{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.4692107545533393,
  "eval_steps": 500,
  "global_step": 8000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004336513443191674,
      "grad_norm": 36.625,
      "learning_rate": 3.902862098872507e-08,
      "loss": 8.4915,
      "step": 10
    },
    {
      "epoch": 0.008673026886383347,
      "grad_norm": 38.0,
      "learning_rate": 8.23937554206418e-08,
      "loss": 7.9652,
      "step": 20
    },
    {
      "epoch": 0.013009540329575022,
      "grad_norm": 27.125,
      "learning_rate": 1.2575888985255857e-07,
      "loss": 7.2548,
      "step": 30
    },
    {
      "epoch": 0.017346053772766695,
      "grad_norm": 24.59375,
      "learning_rate": 1.6912402428447528e-07,
      "loss": 6.6779,
      "step": 40
    },
    {
      "epoch": 0.02168256721595837,
      "grad_norm": 27.5625,
      "learning_rate": 2.1248915871639205e-07,
      "loss": 6.6767,
      "step": 50
    },
    {
      "epoch": 0.026019080659150044,
      "grad_norm": 17.4375,
      "learning_rate": 2.558542931483088e-07,
      "loss": 4.717,
      "step": 60
    },
    {
      "epoch": 0.030355594102341718,
      "grad_norm": 23.109375,
      "learning_rate": 2.9921942758022555e-07,
      "loss": 4.0226,
      "step": 70
    },
    {
      "epoch": 0.03469210754553339,
      "grad_norm": 16.171875,
      "learning_rate": 3.425845620121423e-07,
      "loss": 2.4996,
      "step": 80
    },
    {
      "epoch": 0.03902862098872507,
      "grad_norm": 12.1015625,
      "learning_rate": 3.85949696444059e-07,
      "loss": 1.8746,
      "step": 90
    },
    {
      "epoch": 0.04336513443191674,
      "grad_norm": 16.609375,
      "learning_rate": 4.2931483087597577e-07,
      "loss": 1.5231,
      "step": 100
    },
    {
      "epoch": 0.047701647875108416,
      "grad_norm": 15.890625,
      "learning_rate": 4.726799653078925e-07,
      "loss": 1.4022,
      "step": 110
    },
    {
      "epoch": 0.05203816131830009,
      "grad_norm": 21.21875,
      "learning_rate": 5.160450997398092e-07,
      "loss": 1.9312,
      "step": 120
    },
    {
      "epoch": 0.05637467476149176,
      "grad_norm": 11.140625,
      "learning_rate": 5.59410234171726e-07,
      "loss": 1.2294,
      "step": 130
    },
    {
      "epoch": 0.060711188204683436,
      "grad_norm": 9.5703125,
      "learning_rate": 6.027753686036427e-07,
      "loss": 1.5361,
      "step": 140
    },
    {
      "epoch": 0.06504770164787511,
      "grad_norm": 19.765625,
      "learning_rate": 6.461405030355594e-07,
      "loss": 1.5074,
      "step": 150
    },
    {
      "epoch": 0.06938421509106678,
      "grad_norm": 21.40625,
      "learning_rate": 6.895056374674763e-07,
      "loss": 1.2882,
      "step": 160
    },
    {
      "epoch": 0.07372072853425846,
      "grad_norm": 14.25,
      "learning_rate": 7.32870771899393e-07,
      "loss": 1.451,
      "step": 170
    },
    {
      "epoch": 0.07805724197745013,
      "grad_norm": 16.40625,
      "learning_rate": 7.762359063313096e-07,
      "loss": 1.6452,
      "step": 180
    },
    {
      "epoch": 0.0823937554206418,
      "grad_norm": 15.1484375,
      "learning_rate": 8.196010407632264e-07,
      "loss": 1.3615,
      "step": 190
    },
    {
      "epoch": 0.08673026886383348,
      "grad_norm": 15.4765625,
      "learning_rate": 8.629661751951432e-07,
      "loss": 1.3018,
      "step": 200
    },
    {
      "epoch": 0.09106678230702515,
      "grad_norm": 7.6328125,
      "learning_rate": 9.0633130962706e-07,
      "loss": 1.8447,
      "step": 210
    },
    {
      "epoch": 0.09540329575021683,
      "grad_norm": 15.9609375,
      "learning_rate": 9.496964440589766e-07,
      "loss": 1.5326,
      "step": 220
    },
    {
      "epoch": 0.0997398091934085,
      "grad_norm": 18.4375,
      "learning_rate": 9.930615784908933e-07,
      "loss": 1.6809,
      "step": 230
    },
    {
      "epoch": 0.10407632263660017,
      "grad_norm": 20.75,
      "learning_rate": 1.0364267129228103e-06,
      "loss": 1.6056,
      "step": 240
    },
    {
      "epoch": 0.10841283607979185,
      "grad_norm": 9.6484375,
      "learning_rate": 1.0797918473547268e-06,
      "loss": 1.8439,
      "step": 250
    },
    {
      "epoch": 0.11274934952298352,
      "grad_norm": 18.8125,
      "learning_rate": 1.1231569817866435e-06,
      "loss": 1.6012,
      "step": 260
    },
    {
      "epoch": 0.1170858629661752,
      "grad_norm": 22.34375,
      "learning_rate": 1.1665221162185605e-06,
      "loss": 1.5844,
      "step": 270
    },
    {
      "epoch": 0.12142237640936687,
      "grad_norm": 9.4453125,
      "learning_rate": 1.209887250650477e-06,
      "loss": 1.7287,
      "step": 280
    },
    {
      "epoch": 0.12575888985255854,
      "grad_norm": 14.390625,
      "learning_rate": 1.2532523850823938e-06,
      "loss": 2.0146,
      "step": 290
    },
    {
      "epoch": 0.13009540329575023,
      "grad_norm": 16.296875,
      "learning_rate": 1.2966175195143105e-06,
      "loss": 1.6197,
      "step": 300
    },
    {
      "epoch": 0.1344319167389419,
      "grad_norm": 11.8671875,
      "learning_rate": 1.3399826539462274e-06,
      "loss": 1.4233,
      "step": 310
    },
    {
      "epoch": 0.13876843018213356,
      "grad_norm": 21.859375,
      "learning_rate": 1.383347788378144e-06,
      "loss": 1.9114,
      "step": 320
    },
    {
      "epoch": 0.14310494362532525,
      "grad_norm": 24.515625,
      "learning_rate": 1.426712922810061e-06,
      "loss": 1.3864,
      "step": 330
    },
    {
      "epoch": 0.1474414570685169,
      "grad_norm": 15.265625,
      "learning_rate": 1.4700780572419777e-06,
      "loss": 1.6892,
      "step": 340
    },
    {
      "epoch": 0.15177797051170858,
      "grad_norm": 28.078125,
      "learning_rate": 1.5134431916738942e-06,
      "loss": 1.9963,
      "step": 350
    },
    {
      "epoch": 0.15611448395490027,
      "grad_norm": 16.890625,
      "learning_rate": 1.5568083261058111e-06,
      "loss": 1.7315,
      "step": 360
    },
    {
      "epoch": 0.16045099739809193,
      "grad_norm": 17.015625,
      "learning_rate": 1.6001734605377279e-06,
      "loss": 1.5838,
      "step": 370
    },
    {
      "epoch": 0.1647875108412836,
      "grad_norm": 16.734375,
      "learning_rate": 1.6435385949696444e-06,
      "loss": 1.8812,
      "step": 380
    },
    {
      "epoch": 0.1691240242844753,
      "grad_norm": 19.484375,
      "learning_rate": 1.6869037294015614e-06,
      "loss": 2.0695,
      "step": 390
    },
    {
      "epoch": 0.17346053772766695,
      "grad_norm": 12.71875,
      "learning_rate": 1.730268863833478e-06,
      "loss": 1.7535,
      "step": 400
    },
    {
      "epoch": 0.17779705117085862,
      "grad_norm": 13.34375,
      "learning_rate": 1.7736339982653946e-06,
      "loss": 1.5148,
      "step": 410
    },
    {
      "epoch": 0.1821335646140503,
      "grad_norm": 18.421875,
      "learning_rate": 1.8169991326973116e-06,
      "loss": 2.4693,
      "step": 420
    },
    {
      "epoch": 0.18647007805724197,
      "grad_norm": 18.8125,
      "learning_rate": 1.8603642671292283e-06,
      "loss": 2.1573,
      "step": 430
    },
    {
      "epoch": 0.19080659150043366,
      "grad_norm": 13.9296875,
      "learning_rate": 1.9037294015611448e-06,
      "loss": 2.0659,
      "step": 440
    },
    {
      "epoch": 0.19514310494362533,
      "grad_norm": 28.28125,
      "learning_rate": 1.947094535993062e-06,
      "loss": 2.4561,
      "step": 450
    },
    {
      "epoch": 0.199479618386817,
      "grad_norm": 35.75,
      "learning_rate": 1.9904596704249783e-06,
      "loss": 2.5905,
      "step": 460
    },
    {
      "epoch": 0.20381613183000868,
      "grad_norm": 14.296875,
      "learning_rate": 2.033824804856895e-06,
      "loss": 1.7498,
      "step": 470
    },
    {
      "epoch": 0.20815264527320035,
      "grad_norm": 32.53125,
      "learning_rate": 2.0771899392888122e-06,
      "loss": 2.2018,
      "step": 480
    },
    {
      "epoch": 0.212489158716392,
      "grad_norm": 19.53125,
      "learning_rate": 2.1205550737207285e-06,
      "loss": 2.8259,
      "step": 490
    },
    {
      "epoch": 0.2168256721595837,
      "grad_norm": 11.0234375,
      "learning_rate": 2.1639202081526457e-06,
      "loss": 2.4305,
      "step": 500
    },
    {
      "epoch": 0.2168256721595837,
      "eval_loss": 0.5951342582702637,
      "eval_runtime": 355.4616,
      "eval_samples_per_second": 3.23,
      "eval_steps_per_second": 3.23,
      "step": 500
    },
    {
      "epoch": 0.22116218560277537,
      "grad_norm": 15.8828125,
      "learning_rate": 1.1036426712922812e-06,
      "loss": 2.761,
      "step": 510
    },
    {
      "epoch": 0.22549869904596703,
      "grad_norm": 13.1640625,
      "learning_rate": 1.1253252385082394e-06,
      "loss": 2.3113,
      "step": 520
    },
    {
      "epoch": 0.22983521248915872,
      "grad_norm": 11.2109375,
      "learning_rate": 1.147007805724198e-06,
      "loss": 1.5673,
      "step": 530
    },
    {
      "epoch": 0.2341717259323504,
      "grad_norm": 11.9609375,
      "learning_rate": 1.1686903729401563e-06,
      "loss": 1.4327,
      "step": 540
    },
    {
      "epoch": 0.23850823937554205,
      "grad_norm": 13.1953125,
      "learning_rate": 1.1903729401561145e-06,
      "loss": 1.844,
      "step": 550
    },
    {
      "epoch": 0.24284475281873374,
      "grad_norm": 4.2265625,
      "learning_rate": 1.212055507372073e-06,
      "loss": 1.5323,
      "step": 560
    },
    {
      "epoch": 0.2471812662619254,
      "grad_norm": 12.3828125,
      "learning_rate": 1.2337380745880314e-06,
      "loss": 1.2953,
      "step": 570
    },
    {
      "epoch": 0.2515177797051171,
      "grad_norm": 14.2421875,
      "learning_rate": 1.2554206418039896e-06,
      "loss": 2.0151,
      "step": 580
    },
    {
      "epoch": 0.25585429314830876,
      "grad_norm": 22.03125,
      "learning_rate": 1.2771032090199482e-06,
      "loss": 1.6165,
      "step": 590
    },
    {
      "epoch": 0.26019080659150046,
      "grad_norm": 10.828125,
      "learning_rate": 1.2987857762359065e-06,
      "loss": 1.3585,
      "step": 600
    },
    {
      "epoch": 0.2645273200346921,
      "grad_norm": 16.171875,
      "learning_rate": 1.3204683434518647e-06,
      "loss": 1.6434,
      "step": 610
    },
    {
      "epoch": 0.2688638334778838,
      "grad_norm": 18.171875,
      "learning_rate": 1.3421509106678233e-06,
      "loss": 1.6819,
      "step": 620
    },
    {
      "epoch": 0.2732003469210755,
      "grad_norm": 6.69921875,
      "learning_rate": 1.3638334778837816e-06,
      "loss": 1.1795,
      "step": 630
    },
    {
      "epoch": 0.2775368603642671,
      "grad_norm": 7.046875,
      "learning_rate": 1.3855160450997398e-06,
      "loss": 1.6481,
      "step": 640
    },
    {
      "epoch": 0.2818733738074588,
      "grad_norm": 8.15625,
      "learning_rate": 1.4071986123156984e-06,
      "loss": 1.4091,
      "step": 650
    },
    {
      "epoch": 0.2862098872506505,
      "grad_norm": 27.53125,
      "learning_rate": 1.4288811795316568e-06,
      "loss": 1.7348,
      "step": 660
    },
    {
      "epoch": 0.29054640069384213,
      "grad_norm": 9.28125,
      "learning_rate": 1.450563746747615e-06,
      "loss": 0.9625,
      "step": 670
    },
    {
      "epoch": 0.2948829141370338,
      "grad_norm": 10.9453125,
      "learning_rate": 1.4722463139635735e-06,
      "loss": 1.0473,
      "step": 680
    },
    {
      "epoch": 0.2992194275802255,
      "grad_norm": 10.8828125,
      "learning_rate": 1.4939288811795319e-06,
      "loss": 1.6767,
      "step": 690
    },
    {
      "epoch": 0.30355594102341715,
      "grad_norm": 10.6484375,
      "learning_rate": 1.51561144839549e-06,
      "loss": 1.7264,
      "step": 700
    },
    {
      "epoch": 0.30789245446660884,
      "grad_norm": 14.8203125,
      "learning_rate": 1.5372940156114486e-06,
      "loss": 1.3095,
      "step": 710
    },
    {
      "epoch": 0.31222896790980054,
      "grad_norm": 11.9453125,
      "learning_rate": 1.558976582827407e-06,
      "loss": 1.8803,
      "step": 720
    },
    {
      "epoch": 0.3165654813529922,
      "grad_norm": 11.1796875,
      "learning_rate": 1.5806591500433651e-06,
      "loss": 1.5766,
      "step": 730
    },
    {
      "epoch": 0.32090199479618386,
      "grad_norm": 8.0625,
      "learning_rate": 1.6023417172593237e-06,
      "loss": 1.6594,
      "step": 740
    },
    {
      "epoch": 0.32523850823937556,
      "grad_norm": 5.8203125,
      "learning_rate": 1.624024284475282e-06,
      "loss": 1.5188,
      "step": 750
    },
    {
      "epoch": 0.3295750216825672,
      "grad_norm": 17.359375,
      "learning_rate": 1.6457068516912402e-06,
      "loss": 1.6649,
      "step": 760
    },
    {
      "epoch": 0.3339115351257589,
      "grad_norm": 13.953125,
      "learning_rate": 1.6673894189071988e-06,
      "loss": 1.5696,
      "step": 770
    },
    {
      "epoch": 0.3382480485689506,
      "grad_norm": 14.6796875,
      "learning_rate": 1.6890719861231572e-06,
      "loss": 1.5922,
      "step": 780
    },
    {
      "epoch": 0.3425845620121422,
      "grad_norm": 9.0078125,
      "learning_rate": 1.7107545533391153e-06,
      "loss": 1.5668,
      "step": 790
    },
    {
      "epoch": 0.3469210754553339,
      "grad_norm": 8.0546875,
      "learning_rate": 1.732437120555074e-06,
      "loss": 1.5727,
      "step": 800
    },
    {
      "epoch": 0.3512575888985256,
      "grad_norm": 17.125,
      "learning_rate": 1.7541196877710323e-06,
      "loss": 1.7644,
      "step": 810
    },
    {
      "epoch": 0.35559410234171723,
      "grad_norm": 12.515625,
      "learning_rate": 1.7758022549869905e-06,
      "loss": 1.2897,
      "step": 820
    },
    {
      "epoch": 0.3599306157849089,
      "grad_norm": 9.859375,
      "learning_rate": 1.797484822202949e-06,
      "loss": 1.7956,
      "step": 830
    },
    {
      "epoch": 0.3642671292281006,
      "grad_norm": 8.8984375,
      "learning_rate": 1.8191673894189074e-06,
      "loss": 2.0324,
      "step": 840
    },
    {
      "epoch": 0.3686036426712923,
      "grad_norm": 12.9765625,
      "learning_rate": 1.8408499566348656e-06,
      "loss": 1.4643,
      "step": 850
    },
    {
      "epoch": 0.37294015611448394,
      "grad_norm": 16.890625,
      "learning_rate": 1.8625325238508241e-06,
      "loss": 1.7663,
      "step": 860
    },
    {
      "epoch": 0.37727666955767564,
      "grad_norm": 7.5,
      "learning_rate": 1.8842150910667825e-06,
      "loss": 1.3754,
      "step": 870
    },
    {
      "epoch": 0.38161318300086733,
      "grad_norm": 10.1328125,
      "learning_rate": 1.9058976582827407e-06,
      "loss": 1.9336,
      "step": 880
    },
    {
      "epoch": 0.38594969644405897,
      "grad_norm": 10.921875,
      "learning_rate": 1.9275802254986993e-06,
      "loss": 1.9662,
      "step": 890
    },
    {
      "epoch": 0.39028620988725066,
      "grad_norm": 6.296875,
      "learning_rate": 1.9492627927146576e-06,
      "loss": 1.6599,
      "step": 900
    },
    {
      "epoch": 0.39462272333044235,
      "grad_norm": 24.28125,
      "learning_rate": 1.970945359930616e-06,
      "loss": 1.8978,
      "step": 910
    },
    {
      "epoch": 0.398959236773634,
      "grad_norm": 11.0234375,
      "learning_rate": 1.9926279271465744e-06,
      "loss": 2.2158,
      "step": 920
    },
    {
      "epoch": 0.4032957502168257,
      "grad_norm": 16.109375,
      "learning_rate": 2.0143104943625327e-06,
      "loss": 2.3377,
      "step": 930
    },
    {
      "epoch": 0.40763226366001737,
      "grad_norm": 4.58984375,
