0.858,
      "grad_norm": 0.42547518014907837,
      "learning_rate": 6.017571356669182e-08,
      "loss": 0.2598,
      "step": 4290
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.4948174059391022,
      "learning_rate": 5.8526203570536504e-08,
      "loss": 0.1848,
      "step": 4300
    },
    {
      "epoch": 0.862,
      "grad_norm": 0.18645498156547546,
      "learning_rate": 5.689821038439263e-08,
      "loss": 0.1874,
      "step": 4310
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.848637580871582,
      "learning_rate": 5.529181335435124e-08,
      "loss": 0.1541,
      "step": 4320
    },
    {
      "epoch": 0.866,
      "grad_norm": 0.7719595432281494,
      "learning_rate": 5.37070907739372e-08,
      "loss": 0.2264,
      "step": 4330
    },
    {
      "epoch": 0.868,
      "grad_norm": 0.5386465787887573,
      "learning_rate": 5.2144119880293544e-08,
      "loss": 0.1415,
      "step": 4340
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2296571284532547,
      "learning_rate": 5.060297685041659e-08,
      "loss": 0.181,
      "step": 4350
    },
    {
      "epoch": 0.872,
      "grad_norm": 0.2710702121257782,
      "learning_rate": 4.908373679744315e-08,
      "loss": 0.228,
      "step": 4360
    },
    {
      "epoch": 0.874,
      "grad_norm": 0.18128183484077454,
      "learning_rate": 4.758647376699032e-08,
      "loss": 0.1991,
      "step": 4370
    },
    {
      "epoch": 0.876,
      "grad_norm": 1.9789559841156006,
      "learning_rate": 4.611126073354571e-08,
      "loss": 0.2768,
      "step": 4380
    },
    {
      "epoch": 0.878,
      "grad_norm": 0.7291713953018188,
      "learning_rate": 4.465816959691149e-08,
      "loss": 0.2409,
      "step": 4390
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.0,
      "learning_rate": 4.322727117869951e-08,
      "loss": 0.1278,
      "step": 4400
    },
    {
      "epoch": 0.882,
      "grad_norm": 0.7022453546524048,
      "learning_rate": 4.181863521888018e-08,
      "loss": 0.2995,
      "step": 4410
    },
    {
      "epoch": 0.884,
      "grad_norm": 1.2862244844436646,
      "learning_rate": 4.043233037238281e-08,
      "loss": 0.2753,
      "step": 4420
    },
    {
      "epoch": 0.886,
      "grad_norm": 0.14757445454597473,
      "learning_rate": 3.9068424205749794e-08,
      "loss": 0.2235,
      "step": 4430
    },
    {
      "epoch": 0.888,
      "grad_norm": 2.4390058517456055,
      "learning_rate": 3.7726983193843485e-08,
      "loss": 0.2061,
      "step": 4440
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5575977563858032,
      "learning_rate": 3.6408072716606345e-08,
      "loss": 0.2763,
      "step": 4450
    },
    {
      "epoch": 0.892,
      "grad_norm": 0.5324432849884033,
      "learning_rate": 3.5111757055874326e-08,
      "loss": 0.1996,
      "step": 4460
    },
    {
      "epoch": 0.894,
      "grad_norm": 0.3378039598464966,
      "learning_rate": 3.3838099392243915e-08,
      "loss": 0.2702,
      "step": 4470
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.5573419332504272,
      "learning_rate": 3.258716180199278e-08,
      "loss": 0.1961,
      "step": 4480
    },
    {
      "epoch": 0.898,
      "grad_norm": 0.3722284734249115,
      "learning_rate": 3.135900525405427e-08,
      "loss": 0.1993,
      "step": 4490
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.7958338856697083,
      "learning_rate": 3.015368960704584e-08,
      "loss": 0.2263,
      "step": 4500
    },
    {
      "epoch": 0.902,
      "grad_norm": 0.0,
      "learning_rate": 2.8971273606351655e-08,
      "loss": 0.206,
      "step": 4510
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.4656100273132324,
      "learning_rate": 2.78118148812595e-08,
      "loss": 0.1972,
      "step": 4520
    },
    {
      "epoch": 0.906,
      "grad_norm": 0.3063383102416992,
      "learning_rate": 2.667536994215186e-08,
      "loss": 0.1975,