      "learning_rate": 2.035993061578491e-06,
      "loss": 1.5991,
      "step": 940
    },
    {
      "epoch": 0.411968777103209,
      "grad_norm": 13.7265625,
      "learning_rate": 2.0576756287944495e-06,
      "loss": 1.7245,
      "step": 950
    },
    {
      "epoch": 0.4163052905464007,
      "grad_norm": 16.3125,
      "learning_rate": 2.079358196010408e-06,
      "loss": 1.9573,
      "step": 960
    },
    {
      "epoch": 0.4206418039895924,
      "grad_norm": 10.5234375,
      "learning_rate": 2.101040763226366e-06,
      "loss": 1.5117,
      "step": 970
    },
    {
      "epoch": 0.424978317432784,
      "grad_norm": 17.765625,
      "learning_rate": 2.1227233304423246e-06,
      "loss": 2.0703,
      "step": 980
    },
    {
      "epoch": 0.4293148308759757,
      "grad_norm": 10.1640625,
      "learning_rate": 2.144405897658283e-06,
      "loss": 2.1135,
      "step": 990
    },
    {
      "epoch": 0.4336513443191674,
      "grad_norm": 15.421875,
      "learning_rate": 2.1660884648742413e-06,
      "loss": 1.6969,
      "step": 1000
    },
    {
      "epoch": 0.4336513443191674,
      "eval_loss": 0.46456632018089294,
      "eval_runtime": 353.5184,
      "eval_samples_per_second": 3.247,
      "eval_steps_per_second": 3.247,
      "step": 1000
    },
    {
      "epoch": 0.43798785776235905,
      "grad_norm": 15.5234375,
      "learning_rate": 2.1877710320901997e-06,
      "loss": 1.5506,
      "step": 1010
    },
    {
      "epoch": 0.44232437120555074,
      "grad_norm": 6.51953125,
      "learning_rate": 2.209453599306158e-06,
      "loss": 1.6389,
      "step": 1020
    },
    {
      "epoch": 0.44666088464874243,
      "grad_norm": 10.7109375,
      "learning_rate": 2.2311361665221164e-06,
      "loss": 2.0296,
      "step": 1030
    },
    {
      "epoch": 0.45099739809193407,
      "grad_norm": 15.71875,
      "learning_rate": 2.252818733738075e-06,
      "loss": 2.5187,
      "step": 1040
    },
    {
      "epoch": 0.45533391153512576,
      "grad_norm": 9.1171875,
      "learning_rate": 2.274501300954033e-06,
      "loss": 1.9372,
      "step": 1050
    },
    {
      "epoch": 0.45967042497831745,
      "grad_norm": 10.5546875,
      "learning_rate": 2.2961838681699915e-06,
      "loss": 1.7882,
      "step": 1060
    },
    {
      "epoch": 0.4640069384215091,
      "grad_norm": 15.5,
      "learning_rate": 2.31786643538595e-06,
      "loss": 1.8646,
      "step": 1070
    },
    {
      "epoch": 0.4683434518647008,
      "grad_norm": 12.2578125,
      "learning_rate": 2.3395490026019083e-06,
      "loss": 1.9062,
      "step": 1080
    },
    {
      "epoch": 0.47267996530789247,
      "grad_norm": 16.109375,
      "learning_rate": 2.3612315698178666e-06,
      "loss": 2.2143,
      "step": 1090
    },
    {
      "epoch": 0.4770164787510841,
      "grad_norm": 19.640625,
      "learning_rate": 2.382914137033825e-06,
      "loss": 2.6948,
      "step": 1100
    },
    {
      "epoch": 0.4813529921942758,
      "grad_norm": 18.671875,
      "learning_rate": 2.4045967042497834e-06,
      "loss": 2.4728,
      "step": 1110
    },
    {
      "epoch": 0.4856895056374675,
      "grad_norm": 14.9453125,
      "learning_rate": 2.4262792714657417e-06,
      "loss": 2.085,
      "step": 1120
    },
    {
      "epoch": 0.4900260190806591,
      "grad_norm": 11.3046875,
      "learning_rate": 2.4479618386817e-06,
      "loss": 1.9414,
      "step": 1130
    },
    {
      "epoch": 0.4943625325238508,
      "grad_norm": 17.734375,
      "learning_rate": 2.4696444058976585e-06,
      "loss": 2.5984,
      "step": 1140
    },
    {
      "epoch": 0.4986990459670425,
      "grad_norm": 17.421875,
      "learning_rate": 2.491326973113617e-06,
      "loss": 2.1833,
      "step": 1150
    },
    {
      "epoch": 0.5030355594102341,
      "grad_norm": 23.59375,
      "learning_rate": 2.5130095403295752e-06,
      "loss": 1.8192,
      "step": 1160
    },
    {
      "epoch": 0.5073720728534259,
      "grad_norm": 13.25,
      "learning_rate": 2.534692107545534e-06,
      "loss": 1.9603,
      "step": 1170
    },
    {
      "epoch": 0.5117085862966175,
      "grad_norm": 17.890625,
      "learning_rate": 2.556374674761492e-06,
      "loss": 1.8326,
      "step": 1180
    },
    {
      "epoch": 0.5160450997398092,
      "grad_norm": 9.375,
      "learning_rate": 2.5780572419774503e-06,
      "loss": 1.8499,
      "step": 1190
    },
    {
      "epoch": 0.5203816131830009,
      "grad_norm": 10.8359375,
      "learning_rate": 2.599739809193409e-06,
      "loss": 1.7149,
      "step": 1200
    },
    {
      "epoch": 0.5247181266261925,
      "grad_norm": 15.015625,
      "learning_rate": 2.621422376409367e-06,
      "loss": 1.9407,
      "step": 1210
    },
    {
      "epoch": 0.5290546400693842,
      "grad_norm": 10.9375,
      "learning_rate": 2.6431049436253254e-06,
      "loss": 1.9781,
      "step": 1220
    },
    {
      "epoch": 0.5333911535125759,
      "grad_norm": 16.140625,
      "learning_rate": 2.6647875108412842e-06,
      "loss": 2.3761,
      "step": 1230
    },
    {
      "epoch": 0.5377276669557676,
      "grad_norm": 12.1171875,
      "learning_rate": 2.686470078057242e-06,
      "loss": 2.2025,
      "step": 1240
    },
    {
      "epoch": 0.5420641803989592,
      "grad_norm": 9.7265625,
      "learning_rate": 2.7081526452732006e-06,
      "loss": 1.7392,
      "step": 1250
    },
    {
      "epoch": 0.546400693842151,
      "grad_norm": 14.4765625,
      "learning_rate": 2.7298352124891593e-06,
      "loss": 2.2126,
      "step": 1260
    },
    {
      "epoch": 0.5507372072853426,
      "grad_norm": 12.453125,
      "learning_rate": 2.7515177797051173e-06,
      "loss": 2.1276,
      "step": 1270
    },
    {
      "epoch": 0.5550737207285342,
      "grad_norm": 13.3828125,
      "learning_rate": 2.7732003469210757e-06,
      "loss": 2.4107,
      "step": 1280
    },
    {
      "epoch": 0.559410234171726,
      "grad_norm": 8.5703125,
      "learning_rate": 2.7948829141370345e-06,
      "loss": 1.8659,
      "step": 1290
    },
    {
      "epoch": 0.5637467476149176,
      "grad_norm": 12.5703125,
      "learning_rate": 2.8165654813529924e-06,
      "loss": 2.0755,
      "step": 1300
    },
    {
      "epoch": 0.5680832610581092,
      "grad_norm": 12.875,
      "learning_rate": 2.8382480485689508e-06,
      "loss": 2.6062,
      "step": 1310
    },
    {
      "epoch": 0.572419774501301,
      "grad_norm": 8.8359375,
      "learning_rate": 2.8599306157849096e-06,
      "loss": 2.6835,
      "step": 1320
    },
    {
      "epoch": 0.5767562879444926,
      "grad_norm": 16.9375,
      "learning_rate": 2.8816131830008675e-06,
      "loss": 2.6335,
      "step": 1330
    },
    {
      "epoch": 0.5810928013876843,
      "grad_norm": 14.390625,
      "learning_rate": 2.903295750216826e-06,
      "loss": 2.6873,
      "step": 1340
    },
    {
      "epoch": 0.585429314830876,
      "grad_norm": 20.59375,
      "learning_rate": 2.9249783174327847e-06,
      "loss": 2.377,
      "step": 1350
    },
    {
      "epoch": 0.5897658282740676,
      "grad_norm": 8.390625,
      "learning_rate": 2.9466608846487426e-06,
      "loss": 2.1124,
      "step": 1360
    },
    {
      "epoch": 0.5941023417172593,
      "grad_norm": 15.6484375,
      "learning_rate": 2.968343451864701e-06,
      "loss": 1.9201,
      "step": 1370
    },
    {
      "epoch": 0.598438855160451,
      "grad_norm": 21.65625,
      "learning_rate": 2.9900260190806598e-06,
      "loss": 2.3976,
      "step": 1380
    },
    {
      "epoch": 0.6027753686036427,
      "grad_norm": 18.6875,
      "learning_rate": 3.0117085862966177e-06,
      "loss": 2.4113,
      "step": 1390
    },
    {
      "epoch": 0.6071118820468343,
      "grad_norm": 19.171875,
      "learning_rate": 3.033391153512576e-06,
      "loss": 2.3789,
      "step": 1400
    },
    {
      "epoch": 0.611448395490026,
      "grad_norm": 19.171875,
      "learning_rate": 3.055073720728535e-06,
      "loss": 2.6867,
      "step": 1410
    },
    {
      "epoch": 0.6157849089332177,
      "grad_norm": 9.875,
      "learning_rate": 3.076756287944493e-06,
      "loss": 2.4534,
      "step": 1420
    },
    {
      "epoch": 0.6201214223764093,
      "grad_norm": 18.859375,
      "learning_rate": 3.098438855160451e-06,
      "loss": 2.0599,
      "step": 1430
    },
    {
      "epoch": 0.6244579358196011,
      "grad_norm": 18.703125,
      "learning_rate": 3.12012142237641e-06,
      "loss": 2.407,
      "step": 1440
    },
    {
      "epoch": 0.6287944492627927,
      "grad_norm": 24.125,
      "learning_rate": 3.141803989592368e-06,
      "loss": 1.9136,
      "step": 1450
    },
    {
      "epoch": 0.6331309627059843,
      "grad_norm": 14.2734375,
      "learning_rate": 3.1634865568083263e-06,
      "loss": 2.7491,
      "step": 1460
    },
    {
      "epoch": 0.6374674761491761,
      "grad_norm": 10.1640625,
      "learning_rate": 3.185169124024285e-06,
      "loss": 2.3464,
      "step": 1470
    },
    {
      "epoch": 0.6418039895923677,
      "grad_norm": 15.5703125,
      "learning_rate": 3.206851691240243e-06,
      "loss": 2.1874,
      "step": 1480
    },
    {
      "epoch": 0.6461405030355594,
      "grad_norm": 11.8671875,
      "learning_rate": 3.2285342584562014e-06,
      "loss": 2.9108,
      "step": 1490
    },
    {
      "epoch": 0.6504770164787511,
      "grad_norm": 44.125,
      "learning_rate": 3.25021682567216e-06,
      "loss": 2.0486,
      "step": 1500
    },
    {
      "epoch": 0.6504770164787511,
      "eval_loss": 0.7500036358833313,
      "eval_runtime": 354.5046,
      "eval_samples_per_second": 3.238,
      "eval_steps_per_second": 3.238,
      "step": 1500
    },
    {
      "epoch": 0.6548135299219427,
      "grad_norm": 27.15625,
      "learning_rate": 3.271899392888118e-06,
      "loss": 2.869,
      "step": 1510
    },
    {
      "epoch": 0.6591500433651344,
      "grad_norm": 13.171875,
      "learning_rate": 3.2935819601040765e-06,
      "loss": 2.3973,
      "step": 1520
    },
    {
      "epoch": 0.6634865568083261,
      "grad_norm": 10.171875,
      "learning_rate": 3.3152645273200353e-06,
      "loss": 2.8716,
      "step": 1530
    },
    {
      "epoch": 0.6678230702515178,
      "grad_norm": 11.2890625,
      "learning_rate": 3.3369470945359933e-06,
      "loss": 2.8762,
      "step": 1540
    },
    {
      "epoch": 0.6721595836947094,
      "grad_norm": 16.828125,
      "learning_rate": 3.3586296617519516e-06,
      "loss": 2.6887,
      "step": 1550
    },
    {
      "epoch": 0.6764960971379012,
      "grad_norm": 7.14453125,
      "learning_rate": 3.3803122289679104e-06,
      "loss": 2.1227,
      "step": 1560
    },
    {
      "epoch": 0.6808326105810928,
      "grad_norm": 14.046875,
      "learning_rate": 3.4019947961838684e-06,
      "loss": 2.6509,
      "step": 1570
    },
    {
      "epoch": 0.6851691240242844,
      "grad_norm": 16.71875,
      "learning_rate": 3.4236773633998267e-06,
      "loss": 3.5994,
      "step": 1580
    },
    {
      "epoch": 0.6895056374674762,
      "grad_norm": 16.015625,
      "learning_rate": 3.4453599306157855e-06,
      "loss": 2.4302,
      "step": 1590
    },
    {
      "epoch": 0.6938421509106678,
      "grad_norm": 11.171875,
      "learning_rate": 3.4670424978317435e-06,
      "loss": 3.5905,
      "step": 1600
    },
    {
      "epoch": 0.6981786643538594,
      "grad_norm": 15.078125,
      "learning_rate": 3.488725065047702e-06,
      "loss": 2.5892,
      "step": 1610
    },
    {
      "epoch": 0.7025151777970512,
      "grad_norm": 18.328125,
      "learning_rate": 3.5104076322636606e-06,
      "loss": 3.3452,
      "step": 1620
    },
    {
      "epoch": 0.7068516912402428,
      "grad_norm": 25.171875,
      "learning_rate": 3.5320901994796186e-06,
      "loss": 2.6696,
      "step": 1630
    },
    {
      "epoch": 0.7111882046834345,
      "grad_norm": 24.0,
      "learning_rate": 3.553772766695577e-06,
      "loss": 3.2285,
      "step": 1640
    },
    {
      "epoch": 0.7155247181266262,
      "grad_norm": 22.09375,
      "learning_rate": 3.5754553339115358e-06,
      "loss": 3.1396,
      "step": 1650
    },
    {
      "epoch": 0.7198612315698178,
      "grad_norm": 13.25,
      "learning_rate": 3.5971379011274937e-06,
      "loss": 3.1332,
      "step": 1660
    },
    {
      "epoch": 0.7241977450130095,
      "grad_norm": 19.546875,
      "learning_rate": 3.618820468343452e-06,
      "loss": 3.6973,
      "step": 1670
    },
    {
      "epoch": 0.7285342584562012,
      "grad_norm": 19.21875,
      "learning_rate": 3.640503035559411e-06,
      "loss": 2.5111,
      "step": 1680
    },
    {
      "epoch": 0.7328707718993929,
      "grad_norm": 15.4765625,
      "learning_rate": 3.662185602775369e-06,
      "loss": 2.3382,
      "step": 1690
    },
    {
      "epoch": 0.7372072853425846,
      "grad_norm": 15.1484375,
      "learning_rate": 3.683868169991327e-06,
      "loss": 2.47,
      "step": 1700
    },
    {
      "epoch": 0.7415437987857763,
      "grad_norm": 17.265625,
      "learning_rate": 3.7055507372072855e-06,
      "loss": 3.5493,
      "step": 1710
    },
    {
      "epoch": 0.7458803122289679,
      "grad_norm": 19.578125,
      "learning_rate": 3.727233304423244e-06,
      "loss": 2.856,
      "step": 1720
    },
    {
      "epoch": 0.7502168256721596,
      "grad_norm": 14.21875,
      "learning_rate": 3.7489158716392023e-06,
      "loss": 3.0997,
      "step": 1730
    },
    {
      "epoch": 0.7545533391153513,
      "grad_norm": 19.328125,
      "learning_rate": 3.7705984388551607e-06,
      "loss": 3.0078,
      "step": 1740
    },
    {
      "epoch": 0.7588898525585429,
      "grad_norm": 13.0703125,
      "learning_rate": 3.792281006071119e-06,
      "loss": 2.4847,
      "step": 1750
    },
    {
      "epoch": 0.7632263660017347,
      "grad_norm": 13.4453125,
      "learning_rate": 3.8139635732870774e-06,
      "loss": 2.3696,
      "step": 1760
    },
    {
      "epoch": 0.7675628794449263,
      "grad_norm": 20.03125,
      "learning_rate": 3.835646140503036e-06,
      "loss": 3.774,
      "step": 1770
    },
    {
      "epoch": 0.7718993928881179,
      "grad_norm": 19.015625,
      "learning_rate": 3.857328707718994e-06,
      "loss": 3.134,
      "step": 1780
    },
    {
      "epoch": 0.7762359063313097,
      "grad_norm": 25.109375,
      "learning_rate": 3.879011274934953e-06,
      "loss": 4.3999,
      "step": 1790
    },
    {
      "epoch": 0.7805724197745013,
      "grad_norm": 26.078125,
      "learning_rate": 3.900693842150911e-06,
      "loss": 3.3968,
      "step": 1800
    },
    {
      "epoch": 0.784908933217693,
      "grad_norm": 15.859375,
      "learning_rate": 3.92237640936687e-06,
      "loss": 2.3863,
      "step": 1810
    },
    {
      "epoch": 0.7892454466608847,
      "grad_norm": 15.5078125,
      "learning_rate": 3.944058976582828e-06,
      "loss": 2.9301,
      "step": 1820
    },
    {
      "epoch": 0.7935819601040763,
      "grad_norm": 11.9453125,
      "learning_rate": 3.965741543798786e-06,
      "loss": 3.35,
      "step": 1830
    },
    {
      "epoch": 0.797918473547268,
      "grad_norm": 15.9296875,
      "learning_rate": 3.987424111014744e-06,
      "loss": 3.268,
      "step": 1840
    },
    {
      "epoch": 0.8022549869904597,
      "grad_norm": 17.40625,
      "learning_rate": 4.009106678230703e-06,
      "loss": 3.6494,
      "step": 1850
    },
    {
      "epoch": 0.8065915004336514,
      "grad_norm": 8.546875,
      "learning_rate": 4.030789245446661e-06,
      "loss": 3.9324,
      "step": 1860
    },
    {
      "epoch": 0.810928013876843,
      "grad_norm": 28.578125,
      "learning_rate": 4.05247181266262e-06,
      "loss": 3.7614,
      "step": 1870
    },
    {
      "epoch": 0.8152645273200347,
      "grad_norm": 16.296875,
      "learning_rate": 4.074154379878578e-06,
      "loss": 3.4249,
      "step": 1880
    },
    {
      "epoch": 0.8196010407632264,
      "grad_norm": 30.609375,
      "learning_rate": 4.095836947094537e-06,
      "loss": 3.3493,
      "step": 1890
    },
    {
      "epoch": 0.823937554206418,
      "grad_norm": 15.2109375,
      "learning_rate": 4.1175195143104946e-06,
      "loss": 3.7871,
      "step": 1900
    },
    {
      "epoch": 0.8282740676496098,
      "grad_norm": 22.75,
      "learning_rate": 4.139202081526453e-06,
      "loss": 3.8969,
      "step": 1910
    },
    {
      "epoch": 0.8326105810928014,
      "grad_norm": 23.109375,
      "learning_rate": 4.160884648742411e-06,
      "loss": 4.732,
      "step": 1920
    },
    {
      "epoch": 0.836947094535993,
      "grad_norm": 11.4921875,
      "learning_rate": 4.18256721595837e-06,
      "loss": 3.3877,
      "step": 1930
    },
    {
      "epoch": 0.8412836079791848,
      "grad_norm": 18.4375,
      "learning_rate": 4.204249783174328e-06,
      "loss": 3.4235,
      "step": 1940
    },
    {
      "epoch": 0.8456201214223764,
      "grad_norm": 17.890625,
      "learning_rate": 4.225932350390287e-06,
      "loss": 3.4651,
      "step": 1950
    },
    {
      "epoch": 0.849956634865568,
      "grad_norm": 24.359375,
      "learning_rate": 4.247614917606245e-06,
      "loss": 3.733,
      "step": 1960
    },
    {
      "epoch": 0.8542931483087598,
      "grad_norm": 26.203125,
      "learning_rate": 4.2692974848222036e-06,
      "loss": 3.7384,
      "step": 1970
    },
    {
      "epoch": 0.8586296617519514,
      "grad_norm": 13.2421875,
      "learning_rate": 4.2909800520381615e-06,
      "loss": 4.1869,
      "step": 1980
    },
    {
      "epoch": 0.8629661751951431,
      "grad_norm": 13.625,
      "learning_rate": 4.3126626192541195e-06,
      "loss": 3.0181,
      "step": 1990
    },
    {
      "epoch": 0.8673026886383348,
      "grad_norm": 17.859375,
      "learning_rate": 4.334345186470078e-06,
      "loss": 4.5918,
      "step": 2000
    },
    {
      "epoch": 0.8673026886383348,
      "eval_loss": 0.9994213581085205,
      "eval_runtime": 361.1972,
      "eval_samples_per_second": 3.178,
      "eval_steps_per_second": 3.178,
      "step": 2000
    },
    {
      "epoch": 0.8716392020815265,
      "grad_norm": 15.1875,
      "learning_rate": 5e-07,
      "loss": 3.2362,
      "step": 2010
    },
    {
      "epoch": 0.8759757155247181,
      "grad_norm": 14.4296875,
      "learning_rate": 5e-07,
      "loss": 3.2054,
      "step": 2020
    },
    {
      "epoch": 0.8803122289679098,
      "grad_norm": 12.7734375,
      "learning_rate": 5e-07,
      "loss": 3.2145,
      "step": 2030
    },
    {
      "epoch": 0.8846487424111015,
      "grad_norm": 11.140625,
      "learning_rate": 5e-07,
      "loss": 3.1574,
      "step": 2040
    },
    {
      "epoch": 0.8889852558542931,
      "grad_norm": 14.5859375,
      "learning_rate": 5e-07,
      "loss": 3.1251,
      "step": 2050
    },
    {
      "epoch": 0.8933217692974849,
      "grad_norm": 10.6328125,
      "learning_rate": 5e-07,
      "loss": 3.301,
      "step": 2060
    },
    {
      "epoch": 0.8976582827406765,
      "grad_norm": 14.640625,
      "learning_rate": 5e-07,
      "loss": 2.3001,
      "step": 2070
    },
    {
      "epoch": 0.9019947961838681,
      "grad_norm": 28.171875,
      "learning_rate": 5e-07,
      "loss": 2.332,
      "step": 2080
    },
    {
      "epoch": 0.9063313096270599,
      "grad_norm": 12.2421875,
      "learning_rate": 5e-07,
      "loss": 2.3091,
      "step": 2090
    },
    {
      "epoch": 0.9106678230702515,
      "grad_norm": 15.3359375,
      "learning_rate": 5e-07,
      "loss": 2.5537,
      "step": 2100
    },
    {
      "epoch": 0.9150043365134432,
      "grad_norm": 15.3125,
      "learning_rate": 5e-07,
      "loss": 3.309,
      "step": 2110
    },
    {
      "epoch": 0.9193408499566349,
      "grad_norm": 12.84375,
      "learning_rate": 5e-07,
      "loss": 2.4731,
      "step": 2120
    },
    {
      "epoch": 0.9236773633998265,
      "grad_norm": 14.2890625,
      "learning_rate": 5e-07,
      "loss": 3.0001,
      "step": 2130
    },
    {
      "epoch": 0.9280138768430182,
      "grad_norm": 7.51171875,
      "learning_rate": 5e-07,
      "loss": 1.5443,
      "step": 2140
    },
    {
      "epoch": 0.9323503902862099,
      "grad_norm": 7.53125,
      "learning_rate": 5e-07,
      "loss": 1.6267,
      "step": 2150
    },
    {
      "epoch": 0.9366869037294016,
      "grad_norm": 16.03125,
      "learning_rate": 5e-07,
      "loss": 2.0728,
      "step": 2160
    },
    {
      "epoch": 0.9410234171725932,
      "grad_norm": 7.9453125,
      "learning_rate": 5e-07,
      "loss": 2.454,
      "step": 2170
    },
    {
      "epoch": 0.9453599306157849,
      "grad_norm": 9.3671875,
      "learning_rate": 5e-07,
      "loss": 2.2192,
      "step": 2180
    },
    {
      "epoch": 0.9496964440589766,
      "grad_norm": 14.1328125,
      "learning_rate": 5e-07,
      "loss": 2.3235,
      "step": 2190
    },
    {
      "epoch": 0.9540329575021682,
      "grad_norm": 11.5,
      "learning_rate": 5e-07,
      "loss": 1.298,
      "step": 2200
    },
    {
      "epoch": 0.95836947094536,
      "grad_norm": 11.203125,
      "learning_rate": 5e-07,
      "loss": 1.8626,
      "step": 2210
    },
    {
      "epoch": 0.9627059843885516,
      "grad_norm": 9.421875,
      "learning_rate": 5e-07,
      "loss": 1.9578,
      "step": 2220
    },
    {
      "epoch": 0.9670424978317432,
      "grad_norm": 10.6015625,
      "learning_rate": 5e-07,
      "loss": 1.6782,
      "step": 2230
    },
    {
      "epoch": 0.971379011274935,
      "grad_norm": 6.91796875,
      "learning_rate": 5e-07,
      "loss": 1.6793,
      "step": 2240
    },
    {
      "epoch": 0.9757155247181266,
      "grad_norm": 12.1171875,
      "learning_rate": 5e-07,
      "loss": 1.7789,
      "step": 2250
    },
    {
      "epoch": 0.9800520381613183,
      "grad_norm": 11.2578125,
      "learning_rate": 5e-07,
      "loss": 2.0151,
      "step": 2260
    },
    {
      "epoch": 0.98438855160451,
      "grad_norm": 5.65625,
      "learning_rate": 5e-07,
      "loss": 1.59,
      "step": 2270
    },
    {
      "epoch": 0.9887250650477016,
      "grad_norm": 14.2890625,
      "learning_rate": 5e-07,
      "loss": 1.853,
      "step": 2280
    },
    {
      "epoch": 0.9930615784908933,
      "grad_norm": 6.1953125,
      "learning_rate": 5e-07,
      "loss": 1.1025,
      "step": 2290
    },
    {
      "epoch": 0.997398091934085,
      "grad_norm": 5.95703125,
      "learning_rate": 5e-07,
      "loss": 1.4678,
      "step": 2300
    },
    {
      "epoch": 1.0017346053772767,
      "grad_norm": 10.7109375,
      "learning_rate": 5e-07,
      "loss": 1.9323,
      "step": 2310
    },
    {
      "epoch": 1.0060711188204683,
      "grad_norm": 11.203125,
      "learning_rate": 5e-07,
      "loss": 1.7482,
      "step": 2320
    },
    {
      "epoch": 1.01040763226366,
      "grad_norm": 9.5234375,
      "learning_rate": 5e-07,
      "loss": 1.3478,
      "step": 2330
    },
    {
      "epoch": 1.0147441457068518,
      "grad_norm": 18.265625,
      "learning_rate": 5e-07,
      "loss": 1.6006,
      "step": 2340
    },
    {
      "epoch": 1.0190806591500434,
      "grad_norm": 8.4140625,
      "learning_rate": 5e-07,
      "loss": 1.5607,
      "step": 2350
    },
    {
      "epoch": 1.023417172593235,
      "grad_norm": 8.71875,
      "learning_rate": 5e-07,
      "loss": 1.3238,
      "step": 2360
    },
    {
      "epoch": 1.0277536860364267,
      "grad_norm": 5.93359375,
      "learning_rate": 5e-07,
      "loss": 1.4028,
      "step": 2370
    },
    {
      "epoch": 1.0320901994796183,
      "grad_norm": 3.857421875,
      "learning_rate": 5e-07,
      "loss": 1.3975,
      "step": 2380
    },
    {
      "epoch": 1.03642671292281,
      "grad_norm": 6.97265625,
      "learning_rate": 5e-07,
      "loss": 1.5728,
      "step": 2390
    },
    {
      "epoch": 1.0407632263660018,
      "grad_norm": 9.15625,
      "learning_rate": 5e-07,
      "loss": 1.1999,
      "step": 2400
    },
    {
      "epoch": 1.0450997398091935,
      "grad_norm": 8.3125,
      "learning_rate": 5e-07,
      "loss": 1.6371,
      "step": 2410
    },
    {
      "epoch": 1.049436253252385,
      "grad_norm": 13.6640625,
      "learning_rate": 5e-07,
      "loss": 1.0108,
      "step": 2420
    },
    {
      "epoch": 1.0537727666955767,
      "grad_norm": 12.7890625,
      "learning_rate": 5e-07,
      "loss": 1.56,
      "step": 2430
    },
    {
      "epoch": 1.0581092801387684,
      "grad_norm": 7.0390625,
      "learning_rate": 5e-07,
      "loss": 1.4666,
      "step": 2440
    },
    {
      "epoch": 1.06244579358196,
      "grad_norm": 7.8359375,
      "learning_rate": 5e-07,
      "loss": 1.2073,
      "step": 2450
    },
    {
      "epoch": 1.0667823070251519,
      "grad_norm": 5.77734375,
      "learning_rate": 5e-07,
      "loss": 1.8587,
      "step": 2460
    },
    {
      "epoch": 1.0711188204683435,
      "grad_norm": 9.0703125,
      "learning_rate": 5e-07,
      "loss": 1.1502,
      "step": 2470
    },
    {
      "epoch": 1.0754553339115351,
      "grad_norm": 7.125,
      "learning_rate": 5e-07,
      "loss": 1.0614,
      "step": 2480
    },
    {
      "epoch": 1.0797918473547268,
      "grad_norm": 5.78125,
      "learning_rate": 5e-07,
      "loss": 1.315,
      "step": 2490
    },
    {
      "epoch": 1.0841283607979184,
      "grad_norm": 7.54296875,
      "learning_rate": 5e-07,
      "loss": 1.2356,
      "step": 2500
    },
    {
      "epoch": 1.0841283607979184,
      "eval_loss": 0.36064690351486206,
      "eval_runtime": 356.8626,
      "eval_samples_per_second": 3.217,
      "eval_steps_per_second": 3.217,
      "step": 2500
    },
    {
      "epoch": 1.08846487424111,
      "grad_norm": 6.21875,
      "learning_rate": 5e-07,
      "loss": 1.3064,
      "step": 2510
    },
    {
      "epoch": 1.092801387684302,
      "grad_norm": 10.2109375,
      "learning_rate": 5e-07,
      "loss": 1.3042,
      "step": 2520
    },
    {
      "epoch": 1.0971379011274935,
      "grad_norm": 8.5390625,
      "learning_rate": 5e-07,
      "loss": 1.221,
      "step": 2530
    },
    {
      "epoch": 1.1014744145706852,
      "grad_norm": 7.03515625,
      "learning_rate": 5e-07,
      "loss": 1.1928,
      "step": 2540
    },
    {
      "epoch": 1.1058109280138768,
      "grad_norm": 2.333984375,
      "learning_rate": 5e-07,
      "loss": 1.1142,
      "step": 2550
    },
    {
      "epoch": 1.1101474414570685,
      "grad_norm": 8.2578125,
      "learning_rate": 5e-07,
      "loss": 1.8197,
      "step": 2560
    },
    {
      "epoch": 1.11448395490026,
      "grad_norm": 2.962890625,
      "learning_rate": 5e-07,
      "loss": 1.1441,
      "step": 2570
    },
    {
      "epoch": 1.118820468343452,
      "grad_norm": 9.03125,
      "learning_rate": 5e-07,
      "loss": 1.3163,
      "step": 2580
    },
    {
      "epoch": 1.1231569817866436,
      "grad_norm": 7.5234375,
      "learning_rate": 5e-07,
      "loss": 1.3727,
      "step": 2590
    },
    {
      "epoch": 1.1274934952298352,
      "grad_norm": 6.65625,
      "learning_rate": 5e-07,
      "loss": 1.5097,
      "step": 2600
    },
    {
      "epoch": 1.1318300086730269,
      "grad_norm": 10.1875,
      "learning_rate": 5e-07,
      "loss": 1.483,
      "step": 2610
    },
    {
      "epoch": 1.1361665221162185,
      "grad_norm": 5.85546875,
      "learning_rate": 5e-07,
      "loss": 1.0307,
      "step": 2620
    },
    {
      "epoch": 1.1405030355594103,
      "grad_norm": 4.875,
      "learning_rate": 5e-07,
      "loss": 1.2112,
      "step": 2630
    },
    {
      "epoch": 1.144839549002602,
      "grad_norm": 7.4921875,
      "learning_rate": 5e-07,
      "loss": 1.1523,
      "step": 2640
    },
    {
      "epoch": 1.1491760624457936,
      "grad_norm": 8.2421875,
      "learning_rate": 5e-07,
      "loss": 1.2401,
      "step": 2650
    },
    {
      "epoch": 1.1535125758889853,
      "grad_norm": 5.12890625,
      "learning_rate": 5e-07,
      "loss": 1.1115,
      "step": 2660
    },
    {
      "epoch": 1.157849089332177,
      "grad_norm": 6.1796875,
      "learning_rate": 5e-07,
      "loss": 1.032,
      "step": 2670
    },
    {
      "epoch": 1.1621856027753685,
      "grad_norm": 6.95703125,
      "learning_rate": 5e-07,
      "loss": 1.5435,
      "step": 2680
    },
    {
      "epoch": 1.1665221162185602,
      "grad_norm": 12.7421875,
      "learning_rate": 5e-07,
      "loss": 1.0064,
      "step": 2690
    },
    {
      "epoch": 1.170858629661752,
      "grad_norm": 3.93359375,
      "learning_rate": 5e-07,
      "loss": 1.0834,
      "step": 2700
    },
    {
      "epoch": 1.1751951431049437,
      "grad_norm": 7.98828125,
      "learning_rate": 5e-07,
      "loss": 0.7148,
      "step": 2710
    },
    {
      "epoch": 1.1795316565481353,
      "grad_norm": 7.375,
      "learning_rate": 5e-07,
      "loss": 1.336,
      "step": 2720
    },
    {
      "epoch": 1.183868169991327,
      "grad_norm": 6.12890625,
      "learning_rate": 5e-07,
      "loss": 1.6278,
      "step": 2730
    },
    {
      "epoch": 1.1882046834345186,
      "grad_norm": 5.8359375,
      "learning_rate": 5e-07,
      "loss": 1.1738,
      "step": 2740
    },
    {
      "epoch": 1.1925411968777104,
      "grad_norm": 7.11328125,
      "learning_rate": 5e-07,
      "loss": 1.1472,
      "step": 2750
    },
    {
      "epoch": 1.196877710320902,
      "grad_norm": 9.0234375,
      "learning_rate": 5e-07,
      "loss": 1.1923,
      "step": 2760
    },
    {
      "epoch": 1.2012142237640937,
      "grad_norm": 3.953125,
      "learning_rate": 5e-07,
      "loss": 0.9557,
      "step": 2770
    },
    {
      "epoch": 1.2055507372072853,
      "grad_norm": 4.875,
      "learning_rate": 5e-07,
      "loss": 1.07,
      "step": 2780
    },
    {
      "epoch": 1.209887250650477,
      "grad_norm": 10.1015625,
      "learning_rate": 5e-07,
      "loss": 1.108,
      "step": 2790
    },
    {
      "epoch": 1.2142237640936686,
      "grad_norm": 8.953125,
      "learning_rate": 5e-07,
      "loss": 0.9251,
      "step": 2800
    },
    {
      "epoch": 1.2185602775368602,
      "grad_norm": 6.83984375,
      "learning_rate": 5e-07,
      "loss": 1.0999,
      "step": 2810
    },
    {
      "epoch": 1.222896790980052,
      "grad_norm": 4.93359375,
      "learning_rate": 5e-07,
      "loss": 0.9123,
      "step": 2820
    },
    {
      "epoch": 1.2272333044232437,
      "grad_norm": 6.86328125,
      "learning_rate": 5e-07,
      "loss": 1.27,
      "step": 2830
    },
    {
      "epoch": 1.2315698178664354,
      "grad_norm": 5.4140625,
      "learning_rate": 5e-07,
      "loss": 0.9448,
      "step": 2840
    },
    {
      "epoch": 1.235906331309627,
      "grad_norm": 5.96484375,
      "learning_rate": 5e-07,
      "loss": 1.0828,
      "step": 2850
    },
    {
      "epoch": 1.2402428447528187,
      "grad_norm": 8.359375,
      "learning_rate": 5e-07,
      "loss": 1.0442,
      "step": 2860
    },
    {
      "epoch": 1.2445793581960105,
      "grad_norm": 3.884765625,
      "learning_rate": 5e-07,
      "loss": 1.0766,
      "step": 2870
    },
    {
      "epoch": 1.2489158716392021,
      "grad_norm": 3.455078125,
      "learning_rate": 5e-07,
      "loss": 0.9912,