      "step": 4530
    },
    {
      "epoch": 0.908,
      "grad_norm": 1.4156655073165894,
      "learning_rate": 2.5561994177751732e-08,
      "loss": 0.2039,
      "step": 4540
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.23086102306842804,
      "learning_rate": 2.4471741852423233e-08,
      "loss": 0.1558,
      "step": 4550
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.0,
      "learning_rate": 2.3404666103526537e-08,
      "loss": 0.1516,
      "step": 4560
    },
    {
      "epoch": 0.914,
      "grad_norm": 0.9894579648971558,
      "learning_rate": 2.2360818938828187e-08,
      "loss": 0.27,
      "step": 4570
    },
    {
      "epoch": 0.916,
      "grad_norm": 0.47403454780578613,
      "learning_rate": 2.1340251233966377e-08,
      "loss": 0.1885,
      "step": 4580
    },
    {
      "epoch": 0.918,
      "grad_norm": 1.002657175064087,
      "learning_rate": 2.0343012729971243e-08,
      "loss": 0.1384,
      "step": 4590
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.4796471893787384,
      "learning_rate": 1.936915203084055e-08,
      "loss": 0.1593,
      "step": 4600
    },
    {
      "epoch": 0.922,
      "grad_norm": 0.39311739802360535,
      "learning_rate": 1.8418716601170947e-08,
      "loss": 0.1925,
      "step": 4610
    },
    {
      "epoch": 0.924,
      "grad_norm": 0.5048733353614807,
      "learning_rate": 1.7491752763844292e-08,
      "loss": 0.3124,
      "step": 4620
    },
    {
      "epoch": 0.926,
      "grad_norm": 0.5101653933525085,
      "learning_rate": 1.658830569777031e-08,
      "loss": 0.1582,
      "step": 4630
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.3665623664855957,
      "learning_rate": 1.570841943568446e-08,
      "loss": 0.1796,
      "step": 4640
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.20403508841991425,
      "learning_rate": 1.4852136862001763e-08,
      "loss": 0.3377,
      "step": 4650
    },
    {
      "epoch": 0.932,
      "grad_norm": 0.0,
      "learning_rate": 1.4019499710726911e-08,
      "loss": 0.1866,
      "step": 4660
    },
    {
      "epoch": 0.934,
      "grad_norm": 0.0,
      "learning_rate": 1.3210548563419855e-08,
      "loss": 0.1837,
      "step": 4670
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.0,
      "learning_rate": 1.2425322847218367e-08,
      "loss": 0.1362,
      "step": 4680
    },
    {
      "epoch": 0.938,
      "grad_norm": 0.3222580850124359,
      "learning_rate": 1.166386083291604e-08,
      "loss": 0.2977,
      "step": 4690
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.8778265118598938,
      "learning_rate": 1.0926199633097154e-08,
      "loss": 0.2056,
      "step": 4700
    },
    {
      "epoch": 0.942,
      "grad_norm": 0.33410125970840454,
      "learning_rate": 1.0212375200327972e-08,
      "loss": 0.244,
      "step": 4710
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.2924971878528595,
      "learning_rate": 9.522422325404233e-09,
      "loss": 0.1457,
      "step": 4720
    },
    {
      "epoch": 0.946,
      "grad_norm": 0.0,
      "learning_rate": 8.856374635655695e-09,
      "loss": 0.2427,
      "step": 4730
    },
    {
      "epoch": 0.948,
      "grad_norm": 0.0,
      "learning_rate": 8.214264593307096e-09,
      "loss": 0.2146,
      "step": 4740
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.33636531233787537,
      "learning_rate": 7.59612349389599e-09,
      "loss": 0.1394,
      "step": 4750
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.0,
      "learning_rate": 7.0019814647475636e-09,
      "loss": 0.0868,
      "step": 4760
    },
    {
      "epoch": 0.954,
      "grad_norm": 1.4758881330490112,
      "learning_rate": 6.431867463506046e-09,
      "loss": 0.2783,
      "step": 4770
    },
    {
      "epoch": 0.956,
      "grad_norm": 0.6460458040237427,
      "learning_rate": 5.8858092767236076e-09,
      "loss": 0.2321,
      "step": 4780
    },
    {