      "step": 2880
    },
    {
      "epoch": 1.2532523850823938,
      "grad_norm": 6.5,
      "learning_rate": 5e-07,
      "loss": 1.2884,
      "step": 2890
    },
    {
      "epoch": 1.2575888985255854,
      "grad_norm": 4.0859375,
      "learning_rate": 5e-07,
      "loss": 0.8357,
      "step": 2900
    },
    {
      "epoch": 1.261925411968777,
      "grad_norm": 9.671875,
      "learning_rate": 5e-07,
      "loss": 1.1782,
      "step": 2910
    },
    {
      "epoch": 1.266261925411969,
      "grad_norm": 8.390625,
      "learning_rate": 5e-07,
      "loss": 1.1043,
      "step": 2920
    },
    {
      "epoch": 1.2705984388551603,
      "grad_norm": 7.21484375,
      "learning_rate": 5e-07,
      "loss": 1.0992,
      "step": 2930
    },
    {
      "epoch": 1.2749349522983522,
      "grad_norm": 5.23046875,
      "learning_rate": 5e-07,
      "loss": 1.3262,
      "step": 2940
    },
    {
      "epoch": 1.2792714657415438,
      "grad_norm": 7.80078125,
      "learning_rate": 5e-07,
      "loss": 1.1803,
      "step": 2950
    },
    {
      "epoch": 1.2836079791847355,
      "grad_norm": 18.796875,
      "learning_rate": 5e-07,
      "loss": 1.0296,
      "step": 2960
    },
    {
      "epoch": 1.287944492627927,
      "grad_norm": 7.0546875,
      "learning_rate": 5e-07,
      "loss": 1.1805,
      "step": 2970
    },
    {
      "epoch": 1.2922810060711187,
      "grad_norm": 7.32421875,
      "learning_rate": 5e-07,
      "loss": 1.2959,
      "step": 2980
    },
    {
      "epoch": 1.2966175195143106,
      "grad_norm": 12.6171875,
      "learning_rate": 5e-07,
      "loss": 1.0241,
      "step": 2990
    },
    {
      "epoch": 1.3009540329575022,
      "grad_norm": 3.474609375,
      "learning_rate": 5e-07,
      "loss": 0.8864,
      "step": 3000
    },
    {
      "epoch": 1.3009540329575022,
      "eval_loss": 0.29353854060173035,
      "eval_runtime": 357.8433,
      "eval_samples_per_second": 3.208,
      "eval_steps_per_second": 3.208,
      "step": 3000
    },
    {
      "epoch": 1.3052905464006939,
      "grad_norm": 4.359375,
      "learning_rate": 5e-07,
      "loss": 0.9774,
      "step": 3010
    },
    {
      "epoch": 1.3096270598438855,
      "grad_norm": 7.203125,
      "learning_rate": 5e-07,
      "loss": 1.0685,
      "step": 3020
    },
    {
      "epoch": 1.3139635732870771,
      "grad_norm": 7.5078125,
      "learning_rate": 5e-07,
      "loss": 1.2891,
      "step": 3030
    },
    {
      "epoch": 1.318300086730269,
      "grad_norm": 5.50390625,
      "learning_rate": 5e-07,
      "loss": 1.148,
      "step": 3040
    },
    {
      "epoch": 1.3226366001734604,
      "grad_norm": 3.953125,
      "learning_rate": 5e-07,
      "loss": 0.8926,
      "step": 3050
    },
    {
      "epoch": 1.3269731136166523,
      "grad_norm": 6.84375,
      "learning_rate": 5e-07,
      "loss": 0.6936,
      "step": 3060
    },
    {
      "epoch": 1.331309627059844,
      "grad_norm": 4.01171875,
      "learning_rate": 5e-07,
      "loss": 0.9083,
      "step": 3070
    },
    {
      "epoch": 1.3356461405030355,
      "grad_norm": 12.140625,
      "learning_rate": 5e-07,
      "loss": 0.7945,
      "step": 3080
    },
    {
      "epoch": 1.3399826539462272,
      "grad_norm": 5.07421875,
      "learning_rate": 5e-07,
      "loss": 0.9535,
      "step": 3090
    },
    {
      "epoch": 1.3443191673894188,
      "grad_norm": 3.876953125,
      "learning_rate": 5e-07,
      "loss": 0.8825,
      "step": 3100
    },
    {
      "epoch": 1.3486556808326107,
      "grad_norm": 22.46875,
      "learning_rate": 5e-07,
      "loss": 1.0912,
      "step": 3110
    },
    {
      "epoch": 1.3529921942758023,
      "grad_norm": 6.06640625,
      "learning_rate": 5e-07,
      "loss": 1.1039,
      "step": 3120
    },
    {
      "epoch": 1.357328707718994,
      "grad_norm": 6.0390625,
      "learning_rate": 5e-07,
      "loss": 0.7186,
      "step": 3130
    },
    {
      "epoch": 1.3616652211621856,
      "grad_norm": 10.125,
      "learning_rate": 5e-07,
      "loss": 1.0916,
      "step": 3140
    },
    {
      "epoch": 1.3660017346053772,
      "grad_norm": 2.564453125,
      "learning_rate": 5e-07,
      "loss": 1.0973,
      "step": 3150
    },
    {
      "epoch": 1.370338248048569,
      "grad_norm": 10.53125,
      "learning_rate": 5e-07,
      "loss": 0.7901,
      "step": 3160
    },
    {
      "epoch": 1.3746747614917605,
      "grad_norm": 3.259765625,
      "learning_rate": 5e-07,
      "loss": 1.1929,
      "step": 3170
    },
    {
      "epoch": 1.3790112749349523,
      "grad_norm": 6.39453125,
      "learning_rate": 5e-07,
      "loss": 0.6212,
      "step": 3180
    },
    {
      "epoch": 1.383347788378144,
      "grad_norm": 6.1171875,
      "learning_rate": 5e-07,
      "loss": 1.2393,
      "step": 3190
    },
    {
      "epoch": 1.3876843018213356,
      "grad_norm": 6.34765625,
      "learning_rate": 5e-07,
      "loss": 1.2205,
      "step": 3200
    },
    {
      "epoch": 1.3920208152645273,
      "grad_norm": 9.046875,
      "learning_rate": 5e-07,
      "loss": 0.892,
      "step": 3210
    },
    {
      "epoch": 1.396357328707719,
      "grad_norm": 6.30078125,
      "learning_rate": 5e-07,
      "loss": 1.6091,
      "step": 3220
    },
    {
      "epoch": 1.4006938421509108,
      "grad_norm": 9.0,
      "learning_rate": 5e-07,
      "loss": 1.012,
      "step": 3230
    },
    {
      "epoch": 1.4050303555941024,
      "grad_norm": 20.96875,
      "learning_rate": 5e-07,
      "loss": 1.5016,
      "step": 3240
    },
    {
      "epoch": 1.409366869037294,
      "grad_norm": 10.640625,
      "learning_rate": 5e-07,
      "loss": 0.8426,
      "step": 3250
    },
    {
      "epoch": 1.4137033824804857,
      "grad_norm": 3.755859375,
      "learning_rate": 5e-07,
      "loss": 0.7766,
      "step": 3260
    },
    {
      "epoch": 1.4180398959236773,
      "grad_norm": 4.4296875,
      "learning_rate": 5e-07,
      "loss": 1.1471,
      "step": 3270
    },
    {
      "epoch": 1.4223764093668692,
      "grad_norm": 7.21875,
      "learning_rate": 5e-07,
      "loss": 0.7287,
      "step": 3280
    },
    {
      "epoch": 1.4267129228100608,
      "grad_norm": 7.6328125,
      "learning_rate": 5e-07,
      "loss": 0.9941,
      "step": 3290
    },
    {
      "epoch": 1.4310494362532524,
      "grad_norm": 6.33984375,
      "learning_rate": 5e-07,
      "loss": 1.1628,
      "step": 3300
    },
    {
      "epoch": 1.435385949696444,
      "grad_norm": 5.48828125,
      "learning_rate": 5e-07,
      "loss": 0.9256,
      "step": 3310
    },
    {
      "epoch": 1.4397224631396357,
      "grad_norm": 5.12890625,
      "learning_rate": 5e-07,
      "loss": 0.6051,
      "step": 3320
    },
    {
      "epoch": 1.4440589765828273,
      "grad_norm": 3.7578125,
      "learning_rate": 5e-07,
      "loss": 0.8552,
      "step": 3330
    },
    {
      "epoch": 1.448395490026019,
      "grad_norm": 13.9765625,
      "learning_rate": 5e-07,
      "loss": 1.331,
      "step": 3340
    },
    {
      "epoch": 1.4527320034692108,
      "grad_norm": 15.28125,
      "learning_rate": 5e-07,
      "loss": 1.4595,
      "step": 3350
    },
    {
      "epoch": 1.4570685169124025,
      "grad_norm": 7.078125,
      "learning_rate": 5e-07,
      "loss": 0.9543,
      "step": 3360
    },
    {
      "epoch": 1.461405030355594,
      "grad_norm": 4.80859375,
      "learning_rate": 5e-07,
      "loss": 0.8716,
      "step": 3370
    },
    {
      "epoch": 1.4657415437987857,
      "grad_norm": 5.328125,
      "learning_rate": 5e-07,
      "loss": 0.7755,
      "step": 3380
    },
    {
      "epoch": 1.4700780572419774,
      "grad_norm": 5.6328125,
      "learning_rate": 5e-07,
      "loss": 1.1219,
      "step": 3390
    },
    {
      "epoch": 1.4744145706851692,
      "grad_norm": 9.2421875,
      "learning_rate": 5e-07,
      "loss": 0.967,
      "step": 3400
    },
    {
      "epoch": 1.4787510841283609,
      "grad_norm": 8.6171875,
      "learning_rate": 5e-07,
      "loss": 1.2634,
      "step": 3410
    },
    {
      "epoch": 1.4830875975715525,
      "grad_norm": 9.40625,
      "learning_rate": 5e-07,
      "loss": 1.037,
      "step": 3420
    },
    {
      "epoch": 1.4874241110147441,
      "grad_norm": 2.265625,
      "learning_rate": 5e-07,
      "loss": 1.1226,
      "step": 3430
    },
    {
      "epoch": 1.4917606244579358,
      "grad_norm": 12.578125,
      "learning_rate": 5e-07,
      "loss": 1.2653,
      "step": 3440
    },
    {
      "epoch": 1.4960971379011274,
      "grad_norm": 4.73046875,
      "learning_rate": 5e-07,
      "loss": 1.1246,
      "step": 3450
    },
    {
      "epoch": 1.500433651344319,
      "grad_norm": 4.2421875,
      "learning_rate": 5e-07,
      "loss": 0.8926,
      "step": 3460
    },
    {
      "epoch": 1.504770164787511,
      "grad_norm": 11.8359375,
      "learning_rate": 5e-07,
      "loss": 0.9385,
      "step": 3470
    },
    {
      "epoch": 1.5091066782307025,
      "grad_norm": 16.125,
      "learning_rate": 5e-07,
      "loss": 0.7588,
      "step": 3480
    },
    {
      "epoch": 1.5134431916738942,
      "grad_norm": 2.654296875,
      "learning_rate": 5e-07,
      "loss": 0.6703,
      "step": 3490
    },
    {
      "epoch": 1.5177797051170858,
      "grad_norm": 5.61328125,
      "learning_rate": 5e-07,
      "loss": 0.7913,
      "step": 3500
    },
    {
      "epoch": 1.5177797051170858,
      "eval_loss": 0.2694127559661865,
      "eval_runtime": 358.9506,
      "eval_samples_per_second": 3.198,
      "eval_steps_per_second": 3.198,
      "step": 3500
    },
    {
      "epoch": 1.5221162185602775,
      "grad_norm": 4.296875,
      "learning_rate": 5e-07,
      "loss": 0.6551,
      "step": 3510
    },
    {
      "epoch": 1.5264527320034693,
      "grad_norm": 5.0,
      "learning_rate": 5e-07,
      "loss": 0.8813,
      "step": 3520
    },
    {
      "epoch": 1.5307892454466607,
      "grad_norm": 5.875,
      "learning_rate": 5e-07,
      "loss": 0.9077,
      "step": 3530
    },
    {
      "epoch": 1.5351257588898526,
      "grad_norm": 9.25,
      "learning_rate": 5e-07,
      "loss": 0.9285,
      "step": 3540
    },
    {
      "epoch": 1.5394622723330442,
      "grad_norm": 4.15625,
      "learning_rate": 5e-07,
      "loss": 0.8254,
      "step": 3550
    },
    {
      "epoch": 1.5437987857762359,
      "grad_norm": 6.73828125,
      "learning_rate": 5e-07,
      "loss": 0.7482,
      "step": 3560
    },
    {
      "epoch": 1.5481352992194277,
      "grad_norm": 4.375,
      "learning_rate": 5e-07,
      "loss": 0.7,
      "step": 3570
    },
    {
      "epoch": 1.5524718126626191,
      "grad_norm": 3.751953125,
      "learning_rate": 5e-07,
      "loss": 0.7055,
      "step": 3580
    },
    {
      "epoch": 1.556808326105811,
      "grad_norm": 4.58984375,
      "learning_rate": 5e-07,
      "loss": 0.9137,
      "step": 3590
    },
    {
      "epoch": 1.5611448395490026,
      "grad_norm": 4.8203125,
      "learning_rate": 5e-07,
      "loss": 0.841,
      "step": 3600
    },
    {
      "epoch": 1.5654813529921943,
      "grad_norm": 2.205078125,
      "learning_rate": 5e-07,
      "loss": 0.9595,
      "step": 3610
    },
    {
      "epoch": 1.569817866435386,
      "grad_norm": 3.353515625,
      "learning_rate": 5e-07,
      "loss": 1.0338,
      "step": 3620
    },
    {
      "epoch": 1.5741543798785775,
      "grad_norm": 12.7265625,
      "learning_rate": 5e-07,
      "loss": 0.7508,
      "step": 3630
    },
    {
      "epoch": 1.5784908933217694,
      "grad_norm": 12.3125,
      "learning_rate": 5e-07,
      "loss": 1.1762,
      "step": 3640
    },
    {
      "epoch": 1.5828274067649608,
      "grad_norm": 3.216796875,
      "learning_rate": 5e-07,
      "loss": 0.9823,
      "step": 3650
    },
    {
      "epoch": 1.5871639202081527,
      "grad_norm": 4.51953125,
      "learning_rate": 5e-07,
      "loss": 0.9145,
      "step": 3660
    },
    {
      "epoch": 1.5915004336513443,
      "grad_norm": 5.359375,
      "learning_rate": 5e-07,
      "loss": 0.8734,
      "step": 3670
    },
    {
      "epoch": 1.595836947094536,
      "grad_norm": 6.43359375,
      "learning_rate": 5e-07,
      "loss": 0.8589,
      "step": 3680
    },
    {
      "epoch": 1.6001734605377278,
      "grad_norm": 2.732421875,
      "learning_rate": 5e-07,
      "loss": 0.8493,
      "step": 3690
    },
    {
      "epoch": 1.6045099739809192,
      "grad_norm": 23.03125,
      "learning_rate": 5e-07,
      "loss": 0.8376,
      "step": 3700
    },
    {
      "epoch": 1.608846487424111,
      "grad_norm": 6.48828125,
      "learning_rate": 5e-07,
      "loss": 1.0581,
      "step": 3710
    },
    {
      "epoch": 1.6131830008673027,
      "grad_norm": 2.912109375,
      "learning_rate": 5e-07,
      "loss": 1.0817,
      "step": 3720
    },
    {
      "epoch": 1.6175195143104943,
      "grad_norm": 4.1640625,
      "learning_rate": 5e-07,
      "loss": 0.7223,
      "step": 3730
    },
    {
      "epoch": 1.6218560277536862,
      "grad_norm": 6.44921875,
      "learning_rate": 5e-07,
      "loss": 0.741,
      "step": 3740
    },
    {
      "epoch": 1.6261925411968776,
      "grad_norm": 5.5,
      "learning_rate": 5e-07,
      "loss": 0.7666,
      "step": 3750
    },
    {
      "epoch": 1.6305290546400695,
      "grad_norm": 6.53125,
      "learning_rate": 5e-07,
      "loss": 1.176,
      "step": 3760
    },
    {
      "epoch": 1.6348655680832611,
      "grad_norm": 6.79296875,
      "learning_rate": 5e-07,
      "loss": 0.6085,
      "step": 3770
    },
    {
      "epoch": 1.6392020815264527,
      "grad_norm": 9.921875,
      "learning_rate": 5e-07,
      "loss": 0.7271,
      "step": 3780
    },
    {
      "epoch": 1.6435385949696444,
      "grad_norm": 15.7890625,
      "learning_rate": 5e-07,
      "loss": 1.2186,
      "step": 3790
    },
    {
      "epoch": 1.647875108412836,
      "grad_norm": 5.95703125,
      "learning_rate": 5e-07,
      "loss": 0.844,
      "step": 3800
    },
    {
      "epoch": 1.6522116218560279,
      "grad_norm": 4.9140625,
      "learning_rate": 5e-07,
      "loss": 0.4299,
      "step": 3810
    },
    {
      "epoch": 1.6565481352992193,
      "grad_norm": 36.96875,
      "learning_rate": 5e-07,
      "loss": 0.9919,
      "step": 3820
    },
    {
      "epoch": 1.6608846487424112,
      "grad_norm": 8.0703125,
      "learning_rate": 5e-07,
      "loss": 1.058,
      "step": 3830
    },
    {
      "epoch": 1.6652211621856028,
      "grad_norm": 8.140625,
      "learning_rate": 5e-07,
      "loss": 0.7604,
      "step": 3840
    },
    {
      "epoch": 1.6695576756287944,
      "grad_norm": 7.765625,
      "learning_rate": 5e-07,
      "loss": 1.1248,
      "step": 3850
    },
    {
      "epoch": 1.6738941890719863,
      "grad_norm": 4.609375,
      "learning_rate": 5e-07,
      "loss": 0.6629,
      "step": 3860
    },
    {
      "epoch": 1.6782307025151777,
      "grad_norm": 7.1875,
      "learning_rate": 5e-07,
      "loss": 0.8131,
      "step": 3870
    },
    {
      "epoch": 1.6825672159583696,
      "grad_norm": 14.328125,
      "learning_rate": 5e-07,
      "loss": 0.9518,
      "step": 3880
    },
    {
      "epoch": 1.6869037294015612,
      "grad_norm": 3.33203125,
      "learning_rate": 5e-07,
      "loss": 0.7595,
      "step": 3890
    },
    {
      "epoch": 1.6912402428447528,
      "grad_norm": 9.9921875,
      "learning_rate": 5e-07,
      "loss": 0.755,
      "step": 3900
    },
    {
      "epoch": 1.6955767562879445,
      "grad_norm": 4.34765625,
      "learning_rate": 5e-07,
      "loss": 1.1276,
      "step": 3910
    },
    {
      "epoch": 1.699913269731136,
      "grad_norm": 6.44921875,
      "learning_rate": 5e-07,
      "loss": 1.076,
      "step": 3920
    },
    {
      "epoch": 1.704249783174328,
      "grad_norm": 5.00390625,
      "learning_rate": 5e-07,
      "loss": 0.7449,
      "step": 3930
    },
    {
      "epoch": 1.7085862966175194,
      "grad_norm": 7.41015625,
      "learning_rate": 5e-07,
      "loss": 0.9696,
      "step": 3940
    },
    {
      "epoch": 1.7129228100607112,
      "grad_norm": 7.30078125,
      "learning_rate": 5e-07,
      "loss": 0.9504,
      "step": 3950
    },
    {
      "epoch": 1.7172593235039029,
      "grad_norm": 5.55078125,
      "learning_rate": 5e-07,
      "loss": 0.8228,
      "step": 3960
    },
    {
      "epoch": 1.7215958369470945,
      "grad_norm": 7.86328125,
      "learning_rate": 5e-07,
      "loss": 0.9157,
      "step": 3970
    },
    {
      "epoch": 1.7259323503902864,
      "grad_norm": 4.56640625,
      "learning_rate": 5e-07,
      "loss": 1.0846,
      "step": 3980
    },
    {
      "epoch": 1.7302688638334778,
      "grad_norm": 4.33984375,
      "learning_rate": 5e-07,
      "loss": 0.604,
      "step": 3990
    },
    {
      "epoch": 1.7346053772766696,
      "grad_norm": 5.2734375,
      "learning_rate": 5e-07,
      "loss": 0.6006,
      "step": 4000
    },
    {
      "epoch": 1.7346053772766696,
      "eval_loss": 0.2493589222431183,
      "eval_runtime": 357.8166,
      "eval_samples_per_second": 3.208,
      "eval_steps_per_second": 3.208,
      "step": 4000
    },
    {
      "epoch": 1.7389418907198613,
      "grad_norm": 3.8125,
      "learning_rate": 5e-07,
      "loss": 0.6714,
      "step": 4010
    },
    {
      "epoch": 1.743278404163053,
      "grad_norm": 7.4765625,
      "learning_rate": 5e-07,
      "loss": 1.2386,
      "step": 4020
    },
    {
      "epoch": 1.7476149176062445,
      "grad_norm": 3.744140625,
      "learning_rate": 5e-07,
      "loss": 0.7444,
      "step": 4030
    },
    {
      "epoch": 1.7519514310494362,
      "grad_norm": 4.83984375,
      "learning_rate": 5e-07,
      "loss": 0.6483,
      "step": 4040
    },
    {
      "epoch": 1.756287944492628,
      "grad_norm": 9.0859375,
      "learning_rate": 5e-07,
      "loss": 0.7879,
      "step": 4050
    },
    {
      "epoch": 1.7606244579358195,
      "grad_norm": 2.611328125,
      "learning_rate": 5e-07,
      "loss": 0.5884,
      "step": 4060
    },
    {
      "epoch": 1.7649609713790113,
      "grad_norm": 7.296875,
      "learning_rate": 5e-07,
      "loss": 0.7504,
      "step": 4070
    },
    {
      "epoch": 1.769297484822203,
      "grad_norm": 5.29296875,
      "learning_rate": 5e-07,
      "loss": 1.0075,
      "step": 4080
    },
    {
      "epoch": 1.7736339982653946,
      "grad_norm": 11.0390625,
      "learning_rate": 5e-07,
      "loss": 1.0423,
      "step": 4090
    },
    {
      "epoch": 1.7779705117085864,
      "grad_norm": 7.13671875,
      "learning_rate": 5e-07,
      "loss": 0.8914,
      "step": 4100
    },
    {
      "epoch": 1.7823070251517779,
      "grad_norm": 5.0859375,
      "learning_rate": 5e-07,
      "loss": 0.9347,
      "step": 4110
    },
    {
      "epoch": 1.7866435385949697,
      "grad_norm": 2.689453125,
      "learning_rate": 5e-07,
      "loss": 0.6376,
      "step": 4120
    },
    {
      "epoch": 1.7909800520381614,
      "grad_norm": 6.59375,
      "learning_rate": 5e-07,
      "loss": 0.7078,
      "step": 4130
    },
    {
      "epoch": 1.795316565481353,
      "grad_norm": 4.05859375,
      "learning_rate": 5e-07,
      "loss": 0.7142,
      "step": 4140
    },
    {
      "epoch": 1.7996530789245446,
      "grad_norm": 5.2421875,
      "learning_rate": 5e-07,
      "loss": 0.8456,
      "step": 4150
    },
    {
      "epoch": 1.8039895923677363,
      "grad_norm": 8.0390625,
      "learning_rate": 5e-07,
      "loss": 0.8924,
      "step": 4160
    },
    {
      "epoch": 1.8083261058109281,
      "grad_norm": 5.5546875,
      "learning_rate": 5e-07,
      "loss": 0.8467,
      "step": 4170
    },
    {
      "epoch": 1.8126626192541195,
      "grad_norm": 10.9296875,
      "learning_rate": 5e-07,
      "loss": 0.8604,
      "step": 4180
    },
    {
      "epoch": 1.8169991326973114,
      "grad_norm": 6.5234375,
      "learning_rate": 5e-07,
      "loss": 0.7209,
      "step": 4190
    },
    {
      "epoch": 1.821335646140503,
      "grad_norm": 7.9140625,
      "learning_rate": 5e-07,
      "loss": 0.729,
      "step": 4200
    },
    {
      "epoch": 1.8256721595836947,
      "grad_norm": 10.1953125,
      "learning_rate": 5e-07,
      "loss": 1.0535,
      "step": 4210
    },
    {
      "epoch": 1.8300086730268865,
      "grad_norm": 8.3125,
      "learning_rate": 5e-07,
      "loss": 0.8409,
      "step": 4220
    },
    {
      "epoch": 1.834345186470078,
      "grad_norm": 6.22265625,
      "learning_rate": 5e-07,
      "loss": 0.8111,
      "step": 4230
    },
    {
      "epoch": 1.8386816999132698,
      "grad_norm": 9.4375,
      "learning_rate": 5e-07,
      "loss": 0.7562,
      "step": 4240
    },
    {
      "epoch": 1.8430182133564614,
      "grad_norm": 9.71875,
      "learning_rate": 5e-07,
      "loss": 0.739,
      "step": 4250
    },
    {
      "epoch": 1.847354726799653,
      "grad_norm": 3.982421875,
      "learning_rate": 5e-07,
      "loss": 0.7588,
      "step": 4260
    },
    {
      "epoch": 1.8516912402428447,
      "grad_norm": 3.779296875,
      "learning_rate": 5e-07,
      "loss": 0.6939,
      "step": 4270
    },
    {
      "epoch": 1.8560277536860363,
      "grad_norm": 8.0546875,
      "learning_rate": 5e-07,
      "loss": 0.7679,
      "step": 4280
    },
    {
      "epoch": 1.8603642671292282,
      "grad_norm": 3.904296875,
      "learning_rate": 5e-07,
      "loss": 0.6234,
      "step": 4290
    },
    {
      "epoch": 1.8647007805724196,
      "grad_norm": 6.015625,
      "learning_rate": 5e-07,
      "loss": 0.7342,
      "step": 4300
    },
    {
      "epoch": 1.8690372940156115,
      "grad_norm": 3.12890625,
      "learning_rate": 5e-07,
      "loss": 1.4717,
      "step": 4310
    },
    {
      "epoch": 1.873373807458803,
      "grad_norm": 4.7578125,
      "learning_rate": 5e-07,
      "loss": 0.7458,
      "step": 4320
    },
    {
      "epoch": 1.8777103209019947,
      "grad_norm": 6.78125,
      "learning_rate": 5e-07,
      "loss": 1.0065,
      "step": 4330
    },
    {
      "epoch": 1.8820468343451866,
      "grad_norm": 11.1015625,
      "learning_rate": 5e-07,
      "loss": 0.9169,
      "step": 4340
    },
    {
      "epoch": 1.886383347788378,
      "grad_norm": 3.345703125,
      "learning_rate": 5e-07,
      "loss": 0.7996,
      "step": 4350
    },
    {
      "epoch": 1.8907198612315699,
      "grad_norm": 8.5625,
      "learning_rate": 5e-07,
      "loss": 1.2083,
      "step": 4360
    },
    {
      "epoch": 1.8950563746747615,
      "grad_norm": 4.1484375,
      "learning_rate": 5e-07,
      "loss": 0.6529,
      "step": 4370
    },
    {
      "epoch": 1.8993928881179531,
      "grad_norm": 5.32421875,
      "learning_rate": 5e-07,
      "loss": 0.7471,
      "step": 4380
    },
    {
      "epoch": 1.9037294015611448,
      "grad_norm": 4.9765625,
      "learning_rate": 5e-07,
      "loss": 0.8651,
      "step": 4390
    },
    {
      "epoch": 1.9080659150043364,
      "grad_norm": 3.98828125,
      "learning_rate": 5e-07,
      "loss": 0.6603,
      "step": 4400
    },
    {
      "epoch": 1.9124024284475283,
      "grad_norm": 6.40234375,
      "learning_rate": 5e-07,
      "loss": 1.1579,
      "step": 4410
    },
    {
      "epoch": 1.91673894189072,
      "grad_norm": 9.640625,
      "learning_rate": 5e-07,
      "loss": 0.9742,
      "step": 4420
    },
    {
      "epoch": 1.9210754553339116,
      "grad_norm": 6.2890625,
      "learning_rate": 5e-07,
      "loss": 1.2195,
      "step": 4430
    },
    {
      "epoch": 1.9254119687771032,
      "grad_norm": 7.2265625,
      "learning_rate": 5e-07,
      "loss": 0.7349,
      "step": 4440
    },
    {
      "epoch": 1.9297484822202948,
      "grad_norm": 5.96484375,
      "learning_rate": 5e-07,
      "loss": 0.798,
      "step": 4450
    },
    {
      "epoch": 1.9340849956634867,
      "grad_norm": 4.34765625,
      "learning_rate": 5e-07,
      "loss": 0.6273,
      "step": 4460
    },
    {
      "epoch": 1.938421509106678,
      "grad_norm": 8.4609375,
      "learning_rate": 5e-07,
      "loss": 1.0357,
      "step": 4470
    },
    {
      "epoch": 1.94275802254987,
      "grad_norm": 5.06640625,
      "learning_rate": 5e-07,
      "loss": 0.7975,
      "step": 4480
    },
    {
      "epoch": 1.9470945359930616,
      "grad_norm": 7.26171875,
      "learning_rate": 5e-07,
      "loss": 0.9741,
      "step": 4490
    },
    {
      "epoch": 1.9514310494362532,
      "grad_norm": 4.046875,
      "learning_rate": 5e-07,
      "loss": 0.9937,
      "step": 4500
    },
    {
      "epoch": 1.9514310494362532,
      "eval_loss": 0.2373046875,
      "eval_runtime": 358.8486,
      "eval_samples_per_second": 3.199,
      "eval_steps_per_second": 3.199,
      "step": 4500
    },
    {
      "epoch": 1.955767562879445,
      "grad_norm": 2.248046875,
      "learning_rate": 5e-07,
      "loss": 0.8049,
      "step": 4510
    },
    {
      "epoch": 1.9601040763226365,
      "grad_norm": 8.890625,
      "learning_rate": 5e-07,
      "loss": 0.8718,
      "step": 4520
    },
    {
      "epoch": 1.9644405897658284,
      "grad_norm": 6.8984375,
      "learning_rate": 5e-07,
      "loss": 0.7798,
      "step": 4530
    },
    {
      "epoch": 1.96877710320902,
      "grad_norm": 5.05859375,
      "learning_rate": 5e-07,
      "loss": 0.8669,
      "step": 4540
    },
    {
      "epoch": 1.9731136166522116,
      "grad_norm": 9.140625,
      "learning_rate": 5e-07,
      "loss": 0.7444,
      "step": 4550
    },
    {
      "epoch": 1.9774501300954033,
      "grad_norm": 7.0078125,
      "learning_rate": 5e-07,
      "loss": 0.9435,
      "step": 4560
    },
    {
      "epoch": 1.981786643538595,
      "grad_norm": 4.2734375,
      "learning_rate": 5e-07,
      "loss": 0.6953,
      "step": 4570
    },
    {
      "epoch": 1.9861231569817868,
      "grad_norm": 22.875,
      "learning_rate": 5e-07,
      "loss": 0.86,
      "step": 4580
    },
    {
      "epoch": 1.9904596704249782,
      "grad_norm": 4.78515625,
      "learning_rate": 5e-07,
      "loss": 0.8235,
      "step": 4590
    },
    {
      "epoch": 1.99479618386817,
      "grad_norm": 4.78125,
      "learning_rate": 5e-07,
      "loss": 0.5702,
      "step": 4600
    },
    {
      "epoch": 1.9991326973113617,
      "grad_norm": 4.54296875,
      "learning_rate": 5e-07,
      "loss": 0.8109,
      "step": 4610
    },
    {
      "epoch": 2.0034692107545533,
      "grad_norm": 5.83984375,
      "learning_rate": 5e-07,
      "loss": 0.6309,
      "step": 4620
    },
    {
      "epoch": 2.007805724197745,
      "grad_norm": 6.140625,
      "learning_rate": 5e-07,
      "loss": 0.9938,
      "step": 4630
    },
    {
      "epoch": 2.0121422376409366,
      "grad_norm": 4.265625,
      "learning_rate": 5e-07,
      "loss": 0.5045,
      "step": 4640
    },
    {
      "epoch": 2.0164787510841284,
      "grad_norm": 3.53125,
      "learning_rate": 5e-07,
      "loss": 0.4984,
      "step": 4650
    },
    {
      "epoch": 2.02081526452732,
      "grad_norm": 1.77734375,
      "learning_rate": 5e-07,
      "loss": 0.4955,
      "step": 4660
    },
    {
      "epoch": 2.0251517779705117,
      "grad_norm": 6.1875,
      "learning_rate": 5e-07,
      "loss": 0.588,
      "step": 4670
    },
    {
      "epoch": 2.0294882914137036,
      "grad_norm": 7.10546875,
      "learning_rate": 5e-07,
      "loss": 0.7382,
      "step": 4680
    },
    {
      "epoch": 2.033824804856895,
      "grad_norm": 5.19140625,
      "learning_rate": 5e-07,
      "loss": 0.6366,
      "step": 4690
    },
    {
      "epoch": 2.038161318300087,
      "grad_norm": 5.4609375,
      "learning_rate": 5e-07,
      "loss": 0.5693,
      "step": 4700
    },
    {
      "epoch": 2.0424978317432783,
      "grad_norm": 3.65625,
      "learning_rate": 5e-07,
      "loss": 0.5193,
      "step": 4710
    },
    {
      "epoch": 2.04683434518647,
      "grad_norm": 6.03125,
      "learning_rate": 5e-07,
      "loss": 0.6624,
      "step": 4720
    },
    {
      "epoch": 2.0511708586296615,
      "grad_norm": 2.94921875,
      "learning_rate": 5e-07,
      "loss": 0.5848,
      "step": 4730
    },
    {
      "epoch": 2.0555073720728534,
      "grad_norm": 11.875,
      "learning_rate": 5e-07,
      "loss": 0.9366,
      "step": 4740
    },
    {
      "epoch": 2.0598438855160452,
      "grad_norm": 3.880859375,
      "learning_rate": 5e-07,
      "loss": 0.5543,
      "step": 4750
    },
    {
      "epoch": 2.0641803989592367,
      "grad_norm": 4.33984375,
      "learning_rate": 5e-07,
      "loss": 0.3779,
      "step": 4760
    },
    {
      "epoch": 2.0685169124024285,
      "grad_norm": 8.0390625,
      "learning_rate": 5e-07,
      "loss": 0.6945,
      "step": 4770
    },
    {
      "epoch": 2.07285342584562,
      "grad_norm": 2.064453125,
      "learning_rate": 5e-07,
      "loss": 0.6571,
      "step": 4780
    },
    {
      "epoch": 2.077189939288812,
      "grad_norm": 2.3671875,
      "learning_rate": 5e-07,
      "loss": 0.7887,
      "step": 4790
    },
    {
      "epoch": 2.0815264527320037,
      "grad_norm": 5.6796875,
      "learning_rate": 5e-07,
      "loss": 0.5742,
      "step": 4800
    },
    {
      "epoch": 2.085862966175195,
      "grad_norm": 4.6484375,
      "learning_rate": 5e-07,
      "loss": 0.5191,
      "step": 4810
    },
    {
      "epoch": 2.090199479618387,
      "grad_norm": 3.31640625,
      "learning_rate": 5e-07,
      "loss": 0.7182,
      "step": 4820
    },
    {
      "epoch": 2.0945359930615783,
      "grad_norm": 2.5703125,
      "learning_rate": 5e-07,
      "loss": 0.5177,
      "step": 4830
    },
    {
      "epoch": 2.09887250650477,
      "grad_norm": 5.01171875,
      "learning_rate": 5e-07,
      "loss": 0.692,
      "step": 4840
    },
    {
      "epoch": 2.103209019947962,
      "grad_norm": 5.6875,
      "learning_rate": 5e-07,
      "loss": 0.6262,
      "step": 4850
    },
    {
      "epoch": 2.1075455333911535,
      "grad_norm": 6.80859375,
      "learning_rate": 5e-07,
      "loss": 0.684,
      "step": 4860
    },
    {
      "epoch": 2.1118820468343453,
      "grad_norm": 5.609375,
      "learning_rate": 5e-07,
      "loss": 0.5868,
      "step": 4870
    },
    {
      "epoch": 2.1162185602775367,
      "grad_norm": 5.02734375,
      "learning_rate": 5e-07,
      "loss": 0.7363,
      "step": 4880
    },
    {
      "epoch": 2.1205550737207286,
      "grad_norm": 6.2890625,
      "learning_rate": 5e-07,
      "loss": 0.7488,
      "step": 4890
    },
    {
      "epoch": 2.12489158716392,
      "grad_norm": 2.9296875,
      "learning_rate": 5e-07,
      "loss": 0.5751,
      "step": 4900
    },
    {
      "epoch": 2.129228100607112,
      "grad_norm": 6.37890625,
      "learning_rate": 5e-07,
      "loss": 0.6927,
      "step": 4910
    },
    {
      "epoch": 2.1335646140503037,
      "grad_norm": 5.88671875,
      "learning_rate": 5e-07,
      "loss": 0.6593,
      "step": 4920
    },
    {
      "epoch": 2.137901127493495,
      "grad_norm": 6.28515625,
      "learning_rate": 5e-07,
      "loss": 0.5186,
      "step": 4930
    },
    {
      "epoch": 2.142237640936687,
      "grad_norm": 9.3125,
      "learning_rate": 5e-07,
      "loss": 0.7081,