      "epoch": 0.958,
      "grad_norm": 0.5238158702850342,
      "learning_rate": 5.3638335185058335e-09,
      "loss": 0.2235,
      "step": 4790
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.5409115552902222,
      "learning_rate": 4.865965629214819e-09,
      "loss": 0.2579,
      "step": 4800
    },
    {
      "epoch": 0.962,
      "grad_norm": 0.5042092800140381,
      "learning_rate": 4.3922298742291585e-09,
      "loss": 0.1759,
      "step": 4810
    },
    {
      "epoch": 0.964,
      "grad_norm": 0.4713117182254791,
      "learning_rate": 3.9426493427611175e-09,
      "loss": 0.2138,
      "step": 4820
    },
    {
      "epoch": 0.966,
      "grad_norm": 0.0,
      "learning_rate": 3.5172459467315286e-09,
      "loss": 0.1715,
      "step": 4830
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.8907610177993774,
      "learning_rate": 3.116040419701815e-09,
      "loss": 0.1743,
      "step": 4840
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.20034664869308472,
      "learning_rate": 2.739052315863355e-09,
      "loss": 0.2379,
      "step": 4850
    },
    {
      "epoch": 0.972,
      "grad_norm": 0.3425753712654114,
      "learning_rate": 2.3863000090844076e-09,
      "loss": 0.2835,
      "step": 4860
    },
    {
      "epoch": 0.974,
      "grad_norm": 0.7830394506454468,
      "learning_rate": 2.057800692014833e-09,
      "loss": 0.1899,
      "step": 4870
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.5811526775360107,
      "learning_rate": 1.7535703752478147e-09,
      "loss": 0.3774,
      "step": 4880
    },
    {
      "epoch": 0.978,
      "grad_norm": 0.09480271488428116,
      "learning_rate": 1.4736238865398765e-09,
      "loss": 0.189,
      "step": 4890
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.22772358357906342,
      "learning_rate": 1.217974870087901e-09,
      "loss": 0.196,
      "step": 4900
    },
    {
      "epoch": 0.982,
      "grad_norm": 0.4435659348964691,
      "learning_rate": 9.866357858642205e-10,
      "loss": 0.3854,
      "step": 4910
    },
    {
      "epoch": 0.984,
      "grad_norm": 0.30677875876426697,
      "learning_rate": 7.79617909009489e-10,
      "loss": 0.3105,
      "step": 4920
    },
    {
      "epoch": 0.986,
      "grad_norm": 0.5438311696052551,
      "learning_rate": 5.969313292830125e-10,
      "loss": 0.2419,
      "step": 4930
    },
    {
      "epoch": 0.988,
      "grad_norm": 0.39847153425216675,
      "learning_rate": 4.3858495057080836e-10,
      "loss": 0.2969,
      "step": 4940
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.0,
      "learning_rate": 3.0458649045211894e-10,
      "loss": 0.1753,
      "step": 4950
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.3326397240161896,
      "learning_rate": 1.9494247982282387e-10,
      "loss": 0.1342,
      "step": 4960
    },
    {
      "epoch": 0.994,
      "grad_norm": 1.0523000955581665,
      "learning_rate": 1.0965826257725019e-10,
      "loss": 0.2628,
      "step": 4970
    },
    {
      "epoch": 0.996,
      "grad_norm": 0.48323336243629456,
      "learning_rate": 4.873799534788059e-11,
      "loss": 0.1767,
      "step": 4980
    },
    {
      "epoch": 0.998,
      "grad_norm": 0.9073156118392944,
      "learning_rate": 1.2184647302626582e-11,
      "loss": 0.1813,
      "step": 4990
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6704990267753601,
      "learning_rate": 0.0,
      "loss": 0.2975,
      "step": 5000
    },
    {
      "epoch": 1.0,
      "step": 5000,
      "total_flos": 2.3466648993521664e+16,
      "train_loss": 0.3158116644740105,
      "train_runtime": 807.301,
      "train_samples_per_second": 6.193,
      "train_steps_per_second": 6.193
    }
  ],
  "logging_steps": 10,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 4000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.3466648993521664e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}