      "step": 4940
    },
    {
      "epoch": 2.1465741543798784,
      "grad_norm": 3.65234375,
      "learning_rate": 5e-07,
      "loss": 0.5419,
      "step": 4950
    },
    {
      "epoch": 2.1509106678230703,
      "grad_norm": 5.65625,
      "learning_rate": 5e-07,
      "loss": 0.8378,
      "step": 4960
    },
    {
      "epoch": 2.1552471812662617,
      "grad_norm": 3.12890625,
      "learning_rate": 5e-07,
      "loss": 0.6194,
      "step": 4970
    },
    {
      "epoch": 2.1595836947094535,
      "grad_norm": 7.55859375,
      "learning_rate": 5e-07,
      "loss": 0.7077,
      "step": 4980
    },
    {
      "epoch": 2.1639202081526454,
      "grad_norm": 59.21875,
      "learning_rate": 5e-07,
      "loss": 1.1875,
      "step": 4990
    },
    {
      "epoch": 2.168256721595837,
      "grad_norm": 6.77734375,
      "learning_rate": 5e-07,
      "loss": 0.6303,
      "step": 5000
    },
    {
      "epoch": 2.168256721595837,
      "eval_loss": 0.23754462599754333,
      "eval_runtime": 355.9274,
      "eval_samples_per_second": 3.225,
      "eval_steps_per_second": 3.225,
      "step": 5000
    },
    {
      "epoch": 2.1725932350390287,
      "grad_norm": 5.87890625,
      "learning_rate": 5e-07,
      "loss": 0.8481,
      "step": 5010
    },
    {
      "epoch": 2.17692974848222,
      "grad_norm": 2.375,
      "learning_rate": 5e-07,
      "loss": 0.4243,
      "step": 5020
    },
    {
      "epoch": 2.181266261925412,
      "grad_norm": 3.126953125,
      "learning_rate": 5e-07,
      "loss": 0.7743,
      "step": 5030
    },
    {
      "epoch": 2.185602775368604,
      "grad_norm": 2.826171875,
      "learning_rate": 5e-07,
      "loss": 0.7762,
      "step": 5040
    },
    {
      "epoch": 2.1899392888117952,
      "grad_norm": 5.55859375,
      "learning_rate": 5e-07,
      "loss": 0.9882,
      "step": 5050
    },
    {
      "epoch": 2.194275802254987,
      "grad_norm": 8.375,
      "learning_rate": 5e-07,
      "loss": 0.8427,
      "step": 5060
    },
    {
      "epoch": 2.1986123156981785,
      "grad_norm": 5.703125,
      "learning_rate": 5e-07,
      "loss": 0.729,
      "step": 5070
    },
    {
      "epoch": 2.2029488291413704,
      "grad_norm": 5.39453125,
      "learning_rate": 5e-07,
      "loss": 0.6213,
      "step": 5080
    },
    {
      "epoch": 2.207285342584562,
      "grad_norm": 7.0625,
      "learning_rate": 5e-07,
      "loss": 0.5743,
      "step": 5090
    },
    {
      "epoch": 2.2116218560277536,
      "grad_norm": 7.20703125,
      "learning_rate": 5e-07,
      "loss": 0.5636,
      "step": 5100
    },
    {
      "epoch": 2.2159583694709455,
      "grad_norm": 3.630859375,
      "learning_rate": 5e-07,
      "loss": 0.5747,
      "step": 5110
    },
    {
      "epoch": 2.220294882914137,
      "grad_norm": 3.55859375,
      "learning_rate": 5e-07,
      "loss": 0.7615,
      "step": 5120
    },
    {
      "epoch": 2.2246313963573288,
      "grad_norm": 5.01953125,
      "learning_rate": 5e-07,
      "loss": 0.9233,
      "step": 5130
    },
    {
      "epoch": 2.22896790980052,
      "grad_norm": 3.099609375,
      "learning_rate": 5e-07,
      "loss": 0.7782,
      "step": 5140
    },
    {
      "epoch": 2.233304423243712,
      "grad_norm": 3.3359375,
      "learning_rate": 5e-07,
      "loss": 0.5533,
      "step": 5150
    },
    {
      "epoch": 2.237640936686904,
      "grad_norm": 4.1875,
      "learning_rate": 5e-07,
      "loss": 1.211,
      "step": 5160
    },
    {
      "epoch": 2.2419774501300953,
      "grad_norm": 2.677734375,
      "learning_rate": 5e-07,
      "loss": 0.7955,
      "step": 5170
    },
    {
      "epoch": 2.246313963573287,
      "grad_norm": 6.0859375,
      "learning_rate": 5e-07,
      "loss": 0.5952,
      "step": 5180
    },
    {
      "epoch": 2.2506504770164786,
      "grad_norm": 4.32421875,
      "learning_rate": 5e-07,
      "loss": 1.1031,
      "step": 5190
    },
    {
      "epoch": 2.2549869904596704,
      "grad_norm": 4.41796875,
      "learning_rate": 5e-07,
      "loss": 0.7537,
      "step": 5200
    },
    {
      "epoch": 2.259323503902862,
      "grad_norm": 5.484375,
      "learning_rate": 5e-07,
      "loss": 0.806,
      "step": 5210
    },
    {
      "epoch": 2.2636600173460537,
      "grad_norm": 10.2109375,
      "learning_rate": 5e-07,
      "loss": 0.9878,
      "step": 5220
    },
    {
      "epoch": 2.2679965307892456,
      "grad_norm": 4.6328125,
      "learning_rate": 5e-07,
      "loss": 0.8093,
      "step": 5230
    },
    {
      "epoch": 2.272333044232437,
      "grad_norm": 5.51953125,
      "learning_rate": 5e-07,
      "loss": 0.7418,
      "step": 5240
    },
    {
      "epoch": 2.276669557675629,
      "grad_norm": 6.67578125,
      "learning_rate": 5e-07,
      "loss": 0.67,
      "step": 5250
    },
    {
      "epoch": 2.2810060711188207,
      "grad_norm": 5.2421875,
      "learning_rate": 5e-07,
      "loss": 0.7535,
      "step": 5260
    },
    {
      "epoch": 2.285342584562012,
      "grad_norm": 3.99609375,
      "learning_rate": 5e-07,
      "loss": 0.7558,
      "step": 5270
    },
    {
      "epoch": 2.289679098005204,
      "grad_norm": 4.0078125,
      "learning_rate": 5e-07,
      "loss": 0.5411,
      "step": 5280
    },
    {
      "epoch": 2.2940156114483954,
      "grad_norm": 4.6328125,
      "learning_rate": 5e-07,
      "loss": 0.4449,
      "step": 5290
    },
    {
      "epoch": 2.2983521248915872,
      "grad_norm": 11.046875,
      "learning_rate": 5e-07,
      "loss": 0.7707,
      "step": 5300
    },
    {
      "epoch": 2.3026886383347787,
      "grad_norm": 7.91015625,
      "learning_rate": 5e-07,
      "loss": 0.8648,
      "step": 5310
    },
    {
      "epoch": 2.3070251517779705,
      "grad_norm": 6.9921875,
      "learning_rate": 5e-07,
      "loss": 0.5995,
      "step": 5320
    },
    {
      "epoch": 2.3113616652211624,
      "grad_norm": 4.66796875,
      "learning_rate": 5e-07,
      "loss": 0.8122,
      "step": 5330
    },
    {
      "epoch": 2.315698178664354,
      "grad_norm": 6.625,
      "learning_rate": 5e-07,
      "loss": 0.771,
      "step": 5340
    },
    {
      "epoch": 2.3200346921075456,
      "grad_norm": 3.4609375,
      "learning_rate": 5e-07,
      "loss": 0.539,
      "step": 5350
    },
    {
      "epoch": 2.324371205550737,
      "grad_norm": 6.046875,
      "learning_rate": 5e-07,
      "loss": 0.491,
      "step": 5360
    },
    {
      "epoch": 2.328707718993929,
      "grad_norm": 33.9375,
      "learning_rate": 5e-07,
      "loss": 0.537,
      "step": 5370
    },
    {
      "epoch": 2.3330442324371203,
      "grad_norm": 8.5078125,
      "learning_rate": 5e-07,
      "loss": 0.5656,
      "step": 5380
    },
    {
      "epoch": 2.337380745880312,
      "grad_norm": 7.81640625,
      "learning_rate": 5e-07,
      "loss": 0.4807,
      "step": 5390
    },
    {
      "epoch": 2.341717259323504,
      "grad_norm": 3.3828125,
      "learning_rate": 5e-07,
      "loss": 0.5002,
      "step": 5400
    },
    {
      "epoch": 2.3460537727666955,
      "grad_norm": 6.99609375,
      "learning_rate": 5e-07,
      "loss": 0.6247,
      "step": 5410
    },
    {
      "epoch": 2.3503902862098873,
      "grad_norm": 6.67578125,
      "learning_rate": 5e-07,
      "loss": 0.6208,
      "step": 5420
    },
    {
      "epoch": 2.3547267996530787,
      "grad_norm": 6.57421875,
      "learning_rate": 5e-07,
      "loss": 0.5956,
      "step": 5430
    },
    {
      "epoch": 2.3590633130962706,
      "grad_norm": 4.390625,
      "learning_rate": 5e-07,
      "loss": 0.632,
      "step": 5440
    },
    {
      "epoch": 2.3633998265394625,
      "grad_norm": 5.5078125,
      "learning_rate": 5e-07,
      "loss": 0.6723,
      "step": 5450
    },
    {
      "epoch": 2.367736339982654,
      "grad_norm": 3.119140625,
      "learning_rate": 5e-07,
      "loss": 0.5895,
      "step": 5460
    },
    {
      "epoch": 2.3720728534258457,
      "grad_norm": 1.865234375,
      "learning_rate": 5e-07,
      "loss": 0.6393,
      "step": 5470
    },
    {
      "epoch": 2.376409366869037,
      "grad_norm": 6.12890625,
      "learning_rate": 5e-07,
      "loss": 0.538,
      "step": 5480
    },
    {
      "epoch": 2.380745880312229,
      "grad_norm": 12.890625,
      "learning_rate": 5e-07,
      "loss": 0.8481,
      "step": 5490
    },
    {
      "epoch": 2.385082393755421,
      "grad_norm": 13.2578125,
      "learning_rate": 5e-07,
      "loss": 0.7316,
      "step": 5500
    },
    {
      "epoch": 2.385082393755421,
      "eval_loss": 0.2347976714372635,
      "eval_runtime": 355.7511,
      "eval_samples_per_second": 3.227,
      "eval_steps_per_second": 3.227,
      "step": 5500
    },
    {
      "epoch": 2.3894189071986123,
      "grad_norm": 2.291015625,
      "learning_rate": 5e-07,
      "loss": 0.6471,
      "step": 5510
    },
    {
      "epoch": 2.393755420641804,
      "grad_norm": 6.1171875,
      "learning_rate": 5e-07,
      "loss": 0.5113,
      "step": 5520
    },
    {
      "epoch": 2.3980919340849955,
      "grad_norm": 5.3046875,
      "learning_rate": 5e-07,
      "loss": 1.2012,
      "step": 5530
    },
    {
      "epoch": 2.4024284475281874,
      "grad_norm": 6.234375,
      "learning_rate": 5e-07,
      "loss": 1.0265,
      "step": 5540
    },
    {
      "epoch": 2.406764960971379,
      "grad_norm": 2.275390625,
      "learning_rate": 5e-07,
      "loss": 0.5844,
      "step": 5550
    },
    {
      "epoch": 2.4111014744145707,
      "grad_norm": 2.5546875,
      "learning_rate": 5e-07,
      "loss": 0.5607,
      "step": 5560
    },
    {
      "epoch": 2.4154379878577625,
      "grad_norm": 2.28125,
      "learning_rate": 5e-07,
      "loss": 0.5948,
      "step": 5570
    },
    {
      "epoch": 2.419774501300954,
      "grad_norm": 5.1796875,
      "learning_rate": 5e-07,
      "loss": 0.5848,
      "step": 5580
    },
    {
      "epoch": 2.424111014744146,
      "grad_norm": 5.46484375,
      "learning_rate": 5e-07,
      "loss": 0.4521,
      "step": 5590
    },
    {
      "epoch": 2.4284475281873372,
      "grad_norm": 6.046875,
      "learning_rate": 5e-07,
      "loss": 0.7646,
      "step": 5600
    },
    {
      "epoch": 2.432784041630529,
      "grad_norm": 5.48828125,
      "learning_rate": 5e-07,
      "loss": 0.653,
      "step": 5610
    },
    {
      "epoch": 2.4371205550737205,
      "grad_norm": 5.27734375,
      "learning_rate": 5e-07,
      "loss": 0.5144,
      "step": 5620
    },
    {
      "epoch": 2.4414570685169124,
      "grad_norm": 0.81298828125,
      "learning_rate": 5e-07,
      "loss": 0.4716,
      "step": 5630
    },
    {
      "epoch": 2.445793581960104,
      "grad_norm": 4.046875,
      "learning_rate": 5e-07,
      "loss": 0.7314,
      "step": 5640
    },
    {
      "epoch": 2.4501300954032956,
      "grad_norm": 6.53515625,
      "learning_rate": 5e-07,
      "loss": 0.8207,
      "step": 5650
    },
    {
      "epoch": 2.4544666088464875,
      "grad_norm": 4.37109375,
      "learning_rate": 5e-07,
      "loss": 0.8742,
      "step": 5660
    },
    {
      "epoch": 2.4588031222896793,
      "grad_norm": 5.65625,
      "learning_rate": 5e-07,
      "loss": 0.6526,
      "step": 5670
    },
    {
      "epoch": 2.4631396357328708,
      "grad_norm": 3.833984375,
      "learning_rate": 5e-07,
      "loss": 0.5635,
      "step": 5680
    },
    {
      "epoch": 2.4674761491760626,
      "grad_norm": 3.97265625,
      "learning_rate": 5e-07,
      "loss": 0.5125,
      "step": 5690
    },
    {
      "epoch": 2.471812662619254,
      "grad_norm": 3.69140625,
      "learning_rate": 5e-07,
      "loss": 0.726,
      "step": 5700
    },
    {
      "epoch": 2.476149176062446,
      "grad_norm": 7.29296875,
      "learning_rate": 5e-07,
      "loss": 0.9367,
      "step": 5710
    },
    {
      "epoch": 2.4804856895056373,
      "grad_norm": 7.7734375,
      "learning_rate": 5e-07,
      "loss": 0.7098,
      "step": 5720
    },
    {
      "epoch": 2.484822202948829,
      "grad_norm": 2.861328125,
      "learning_rate": 5e-07,
      "loss": 0.5552,
      "step": 5730
    },
    {
      "epoch": 2.489158716392021,
      "grad_norm": 10.75,
      "learning_rate": 5e-07,
      "loss": 0.83,
      "step": 5740
    },
    {
      "epoch": 2.4934952298352124,
      "grad_norm": 3.9296875,
      "learning_rate": 5e-07,
      "loss": 0.7866,
      "step": 5750
    },
    {
      "epoch": 2.4978317432784043,
      "grad_norm": 4.9453125,
      "learning_rate": 5e-07,
      "loss": 0.7947,
      "step": 5760
    },
    {
      "epoch": 2.5021682567215957,
      "grad_norm": 5.22265625,
      "learning_rate": 5e-07,
      "loss": 0.5734,
      "step": 5770
    },
    {
      "epoch": 2.5065047701647876,
      "grad_norm": 4.52734375,
      "learning_rate": 5e-07,
      "loss": 0.4701,
      "step": 5780
    },
    {
      "epoch": 2.510841283607979,
      "grad_norm": 7.2734375,
      "learning_rate": 5e-07,
      "loss": 0.5661,
      "step": 5790
    },
    {
      "epoch": 2.515177797051171,
      "grad_norm": 7.5390625,
      "learning_rate": 5e-07,
      "loss": 0.423,
      "step": 5800
    },
    {
      "epoch": 2.5195143104943627,
      "grad_norm": 9.703125,
      "learning_rate": 5e-07,
      "loss": 0.482,
      "step": 5810
    },
    {
      "epoch": 2.523850823937554,
      "grad_norm": 6.80078125,
      "learning_rate": 5e-07,
      "loss": 0.5346,
      "step": 5820
    },
    {
      "epoch": 2.528187337380746,
      "grad_norm": 4.2578125,
      "learning_rate": 5e-07,
      "loss": 0.4976,
      "step": 5830
    },
    {
      "epoch": 2.532523850823938,
      "grad_norm": 6.35546875,
      "learning_rate": 5e-07,
      "loss": 0.6387,
      "step": 5840
    },
    {
      "epoch": 2.5368603642671292,
      "grad_norm": 2.625,
      "learning_rate": 5e-07,
      "loss": 0.534,
      "step": 5850
    },
    {
      "epoch": 2.5411968777103207,
      "grad_norm": 6.40625,
      "learning_rate": 5e-07,
      "loss": 0.4314,
      "step": 5860
    },
    {
      "epoch": 2.5455333911535125,
      "grad_norm": 4.17578125,
      "learning_rate": 5e-07,
      "loss": 0.4244,
      "step": 5870
    },
    {
      "epoch": 2.5498699045967044,
      "grad_norm": 16.0,
      "learning_rate": 5e-07,
      "loss": 0.5925,
      "step": 5880
    },
    {
      "epoch": 2.554206418039896,
      "grad_norm": 8.25,
      "learning_rate": 5e-07,
      "loss": 0.4992,
      "step": 5890
    },
    {
      "epoch": 2.5585429314830876,
      "grad_norm": 3.38671875,
      "learning_rate": 5e-07,
      "loss": 0.4373,
      "step": 5900
    },
    {
      "epoch": 2.5628794449262795,
      "grad_norm": 4.98046875,
      "learning_rate": 5e-07,
      "loss": 0.5983,
      "step": 5910
    },
    {
      "epoch": 2.567215958369471,
      "grad_norm": 2.572265625,
      "learning_rate": 5e-07,
      "loss": 0.7139,
      "step": 5920
    },
    {
      "epoch": 2.5715524718126628,
      "grad_norm": 4.2265625,
      "learning_rate": 5e-07,
      "loss": 0.6596,
      "step": 5930
    },
    {
      "epoch": 2.575888985255854,
      "grad_norm": 5.4609375,
      "learning_rate": 5e-07,
      "loss": 0.6405,
      "step": 5940
    },
    {
      "epoch": 2.580225498699046,
      "grad_norm": 13.40625,
      "learning_rate": 5e-07,
      "loss": 0.7878,
      "step": 5950
    },
    {
      "epoch": 2.5845620121422375,
      "grad_norm": 6.4140625,
      "learning_rate": 5e-07,
      "loss": 0.6556,
      "step": 5960
    },
    {
      "epoch": 2.5888985255854293,
      "grad_norm": 2.671875,
      "learning_rate": 5e-07,
      "loss": 0.53,
      "step": 5970
    },
    {
      "epoch": 2.593235039028621,
      "grad_norm": 2.458984375,
      "learning_rate": 5e-07,
      "loss": 0.3987,
      "step": 5980
    },
    {
      "epoch": 2.5975715524718126,
      "grad_norm": 4.78125,
      "learning_rate": 5e-07,
      "loss": 0.8542,
      "step": 5990
    },
    {
      "epoch": 2.6019080659150045,
      "grad_norm": 3.7890625,
      "learning_rate": 5e-07,
      "loss": 0.4588,
      "step": 6000
    },
    {
      "epoch": 2.6019080659150045,
      "eval_loss": 0.24308811128139496,
      "eval_runtime": 358.9752,
      "eval_samples_per_second": 3.198,
      "eval_steps_per_second": 3.198,
      "step": 6000
    },
    {
      "epoch": 2.606244579358196,
      "grad_norm": 8.109375,
      "learning_rate": 5e-07,
      "loss": 0.619,
      "step": 6010
    },
    {
      "epoch": 2.6105810928013877,
      "grad_norm": 5.0390625,
      "learning_rate": 5e-07,
      "loss": 0.6466,
      "step": 6020
    },
    {
      "epoch": 2.614917606244579,
      "grad_norm": 7.73828125,
      "learning_rate": 5e-07,
      "loss": 0.6927,
      "step": 6030
    },
    {
      "epoch": 2.619254119687771,
      "grad_norm": 6.26171875,
      "learning_rate": 5e-07,
      "loss": 0.436,
      "step": 6040
    },
    {
      "epoch": 2.623590633130963,
      "grad_norm": 3.158203125,
      "learning_rate": 5e-07,
      "loss": 0.4832,
      "step": 6050
    },
    {
      "epoch": 2.6279271465741543,
      "grad_norm": 1.990234375,
      "learning_rate": 5e-07,
      "loss": 0.4336,
      "step": 6060
    },
    {
      "epoch": 2.632263660017346,
      "grad_norm": 3.455078125,
      "learning_rate": 5e-07,
      "loss": 0.8438,
      "step": 6070
    },
    {
      "epoch": 2.636600173460538,
      "grad_norm": 3.828125,
      "learning_rate": 5e-07,
      "loss": 0.488,
      "step": 6080
    },
    {
      "epoch": 2.6409366869037294,
      "grad_norm": 10.140625,
      "learning_rate": 5e-07,
      "loss": 0.4691,
      "step": 6090
    },
    {
      "epoch": 2.645273200346921,
      "grad_norm": 6.46484375,
      "learning_rate": 5e-07,
      "loss": 0.9632,
      "step": 6100
    },
    {
      "epoch": 2.6496097137901127,
      "grad_norm": 5.5703125,
      "learning_rate": 5e-07,
      "loss": 0.4339,
      "step": 6110
    },
    {
      "epoch": 2.6539462272333045,
      "grad_norm": 4.2421875,
      "learning_rate": 5e-07,
      "loss": 0.4046,
      "step": 6120
    },
    {
      "epoch": 2.658282740676496,
      "grad_norm": 5.37890625,
      "learning_rate": 5e-07,
      "loss": 0.55,
      "step": 6130
    },
    {
      "epoch": 2.662619254119688,
      "grad_norm": 3.14453125,
      "learning_rate": 5e-07,
      "loss": 0.5741,
      "step": 6140
    },
    {
      "epoch": 2.6669557675628797,
      "grad_norm": 12.15625,
      "learning_rate": 5e-07,
      "loss": 0.9115,
      "step": 6150
    },
    {
      "epoch": 2.671292281006071,
      "grad_norm": 3.884765625,
      "learning_rate": 5e-07,
      "loss": 0.4068,
      "step": 6160
    },
    {
      "epoch": 2.675628794449263,
      "grad_norm": 4.6640625,
      "learning_rate": 5e-07,
      "loss": 0.4277,
      "step": 6170
    },
    {
      "epoch": 2.6799653078924544,
      "grad_norm": 4.80859375,
      "learning_rate": 5e-07,
      "loss": 0.4792,
      "step": 6180
    },
    {
      "epoch": 2.684301821335646,
      "grad_norm": 3.59375,
      "learning_rate": 5e-07,
      "loss": 0.6432,
      "step": 6190
    },
    {
      "epoch": 2.6886383347788376,
      "grad_norm": 8.5,
      "learning_rate": 5e-07,
      "loss": 0.4541,
      "step": 6200
    },
    {
      "epoch": 2.6929748482220295,
      "grad_norm": 3.353515625,
      "learning_rate": 5e-07,
      "loss": 0.6624,
      "step": 6210
    },
    {
      "epoch": 2.6973113616652213,
      "grad_norm": 3.236328125,
      "learning_rate": 5e-07,
      "loss": 0.7516,
      "step": 6220
    },
    {
      "epoch": 2.7016478751084128,
      "grad_norm": 5.4140625,
      "learning_rate": 5e-07,
      "loss": 0.7526,
      "step": 6230
    },
    {
      "epoch": 2.7059843885516046,
      "grad_norm": 7.0859375,
      "learning_rate": 5e-07,
      "loss": 0.4969,
      "step": 6240
    },
    {
      "epoch": 2.710320901994796,
      "grad_norm": 3.01953125,
      "learning_rate": 5e-07,
      "loss": 0.5736,
      "step": 6250
    },
    {
      "epoch": 2.714657415437988,
      "grad_norm": 5.2265625,
      "learning_rate": 5e-07,
      "loss": 0.5975,
      "step": 6260
    },
    {
      "epoch": 2.7189939288811793,
      "grad_norm": 4.87109375,
      "learning_rate": 5e-07,
      "loss": 0.5656,
      "step": 6270
    },
    {
      "epoch": 2.723330442324371,
      "grad_norm": 10.890625,
      "learning_rate": 5e-07,
      "loss": 0.6976,
      "step": 6280
    },
    {
      "epoch": 2.727666955767563,
      "grad_norm": 4.15625,
      "learning_rate": 5e-07,
      "loss": 0.5417,
      "step": 6290
    },
    {
      "epoch": 2.7320034692107544,
      "grad_norm": 4.12109375,
      "learning_rate": 5e-07,
      "loss": 0.4496,
      "step": 6300
    },
    {
      "epoch": 2.7363399826539463,
      "grad_norm": 3.986328125,
      "learning_rate": 5e-07,
      "loss": 0.4435,
      "step": 6310
    },
    {
      "epoch": 2.740676496097138,
      "grad_norm": 3.19921875,
      "learning_rate": 5e-07,
      "loss": 0.4794,
      "step": 6320
    },
    {
      "epoch": 2.7450130095403296,
      "grad_norm": 6.41796875,
      "learning_rate": 5e-07,
      "loss": 0.8484,
      "step": 6330
    },
    {
      "epoch": 2.749349522983521,
      "grad_norm": 3.611328125,
      "learning_rate": 5e-07,
      "loss": 0.5335,
      "step": 6340
    },
    {
      "epoch": 2.753686036426713,
      "grad_norm": 2.078125,
      "learning_rate": 5e-07,
      "loss": 0.645,
      "step": 6350
    },
    {
      "epoch": 2.7580225498699047,
      "grad_norm": 6.62109375,
      "learning_rate": 5e-07,
      "loss": 0.4124,
      "step": 6360
    },
    {
      "epoch": 2.762359063313096,
      "grad_norm": 2.4921875,
      "learning_rate": 5e-07,
      "loss": 0.3303,
      "step": 6370
    },
    {
      "epoch": 2.766695576756288,
      "grad_norm": 6.05859375,
      "learning_rate": 5e-07,
      "loss": 0.5787,
      "step": 6380
    },
    {
      "epoch": 2.77103209019948,
      "grad_norm": 3.328125,
      "learning_rate": 5e-07,
      "loss": 0.4759,
      "step": 6390
    },
    {
      "epoch": 2.7753686036426712,
      "grad_norm": 6.46875,
      "learning_rate": 5e-07,
      "loss": 0.9266,
      "step": 6400
    },
    {
      "epoch": 2.779705117085863,
      "grad_norm": 6.95703125,
      "learning_rate": 5e-07,
      "loss": 0.7322,
      "step": 6410
    },
    {
      "epoch": 2.7840416305290545,
      "grad_norm": 9.703125,
      "learning_rate": 5e-07,
      "loss": 0.4155,
      "step": 6420
    },
    {
      "epoch": 2.7883781439722464,
      "grad_norm": 5.01171875,
      "learning_rate": 5e-07,
      "loss": 0.4258,
      "step": 6430
    },
    {
      "epoch": 2.792714657415438,
      "grad_norm": 10.5234375,
      "learning_rate": 5e-07,
      "loss": 0.5638,
      "step": 6440
    },
    {
      "epoch": 2.7970511708586296,
      "grad_norm": 1.9140625,
      "learning_rate": 5e-07,
      "loss": 0.4245,
      "step": 6450
    },
    {
      "epoch": 2.8013876843018215,
      "grad_norm": 6.1171875,
      "learning_rate": 5e-07,
      "loss": 0.6251,
      "step": 6460
    },
    {
      "epoch": 2.805724197745013,
      "grad_norm": 12.859375,
      "learning_rate": 5e-07,
      "loss": 0.5498,
      "step": 6470
    },
    {
      "epoch": 2.8100607111882048,
      "grad_norm": 15.3203125,
      "learning_rate": 5e-07,
      "loss": 0.5401,
      "step": 6480
    },
    {
      "epoch": 2.8143972246313966,
      "grad_norm": 4.22265625,
      "learning_rate": 5e-07,
      "loss": 0.621,
      "step": 6490
    },
    {
      "epoch": 2.818733738074588,
      "grad_norm": 5.21484375,
      "learning_rate": 5e-07,
      "loss": 0.4425,
      "step": 6500
    },
    {
      "epoch": 2.818733738074588,
      "eval_loss": 0.23988975584506989,
      "eval_runtime": 358.0686,
      "eval_samples_per_second": 3.206,
      "eval_steps_per_second": 3.206,
      "step": 6500
    },
    {
      "epoch": 2.8230702515177795,
      "grad_norm": 2.630859375,
      "learning_rate": 2e-07,
      "loss": 0.5771,
      "step": 6510
    },
    {
      "epoch": 2.8274067649609713,
      "grad_norm": 3.736328125,
      "learning_rate": 2e-07,
      "loss": 0.6044,
      "step": 6520
    },
    {
      "epoch": 2.831743278404163,
      "grad_norm": 1.0732421875,
      "learning_rate": 2e-07,
      "loss": 0.6372,
      "step": 6530
    },
    {
      "epoch": 2.8360797918473546,
      "grad_norm": 6.96875,
      "learning_rate": 2e-07,
      "loss": 0.5307,
      "step": 6540
    },
    {
      "epoch": 2.8404163052905465,
      "grad_norm": 4.5078125,
      "learning_rate": 2e-07,
      "loss": 0.4742,
      "step": 6550
    },
    {
      "epoch": 2.8447528187337383,
      "grad_norm": 7.68359375,
      "learning_rate": 2e-07,
      "loss": 0.5437,
      "step": 6560
    },
    {
      "epoch": 2.8490893321769297,
      "grad_norm": 4.55859375,
      "learning_rate": 2e-07,
      "loss": 0.4077,
      "step": 6570
    },
    {
      "epoch": 2.8534258456201216,
      "grad_norm": 3.0625,
      "learning_rate": 2e-07,
      "loss": 0.4372,
      "step": 6580
    },
    {
      "epoch": 2.857762359063313,
      "grad_norm": 5.2109375,
      "learning_rate": 2e-07,
      "loss": 0.539,
      "step": 6590
    },
    {
      "epoch": 2.862098872506505,
      "grad_norm": 13.59375,
      "learning_rate": 2e-07,
      "loss": 0.5104,
      "step": 6600
    },
    {
      "epoch": 2.8664353859496963,
      "grad_norm": 11.3671875,
      "learning_rate": 2e-07,
      "loss": 0.6664,
      "step": 6610
    },
    {
      "epoch": 2.870771899392888,
      "grad_norm": 2.939453125,
      "learning_rate": 2e-07,
      "loss": 0.864,
      "step": 6620
    },
    {
      "epoch": 2.87510841283608,
      "grad_norm": 2.765625,
      "learning_rate": 2e-07,
      "loss": 0.6108,
      "step": 6630
    },
    {
      "epoch": 2.8794449262792714,
      "grad_norm": 8.0078125,
      "learning_rate": 2e-07,
      "loss": 0.5593,
      "step": 6640
    },
    {
      "epoch": 2.8837814397224633,
      "grad_norm": 6.0078125,
      "learning_rate": 2e-07,
      "loss": 0.4876,
      "step": 6650
    },
    {
      "epoch": 2.8881179531656547,
      "grad_norm": 4.52734375,
      "learning_rate": 2e-07,
      "loss": 0.6114,
      "step": 6660
    },
    {
      "epoch": 2.8924544666088465,
      "grad_norm": 5.15234375,
      "learning_rate": 2e-07,
      "loss": 0.6783,
      "step": 6670
    },
    {
      "epoch": 2.896790980052038,
      "grad_norm": 5.3046875,
      "learning_rate": 2e-07,
      "loss": 0.3414,
      "step": 6680
    },
    {
      "epoch": 2.90112749349523,
      "grad_norm": 2.97265625,
      "learning_rate": 2e-07,
      "loss": 0.3697,
      "step": 6690
    },
    {
      "epoch": 2.9054640069384217,
      "grad_norm": 5.43359375,
      "learning_rate": 2e-07,
      "loss": 0.5389,
      "step": 6700
    },
    {
      "epoch": 2.909800520381613,
      "grad_norm": 6.79296875,
      "learning_rate": 2e-07,
      "loss": 0.4314,
      "step": 6710
    },
    {
      "epoch": 2.914137033824805,
      "grad_norm": 6.23828125,
      "learning_rate": 2e-07,
      "loss": 0.6803,
      "step": 6720
    },
    {
      "epoch": 2.918473547267997,
      "grad_norm": 12.578125,
      "learning_rate": 2e-07,
      "loss": 0.572,
      "step": 6730
    },
    {
      "epoch": 2.922810060711188,
      "grad_norm": 1.78515625,
      "learning_rate": 2e-07,
      "loss": 0.845,
      "step": 6740
    },
    {
      "epoch": 2.9271465741543796,
      "grad_norm": 2.580078125,
      "learning_rate": 2e-07,
      "loss": 0.3181,
      "step": 6750
    },
    {
      "epoch": 2.9314830875975715,
      "grad_norm": 3.716796875,
      "learning_rate": 2e-07,
      "loss": 0.6249,
      "step": 6760
    },
    {
      "epoch": 2.9358196010407633,
      "grad_norm": 3.58203125,
      "learning_rate": 2e-07,
      "loss": 0.3803,
      "step": 6770
    },
    {
      "epoch": 2.9401561144839548,
      "grad_norm": 2.794921875,
      "learning_rate": 2e-07,
      "loss": 0.6771,
      "step": 6780
    },
    {
      "epoch": 2.9444926279271466,
      "grad_norm": 5.37109375,
      "learning_rate": 2e-07,
      "loss": 0.6627,
      "step": 6790
    },
    {
      "epoch": 2.9488291413703385,
      "grad_norm": 4.640625,
      "learning_rate": 2e-07,
      "loss": 0.5273,
      "step": 6800
    },
    {
      "epoch": 2.95316565481353,
      "grad_norm": 1.921875,
      "learning_rate": 2e-07,
      "loss": 0.5487,
      "step": 6810
    },
    {
      "epoch": 2.9575021682567217,
      "grad_norm": 2.58984375,
      "learning_rate": 2e-07,
      "loss": 0.5164,
      "step": 6820
    },
    {
      "epoch": 2.961838681699913,
      "grad_norm": 3.876953125,
      "learning_rate": 2e-07,
      "loss": 0.5548,
      "step": 6830
    },
    {
      "epoch": 2.966175195143105,
      "grad_norm": 6.5625,
      "learning_rate": 2e-07,
      "loss": 0.5215,
      "step": 6840
    },
    {
      "epoch": 2.9705117085862964,
      "grad_norm": 5.1875,
      "learning_rate": 2e-07,
      "loss": 0.5623,
      "step": 6850
    },
    {
      "epoch": 2.9748482220294883,
      "grad_norm": 6.54296875,
      "learning_rate": 2e-07,
      "loss": 0.4787,
      "step": 6860
    },
    {
      "epoch": 2.97918473547268,
      "grad_norm": 9.7578125,
      "learning_rate": 2e-07,
      "loss": 0.5124,
      "step": 6870
    },
    {
      "epoch": 2.9835212489158716,
      "grad_norm": 5.59375,
      "learning_rate": 2e-07,
      "loss": 0.4726,
      "step": 6880
    },
    {
      "epoch": 2.9878577623590634,
      "grad_norm": 5.2265625,
      "learning_rate": 2e-07,
      "loss": 0.4183,
      "step": 6890
    },
    {
      "epoch": 2.992194275802255,
      "grad_norm": 1.85546875,
      "learning_rate": 2e-07,
      "loss": 0.4258,
      "step": 6900
    },
    {
      "epoch": 2.9965307892454467,
      "grad_norm": 3.8046875,
      "learning_rate": 2e-07,
      "loss": 0.4386,
      "step": 6910
    },
    {
      "epoch": 3.0008673026886385,
      "grad_norm": 6.58203125,
      "learning_rate": 2e-07,
      "loss": 0.4365,
      "step": 6920
    },
    {
      "epoch": 3.00520381613183,
      "grad_norm": 12.7421875,
      "learning_rate": 2e-07,
      "loss": 0.39,
      "step": 6930
    },
    {
      "epoch": 3.009540329575022,
      "grad_norm": 6.8125,
      "learning_rate": 2e-07,
      "loss": 0.3119,
      "step": 6940
    },
    {
      "epoch": 3.0138768430182132,
      "grad_norm": 3.3203125,
      "learning_rate": 2e-07,
      "loss": 0.365,
      "step": 6950
    },
    {
      "epoch": 3.018213356461405,
      "grad_norm": 1.4482421875,
      "learning_rate": 2e-07,
      "loss": 0.4498,
      "step": 6960
    },
    {
      "epoch": 3.0225498699045965,
      "grad_norm": 2.93359375,
      "learning_rate": 2e-07,
      "loss": 0.7906,
      "step": 6970
    },
    {
      "epoch": 3.0268863833477884,
      "grad_norm": 3.09375,
      "learning_rate": 2e-07,
      "loss": 0.3584,
      "step": 6980
    },
    {
      "epoch": 3.0312228967909802,
      "grad_norm": 2.9921875,
      "learning_rate": 2e-07,
      "loss": 0.3418,
      "step": 6990
    },
    {
      "epoch": 3.0355594102341716,
      "grad_norm": 2.271484375,
      "learning_rate": 2e-07,
      "loss": 0.3432,
      "step": 7000
    },
    {
      "epoch": 3.0355594102341716,
      "eval_loss": 0.21796274185180664,
      "eval_runtime": 362.0862,
      "eval_samples_per_second": 3.171,
      "eval_steps_per_second": 3.171,
      "step": 7000
    },
    {
      "epoch": 3.0398959236773635,
      "grad_norm": 4.87890625,
      "learning_rate": 2e-07,
      "loss": 0.5921,
      "step": 7010
    },
    {
      "epoch": 3.044232437120555,
      "grad_norm": 3.90625,
      "learning_rate": 2e-07,
      "loss": 0.557,
      "step": 7020
    },
    {
      "epoch": 3.0485689505637468,
      "grad_norm": 5.6640625,
      "learning_rate": 2e-07,
      "loss": 0.217,
      "step": 7030
    },
    {
      "epoch": 3.0529054640069386,
      "grad_norm": 1.6357421875,
      "learning_rate": 2e-07,
      "loss": 0.7889,
      "step": 7040
    },
    {
      "epoch": 3.05724197745013,
      "grad_norm": 5.26171875,
      "learning_rate": 2e-07,
      "loss": 0.4197,
      "step": 7050
    },
    {
      "epoch": 3.061578490893322,
      "grad_norm": 1.45703125,
      "learning_rate": 2e-07,
      "loss": 0.4513,
      "step": 7060
    },
    {
      "epoch": 3.0659150043365133,
      "grad_norm": 3.951171875,
      "learning_rate": 2e-07,
      "loss": 0.3803,
      "step": 7070
    },
    {
      "epoch": 3.070251517779705,
      "grad_norm": 2.982421875,
      "learning_rate": 2e-07,
      "loss": 0.4552,
      "step": 7080
    },
    {
      "epoch": 3.0745880312228966,
      "grad_norm": 4.66015625,
      "learning_rate": 2e-07,
      "loss": 0.4868,
      "step": 7090
    },
    {
      "epoch": 3.0789245446660884,
      "grad_norm": 2.876953125,
      "learning_rate": 2e-07,
      "loss": 0.4554,
      "step": 7100
    },
    {
      "epoch": 3.0832610581092803,
      "grad_norm": 6.75,
      "learning_rate": 2e-07,
      "loss": 0.2638,
      "step": 7110
    },
    {
      "epoch": 3.0875975715524717,
      "grad_norm": 4.25390625,
      "learning_rate": 2e-07,
      "loss": 0.3427,
      "step": 7120
    },
    {
      "epoch": 3.0919340849956636,
      "grad_norm": 2.958984375,
      "learning_rate": 2e-07,
      "loss": 0.3981,
      "step": 7130
    },
    {
      "epoch": 3.096270598438855,
      "grad_norm": 2.365234375,
      "learning_rate": 2e-07,
      "loss": 0.6161,
      "step": 7140
    },
    {
      "epoch": 3.100607111882047,
      "grad_norm": 3.91015625,
      "learning_rate": 2e-07,
      "loss": 0.345,
      "step": 7150
    },
    {
      "epoch": 3.1049436253252387,
      "grad_norm": 10.078125,
      "learning_rate": 2e-07,
      "loss": 0.2897,
      "step": 7160
    },
    {
      "epoch": 3.10928013876843,
      "grad_norm": 2.703125,
      "learning_rate": 2e-07,
      "loss": 0.5189,
      "step": 7170
    },
    {
      "epoch": 3.113616652211622,
      "grad_norm": 4.37890625,
      "learning_rate": 2e-07,
      "loss": 0.4248,
      "step": 7180
    },
    {
      "epoch": 3.1179531656548134,
      "grad_norm": 3.4765625,
      "learning_rate": 2e-07,
      "loss": 0.3104,
      "step": 7190
    },
    {
      "epoch": 3.1222896790980053,
      "grad_norm": 4.94921875,
      "learning_rate": 2e-07,
      "loss": 0.3171,
      "step": 7200
    },
    {
      "epoch": 3.126626192541197,
      "grad_norm": 1.326171875,
      "learning_rate": 2e-07,
      "loss": 0.4105,
      "step": 7210
    },
    {
      "epoch": 3.1309627059843885,
      "grad_norm": 3.4296875,
      "learning_rate": 2e-07,
      "loss": 0.4774,
      "step": 7220
    },
    {
      "epoch": 3.1352992194275804,
      "grad_norm": 2.41015625,
      "learning_rate": 2e-07,
      "loss": 0.4051,
      "step": 7230
    },
    {
      "epoch": 3.139635732870772,
      "grad_norm": 3.28515625,
      "learning_rate": 2e-07,
      "loss": 0.645,
      "step": 7240
    },
    {
      "epoch": 3.1439722463139637,
      "grad_norm": 3.939453125,
      "learning_rate": 2e-07,
      "loss": 0.3104,
      "step": 7250
    },
    {
      "epoch": 3.148308759757155,
      "grad_norm": 3.919921875,
      "learning_rate": 2e-07,
      "loss": 0.4191,
      "step": 7260
    },
    {
      "epoch": 3.152645273200347,
      "grad_norm": 1.884765625,
      "learning_rate": 2e-07,
      "loss": 0.4937,
      "step": 7270
    },
    {
      "epoch": 3.156981786643539,
      "grad_norm": 2.615234375,
      "learning_rate": 2e-07,
      "loss": 0.3462,
      "step": 7280
    },
    {
      "epoch": 3.16131830008673,
      "grad_norm": 3.8515625,
      "learning_rate": 2e-07,
      "loss": 0.7002,
      "step": 7290
    },
    {
      "epoch": 3.165654813529922,
      "grad_norm": 2.994140625,
      "learning_rate": 2e-07,
      "loss": 0.298,
      "step": 7300
    },
    {
      "epoch": 3.1699913269731135,
      "grad_norm": 5.84375,
      "learning_rate": 2e-07,
      "loss": 0.3524,
      "step": 7310
    },
    {
      "epoch": 3.1743278404163053,
      "grad_norm": 2.716796875,
      "learning_rate": 2e-07,
      "loss": 0.3881,
      "step": 7320
    },
    {
      "epoch": 3.1786643538594967,
      "grad_norm": 2.267578125,
      "learning_rate": 2e-07,
      "loss": 0.4421,
      "step": 7330
    },
    {
      "epoch": 3.1830008673026886,
      "grad_norm": 2.578125,
      "learning_rate": 2e-07,
      "loss": 0.241,
      "step": 7340
    },
    {
      "epoch": 3.1873373807458805,
      "grad_norm": 3.607421875,
      "learning_rate": 2e-07,
      "loss": 0.4516,
      "step": 7350
    },
    {
      "epoch": 3.191673894189072,
      "grad_norm": 3.646484375,
      "learning_rate": 2e-07,
      "loss": 0.4759,
      "step": 7360
    },
    {
      "epoch": 3.1960104076322637,
      "grad_norm": 2.017578125,
      "learning_rate": 2e-07,
      "loss": 0.528,
      "step": 7370
    },
    {
      "epoch": 3.200346921075455,
      "grad_norm": 4.69140625,
      "learning_rate": 2e-07,
      "loss": 0.6042,
      "step": 7380
    },
    {
      "epoch": 3.204683434518647,
      "grad_norm": 8.28125,
      "learning_rate": 2e-07,
      "loss": 0.5694,
      "step": 7390
    },
    {
      "epoch": 3.209019947961839,
      "grad_norm": 2.5546875,
      "learning_rate": 2e-07,
      "loss": 0.2893,
      "step": 7400
    },
    {
      "epoch": 3.2133564614050303,
      "grad_norm": 8.28125,
      "learning_rate": 2e-07,
      "loss": 0.2969,
      "step": 7410
    },
    {
      "epoch": 3.217692974848222,
      "grad_norm": 3.166015625,
      "learning_rate": 2e-07,
      "loss": 0.3895,
      "step": 7420
    },
    {
      "epoch": 3.2220294882914136,
      "grad_norm": 1.837890625,
      "learning_rate": 2e-07,
      "loss": 0.3902,
      "step": 7430
    },
    {
      "epoch": 3.2263660017346054,
      "grad_norm": 5.79296875,
      "learning_rate": 2e-07,
      "loss": 0.4746,
      "step": 7440
    },
    {
      "epoch": 3.2307025151777973,
      "grad_norm": 5.5,
      "learning_rate": 2e-07,
      "loss": 0.4968,
      "step": 7450
    },
    {
      "epoch": 3.2350390286209887,
      "grad_norm": 1.65234375,
      "learning_rate": 2e-07,
      "loss": 0.2375,
      "step": 7460
    },
    {
      "epoch": 3.2393755420641805,
      "grad_norm": 5.39453125,
      "learning_rate": 2e-07,
      "loss": 0.396,
      "step": 7470
    },
    {
      "epoch": 3.243712055507372,
      "grad_norm": 5.44921875,
      "learning_rate": 2e-07,
      "loss": 0.3022,
      "step": 7480
    },
    {
      "epoch": 3.248048568950564,
      "grad_norm": 4.95703125,
      "learning_rate": 2e-07,
      "loss": 0.2019,
      "step": 7490
    },
    {
      "epoch": 3.2523850823937552,
      "grad_norm": 3.322265625,
      "learning_rate": 2e-07,
      "loss": 0.6658,
      "step": 7500
    },
    {
      "epoch": 3.2523850823937552,
      "eval_loss": 0.22025227546691895,
      "eval_runtime": 361.4148,
      "eval_samples_per_second": 3.176,
      "eval_steps_per_second": 3.176,
      "step": 7500
    },
    {
      "epoch": 3.256721595836947,
      "grad_norm": 2.9921875,
      "learning_rate": 2e-07,
      "loss": 0.4797,
      "step": 7510
    },
    {
      "epoch": 3.261058109280139,
      "grad_norm": 6.2109375,
      "learning_rate": 2e-07,
      "loss": 0.4391,
      "step": 7520
    },
    {
      "epoch": 3.2653946227233304,
      "grad_norm": 4.65625,
      "learning_rate": 2e-07,
      "loss": 0.5165,
      "step": 7530
    },
    {
      "epoch": 3.2697311361665222,
      "grad_norm": 2.013671875,
      "learning_rate": 2e-07,
      "loss": 0.3324,
      "step": 7540
    },
    {
      "epoch": 3.2740676496097136,
      "grad_norm": 3.3203125,
      "learning_rate": 2e-07,
      "loss": 0.3766,
      "step": 7550
    },
    {
      "epoch": 3.2784041630529055,
      "grad_norm": 1.9072265625,
      "learning_rate": 2e-07,
      "loss": 0.3006,
      "step": 7560
    },
    {
      "epoch": 3.282740676496097,
      "grad_norm": 6.7890625,
      "learning_rate": 2e-07,
      "loss": 0.5033,
      "step": 7570
    },
    {
      "epoch": 3.2870771899392888,
      "grad_norm": 4.203125,
      "learning_rate": 2e-07,
      "loss": 0.4685,
      "step": 7580
    },
    {
      "epoch": 3.2914137033824806,
      "grad_norm": 3.947265625,
      "learning_rate": 2e-07,
      "loss": 0.6479,
      "step": 7590
    },
    {
      "epoch": 3.295750216825672,
      "grad_norm": 2.203125,
      "learning_rate": 2e-07,
      "loss": 0.3344,
      "step": 7600
    },
    {
      "epoch": 3.300086730268864,
      "grad_norm": 7.02734375,
      "learning_rate": 2e-07,
      "loss": 0.4241,
      "step": 7610
    },
    {
      "epoch": 3.3044232437120553,
      "grad_norm": 4.0703125,
      "learning_rate": 2e-07,
      "loss": 0.3515,
      "step": 7620
    },
    {
      "epoch": 3.308759757155247,
      "grad_norm": 5.25,
      "learning_rate": 2e-07,
      "loss": 0.5599,
      "step": 7630
    },
    {
      "epoch": 3.313096270598439,
      "grad_norm": 2.361328125,
      "learning_rate": 2e-07,
      "loss": 0.3572,
      "step": 7640
    },
    {
      "epoch": 3.3174327840416304,
      "grad_norm": 4.21875,
      "learning_rate": 2e-07,
      "loss": 0.3203,
      "step": 7650
    },
    {
      "epoch": 3.3217692974848223,
      "grad_norm": 3.9453125,
      "learning_rate": 2e-07,
      "loss": 0.3434,
      "step": 7660
    },
    {
      "epoch": 3.3261058109280137,
      "grad_norm": 3.09765625,
      "learning_rate": 2e-07,
      "loss": 0.3763,
      "step": 7670
    },
    {
      "epoch": 3.3304423243712056,
      "grad_norm": 1.89453125,
      "learning_rate": 2e-07,
      "loss": 0.4897,
      "step": 7680
    },
    {
      "epoch": 3.3347788378143974,
      "grad_norm": 5.87890625,
      "learning_rate": 2e-07,
      "loss": 0.4377,
      "step": 7690
    },
    {
      "epoch": 3.339115351257589,
      "grad_norm": 2.501953125,
      "learning_rate": 2e-07,
      "loss": 0.2714,
      "step": 7700
    },
    {
      "epoch": 3.3434518647007807,
      "grad_norm": 4.421875,
      "learning_rate": 2e-07,
      "loss": 0.4043,
      "step": 7710
    },
    {
      "epoch": 3.347788378143972,
      "grad_norm": 1.705078125,
      "learning_rate": 2e-07,
      "loss": 0.5656,
      "step": 7720
    },
    {
      "epoch": 3.352124891587164,
      "grad_norm": 4.26171875,
      "learning_rate": 2e-07,
      "loss": 0.4125,
      "step": 7730
    },
    {
      "epoch": 3.3564614050303554,
      "grad_norm": 4.29296875,
      "learning_rate": 2e-07,
      "loss": 0.541,
      "step": 7740
    },
    {
      "epoch": 3.3607979184735473,
      "grad_norm": 4.5546875,
      "learning_rate": 2e-07,
      "loss": 0.2898,
      "step": 7750
    },
    {
      "epoch": 3.365134431916739,
      "grad_norm": 2.8671875,
      "learning_rate": 2e-07,
      "loss": 0.2838,
      "step": 7760
    },
    {
      "epoch": 3.3694709453599305,
      "grad_norm": 1.4892578125,
      "learning_rate": 2e-07,
      "loss": 0.457,
      "step": 7770
    },
    {
      "epoch": 3.3738074588031224,
      "grad_norm": 4.52734375,
      "learning_rate": 2e-07,
      "loss": 0.4005,
      "step": 7780
    },
    {
      "epoch": 3.378143972246314,
      "grad_norm": 8.09375,
      "learning_rate": 2e-07,
      "loss": 0.4946,
      "step": 7790
    },
    {
      "epoch": 3.3824804856895057,
      "grad_norm": 1.9833984375,
      "learning_rate": 2e-07,
      "loss": 0.5866,
      "step": 7800
    },
    {
      "epoch": 3.386816999132697,
      "grad_norm": 3.560546875,
      "learning_rate": 2e-07,
      "loss": 0.3126,
      "step": 7810
    },
    {
      "epoch": 3.391153512575889,
      "grad_norm": 2.423828125,
      "learning_rate": 2e-07,
      "loss": 0.4233,
      "step": 7820
    },
    {
      "epoch": 3.395490026019081,
      "grad_norm": 5.9296875,
      "learning_rate": 2e-07,
      "loss": 0.4553,
      "step": 7830
    },
    {
      "epoch": 3.399826539462272,
      "grad_norm": 5.59375,
      "learning_rate": 2e-07,
      "loss": 0.3368,
      "step": 7840
    },
    {
      "epoch": 3.404163052905464,
      "grad_norm": 2.2734375,
      "learning_rate": 2e-07,
      "loss": 0.3359,
      "step": 7850
    },
    {
      "epoch": 3.408499566348656,
      "grad_norm": 1.6650390625,
      "learning_rate": 2e-07,
      "loss": 0.2672,
      "step": 7860
    },
    {
      "epoch": 3.4128360797918473,
      "grad_norm": 20.265625,
      "learning_rate": 2e-07,
      "loss": 0.3453,
      "step": 7870
    },
    {
      "epoch": 3.417172593235039,
      "grad_norm": 3.1171875,
      "learning_rate": 2e-07,
      "loss": 0.3565,
      "step": 7880
    },
    {
      "epoch": 3.4215091066782306,
      "grad_norm": 4.4453125,
      "learning_rate": 2e-07,
      "loss": 0.34,
      "step": 7890
    },
    {
      "epoch": 3.4258456201214225,
      "grad_norm": 16.21875,
      "learning_rate": 2e-07,
      "loss": 0.628,
      "step": 7900
    },
    {
      "epoch": 3.430182133564614,
      "grad_norm": 6.12109375,
      "learning_rate": 2e-07,
      "loss": 0.6415,
      "step": 7910
    },
    {
      "epoch": 3.4345186470078057,
      "grad_norm": 3.70703125,
      "learning_rate": 2e-07,
      "loss": 0.3398,
      "step": 7920
    },
    {
      "epoch": 3.4388551604509976,
      "grad_norm": 8.703125,
      "learning_rate": 2e-07,
      "loss": 0.3147,
      "step": 7930
    },
    {
      "epoch": 3.443191673894189,
      "grad_norm": 4.7734375,
      "learning_rate": 2e-07,
      "loss": 0.37,
      "step": 7940
    },
    {
      "epoch": 3.447528187337381,
      "grad_norm": 2.146484375,
      "learning_rate": 2e-07,
      "loss": 0.333,
      "step": 7950
    },
    {
      "epoch": 3.4518647007805723,
      "grad_norm": 6.33984375,
      "learning_rate": 2e-07,
      "loss": 0.3912,
      "step": 7960
    },
    {
      "epoch": 3.456201214223764,
      "grad_norm": 3.708984375,
      "learning_rate": 2e-07,
      "loss": 0.4113,
      "step": 7970
    },
    {
      "epoch": 3.4605377276669556,
      "grad_norm": 3.970703125,
      "learning_rate": 2e-07,
      "loss": 0.3795,
      "step": 7980
    },
    {
      "epoch": 3.4648742411101474,
      "grad_norm": 5.6484375,
      "learning_rate": 2e-07,
      "loss": 0.4278,
      "step": 7990
    },
    {
      "epoch": 3.4692107545533393,
      "grad_norm": 12.859375,
      "learning_rate": 2e-07,
      "loss": 0.4818,
      "step": 8000
    },
    {
      "epoch": 3.4692107545533393,
      "eval_loss": 0.2136317640542984,
      "eval_runtime": 360.1744,
      "eval_samples_per_second": 3.187,
      "eval_steps_per_second": 3.187,
      "step": 8000
    }
  ],
  "logging_steps": 10,
  "max_steps": 23060,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.530560304928047e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}